diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,63021 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.967533863685229,
+  "eval_steps": 500,
+  "global_step": 9000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.3429909468444645,
+      "learning_rate": 3.571428571428572e-08,
+      "loss": 0.7295,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.08536851126392,
+      "learning_rate": 7.142857142857144e-08,
+      "loss": 0.6339,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.741844844961892,
+      "learning_rate": 1.0714285714285716e-07,
+      "loss": 0.7143,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.181704830271974,
+      "learning_rate": 1.4285714285714287e-07,
+      "loss": 0.7099,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.425773064810272,
+      "learning_rate": 1.7857142857142858e-07,
+      "loss": 0.2857,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.129379193967202,
+      "learning_rate": 2.142857142857143e-07,
+      "loss": 0.7138,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.774436266993971,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 0.7577,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.168869949738837,
+      "learning_rate": 2.8571428571428575e-07,
+      "loss": 0.7012,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.601931674938831,
+      "learning_rate": 3.214285714285714e-07,
+      "loss": 0.7398,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.9667239449896705,
+      "learning_rate": 3.5714285714285716e-07,
+      "loss": 0.6223,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.285727464643471,
+      "learning_rate": 3.9285714285714286e-07,
+      "loss": 0.7344,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.77435849110028,
+      "learning_rate": 4.285714285714286e-07,
+      "loss": 0.782,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.9579176453501885,
+      "learning_rate": 4.642857142857143e-07,
+      "loss": 0.6826,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.41500897386890795,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 0.2681,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.630322514804351,
+      "learning_rate": 5.357142857142857e-07,
+      "loss": 0.7241,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.69870703656859,
+      "learning_rate": 5.714285714285715e-07,
+      "loss": 0.6929,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.5362335472787105,
+      "learning_rate": 6.071428571428572e-07,
+      "loss": 0.6869,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.570566762502487,
+      "learning_rate": 6.428571428571428e-07,
+      "loss": 0.6747,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.200351182959851,
+      "learning_rate": 6.785714285714286e-07,
+      "loss": 0.6797,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.863099837977999,
+      "learning_rate": 7.142857142857143e-07,
+      "loss": 0.6627,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.05350684518345,
+      "learning_rate": 7.5e-07,
+      "loss": 0.6403,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.374751607990124,
+      "learning_rate": 7.857142857142857e-07,
+      "loss": 0.703,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.956352059970736,
+      "learning_rate": 8.214285714285715e-07,
+      "loss": 0.6559,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.4161519975788515,
+      "learning_rate": 8.571428571428572e-07,
+      "loss": 0.2652,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.463171096590779,
+      "learning_rate": 8.928571428571429e-07,
+      "loss": 0.6395,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.740168475674717,
+      "learning_rate": 9.285714285714287e-07,
+      "loss": 0.6322,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8624334132888785,
+      "learning_rate": 9.642857142857145e-07,
+      "loss": 0.6646,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.189931169134805,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.6641,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.4178135394410765,
+      "learning_rate": 1.0357142857142859e-06,
+      "loss": 0.6466,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.3997148840808848,
+      "learning_rate": 1.0714285714285714e-06,
+      "loss": 0.2754,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.334538320841567,
+      "learning_rate": 1.1071428571428573e-06,
+      "loss": 0.638,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5628906731356595,
+      "learning_rate": 1.142857142857143e-06,
+      "loss": 0.5836,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.803098938942329,
+      "learning_rate": 1.1785714285714287e-06,
+      "loss": 0.6059,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.500870887520296,
+      "learning_rate": 1.2142857142857144e-06,
+      "loss": 0.6031,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.896041837249922,
+      "learning_rate": 1.25e-06,
+      "loss": 0.6621,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.7875475971664807,
+      "learning_rate": 1.2857142857142856e-06,
+      "loss": 0.6216,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.531782895960642,
+      "learning_rate": 1.3214285714285715e-06,
+      "loss": 0.5628,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6418532069745515,
+      "learning_rate": 1.3571428571428572e-06,
+      "loss": 0.5848,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6143456302713375,
+      "learning_rate": 1.392857142857143e-06,
+      "loss": 0.5695,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.41270312294046946,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.2811,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.970627760591574,
+      "learning_rate": 1.4642857142857145e-06,
+      "loss": 0.5826,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8659490937129366,
+      "learning_rate": 1.5e-06,
+      "loss": 0.639,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.706928245230411,
+      "learning_rate": 1.535714285714286e-06,
+      "loss": 0.5587,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.3320273886669565,
+      "learning_rate": 1.5714285714285714e-06,
+      "loss": 0.5712,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.393768132679274,
+      "learning_rate": 1.6071428571428574e-06,
+      "loss": 0.6398,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.4788112467629326,
+      "learning_rate": 1.642857142857143e-06,
+      "loss": 0.6143,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.140136230181165,
+      "learning_rate": 1.6785714285714286e-06,
+      "loss": 0.652,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.027863784027695,
+      "learning_rate": 1.7142857142857145e-06,
+      "loss": 0.6881,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.670825261241078,
+      "learning_rate": 1.75e-06,
+      "loss": 0.6183,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.884187013364047,
+      "learning_rate": 1.7857142857142859e-06,
+      "loss": 0.5311,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8463219562679227,
+      "learning_rate": 1.8214285714285716e-06,
+      "loss": 0.6497,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7848603780728345,
+      "learning_rate": 1.8571428571428573e-06,
+      "loss": 0.6034,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.702489693969101,
+      "learning_rate": 1.892857142857143e-06,
+      "loss": 0.5733,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.221760084597975,
+      "learning_rate": 1.928571428571429e-06,
+      "loss": 0.5862,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7972738625148614,
+      "learning_rate": 1.9642857142857144e-06,
+      "loss": 0.5753,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.112032003263317,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.5833,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4233485040336077,
+      "learning_rate": 2.035714285714286e-06,
+      "loss": 0.2769,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 20.928612055235522,
+      "learning_rate": 2.0714285714285717e-06,
+      "loss": 0.592,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.480913907628851,
+      "learning_rate": 2.1071428571428572e-06,
+      "loss": 0.6019,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.67874703675065,
+      "learning_rate": 2.1428571428571427e-06,
+      "loss": 0.5926,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.555335641656111,
+      "learning_rate": 2.1785714285714286e-06,
+      "loss": 0.5663,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6860254131049768,
+      "learning_rate": 2.2142857142857146e-06,
+      "loss": 0.6072,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0051926276874,
+      "learning_rate": 2.25e-06,
+      "loss": 0.5667,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.841417332722371,
+      "learning_rate": 2.285714285714286e-06,
+      "loss": 0.5537,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1878681602053156,
+      "learning_rate": 2.321428571428572e-06,
+      "loss": 0.597,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.962483713613043,
+      "learning_rate": 2.3571428571428574e-06,
+      "loss": 0.5401,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8492327516266602,
+      "learning_rate": 2.3928571428571433e-06,
+      "loss": 0.5825,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0817157975196445,
+      "learning_rate": 2.428571428571429e-06,
+      "loss": 0.5708,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2790475268466333,
+      "learning_rate": 2.4642857142857147e-06,
+      "loss": 0.5826,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6640844277803364,
+      "learning_rate": 2.5e-06,
+      "loss": 0.5262,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.973201341090745,
+      "learning_rate": 2.5357142857142857e-06,
+      "loss": 0.521,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.44515354668496,
+      "learning_rate": 2.571428571428571e-06,
+      "loss": 0.5324,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.259752558249684,
+      "learning_rate": 2.6071428571428575e-06,
+      "loss": 0.5368,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.429982978840439,
+      "learning_rate": 2.642857142857143e-06,
+      "loss": 0.5573,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9934276085735823,
+      "learning_rate": 2.6785714285714285e-06,
+      "loss": 0.5557,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7620957404449937,
+      "learning_rate": 2.7142857142857144e-06,
+      "loss": 0.6593,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4259623027422044,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.5366,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.286533324357221,
+      "learning_rate": 2.785714285714286e-06,
+      "loss": 0.5504,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1368638923353913,
+      "learning_rate": 2.8214285714285718e-06,
+      "loss": 0.5578,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9013714563682895,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.5831,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.108354975589372,
+      "learning_rate": 2.892857142857143e-06,
+      "loss": 0.5821,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4451119366043494,
+      "learning_rate": 2.928571428571429e-06,
+      "loss": 0.5931,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.021404069843295,
+      "learning_rate": 2.9642857142857146e-06,
+      "loss": 0.556,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.259884791468475,
+      "learning_rate": 3e-06,
+      "loss": 0.5935,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.491905385012073,
+      "learning_rate": 3.0357142857142856e-06,
+      "loss": 0.5194,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.4403782352986125,
+      "learning_rate": 3.071428571428572e-06,
+      "loss": 0.5263,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5050800565755966,
+      "learning_rate": 3.1071428571428574e-06,
+      "loss": 0.5162,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.150059299679612,
+      "learning_rate": 3.142857142857143e-06,
+      "loss": 0.5414,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4519878862313798,
+      "learning_rate": 3.178571428571429e-06,
+      "loss": 0.5223,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3898004025604619,
+      "learning_rate": 3.2142857142857147e-06,
+      "loss": 0.2584,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.844436924819011,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.5436,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.00581737722414,
+      "learning_rate": 3.285714285714286e-06,
+      "loss": 0.5412,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.140302812949141,
+      "learning_rate": 3.3214285714285716e-06,
+      "loss": 0.5516,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.027758008175357,
+      "learning_rate": 3.357142857142857e-06,
+      "loss": 0.561,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.255412797897802,
+      "learning_rate": 3.3928571428571435e-06,
+      "loss": 0.6345,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.81243297197542,
+      "learning_rate": 3.428571428571429e-06,
+      "loss": 0.4642,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.623172045808998,
+      "learning_rate": 3.4642857142857145e-06,
+      "loss": 0.5353,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.24717036170527,
+      "learning_rate": 3.5e-06,
+      "loss": 0.5427,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.206919224292777,
+      "learning_rate": 3.5357142857142863e-06,
+      "loss": 0.5662,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8954277039581346,
+      "learning_rate": 3.5714285714285718e-06,
+      "loss": 0.496,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.2183268536880005,
+      "learning_rate": 3.6071428571428573e-06,
+      "loss": 0.4897,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7079537020343873,
+      "learning_rate": 3.642857142857143e-06,
+      "loss": 0.5797,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1672362992154484,
+      "learning_rate": 3.678571428571429e-06,
+      "loss": 0.5403,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.02243372683841,
+      "learning_rate": 3.7142857142857146e-06,
+      "loss": 0.5303,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.571786166372638,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.5809,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.187675989518087,
+      "learning_rate": 3.785714285714286e-06,
+      "loss": 0.5384,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6793498977297387,
+      "learning_rate": 3.8214285714285715e-06,
+      "loss": 0.5364,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.170107321737625,
+      "learning_rate": 3.857142857142858e-06,
+      "loss": 0.5403,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8384452597677883,
+      "learning_rate": 3.892857142857143e-06,
+      "loss": 0.5909,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.6219682939617925,
+      "learning_rate": 3.928571428571429e-06,
+      "loss": 0.5155,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5419663364831275,
+      "learning_rate": 3.964285714285714e-06,
+      "loss": 0.5406,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.039471098647939,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.5685,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5986249257661194,
+      "learning_rate": 4.035714285714286e-06,
+      "loss": 0.536,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5859586363078435,
+      "learning_rate": 4.071428571428572e-06,
+      "loss": 0.5362,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.448831609661831,
+      "learning_rate": 4.107142857142857e-06,
+      "loss": 0.529,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.662327743446219,
+      "learning_rate": 4.1428571428571435e-06,
+      "loss": 0.5359,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4054858054096825,
+      "learning_rate": 4.178571428571429e-06,
+      "loss": 0.2748,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.683792293890771,
+      "learning_rate": 4.2142857142857145e-06,
+      "loss": 0.5026,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 66.43051642679107,
+      "learning_rate": 4.25e-06,
+      "loss": 0.5867,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4027226651004199,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 0.2618,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.617868204773627,
+      "learning_rate": 4.321428571428572e-06,
+      "loss": 0.5301,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.591807979190813,
+      "learning_rate": 4.357142857142857e-06,
+      "loss": 0.5442,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.82518010813047,
+      "learning_rate": 4.392857142857143e-06,
+      "loss": 0.5675,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.865670452593747,
+      "learning_rate": 4.428571428571429e-06,
+      "loss": 0.5132,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.81537657150533,
+      "learning_rate": 4.464285714285715e-06,
+      "loss": 0.529,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.87718975076854,
+      "learning_rate": 4.5e-06,
+      "loss": 0.5655,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.513375979042196,
+      "learning_rate": 4.5357142857142865e-06,
+      "loss": 0.5715,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 42.72049457121376,
+      "learning_rate": 4.571428571428572e-06,
+      "loss": 0.5481,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 123.33901935646568,
+      "learning_rate": 4.6071428571428574e-06,
+      "loss": 0.5208,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.309561225198576,
+      "learning_rate": 4.642857142857144e-06,
+      "loss": 0.5647,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.781908721984838,
+      "learning_rate": 4.678571428571429e-06,
+      "loss": 0.5187,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.550334273580416,
+      "learning_rate": 4.714285714285715e-06,
+      "loss": 0.5053,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.567182181515489,
+      "learning_rate": 4.75e-06,
+      "loss": 0.5617,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.410195898886369,
+      "learning_rate": 4.785714285714287e-06,
+      "loss": 0.283,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6527135018582495,
+      "learning_rate": 4.821428571428572e-06,
+      "loss": 0.4543,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.396792260435254,
+      "learning_rate": 4.857142857142858e-06,
+      "loss": 0.5175,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3039903283417846,
+      "learning_rate": 4.892857142857143e-06,
+      "loss": 0.5535,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3463799037214828,
+      "learning_rate": 4.928571428571429e-06,
+      "loss": 0.5573,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.733371296635203,
+      "learning_rate": 4.964285714285715e-06,
+      "loss": 0.5279,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.054592957945209,
+      "learning_rate": 5e-06,
+      "loss": 0.5851,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.700991688419293,
+      "learning_rate": 5.035714285714286e-06,
+      "loss": 0.5389,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.44985816930090117,
+      "learning_rate": 5.071428571428571e-06,
+      "loss": 0.2793,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.694678732515108,
+      "learning_rate": 5.107142857142857e-06,
+      "loss": 0.5306,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4288722987915697,
+      "learning_rate": 5.142857142857142e-06,
+      "loss": 0.2844,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.780589701834514,
+      "learning_rate": 5.1785714285714296e-06,
+      "loss": 0.5861,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.177266606364717,
+      "learning_rate": 5.214285714285715e-06,
+      "loss": 0.5825,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.172154066390874,
+      "learning_rate": 5.2500000000000006e-06,
+      "loss": 0.5198,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6611830656044515,
+      "learning_rate": 5.285714285714286e-06,
+      "loss": 0.5266,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9237450374224596,
+      "learning_rate": 5.3214285714285715e-06,
+      "loss": 0.5682,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.649250316493386,
+      "learning_rate": 5.357142857142857e-06,
+      "loss": 0.5301,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9883544781240468,
+      "learning_rate": 5.392857142857143e-06,
+      "loss": 0.537,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4152368958893654,
+      "learning_rate": 5.428571428571429e-06,
+      "loss": 0.2692,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4915026736561883,
+      "learning_rate": 5.464285714285714e-06,
+      "loss": 0.6079,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.27076237258208,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 0.5512,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.37968820157114963,
+      "learning_rate": 5.535714285714286e-06,
+      "loss": 0.2632,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.227756058253568,
+      "learning_rate": 5.571428571428572e-06,
+      "loss": 0.5489,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.056098221034406,
+      "learning_rate": 5.607142857142858e-06,
+      "loss": 0.5225,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.47045678069264,
+      "learning_rate": 5.6428571428571435e-06,
+      "loss": 0.5087,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6229258900244283,
+      "learning_rate": 5.678571428571429e-06,
+      "loss": 0.5509,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7922512745621924,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.5041,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.175391589091643,
+      "learning_rate": 5.75e-06,
+      "loss": 0.557,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.404619699403472,
+      "learning_rate": 5.785714285714286e-06,
+      "loss": 0.54,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.549853135326141,
+      "learning_rate": 5.821428571428573e-06,
+      "loss": 0.4918,
+      "step": 163
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9871944515635334,
+      "learning_rate": 5.857142857142858e-06,
+      "loss": 0.5286,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8848445582591196,
+      "learning_rate": 5.892857142857144e-06,
+      "loss": 0.5339,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.917651659415263,
+      "learning_rate": 5.928571428571429e-06,
+      "loss": 0.5383,
+      "step": 166
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.43590916100088073,
+      "learning_rate": 5.964285714285715e-06,
+      "loss": 0.2801,
+      "step": 167
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.501081049076598,
+      "learning_rate": 6e-06,
+      "loss": 0.5519,
+      "step": 168
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.801401008891093,
+      "learning_rate": 6.035714285714286e-06,
+      "loss": 0.5522,
+      "step": 169
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.966761617134997,
+      "learning_rate": 6.071428571428571e-06,
+      "loss": 0.5436,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8923196974711334,
+      "learning_rate": 6.107142857142858e-06,
+      "loss": 0.5597,
+      "step": 171
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9502846533209937,
+      "learning_rate": 6.142857142857144e-06,
+      "loss": 0.5569,
+      "step": 172
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.493426082802722,
+      "learning_rate": 6.178571428571429e-06,
+      "loss": 0.5273,
+      "step": 173
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9142105102611815,
+      "learning_rate": 6.214285714285715e-06,
+      "loss": 0.5773,
+      "step": 174
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3919820846675135,
+      "learning_rate": 6.25e-06,
+      "loss": 0.559,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.758524940331889,
+      "learning_rate": 6.285714285714286e-06,
+      "loss": 0.5678,
+      "step": 176
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5859269584209703,
+      "learning_rate": 6.321428571428571e-06,
+      "loss": 0.5521,
+      "step": 177
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0207046004846285,
+      "learning_rate": 6.357142857142858e-06,
+      "loss": 0.4989,
+      "step": 178
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.365182617578441,
+      "learning_rate": 6.392857142857143e-06,
+      "loss": 0.5018,
+      "step": 179
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.483902606199555,
+      "learning_rate": 6.4285714285714295e-06,
+      "loss": 0.6055,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.429490157237544,
+      "learning_rate": 6.464285714285715e-06,
+      "loss": 0.571,
+      "step": 181
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.11050273899657,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 0.5576,
+      "step": 182
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.492177167751906,
+      "learning_rate": 6.535714285714286e-06,
+      "loss": 0.5344,
+      "step": 183
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.549456965570582,
+      "learning_rate": 6.571428571428572e-06,
+      "loss": 0.5485,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.240076869241682,
+      "learning_rate": 6.607142857142858e-06,
+      "loss": 0.5687,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.177147037563416,
+      "learning_rate": 6.642857142857143e-06,
+      "loss": 0.5001,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.170418733806601,
+      "learning_rate": 6.678571428571429e-06,
+      "loss": 0.553,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.746520584987548,
+      "learning_rate": 6.714285714285714e-06,
+      "loss": 0.5171,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.000028456168399,
+      "learning_rate": 6.750000000000001e-06,
+      "loss": 0.5595,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.562614410090741,
+      "learning_rate": 6.785714285714287e-06,
+      "loss": 0.5414,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.837333280972769,
+      "learning_rate": 6.8214285714285724e-06,
+      "loss": 0.5753,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.745874141258343,
+      "learning_rate": 6.857142857142858e-06,
+      "loss": 0.5859,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.320994681076247,
+      "learning_rate": 6.892857142857143e-06,
+      "loss": 0.5049,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4861616939320776,
+      "learning_rate": 6.928571428571429e-06,
+      "loss": 0.5397,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.154673807197627,
+      "learning_rate": 6.964285714285714e-06,
+      "loss": 0.5161,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.468145441161037,
+      "learning_rate": 7e-06,
+      "loss": 0.5436,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.361994937382034,
+      "learning_rate": 7.035714285714287e-06,
+      "loss": 0.5008,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4347526484171227,
+      "learning_rate": 7.0714285714285726e-06,
+      "loss": 0.5396,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9270738884115417,
+      "learning_rate": 7.107142857142858e-06,
+      "loss": 0.5813,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6005105185348376,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 0.4904,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1905773837586913,
+      "learning_rate": 7.178571428571429e-06,
+      "loss": 0.5455,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.692327283072809,
+      "learning_rate": 7.2142857142857145e-06,
+      "loss": 0.4802,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.371185643584461,
+      "learning_rate": 7.25e-06,
+      "loss": 0.5141,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.426270383139635,
+      "learning_rate": 7.285714285714286e-06,
+      "loss": 0.5351,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.678060855795258,
+      "learning_rate": 7.321428571428572e-06,
+      "loss": 0.5218,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8609732599172255,
+      "learning_rate": 7.357142857142858e-06,
+      "loss": 0.5821,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8528986794414397,
+      "learning_rate": 7.392857142857144e-06,
+      "loss": 0.4964,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.46744877238430504,
+      "learning_rate": 7.428571428571429e-06,
+      "loss": 0.2821,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0609715796783368,
+      "learning_rate": 7.464285714285715e-06,
+      "loss": 0.5862,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.962949758605313,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.5045,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9060501632894096,
+      "learning_rate": 7.5357142857142865e-06,
+      "loss": 0.5206,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.272300538616481,
+      "learning_rate": 7.571428571428572e-06,
+      "loss": 0.5677,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1639625859982496,
+      "learning_rate": 7.6071428571428575e-06,
+      "loss": 0.5631,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.602004006540063,
+      "learning_rate": 7.642857142857143e-06,
+      "loss": 0.5466,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7530898556502503,
+      "learning_rate": 7.67857142857143e-06,
+      "loss": 0.5134,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.602745535270071,
+      "learning_rate": 7.714285714285716e-06,
+      "loss": 0.5354,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.443008536300093,
+      "learning_rate": 7.75e-06,
+      "loss": 0.5697,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.971513193873936,
+      "learning_rate": 7.785714285714287e-06,
+      "loss": 0.5146,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3830622348472543,
+      "learning_rate": 7.821428571428571e-06,
+      "loss": 0.5554,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.620826321646173,
+      "learning_rate": 7.857142857142858e-06,
+      "loss": 0.4855,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.107857468660166,
+      "learning_rate": 7.892857142857144e-06,
+      "loss": 0.503,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7700210610808496,
+      "learning_rate": 7.928571428571429e-06,
+      "loss": 0.5118,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6486389633649474,
+      "learning_rate": 7.964285714285715e-06,
+      "loss": 0.2822,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6773357718112036,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.5734,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7695915559066018,
+      "learning_rate": 8.035714285714286e-06,
+      "loss": 0.491,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.44809010659209053,
+      "learning_rate": 8.071428571428572e-06,
+      "loss": 0.2695,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.409630585090655,
+      "learning_rate": 8.107142857142859e-06,
+      "loss": 0.5013,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.592428228550645,
+      "learning_rate": 8.142857142857143e-06,
+      "loss": 0.5429,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.573948027335371,
+      "learning_rate": 8.17857142857143e-06,
+      "loss": 0.5491,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.864031445331459,
+      "learning_rate": 8.214285714285714e-06,
+      "loss": 0.5684,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.074207496075294,
+      "learning_rate": 8.25e-06,
+      "loss": 0.5765,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.839360795885173,
+      "learning_rate": 8.285714285714287e-06,
+      "loss": 0.5222,
+      "step": 232
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.976565969661428,
+      "learning_rate": 8.321428571428573e-06,
+      "loss": 0.5717,
+      "step": 233
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.007292217954472,
+      "learning_rate": 8.357142857142858e-06,
+      "loss": 0.5557,
+      "step": 234
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.308386173913218,
+      "learning_rate": 8.392857142857144e-06,
+      "loss": 0.5071,
+      "step": 235
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7019492070308929,
+      "learning_rate": 8.428571428571429e-06,
+      "loss": 0.2904,
+      "step": 236
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.1991502173743385,
+      "learning_rate": 8.464285714285715e-06,
+      "loss": 0.5258,
+      "step": 237
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8613625343729323,
+      "learning_rate": 8.5e-06,
+      "loss": 0.5415,
+      "step": 238
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.8189181083362245,
+      "learning_rate": 8.535714285714286e-06,
+      "loss": 0.5231,
+      "step": 239
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3875520860789674,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.5797,
+      "step": 240
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.482337153619518,
+      "learning_rate": 8.607142857142859e-06,
+      "loss": 0.5505,
+      "step": 241
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.016463116691625,
+      "learning_rate": 8.642857142857144e-06,
+      "loss": 0.5123,
+      "step": 242
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1585596237118327,
+      "learning_rate": 8.67857142857143e-06,
+      "loss": 0.5089,
+      "step": 243
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.021446208312963,
+      "learning_rate": 8.714285714285715e-06,
+      "loss": 0.5741,
+      "step": 244
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.3178429067713155,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 0.5032,
+      "step": 245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.465007732847373,
+      "learning_rate": 8.785714285714286e-06,
+      "loss": 0.5379,
+      "step": 246
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.310496480168842,
+      "learning_rate": 8.821428571428572e-06,
+      "loss": 0.4902,
+      "step": 247
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.575735675294582,
+      "learning_rate": 8.857142857142858e-06,
+      "loss": 0.522,
+      "step": 248
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.776220734616733,
+      "learning_rate": 8.892857142857143e-06,
+      "loss": 0.4627,
+      "step": 249
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.624573893364172,
+      "learning_rate": 8.92857142857143e-06,
+      "loss": 0.5043,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.565274974042856,
+      "learning_rate": 8.964285714285716e-06,
+      "loss": 0.5416,
+      "step": 251
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7713906613892654,
+      "learning_rate": 9e-06,
+      "loss": 0.5582,
+      "step": 252
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3173933014526296,
+      "learning_rate": 9.035714285714287e-06,
+      "loss": 0.5411,
+      "step": 253
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.223378152327329,
+      "learning_rate": 9.071428571428573e-06,
+      "loss": 0.5279,
+      "step": 254
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.773356428361358,
+      "learning_rate": 9.107142857142858e-06,
+      "loss": 0.5823,
+      "step": 255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4264596225756425,
+      "learning_rate": 9.142857142857144e-06,
+      "loss": 0.5525,
+      "step": 256
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.5285704209606505,
+      "learning_rate": 9.178571428571429e-06,
+      "loss": 0.562,
+      "step": 257
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.07316256685808,
+      "learning_rate": 9.214285714285715e-06,
+      "loss": 0.5129,
+      "step": 258
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.825432350471509,
+      "learning_rate": 9.250000000000001e-06,
+      "loss": 0.5079,
+      "step": 259
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3181455198769254,
+      "learning_rate": 9.285714285714288e-06,
+      "loss": 0.5506,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.600615318513523,
+      "learning_rate": 9.321428571428572e-06,
+      "loss": 0.5369,
+      "step": 261
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9450589607724148,
+      "learning_rate": 9.357142857142859e-06,
+      "loss": 0.3056,
+      "step": 262
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.827427343172781,
+      "learning_rate": 9.392857142857143e-06,
+      "loss": 0.524,
+      "step": 263
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.250151160160805,
+      "learning_rate": 9.42857142857143e-06,
+      "loss": 0.4748,
+      "step": 264
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4316183273647423,
+      "learning_rate": 9.464285714285714e-06,
+      "loss": 0.5738,
+      "step": 265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5599688202309212,
+      "learning_rate": 9.5e-06,
+      "loss": 0.2836,
+      "step": 266
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.1117592711923345,
+      "learning_rate": 9.535714285714287e-06,
+      "loss": 0.5404,
+      "step": 267
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9423093597758165,
+      "learning_rate": 9.571428571428573e-06,
+      "loss": 0.5055,
+      "step": 268
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.949650759205794,
+      "learning_rate": 9.607142857142858e-06,
+      "loss": 0.5173,
+      "step": 269
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0301490372048656,
+      "learning_rate": 9.642857142857144e-06,
+      "loss": 0.5218,
+      "step": 270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7767950010358087,
+      "learning_rate": 9.678571428571429e-06,
+      "loss": 0.4877,
+      "step": 271
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6732344994933961,
+      "learning_rate": 9.714285714285715e-06,
+      "loss": 0.3064,
+      "step": 272
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8845943914668575,
+      "learning_rate": 9.75e-06,
+      "loss": 0.5091,
+      "step": 273
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0978279231789516,
+      "learning_rate": 9.785714285714286e-06,
+      "loss": 0.4888,
+      "step": 274
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.916493572459872,
+      "learning_rate": 9.821428571428573e-06,
+      "loss": 0.5244,
+      "step": 275
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.437869560617859,
+      "learning_rate": 9.857142857142859e-06,
+      "loss": 0.5097,
+      "step": 276
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3228775798093313,
+      "learning_rate": 9.892857142857143e-06,
+      "loss": 0.4994,
+      "step": 277
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5233913694365816,
+      "learning_rate": 9.92857142857143e-06,
+      "loss": 0.3256,
+      "step": 278
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4832453956013481,
+      "learning_rate": 9.964285714285714e-06,
+      "loss": 0.2721,
+      "step": 279
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3990049252849548,
+      "learning_rate": 1e-05,
+      "loss": 0.2759,
+      "step": 280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7119133200093533,
+      "learning_rate": 9.999999696866382e-06,
+      "loss": 0.5279,
+      "step": 281
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.452383294590257,
+      "learning_rate": 9.999998787465562e-06,
+      "loss": 0.5824,
+      "step": 282
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.936022991790363,
+      "learning_rate": 9.999997271797653e-06,
+      "loss": 0.5941,
+      "step": 283
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.753056929921005,
+      "learning_rate": 9.999995149862836e-06,
+      "loss": 0.4969,
+      "step": 284
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.785890906593368,
+      "learning_rate": 9.999992421661369e-06,
+      "loss": 0.5488,
+      "step": 285
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7997414803418663,
+      "learning_rate": 9.999989087193582e-06,
+      "loss": 0.5497,
+      "step": 286
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.061663771560861,
+      "learning_rate": 9.999985146459882e-06,
+      "loss": 0.5691,
+      "step": 287
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.357630846080074,
+      "learning_rate": 9.999980599460747e-06,
+      "loss": 0.516,
+      "step": 288
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.769300325816736,
+      "learning_rate": 9.999975446196725e-06,
+      "loss": 0.595,
+      "step": 289
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.259289028326002,
+      "learning_rate": 9.999969686668442e-06,
+      "loss": 0.5447,
+      "step": 290
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.929793042677327,
+      "learning_rate": 9.9999633208766e-06,
+      "loss": 0.5158,
+      "step": 291
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8919303814249933,
+      "learning_rate": 9.999956348821965e-06,
+      "loss": 0.5417,
+      "step": 292
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.57408120766393,
+      "learning_rate": 9.999948770505386e-06,
+      "loss": 0.5286,
+      "step": 293
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3423102601926855,
+      "learning_rate": 9.99994058592778e-06,
+      "loss": 0.5776,
+      "step": 294
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0866213114889987,
+      "learning_rate": 9.999931795090143e-06,
+      "loss": 0.5159,
+      "step": 295
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3457183064136533,
+      "learning_rate": 9.999922397993537e-06,
+      "loss": 0.4948,
+      "step": 296
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1008176767249154,
+      "learning_rate": 9.999912394639105e-06,
+      "loss": 0.5047,
+      "step": 297
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0659090071259203,
+      "learning_rate": 9.999901785028055e-06,
+      "loss": 0.4969,
+      "step": 298
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9331981386586308,
+      "learning_rate": 9.999890569161676e-06,
+      "loss": 0.5046,
+      "step": 299
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.187693237901873,
+      "learning_rate": 9.99987874704133e-06,
+      "loss": 0.3347,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.484885646572625,
+      "learning_rate": 9.99986631866845e-06,
+      "loss": 0.5775,
+      "step": 301
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6604750547424905,
+      "learning_rate": 9.999853284044539e-06,
+      "loss": 0.5224,
+      "step": 302
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.99652793453213,
+      "learning_rate": 9.99983964317118e-06,
+      "loss": 0.5212,
+      "step": 303
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.719369992063939,
+      "learning_rate": 9.99982539605003e-06,
+      "loss": 0.5549,
+      "step": 304
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.235510791424286,
+      "learning_rate": 9.999810542682811e-06,
+      "loss": 0.5728,
+      "step": 305
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6876611055529716,
+      "learning_rate": 9.999795083071328e-06,
+      "loss": 0.302,
+      "step": 306
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9020810407363005,
+      "learning_rate": 9.999779017217454e-06,
+      "loss": 0.5006,
+      "step": 307
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.567752062992729,
+      "learning_rate": 9.999762345123137e-06,
+      "loss": 0.5953,
+      "step": 308
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7150354923946245,
+      "learning_rate": 9.999745066790398e-06,
+      "loss": 0.6027,
+      "step": 309
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5261060437216579,
+      "learning_rate": 9.999727182221335e-06,
+      "loss": 0.2931,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3350086146617444,
+      "learning_rate": 9.999708691418113e-06,
+      "loss": 0.5425,
+      "step": 311
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.011298993794657,
+      "learning_rate": 9.999689594382977e-06,
+      "loss": 0.516,
+      "step": 312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2613823465353047,
+      "learning_rate": 9.99966989111824e-06,
+      "loss": 0.5605,
+      "step": 313
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.149164003076677,
+      "learning_rate": 9.999649581626292e-06,
+      "loss": 0.5247,
+      "step": 314
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.000814750033862,
+      "learning_rate": 9.999628665909597e-06,
+      "loss": 0.5719,
+      "step": 315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2506049662349263,
+      "learning_rate": 9.999607143970689e-06,
+      "loss": 0.5207,
+      "step": 316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5555692191289926,
+      "learning_rate": 9.99958501581218e-06,
+      "loss": 0.3017,
+      "step": 317
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.657602661855044,
+      "learning_rate": 9.999562281436749e-06,
+      "loss": 0.5199,
+      "step": 318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.98162738022858,
+      "learning_rate": 9.999538940847156e-06,
+      "loss": 0.534,
+      "step": 319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3135053647047763,
+      "learning_rate": 9.999514994046232e-06,
+      "loss": 0.5631,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9386553741285684,
+      "learning_rate": 9.999490441036877e-06,
+      "loss": 0.5644,
+      "step": 321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2811445104378767,
+      "learning_rate": 9.999465281822072e-06,
+      "loss": 0.5475,
+      "step": 322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.1335788942809275,
+      "learning_rate": 9.999439516404864e-06,
+      "loss": 0.5637,
+      "step": 323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5003132364784513,
+      "learning_rate": 9.99941314478838e-06,
+      "loss": 0.5669,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.025419375838231,
+      "learning_rate": 9.999386166975818e-06,
+      "loss": 0.5324,
+      "step": 325
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.974496694999522,
+      "learning_rate": 9.999358582970444e-06,
+      "loss": 0.5876,
+      "step": 326
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0017755704317723,
+      "learning_rate": 9.99933039277561e-06,
+      "loss": 0.5325,
+      "step": 327
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3187828265158172,
+      "learning_rate": 9.999301596394728e-06,
+      "loss": 0.5883,
+      "step": 328
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.545874779579726,
+      "learning_rate": 9.999272193831294e-06,
+      "loss": 0.5643,
+      "step": 329
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.427997812985579,
+      "learning_rate": 9.999242185088871e-06,
+      "loss": 0.5985,
+      "step": 330
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.750461021203385,
+      "learning_rate": 9.999211570171098e-06,
+      "loss": 0.5609,
+      "step": 331
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9789339173013323,
+      "learning_rate": 9.999180349081688e-06,
+      "loss": 0.5364,
+      "step": 332
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.861642092128247,
+      "learning_rate": 9.999148521824424e-06,
+      "loss": 0.5402,
+      "step": 333
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.145206729748671,
+      "learning_rate": 9.999116088403168e-06,
+      "loss": 0.5569,
+      "step": 334
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0801041692843896,
+      "learning_rate": 9.999083048821852e-06,
+      "loss": 0.5817,
+      "step": 335
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1525766468571508,
+      "learning_rate": 9.999049403084481e-06,
+      "loss": 0.5669,
+      "step": 336
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.265583807527409,
+      "learning_rate": 9.999015151195137e-06,
+      "loss": 0.5396,
+      "step": 337
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4996952300762345,
+      "learning_rate": 9.99898029315797e-06,
+      "loss": 0.5396,
+      "step": 338
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.543376200206948,
+      "learning_rate": 9.998944828977209e-06,
+      "loss": 0.5114,
+      "step": 339
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7691564872396985,
+      "learning_rate": 9.998908758657153e-06,
+      "loss": 0.539,
+      "step": 340
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.490784281342028,
+      "learning_rate": 9.998872082202178e-06,
+      "loss": 0.5353,
+      "step": 341
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.466186329134036,
+      "learning_rate": 9.998834799616729e-06,
+      "loss": 0.5151,
+      "step": 342
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.977799241885556,
+      "learning_rate": 9.998796910905326e-06,
+      "loss": 0.585,
+      "step": 343
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4809556091788476,
+      "learning_rate": 9.998758416072563e-06,
+      "loss": 0.5084,
+      "step": 344
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.518960926490384,
+      "learning_rate": 9.99871931512311e-06,
+      "loss": 0.5459,
+      "step": 345
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.38297353278143,
+      "learning_rate": 9.998679608061706e-06,
+      "loss": 0.5023,
+      "step": 346
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6483796074269415,
+      "learning_rate": 9.998639294893168e-06,
+      "loss": 0.5231,
+      "step": 347
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.916471258143763,
+      "learning_rate": 9.998598375622382e-06,
+      "loss": 0.4947,
+      "step": 348
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.529447817831172,
+      "learning_rate": 9.99855685025431e-06,
+      "loss": 0.5332,
+      "step": 349
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8220651943901713,
+      "learning_rate": 9.998514718793986e-06,
+      "loss": 0.5399,
+      "step": 350
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.744512047782731,
+      "learning_rate": 9.998471981246522e-06,
+      "loss": 0.4838,
+      "step": 351
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.231385216396646,
+      "learning_rate": 9.998428637617096e-06,
+      "loss": 0.579,
+      "step": 352
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.111071139269724,
+      "learning_rate": 9.998384687910968e-06,
+      "loss": 0.5538,
+      "step": 353
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0273862021115088,
+      "learning_rate": 9.998340132133462e-06,
+      "loss": 0.5457,
+      "step": 354
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.943646972063705,
+      "learning_rate": 9.998294970289984e-06,
+      "loss": 0.5494,
+      "step": 355
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3628729345823154,
+      "learning_rate": 9.998249202386008e-06,
+      "loss": 0.5169,
+      "step": 356
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.82950828318395,
+      "learning_rate": 9.998202828427087e-06,
+      "loss": 0.5929,
+      "step": 357
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7072528536534697,
+      "learning_rate": 9.99815584841884e-06,
+      "loss": 0.5405,
+      "step": 358
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7394565744098878,
+      "learning_rate": 9.998108262366965e-06,
+      "loss": 0.5642,
+      "step": 359
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5185254276201032,
+      "learning_rate": 9.998060070277232e-06,
+      "loss": 0.5232,
+      "step": 360
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1479899581881488,
+      "learning_rate": 9.998011272155486e-06,
+      "loss": 0.5129,
+      "step": 361
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.100467572338726,
+      "learning_rate": 9.997961868007641e-06,
+      "loss": 0.522,
+      "step": 362
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5823525778975401,
+      "learning_rate": 9.99791185783969e-06,
+      "loss": 0.3156,
+      "step": 363
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.460816574776875,
+      "learning_rate": 9.997861241657695e-06,
+      "loss": 0.5587,
+      "step": 364
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.306689846369012,
+      "learning_rate": 9.997810019467794e-06,
+      "loss": 0.5648,
+      "step": 365
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1678428584227585,
+      "learning_rate": 9.997758191276199e-06,
+      "loss": 0.5292,
+      "step": 366
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9183391925559037,
+      "learning_rate": 9.997705757089193e-06,
+      "loss": 0.5349,
+      "step": 367
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.757645628525963,
+      "learning_rate": 9.997652716913135e-06,
+      "loss": 0.5589,
+      "step": 368
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.275480693306976,
+      "learning_rate": 9.997599070754454e-06,
+      "loss": 0.5375,
+      "step": 369
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.572326642298418,
+      "learning_rate": 9.997544818619658e-06,
+      "loss": 0.5486,
+      "step": 370
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.391233683734037,
+      "learning_rate": 9.99748996051532e-06,
+      "loss": 0.5854,
+      "step": 371
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.489916852427161,
+      "learning_rate": 9.997434496448099e-06,
+      "loss": 0.5622,
+      "step": 372
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.093618477210969,
+      "learning_rate": 9.997378426424716e-06,
+      "loss": 0.5151,
+      "step": 373
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.744931911950113,
+      "learning_rate": 9.99732175045197e-06,
+      "loss": 0.5675,
+      "step": 374
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.441384272664658,
+      "learning_rate": 9.997264468536734e-06,
+      "loss": 0.541,
+      "step": 375
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.925662998797883,
+      "learning_rate": 9.99720658068595e-06,
+      "loss": 0.5637,
+      "step": 376
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8541502718031921,
+      "learning_rate": 9.997148086906643e-06,
+      "loss": 0.3164,
+      "step": 377
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5050758031788236,
+      "learning_rate": 9.997088987205902e-06,
+      "loss": 0.5834,
+      "step": 378
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.746449340789178,
+      "learning_rate": 9.997029281590893e-06,
+      "loss": 0.5306,
+      "step": 379
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.492589395884123,
+      "learning_rate": 9.996968970068857e-06,
+      "loss": 0.5413,
+      "step": 380
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.1181345788051855,
+      "learning_rate": 9.996908052647107e-06,
+      "loss": 0.5858,
+      "step": 381
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5513337932638107,
+      "learning_rate": 9.996846529333028e-06,
+      "loss": 0.2931,
+      "step": 382
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8924376102054223,
+      "learning_rate": 9.99678440013408e-06,
+      "loss": 0.5095,
+      "step": 383
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.444605897547458,
+      "learning_rate": 9.996721665057796e-06,
+      "loss": 0.478,
+      "step": 384
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.892075191362373,
+      "learning_rate": 9.996658324111786e-06,
+      "loss": 0.5169,
+      "step": 385
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.35962042998851,
+      "learning_rate": 9.996594377303727e-06,
+      "loss": 0.5403,
+      "step": 386
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8237731969800235,
+      "learning_rate": 9.996529824641375e-06,
+      "loss": 0.4909,
+      "step": 387
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.832048696150894,
+      "learning_rate": 9.996464666132554e-06,
+      "loss": 0.5372,
+      "step": 388
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.438162921792171,
+      "learning_rate": 9.996398901785166e-06,
+      "loss": 0.5853,
+      "step": 389
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2890671877856366,
+      "learning_rate": 9.996332531607189e-06,
+      "loss": 0.5115,
+      "step": 390
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.524513600142909,
+      "learning_rate": 9.996265555606666e-06,
+      "loss": 0.3177,
+      "step": 391
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9915465581312746,
+      "learning_rate": 9.996197973791722e-06,
+      "loss": 0.5127,
+      "step": 392
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.733351774864903,
+      "learning_rate": 9.996129786170545e-06,
+      "loss": 0.5413,
+      "step": 393
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.976392949128436,
+      "learning_rate": 9.996060992751411e-06,
+      "loss": 0.4854,
+      "step": 394
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8143304276595624,
+      "learning_rate": 9.995991593542655e-06,
+      "loss": 0.5661,
+      "step": 395
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.557840045474204,
+      "learning_rate": 9.995921588552694e-06,
+      "loss": 0.5256,
+      "step": 396
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.451528694558367,
+      "learning_rate": 9.99585097779002e-06,
+      "loss": 0.5587,
+      "step": 397
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.500878580232607,
+      "learning_rate": 9.995779761263188e-06,
+      "loss": 0.5387,
+      "step": 398
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.612742652632811,
+      "learning_rate": 9.995707938980841e-06,
+      "loss": 0.5225,
+      "step": 399
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.670882313515096,
+      "learning_rate": 9.99563551095168e-06,
+      "loss": 0.5114,
+      "step": 400
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5267088757989296,
+      "learning_rate": 9.995562477184492e-06,
+      "loss": 0.3034,
+      "step": 401
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.830955542985375,
+      "learning_rate": 9.995488837688132e-06,
+      "loss": 0.5499,
+      "step": 402
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.089260886599526,
+      "learning_rate": 9.995414592471528e-06,
+      "loss": 0.5301,
+      "step": 403
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.234147850469226,
+      "learning_rate": 9.995339741543682e-06,
+      "loss": 0.5302,
+      "step": 404
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.067222404053777,
+      "learning_rate": 9.995264284913671e-06,
+      "loss": 0.5536,
+      "step": 405
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.742505306588949,
+      "learning_rate": 9.995188222590646e-06,
+      "loss": 0.5203,
+      "step": 406
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1655664938542434,
+      "learning_rate": 9.995111554583825e-06,
+      "loss": 0.5672,
+      "step": 407
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.783104216863022,
+      "learning_rate": 9.995034280902508e-06,
+      "loss": 0.5648,
+      "step": 408
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.874204560977364,
+      "learning_rate": 9.994956401556067e-06,
+      "loss": 0.4638,
+      "step": 409
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.080849811376308,
+      "learning_rate": 9.994877916553937e-06,
+      "loss": 0.5786,
+      "step": 410
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2901703077227427,
+      "learning_rate": 9.994798825905644e-06,
+      "loss": 0.5786,
+      "step": 411
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.49985559788896955,
+      "learning_rate": 9.99471912962077e-06,
+      "loss": 0.3017,
+      "step": 412
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.772217338113931,
+      "learning_rate": 9.994638827708986e-06,
+      "loss": 0.516,
+      "step": 413
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.131604909360112,
+      "learning_rate": 9.994557920180024e-06,
+      "loss": 0.5411,
+      "step": 414
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.847320860747283,
+      "learning_rate": 9.994476407043694e-06,
+      "loss": 0.5063,
+      "step": 415
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.119165572439268,
+      "learning_rate": 9.99439428830988e-06,
+      "loss": 0.558,
+      "step": 416
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.32054724356699,
+      "learning_rate": 9.994311563988543e-06,
+      "loss": 0.4843,
+      "step": 417
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.127722797584064,
+      "learning_rate": 9.99422823408971e-06,
+      "loss": 0.5571,
+      "step": 418
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.706333074196017,
+      "learning_rate": 9.994144298623485e-06,
+      "loss": 0.5173,
+      "step": 419
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4662105101113834,
+      "learning_rate": 9.994059757600047e-06,
+      "loss": 0.3101,
+      "step": 420
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.43258882952541755,
+      "learning_rate": 9.993974611029646e-06,
+      "loss": 0.2902,
+      "step": 421
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6044697379200827,
+      "learning_rate": 9.993888858922606e-06,
+      "loss": 0.5122,
+      "step": 422
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.383662754249409,
+      "learning_rate": 9.993802501289327e-06,
+      "loss": 0.5114,
+      "step": 423
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4488989283116247,
+      "learning_rate": 9.993715538140277e-06,
+      "loss": 0.292,
+      "step": 424
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.39799361668284977,
+      "learning_rate": 9.993627969486003e-06,
+      "loss": 0.2767,
+      "step": 425
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.42720049724416564,
+      "learning_rate": 9.993539795337122e-06,
+      "loss": 0.2813,
+      "step": 426
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6507561854976847,
+      "learning_rate": 9.993451015704324e-06,
+      "loss": 0.4926,
+      "step": 427
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3999398797929541,
+      "learning_rate": 9.993361630598377e-06,
+      "loss": 0.3144,
+      "step": 428
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3883836090090217,
+      "learning_rate": 9.993271640030116e-06,
+      "loss": 0.2918,
+      "step": 429
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.53358548288134,
+      "learning_rate": 9.993181044010456e-06,
+      "loss": 0.5368,
+      "step": 430
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.688825094564993,
+      "learning_rate": 9.993089842550379e-06,
+      "loss": 0.5253,
+      "step": 431
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.204652240270434,
+      "learning_rate": 9.992998035660946e-06,
+      "loss": 0.5511,
+      "step": 432
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.54604343589328,
+      "learning_rate": 9.992905623353287e-06,
+      "loss": 0.4978,
+      "step": 433
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.204140210205881,
+      "learning_rate": 9.992812605638606e-06,
+      "loss": 0.5118,
+      "step": 434
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.180722617953986,
+      "learning_rate": 9.992718982528187e-06,
+      "loss": 0.5072,
+      "step": 435
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.669378893189453,
+      "learning_rate": 9.992624754033377e-06,
+      "loss": 0.5416,
+      "step": 436
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.864898979182075,
+      "learning_rate": 9.992529920165603e-06,
+      "loss": 0.4911,
+      "step": 437
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7262472833912197,
+      "learning_rate": 9.992434480936365e-06,
+      "loss": 0.5425,
+      "step": 438
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.070193947568709,
+      "learning_rate": 9.992338436357235e-06,
+      "loss": 0.5157,
+      "step": 439
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.808865543912254,
+      "learning_rate": 9.992241786439859e-06,
+      "loss": 0.5391,
+      "step": 440
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6245434717216933,
+      "learning_rate": 9.992144531195955e-06,
+      "loss": 0.5171,
+      "step": 441
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.199251059453519,
+      "learning_rate": 9.992046670637316e-06,
+      "loss": 0.5596,
+      "step": 442
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.27256122917303,
+      "learning_rate": 9.991948204775808e-06,
+      "loss": 0.5309,
+      "step": 443
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1587495800861554,
+      "learning_rate": 9.99184913362337e-06,
+      "loss": 0.5623,
+      "step": 444
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.587953589310219,
+      "learning_rate": 9.991749457192015e-06,
+      "loss": 0.5598,
+      "step": 445
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.5213402891394985,
+      "learning_rate": 9.991649175493829e-06,
+      "loss": 0.5124,
+      "step": 446
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6366543169299943,
+      "learning_rate": 9.991548288540971e-06,
+      "loss": 0.5156,
+      "step": 447
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.326239938755768,
+      "learning_rate": 9.991446796345676e-06,
+      "loss": 0.5493,
+      "step": 448
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.057405748414013,
+      "learning_rate": 9.991344698920248e-06,
+      "loss": 0.582,
+      "step": 449
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.582747031193449,
+      "learning_rate": 9.991241996277068e-06,
+      "loss": 0.4904,
+      "step": 450
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.194981308492735,
+      "learning_rate": 9.991138688428588e-06,
+      "loss": 0.5776,
+      "step": 451
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.912247597382538,
+      "learning_rate": 9.991034775387336e-06,
+      "loss": 0.5152,
+      "step": 452
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.7668005024964275,
+      "learning_rate": 9.99093025716591e-06,
+      "loss": 0.488,
+      "step": 453
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7952800977287833,
+      "learning_rate": 9.990825133776983e-06,
+      "loss": 0.5541,
+      "step": 454
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.778119977733048,
+      "learning_rate": 9.990719405233302e-06,
+      "loss": 0.5152,
+      "step": 455
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.773214801265035,
+      "learning_rate": 9.99061307154769e-06,
+      "loss": 0.6025,
+      "step": 456
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4144760510317926,
+      "learning_rate": 9.990506132733036e-06,
+      "loss": 0.5709,
+      "step": 457
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6083088143241259,
+      "learning_rate": 9.99039858880231e-06,
+      "loss": 0.3225,
+      "step": 458
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.124065524121391,
+      "learning_rate": 9.99029043976855e-06,
+      "loss": 0.5223,
+      "step": 459
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0147385011046053,
+      "learning_rate": 9.990181685644869e-06,
+      "loss": 0.5517,
+      "step": 460
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.530514260009234,
+      "learning_rate": 9.990072326444456e-06,
+      "loss": 0.5024,
+      "step": 461
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.808586227277841,
+      "learning_rate": 9.98996236218057e-06,
+      "loss": 0.568,
+      "step": 462
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.564936458265453,
+      "learning_rate": 9.989851792866544e-06,
+      "loss": 0.5161,
+      "step": 463
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.128819333066748,
+      "learning_rate": 9.989740618515787e-06,
+      "loss": 0.4787,
+      "step": 464
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.782507906244099,
+      "learning_rate": 9.989628839141777e-06,
+      "loss": 0.5387,
+      "step": 465
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.06728537248518,
+      "learning_rate": 9.989516454758066e-06,
+      "loss": 0.5338,
+      "step": 466
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7539670216553893,
+      "learning_rate": 9.989403465378286e-06,
+      "loss": 0.545,
+      "step": 467
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6740234503476985,
+      "learning_rate": 9.989289871016134e-06,
+      "loss": 0.3213,
+      "step": 468
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.036313311307471,
+      "learning_rate": 9.989175671685383e-06,
+      "loss": 0.5069,
+      "step": 469
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44965860636792737,
+      "learning_rate": 9.989060867399884e-06,
+      "loss": 0.2929,
+      "step": 470
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.516520619490254,
+      "learning_rate": 9.988945458173553e-06,
+      "loss": 0.5895,
+      "step": 471
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.233309663558766,
+      "learning_rate": 9.988829444020388e-06,
+      "loss": 0.5385,
+      "step": 472
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4321775692753502,
+      "learning_rate": 9.988712824954452e-06,
+      "loss": 0.5591,
+      "step": 473
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.456467120748466,
+      "learning_rate": 9.988595600989887e-06,
+      "loss": 0.56,
+      "step": 474
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.263451941593786,
+      "learning_rate": 9.988477772140908e-06,
+      "loss": 0.5571,
+      "step": 475
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7075787302188665,
+      "learning_rate": 9.988359338421801e-06,
+      "loss": 0.5447,
+      "step": 476
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.026605308569469,
+      "learning_rate": 9.988240299846927e-06,
+      "loss": 0.5339,
+      "step": 477
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.039954636247844,
+      "learning_rate": 9.988120656430718e-06,
+      "loss": 0.3145,
+      "step": 478
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.981865139417412,
+      "learning_rate": 9.988000408187684e-06,
+      "loss": 0.5224,
+      "step": 479
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.081547578391394,
+      "learning_rate": 9.987879555132404e-06,
+      "loss": 0.5118,
+      "step": 480
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.849320918500257,
+      "learning_rate": 9.987758097279535e-06,
+      "loss": 0.5237,
+      "step": 481
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7228939211256957,
+      "learning_rate": 9.987636034643798e-06,
+      "loss": 0.5377,
+      "step": 482
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 26.636450326258927,
+      "learning_rate": 9.987513367239995e-06,
+      "loss": 0.5194,
+      "step": 483
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.055317012868984,
+      "learning_rate": 9.987390095083005e-06,
+      "loss": 0.5019,
+      "step": 484
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9445826868852363,
+      "learning_rate": 9.987266218187771e-06,
+      "loss": 0.4759,
+      "step": 485
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.220630013400035,
+      "learning_rate": 9.987141736569314e-06,
+      "loss": 0.5362,
+      "step": 486
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.768405569650566,
+      "learning_rate": 9.98701665024273e-06,
+      "loss": 0.5024,
+      "step": 487
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5913217519324876,
+      "learning_rate": 9.986890959223181e-06,
+      "loss": 0.5186,
+      "step": 488
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0786323018587862,
+      "learning_rate": 9.986764663525913e-06,
+      "loss": 0.5713,
+      "step": 489
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.666902457916991,
+      "learning_rate": 9.986637763166236e-06,
+      "loss": 0.5699,
+      "step": 490
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9003264852593627,
+      "learning_rate": 9.986510258159541e-06,
+      "loss": 0.5157,
+      "step": 491
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 18.86380819485211,
+      "learning_rate": 9.986382148521283e-06,
+      "loss": 0.5355,
+      "step": 492
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.172870299198284,
+      "learning_rate": 9.986253434267002e-06,
+      "loss": 0.4878,
+      "step": 493
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4620383003904105,
+      "learning_rate": 9.9861241154123e-06,
+      "loss": 0.5334,
+      "step": 494
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0278673320867164,
+      "learning_rate": 9.98599419197286e-06,
+      "loss": 0.5634,
+      "step": 495
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.608599982903994,
+      "learning_rate": 9.985863663964434e-06,
+      "loss": 0.5449,
+      "step": 496
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.767757899519765,
+      "learning_rate": 9.98573253140285e-06,
+      "loss": 0.5358,
+      "step": 497
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.099391693587422,
+      "learning_rate": 9.98560079430401e-06,
+      "loss": 0.5598,
+      "step": 498
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9147331340227668,
+      "learning_rate": 9.985468452683882e-06,
+      "loss": 0.3247,
+      "step": 499
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8079443409722772,
+      "learning_rate": 9.98533550655852e-06,
+      "loss": 0.3301,
+      "step": 500
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.939443521021169,
+      "learning_rate": 9.985201955944039e-06,
+      "loss": 0.5056,
+      "step": 501
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.2191369159520855,
+      "learning_rate": 9.985067800856636e-06,
+      "loss": 0.5659,
+      "step": 502
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.484560996424138,
+      "learning_rate": 9.984933041312575e-06,
+      "loss": 0.5272,
+      "step": 503
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6771613270865884,
+      "learning_rate": 9.984797677328195e-06,
+      "loss": 0.4791,
+      "step": 504
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.664255277432107,
+      "learning_rate": 9.984661708919913e-06,
+      "loss": 0.5619,
+      "step": 505
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.060539286076406,
+      "learning_rate": 9.984525136104214e-06,
+      "loss": 0.5566,
+      "step": 506
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.305745988149313,
+      "learning_rate": 9.98438795889766e-06,
+      "loss": 0.5248,
+      "step": 507
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.213481907356607,
+      "learning_rate": 9.984250177316881e-06,
+      "loss": 0.4986,
+      "step": 508
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5265652524862654,
+      "learning_rate": 9.984111791378583e-06,
+      "loss": 0.5651,
+      "step": 509
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.664000845433132,
+      "learning_rate": 9.983972801099548e-06,
+      "loss": 0.5375,
+      "step": 510
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6964974238414996,
+      "learning_rate": 9.98383320649663e-06,
+      "loss": 0.5387,
+      "step": 511
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.146181424580022,
+      "learning_rate": 9.983693007586754e-06,
+      "loss": 0.4523,
+      "step": 512
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7270321763413072,
+      "learning_rate": 9.983552204386916e-06,
+      "loss": 0.4912,
+      "step": 513
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0719489073487067,
+      "learning_rate": 9.983410796914197e-06,
+      "loss": 0.5199,
+      "step": 514
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.541301155317855,
+      "learning_rate": 9.983268785185736e-06,
+      "loss": 0.559,
+      "step": 515
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.654250225710787,
+      "learning_rate": 9.983126169218754e-06,
+      "loss": 0.5742,
+      "step": 516
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.196488646309432,
+      "learning_rate": 9.982982949030546e-06,
+      "loss": 0.5509,
+      "step": 517
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.539568074045498,
+      "learning_rate": 9.982839124638474e-06,
+      "loss": 0.5736,
+      "step": 518
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.462824928058448,
+      "learning_rate": 9.982694696059982e-06,
+      "loss": 0.5626,
+      "step": 519
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.005484969499044,
+      "learning_rate": 9.982549663312582e-06,
+      "loss": 0.4804,
+      "step": 520
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.282084565973807,
+      "learning_rate": 9.982404026413854e-06,
+      "loss": 0.5343,
+      "step": 521
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.252999726526006,
+      "learning_rate": 9.982257785381464e-06,
+      "loss": 0.5068,
+      "step": 522
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.540438238694213,
+      "learning_rate": 9.982110940233139e-06,
+      "loss": 0.6013,
+      "step": 523
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.920980536592734,
+      "learning_rate": 9.981963490986687e-06,
+      "loss": 0.5041,
+      "step": 524
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.752321859113102,
+      "learning_rate": 9.981815437659987e-06,
+      "loss": 0.5427,
+      "step": 525
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5115260768607164,
+      "learning_rate": 9.98166678027099e-06,
+      "loss": 0.494,
+      "step": 526
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1842175380423234,
+      "learning_rate": 9.98151751883772e-06,
+      "loss": 0.5377,
+      "step": 527
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2333000007810355,
+      "learning_rate": 9.981367653378278e-06,
+      "loss": 0.5632,
+      "step": 528
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6834035896245587,
+      "learning_rate": 9.981217183910834e-06,
+      "loss": 0.5455,
+      "step": 529
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.097352274741521,
+      "learning_rate": 9.981066110453634e-06,
+      "loss": 0.5634,
+      "step": 530
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.447024334220751,
+      "learning_rate": 9.980914433024997e-06,
+      "loss": 0.5107,
+      "step": 531
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.097472841943782,
+      "learning_rate": 9.980762151643313e-06,
+      "loss": 0.5454,
+      "step": 532
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8154445769282868,
+      "learning_rate": 9.980609266327046e-06,
+      "loss": 0.5056,
+      "step": 533
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.498607197202146,
+      "learning_rate": 9.980455777094733e-06,
+      "loss": 0.526,
+      "step": 534
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4572801136677365,
+      "learning_rate": 9.98030168396499e-06,
+      "loss": 0.5345,
+      "step": 535
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.313819979147645,
+      "learning_rate": 9.980146986956495e-06,
+      "loss": 0.5288,
+      "step": 536
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5966708487057697,
+      "learning_rate": 9.979991686088012e-06,
+      "loss": 0.5245,
+      "step": 537
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.304970518423037,
+      "learning_rate": 9.979835781378366e-06,
+      "loss": 0.5573,
+      "step": 538
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.528255604569721,
+      "learning_rate": 9.979679272846463e-06,
+      "loss": 0.5231,
+      "step": 539
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.86363152922509,
+      "learning_rate": 9.979522160511282e-06,
+      "loss": 0.5553,
+      "step": 540
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.264572614842528,
+      "learning_rate": 9.979364444391873e-06,
+      "loss": 0.5137,
+      "step": 541
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.017604666834211,
+      "learning_rate": 9.979206124507357e-06,
+      "loss": 0.4953,
+      "step": 542
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.745400976099307,
+      "learning_rate": 9.979047200876932e-06,
+      "loss": 0.5175,
+      "step": 543
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.437715836181326,
+      "learning_rate": 9.97888767351987e-06,
+      "loss": 0.5464,
+      "step": 544
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7048808271294416,
+      "learning_rate": 9.978727542455512e-06,
+      "loss": 0.5437,
+      "step": 545
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.835617631572047,
+      "learning_rate": 9.978566807703275e-06,
+      "loss": 0.3938,
+      "step": 546
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.456143422130274,
+      "learning_rate": 9.978405469282648e-06,
+      "loss": 0.4966,
+      "step": 547
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3220812922802496,
+      "learning_rate": 9.978243527213196e-06,
+      "loss": 0.3552,
+      "step": 548
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8363270964250553,
+      "learning_rate": 9.978080981514554e-06,
+      "loss": 0.5319,
+      "step": 549
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.0562771022040005,
+      "learning_rate": 9.977917832206431e-06,
+      "loss": 0.5224,
+      "step": 550
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.451410384013402,
+      "learning_rate": 9.977754079308608e-06,
+      "loss": 0.5258,
+      "step": 551
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.347217973510809,
+      "learning_rate": 9.977589722840943e-06,
+      "loss": 0.5432,
+      "step": 552
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.540594042168321,
+      "learning_rate": 9.977424762823364e-06,
+      "loss": 0.4619,
+      "step": 553
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6780025765701634,
+      "learning_rate": 9.977259199275871e-06,
+      "loss": 0.5158,
+      "step": 554
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.473611311453498,
+      "learning_rate": 9.977093032218544e-06,
+      "loss": 0.5944,
+      "step": 555
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5524985609501734,
+      "learning_rate": 9.976926261671524e-06,
+      "loss": 0.5489,
+      "step": 556
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.750093570662159,
+      "learning_rate": 9.97675888765504e-06,
+      "loss": 0.5316,
+      "step": 557
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2348302564030456,
+      "learning_rate": 9.976590910189382e-06,
+      "loss": 0.5482,
+      "step": 558
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6014916142279185,
+      "learning_rate": 9.976422329294919e-06,
+      "loss": 0.5295,
+      "step": 559
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.405439797915125,
+      "learning_rate": 9.976253144992094e-06,
+      "loss": 0.5661,
+      "step": 560
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.475032914327279,
+      "learning_rate": 9.976083357301419e-06,
+      "loss": 0.5356,
+      "step": 561
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1143196725149864,
+      "learning_rate": 9.97591296624348e-06,
+      "loss": 0.5462,
+      "step": 562
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.399193981697957,
+      "learning_rate": 9.97574197183894e-06,
+      "loss": 0.5243,
+      "step": 563
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.678630852237161,
+      "learning_rate": 9.975570374108531e-06,
+      "loss": 0.5142,
+      "step": 564
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.034935090939391,
+      "learning_rate": 9.975398173073064e-06,
+      "loss": 0.5442,
+      "step": 565
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2578757632885367,
+      "learning_rate": 9.975225368753412e-06,
+      "loss": 0.5583,
+      "step": 566
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3911678974179456,
+      "learning_rate": 9.975051961170533e-06,
+      "loss": 0.5159,
+      "step": 567
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6856245632145934,
+      "learning_rate": 9.974877950345453e-06,
+      "loss": 0.5392,
+      "step": 568
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6854259352738197,
+      "learning_rate": 9.97470333629927e-06,
+      "loss": 0.5771,
+      "step": 569
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.16339486445922,
+      "learning_rate": 9.974528119053158e-06,
+      "loss": 0.5931,
+      "step": 570
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.223460368179565,
+      "learning_rate": 9.97435229862836e-06,
+      "loss": 0.589,
+      "step": 571
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.600940076415167,
+      "learning_rate": 9.974175875046197e-06,
+      "loss": 0.4665,
+      "step": 572
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.666145879375337,
+      "learning_rate": 9.973998848328062e-06,
+      "loss": 0.5527,
+      "step": 573
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.46517515201637,
+      "learning_rate": 9.973821218495417e-06,
+      "loss": 0.4736,
+      "step": 574
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3723960792964767,
+      "learning_rate": 9.973642985569803e-06,
+      "loss": 0.4905,
+      "step": 575
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.016249870118298,
+      "learning_rate": 9.973464149572829e-06,
+      "loss": 0.3587,
+      "step": 576
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8227173908155496,
+      "learning_rate": 9.97328471052618e-06,
+      "loss": 0.5319,
+      "step": 577
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6887676654730489,
+      "learning_rate": 9.973104668451617e-06,
+      "loss": 0.3175,
+      "step": 578
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.380668977193693,
+      "learning_rate": 9.972924023370967e-06,
+      "loss": 0.5708,
+      "step": 579
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6462546289224265,
+      "learning_rate": 9.972742775306135e-06,
+      "loss": 0.5864,
+      "step": 580
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.790019717941242,
+      "learning_rate": 9.972560924279098e-06,
+      "loss": 0.5715,
+      "step": 581
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.176268895550913,
+      "learning_rate": 9.972378470311906e-06,
+      "loss": 0.5602,
+      "step": 582
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.767712768995422,
+      "learning_rate": 9.97219541342668e-06,
+      "loss": 0.5992,
+      "step": 583
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1780307460367685,
+      "learning_rate": 9.97201175364562e-06,
+      "loss": 0.5527,
+      "step": 584
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1383218299516003,
+      "learning_rate": 9.971827490990993e-06,
+      "loss": 0.5637,
+      "step": 585
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.754079216468853,
+      "learning_rate": 9.971642625485144e-06,
+      "loss": 0.5546,
+      "step": 586
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.848446895200657,
+      "learning_rate": 9.971457157150485e-06,
+      "loss": 0.5513,
+      "step": 587
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8051532959362744,
+      "learning_rate": 9.971271086009509e-06,
+      "loss": 0.5674,
+      "step": 588
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.925411055590649,
+      "learning_rate": 9.971084412084773e-06,
+      "loss": 0.5481,
+      "step": 589
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7737996795332406,
+      "learning_rate": 9.970897135398914e-06,
+      "loss": 0.3551,
+      "step": 590
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6098005565981709,
+      "learning_rate": 9.97070925597464e-06,
+      "loss": 0.3028,
+      "step": 591
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.980891720384619,
+      "learning_rate": 9.970520773834734e-06,
+      "loss": 0.5039,
+      "step": 592
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46114136194096167,
+      "learning_rate": 9.970331689002046e-06,
+      "loss": 0.3061,
+      "step": 593
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6123561975905485,
+      "learning_rate": 9.970142001499505e-06,
+      "loss": 0.5363,
+      "step": 594
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.484343496200627,
+      "learning_rate": 9.969951711350114e-06,
+      "loss": 0.544,
+      "step": 595
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2656894447539275,
+      "learning_rate": 9.969760818576943e-06,
+      "loss": 0.5076,
+      "step": 596
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0372590283419627,
+      "learning_rate": 9.969569323203138e-06,
+      "loss": 0.5371,
+      "step": 597
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6724624289185117,
+      "learning_rate": 9.96937722525192e-06,
+      "loss": 0.5462,
+      "step": 598
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1396309636887727,
+      "learning_rate": 9.969184524746585e-06,
+      "loss": 0.5564,
+      "step": 599
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1712463195055047,
+      "learning_rate": 9.96899122171049e-06,
+      "loss": 0.6062,
+      "step": 600
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.07696953112035,
+      "learning_rate": 9.968797316167083e-06,
+      "loss": 0.5523,
+      "step": 601
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8229687088716413,
+      "learning_rate": 9.96860280813987e-06,
+      "loss": 0.5604,
+      "step": 602
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8311600288886243,
+      "learning_rate": 9.968407697652436e-06,
+      "loss": 0.343,
+      "step": 603
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.804451550199023,
+      "learning_rate": 9.96821198472844e-06,
+      "loss": 0.5298,
+      "step": 604
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.257631559935912,
+      "learning_rate": 9.968015669391613e-06,
+      "loss": 0.5387,
+      "step": 605
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1651571562511767,
+      "learning_rate": 9.967818751665759e-06,
+      "loss": 0.5791,
+      "step": 606
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.01551909800727,
+      "learning_rate": 9.967621231574754e-06,
+      "loss": 0.514,
+      "step": 607
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.40170698128064,
+      "learning_rate": 9.96742310914255e-06,
+      "loss": 0.5585,
+      "step": 608
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.3996582499889785,
+      "learning_rate": 9.967224384393168e-06,
+      "loss": 0.5669,
+      "step": 609
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8314979071498723,
+      "learning_rate": 9.967025057350705e-06,
+      "loss": 0.4908,
+      "step": 610
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.435202597988599,
+      "learning_rate": 9.96682512803933e-06,
+      "loss": 0.5311,
+      "step": 611
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.44266239483541,
+      "learning_rate": 9.966624596483284e-06,
+      "loss": 0.575,
+      "step": 612
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5901821397909455,
+      "learning_rate": 9.966423462706885e-06,
+      "loss": 0.5302,
+      "step": 613
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.901006288941655,
+      "learning_rate": 9.966221726734518e-06,
+      "loss": 0.5802,
+      "step": 614
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7443731497510826,
+      "learning_rate": 9.966019388590645e-06,
+      "loss": 0.5813,
+      "step": 615
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.757749827557112,
+      "learning_rate": 9.965816448299801e-06,
+      "loss": 0.5329,
+      "step": 616
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.409820562077231,
+      "learning_rate": 9.965612905886593e-06,
+      "loss": 0.5513,
+      "step": 617
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.281929314132998,
+      "learning_rate": 9.965408761375702e-06,
+      "loss": 0.5066,
+      "step": 618
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.392657701176082,
+      "learning_rate": 9.965204014791879e-06,
+      "loss": 0.5119,
+      "step": 619
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.750745550278187,
+      "learning_rate": 9.964998666159952e-06,
+      "loss": 0.5347,
+      "step": 620
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.562025327617273,
+      "learning_rate": 9.96479271550482e-06,
+      "loss": 0.4893,
+      "step": 621
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.8999762659060595,
+      "learning_rate": 9.964586162851455e-06,
+      "loss": 0.5157,
+      "step": 622
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.432847983714552,
+      "learning_rate": 9.964379008224902e-06,
+      "loss": 0.5383,
+      "step": 623
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8846038629241746,
+      "learning_rate": 9.96417125165028e-06,
+      "loss": 0.5595,
+      "step": 624
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.254693629684428,
+      "learning_rate": 9.963962893152779e-06,
+      "loss": 0.5338,
+      "step": 625
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4654719273176244,
+      "learning_rate": 9.963753932757663e-06,
+      "loss": 0.5116,
+      "step": 626
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1819842035060666,
+      "learning_rate": 9.96354437049027e-06,
+      "loss": 0.4902,
+      "step": 627
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.077945278253239,
+      "learning_rate": 9.963334206376011e-06,
+      "loss": 0.5294,
+      "step": 628
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.49627850259544154,
+      "learning_rate": 9.96312344044037e-06,
+      "loss": 0.3126,
+      "step": 629
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.963490618549204,
+      "learning_rate": 9.962912072708898e-06,
+      "loss": 0.5124,
+      "step": 630
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6832464238726432,
+      "learning_rate": 9.96270010320723e-06,
+      "loss": 0.5592,
+      "step": 631
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.896136222153313,
+      "learning_rate": 9.962487531961064e-06,
+      "loss": 0.5549,
+      "step": 632
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5922522509304486,
+      "learning_rate": 9.962274358996178e-06,
+      "loss": 0.4986,
+      "step": 633
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.283601813245235,
+      "learning_rate": 9.962060584338416e-06,
+      "loss": 0.5334,
+      "step": 634
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1679629225684844,
+      "learning_rate": 9.961846208013703e-06,
+      "loss": 0.5502,
+      "step": 635
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2514022839294277,
+      "learning_rate": 9.961631230048032e-06,
+      "loss": 0.5266,
+      "step": 636
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.46091194325966195,
+      "learning_rate": 9.961415650467468e-06,
+      "loss": 0.3131,
+      "step": 637
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.951556220579447,
+      "learning_rate": 9.961199469298151e-06,
+      "loss": 0.5754,
+      "step": 638
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4875104047133254,
+      "learning_rate": 9.960982686566295e-06,
+      "loss": 0.5156,
+      "step": 639
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3657494631614058,
+      "learning_rate": 9.960765302298185e-06,
+      "loss": 0.5089,
+      "step": 640
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.325900437016625,
+      "learning_rate": 9.960547316520182e-06,
+      "loss": 0.5569,
+      "step": 641
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4018451681478655,
+      "learning_rate": 9.960328729258713e-06,
+      "loss": 0.3141,
+      "step": 642
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.754103900717141,
+      "learning_rate": 9.960109540540283e-06,
+      "loss": 0.5797,
+      "step": 643
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9717326355068123,
+      "learning_rate": 9.959889750391474e-06,
+      "loss": 0.5511,
+      "step": 644
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6833240019542393,
+      "learning_rate": 9.959669358838932e-06,
+      "loss": 0.5333,
+      "step": 645
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.261135233538854,
+      "learning_rate": 9.959448365909383e-06,
+      "loss": 0.51,
+      "step": 646
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.640195398740264,
+      "learning_rate": 9.959226771629623e-06,
+      "loss": 0.4662,
+      "step": 647
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8116594216138786,
+      "learning_rate": 9.959004576026516e-06,
+      "loss": 0.517,
+      "step": 648
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.205470720983286,
+      "learning_rate": 9.95878177912701e-06,
+      "loss": 0.5015,
+      "step": 649
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43758942617771623,
+      "learning_rate": 9.958558380958118e-06,
+      "loss": 0.2997,
+      "step": 650
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.772394561266011,
+      "learning_rate": 9.958334381546927e-06,
+      "loss": 0.5651,
+      "step": 651
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.866512432338939,
+      "learning_rate": 9.958109780920598e-06,
+      "loss": 0.4874,
+      "step": 652
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9145577252979646,
+      "learning_rate": 9.957884579106365e-06,
+      "loss": 0.5439,
+      "step": 653
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.362064559253772,
+      "learning_rate": 9.957658776131536e-06,
+      "loss": 0.5359,
+      "step": 654
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.962920028441446,
+      "learning_rate": 9.957432372023487e-06,
+      "loss": 0.5597,
+      "step": 655
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3923317982461118,
+      "learning_rate": 9.957205366809673e-06,
+      "loss": 0.3156,
+      "step": 656
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.40737016036122925,
+      "learning_rate": 9.956977760517617e-06,
+      "loss": 0.3038,
+      "step": 657
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.675059791416387,
+      "learning_rate": 9.956749553174919e-06,
+      "loss": 0.5159,
+      "step": 658
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.532453316799696,
+      "learning_rate": 9.956520744809249e-06,
+      "loss": 0.5237,
+      "step": 659
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4755632321034953,
+      "learning_rate": 9.956291335448351e-06,
+      "loss": 0.5427,
+      "step": 660
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4141503753398195,
+      "learning_rate": 9.95606132512004e-06,
+      "loss": 0.5286,
+      "step": 661
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7123531224254274,
+      "learning_rate": 9.95583071385221e-06,
+      "loss": 0.5808,
+      "step": 662
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4131677252708252,
+      "learning_rate": 9.95559950167282e-06,
+      "loss": 0.5416,
+      "step": 663
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.112848818872381,
+      "learning_rate": 9.955367688609904e-06,
+      "loss": 0.5722,
+      "step": 664
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.111952136156833,
+      "learning_rate": 9.955135274691575e-06,
+      "loss": 0.5575,
+      "step": 665
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3218253869904233,
+      "learning_rate": 9.954902259946008e-06,
+      "loss": 0.5158,
+      "step": 666
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5020801013742515,
+      "learning_rate": 9.954668644401462e-06,
+      "loss": 0.5797,
+      "step": 667
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9078217487692166,
+      "learning_rate": 9.95443442808626e-06,
+      "loss": 0.5531,
+      "step": 668
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.356726551615006,
+      "learning_rate": 9.954199611028803e-06,
+      "loss": 0.5164,
+      "step": 669
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.013036259282614,
+      "learning_rate": 9.953964193257563e-06,
+      "loss": 0.5588,
+      "step": 670
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.545771192539764,
+      "learning_rate": 9.953728174801089e-06,
+      "loss": 0.5489,
+      "step": 671
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0188749926618033,
+      "learning_rate": 9.953491555687994e-06,
+      "loss": 0.5889,
+      "step": 672
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.51874657963977,
+      "learning_rate": 9.95325433594697e-06,
+      "loss": 0.5055,
+      "step": 673
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5376420416483615,
+      "learning_rate": 9.95301651560678e-06,
+      "loss": 0.3046,
+      "step": 674
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4872458808142259,
+      "learning_rate": 9.952778094696263e-06,
+      "loss": 0.2959,
+      "step": 675
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.407810720814409,
+      "learning_rate": 9.952539073244326e-06,
+      "loss": 0.5299,
+      "step": 676
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4114011145154149,
+      "learning_rate": 9.952299451279956e-06,
+      "loss": 0.3054,
+      "step": 677
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7930397248042786,
+      "learning_rate": 9.952059228832202e-06,
+      "loss": 0.5859,
+      "step": 678
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1596500985586173,
+      "learning_rate": 9.951818405930194e-06,
+      "loss": 0.4833,
+      "step": 679
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.717317879169971,
+      "learning_rate": 9.951576982603133e-06,
+      "loss": 0.5788,
+      "step": 680
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5866263016192783,
+      "learning_rate": 9.951334958880292e-06,
+      "loss": 0.5612,
+      "step": 681
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.594283357642824,
+      "learning_rate": 9.951092334791018e-06,
+      "loss": 0.511,
+      "step": 682
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.065178591694827,
+      "learning_rate": 9.950849110364729e-06,
+      "loss": 0.5544,
+      "step": 683
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.785545709421961,
+      "learning_rate": 9.950605285630917e-06,
+      "loss": 0.5348,
+      "step": 684
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.367118491390597,
+      "learning_rate": 9.950360860619147e-06,
+      "loss": 0.5391,
+      "step": 685
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.377343897492596,
+      "learning_rate": 9.950115835359055e-06,
+      "loss": 0.5143,
+      "step": 686
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6953422930511235,
+      "learning_rate": 9.949870209880355e-06,
+      "loss": 0.5649,
+      "step": 687
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.8926363835702,
+      "learning_rate": 9.949623984212825e-06,
+      "loss": 0.5167,
+      "step": 688
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.460888300012385,
+      "learning_rate": 9.949377158386324e-06,
+      "loss": 0.5379,
+      "step": 689
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1820517288670924,
+      "learning_rate": 9.949129732430779e-06,
+      "loss": 0.5602,
+      "step": 690
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3848103990494987,
+      "learning_rate": 9.948881706376193e-06,
+      "loss": 0.5382,
+      "step": 691
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6887016380147245,
+      "learning_rate": 9.948633080252637e-06,
+      "loss": 0.3221,
+      "step": 692
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.588578913388952,
+      "learning_rate": 9.94838385409026e-06,
+      "loss": 0.5163,
+      "step": 693
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5257800301595572,
+      "learning_rate": 9.94813402791928e-06,
+      "loss": 0.3029,
+      "step": 694
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.007198275756981,
+      "learning_rate": 9.947883601769992e-06,
+      "loss": 0.5071,
+      "step": 695
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.156184561849983,
+      "learning_rate": 9.947632575672758e-06,
+      "loss": 0.5353,
+      "step": 696
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.827705481223566,
+      "learning_rate": 9.947380949658018e-06,
+      "loss": 0.5131,
+      "step": 697
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1187598152831004,
+      "learning_rate": 9.947128723756282e-06,
+      "loss": 0.4901,
+      "step": 698
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5683989285724226,
+      "learning_rate": 9.946875897998131e-06,
+      "loss": 0.3163,
+      "step": 699
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.846308254645736,
+      "learning_rate": 9.946622472414224e-06,
+      "loss": 0.5053,
+      "step": 700
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5317655931511314,
+      "learning_rate": 9.946368447035291e-06,
+      "loss": 0.5243,
+      "step": 701
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 22.770944530255502,
+      "learning_rate": 9.946113821892129e-06,
+      "loss": 0.4753,
+      "step": 702
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4999185152473992,
+      "learning_rate": 9.945858597015614e-06,
+      "loss": 0.305,
+      "step": 703
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8676166301261157,
+      "learning_rate": 9.945602772436693e-06,
+      "loss": 0.5405,
+      "step": 704
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.979273335727193,
+      "learning_rate": 9.945346348186385e-06,
+      "loss": 0.5396,
+      "step": 705
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4502079520615214,
+      "learning_rate": 9.945089324295785e-06,
+      "loss": 0.524,
+      "step": 706
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6102647162207187,
+      "learning_rate": 9.944831700796055e-06,
+      "loss": 0.5349,
+      "step": 707
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7769160434284297,
+      "learning_rate": 9.944573477718435e-06,
+      "loss": 0.5893,
+      "step": 708
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.915921810591561,
+      "learning_rate": 9.944314655094233e-06,
+      "loss": 0.522,
+      "step": 709
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.627637884365815,
+      "learning_rate": 9.944055232954834e-06,
+      "loss": 0.6113,
+      "step": 710
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0722381215515213,
+      "learning_rate": 9.943795211331692e-06,
+      "loss": 0.5229,
+      "step": 711
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.940064997555016,
+      "learning_rate": 9.943534590256338e-06,
+      "loss": 0.5685,
+      "step": 712
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5647737083189155,
+      "learning_rate": 9.943273369760373e-06,
+      "loss": 0.3122,
+      "step": 713
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.779339319032247,
+      "learning_rate": 9.943011549875468e-06,
+      "loss": 0.5261,
+      "step": 714
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4492015641441565,
+      "learning_rate": 9.942749130633371e-06,
+      "loss": 0.303,
+      "step": 715
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4924915824870144,
+      "learning_rate": 9.942486112065903e-06,
+      "loss": 0.5512,
+      "step": 716
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2657463123477264,
+      "learning_rate": 9.942222494204956e-06,
+      "loss": 0.5245,
+      "step": 717
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.953755757454533,
+      "learning_rate": 9.941958277082491e-06,
+      "loss": 0.5264,
+      "step": 718
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3068797713427567,
+      "learning_rate": 9.941693460730549e-06,
+      "loss": 0.5453,
+      "step": 719
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6876048293671593,
+      "learning_rate": 9.941428045181236e-06,
+      "loss": 0.5329,
+      "step": 720
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.665270010924138,
+      "learning_rate": 9.941162030466738e-06,
+      "loss": 0.5178,
+      "step": 721
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.558670552673825,
+      "learning_rate": 9.940895416619308e-06,
+      "loss": 0.5692,
+      "step": 722
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5933459163519134,
+      "learning_rate": 9.940628203671277e-06,
+      "loss": 0.3077,
+      "step": 723
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.234951348525082,
+      "learning_rate": 9.940360391655042e-06,
+      "loss": 0.5504,
+      "step": 724
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8397746526811245,
+      "learning_rate": 9.940091980603077e-06,
+      "loss": 0.5321,
+      "step": 725
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8694218162088234,
+      "learning_rate": 9.93982297054793e-06,
+      "loss": 0.484,
+      "step": 726
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.213811184574513,
+      "learning_rate": 9.939553361522216e-06,
+      "loss": 0.5347,
+      "step": 727
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9149960332499196,
+      "learning_rate": 9.939283153558628e-06,
+      "loss": 0.5202,
+      "step": 728
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.944313090449764,
+      "learning_rate": 9.93901234668993e-06,
+      "loss": 0.5148,
+      "step": 729
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.073140123522265,
+      "learning_rate": 9.938740940948958e-06,
+      "loss": 0.5075,
+      "step": 730
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8476111410653515,
+      "learning_rate": 9.93846893636862e-06,
+      "loss": 0.5326,
+      "step": 731
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.482084951246496,
+      "learning_rate": 9.938196332981897e-06,
+      "loss": 0.5642,
+      "step": 732
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.342025140827017,
+      "learning_rate": 9.937923130821845e-06,
+      "loss": 0.5128,
+      "step": 733
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4306803722019485,
+      "learning_rate": 9.93764932992159e-06,
+      "loss": 0.4859,
+      "step": 734
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.20922562460637,
+      "learning_rate": 9.937374930314329e-06,
+      "loss": 0.4965,
+      "step": 735
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.079100092770074,
+      "learning_rate": 9.937099932033338e-06,
+      "loss": 0.5084,
+      "step": 736
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.032655486904309,
+      "learning_rate": 9.93682433511196e-06,
+      "loss": 0.5242,
+      "step": 737
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.952722745384609,
+      "learning_rate": 9.93654813958361e-06,
+      "loss": 0.5766,
+      "step": 738
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4775576737428895,
+      "learning_rate": 9.93627134548178e-06,
+      "loss": 0.5467,
+      "step": 739
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.490029918962867,
+      "learning_rate": 9.935993952840033e-06,
+      "loss": 0.305,
+      "step": 740
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.58577349783184,
+      "learning_rate": 9.935715961691999e-06,
+      "loss": 0.5172,
+      "step": 741
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.45598422174922276,
+      "learning_rate": 9.93543737207139e-06,
+      "loss": 0.3109,
+      "step": 742
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.41526831062359143,
+      "learning_rate": 9.935158184011986e-06,
+      "loss": 0.3141,
+      "step": 743
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.772329017058322,
+      "learning_rate": 9.934878397547636e-06,
+      "loss": 0.507,
+      "step": 744
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6682711490353856,
+      "learning_rate": 9.934598012712268e-06,
+      "loss": 0.5119,
+      "step": 745
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 71.73069722428899,
+      "learning_rate": 9.93431702953988e-06,
+      "loss": 0.5304,
+      "step": 746
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.367987811778251,
+      "learning_rate": 9.934035448064538e-06,
+      "loss": 0.5354,
+      "step": 747
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0167208183569296,
+      "learning_rate": 9.933753268320391e-06,
+      "loss": 0.5144,
+      "step": 748
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.283296196099581,
+      "learning_rate": 9.933470490341651e-06,
+      "loss": 0.5576,
+      "step": 749
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0836446066753798,
+      "learning_rate": 9.933187114162604e-06,
+      "loss": 0.5942,
+      "step": 750
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.812876907378693,
+      "learning_rate": 9.932903139817612e-06,
+      "loss": 0.5685,
+      "step": 751
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.162408791262869,
+      "learning_rate": 9.932618567341107e-06,
+      "loss": 0.5272,
+      "step": 752
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.657024440528312,
+      "learning_rate": 9.932333396767597e-06,
+      "loss": 0.4943,
+      "step": 753
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7628344125892044,
+      "learning_rate": 9.932047628131659e-06,
+      "loss": 0.4918,
+      "step": 754
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0590343633029042,
+      "learning_rate": 9.931761261467943e-06,
+      "loss": 0.5616,
+      "step": 755
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.776545907597171,
+      "learning_rate": 9.93147429681117e-06,
+      "loss": 0.5213,
+      "step": 756
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.910522670525012,
+      "learning_rate": 9.931186734196137e-06,
+      "loss": 0.5061,
+      "step": 757
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6794636366013456,
+      "learning_rate": 9.930898573657712e-06,
+      "loss": 0.535,
+      "step": 758
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.071499369546938,
+      "learning_rate": 9.930609815230838e-06,
+      "loss": 0.5552,
+      "step": 759
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9795077829775286,
+      "learning_rate": 9.930320458950525e-06,
+      "loss": 0.3421,
+      "step": 760
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.052178035301522,
+      "learning_rate": 9.930030504851857e-06,
+      "loss": 0.5846,
+      "step": 761
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.756803475356955,
+      "learning_rate": 9.929739952969995e-06,
+      "loss": 0.5279,
+      "step": 762
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.41640432839740277,
+      "learning_rate": 9.929448803340167e-06,
+      "loss": 0.2918,
+      "step": 763
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8875108853070874,
+      "learning_rate": 9.929157055997677e-06,
+      "loss": 0.5655,
+      "step": 764
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1206225214062115,
+      "learning_rate": 9.928864710977903e-06,
+      "loss": 0.5269,
+      "step": 765
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9695905978979495,
+      "learning_rate": 9.928571768316289e-06,
+      "loss": 0.5293,
+      "step": 766
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6312200044110194,
+      "learning_rate": 9.928278228048358e-06,
+      "loss": 0.5851,
+      "step": 767
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.035678863357291,
+      "learning_rate": 9.927984090209698e-06,
+      "loss": 0.592,
+      "step": 768
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7285037603921037,
+      "learning_rate": 9.927689354835983e-06,
+      "loss": 0.3205,
+      "step": 769
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.708154516972819,
+      "learning_rate": 9.927394021962942e-06,
+      "loss": 0.5838,
+      "step": 770
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6598415589030404,
+      "learning_rate": 9.927098091626388e-06,
+      "loss": 0.5921,
+      "step": 771
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.88416573795974,
+      "learning_rate": 9.926801563862205e-06,
+      "loss": 0.5353,
+      "step": 772
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6225904486063607,
+      "learning_rate": 9.926504438706348e-06,
+      "loss": 0.5662,
+      "step": 773
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.387427664642408,
+      "learning_rate": 9.926206716194842e-06,
+      "loss": 0.5573,
+      "step": 774
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.521505782917352,
+      "learning_rate": 9.92590839636379e-06,
+      "loss": 0.4958,
+      "step": 775
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.931708792212651,
+      "learning_rate": 9.925609479249363e-06,
+      "loss": 0.5448,
+      "step": 776
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.089881283870158,
+      "learning_rate": 9.925309964887803e-06,
+      "loss": 0.5218,
+      "step": 777
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.267311375420507,
+      "learning_rate": 9.925009853315433e-06,
+      "loss": 0.5753,
+      "step": 778
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.25912511697125,
+      "learning_rate": 9.924709144568635e-06,
+      "loss": 0.5288,
+      "step": 779
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.714618407281405,
+      "learning_rate": 9.924407838683878e-06,
+      "loss": 0.5341,
+      "step": 780
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6768304814834227,
+      "learning_rate": 9.924105935697694e-06,
+      "loss": 0.532,
+      "step": 781
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6622075220486225,
+      "learning_rate": 9.923803435646688e-06,
+      "loss": 0.5204,
+      "step": 782
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.952192219954515,
+      "learning_rate": 9.923500338567542e-06,
+      "loss": 0.5018,
+      "step": 783
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.309338912727521,
+      "learning_rate": 9.923196644497004e-06,
+      "loss": 0.5086,
+      "step": 784
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.87658092496677,
+      "learning_rate": 9.9228923534719e-06,
+      "loss": 0.5697,
+      "step": 785
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9169418708639774,
+      "learning_rate": 9.922587465529128e-06,
+      "loss": 0.5862,
+      "step": 786
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8568211644697223,
+      "learning_rate": 9.922281980705654e-06,
+      "loss": 0.5028,
+      "step": 787
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5484101626144877,
+      "learning_rate": 9.921975899038521e-06,
+      "loss": 0.4823,
+      "step": 788
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.5876948164407425,
+      "learning_rate": 9.92166922056484e-06,
+      "loss": 0.5258,
+      "step": 789
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7726706483993677,
+      "learning_rate": 9.921361945321801e-06,
+      "loss": 0.5611,
+      "step": 790
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.414654887024615,
+      "learning_rate": 9.921054073346658e-06,
+      "loss": 0.4947,
+      "step": 791
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7666068702468154,
+      "learning_rate": 9.920745604676744e-06,
+      "loss": 0.5232,
+      "step": 792
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8739030931396834,
+      "learning_rate": 9.92043653934946e-06,
+      "loss": 0.5665,
+      "step": 793
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.162324951841285,
+      "learning_rate": 9.920126877402283e-06,
+      "loss": 0.5084,
+      "step": 794
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.882817578035477,
+      "learning_rate": 9.91981661887276e-06,
+      "loss": 0.497,
+      "step": 795
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.247030853607853,
+      "learning_rate": 9.919505763798508e-06,
+      "loss": 0.5635,
+      "step": 796
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.457349808806143,
+      "learning_rate": 9.919194312217227e-06,
+      "loss": 0.4655,
+      "step": 797
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7469570217372774,
+      "learning_rate": 9.918882264166672e-06,
+      "loss": 0.5313,
+      "step": 798
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.245450632932096,
+      "learning_rate": 9.918569619684687e-06,
+      "loss": 0.4999,
+      "step": 799
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8016790843580006,
+      "learning_rate": 9.918256378809178e-06,
+      "loss": 0.5656,
+      "step": 800
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0527000917632434,
+      "learning_rate": 9.917942541578127e-06,
+      "loss": 0.5161,
+      "step": 801
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4885522757249546,
+      "learning_rate": 9.917628108029589e-06,
+      "loss": 0.3181,
+      "step": 802
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.758941821762297,
+      "learning_rate": 9.91731307820169e-06,
+      "loss": 0.5046,
+      "step": 803
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.708899807757208,
+      "learning_rate": 9.916997452132625e-06,
+      "loss": 0.5318,
+      "step": 804
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.253529598594698,
+      "learning_rate": 9.91668122986067e-06,
+      "loss": 0.5527,
+      "step": 805
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.39783407443025326,
+      "learning_rate": 9.916364411424164e-06,
+      "loss": 0.3012,
+      "step": 806
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.4878144538009925,
+      "learning_rate": 9.916046996861524e-06,
+      "loss": 0.4977,
+      "step": 807
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.384843493516561,
+      "learning_rate": 9.915728986211237e-06,
+      "loss": 0.5271,
+      "step": 808
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 63.01291610773949,
+      "learning_rate": 9.915410379511866e-06,
+      "loss": 0.5389,
+      "step": 809
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9832454185301227,
+      "learning_rate": 9.915091176802037e-06,
+      "loss": 0.5283,
+      "step": 810
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4347788607263533,
+      "learning_rate": 9.91477137812046e-06,
+      "loss": 0.3168,
+      "step": 811
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6890445579570534,
+      "learning_rate": 9.914450983505908e-06,
+      "loss": 0.55,
+      "step": 812
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.991722873841082,
+      "learning_rate": 9.914129992997232e-06,
+      "loss": 0.4963,
+      "step": 813
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4372558618791543,
+      "learning_rate": 9.913808406633354e-06,
+      "loss": 0.3192,
+      "step": 814
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3644209937238267,
+      "learning_rate": 9.913486224453266e-06,
+      "loss": 0.5458,
+      "step": 815
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1478814098014203,
+      "learning_rate": 9.913163446496034e-06,
+      "loss": 0.4909,
+      "step": 816
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7220252515775862,
+      "learning_rate": 9.912840072800795e-06,
+      "loss": 0.5783,
+      "step": 817
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8583440133365667,
+      "learning_rate": 9.912516103406759e-06,
+      "loss": 0.5124,
+      "step": 818
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3825480556279457,
+      "learning_rate": 9.912191538353213e-06,
+      "loss": 0.5134,
+      "step": 819
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.60160894296651,
+      "learning_rate": 9.911866377679505e-06,
+      "loss": 0.4787,
+      "step": 820
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.092750431950959,
+      "learning_rate": 9.911540621425067e-06,
+      "loss": 0.518,
+      "step": 821
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4360222716243771,
+      "learning_rate": 9.911214269629394e-06,
+      "loss": 0.3137,
+      "step": 822
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.0792653525149545,
+      "learning_rate": 9.91088732233206e-06,
+      "loss": 0.5135,
+      "step": 823
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1624063099221202,
+      "learning_rate": 9.910559779572708e-06,
+      "loss": 0.5572,
+      "step": 824
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.154454248484027,
+      "learning_rate": 9.910231641391053e-06,
+      "loss": 0.5506,
+      "step": 825
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5041761037631964,
+      "learning_rate": 9.909902907826884e-06,
+      "loss": 0.5466,
+      "step": 826
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.415879227327702,
+      "learning_rate": 9.90957357892006e-06,
+      "loss": 0.5692,
+      "step": 827
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7136547859861704,
+      "learning_rate": 9.909243654710514e-06,
+      "loss": 0.5683,
+      "step": 828
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7221139696877277,
+      "learning_rate": 9.90891313523825e-06,
+      "loss": 0.5107,
+      "step": 829
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.577903039573576,
+      "learning_rate": 9.908582020543345e-06,
+      "loss": 0.5268,
+      "step": 830
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.059453696035629,
+      "learning_rate": 9.908250310665947e-06,
+      "loss": 0.5601,
+      "step": 831
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.67025329871377,
+      "learning_rate": 9.907918005646277e-06,
+      "loss": 0.548,
+      "step": 832
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.271226105826306,
+      "learning_rate": 9.90758510552463e-06,
+      "loss": 0.497,
+      "step": 833
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.501389672067488,
+      "learning_rate": 9.907251610341369e-06,
+      "loss": 0.5222,
+      "step": 834
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.160222972498446,
+      "learning_rate": 9.906917520136932e-06,
+      "loss": 0.5604,
+      "step": 835
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5504851563054136,
+      "learning_rate": 9.90658283495183e-06,
+      "loss": 0.5467,
+      "step": 836
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4911565638084747,
+      "learning_rate": 9.906247554826645e-06,
+      "loss": 0.4923,
+      "step": 837
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.048017962156097,
+      "learning_rate": 9.905911679802028e-06,
+      "loss": 0.5481,
+      "step": 838
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5811479139920973,
+      "learning_rate": 9.905575209918705e-06,
+      "loss": 0.3138,
+      "step": 839
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7259132845405443,
+      "learning_rate": 9.905238145217477e-06,
+      "loss": 0.5167,
+      "step": 840
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6678993910677526,
+      "learning_rate": 9.904900485739215e-06,
+      "loss": 0.5664,
+      "step": 841
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5998010062268033,
+      "learning_rate": 9.904562231524856e-06,
+      "loss": 0.477,
+      "step": 842
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.56367090944268,
+      "learning_rate": 9.904223382615418e-06,
+      "loss": 0.5772,
+      "step": 843
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.694748738735068,
+      "learning_rate": 9.90388393905199e-06,
+      "loss": 0.5774,
+      "step": 844
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.137314108495832,
+      "learning_rate": 9.903543900875727e-06,
+      "loss": 0.5544,
+      "step": 845
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.939567951492156,
+      "learning_rate": 9.903203268127861e-06,
+      "loss": 0.4669,
+      "step": 846
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.607298149046012,
+      "learning_rate": 9.902862040849694e-06,
+      "loss": 0.586,
+      "step": 847
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.555370203420079,
+      "learning_rate": 9.902520219082604e-06,
+      "loss": 0.551,
+      "step": 848
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.988227724582186,
+      "learning_rate": 9.902177802868033e-06,
+      "loss": 0.5491,
+      "step": 849
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.096792366403509,
+      "learning_rate": 9.901834792247504e-06,
+      "loss": 0.4923,
+      "step": 850
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.761576178470272,
+      "learning_rate": 9.901491187262608e-06,
+      "loss": 0.4652,
+      "step": 851
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.960589638549929,
+      "learning_rate": 9.901146987955008e-06,
+      "loss": 0.5305,
+      "step": 852
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5655563582530783,
+      "learning_rate": 9.900802194366439e-06,
+      "loss": 0.5489,
+      "step": 853
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5298273944367016,
+      "learning_rate": 9.900456806538707e-06,
+      "loss": 0.4751,
+      "step": 854
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.505671210743426,
+      "learning_rate": 9.900110824513693e-06,
+      "loss": 0.518,
+      "step": 855
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.474738762350817,
+      "learning_rate": 9.899764248333348e-06,
+      "loss": 0.5584,
+      "step": 856
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7260249446987337,
+      "learning_rate": 9.899417078039695e-06,
+      "loss": 0.5459,
+      "step": 857
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.921706100067747,
+      "learning_rate": 9.899069313674832e-06,
+      "loss": 0.5285,
+      "step": 858
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9196901213703543,
+      "learning_rate": 9.898720955280926e-06,
+      "loss": 0.465,
+      "step": 859
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.439941395170014,
+      "learning_rate": 9.898372002900214e-06,
+      "loss": 0.5332,
+      "step": 860
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.75227317436376,
+      "learning_rate": 9.89802245657501e-06,
+      "loss": 0.5245,
+      "step": 861
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.165047629449932,
+      "learning_rate": 9.897672316347697e-06,
+      "loss": 0.5066,
+      "step": 862
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5389842038827366,
+      "learning_rate": 9.897321582260731e-06,
+      "loss": 0.5082,
+      "step": 863
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.132736382219138,
+      "learning_rate": 9.896970254356638e-06,
+      "loss": 0.5408,
+      "step": 864
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8873061759208194,
+      "learning_rate": 9.896618332678021e-06,
+      "loss": 0.5595,
+      "step": 865
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.8093405610248405,
+      "learning_rate": 9.89626581726755e-06,
+      "loss": 0.527,
+      "step": 866
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7860845360422704,
+      "learning_rate": 9.895912708167968e-06,
+      "loss": 0.5171,
+      "step": 867
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4131162424696315,
+      "learning_rate": 9.89555900542209e-06,
+      "loss": 0.513,
+      "step": 868
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5021973626726965,
+      "learning_rate": 9.895204709072807e-06,
+      "loss": 0.5079,
+      "step": 869
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2881648655935938,
+      "learning_rate": 9.894849819163076e-06,
+      "loss": 0.5461,
+      "step": 870
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.9195284357198386,
+      "learning_rate": 9.894494335735929e-06,
+      "loss": 0.5454,
+      "step": 871
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.644375692212634,
+      "learning_rate": 9.894138258834471e-06,
+      "loss": 0.5202,
+      "step": 872
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.337629721572819,
+      "learning_rate": 9.893781588501876e-06,
+      "loss": 0.572,
+      "step": 873
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3827858827727875,
+      "learning_rate": 9.893424324781392e-06,
+      "loss": 0.5013,
+      "step": 874
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.04857539538286,
+      "learning_rate": 9.893066467716337e-06,
+      "loss": 0.5558,
+      "step": 875
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.553770359230841,
+      "learning_rate": 9.892708017350106e-06,
+      "loss": 0.531,
+      "step": 876
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6333524869476066,
+      "learning_rate": 9.892348973726159e-06,
+      "loss": 0.5521,
+      "step": 877
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.807593432522117,
+      "learning_rate": 9.891989336888033e-06,
+      "loss": 0.5399,
+      "step": 878
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.51834156945878,
+      "learning_rate": 9.891629106879333e-06,
+      "loss": 0.4792,
+      "step": 879
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.420076170612479,
+      "learning_rate": 9.891268283743742e-06,
+      "loss": 0.5407,
+      "step": 880
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4507856303913944,
+      "learning_rate": 9.890906867525007e-06,
+      "loss": 0.5258,
+      "step": 881
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.698143933792585,
+      "learning_rate": 9.890544858266955e-06,
+      "loss": 0.5508,
+      "step": 882
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5106849735761996,
+      "learning_rate": 9.890182256013477e-06,
+      "loss": 0.473,
+      "step": 883
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2210339402369903,
+      "learning_rate": 9.889819060808541e-06,
+      "loss": 0.5083,
+      "step": 884
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.46994898819142067,
+      "learning_rate": 9.889455272696187e-06,
+      "loss": 0.2953,
+      "step": 885
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0440872143887847,
+      "learning_rate": 9.889090891720526e-06,
+      "loss": 0.4852,
+      "step": 886
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.040542418172153,
+      "learning_rate": 9.888725917925736e-06,
+      "loss": 0.4766,
+      "step": 887
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.470193451892504,
+      "learning_rate": 9.888360351356077e-06,
+      "loss": 0.5132,
+      "step": 888
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.211305724874561,
+      "learning_rate": 9.887994192055872e-06,
+      "loss": 0.5795,
+      "step": 889
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0074798209204703,
+      "learning_rate": 9.887627440069519e-06,
+      "loss": 0.5672,
+      "step": 890
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3329686235958444,
+      "learning_rate": 9.887260095441489e-06,
+      "loss": 0.5629,
+      "step": 891
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.496800949060134,
+      "learning_rate": 9.886892158216324e-06,
+      "loss": 0.5498,
+      "step": 892
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9971129947474697,
+      "learning_rate": 9.886523628438635e-06,
+      "loss": 0.511,
+      "step": 893
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.98948394380757,
+      "learning_rate": 9.886154506153112e-06,
+      "loss": 0.4743,
+      "step": 894
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.411773371719376,
+      "learning_rate": 9.88578479140451e-06,
+      "loss": 0.5073,
+      "step": 895
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.437262953534369,
+      "learning_rate": 9.885414484237657e-06,
+      "loss": 0.5405,
+      "step": 896
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.35029557531425,
+      "learning_rate": 9.885043584697457e-06,
+      "loss": 0.5365,
+      "step": 897
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.470775062244829,
+      "learning_rate": 9.88467209282888e-06,
+      "loss": 0.5658,
+      "step": 898
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.590639281776545,
+      "learning_rate": 9.88430000867697e-06,
+      "loss": 0.5347,
+      "step": 899
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.452295556322076,
+      "learning_rate": 9.883927332286849e-06,
+      "loss": 0.3093,
+      "step": 900
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.275573948966051,
+      "learning_rate": 9.883554063703697e-06,
+      "loss": 0.4997,
+      "step": 901
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3449909908211946,
+      "learning_rate": 9.883180202972781e-06,
+      "loss": 0.5665,
+      "step": 902
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5499824503056914,
+      "learning_rate": 9.882805750139432e-06,
+      "loss": 0.5075,
+      "step": 903
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.220681534195481,
+      "learning_rate": 9.88243070524905e-06,
+      "loss": 0.5523,
+      "step": 904
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.91562871005701,
+      "learning_rate": 9.882055068347114e-06,
+      "loss": 0.5672,
+      "step": 905
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2240786574420097,
+      "learning_rate": 9.881678839479169e-06,
+      "loss": 0.5279,
+      "step": 906
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.777037577011477,
+      "learning_rate": 9.881302018690834e-06,
+      "loss": 0.5466,
+      "step": 907
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.357819341136209,
+      "learning_rate": 9.880924606027802e-06,
+      "loss": 0.556,
+      "step": 908
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.989131371115194,
+      "learning_rate": 9.880546601535835e-06,
+      "loss": 0.4746,
+      "step": 909
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4815449682029076,
+      "learning_rate": 9.880168005260766e-06,
+      "loss": 0.5818,
+      "step": 910
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7157741073191084,
+      "learning_rate": 9.879788817248503e-06,
+      "loss": 0.5368,
+      "step": 911
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.169566366811378,
+      "learning_rate": 9.879409037545022e-06,
+      "loss": 0.5135,
+      "step": 912
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9155166437601543,
+      "learning_rate": 9.879028666196373e-06,
+      "loss": 0.5542,
+      "step": 913
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.073623866276996,
+      "learning_rate": 9.878647703248678e-06,
+      "loss": 0.4963,
+      "step": 914
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.467396875577332,
+      "learning_rate": 9.878266148748129e-06,
+      "loss": 0.5345,
+      "step": 915
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4812603625565745,
+      "learning_rate": 9.877884002740992e-06,
+      "loss": 0.3178,
+      "step": 916
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1579952180153366,
+      "learning_rate": 9.877501265273603e-06,
+      "loss": 0.5423,
+      "step": 917
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.047287741262805,
+      "learning_rate": 9.87711793639237e-06,
+      "loss": 0.5473,
+      "step": 918
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.682587655119555,
+      "learning_rate": 9.876734016143775e-06,
+      "loss": 0.5124,
+      "step": 919
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.096103554528039,
+      "learning_rate": 9.876349504574365e-06,
+      "loss": 0.5269,
+      "step": 920
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6595894888683693,
+      "learning_rate": 9.875964401730769e-06,
+      "loss": 0.4875,
+      "step": 921
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.699082939954343,
+      "learning_rate": 9.875578707659678e-06,
+      "loss": 0.5211,
+      "step": 922
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.10770951798126,
+      "learning_rate": 9.87519242240786e-06,
+      "loss": 0.5243,
+      "step": 923
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.133061210176109,
+      "learning_rate": 9.874805546022152e-06,
+      "loss": 0.6268,
+      "step": 924
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4420348646721921,
+      "learning_rate": 9.874418078549467e-06,
+      "loss": 0.3153,
+      "step": 925
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8302234065684226,
+      "learning_rate": 9.874030020036787e-06,
+      "loss": 0.5191,
+      "step": 926
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.456933171427221,
+      "learning_rate": 9.873641370531161e-06,
+      "loss": 0.5443,
+      "step": 927
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3900854178282884,
+      "learning_rate": 9.873252130079718e-06,
+      "loss": 0.5733,
+      "step": 928
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5851876713064765,
+      "learning_rate": 9.872862298729653e-06,
+      "loss": 0.5168,
+      "step": 929
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3922281052325148,
+      "learning_rate": 9.872471876528235e-06,
+      "loss": 0.3031,
+      "step": 930
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.470279824906071,
+      "learning_rate": 9.872080863522805e-06,
+      "loss": 0.5285,
+      "step": 931
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7489836358166415,
+      "learning_rate": 9.871689259760772e-06,
+      "loss": 0.4778,
+      "step": 932
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.889340965806428,
+      "learning_rate": 9.871297065289623e-06,
+      "loss": 0.5121,
+      "step": 933
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.38723907076015335,
+      "learning_rate": 9.87090428015691e-06,
+      "loss": 0.2999,
+      "step": 934
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.9386513087261585,
+      "learning_rate": 9.870510904410262e-06,
+      "loss": 0.514,
+      "step": 935
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.085597342575921,
+      "learning_rate": 9.870116938097374e-06,
+      "loss": 0.5597,
+      "step": 936
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.037435169227898,
+      "learning_rate": 9.869722381266016e-06,
+      "loss": 0.4977,
+      "step": 937
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1170369073222832,
+      "learning_rate": 9.869327233964032e-06,
+      "loss": 0.5407,
+      "step": 938
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.131975494024026,
+      "learning_rate": 9.868931496239335e-06,
+      "loss": 0.4649,
+      "step": 939
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.196973178951487,
+      "learning_rate": 9.868535168139907e-06,
+      "loss": 0.5586,
+      "step": 940
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4195001424133085,
+      "learning_rate": 9.868138249713805e-06,
+      "loss": 0.5777,
+      "step": 941
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.489515888988623,
+      "learning_rate": 9.86774074100916e-06,
+      "loss": 0.5129,
+      "step": 942
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.792785083107621,
+      "learning_rate": 9.867342642074165e-06,
+      "loss": 0.5234,
+      "step": 943
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.682188108587573,
+      "learning_rate": 9.866943952957097e-06,
+      "loss": 0.5538,
+      "step": 944
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.053253062754857,
+      "learning_rate": 9.866544673706294e-06,
+      "loss": 0.5122,
+      "step": 945
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0367664646683576,
+      "learning_rate": 9.866144804370172e-06,
+      "loss": 0.524,
+      "step": 946
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.097253289886195,
+      "learning_rate": 9.865744344997217e-06,
+      "loss": 0.52,
+      "step": 947
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.902831464696978,
+      "learning_rate": 9.865343295635986e-06,
+      "loss": 0.4781,
+      "step": 948
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.0649952214434855,
+      "learning_rate": 9.864941656335105e-06,
+      "loss": 0.4911,
+      "step": 949
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.579266790479727,
+      "learning_rate": 9.864539427143277e-06,
+      "loss": 0.4872,
+      "step": 950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 22.394425305610277,
+      "learning_rate": 9.864136608109274e-06,
+      "loss": 0.5211,
+      "step": 951
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.094745845694956,
+      "learning_rate": 9.863733199281938e-06,
+      "loss": 0.5518,
+      "step": 952
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.483180134776358,
+      "learning_rate": 9.863329200710183e-06,
+      "loss": 0.5162,
+      "step": 953
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.713704199112946,
+      "learning_rate": 9.862924612442996e-06,
+      "loss": 0.5389,
+      "step": 954
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6345564141007567,
+      "learning_rate": 9.862519434529435e-06,
+      "loss": 0.52,
+      "step": 955
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3107504193369355,
+      "learning_rate": 9.862113667018628e-06,
+      "loss": 0.5428,
+      "step": 956
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.621799208932092,
+      "learning_rate": 9.861707309959777e-06,
+      "loss": 0.5476,
+      "step": 957
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1550490909571893,
+      "learning_rate": 9.861300363402154e-06,
+      "loss": 0.5394,
+      "step": 958
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7246270993472135,
+      "learning_rate": 9.860892827395104e-06,
+      "loss": 0.5155,
+      "step": 959
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.876761494828879,
+      "learning_rate": 9.860484701988038e-06,
+      "loss": 0.5524,
+      "step": 960
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6676180047475655,
+      "learning_rate": 9.860075987230446e-06,
+      "loss": 0.571,
+      "step": 961
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.592109378419558,
+      "learning_rate": 9.859666683171886e-06,
+      "loss": 0.4829,
+      "step": 962
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3267450986248677,
+      "learning_rate": 9.859256789861986e-06,
+      "loss": 0.5232,
+      "step": 963
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.362942773379264,
+      "learning_rate": 9.858846307350449e-06,
+      "loss": 0.5416,
+      "step": 964
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3539854567315444,
+      "learning_rate": 9.858435235687046e-06,
+      "loss": 0.5315,
+      "step": 965
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3803528097816002,
+      "learning_rate": 9.85802357492162e-06,
+      "loss": 0.5622,
+      "step": 966
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.687397267771512,
+      "learning_rate": 9.857611325104088e-06,
+      "loss": 0.5682,
+      "step": 967
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.384627600413921,
+      "learning_rate": 9.857198486284436e-06,
+      "loss": 0.5579,
+      "step": 968
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5386175415801658,
+      "learning_rate": 9.856785058512722e-06,
+      "loss": 0.3022,
+      "step": 969
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.127388414514686,
+      "learning_rate": 9.856371041839077e-06,
+      "loss": 0.4965,
+      "step": 970
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3029111594339846,
+      "learning_rate": 9.8559564363137e-06,
+      "loss": 0.5976,
+      "step": 971
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6474548693793816,
+      "learning_rate": 9.855541241986864e-06,
+      "loss": 0.5351,
+      "step": 972
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4226648587401196,
+      "learning_rate": 9.855125458908912e-06,
+      "loss": 0.4798,
+      "step": 973
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.134068264126652,
+      "learning_rate": 9.854709087130261e-06,
+      "loss": 0.5933,
+      "step": 974
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7781164998319245,
+      "learning_rate": 9.854292126701397e-06,
+      "loss": 0.5436,
+      "step": 975
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.738102295178653,
+      "learning_rate": 9.853874577672875e-06,
+      "loss": 0.5004,
+      "step": 976
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5986372104594588,
+      "learning_rate": 9.853456440095329e-06,
+      "loss": 0.3253,
+      "step": 977
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.383320893508117,
+      "learning_rate": 9.853037714019455e-06,
+      "loss": 0.4966,
+      "step": 978
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1609032225285,
+      "learning_rate": 9.85261839949603e-06,
+      "loss": 0.566,
+      "step": 979
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.643199873717227,
+      "learning_rate": 9.852198496575893e-06,
+      "loss": 0.5515,
+      "step": 980
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6339986464175422,
+      "learning_rate": 9.85177800530996e-06,
+      "loss": 0.5396,
+      "step": 981
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5385221890299845,
+      "learning_rate": 9.851356925749218e-06,
+      "loss": 0.5914,
+      "step": 982
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9619885589894168,
+      "learning_rate": 9.850935257944722e-06,
+      "loss": 0.4819,
+      "step": 983
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.610250370020852,
+      "learning_rate": 9.850513001947604e-06,
+      "loss": 0.5128,
+      "step": 984
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.479850595968732,
+      "learning_rate": 9.850090157809062e-06,
+      "loss": 0.5792,
+      "step": 985
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.307210787237074,
+      "learning_rate": 9.849666725580366e-06,
+      "loss": 0.5338,
+      "step": 986
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.692007248298248,
+      "learning_rate": 9.849242705312863e-06,
+      "loss": 0.5472,
+      "step": 987
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1783900542212784,
+      "learning_rate": 9.848818097057962e-06,
+      "loss": 0.5178,
+      "step": 988
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5378300495775874,
+      "learning_rate": 9.848392900867151e-06,
+      "loss": 0.5695,
+      "step": 989
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.709140106168337,
+      "learning_rate": 9.847967116791985e-06,
+      "loss": 0.5376,
+      "step": 990
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8326740988612777,
+      "learning_rate": 9.847540744884094e-06,
+      "loss": 0.5663,
+      "step": 991
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4608709478578374,
+      "learning_rate": 9.847113785195175e-06,
+      "loss": 0.5088,
+      "step": 992
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.256283690631253,
+      "learning_rate": 9.846686237776999e-06,
+      "loss": 0.59,
+      "step": 993
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.629150374892115,
+      "learning_rate": 9.846258102681407e-06,
+      "loss": 0.4983,
+      "step": 994
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4748399486061157,
+      "learning_rate": 9.845829379960312e-06,
+      "loss": 0.4853,
+      "step": 995
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9723005119747765,
+      "learning_rate": 9.845400069665701e-06,
+      "loss": 0.5441,
+      "step": 996
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.47311709216153164,
+      "learning_rate": 9.844970171849625e-06,
+      "loss": 0.3158,
+      "step": 997
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.776730865185363,
+      "learning_rate": 9.844539686564213e-06,
+      "loss": 0.5513,
+      "step": 998
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.541610955320572,
+      "learning_rate": 9.844108613861663e-06,
+      "loss": 0.5462,
+      "step": 999
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.522113600892648,
+      "learning_rate": 9.843676953794242e-06,
+      "loss": 0.5166,
+      "step": 1000
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.898840111096932,
+      "learning_rate": 9.843244706414292e-06,
+      "loss": 0.5123,
+      "step": 1001
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.757109980248497,
+      "learning_rate": 9.842811871774224e-06,
+      "loss": 0.5097,
+      "step": 1002
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.027785560269563,
+      "learning_rate": 9.842378449926521e-06,
+      "loss": 0.492,
+      "step": 1003
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1852303118108733,
+      "learning_rate": 9.841944440923738e-06,
+      "loss": 0.5495,
+      "step": 1004
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3233983741471107,
+      "learning_rate": 9.841509844818498e-06,
+      "loss": 0.521,
+      "step": 1005
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.444400561030541,
+      "learning_rate": 9.841074661663497e-06,
+      "loss": 0.5137,
+      "step": 1006
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4857317092897961,
+      "learning_rate": 9.840638891511505e-06,
+      "loss": 0.2959,
+      "step": 1007
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.780127730997902,
+      "learning_rate": 9.840202534415358e-06,
+      "loss": 0.5261,
+      "step": 1008
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8324009405658015,
+      "learning_rate": 9.839765590427968e-06,
+      "loss": 0.5392,
+      "step": 1009
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.710342996439533,
+      "learning_rate": 9.839328059602316e-06,
+      "loss": 0.5031,
+      "step": 1010
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1090913871974855,
+      "learning_rate": 9.838889941991451e-06,
+      "loss": 0.5086,
+      "step": 1011
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8947420553816285,
+      "learning_rate": 9.838451237648499e-06,
+      "loss": 0.5304,
+      "step": 1012
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0109828297197097,
+      "learning_rate": 9.838011946626654e-06,
+      "loss": 0.5759,
+      "step": 1013
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7749383353953556,
+      "learning_rate": 9.83757206897918e-06,
+      "loss": 0.5768,
+      "step": 1014
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0625238902887104,
+      "learning_rate": 9.837131604759416e-06,
+      "loss": 0.5435,
+      "step": 1015
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.373551428074886,
+      "learning_rate": 9.836690554020769e-06,
+      "loss": 0.5333,
+      "step": 1016
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.635731700789115,
+      "learning_rate": 9.836248916816718e-06,
+      "loss": 0.5037,
+      "step": 1017
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.297467023695668,
+      "learning_rate": 9.835806693200812e-06,
+      "loss": 0.5074,
+      "step": 1018
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0076228527735056,
+      "learning_rate": 9.835363883226675e-06,
+      "loss": 0.5231,
+      "step": 1019
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.818489263242699,
+      "learning_rate": 9.834920486947994e-06,
+      "loss": 0.5193,
+      "step": 1020
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8117567235212104,
+      "learning_rate": 9.834476504418537e-06,
+      "loss": 0.5527,
+      "step": 1021
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7221488678585946,
+      "learning_rate": 9.834031935692136e-06,
+      "loss": 0.5126,
+      "step": 1022
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.5484454235889435,
+      "learning_rate": 9.833586780822698e-06,
+      "loss": 0.5294,
+      "step": 1023
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2971151789959694,
+      "learning_rate": 9.8331410398642e-06,
+      "loss": 0.5756,
+      "step": 1024
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.757954075691712,
+      "learning_rate": 9.832694712870687e-06,
+      "loss": 0.4987,
+      "step": 1025
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.670883764958368,
+      "learning_rate": 9.83224779989628e-06,
+      "loss": 0.5513,
+      "step": 1026
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.360125218834206,
+      "learning_rate": 9.831800300995167e-06,
+      "loss": 0.5763,
+      "step": 1027
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.709522939540928,
+      "learning_rate": 9.83135221622161e-06,
+      "loss": 0.5647,
+      "step": 1028
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3827429978689127,
+      "learning_rate": 9.830903545629942e-06,
+      "loss": 0.5067,
+      "step": 1029
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7208111521066765,
+      "learning_rate": 9.830454289274563e-06,
+      "loss": 0.4717,
+      "step": 1030
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.544984513267673,
+      "learning_rate": 9.830004447209949e-06,
+      "loss": 0.5373,
+      "step": 1031
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4619575497702706,
+      "learning_rate": 9.829554019490644e-06,
+      "loss": 0.535,
+      "step": 1032
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.351010988159188,
+      "learning_rate": 9.829103006171264e-06,
+      "loss": 0.5409,
+      "step": 1033
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3959748344310627,
+      "learning_rate": 9.828651407306495e-06,
+      "loss": 0.5443,
+      "step": 1034
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.691208500664219,
+      "learning_rate": 9.828199222951099e-06,
+      "loss": 0.5456,
+      "step": 1035
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5206389456933693,
+      "learning_rate": 9.8277464531599e-06,
+      "loss": 0.5198,
+      "step": 1036
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4648007394145014,
+      "learning_rate": 9.8272930979878e-06,
+      "loss": 0.3211,
+      "step": 1037
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3478252020221633,
+      "learning_rate": 9.826839157489768e-06,
+      "loss": 0.5203,
+      "step": 1038
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.15538349647015,
+      "learning_rate": 9.826384631720849e-06,
+      "loss": 0.5427,
+      "step": 1039
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4624823448813182,
+      "learning_rate": 9.825929520736155e-06,
+      "loss": 0.5575,
+      "step": 1040
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 16.402392337520524,
+      "learning_rate": 9.825473824590867e-06,
+      "loss": 0.5402,
+      "step": 1041
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.73197213046868,
+      "learning_rate": 9.825017543340244e-06,
+      "loss": 0.5363,
+      "step": 1042
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9600496089783097,
+      "learning_rate": 9.82456067703961e-06,
+      "loss": 0.5389,
+      "step": 1043
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1924330789034716,
+      "learning_rate": 9.82410322574436e-06,
+      "loss": 0.544,
+      "step": 1044
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.610568594669937,
+      "learning_rate": 9.823645189509961e-06,
+      "loss": 0.4995,
+      "step": 1045
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5947362987752984,
+      "learning_rate": 9.823186568391957e-06,
+      "loss": 0.5128,
+      "step": 1046
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.250522802309498,
+      "learning_rate": 9.82272736244595e-06,
+      "loss": 0.5429,
+      "step": 1047
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.348146820673165,
+      "learning_rate": 9.822267571727624e-06,
+      "loss": 0.5119,
+      "step": 1048
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.90195827102896,
+      "learning_rate": 9.821807196292732e-06,
+      "loss": 0.5359,
+      "step": 1049
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4713555188124159,
+      "learning_rate": 9.821346236197092e-06,
+      "loss": 0.3059,
+      "step": 1050
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.941704825598857,
+      "learning_rate": 9.820884691496601e-06,
+      "loss": 0.5437,
+      "step": 1051
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2727227382777104,
+      "learning_rate": 9.82042256224722e-06,
+      "loss": 0.5083,
+      "step": 1052
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.072769946539609,
+      "learning_rate": 9.819959848504986e-06,
+      "loss": 0.4918,
+      "step": 1053
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.746788049837481,
+      "learning_rate": 9.819496550326002e-06,
+      "loss": 0.5142,
+      "step": 1054
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.626982443715591,
+      "learning_rate": 9.819032667766447e-06,
+      "loss": 0.4895,
+      "step": 1055
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 19.94000263523307,
+      "learning_rate": 9.818568200882566e-06,
+      "loss": 0.5458,
+      "step": 1056
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.659694667599065,
+      "learning_rate": 9.81810314973068e-06,
+      "loss": 0.5483,
+      "step": 1057
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4364529928160736,
+      "learning_rate": 9.817637514367175e-06,
+      "loss": 0.299,
+      "step": 1058
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.44428619736583264,
+      "learning_rate": 9.817171294848514e-06,
+      "loss": 0.3106,
+      "step": 1059
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.071751955585563,
+      "learning_rate": 9.816704491231226e-06,
+      "loss": 0.5524,
+      "step": 1060
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4510107148427296,
+      "learning_rate": 9.816237103571913e-06,
+      "loss": 0.3191,
+      "step": 1061
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4033904666040553,
+      "learning_rate": 9.815769131927247e-06,
+      "loss": 0.528,
+      "step": 1062
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1879780449839337,
+      "learning_rate": 9.81530057635397e-06,
+      "loss": 0.4827,
+      "step": 1063
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.983920709038129,
+      "learning_rate": 9.814831436908897e-06,
+      "loss": 0.5819,
+      "step": 1064
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.799200077383184,
+      "learning_rate": 9.814361713648914e-06,
+      "loss": 0.5238,
+      "step": 1065
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4171847033405682,
+      "learning_rate": 9.813891406630975e-06,
+      "loss": 0.5472,
+      "step": 1066
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.10604518694355,
+      "learning_rate": 9.81342051591211e-06,
+      "loss": 0.5132,
+      "step": 1067
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.509759246653475,
+      "learning_rate": 9.812949041549408e-06,
+      "loss": 0.5422,
+      "step": 1068
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.472950753188651,
+      "learning_rate": 9.812476983600046e-06,
+      "loss": 0.5399,
+      "step": 1069
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.426940041816619,
+      "learning_rate": 9.812004342121257e-06,
+      "loss": 0.5531,
+      "step": 1070
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.634523266637475,
+      "learning_rate": 9.811531117170353e-06,
+      "loss": 0.5689,
+      "step": 1071
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.967859496029477,
+      "learning_rate": 9.811057308804711e-06,
+      "loss": 0.5721,
+      "step": 1072
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4908128910616822,
+      "learning_rate": 9.810582917081786e-06,
+      "loss": 0.6015,
+      "step": 1073
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6464029187052844,
+      "learning_rate": 9.810107942059097e-06,
+      "loss": 0.5358,
+      "step": 1074
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.99135690285684,
+      "learning_rate": 9.809632383794236e-06,
+      "loss": 0.5709,
+      "step": 1075
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1012438420287323,
+      "learning_rate": 9.809156242344868e-06,
+      "loss": 0.555,
+      "step": 1076
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.451642716497322,
+      "learning_rate": 9.808679517768727e-06,
+      "loss": 0.5229,
+      "step": 1077
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1535227494983396,
+      "learning_rate": 9.808202210123617e-06,
+      "loss": 0.5208,
+      "step": 1078
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.217561832486134,
+      "learning_rate": 9.80772431946741e-06,
+      "loss": 0.5267,
+      "step": 1079
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1908553380198628,
+      "learning_rate": 9.807245845858056e-06,
+      "loss": 0.5158,
+      "step": 1080
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.81048071357809,
+      "learning_rate": 9.80676678935357e-06,
+      "loss": 0.551,
+      "step": 1081
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2824270683183,
+      "learning_rate": 9.80628715001204e-06,
+      "loss": 0.5126,
+      "step": 1082
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.343337293051234,
+      "learning_rate": 9.805806927891622e-06,
+      "loss": 0.5111,
+      "step": 1083
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.091495135695928,
+      "learning_rate": 9.805326123050547e-06,
+      "loss": 0.4916,
+      "step": 1084
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.984933487733677,
+      "learning_rate": 9.804844735547113e-06,
+      "loss": 0.5462,
+      "step": 1085
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7876209035689712,
+      "learning_rate": 9.804362765439688e-06,
+      "loss": 0.3333,
+      "step": 1086
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.010635202413938,
+      "learning_rate": 9.803880212786716e-06,
+      "loss": 0.5332,
+      "step": 1087
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9813490546994865,
+      "learning_rate": 9.803397077646706e-06,
+      "loss": 0.5243,
+      "step": 1088
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.320801690237514,
+      "learning_rate": 9.80291336007824e-06,
+      "loss": 0.5134,
+      "step": 1089
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.814510370414608,
+      "learning_rate": 9.802429060139973e-06,
+      "loss": 0.5183,
+      "step": 1090
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1136992117138163,
+      "learning_rate": 9.801944177890626e-06,
+      "loss": 0.5691,
+      "step": 1091
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.311081441194139,
+      "learning_rate": 9.801458713388991e-06,
+      "loss": 0.5347,
+      "step": 1092
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.677571602000118,
+      "learning_rate": 9.800972666693936e-06,
+      "loss": 0.4887,
+      "step": 1093
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.745368384264501,
+      "learning_rate": 9.800486037864393e-06,
+      "loss": 0.5266,
+      "step": 1094
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6781275041424397,
+      "learning_rate": 9.799998826959366e-06,
+      "loss": 0.3105,
+      "step": 1095
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.0786129646366875,
+      "learning_rate": 9.799511034037934e-06,
+      "loss": 0.5157,
+      "step": 1096
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.367945303419398,
+      "learning_rate": 9.799022659159243e-06,
+      "loss": 0.5028,
+      "step": 1097
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.262873170109456,
+      "learning_rate": 9.798533702382512e-06,
+      "loss": 0.5547,
+      "step": 1098
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.982842946448107,
+      "learning_rate": 9.798044163767022e-06,
+      "loss": 0.5384,
+      "step": 1099
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0045522787847028,
+      "learning_rate": 9.797554043372139e-06,
+      "loss": 0.5541,
+      "step": 1100
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.674684068049811,
+      "learning_rate": 9.79706334125729e-06,
+      "loss": 0.5327,
+      "step": 1101
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9754554900313948,
+      "learning_rate": 9.796572057481969e-06,
+      "loss": 0.5588,
+      "step": 1102
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9225079546262465,
+      "learning_rate": 9.79608019210575e-06,
+      "loss": 0.5523,
+      "step": 1103
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8442060883007754,
+      "learning_rate": 9.795587745188277e-06,
+      "loss": 0.5232,
+      "step": 1104
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0671802035289613,
+      "learning_rate": 9.795094716789252e-06,
+      "loss": 0.5493,
+      "step": 1105
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6485128533891027,
+      "learning_rate": 9.794601106968466e-06,
+      "loss": 0.5166,
+      "step": 1106
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.737096200702251,
+      "learning_rate": 9.794106915785764e-06,
+      "loss": 0.5166,
+      "step": 1107
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5937706206224425,
+      "learning_rate": 9.79361214330107e-06,
+      "loss": 0.5556,
+      "step": 1108
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4099782375542316,
+      "learning_rate": 9.793116789574379e-06,
+      "loss": 0.5551,
+      "step": 1109
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.885971086325263,
+      "learning_rate": 9.792620854665752e-06,
+      "loss": 0.4949,
+      "step": 1110
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.149869869316785,
+      "learning_rate": 9.792124338635325e-06,
+      "loss": 0.544,
+      "step": 1111
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.76900938914313,
+      "learning_rate": 9.7916272415433e-06,
+      "loss": 0.5972,
+      "step": 1112
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.977873582196192,
+      "learning_rate": 9.791129563449953e-06,
+      "loss": 0.5306,
+      "step": 1113
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.1240001955338705,
+      "learning_rate": 9.790631304415627e-06,
+      "loss": 0.5557,
+      "step": 1114
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.698903864111224,
+      "learning_rate": 9.790132464500741e-06,
+      "loss": 0.5565,
+      "step": 1115
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5367019677229306,
+      "learning_rate": 9.789633043765781e-06,
+      "loss": 0.3101,
+      "step": 1116
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4696541434924939,
+      "learning_rate": 9.7891330422713e-06,
+      "loss": 0.3061,
+      "step": 1117
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.402928253256138,
+      "learning_rate": 9.788632460077928e-06,
+      "loss": 0.5522,
+      "step": 1118
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.362997234320723,
+      "learning_rate": 9.788131297246362e-06,
+      "loss": 0.5449,
+      "step": 1119
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.35274701267931,
+      "learning_rate": 9.787629553837368e-06,
+      "loss": 0.5522,
+      "step": 1120
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2780700609567823,
+      "learning_rate": 9.787127229911783e-06,
+      "loss": 0.5141,
+      "step": 1121
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7434919771655815,
+      "learning_rate": 9.786624325530521e-06,
+      "loss": 0.5309,
+      "step": 1122
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.346778967407015,
+      "learning_rate": 9.786120840754557e-06,
+      "loss": 0.5217,
+      "step": 1123
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6304104576547926,
+      "learning_rate": 9.78561677564494e-06,
+      "loss": 0.3051,
+      "step": 1124
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 54.6306558185081,
+      "learning_rate": 9.78511213026279e-06,
+      "loss": 0.5376,
+      "step": 1125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6297603828086986,
+      "learning_rate": 9.784606904669298e-06,
+      "loss": 0.5642,
+      "step": 1126
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.16609710877351,
+      "learning_rate": 9.784101098925723e-06,
+      "loss": 0.5474,
+      "step": 1127
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3920070460720235,
+      "learning_rate": 9.783594713093399e-06,
+      "loss": 0.5182,
+      "step": 1128
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.42337134723578207,
+      "learning_rate": 9.783087747233722e-06,
+      "loss": 0.3179,
+      "step": 1129
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.47110559625524695,
+      "learning_rate": 9.782580201408165e-06,
+      "loss": 0.3001,
+      "step": 1130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.211383439848518,
+      "learning_rate": 9.782072075678273e-06,
+      "loss": 0.4972,
+      "step": 1131
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.560303521680692,
+      "learning_rate": 9.781563370105654e-06,
+      "loss": 0.5187,
+      "step": 1132
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.557439828624445,
+      "learning_rate": 9.781054084751993e-06,
+      "loss": 0.4635,
+      "step": 1133
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.991548120522923,
+      "learning_rate": 9.78054421967904e-06,
+      "loss": 0.4887,
+      "step": 1134
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.838359682219407,
+      "learning_rate": 9.78003377494862e-06,
+      "loss": 0.5086,
+      "step": 1135
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1755691331658604,
+      "learning_rate": 9.779522750622626e-06,
+      "loss": 0.5705,
+      "step": 1136
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.171465153310783,
+      "learning_rate": 9.77901114676302e-06,
+      "loss": 0.5376,
+      "step": 1137
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.414602647045598,
+      "learning_rate": 9.778498963431838e-06,
+      "loss": 0.4732,
+      "step": 1138
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0159953833282827,
+      "learning_rate": 9.77798620069118e-06,
+      "loss": 0.5139,
+      "step": 1139
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.021411823253752,
+      "learning_rate": 9.777472858603227e-06,
+      "loss": 0.5582,
+      "step": 1140
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8163225097375872,
+      "learning_rate": 9.776958937230216e-06,
+      "loss": 0.4982,
+      "step": 1141
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.593917842133048,
+      "learning_rate": 9.776444436634467e-06,
+      "loss": 0.4914,
+      "step": 1142
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.134996704478066,
+      "learning_rate": 9.775929356878362e-06,
+      "loss": 0.5661,
+      "step": 1143
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1824799742472103,
+      "learning_rate": 9.775413698024358e-06,
+      "loss": 0.5333,
+      "step": 1144
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8013971853995354,
+      "learning_rate": 9.77489746013498e-06,
+      "loss": 0.5607,
+      "step": 1145
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.600410366159765,
+      "learning_rate": 9.774380643272824e-06,
+      "loss": 0.3276,
+      "step": 1146
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.204175330198102,
+      "learning_rate": 9.773863247500554e-06,
+      "loss": 0.5438,
+      "step": 1147
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1422354538612867,
+      "learning_rate": 9.773345272880908e-06,
+      "loss": 0.5137,
+      "step": 1148
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.40516495810484,
+      "learning_rate": 9.77282671947669e-06,
+      "loss": 0.5338,
+      "step": 1149
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40735475361878887,
+      "learning_rate": 9.77230758735078e-06,
+      "loss": 0.3037,
+      "step": 1150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.299843951025853,
+      "learning_rate": 9.771787876566121e-06,
+      "loss": 0.4718,
+      "step": 1151
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.448347791199286,
+      "learning_rate": 9.771267587185732e-06,
+      "loss": 0.5106,
+      "step": 1152
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.75031452294923,
+      "learning_rate": 9.770746719272696e-06,
+      "loss": 0.502,
+      "step": 1153
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.496628712354576,
+      "learning_rate": 9.770225272890177e-06,
+      "loss": 0.5226,
+      "step": 1154
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7234650900694537,
+      "learning_rate": 9.769703248101397e-06,
+      "loss": 0.4835,
+      "step": 1155
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0902200021205397,
+      "learning_rate": 9.769180644969653e-06,
+      "loss": 0.5542,
+      "step": 1156
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4903105435427925,
+      "learning_rate": 9.768657463558316e-06,
+      "loss": 0.3176,
+      "step": 1157
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9439734384390666,
+      "learning_rate": 9.76813370393082e-06,
+      "loss": 0.5594,
+      "step": 1158
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.803100605859909,
+      "learning_rate": 9.767609366150675e-06,
+      "loss": 0.506,
+      "step": 1159
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0895827294870286,
+      "learning_rate": 9.767084450281457e-06,
+      "loss": 0.576,
+      "step": 1160
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5438175383178354,
+      "learning_rate": 9.766558956386814e-06,
+      "loss": 0.54,
+      "step": 1161
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.446390394401688,
+      "learning_rate": 9.766032884530466e-06,
+      "loss": 0.5304,
+      "step": 1162
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7220833997717024,
+      "learning_rate": 9.765506234776201e-06,
+      "loss": 0.5269,
+      "step": 1163
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.819035065066364,
+      "learning_rate": 9.764979007187874e-06,
+      "loss": 0.5523,
+      "step": 1164
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.577743647999412,
+      "learning_rate": 9.764451201829415e-06,
+      "loss": 0.5425,
+      "step": 1165
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.1359205953061,
+      "learning_rate": 9.763922818764824e-06,
+      "loss": 0.5279,
+      "step": 1166
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8292353931630454,
+      "learning_rate": 9.763393858058165e-06,
+      "loss": 0.4739,
+      "step": 1167
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.003032715248175,
+      "learning_rate": 9.76286431977358e-06,
+      "loss": 0.4909,
+      "step": 1168
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.48382871935652,
+      "learning_rate": 9.762334203975277e-06,
+      "loss": 0.2988,
+      "step": 1169
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.504660090934234,
+      "learning_rate": 9.761803510727532e-06,
+      "loss": 0.5487,
+      "step": 1170
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.363328369324873,
+      "learning_rate": 9.761272240094695e-06,
+      "loss": 0.5595,
+      "step": 1171
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9851791524216444,
+      "learning_rate": 9.760740392141186e-06,
+      "loss": 0.527,
+      "step": 1172
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4054173433667769,
+      "learning_rate": 9.76020796693149e-06,
+      "loss": 0.2919,
+      "step": 1173
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.546756548503936,
+      "learning_rate": 9.759674964530167e-06,
+      "loss": 0.5423,
+      "step": 1174
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1340775537109455,
+      "learning_rate": 9.759141385001846e-06,
+      "loss": 0.5446,
+      "step": 1175
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2592304791135693,
+      "learning_rate": 9.758607228411226e-06,
+      "loss": 0.5554,
+      "step": 1176
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.315525490182364,
+      "learning_rate": 9.758072494823072e-06,
+      "loss": 0.4845,
+      "step": 1177
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.658623877993458,
+      "learning_rate": 9.757537184302225e-06,
+      "loss": 0.4718,
+      "step": 1178
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.062320342248418,
+      "learning_rate": 9.757001296913594e-06,
+      "loss": 0.5212,
+      "step": 1179
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.547718077010988,
+      "learning_rate": 9.756464832722155e-06,
+      "loss": 0.5561,
+      "step": 1180
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9079278981850467,
+      "learning_rate": 9.755927791792956e-06,
+      "loss": 0.5395,
+      "step": 1181
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.075234271506065,
+      "learning_rate": 9.755390174191117e-06,
+      "loss": 0.5594,
+      "step": 1182
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1763315237406275,
+      "learning_rate": 9.754851979981827e-06,
+      "loss": 0.5399,
+      "step": 1183
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2340128110001705,
+      "learning_rate": 9.75431320923034e-06,
+      "loss": 0.5447,
+      "step": 1184
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.366966957134144,
+      "learning_rate": 9.753773862001985e-06,
+      "loss": 0.543,
+      "step": 1185
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.504435333808244,
+      "learning_rate": 9.753233938362161e-06,
+      "loss": 0.4984,
+      "step": 1186
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6389161733395476,
+      "learning_rate": 9.752693438376337e-06,
+      "loss": 0.5454,
+      "step": 1187
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4645224284612879,
+      "learning_rate": 9.752152362110046e-06,
+      "loss": 0.3068,
+      "step": 1188
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40488857497409236,
+      "learning_rate": 9.751610709628898e-06,
+      "loss": 0.3053,
+      "step": 1189
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.158108073826769,
+      "learning_rate": 9.751068480998572e-06,
+      "loss": 0.562,
+      "step": 1190
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4497722256794265,
+      "learning_rate": 9.750525676284813e-06,
+      "loss": 0.5278,
+      "step": 1191
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7629399349582098,
+      "learning_rate": 9.749982295553437e-06,
+      "loss": 0.5308,
+      "step": 1192
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3840285531051673,
+      "learning_rate": 9.749438338870332e-06,
+      "loss": 0.5595,
+      "step": 1193
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.792181651342754,
+      "learning_rate": 9.748893806301455e-06,
+      "loss": 0.5293,
+      "step": 1194
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1921515714564457,
+      "learning_rate": 9.748348697912833e-06,
+      "loss": 0.5406,
+      "step": 1195
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5177293434234235,
+      "learning_rate": 9.74780301377056e-06,
+      "loss": 0.308,
+      "step": 1196
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.954921811677318,
+      "learning_rate": 9.747256753940803e-06,
+      "loss": 0.5497,
+      "step": 1197
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.190773506971967,
+      "learning_rate": 9.746709918489801e-06,
+      "loss": 0.5066,
+      "step": 1198
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4454320974682857,
+      "learning_rate": 9.746162507483854e-06,
+      "loss": 0.5312,
+      "step": 1199
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.780020527173888,
+      "learning_rate": 9.745614520989341e-06,
+      "loss": 0.5414,
+      "step": 1200
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.186176692858363,
+      "learning_rate": 9.745065959072708e-06,
+      "loss": 0.5031,
+      "step": 1201
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5159363305115994,
+      "learning_rate": 9.74451682180047e-06,
+      "loss": 0.5112,
+      "step": 1202
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1076783504151946,
+      "learning_rate": 9.743967109239207e-06,
+      "loss": 0.5015,
+      "step": 1203
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4543352308559228,
+      "learning_rate": 9.743416821455577e-06,
+      "loss": 0.5492,
+      "step": 1204
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.371695969111709,
+      "learning_rate": 9.742865958516307e-06,
+      "loss": 0.529,
+      "step": 1205
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2131054668517645,
+      "learning_rate": 9.742314520488188e-06,
+      "loss": 0.4943,
+      "step": 1206
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.372971885417964,
+      "learning_rate": 9.741762507438084e-06,
+      "loss": 0.5391,
+      "step": 1207
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.174829014411358,
+      "learning_rate": 9.741209919432928e-06,
+      "loss": 0.5443,
+      "step": 1208
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.766240893442258,
+      "learning_rate": 9.740656756539724e-06,
+      "loss": 0.4915,
+      "step": 1209
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.304285795929697,
+      "learning_rate": 9.740103018825544e-06,
+      "loss": 0.5159,
+      "step": 1210
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.267829378526681,
+      "learning_rate": 9.739548706357533e-06,
+      "loss": 0.5196,
+      "step": 1211
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4575888499007,
+      "learning_rate": 9.7389938192029e-06,
+      "loss": 0.5257,
+      "step": 1212
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.90782177106157,
+      "learning_rate": 9.73843835742893e-06,
+      "loss": 0.5222,
+      "step": 1213
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.29800015816317,
+      "learning_rate": 9.737882321102973e-06,
+      "loss": 0.5799,
+      "step": 1214
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5635673302263884,
+      "learning_rate": 9.73732571029245e-06,
+      "loss": 0.5353,
+      "step": 1215
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7426307202061513,
+      "learning_rate": 9.736768525064852e-06,
+      "loss": 0.474,
+      "step": 1216
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1450490267348274,
+      "learning_rate": 9.736210765487741e-06,
+      "loss": 0.5347,
+      "step": 1217
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.902317422362506,
+      "learning_rate": 9.735652431628748e-06,
+      "loss": 0.5508,
+      "step": 1218
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2096193498029004,
+      "learning_rate": 9.73509352355557e-06,
+      "loss": 0.5237,
+      "step": 1219
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8695419912103626,
+      "learning_rate": 9.734534041335976e-06,
+      "loss": 0.5066,
+      "step": 1220
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.207058928314355,
+      "learning_rate": 9.73397398503781e-06,
+      "loss": 0.51,
+      "step": 1221
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5441272511824042,
+      "learning_rate": 9.733413354728978e-06,
+      "loss": 0.4964,
+      "step": 1222
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.52800905515106,
+      "learning_rate": 9.732852150477456e-06,
+      "loss": 0.565,
+      "step": 1223
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3017442978805955,
+      "learning_rate": 9.732290372351295e-06,
+      "loss": 0.4878,
+      "step": 1224
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9163572350082196,
+      "learning_rate": 9.731728020418612e-06,
+      "loss": 0.5008,
+      "step": 1225
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.090864552893475,
+      "learning_rate": 9.731165094747593e-06,
+      "loss": 0.441,
+      "step": 1226
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7143933996424305,
+      "learning_rate": 9.730601595406497e-06,
+      "loss": 0.5272,
+      "step": 1227
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.530185776875611,
+      "learning_rate": 9.730037522463647e-06,
+      "loss": 0.5594,
+      "step": 1228
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8546893205858828,
+      "learning_rate": 9.72947287598744e-06,
+      "loss": 0.5401,
+      "step": 1229
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.750601301669908,
+      "learning_rate": 9.728907656046344e-06,
+      "loss": 0.5302,
+      "step": 1230
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.198781768145728,
+      "learning_rate": 9.72834186270889e-06,
+      "loss": 0.5606,
+      "step": 1231
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4598331202559636,
+      "learning_rate": 9.727775496043685e-06,
+      "loss": 0.5766,
+      "step": 1232
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3745935892538554,
+      "learning_rate": 9.727208556119401e-06,
+      "loss": 0.5687,
+      "step": 1233
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.46870419113172956,
+      "learning_rate": 9.726641043004782e-06,
+      "loss": 0.3118,
+      "step": 1234
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.685438116439696,
+      "learning_rate": 9.726072956768644e-06,
+      "loss": 0.4962,
+      "step": 1235
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.635004244288688,
+      "learning_rate": 9.725504297479865e-06,
+      "loss": 0.5453,
+      "step": 1236
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.536954295143957,
+      "learning_rate": 9.7249350652074e-06,
+      "loss": 0.577,
+      "step": 1237
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9948613206338393,
+      "learning_rate": 9.724365260020267e-06,
+      "loss": 0.546,
+      "step": 1238
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8751869323624897,
+      "learning_rate": 9.723794881987561e-06,
+      "loss": 0.5504,
+      "step": 1239
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 19.411878536685695,
+      "learning_rate": 9.723223931178439e-06,
+      "loss": 0.5322,
+      "step": 1240
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8264894179833524,
+      "learning_rate": 9.722652407662129e-06,
+      "loss": 0.5288,
+      "step": 1241
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.834648980138129,
+      "learning_rate": 9.722080311507938e-06,
+      "loss": 0.5516,
+      "step": 1242
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 20.251482959292595,
+      "learning_rate": 9.721507642785225e-06,
+      "loss": 0.5094,
+      "step": 1243
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.252308856119477,
+      "learning_rate": 9.720934401563438e-06,
+      "loss": 0.5887,
+      "step": 1244
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.426936249719257,
+      "learning_rate": 9.720360587912075e-06,
+      "loss": 0.6084,
+      "step": 1245
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.608793170698113,
+      "learning_rate": 9.71978620190072e-06,
+      "loss": 0.5386,
+      "step": 1246
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5021803589085256,
+      "learning_rate": 9.719211243599015e-06,
+      "loss": 0.4879,
+      "step": 1247
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.481551759407057,
+      "learning_rate": 9.718635713076677e-06,
+      "loss": 0.5794,
+      "step": 1248
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9714897505064872,
+      "learning_rate": 9.718059610403493e-06,
+      "loss": 0.5434,
+      "step": 1249
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4975735308149125,
+      "learning_rate": 9.717482935649313e-06,
+      "loss": 0.5499,
+      "step": 1250
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.517323259996049,
+      "learning_rate": 9.716905688884065e-06,
+      "loss": 0.5027,
+      "step": 1251
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.428958562743564,
+      "learning_rate": 9.716327870177739e-06,
+      "loss": 0.5031,
+      "step": 1252
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.601849118808271,
+      "learning_rate": 9.7157494796004e-06,
+      "loss": 0.5068,
+      "step": 1253
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.714275986794909,
+      "learning_rate": 9.71517051722218e-06,
+      "loss": 0.5492,
+      "step": 1254
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.251969595582135,
+      "learning_rate": 9.714590983113279e-06,
+      "loss": 0.5232,
+      "step": 1255
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7577069522342095,
+      "learning_rate": 9.714010877343967e-06,
+      "loss": 0.5106,
+      "step": 1256
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.580827030052822,
+      "learning_rate": 9.713430199984583e-06,
+      "loss": 0.5531,
+      "step": 1257
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.443358331093899,
+      "learning_rate": 9.71284895110554e-06,
+      "loss": 0.5521,
+      "step": 1258
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.64536471854187,
+      "learning_rate": 9.712267130777313e-06,
+      "loss": 0.5116,
+      "step": 1259
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4202506622021893,
+      "learning_rate": 9.71168473907045e-06,
+      "loss": 0.5475,
+      "step": 1260
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.354677666274275,
+      "learning_rate": 9.71110177605557e-06,
+      "loss": 0.5459,
+      "step": 1261
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9558721961493286,
+      "learning_rate": 9.710518241803357e-06,
+      "loss": 0.5953,
+      "step": 1262
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4326134061610225,
+      "learning_rate": 9.70993413638457e-06,
+      "loss": 0.5674,
+      "step": 1263
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6210227877099177,
+      "learning_rate": 9.709349459870028e-06,
+      "loss": 0.5421,
+      "step": 1264
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1645833336107065,
+      "learning_rate": 9.70876421233063e-06,
+      "loss": 0.5382,
+      "step": 1265
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.382119787451024,
+      "learning_rate": 9.708178393837338e-06,
+      "loss": 0.5322,
+      "step": 1266
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.533870796048674,
+      "learning_rate": 9.707592004461183e-06,
+      "loss": 0.5213,
+      "step": 1267
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.034781170926504,
+      "learning_rate": 9.707005044273268e-06,
+      "loss": 0.5102,
+      "step": 1268
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8654931164100366,
+      "learning_rate": 9.706417513344764e-06,
+      "loss": 0.5503,
+      "step": 1269
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.619937483855922,
+      "learning_rate": 9.705829411746912e-06,
+      "loss": 0.5183,
+      "step": 1270
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.476933870987792,
+      "learning_rate": 9.70524073955102e-06,
+      "loss": 0.5434,
+      "step": 1271
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.416735160312356,
+      "learning_rate": 9.704651496828466e-06,
+      "loss": 0.57,
+      "step": 1272
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6905412452360653,
+      "learning_rate": 9.704061683650701e-06,
+      "loss": 0.4835,
+      "step": 1273
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7189440898354191,
+      "learning_rate": 9.703471300089237e-06,
+      "loss": 0.4869,
+      "step": 1274
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.46856363032841375,
+      "learning_rate": 9.702880346215664e-06,
+      "loss": 0.3023,
+      "step": 1275
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6403917733175946,
+      "learning_rate": 9.702288822101634e-06,
+      "loss": 0.5196,
+      "step": 1276
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1389767482269733,
+      "learning_rate": 9.701696727818875e-06,
+      "loss": 0.4938,
+      "step": 1277
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.021728685133597,
+      "learning_rate": 9.701104063439177e-06,
+      "loss": 0.5483,
+      "step": 1278
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2615719556604774,
+      "learning_rate": 9.700510829034404e-06,
+      "loss": 0.5166,
+      "step": 1279
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9255859692855353,
+      "learning_rate": 9.699917024676489e-06,
+      "loss": 0.4992,
+      "step": 1280
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.754539800261354,
+      "learning_rate": 9.699322650437433e-06,
+      "loss": 0.481,
+      "step": 1281
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.448657947932157,
+      "learning_rate": 9.698727706389301e-06,
+      "loss": 0.4954,
+      "step": 1282
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.399464001350794,
+      "learning_rate": 9.698132192604239e-06,
+      "loss": 0.4613,
+      "step": 1283
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8184265768743724,
+      "learning_rate": 9.69753610915445e-06,
+      "loss": 0.53,
+      "step": 1284
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4088333468198906,
+      "learning_rate": 9.696939456112213e-06,
+      "loss": 0.516,
+      "step": 1285
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1537240539958837,
+      "learning_rate": 9.696342233549874e-06,
+      "loss": 0.5174,
+      "step": 1286
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5408474179138087,
+      "learning_rate": 9.695744441539849e-06,
+      "loss": 0.5541,
+      "step": 1287
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5987042739580883,
+      "learning_rate": 9.695146080154622e-06,
+      "loss": 0.5535,
+      "step": 1288
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7244718900064275,
+      "learning_rate": 9.694547149466745e-06,
+      "loss": 0.4973,
+      "step": 1289
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2120171372522672,
+      "learning_rate": 9.693947649548843e-06,
+      "loss": 0.5624,
+      "step": 1290
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.539944282298737,
+      "learning_rate": 9.693347580473604e-06,
+      "loss": 0.5086,
+      "step": 1291
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.2316043477407925,
+      "learning_rate": 9.692746942313792e-06,
+      "loss": 0.5609,
+      "step": 1292
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.48097610167309607,
+      "learning_rate": 9.692145735142236e-06,
+      "loss": 0.3178,
+      "step": 1293
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6778354833939555,
+      "learning_rate": 9.691543959031831e-06,
+      "loss": 0.6211,
+      "step": 1294
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.603840945010066,
+      "learning_rate": 9.690941614055547e-06,
+      "loss": 0.591,
+      "step": 1295
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0371943451008536,
+      "learning_rate": 9.690338700286422e-06,
+      "loss": 0.5555,
+      "step": 1296
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6105865922139504,
+      "learning_rate": 9.689735217797559e-06,
+      "loss": 0.4833,
+      "step": 1297
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3715070293616285,
+      "learning_rate": 9.689131166662131e-06,
+      "loss": 0.5166,
+      "step": 1298
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.000255726912597,
+      "learning_rate": 9.688526546953384e-06,
+      "loss": 0.5074,
+      "step": 1299
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4289704748593506,
+      "learning_rate": 9.68792135874463e-06,
+      "loss": 0.5303,
+      "step": 1300
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.902906124906545,
+      "learning_rate": 9.687315602109248e-06,
+      "loss": 0.4967,
+      "step": 1301
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2177812989509453,
+      "learning_rate": 9.686709277120692e-06,
+      "loss": 0.5448,
+      "step": 1302
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.917358423660369,
+      "learning_rate": 9.686102383852478e-06,
+      "loss": 0.5186,
+      "step": 1303
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.202012217847628,
+      "learning_rate": 9.685494922378194e-06,
+      "loss": 0.5729,
+      "step": 1304
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2423603257251754,
+      "learning_rate": 9.684886892771496e-06,
+      "loss": 0.5406,
+      "step": 1305
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.296831687686048,
+      "learning_rate": 9.684278295106112e-06,
+      "loss": 0.5171,
+      "step": 1306
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9317326832201136,
+      "learning_rate": 9.683669129455837e-06,
+      "loss": 0.5667,
+      "step": 1307
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9137020935913265,
+      "learning_rate": 9.683059395894531e-06,
+      "loss": 0.5241,
+      "step": 1308
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1952685167015424,
+      "learning_rate": 9.682449094496129e-06,
+      "loss": 0.5407,
+      "step": 1309
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4203096329244382,
+      "learning_rate": 9.681838225334632e-06,
+      "loss": 0.5127,
+      "step": 1310
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.163903546038646,
+      "learning_rate": 9.681226788484108e-06,
+      "loss": 0.4963,
+      "step": 1311
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0648370124211604,
+      "learning_rate": 9.680614784018699e-06,
+      "loss": 0.5233,
+      "step": 1312
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.948912621629233,
+      "learning_rate": 9.68000221201261e-06,
+      "loss": 0.539,
+      "step": 1313
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6639393773191538,
+      "learning_rate": 9.67938907254012e-06,
+      "loss": 0.5677,
+      "step": 1314
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4922915561741643,
+      "learning_rate": 9.67877536567557e-06,
+      "loss": 0.3369,
+      "step": 1315
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.712490398485277,
+      "learning_rate": 9.678161091493378e-06,
+      "loss": 0.5625,
+      "step": 1316
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.532223964991713,
+      "learning_rate": 9.677546250068025e-06,
+      "loss": 0.542,
+      "step": 1317
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.392507799256179,
+      "learning_rate": 9.676930841474064e-06,
+      "loss": 0.5839,
+      "step": 1318
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4288104475595016,
+      "learning_rate": 9.676314865786114e-06,
+      "loss": 0.302,
+      "step": 1319
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.196829770687926,
+      "learning_rate": 9.675698323078865e-06,
+      "loss": 0.5445,
+      "step": 1320
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.555233583219232,
+      "learning_rate": 9.675081213427076e-06,
+      "loss": 0.6118,
+      "step": 1321
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.978189497496224,
+      "learning_rate": 9.67446353690557e-06,
+      "loss": 0.5391,
+      "step": 1322
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9180821898149043,
+      "learning_rate": 9.673845293589246e-06,
+      "loss": 0.545,
+      "step": 1323
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9378346799881117,
+      "learning_rate": 9.673226483553068e-06,
+      "loss": 0.4924,
+      "step": 1324
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.560254289416233,
+      "learning_rate": 9.672607106872067e-06,
+      "loss": 0.5572,
+      "step": 1325
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7241531331523414,
+      "learning_rate": 9.671987163621344e-06,
+      "loss": 0.557,
+      "step": 1326
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4656180216088064,
+      "learning_rate": 9.671366653876071e-06,
+      "loss": 0.5948,
+      "step": 1327
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6117354150981873,
+      "learning_rate": 9.670745577711488e-06,
+      "loss": 0.5293,
+      "step": 1328
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8152871446814527,
+      "learning_rate": 9.670123935202903e-06,
+      "loss": 0.5995,
+      "step": 1329
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.942917657331075,
+      "learning_rate": 9.669501726425687e-06,
+      "loss": 0.5078,
+      "step": 1330
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7742970065377044,
+      "learning_rate": 9.668878951455289e-06,
+      "loss": 0.5346,
+      "step": 1331
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5038551315297554,
+      "learning_rate": 9.668255610367223e-06,
+      "loss": 0.3036,
+      "step": 1332
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4428846002167544,
+      "learning_rate": 9.66763170323707e-06,
+      "loss": 0.3075,
+      "step": 1333
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.098453961363088,
+      "learning_rate": 9.66700723014048e-06,
+      "loss": 0.4851,
+      "step": 1334
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5948073111685166,
+      "learning_rate": 9.666382191153175e-06,
+      "loss": 0.4886,
+      "step": 1335
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3800806567186763,
+      "learning_rate": 9.665756586350942e-06,
+      "loss": 0.5018,
+      "step": 1336
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.778711760807915,
+      "learning_rate": 9.665130415809636e-06,
+      "loss": 0.5543,
+      "step": 1337
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.200372477354754,
+      "learning_rate": 9.664503679605186e-06,
+      "loss": 0.5458,
+      "step": 1338
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6030868219882242,
+      "learning_rate": 9.663876377813583e-06,
+      "loss": 0.3376,
+      "step": 1339
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.646119763476875,
+      "learning_rate": 9.663248510510891e-06,
+      "loss": 0.52,
+      "step": 1340
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.567376057386586,
+      "learning_rate": 9.66262007777324e-06,
+      "loss": 0.5263,
+      "step": 1341
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5444386521299807,
+      "learning_rate": 9.661991079676829e-06,
+      "loss": 0.5492,
+      "step": 1342
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9913921908992673,
+      "learning_rate": 9.661361516297929e-06,
+      "loss": 0.5556,
+      "step": 1343
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3057234185273052,
+      "learning_rate": 9.660731387712874e-06,
+      "loss": 0.5237,
+      "step": 1344
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4101407933270865,
+      "learning_rate": 9.66010069399807e-06,
+      "loss": 0.3046,
+      "step": 1345
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5188741064562206,
+      "learning_rate": 9.659469435229992e-06,
+      "loss": 0.5703,
+      "step": 1346
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6980309497161428,
+      "learning_rate": 9.658837611485181e-06,
+      "loss": 0.4809,
+      "step": 1347
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.633866255468159,
+      "learning_rate": 9.658205222840249e-06,
+      "loss": 0.5632,
+      "step": 1348
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.520217938301737,
+      "learning_rate": 9.657572269371873e-06,
+      "loss": 0.559,
+      "step": 1349
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6036108904384827,
+      "learning_rate": 9.656938751156803e-06,
+      "loss": 0.5038,
+      "step": 1350
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.995517854694486,
+      "learning_rate": 9.656304668271855e-06,
+      "loss": 0.5278,
+      "step": 1351
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0199968459825945,
+      "learning_rate": 9.655670020793911e-06,
+      "loss": 0.4954,
+      "step": 1352
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9338041426336923,
+      "learning_rate": 9.655034808799928e-06,
+      "loss": 0.5259,
+      "step": 1353
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4164058832234008,
+      "learning_rate": 9.654399032366925e-06,
+      "loss": 0.5005,
+      "step": 1354
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.49532473758300577,
+      "learning_rate": 9.653762691571995e-06,
+      "loss": 0.3089,
+      "step": 1355
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9740024302769328,
+      "learning_rate": 9.653125786492295e-06,
+      "loss": 0.5696,
+      "step": 1356
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.781478739228713,
+      "learning_rate": 9.652488317205048e-06,
+      "loss": 0.5352,
+      "step": 1357
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3054459051293876,
+      "learning_rate": 9.651850283787557e-06,
+      "loss": 0.5248,
+      "step": 1358
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2299830654973967,
+      "learning_rate": 9.65121168631718e-06,
+      "loss": 0.5067,
+      "step": 1359
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.150983290662543,
+      "learning_rate": 9.65057252487135e-06,
+      "loss": 0.5394,
+      "step": 1360
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7116543193519367,
+      "learning_rate": 9.64993279952757e-06,
+      "loss": 0.5271,
+      "step": 1361
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5740083557375577,
+      "learning_rate": 9.649292510363408e-06,
+      "loss": 0.5376,
+      "step": 1362
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4892272981530845,
+      "learning_rate": 9.648651657456498e-06,
+      "loss": 0.527,
+      "step": 1363
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.093988512971543,
+      "learning_rate": 9.64801024088455e-06,
+      "loss": 0.5229,
+      "step": 1364
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.63110235420523,
+      "learning_rate": 9.647368260725336e-06,
+      "loss": 0.4933,
+      "step": 1365
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.351031106702062,
+      "learning_rate": 9.646725717056697e-06,
+      "loss": 0.5309,
+      "step": 1366
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.751442650224608,
+      "learning_rate": 9.646082609956546e-06,
+      "loss": 0.5131,
+      "step": 1367
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.955726769906433,
+      "learning_rate": 9.645438939502863e-06,
+      "loss": 0.574,
+      "step": 1368
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8642060728141354,
+      "learning_rate": 9.644794705773691e-06,
+      "loss": 0.5393,
+      "step": 1369
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4486065675007111,
+      "learning_rate": 9.644149908847149e-06,
+      "loss": 0.3074,
+      "step": 1370
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3740517360495264,
+      "learning_rate": 9.64350454880142e-06,
+      "loss": 0.5107,
+      "step": 1371
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4288977284960787,
+      "learning_rate": 9.642858625714753e-06,
+      "loss": 0.537,
+      "step": 1372
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.1614282475038555,
+      "learning_rate": 9.642212139665474e-06,
+      "loss": 0.5214,
+      "step": 1373
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0460397736349414,
+      "learning_rate": 9.641565090731967e-06,
+      "loss": 0.6159,
+      "step": 1374
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.834076942664928,
+      "learning_rate": 9.640917478992693e-06,
+      "loss": 0.5407,
+      "step": 1375
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.444817721173517,
+      "learning_rate": 9.640269304526175e-06,
+      "loss": 0.5189,
+      "step": 1376
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8564081987495267,
+      "learning_rate": 9.639620567411005e-06,
+      "loss": 0.501,
+      "step": 1377
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6228497142896936,
+      "learning_rate": 9.638971267725847e-06,
+      "loss": 0.5713,
+      "step": 1378
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.64154783878598,
+      "learning_rate": 9.638321405549429e-06,
+      "loss": 0.5517,
+      "step": 1379
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.838732933020347,
+      "learning_rate": 9.63767098096055e-06,
+      "loss": 0.4227,
+      "step": 1380
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.274122993779472,
+      "learning_rate": 9.637019994038076e-06,
+      "loss": 0.4425,
+      "step": 1381
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.60116398461741,
+      "learning_rate": 9.636368444860942e-06,
+      "loss": 0.5583,
+      "step": 1382
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.519662871819063,
+      "learning_rate": 9.63571633350815e-06,
+      "loss": 0.4977,
+      "step": 1383
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.7174728490794635,
+      "learning_rate": 9.635063660058772e-06,
+      "loss": 0.5488,
+      "step": 1384
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.459877178378388,
+      "learning_rate": 9.634410424591942e-06,
+      "loss": 0.5732,
+      "step": 1385
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1555962487716616,
+      "learning_rate": 9.633756627186874e-06,
+      "loss": 0.5223,
+      "step": 1386
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.211486225300507,
+      "learning_rate": 9.63310226792284e-06,
+      "loss": 0.5068,
+      "step": 1387
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.248399266687858,
+      "learning_rate": 9.632447346879182e-06,
+      "loss": 0.5803,
+      "step": 1388
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.169998215978903,
+      "learning_rate": 9.631791864135313e-06,
+      "loss": 0.5377,
+      "step": 1389
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6644933479182704,
+      "learning_rate": 9.631135819770711e-06,
+      "loss": 0.5251,
+      "step": 1390
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.45508212660194,
+      "learning_rate": 9.630479213864928e-06,
+      "loss": 0.5307,
+      "step": 1391
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.938073567119678,
+      "learning_rate": 9.629822046497575e-06,
+      "loss": 0.5102,
+      "step": 1392
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0705400312642412,
+      "learning_rate": 9.629164317748336e-06,
+      "loss": 0.5212,
+      "step": 1393
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.992372724243332,
+      "learning_rate": 9.628506027696967e-06,
+      "loss": 0.4702,
+      "step": 1394
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6926004461878033,
+      "learning_rate": 9.627847176423283e-06,
+      "loss": 0.5016,
+      "step": 1395
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3395875734268485,
+      "learning_rate": 9.627187764007175e-06,
+      "loss": 0.5141,
+      "step": 1396
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.887676928897136,
+      "learning_rate": 9.6265277905286e-06,
+      "loss": 0.5148,
+      "step": 1397
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.374558124756428,
+      "learning_rate": 9.625867256067577e-06,
+      "loss": 0.5299,
+      "step": 1398
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.979792404781306,
+      "learning_rate": 9.625206160704204e-06,
+      "loss": 0.5385,
+      "step": 1399
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.890895242342888,
+      "learning_rate": 9.624544504518637e-06,
+      "loss": 0.5373,
+      "step": 1400
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.3701009744515105,
+      "learning_rate": 9.623882287591106e-06,
+      "loss": 0.4927,
+      "step": 1401
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0260659236217355,
+      "learning_rate": 9.623219510001907e-06,
+      "loss": 0.5509,
+      "step": 1402
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7114129780235676,
+      "learning_rate": 9.622556171831404e-06,
+      "loss": 0.5155,
+      "step": 1403
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.357547239955236,
+      "learning_rate": 9.621892273160028e-06,
+      "loss": 0.5571,
+      "step": 1404
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3686005392089124,
+      "learning_rate": 9.62122781406828e-06,
+      "loss": 0.4961,
+      "step": 1405
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.988383179494459,
+      "learning_rate": 9.62056279463673e-06,
+      "loss": 0.5034,
+      "step": 1406
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8189438153000053,
+      "learning_rate": 9.619897214946007e-06,
+      "loss": 0.4741,
+      "step": 1407
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9786727175933703,
+      "learning_rate": 9.619231075076824e-06,
+      "loss": 0.5478,
+      "step": 1408
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9361251275610267,
+      "learning_rate": 9.618564375109946e-06,
+      "loss": 0.5619,
+      "step": 1409
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1932979762225293,
+      "learning_rate": 9.617897115126216e-06,
+      "loss": 0.5839,
+      "step": 1410
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.944222869519505,
+      "learning_rate": 9.617229295206538e-06,
+      "loss": 0.5195,
+      "step": 1411
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.722642494149759,
+      "learning_rate": 9.616560915431892e-06,
+      "loss": 0.5584,
+      "step": 1412
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.788084568218789,
+      "learning_rate": 9.61589197588332e-06,
+      "loss": 0.531,
+      "step": 1413
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0058679216855144,
+      "learning_rate": 9.61522247664193e-06,
+      "loss": 0.57,
+      "step": 1414
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.363878446828729,
+      "learning_rate": 9.614552417788906e-06,
+      "loss": 0.5277,
+      "step": 1415
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.248042057219263,
+      "learning_rate": 9.613881799405492e-06,
+      "loss": 0.5371,
+      "step": 1416
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.584844156439413,
+      "learning_rate": 9.613210621573003e-06,
+      "loss": 0.5436,
+      "step": 1417
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.003603604963202,
+      "learning_rate": 9.612538884372821e-06,
+      "loss": 0.5547,
+      "step": 1418
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3511896687112075,
+      "learning_rate": 9.611866587886399e-06,
+      "loss": 0.5402,
+      "step": 1419
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.632642113272902,
+      "learning_rate": 9.611193732195254e-06,
+      "loss": 0.5442,
+      "step": 1420
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5030130146347205,
+      "learning_rate": 9.61052031738097e-06,
+      "loss": 0.3287,
+      "step": 1421
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.561281606243076,
+      "learning_rate": 9.609846343525204e-06,
+      "loss": 0.5094,
+      "step": 1422
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8793266734366942,
+      "learning_rate": 9.609171810709676e-06,
+      "loss": 0.5082,
+      "step": 1423
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38487188212650714,
+      "learning_rate": 9.608496719016176e-06,
+      "loss": 0.3096,
+      "step": 1424
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.517918811590563,
+      "learning_rate": 9.60782106852656e-06,
+      "loss": 0.5123,
+      "step": 1425
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.45808353231926113,
+      "learning_rate": 9.607144859322755e-06,
+      "loss": 0.3072,
+      "step": 1426
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.069856662958971,
+      "learning_rate": 9.60646809148675e-06,
+      "loss": 0.5518,
+      "step": 1427
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6635550764913387,
+      "learning_rate": 9.60579076510061e-06,
+      "loss": 0.5133,
+      "step": 1428
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4265988688810982,
+      "learning_rate": 9.605112880246463e-06,
+      "loss": 0.3078,
+      "step": 1429
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.543239747121528,
+      "learning_rate": 9.6044344370065e-06,
+      "loss": 0.4857,
+      "step": 1430
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.537432432779643,
+      "learning_rate": 9.603755435462989e-06,
+      "loss": 0.5322,
+      "step": 1431
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5179654557661735,
+      "learning_rate": 9.60307587569826e-06,
+      "loss": 0.4418,
+      "step": 1432
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.587334445404809,
+      "learning_rate": 9.602395757794712e-06,
+      "loss": 0.5004,
+      "step": 1433
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.69662027045739,
+      "learning_rate": 9.60171508183481e-06,
+      "loss": 0.523,
+      "step": 1434
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.278323680175719,
+      "learning_rate": 9.601033847901092e-06,
+      "loss": 0.5273,
+      "step": 1435
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9066146766074463,
+      "learning_rate": 9.600352056076154e-06,
+      "loss": 0.5538,
+      "step": 1436
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4252735377690194,
+      "learning_rate": 9.599669706442674e-06,
+      "loss": 0.3216,
+      "step": 1437
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.812402411459362,
+      "learning_rate": 9.59898679908338e-06,
+      "loss": 0.5056,
+      "step": 1438
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.774004268736158,
+      "learning_rate": 9.598303334081084e-06,
+      "loss": 0.4844,
+      "step": 1439
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6534485328562023,
+      "learning_rate": 9.597619311518658e-06,
+      "loss": 0.5662,
+      "step": 1440
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.177531510757156,
+      "learning_rate": 9.596934731479038e-06,
+      "loss": 0.5232,
+      "step": 1441
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5885844876244777,
+      "learning_rate": 9.596249594045233e-06,
+      "loss": 0.5159,
+      "step": 1442
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7801488888273793,
+      "learning_rate": 9.59556389930032e-06,
+      "loss": 0.5112,
+      "step": 1443
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3939869343421967,
+      "learning_rate": 9.594877647327441e-06,
+      "loss": 0.3028,
+      "step": 1444
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.072857199460597,
+      "learning_rate": 9.594190838209806e-06,
+      "loss": 0.5005,
+      "step": 1445
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3109799540478715,
+      "learning_rate": 9.593503472030693e-06,
+      "loss": 0.6036,
+      "step": 1446
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.301913395161588,
+      "learning_rate": 9.592815548873448e-06,
+      "loss": 0.4987,
+      "step": 1447
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.709887499180956,
+      "learning_rate": 9.592127068821484e-06,
+      "loss": 0.6079,
+      "step": 1448
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.936662461109622,
+      "learning_rate": 9.591438031958282e-06,
+      "loss": 0.6045,
+      "step": 1449
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1855877728929682,
+      "learning_rate": 9.59074843836739e-06,
+      "loss": 0.5081,
+      "step": 1450
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6033850800693417,
+      "learning_rate": 9.59005828813242e-06,
+      "loss": 0.5598,
+      "step": 1451
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4422472515426077,
+      "learning_rate": 9.589367581337062e-06,
+      "loss": 0.485,
+      "step": 1452
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9878591533792003,
+      "learning_rate": 9.58867631806506e-06,
+      "loss": 0.5189,
+      "step": 1453
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.553111395616161,
+      "learning_rate": 9.587984498400236e-06,
+      "loss": 0.5697,
+      "step": 1454
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.202037641499834,
+      "learning_rate": 9.587292122426475e-06,
+      "loss": 0.5406,
+      "step": 1455
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.268230119344162,
+      "learning_rate": 9.586599190227728e-06,
+      "loss": 0.5466,
+      "step": 1456
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.523350204857247,
+      "learning_rate": 9.585905701888019e-06,
+      "loss": 0.5302,
+      "step": 1457
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.304803282992968,
+      "learning_rate": 9.58521165749143e-06,
+      "loss": 0.5382,
+      "step": 1458
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.792926214874983,
+      "learning_rate": 9.584517057122122e-06,
+      "loss": 0.5042,
+      "step": 1459
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.899353320200434,
+      "learning_rate": 9.583821900864316e-06,
+      "loss": 0.5142,
+      "step": 1460
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1628152252803394,
+      "learning_rate": 9.583126188802301e-06,
+      "loss": 0.5371,
+      "step": 1461
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.87790568028764,
+      "learning_rate": 9.582429921020436e-06,
+      "loss": 0.5469,
+      "step": 1462
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.503782528934701,
+      "learning_rate": 9.581733097603146e-06,
+      "loss": 0.4903,
+      "step": 1463
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8124271708119073,
+      "learning_rate": 9.581035718634919e-06,
+      "loss": 0.565,
+      "step": 1464
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4830675364687127,
+      "learning_rate": 9.58033778420032e-06,
+      "loss": 0.5304,
+      "step": 1465
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.108254800452208,
+      "learning_rate": 9.579639294383974e-06,
+      "loss": 0.5062,
+      "step": 1466
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5831359045070563,
+      "learning_rate": 9.578940249270573e-06,
+      "loss": 0.5444,
+      "step": 1467
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.570468704028667,
+      "learning_rate": 9.578240648944882e-06,
+      "loss": 0.5293,
+      "step": 1468
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4693944011358733,
+      "learning_rate": 9.577540493491731e-06,
+      "loss": 0.5242,
+      "step": 1469
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6193520449650447,
+      "learning_rate": 9.576839782996013e-06,
+      "loss": 0.5501,
+      "step": 1470
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.547745899394878,
+      "learning_rate": 9.576138517542691e-06,
+      "loss": 0.5707,
+      "step": 1471
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1931836660622515,
+      "learning_rate": 9.575436697216798e-06,
+      "loss": 0.4978,
+      "step": 1472
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.9570937660261025,
+      "learning_rate": 9.574734322103431e-06,
+      "loss": 0.5691,
+      "step": 1473
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4605207839706409,
+      "learning_rate": 9.574031392287756e-06,
+      "loss": 0.3068,
+      "step": 1474
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.110648044901402,
+      "learning_rate": 9.573327907855006e-06,
+      "loss": 0.5278,
+      "step": 1475
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.283475020250859,
+      "learning_rate": 9.572623868890482e-06,
+      "loss": 0.4912,
+      "step": 1476
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.057673003547701,
+      "learning_rate": 9.571919275479549e-06,
+      "loss": 0.5383,
+      "step": 1477
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4248871329111702,
+      "learning_rate": 9.57121412770764e-06,
+      "loss": 0.3053,
+      "step": 1478
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.193403392061226,
+      "learning_rate": 9.570508425660261e-06,
+      "loss": 0.5061,
+      "step": 1479
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.409833779866407,
+      "learning_rate": 9.569802169422978e-06,
+      "loss": 0.3105,
+      "step": 1480
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.851305878629524,
+      "learning_rate": 9.569095359081426e-06,
+      "loss": 0.5383,
+      "step": 1481
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5917156035016644,
+      "learning_rate": 9.56838799472131e-06,
+      "loss": 0.5026,
+      "step": 1482
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.762340432662573,
+      "learning_rate": 9.567680076428401e-06,
+      "loss": 0.5232,
+      "step": 1483
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4568883839156495,
+      "learning_rate": 9.566971604288535e-06,
+      "loss": 0.2954,
+      "step": 1484
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.41681243485749614,
+      "learning_rate": 9.566262578387616e-06,
+      "loss": 0.308,
+      "step": 1485
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.536896823608822,
+      "learning_rate": 9.56555299881162e-06,
+      "loss": 0.521,
+      "step": 1486
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.012183981936292,
+      "learning_rate": 9.564842865646581e-06,
+      "loss": 0.5747,
+      "step": 1487
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6773004303290833,
+      "learning_rate": 9.564132178978607e-06,
+      "loss": 0.5262,
+      "step": 1488
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.716611459948215,
+      "learning_rate": 9.563420938893872e-06,
+      "loss": 0.547,
+      "step": 1489
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.129279457152869,
+      "learning_rate": 9.562709145478616e-06,
+      "loss": 0.4931,
+      "step": 1490
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6102958363092217,
+      "learning_rate": 9.561996798819146e-06,
+      "loss": 0.5132,
+      "step": 1491
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.067234353511963,
+      "learning_rate": 9.561283899001837e-06,
+      "loss": 0.5186,
+      "step": 1492
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4318211033931443,
+      "learning_rate": 9.560570446113128e-06,
+      "loss": 0.3244,
+      "step": 1493
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.197262000085641,
+      "learning_rate": 9.559856440239531e-06,
+      "loss": 0.5127,
+      "step": 1494
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.164214935237647,
+      "learning_rate": 9.55914188146762e-06,
+      "loss": 0.4745,
+      "step": 1495
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.019573575213979,
+      "learning_rate": 9.558426769884039e-06,
+      "loss": 0.5285,
+      "step": 1496
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3472353446835044,
+      "learning_rate": 9.557711105575496e-06,
+      "loss": 0.5261,
+      "step": 1497
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9416045358820235,
+      "learning_rate": 9.55699488862877e-06,
+      "loss": 0.493,
+      "step": 1498
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7202086657983857,
+      "learning_rate": 9.556278119130702e-06,
+      "loss": 0.5706,
+      "step": 1499
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.154475267231696,
+      "learning_rate": 9.555560797168203e-06,
+      "loss": 0.535,
+      "step": 1500
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3889008552783646,
+      "learning_rate": 9.554842922828254e-06,
+      "loss": 0.5299,
+      "step": 1501
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6265950290702564,
+      "learning_rate": 9.554124496197899e-06,
+      "loss": 0.5242,
+      "step": 1502
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.424035134083923,
+      "learning_rate": 9.553405517364246e-06,
+      "loss": 0.5665,
+      "step": 1503
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.47912596948256786,
+      "learning_rate": 9.552685986414477e-06,
+      "loss": 0.3041,
+      "step": 1504
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.144553999609455,
+      "learning_rate": 9.551965903435836e-06,
+      "loss": 0.5071,
+      "step": 1505
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.484273460588695,
+      "learning_rate": 9.551245268515637e-06,
+      "loss": 0.5429,
+      "step": 1506
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.205154448855923,
+      "learning_rate": 9.550524081741256e-06,
+      "loss": 0.5494,
+      "step": 1507
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.418772062474472,
+      "learning_rate": 9.549802343200146e-06,
+      "loss": 0.5016,
+      "step": 1508
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1126223334568,
+      "learning_rate": 9.549080052979814e-06,
+      "loss": 0.5068,
+      "step": 1509
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4079822232632724,
+      "learning_rate": 9.548357211167842e-06,
+      "loss": 0.3116,
+      "step": 1510
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2784062537378835,
+      "learning_rate": 9.54763381785188e-06,
+      "loss": 0.5562,
+      "step": 1511
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3855829640163346,
+      "learning_rate": 9.546909873119637e-06,
+      "loss": 0.2937,
+      "step": 1512
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4459879941505528,
+      "learning_rate": 9.5461853770589e-06,
+      "loss": 0.3166,
+      "step": 1513
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1508994002892945,
+      "learning_rate": 9.545460329757508e-06,
+      "loss": 0.475,
+      "step": 1514
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9232371856653376,
+      "learning_rate": 9.544734731303383e-06,
+      "loss": 0.507,
+      "step": 1515
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8610572741189846,
+      "learning_rate": 9.544008581784503e-06,
+      "loss": 0.5433,
+      "step": 1516
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4377951460663121,
+      "learning_rate": 9.543281881288918e-06,
+      "loss": 0.3,
+      "step": 1517
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9720638046820125,
+      "learning_rate": 9.542554629904742e-06,
+      "loss": 0.5709,
+      "step": 1518
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1990739299716657,
+      "learning_rate": 9.541826827720156e-06,
+      "loss": 0.4847,
+      "step": 1519
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8354378750051064,
+      "learning_rate": 9.541098474823408e-06,
+      "loss": 0.5196,
+      "step": 1520
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3593292177271628,
+      "learning_rate": 9.540369571302816e-06,
+      "loss": 0.5553,
+      "step": 1521
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.04471529587606,
+      "learning_rate": 9.539640117246758e-06,
+      "loss": 0.5219,
+      "step": 1522
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.895483058305233,
+      "learning_rate": 9.538910112743689e-06,
+      "loss": 0.5061,
+      "step": 1523
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.343236981245239,
+      "learning_rate": 9.538179557882119e-06,
+      "loss": 0.4673,
+      "step": 1524
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6562980535999996,
+      "learning_rate": 9.53744845275063e-06,
+      "loss": 0.5003,
+      "step": 1525
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.546742602080648,
+      "learning_rate": 9.536716797437876e-06,
+      "loss": 0.5702,
+      "step": 1526
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2203244773808883,
+      "learning_rate": 9.535984592032568e-06,
+      "loss": 0.4884,
+      "step": 1527
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.26986282821875,
+      "learning_rate": 9.535251836623491e-06,
+      "loss": 0.5186,
+      "step": 1528
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.990312426849362,
+      "learning_rate": 9.534518531299494e-06,
+      "loss": 0.5337,
+      "step": 1529
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4686009467553461,
+      "learning_rate": 9.533784676149493e-06,
+      "loss": 0.3104,
+      "step": 1530
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4297067303078612,
+      "learning_rate": 9.533050271262468e-06,
+      "loss": 0.302,
+      "step": 1531
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5914131390129143,
+      "learning_rate": 9.53231531672747e-06,
+      "loss": 0.5409,
+      "step": 1532
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4676767363598606,
+      "learning_rate": 9.531579812633616e-06,
+      "loss": 0.5513,
+      "step": 1533
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0017610083508304,
+      "learning_rate": 9.530843759070085e-06,
+      "loss": 0.4932,
+      "step": 1534
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.964742219523941,
+      "learning_rate": 9.530107156126129e-06,
+      "loss": 0.5291,
+      "step": 1535
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1952664471089793,
+      "learning_rate": 9.529370003891063e-06,
+      "loss": 0.5363,
+      "step": 1536
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3519200104801494,
+      "learning_rate": 9.528632302454267e-06,
+      "loss": 0.5146,
+      "step": 1537
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.520215703803163,
+      "learning_rate": 9.527894051905194e-06,
+      "loss": 0.6145,
+      "step": 1538
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.286816683483525,
+      "learning_rate": 9.527155252333357e-06,
+      "loss": 0.5317,
+      "step": 1539
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.321914637464794,
+      "learning_rate": 9.526415903828338e-06,
+      "loss": 0.5262,
+      "step": 1540
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8251617213106677,
+      "learning_rate": 9.525676006479785e-06,
+      "loss": 0.5559,
+      "step": 1541
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.939827848012452,
+      "learning_rate": 9.524935560377414e-06,
+      "loss": 0.4865,
+      "step": 1542
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.639775612634929,
+      "learning_rate": 9.524194565611007e-06,
+      "loss": 0.498,
+      "step": 1543
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2733580879579045,
+      "learning_rate": 9.52345302227041e-06,
+      "loss": 0.5614,
+      "step": 1544
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.501705078056028,
+      "learning_rate": 9.522710930445542e-06,
+      "loss": 0.5311,
+      "step": 1545
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3915448115799,
+      "learning_rate": 9.52196829022638e-06,
+      "loss": 0.5543,
+      "step": 1546
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.459047028551328,
+      "learning_rate": 9.521225101702973e-06,
+      "loss": 0.5173,
+      "step": 1547
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.571358307383417,
+      "learning_rate": 9.520481364965436e-06,
+      "loss": 0.522,
+      "step": 1548
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.538826785516368,
+      "learning_rate": 9.519737080103949e-06,
+      "loss": 0.4827,
+      "step": 1549
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.294027139149512,
+      "learning_rate": 9.518992247208757e-06,
+      "loss": 0.6315,
+      "step": 1550
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5320802892301717,
+      "learning_rate": 9.51824686637018e-06,
+      "loss": 0.3218,
+      "step": 1551
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.978476750213898,
+      "learning_rate": 9.517500937678591e-06,
+      "loss": 0.4894,
+      "step": 1552
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8676520776182266,
+      "learning_rate": 9.51675446122444e-06,
+      "loss": 0.5144,
+      "step": 1553
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1634687500204626,
+      "learning_rate": 9.516007437098238e-06,
+      "loss": 0.4914,
+      "step": 1554
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.485246838260815,
+      "learning_rate": 9.515259865390565e-06,
+      "loss": 0.5108,
+      "step": 1555
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.42022584764607,
+      "learning_rate": 9.514511746192067e-06,
+      "loss": 0.5062,
+      "step": 1556
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4337834599408943,
+      "learning_rate": 9.513763079593457e-06,
+      "loss": 0.5348,
+      "step": 1557
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3097947870527937,
+      "learning_rate": 9.513013865685511e-06,
+      "loss": 0.543,
+      "step": 1558
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.336918292796128,
+      "learning_rate": 9.512264104559077e-06,
+      "loss": 0.5189,
+      "step": 1559
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5677237090356897,
+      "learning_rate": 9.511513796305063e-06,
+      "loss": 0.5357,
+      "step": 1560
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7755874720423592,
+      "learning_rate": 9.510762941014446e-06,
+      "loss": 0.5133,
+      "step": 1561
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.151382638972247,
+      "learning_rate": 9.510011538778275e-06,
+      "loss": 0.5083,
+      "step": 1562
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4488763229986232,
+      "learning_rate": 9.509259589687653e-06,
+      "loss": 0.5527,
+      "step": 1563
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.422159442182896,
+      "learning_rate": 9.50850709383376e-06,
+      "loss": 0.5717,
+      "step": 1564
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5338529816740766,
+      "learning_rate": 9.507754051307842e-06,
+      "loss": 0.3,
+      "step": 1565
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.43670843440068,
+      "learning_rate": 9.507000462201202e-06,
+      "loss": 0.5744,
+      "step": 1566
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.43401506212998986,
+      "learning_rate": 9.50624632660522e-06,
+      "loss": 0.2894,
+      "step": 1567
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2561265088442206,
+      "learning_rate": 9.505491644611334e-06,
+      "loss": 0.5231,
+      "step": 1568
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0134596477674114,
+      "learning_rate": 9.504736416311053e-06,
+      "loss": 0.526,
+      "step": 1569
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.229006393544419,
+      "learning_rate": 9.503980641795952e-06,
+      "loss": 0.5145,
+      "step": 1570
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.7415614188573025,
+      "learning_rate": 9.503224321157671e-06,
+      "loss": 0.5193,
+      "step": 1571
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.606253513220701,
+      "learning_rate": 9.502467454487916e-06,
+      "loss": 0.5453,
+      "step": 1572
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.576449698403622,
+      "learning_rate": 9.501710041878457e-06,
+      "loss": 0.5138,
+      "step": 1573
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.843526332351611,
+      "learning_rate": 9.50095208342114e-06,
+      "loss": 0.5305,
+      "step": 1574
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0957820889528587,
+      "learning_rate": 9.500193579207863e-06,
+      "loss": 0.5522,
+      "step": 1575
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2407457904366836,
+      "learning_rate": 9.499434529330601e-06,
+      "loss": 0.5271,
+      "step": 1576
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.581058066401948,
+      "learning_rate": 9.498674933881392e-06,
+      "loss": 0.5353,
+      "step": 1577
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9667389497144674,
+      "learning_rate": 9.497914792952334e-06,
+      "loss": 0.5451,
+      "step": 1578
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.144630996394142,
+      "learning_rate": 9.497154106635604e-06,
+      "loss": 0.5602,
+      "step": 1579
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2445633882609317,
+      "learning_rate": 9.496392875023433e-06,
+      "loss": 0.5223,
+      "step": 1580
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.666332897462789,
+      "learning_rate": 9.495631098208125e-06,
+      "loss": 0.5179,
+      "step": 1581
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.479065585904251,
+      "learning_rate": 9.494868776282047e-06,
+      "loss": 0.5101,
+      "step": 1582
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0419483915815926,
+      "learning_rate": 9.494105909337635e-06,
+      "loss": 0.5529,
+      "step": 1583
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.824488134910987,
+      "learning_rate": 9.493342497467386e-06,
+      "loss": 0.4987,
+      "step": 1584
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.9106712338373955,
+      "learning_rate": 9.49257854076387e-06,
+      "loss": 0.506,
+      "step": 1585
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0871158316497294,
+      "learning_rate": 9.491814039319716e-06,
+      "loss": 0.5737,
+      "step": 1586
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.458404176921433,
+      "learning_rate": 9.491048993227626e-06,
+      "loss": 0.5411,
+      "step": 1587
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.210444820351046,
+      "learning_rate": 9.490283402580362e-06,
+      "loss": 0.5235,
+      "step": 1588
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.067811090460618,
+      "learning_rate": 9.489517267470754e-06,
+      "loss": 0.4696,
+      "step": 1589
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6408974265758367,
+      "learning_rate": 9.488750587991701e-06,
+      "loss": 0.6012,
+      "step": 1590
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1939167268224424,
+      "learning_rate": 9.487983364236163e-06,
+      "loss": 0.5277,
+      "step": 1591
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9699474740434675,
+      "learning_rate": 9.48721559629717e-06,
+      "loss": 0.4315,
+      "step": 1592
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.811821160984221,
+      "learning_rate": 9.486447284267816e-06,
+      "loss": 0.3152,
+      "step": 1593
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4211860886331826,
+      "learning_rate": 9.485678428241264e-06,
+      "loss": 0.5611,
+      "step": 1594
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.24140286658992,
+      "learning_rate": 9.484909028310735e-06,
+      "loss": 0.4945,
+      "step": 1595
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4897628082751751,
+      "learning_rate": 9.484139084569526e-06,
+      "loss": 0.3087,
+      "step": 1596
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9167758904241388,
+      "learning_rate": 9.483368597110993e-06,
+      "loss": 0.5359,
+      "step": 1597
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4900882592072843,
+      "learning_rate": 9.48259756602856e-06,
+      "loss": 0.3248,
+      "step": 1598
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5317921471978976,
+      "learning_rate": 9.48182599141572e-06,
+      "loss": 0.3152,
+      "step": 1599
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5802510460045927,
+      "learning_rate": 9.481053873366027e-06,
+      "loss": 0.5308,
+      "step": 1600
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4205947575929945,
+      "learning_rate": 9.480281211973103e-06,
+      "loss": 0.5138,
+      "step": 1601
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8656021224676658,
+      "learning_rate": 9.479508007330638e-06,
+      "loss": 0.5442,
+      "step": 1602
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.059897287791515,
+      "learning_rate": 9.478734259532382e-06,
+      "loss": 0.5174,
+      "step": 1603
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.957825655297245,
+      "learning_rate": 9.477959968672156e-06,
+      "loss": 0.5371,
+      "step": 1604
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.12488809245207,
+      "learning_rate": 9.477185134843847e-06,
+      "loss": 0.5502,
+      "step": 1605
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4431635622813497,
+      "learning_rate": 9.476409758141404e-06,
+      "loss": 0.5609,
+      "step": 1606
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1397425025569734,
+      "learning_rate": 9.475633838658847e-06,
+      "loss": 0.4837,
+      "step": 1607
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.591425204752494,
+      "learning_rate": 9.474857376490257e-06,
+      "loss": 0.5802,
+      "step": 1608
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4455478145215963,
+      "learning_rate": 9.474080371729782e-06,
+      "loss": 0.5268,
+      "step": 1609
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.764121771718693,
+      "learning_rate": 9.473302824471637e-06,
+      "loss": 0.5278,
+      "step": 1610
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.71777917898482,
+      "learning_rate": 9.472524734810105e-06,
+      "loss": 0.5607,
+      "step": 1611
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8991770098290233,
+      "learning_rate": 9.471746102839527e-06,
+      "loss": 0.5025,
+      "step": 1612
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6418234523225232,
+      "learning_rate": 9.470966928654321e-06,
+      "loss": 0.3374,
+      "step": 1613
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.629371011796463,
+      "learning_rate": 9.470187212348958e-06,
+      "loss": 0.5214,
+      "step": 1614
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.696798241644457,
+      "learning_rate": 9.469406954017985e-06,
+      "loss": 0.5613,
+      "step": 1615
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4629313270651903,
+      "learning_rate": 9.468626153756011e-06,
+      "loss": 0.3259,
+      "step": 1616
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9148771330169194,
+      "learning_rate": 9.46784481165771e-06,
+      "loss": 0.5706,
+      "step": 1617
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.330811571859148,
+      "learning_rate": 9.467062927817822e-06,
+      "loss": 0.5262,
+      "step": 1618
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.46174250665732824,
+      "learning_rate": 9.466280502331156e-06,
+      "loss": 0.3046,
+      "step": 1619
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.346248808121316,
+      "learning_rate": 9.465497535292579e-06,
+      "loss": 0.5651,
+      "step": 1620
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4353195549527933,
+      "learning_rate": 9.464714026797032e-06,
+      "loss": 0.5171,
+      "step": 1621
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9407752707985435,
+      "learning_rate": 9.463929976939516e-06,
+      "loss": 0.5082,
+      "step": 1622
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5349393954135646,
+      "learning_rate": 9.463145385815102e-06,
+      "loss": 0.5237,
+      "step": 1623
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2873850353266603,
+      "learning_rate": 9.462360253518923e-06,
+      "loss": 0.5179,
+      "step": 1624
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7133051601453486,
+      "learning_rate": 9.461574580146179e-06,
+      "loss": 0.5275,
+      "step": 1625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5494147302773866,
+      "learning_rate": 9.460788365792135e-06,
+      "loss": 0.5284,
+      "step": 1626
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9450437822533853,
+      "learning_rate": 9.460001610552125e-06,
+      "loss": 0.5349,
+      "step": 1627
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2241726703686124,
+      "learning_rate": 9.45921431452154e-06,
+      "loss": 0.5111,
+      "step": 1628
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.926804409009167,
+      "learning_rate": 9.45842647779585e-06,
+      "loss": 0.5263,
+      "step": 1629
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9612660634661436,
+      "learning_rate": 9.457638100470577e-06,
+      "loss": 0.4872,
+      "step": 1630
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5802751250483078,
+      "learning_rate": 9.456849182641318e-06,
+      "loss": 0.3082,
+      "step": 1631
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4560988131728734,
+      "learning_rate": 9.45605972440373e-06,
+      "loss": 0.5212,
+      "step": 1632
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.677461700902719,
+      "learning_rate": 9.455269725853538e-06,
+      "loss": 0.5085,
+      "step": 1633
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.416002903224991,
+      "learning_rate": 9.45447918708653e-06,
+      "loss": 0.5271,
+      "step": 1634
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3629553654950497,
+      "learning_rate": 9.453688108198567e-06,
+      "loss": 0.5084,
+      "step": 1635
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4160267782662914,
+      "learning_rate": 9.452896489285563e-06,
+      "loss": 0.5258,
+      "step": 1636
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.746464479586174,
+      "learning_rate": 9.45210433044351e-06,
+      "loss": 0.5259,
+      "step": 1637
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4125778336465415,
+      "learning_rate": 9.451311631768459e-06,
+      "loss": 0.3016,
+      "step": 1638
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3775088733372125,
+      "learning_rate": 9.450518393356527e-06,
+      "loss": 0.5449,
+      "step": 1639
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7404321781823011,
+      "learning_rate": 9.449724615303894e-06,
+      "loss": 0.5069,
+      "step": 1640
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.91606063566569,
+      "learning_rate": 9.448930297706813e-06,
+      "loss": 0.5206,
+      "step": 1641
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7023097018726983,
+      "learning_rate": 9.448135440661595e-06,
+      "loss": 0.565,
+      "step": 1642
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.688307844435922,
+      "learning_rate": 9.44734004426462e-06,
+      "loss": 0.5576,
+      "step": 1643
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.399927049440964,
+      "learning_rate": 9.446544108612332e-06,
+      "loss": 0.5457,
+      "step": 1644
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7801718651992453,
+      "learning_rate": 9.445747633801241e-06,
+      "loss": 0.544,
+      "step": 1645
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0895221627474467,
+      "learning_rate": 9.444950619927924e-06,
+      "loss": 0.5235,
+      "step": 1646
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6022632117751963,
+      "learning_rate": 9.44415306708902e-06,
+      "loss": 0.54,
+      "step": 1647
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3264042489107495,
+      "learning_rate": 9.443354975381233e-06,
+      "loss": 0.4748,
+      "step": 1648
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7469121055851238,
+      "learning_rate": 9.44255634490134e-06,
+      "loss": 0.4853,
+      "step": 1649
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.939606094052758,
+      "learning_rate": 9.441757175746171e-06,
+      "loss": 0.4828,
+      "step": 1650
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.274178783772827,
+      "learning_rate": 9.440957468012633e-06,
+      "loss": 0.5049,
+      "step": 1651
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9100662912533746,
+      "learning_rate": 9.440157221797692e-06,
+      "loss": 0.5294,
+      "step": 1652
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9823864222775844,
+      "learning_rate": 9.43935643719838e-06,
+      "loss": 0.5637,
+      "step": 1653
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.310834494445614,
+      "learning_rate": 9.438555114311796e-06,
+      "loss": 0.5301,
+      "step": 1654
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3920938647453998,
+      "learning_rate": 9.4377532532351e-06,
+      "loss": 0.5041,
+      "step": 1655
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9468637875533443,
+      "learning_rate": 9.436950854065526e-06,
+      "loss": 0.5224,
+      "step": 1656
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9080648272861227,
+      "learning_rate": 9.436147916900361e-06,
+      "loss": 0.4809,
+      "step": 1657
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.169617184341406,
+      "learning_rate": 9.43534444183697e-06,
+      "loss": 0.5092,
+      "step": 1658
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5561885926434513,
+      "learning_rate": 9.434540428972772e-06,
+      "loss": 0.5564,
+      "step": 1659
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2733056909908997,
+      "learning_rate": 9.433735878405262e-06,
+      "loss": 0.4739,
+      "step": 1660
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8871471111519793,
+      "learning_rate": 9.43293079023199e-06,
+      "loss": 0.555,
+      "step": 1661
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.741232542425293,
+      "learning_rate": 9.432125164550576e-06,
+      "loss": 0.5215,
+      "step": 1662
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.033309971148459,
+      "learning_rate": 9.431319001458705e-06,
+      "loss": 0.5828,
+      "step": 1663
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.425797621713783,
+      "learning_rate": 9.43051230105413e-06,
+      "loss": 0.5019,
+      "step": 1664
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.46820835739125305,
+      "learning_rate": 9.429705063434664e-06,
+      "loss": 0.3164,
+      "step": 1665
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.092266950367032,
+      "learning_rate": 9.428897288698189e-06,
+      "loss": 0.5315,
+      "step": 1666
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.458630548994502,
+      "learning_rate": 9.428088976942647e-06,
+      "loss": 0.509,
+      "step": 1667
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.775256257931443,
+      "learning_rate": 9.427280128266049e-06,
+      "loss": 0.5469,
+      "step": 1668
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8999368819238915,
+      "learning_rate": 9.426470742766476e-06,
+      "loss": 0.5185,
+      "step": 1669
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.94905776459023,
+      "learning_rate": 9.425660820542064e-06,
+      "loss": 0.4887,
+      "step": 1670
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9673949619617077,
+      "learning_rate": 9.424850361691019e-06,
+      "loss": 0.5694,
+      "step": 1671
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.989818203511933,
+      "learning_rate": 9.424039366311612e-06,
+      "loss": 0.5112,
+      "step": 1672
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4056604984386101,
+      "learning_rate": 9.423227834502183e-06,
+      "loss": 0.3042,
+      "step": 1673
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1271861595020654,
+      "learning_rate": 9.422415766361125e-06,
+      "loss": 0.4983,
+      "step": 1674
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.527097578998975,
+      "learning_rate": 9.421603161986912e-06,
+      "loss": 0.5526,
+      "step": 1675
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.783656104831014,
+      "learning_rate": 9.420790021478073e-06,
+      "loss": 0.4963,
+      "step": 1676
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.354010637316767,
+      "learning_rate": 9.419976344933201e-06,
+      "loss": 0.5538,
+      "step": 1677
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6462007736659863,
+      "learning_rate": 9.419162132450961e-06,
+      "loss": 0.5149,
+      "step": 1678
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4656736274235453,
+      "learning_rate": 9.418347384130076e-06,
+      "loss": 0.5175,
+      "step": 1679
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.252947938560004,
+      "learning_rate": 9.41753210006934e-06,
+      "loss": 0.5318,
+      "step": 1680
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1021053784346946,
+      "learning_rate": 9.416716280367605e-06,
+      "loss": 0.5767,
+      "step": 1681
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.148076007283585,
+      "learning_rate": 9.415899925123796e-06,
+      "loss": 0.5379,
+      "step": 1682
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2850382670426885,
+      "learning_rate": 9.415083034436896e-06,
+      "loss": 0.5465,
+      "step": 1683
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5805848832431435,
+      "learning_rate": 9.414265608405956e-06,
+      "loss": 0.6038,
+      "step": 1684
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.733047245569631,
+      "learning_rate": 9.413447647130096e-06,
+      "loss": 0.5301,
+      "step": 1685
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.973356038891759,
+      "learning_rate": 9.412629150708493e-06,
+      "loss": 0.5759,
+      "step": 1686
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.223516426663284,
+      "learning_rate": 9.41181011924039e-06,
+      "loss": 0.4842,
+      "step": 1687
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.299749242130541,
+      "learning_rate": 9.4109905528251e-06,
+      "loss": 0.5539,
+      "step": 1688
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.832833842209327,
+      "learning_rate": 9.410170451562002e-06,
+      "loss": 0.4911,
+      "step": 1689
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2902420310149885,
+      "learning_rate": 9.40934981555053e-06,
+      "loss": 0.5528,
+      "step": 1690
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.44053109427136145,
+      "learning_rate": 9.408528644890191e-06,
+      "loss": 0.3274,
+      "step": 1691
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8044914558117777,
+      "learning_rate": 9.407706939680557e-06,
+      "loss": 0.5527,
+      "step": 1692
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1373918788518482,
+      "learning_rate": 9.406884700021261e-06,
+      "loss": 0.5408,
+      "step": 1693
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9351993762082618,
+      "learning_rate": 9.406061926012001e-06,
+      "loss": 0.534,
+      "step": 1694
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.63742728483613,
+      "learning_rate": 9.405238617752543e-06,
+      "loss": 0.5247,
+      "step": 1695
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8052473018629906,
+      "learning_rate": 9.404414775342715e-06,
+      "loss": 0.5394,
+      "step": 1696
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.496611362746242,
+      "learning_rate": 9.403590398882411e-06,
+      "loss": 0.5013,
+      "step": 1697
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.031756152632166,
+      "learning_rate": 9.402765488471592e-06,
+      "loss": 0.5125,
+      "step": 1698
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.180642896955139,
+      "learning_rate": 9.401940044210276e-06,
+      "loss": 0.6037,
+      "step": 1699
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8682534343295734,
+      "learning_rate": 9.401114066198556e-06,
+      "loss": 0.5378,
+      "step": 1700
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1106854652273928,
+      "learning_rate": 9.400287554536582e-06,
+      "loss": 0.518,
+      "step": 1701
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.171298035848983,
+      "learning_rate": 9.399460509324571e-06,
+      "loss": 0.5156,
+      "step": 1702
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.90678593739655,
+      "learning_rate": 9.398632930662805e-06,
+      "loss": 0.5517,
+      "step": 1703
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.363731949978896,
+      "learning_rate": 9.397804818651634e-06,
+      "loss": 0.4971,
+      "step": 1704
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.580988467702731,
+      "learning_rate": 9.396976173391467e-06,
+      "loss": 0.5398,
+      "step": 1705
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.920951749096631,
+      "learning_rate": 9.39614699498278e-06,
+      "loss": 0.547,
+      "step": 1706
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.423336008847475,
+      "learning_rate": 9.395317283526113e-06,
+      "loss": 0.5221,
+      "step": 1707
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4337942475445473,
+      "learning_rate": 9.394487039122074e-06,
+      "loss": 0.3144,
+      "step": 1708
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.482487024463605,
+      "learning_rate": 9.39365626187133e-06,
+      "loss": 0.5182,
+      "step": 1709
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.885603018599122,
+      "learning_rate": 9.392824951874618e-06,
+      "loss": 0.5071,
+      "step": 1710
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3800688188187621,
+      "learning_rate": 9.391993109232735e-06,
+      "loss": 0.319,
+      "step": 1711
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7673761517806934,
+      "learning_rate": 9.391160734046548e-06,
+      "loss": 0.5045,
+      "step": 1712
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8010430976334053,
+      "learning_rate": 9.390327826416981e-06,
+      "loss": 0.5621,
+      "step": 1713
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4029720112938054,
+      "learning_rate": 9.38949438644503e-06,
+      "loss": 0.3123,
+      "step": 1714
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5408608792573277,
+      "learning_rate": 9.388660414231752e-06,
+      "loss": 0.5534,
+      "step": 1715
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8474978288783412,
+      "learning_rate": 9.387825909878269e-06,
+      "loss": 0.528,
+      "step": 1716
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4396046879351507,
+      "learning_rate": 9.386990873485767e-06,
+      "loss": 0.5749,
+      "step": 1717
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8445536046376834,
+      "learning_rate": 9.386155305155498e-06,
+      "loss": 0.5082,
+      "step": 1718
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.584395934616473,
+      "learning_rate": 9.385319204988777e-06,
+      "loss": 0.472,
+      "step": 1719
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8303231105492976,
+      "learning_rate": 9.384482573086983e-06,
+      "loss": 0.5475,
+      "step": 1720
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4376585950846055,
+      "learning_rate": 9.383645409551561e-06,
+      "loss": 0.5202,
+      "step": 1721
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4294644266362538,
+      "learning_rate": 9.382807714484021e-06,
+      "loss": 0.3051,
+      "step": 1722
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4215461342339995,
+      "learning_rate": 9.381969487985936e-06,
+      "loss": 0.3013,
+      "step": 1723
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.563429391321255,
+      "learning_rate": 9.381130730158944e-06,
+      "loss": 0.5385,
+      "step": 1724
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4084721893259875,
+      "learning_rate": 9.380291441104748e-06,
+      "loss": 0.5332,
+      "step": 1725
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.723667296499926,
+      "learning_rate": 9.379451620925112e-06,
+      "loss": 0.5225,
+      "step": 1726
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.125463706625784,
+      "learning_rate": 9.378611269721867e-06,
+      "loss": 0.5174,
+      "step": 1727
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9756108726607513,
+      "learning_rate": 9.377770387596911e-06,
+      "loss": 0.5577,
+      "step": 1728
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2913445505156216,
+      "learning_rate": 9.376928974652205e-06,
+      "loss": 0.5344,
+      "step": 1729
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.48191060488684356,
+      "learning_rate": 9.37608703098977e-06,
+      "loss": 0.3063,
+      "step": 1730
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1020701216689974,
+      "learning_rate": 9.375244556711696e-06,
+      "loss": 0.482,
+      "step": 1731
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1281789270198086,
+      "learning_rate": 9.374401551920135e-06,
+      "loss": 0.5269,
+      "step": 1732
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5376984093604675,
+      "learning_rate": 9.373558016717306e-06,
+      "loss": 0.5159,
+      "step": 1733
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8158501416694657,
+      "learning_rate": 9.37271395120549e-06,
+      "loss": 0.5327,
+      "step": 1734
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4156105171091714,
+      "learning_rate": 9.37186935548703e-06,
+      "loss": 0.3065,
+      "step": 1735
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.625078949477191,
+      "learning_rate": 9.371024229664342e-06,
+      "loss": 0.4958,
+      "step": 1736
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6970996436709975,
+      "learning_rate": 9.370178573839894e-06,
+      "loss": 0.5685,
+      "step": 1737
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4507721968954283,
+      "learning_rate": 9.36933238811623e-06,
+      "loss": 0.3245,
+      "step": 1738
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.567515409182905,
+      "learning_rate": 9.368485672595949e-06,
+      "loss": 0.5449,
+      "step": 1739
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.354469765088864,
+      "learning_rate": 9.36763842738172e-06,
+      "loss": 0.5211,
+      "step": 1740
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7942660467342546,
+      "learning_rate": 9.366790652576274e-06,
+      "loss": 0.509,
+      "step": 1741
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.201653572911632,
+      "learning_rate": 9.365942348282407e-06,
+      "loss": 0.5755,
+      "step": 1742
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.41787652683899806,
+      "learning_rate": 9.365093514602979e-06,
+      "loss": 0.3144,
+      "step": 1743
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2134031000725054,
+      "learning_rate": 9.364244151640913e-06,
+      "loss": 0.5436,
+      "step": 1744
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5997559495851155,
+      "learning_rate": 9.363394259499197e-06,
+      "loss": 0.5047,
+      "step": 1745
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.39876422963035585,
+      "learning_rate": 9.362543838280886e-06,
+      "loss": 0.2995,
+      "step": 1746
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.646498500645803,
+      "learning_rate": 9.361692888089094e-06,
+      "loss": 0.5617,
+      "step": 1747
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4995066317918817,
+      "learning_rate": 9.360841409027002e-06,
+      "loss": 0.5531,
+      "step": 1748
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.412781473694403,
+      "learning_rate": 9.359989401197853e-06,
+      "loss": 0.5459,
+      "step": 1749
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6684051264124427,
+      "learning_rate": 9.359136864704962e-06,
+      "loss": 0.5269,
+      "step": 1750
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.417408043560775,
+      "learning_rate": 9.358283799651694e-06,
+      "loss": 0.5487,
+      "step": 1751
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7400702883178876,
+      "learning_rate": 9.357430206141492e-06,
+      "loss": 0.5147,
+      "step": 1752
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.206405998285153,
+      "learning_rate": 9.356576084277856e-06,
+      "loss": 0.5043,
+      "step": 1753
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.337706058479607,
+      "learning_rate": 9.355721434164349e-06,
+      "loss": 0.5193,
+      "step": 1754
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9275625808027184,
+      "learning_rate": 9.354866255904602e-06,
+      "loss": 0.5161,
+      "step": 1755
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7544012022935362,
+      "learning_rate": 9.354010549602308e-06,
+      "loss": 0.515,
+      "step": 1756
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.620862050724363,
+      "learning_rate": 9.353154315361223e-06,
+      "loss": 0.4944,
+      "step": 1757
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.167826142389979,
+      "learning_rate": 9.352297553285172e-06,
+      "loss": 0.5212,
+      "step": 1758
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.843590442866372,
+      "learning_rate": 9.351440263478036e-06,
+      "loss": 0.5582,
+      "step": 1759
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.054338855146003,
+      "learning_rate": 9.350582446043768e-06,
+      "loss": 0.5209,
+      "step": 1760
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.549421354387745,
+      "learning_rate": 9.34972410108638e-06,
+      "loss": 0.5278,
+      "step": 1761
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.700429307405204,
+      "learning_rate": 9.348865228709947e-06,
+      "loss": 0.5364,
+      "step": 1762
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.311888710226101,
+      "learning_rate": 9.348005829018613e-06,
+      "loss": 0.5261,
+      "step": 1763
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.0431285074519465,
+      "learning_rate": 9.347145902116583e-06,
+      "loss": 0.5522,
+      "step": 1764
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.502393439753071,
+      "learning_rate": 9.346285448108125e-06,
+      "loss": 0.5308,
+      "step": 1765
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0714938684070985,
+      "learning_rate": 9.345424467097572e-06,
+      "loss": 0.5341,
+      "step": 1766
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.391478607882581,
+      "learning_rate": 9.344562959189321e-06,
+      "loss": 0.5391,
+      "step": 1767
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5387037888077546,
+      "learning_rate": 9.343700924487835e-06,
+      "loss": 0.4983,
+      "step": 1768
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3521105601160874,
+      "learning_rate": 9.342838363097634e-06,
+      "loss": 0.5231,
+      "step": 1769
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.204877564115765,
+      "learning_rate": 9.341975275123314e-06,
+      "loss": 0.4969,
+      "step": 1770
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.670264968021724,
+      "learning_rate": 9.34111166066952e-06,
+      "loss": 0.5039,
+      "step": 1771
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.083887150515763,
+      "learning_rate": 9.34024751984097e-06,
+      "loss": 0.553,
+      "step": 1772
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1211701986070306,
+      "learning_rate": 9.339382852742447e-06,
+      "loss": 0.5282,
+      "step": 1773
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.576446705161954,
+      "learning_rate": 9.338517659478792e-06,
+      "loss": 0.5247,
+      "step": 1774
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3683047726878708,
+      "learning_rate": 9.337651940154914e-06,
+      "loss": 0.5201,
+      "step": 1775
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.892019071247031,
+      "learning_rate": 9.336785694875785e-06,
+      "loss": 0.5697,
+      "step": 1776
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8866244415814464,
+      "learning_rate": 9.335918923746438e-06,
+      "loss": 0.4546,
+      "step": 1777
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2547663606316757,
+      "learning_rate": 9.335051626871973e-06,
+      "loss": 0.4986,
+      "step": 1778
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1264879093074334,
+      "learning_rate": 9.334183804357555e-06,
+      "loss": 0.5165,
+      "step": 1779
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9754795942770953,
+      "learning_rate": 9.333315456308407e-06,
+      "loss": 0.5107,
+      "step": 1780
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.430370529394148,
+      "learning_rate": 9.332446582829822e-06,
+      "loss": 0.4903,
+      "step": 1781
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4522001912751901,
+      "learning_rate": 9.33157718402715e-06,
+      "loss": 0.3096,
+      "step": 1782
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6935614412231734,
+      "learning_rate": 9.330707260005814e-06,
+      "loss": 0.5337,
+      "step": 1783
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9865971641518105,
+      "learning_rate": 9.329836810871291e-06,
+      "loss": 0.5147,
+      "step": 1784
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6635453538977614,
+      "learning_rate": 9.328965836729128e-06,
+      "loss": 0.5161,
+      "step": 1785
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.553700429958943,
+      "learning_rate": 9.328094337684933e-06,
+      "loss": 0.5038,
+      "step": 1786
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.279467617905753,
+      "learning_rate": 9.327222313844377e-06,
+      "loss": 0.4697,
+      "step": 1787
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.747965777652959,
+      "learning_rate": 9.326349765313199e-06,
+      "loss": 0.5615,
+      "step": 1788
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.40035494584664394,
+      "learning_rate": 9.325476692197197e-06,
+      "loss": 0.304,
+      "step": 1789
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.367195936297447,
+      "learning_rate": 9.324603094602232e-06,
+      "loss": 0.4837,
+      "step": 1790
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6183992269256495,
+      "learning_rate": 9.323728972634234e-06,
+      "loss": 0.5644,
+      "step": 1791
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9971382840000445,
+      "learning_rate": 9.322854326399192e-06,
+      "loss": 0.5459,
+      "step": 1792
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.777990461445419,
+      "learning_rate": 9.32197915600316e-06,
+      "loss": 0.4963,
+      "step": 1793
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.758563714525283,
+      "learning_rate": 9.321103461552254e-06,
+      "loss": 0.4992,
+      "step": 1794
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.629365938150127,
+      "learning_rate": 9.320227243152657e-06,
+      "loss": 0.5413,
+      "step": 1795
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.170809162284872,
+      "learning_rate": 9.319350500910613e-06,
+      "loss": 0.5717,
+      "step": 1796
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.854863148403779,
+      "learning_rate": 9.318473234932428e-06,
+      "loss": 0.5199,
+      "step": 1797
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.110645460561053,
+      "learning_rate": 9.317595445324477e-06,
+      "loss": 0.4284,
+      "step": 1798
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7632807867473907,
+      "learning_rate": 9.316717132193193e-06,
+      "loss": 0.5379,
+      "step": 1799
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.528325980282886,
+      "learning_rate": 9.315838295645074e-06,
+      "loss": 0.5525,
+      "step": 1800
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3042804559881045,
+      "learning_rate": 9.314958935786684e-06,
+      "loss": 0.5356,
+      "step": 1801
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.794167562608316,
+      "learning_rate": 9.314079052724645e-06,
+      "loss": 0.5327,
+      "step": 1802
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.619179804542359,
+      "learning_rate": 9.313198646565648e-06,
+      "loss": 0.5429,
+      "step": 1803
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3745544952054036,
+      "learning_rate": 9.312317717416448e-06,
+      "loss": 0.5336,
+      "step": 1804
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1819813385129256,
+      "learning_rate": 9.311436265383856e-06,
+      "loss": 0.5212,
+      "step": 1805
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4032309037748925,
+      "learning_rate": 9.310554290574754e-06,
+      "loss": 0.4835,
+      "step": 1806
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7747377495740735,
+      "learning_rate": 9.309671793096082e-06,
+      "loss": 0.5574,
+      "step": 1807
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.391035141436964,
+      "learning_rate": 9.308788773054848e-06,
+      "loss": 0.5389,
+      "step": 1808
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.287838957023062,
+      "learning_rate": 9.307905230558123e-06,
+      "loss": 0.5198,
+      "step": 1809
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4380818904951786,
+      "learning_rate": 9.307021165713034e-06,
+      "loss": 0.3229,
+      "step": 1810
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4503771810670606,
+      "learning_rate": 9.30613657862678e-06,
+      "loss": 0.3159,
+      "step": 1811
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.14973317117769,
+      "learning_rate": 9.305251469406621e-06,
+      "loss": 0.5157,
+      "step": 1812
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.365496288932613,
+      "learning_rate": 9.304365838159882e-06,
+      "loss": 0.5526,
+      "step": 1813
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6578023090778666,
+      "learning_rate": 9.303479684993943e-06,
+      "loss": 0.5447,
+      "step": 1814
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8485532795958446,
+      "learning_rate": 9.302593010016255e-06,
+      "loss": 0.5283,
+      "step": 1815
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.023654331553488,
+      "learning_rate": 9.301705813334332e-06,
+      "loss": 0.484,
+      "step": 1816
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1434654307001386,
+      "learning_rate": 9.30081809505575e-06,
+      "loss": 0.5367,
+      "step": 1817
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.018265455481575,
+      "learning_rate": 9.299929855288145e-06,
+      "loss": 0.4981,
+      "step": 1818
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1504580299617087,
+      "learning_rate": 9.299041094139222e-06,
+      "loss": 0.5289,
+      "step": 1819
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.584973845772277,
+      "learning_rate": 9.298151811716745e-06,
+      "loss": 0.5285,
+      "step": 1820
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.2968029924007825,
+      "learning_rate": 9.297262008128544e-06,
+      "loss": 0.5235,
+      "step": 1821
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2677576575699163,
+      "learning_rate": 9.296371683482508e-06,
+      "loss": 0.5149,
+      "step": 1822
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6995784016033606,
+      "learning_rate": 9.295480837886595e-06,
+      "loss": 0.5234,
+      "step": 1823
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.901629207576652,
+      "learning_rate": 9.294589471448819e-06,
+      "loss": 0.5454,
+      "step": 1824
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8228860217196,
+      "learning_rate": 9.293697584277266e-06,
+      "loss": 0.4999,
+      "step": 1825
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.296719492547129,
+      "learning_rate": 9.292805176480077e-06,
+      "loss": 0.558,
+      "step": 1826
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4698314285991687,
+      "learning_rate": 9.291912248165462e-06,
+      "loss": 0.572,
+      "step": 1827
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.06837692618524,
+      "learning_rate": 9.291018799441692e-06,
+      "loss": 0.5216,
+      "step": 1828
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.65783162598546,
+      "learning_rate": 9.290124830417097e-06,
+      "loss": 0.5115,
+      "step": 1829
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.76117155858484,
+      "learning_rate": 9.289230341200075e-06,
+      "loss": 0.5236,
+      "step": 1830
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5543340054141089,
+      "learning_rate": 9.288335331899089e-06,
+      "loss": 0.3277,
+      "step": 1831
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.893238778653382,
+      "learning_rate": 9.28743980262266e-06,
+      "loss": 0.5703,
+      "step": 1832
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9409480386375946,
+      "learning_rate": 9.286543753479372e-06,
+      "loss": 0.5561,
+      "step": 1833
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.38082204557836,
+      "learning_rate": 9.285647184577878e-06,
+      "loss": 0.5324,
+      "step": 1834
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.077018577375055,
+      "learning_rate": 9.284750096026886e-06,
+      "loss": 0.5231,
+      "step": 1835
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.1518075597155635,
+      "learning_rate": 9.283852487935174e-06,
+      "loss": 0.4699,
+      "step": 1836
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.049409990842314,
+      "learning_rate": 9.282954360411579e-06,
+      "loss": 0.5394,
+      "step": 1837
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.64239414987076,
+      "learning_rate": 9.282055713565e-06,
+      "loss": 0.5271,
+      "step": 1838
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.147668449240302,
+      "learning_rate": 9.28115654750441e-06,
+      "loss": 0.5946,
+      "step": 1839
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1170640235358946,
+      "learning_rate": 9.280256862338822e-06,
+      "loss": 0.4696,
+      "step": 1840
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.216714462464971,
+      "learning_rate": 9.279356658177337e-06,
+      "loss": 0.547,
+      "step": 1841
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5684591710347816,
+      "learning_rate": 9.278455935129103e-06,
+      "loss": 0.307,
+      "step": 1842
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2661320337685167,
+      "learning_rate": 9.277554693303337e-06,
+      "loss": 0.537,
+      "step": 1843
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.006638577764285,
+      "learning_rate": 9.276652932809315e-06,
+      "loss": 0.5344,
+      "step": 1844
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.733125962282889,
+      "learning_rate": 9.275750653756385e-06,
+      "loss": 0.5594,
+      "step": 1845
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.134218151512563,
+      "learning_rate": 9.274847856253946e-06,
+      "loss": 0.5246,
+      "step": 1846
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.46210530468347,
+      "learning_rate": 9.273944540411465e-06,
+      "loss": 0.5308,
+      "step": 1847
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.237029332140171,
+      "learning_rate": 9.273040706338476e-06,
+      "loss": 0.5178,
+      "step": 1848
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1479237287693485,
+      "learning_rate": 9.272136354144569e-06,
+      "loss": 0.5573,
+      "step": 1849
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.349003197576602,
+      "learning_rate": 9.271231483939403e-06,
+      "loss": 0.5308,
+      "step": 1850
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.703101137042431,
+      "learning_rate": 9.27032609583269e-06,
+      "loss": 0.5544,
+      "step": 1851
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3797214155355997,
+      "learning_rate": 9.269420189934219e-06,
+      "loss": 0.5767,
+      "step": 1852
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.234579437518564,
+      "learning_rate": 9.26851376635383e-06,
+      "loss": 0.5853,
+      "step": 1853
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.81065725497869,
+      "learning_rate": 9.267606825201433e-06,
+      "loss": 0.5125,
+      "step": 1854
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.446884385461072,
+      "learning_rate": 9.266699366586992e-06,
+      "loss": 0.5578,
+      "step": 1855
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.666663825430134,
+      "learning_rate": 9.265791390620547e-06,
+      "loss": 0.4923,
+      "step": 1856
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.460646670151559,
+      "learning_rate": 9.26488289741219e-06,
+      "loss": 0.5089,
+      "step": 1857
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.668095874605047,
+      "learning_rate": 9.263973887072074e-06,
+      "loss": 0.5184,
+      "step": 1858
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3104650684932513,
+      "learning_rate": 9.263064359710428e-06,
+      "loss": 0.4823,
+      "step": 1859
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4692947576837628,
+      "learning_rate": 9.262154315437528e-06,
+      "loss": 0.31,
+      "step": 1860
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.031576602123491,
+      "learning_rate": 9.261243754363726e-06,
+      "loss": 0.5689,
+      "step": 1861
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.415187440656014,
+      "learning_rate": 9.260332676599426e-06,
+      "loss": 0.322,
+      "step": 1862
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.649993872455281,
+      "learning_rate": 9.259421082255104e-06,
+      "loss": 0.5183,
+      "step": 1863
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.435348436543113,
+      "learning_rate": 9.258508971441289e-06,
+      "loss": 0.5223,
+      "step": 1864
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8058754893962856,
+      "learning_rate": 9.25759634426858e-06,
+      "loss": 0.4807,
+      "step": 1865
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6449122687950193,
+      "learning_rate": 9.256683200847638e-06,
+      "loss": 0.5199,
+      "step": 1866
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.440038704146631,
+      "learning_rate": 9.255769541289182e-06,
+      "loss": 0.5214,
+      "step": 1867
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.834925025529794,
+      "learning_rate": 9.254855365703997e-06,
+      "loss": 0.527,
+      "step": 1868
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.510337640882862,
+      "learning_rate": 9.253940674202931e-06,
+      "loss": 0.5068,
+      "step": 1869
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.764048464832035,
+      "learning_rate": 9.25302546689689e-06,
+      "loss": 0.5497,
+      "step": 1870
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.604738404937431,
+      "learning_rate": 9.25210974389685e-06,
+      "loss": 0.543,
+      "step": 1871
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4674086951726357,
+      "learning_rate": 9.251193505313845e-06,
+      "loss": 0.5251,
+      "step": 1872
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.521340916220292,
+      "learning_rate": 9.250276751258972e-06,
+      "loss": 0.5328,
+      "step": 1873
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6407783777514307,
+      "learning_rate": 9.249359481843389e-06,
+      "loss": 0.5517,
+      "step": 1874
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.5566996588790465,
+      "learning_rate": 9.24844169717832e-06,
+      "loss": 0.5894,
+      "step": 1875
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3562643192293353,
+      "learning_rate": 9.247523397375047e-06,
+      "loss": 0.597,
+      "step": 1876
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.389203950779508,
+      "learning_rate": 9.24660458254492e-06,
+      "loss": 0.5742,
+      "step": 1877
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.604076029443648,
+      "learning_rate": 9.245685252799346e-06,
+      "loss": 0.3223,
+      "step": 1878
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.342488167063877,
+      "learning_rate": 9.244765408249798e-06,
+      "loss": 0.5225,
+      "step": 1879
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.535381620838086,
+      "learning_rate": 9.243845049007811e-06,
+      "loss": 0.5709,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9741455036667277,
+      "learning_rate": 9.242924175184982e-06,
+      "loss": 0.5241,
+      "step": 1881
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0396950422504827,
+      "learning_rate": 9.242002786892967e-06,
+      "loss": 0.5236,
+      "step": 1882
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5083820189751784,
+      "learning_rate": 9.241080884243492e-06,
+      "loss": 0.5537,
+      "step": 1883
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.63614080371913,
+      "learning_rate": 9.240158467348337e-06,
+      "loss": 0.4883,
+      "step": 1884
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5945216453980415,
+      "learning_rate": 9.239235536319351e-06,
+      "loss": 0.5493,
+      "step": 1885
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.51655475356588,
+      "learning_rate": 9.23831209126844e-06,
+      "loss": 0.2993,
+      "step": 1886
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.432397088154824,
+      "learning_rate": 9.237388132307576e-06,
+      "loss": 0.5285,
+      "step": 1887
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.45111186094335504,
+      "learning_rate": 9.236463659548793e-06,
+      "loss": 0.3098,
+      "step": 1888
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.476623506996522,
+      "learning_rate": 9.235538673104187e-06,
+      "loss": 0.5678,
+      "step": 1889
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8669791971385976,
+      "learning_rate": 9.234613173085913e-06,
+      "loss": 0.5069,
+      "step": 1890
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5475398892508383,
+      "learning_rate": 9.233687159606195e-06,
+      "loss": 0.5629,
+      "step": 1891
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6007441289845494,
+      "learning_rate": 9.232760632777311e-06,
+      "loss": 0.5205,
+      "step": 1892
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.530207424436768,
+      "learning_rate": 9.23183359271161e-06,
+      "loss": 0.5282,
+      "step": 1893
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4767607319459197,
+      "learning_rate": 9.230906039521495e-06,
+      "loss": 0.3043,
+      "step": 1894
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.095940542021617,
+      "learning_rate": 9.229977973319436e-06,
+      "loss": 0.5487,
+      "step": 1895
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.52521576757823,
+      "learning_rate": 9.229049394217966e-06,
+      "loss": 0.5231,
+      "step": 1896
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.966621638422822,
+      "learning_rate": 9.228120302329678e-06,
+      "loss": 0.5398,
+      "step": 1897
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0437453348937273,
+      "learning_rate": 9.227190697767224e-06,
+      "loss": 0.563,
+      "step": 1898
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.705556619159484,
+      "learning_rate": 9.226260580643327e-06,
+      "loss": 0.5654,
+      "step": 1899
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.008331857199472,
+      "learning_rate": 9.225329951070762e-06,
+      "loss": 0.5074,
+      "step": 1900
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3662960075154102,
+      "learning_rate": 9.224398809162376e-06,
+      "loss": 0.5142,
+      "step": 1901
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3179507638839683,
+      "learning_rate": 9.22346715503107e-06,
+      "loss": 0.5273,
+      "step": 1902
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7860297525203537,
+      "learning_rate": 9.222534988789811e-06,
+      "loss": 0.5899,
+      "step": 1903
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.109067349631166,
+      "learning_rate": 9.221602310551627e-06,
+      "loss": 0.5523,
+      "step": 1904
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.31420572228397,
+      "learning_rate": 9.220669120429608e-06,
+      "loss": 0.5855,
+      "step": 1905
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.2329389791687495,
+      "learning_rate": 9.219735418536907e-06,
+      "loss": 0.4992,
+      "step": 1906
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.351185603029692,
+      "learning_rate": 9.21880120498674e-06,
+      "loss": 0.5245,
+      "step": 1907
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.616559353459695,
+      "learning_rate": 9.217866479892383e-06,
+      "loss": 0.5349,
+      "step": 1908
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.927481214457228,
+      "learning_rate": 9.216931243367173e-06,
+      "loss": 0.5745,
+      "step": 1909
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7811617487394127,
+      "learning_rate": 9.215995495524512e-06,
+      "loss": 0.5059,
+      "step": 1910
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.43200772204939325,
+      "learning_rate": 9.215059236477864e-06,
+      "loss": 0.2997,
+      "step": 1911
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.071230326671535,
+      "learning_rate": 9.21412246634075e-06,
+      "loss": 0.5281,
+      "step": 1912
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.066911166275264,
+      "learning_rate": 9.21318518522676e-06,
+      "loss": 0.5665,
+      "step": 1913
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2721412232497653,
+      "learning_rate": 9.21224739324954e-06,
+      "loss": 0.522,
+      "step": 1914
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7499087566423324,
+      "learning_rate": 9.211309090522803e-06,
+      "loss": 0.5148,
+      "step": 1915
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.152461875318631,
+      "learning_rate": 9.21037027716032e-06,
+      "loss": 0.491,
+      "step": 1916
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5354245692165285,
+      "learning_rate": 9.209430953275925e-06,
+      "loss": 0.5191,
+      "step": 1917
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.751741438792439,
+      "learning_rate": 9.208491118983515e-06,
+      "loss": 0.5344,
+      "step": 1918
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.073710319253984,
+      "learning_rate": 9.207550774397048e-06,
+      "loss": 0.521,
+      "step": 1919
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.443121075153324,
+      "learning_rate": 9.206609919630543e-06,
+      "loss": 0.4446,
+      "step": 1920
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8184562349335556,
+      "learning_rate": 9.205668554798084e-06,
+      "loss": 0.5234,
+      "step": 1921
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4180018678537256,
+      "learning_rate": 9.204726680013813e-06,
+      "loss": 0.5236,
+      "step": 1922
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9526051838591707,
+      "learning_rate": 9.203784295391938e-06,
+      "loss": 0.5054,
+      "step": 1923
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.332185896104056,
+      "learning_rate": 9.202841401046722e-06,
+      "loss": 0.4926,
+      "step": 1924
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5328279601384027,
+      "learning_rate": 9.201897997092497e-06,
+      "loss": 0.5655,
+      "step": 1925
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.42675114848006956,
+      "learning_rate": 9.200954083643654e-06,
+      "loss": 0.3036,
+      "step": 1926
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.03719238183766,
+      "learning_rate": 9.200009660814646e-06,
+      "loss": 0.5389,
+      "step": 1927
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8367882554067703,
+      "learning_rate": 9.199064728719988e-06,
+      "loss": 0.5358,
+      "step": 1928
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7313582804775365,
+      "learning_rate": 9.198119287474254e-06,
+      "loss": 0.5197,
+      "step": 1929
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.446774862967875,
+      "learning_rate": 9.197173337192082e-06,
+      "loss": 0.5502,
+      "step": 1930
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.899094571473546,
+      "learning_rate": 9.196226877988174e-06,
+      "loss": 0.5335,
+      "step": 1931
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.901178921025131,
+      "learning_rate": 9.195279909977293e-06,
+      "loss": 0.479,
+      "step": 1932
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.448935433461287,
+      "learning_rate": 9.194332433274256e-06,
+      "loss": 0.5085,
+      "step": 1933
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.118075361360932,
+      "learning_rate": 9.193384447993954e-06,
+      "loss": 0.4799,
+      "step": 1934
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0375463157925537,
+      "learning_rate": 9.192435954251329e-06,
+      "loss": 0.5014,
+      "step": 1935
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.712198499361907,
+      "learning_rate": 9.191486952161393e-06,
+      "loss": 0.5011,
+      "step": 1936
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.150547685015238,
+      "learning_rate": 9.190537441839212e-06,
+      "loss": 0.4918,
+      "step": 1937
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.127731694780296,
+      "learning_rate": 9.189587423399919e-06,
+      "loss": 0.5649,
+      "step": 1938
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.246493539219528,
+      "learning_rate": 9.188636896958708e-06,
+      "loss": 0.5293,
+      "step": 1939
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8973976973856215,
+      "learning_rate": 9.187685862630833e-06,
+      "loss": 0.5076,
+      "step": 1940
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.489393309674177,
+      "learning_rate": 9.18673432053161e-06,
+      "loss": 0.5486,
+      "step": 1941
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.428417646070231,
+      "learning_rate": 9.185782270776416e-06,
+      "loss": 0.5144,
+      "step": 1942
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.258987921143218,
+      "learning_rate": 9.184829713480691e-06,
+      "loss": 0.5779,
+      "step": 1943
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.915229752905856,
+      "learning_rate": 9.183876648759937e-06,
+      "loss": 0.5117,
+      "step": 1944
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.698754650351194,
+      "learning_rate": 9.182923076729716e-06,
+      "loss": 0.5765,
+      "step": 1945
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8401117023916296,
+      "learning_rate": 9.181968997505649e-06,
+      "loss": 0.5624,
+      "step": 1946
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.442355792267787,
+      "learning_rate": 9.181014411203426e-06,
+      "loss": 0.4837,
+      "step": 1947
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5209123009318914,
+      "learning_rate": 9.18005931793879e-06,
+      "loss": 0.4781,
+      "step": 1948
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3121601264296685,
+      "learning_rate": 9.179103717827551e-06,
+      "loss": 0.4876,
+      "step": 1949
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3626209622252055,
+      "learning_rate": 9.178147610985578e-06,
+      "loss": 0.4619,
+      "step": 1950
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4351949701293507,
+      "learning_rate": 9.177190997528805e-06,
+      "loss": 0.5038,
+      "step": 1951
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.246257015012771,
+      "learning_rate": 9.176233877573219e-06,
+      "loss": 0.5231,
+      "step": 1952
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.799737887792141,
+      "learning_rate": 9.175276251234881e-06,
+      "loss": 0.4959,
+      "step": 1953
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.947995024727063,
+      "learning_rate": 9.1743181186299e-06,
+      "loss": 0.547,
+      "step": 1954
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.423955181751932,
+      "learning_rate": 9.17335947987446e-06,
+      "loss": 0.4921,
+      "step": 1955
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.496902684798367,
+      "learning_rate": 9.172400335084793e-06,
+      "loss": 0.5263,
+      "step": 1956
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5836300947414763,
+      "learning_rate": 9.171440684377201e-06,
+      "loss": 0.5126,
+      "step": 1957
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4198771785702187,
+      "learning_rate": 9.170480527868046e-06,
+      "loss": 0.4913,
+      "step": 1958
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.963177170936722,
+      "learning_rate": 9.169519865673748e-06,
+      "loss": 0.5512,
+      "step": 1959
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.226506655180296,
+      "learning_rate": 9.168558697910792e-06,
+      "loss": 0.5406,
+      "step": 1960
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.99385833022134,
+      "learning_rate": 9.167597024695722e-06,
+      "loss": 0.5753,
+      "step": 1961
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.983809801847899,
+      "learning_rate": 9.166634846145146e-06,
+      "loss": 0.5312,
+      "step": 1962
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.541352112007478,
+      "learning_rate": 9.165672162375731e-06,
+      "loss": 0.4411,
+      "step": 1963
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8641734514337904,
+      "learning_rate": 9.164708973504203e-06,
+      "loss": 0.5506,
+      "step": 1964
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.717323643712476,
+      "learning_rate": 9.163745279647356e-06,
+      "loss": 0.5384,
+      "step": 1965
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.45603060626745057,
+      "learning_rate": 9.162781080922039e-06,
+      "loss": 0.3162,
+      "step": 1966
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.436666477329605,
+      "learning_rate": 9.161816377445164e-06,
+      "loss": 0.5234,
+      "step": 1967
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8262011701621073,
+      "learning_rate": 9.160851169333704e-06,
+      "loss": 0.5552,
+      "step": 1968
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.158030470880666,
+      "learning_rate": 9.159885456704696e-06,
+      "loss": 0.5083,
+      "step": 1969
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.423713276195243,
+      "learning_rate": 9.158919239675237e-06,
+      "loss": 0.3096,
+      "step": 1970
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.213173543861646,
+      "learning_rate": 9.157952518362478e-06,
+      "loss": 0.5142,
+      "step": 1971
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.917477205609795,
+      "learning_rate": 9.156985292883645e-06,
+      "loss": 0.5369,
+      "step": 1972
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.611199526220166,
+      "learning_rate": 9.156017563356013e-06,
+      "loss": 0.5365,
+      "step": 1973
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1586453421070515,
+      "learning_rate": 9.155049329896924e-06,
+      "loss": 0.5131,
+      "step": 1974
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6964750408456246,
+      "learning_rate": 9.154080592623777e-06,
+      "loss": 0.5136,
+      "step": 1975
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.1015913798632395,
+      "learning_rate": 9.15311135165404e-06,
+      "loss": 0.5402,
+      "step": 1976
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.294863864669501,
+      "learning_rate": 9.152141607105231e-06,
+      "loss": 0.5915,
+      "step": 1977
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.45960629036268,
+      "learning_rate": 9.15117135909494e-06,
+      "loss": 0.5322,
+      "step": 1978
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4919856724746057,
+      "learning_rate": 9.15020060774081e-06,
+      "loss": 0.3157,
+      "step": 1979
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.816920503983882,
+      "learning_rate": 9.149229353160546e-06,
+      "loss": 0.5057,
+      "step": 1980
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.317972382821018,
+      "learning_rate": 9.14825759547192e-06,
+      "loss": 0.502,
+      "step": 1981
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.33275919598886,
+      "learning_rate": 9.14728533479276e-06,
+      "loss": 0.5054,
+      "step": 1982
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.37975713020830126,
+      "learning_rate": 9.146312571240955e-06,
+      "loss": 0.2939,
+      "step": 1983
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2432041799734734,
+      "learning_rate": 9.145339304934453e-06,
+      "loss": 0.5041,
+      "step": 1984
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9476844944963467,
+      "learning_rate": 9.144365535991273e-06,
+      "loss": 0.496,
+      "step": 1985
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.436348340380141,
+      "learning_rate": 9.143391264529482e-06,
+      "loss": 0.5852,
+      "step": 1986
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.819981713644153,
+      "learning_rate": 9.142416490667217e-06,
+      "loss": 0.4866,
+      "step": 1987
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.607938164874637,
+      "learning_rate": 9.14144121452267e-06,
+      "loss": 0.5095,
+      "step": 1988
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.286104689560544,
+      "learning_rate": 9.1404654362141e-06,
+      "loss": 0.5284,
+      "step": 1989
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7347929283646986,
+      "learning_rate": 9.13948915585982e-06,
+      "loss": 0.5248,
+      "step": 1990
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6216480700212172,
+      "learning_rate": 9.13851237357821e-06,
+      "loss": 0.5509,
+      "step": 1991
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4448338826013744,
+      "learning_rate": 9.137535089487705e-06,
+      "loss": 0.295,
+      "step": 1992
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.134693000687839,
+      "learning_rate": 9.13655730370681e-06,
+      "loss": 0.4972,
+      "step": 1993
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7968259695870565,
+      "learning_rate": 9.135579016354077e-06,
+      "loss": 0.5179,
+      "step": 1994
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5394154302046714,
+      "learning_rate": 9.134600227548133e-06,
+      "loss": 0.5184,
+      "step": 1995
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4014295554634114,
+      "learning_rate": 9.133620937407656e-06,
+      "loss": 0.3123,
+      "step": 1996
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6969578750443124,
+      "learning_rate": 9.13264114605139e-06,
+      "loss": 0.5048,
+      "step": 1997
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7092285909156146,
+      "learning_rate": 9.131660853598139e-06,
+      "loss": 0.5234,
+      "step": 1998
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.654080483079467,
+      "learning_rate": 9.130680060166764e-06,
+      "loss": 0.5115,
+      "step": 1999
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.20737470600392,
+      "learning_rate": 9.129698765876191e-06,
+      "loss": 0.5265,
+      "step": 2000
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.95773105448671,
+      "learning_rate": 9.128716970845407e-06,
+      "loss": 0.5395,
+      "step": 2001
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.027027674191432,
+      "learning_rate": 9.127734675193454e-06,
+      "loss": 0.5282,
+      "step": 2002
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.918391907280234,
+      "learning_rate": 9.126751879039442e-06,
+      "loss": 0.4926,
+      "step": 2003
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.075300919304936,
+      "learning_rate": 9.12576858250254e-06,
+      "loss": 0.5342,
+      "step": 2004
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.48739744507350224,
+      "learning_rate": 9.12478478570197e-06,
+      "loss": 0.3117,
+      "step": 2005
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.247249395190533,
+      "learning_rate": 9.123800488757027e-06,
+      "loss": 0.5677,
+      "step": 2006
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2261195846180764,
+      "learning_rate": 9.122815691787056e-06,
+      "loss": 0.5046,
+      "step": 2007
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6930847551777624,
+      "learning_rate": 9.121830394911471e-06,
+      "loss": 0.473,
+      "step": 2008
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3456321358860044,
+      "learning_rate": 9.120844598249738e-06,
+      "loss": 0.5488,
+      "step": 2009
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.608394577065099,
+      "learning_rate": 9.119858301921391e-06,
+      "loss": 0.5268,
+      "step": 2010
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9224658070445475,
+      "learning_rate": 9.118871506046025e-06,
+      "loss": 0.4995,
+      "step": 2011
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.690971442129256,
+      "learning_rate": 9.117884210743287e-06,
+      "loss": 0.5316,
+      "step": 2012
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.688252349434401,
+      "learning_rate": 9.11689641613289e-06,
+      "loss": 0.5211,
+      "step": 2013
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.40062133096620134,
+      "learning_rate": 9.115908122334611e-06,
+      "loss": 0.3021,
+      "step": 2014
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.709757893236738,
+      "learning_rate": 9.114919329468283e-06,
+      "loss": 0.5403,
+      "step": 2015
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.190592722553825,
+      "learning_rate": 9.113930037653799e-06,
+      "loss": 0.5761,
+      "step": 2016
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.002061672260471,
+      "learning_rate": 9.112940247011116e-06,
+      "loss": 0.5429,
+      "step": 2017
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.231303882227882,
+      "learning_rate": 9.111949957660248e-06,
+      "loss": 0.5334,
+      "step": 2018
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5893286299867877,
+      "learning_rate": 9.110959169721272e-06,
+      "loss": 0.5039,
+      "step": 2019
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.60251352861136,
+      "learning_rate": 9.109967883314323e-06,
+      "loss": 0.5448,
+      "step": 2020
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.548888867866416,
+      "learning_rate": 9.108976098559601e-06,
+      "loss": 0.4888,
+      "step": 2021
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.771557655006352,
+      "learning_rate": 9.107983815577359e-06,
+      "loss": 0.4686,
+      "step": 2022
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.617674976115878,
+      "learning_rate": 9.106991034487917e-06,
+      "loss": 0.4792,
+      "step": 2023
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.7081550954321045,
+      "learning_rate": 9.105997755411656e-06,
+      "loss": 0.5482,
+      "step": 2024
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.764016906115595,
+      "learning_rate": 9.10500397846901e-06,
+      "loss": 0.4566,
+      "step": 2025
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7034188043082046,
+      "learning_rate": 9.104009703780478e-06,
+      "loss": 0.5471,
+      "step": 2026
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2563785273134904,
+      "learning_rate": 9.103014931466622e-06,
+      "loss": 0.5886,
+      "step": 2027
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9190972103442605,
+      "learning_rate": 9.102019661648061e-06,
+      "loss": 0.5449,
+      "step": 2028
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4377951546699324,
+      "learning_rate": 9.101023894445474e-06,
+      "loss": 0.4774,
+      "step": 2029
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3698385997783453,
+      "learning_rate": 9.100027629979599e-06,
+      "loss": 0.475,
+      "step": 2030
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.033680767652265,
+      "learning_rate": 9.099030868371241e-06,
+      "loss": 0.5635,
+      "step": 2031
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.982814301280672,
+      "learning_rate": 9.09803360974126e-06,
+      "loss": 0.5202,
+      "step": 2032
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5705077801054617,
+      "learning_rate": 9.097035854210574e-06,
+      "loss": 0.4866,
+      "step": 2033
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.29919836680443,
+      "learning_rate": 9.096037601900166e-06,
+      "loss": 0.595,
+      "step": 2034
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5491045033003834,
+      "learning_rate": 9.095038852931077e-06,
+      "loss": 0.5116,
+      "step": 2035
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.883023866167453,
+      "learning_rate": 9.09403960742441e-06,
+      "loss": 0.5154,
+      "step": 2036
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.515294009974298,
+      "learning_rate": 9.093039865501328e-06,
+      "loss": 0.5066,
+      "step": 2037
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.406723714300111,
+      "learning_rate": 9.09203962728305e-06,
+      "loss": 0.5272,
+      "step": 2038
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.515359548334093,
+      "learning_rate": 9.09103889289086e-06,
+      "loss": 0.4636,
+      "step": 2039
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1272810407496445,
+      "learning_rate": 9.0900376624461e-06,
+      "loss": 0.4753,
+      "step": 2040
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.34061349084811,
+      "learning_rate": 9.089035936070172e-06,
+      "loss": 0.5529,
+      "step": 2041
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9162196819444466,
+      "learning_rate": 9.08803371388454e-06,
+      "loss": 0.5137,
+      "step": 2042
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2440100796597475,
+      "learning_rate": 9.087030996010728e-06,
+      "loss": 0.4998,
+      "step": 2043
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.012046371264383,
+      "learning_rate": 9.086027782570318e-06,
+      "loss": 0.5435,
+      "step": 2044
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.352717654292368,
+      "learning_rate": 9.085024073684952e-06,
+      "loss": 0.4873,
+      "step": 2045
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.323101858701656,
+      "learning_rate": 9.084019869476332e-06,
+      "loss": 0.5606,
+      "step": 2046
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.0368830046719,
+      "learning_rate": 9.083015170066225e-06,
+      "loss": 0.4954,
+      "step": 2047
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.93448413167972,
+      "learning_rate": 9.082009975576452e-06,
+      "loss": 0.4851,
+      "step": 2048
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6062959851071543,
+      "learning_rate": 9.081004286128896e-06,
+      "loss": 0.5642,
+      "step": 2049
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6944140045633715,
+      "learning_rate": 9.079998101845502e-06,
+      "loss": 0.5011,
+      "step": 2050
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.250508745784272,
+      "learning_rate": 9.07899142284827e-06,
+      "loss": 0.5459,
+      "step": 2051
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0755371722453435,
+      "learning_rate": 9.077984249259268e-06,
+      "loss": 0.5403,
+      "step": 2052
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.167154305092705,
+      "learning_rate": 9.076976581200616e-06,
+      "loss": 0.4927,
+      "step": 2053
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.9946472738544987,
+      "learning_rate": 9.075968418794499e-06,
+      "loss": 0.5623,
+      "step": 2054
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4158766042866198,
+      "learning_rate": 9.074959762163158e-06,
+      "loss": 0.4987,
+      "step": 2055
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0486258434896563,
+      "learning_rate": 9.073950611428896e-06,
+      "loss": 0.5348,
+      "step": 2056
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.300395179005123,
+      "learning_rate": 9.07294096671408e-06,
+      "loss": 0.5325,
+      "step": 2057
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2701997224299126,
+      "learning_rate": 9.071930828141128e-06,
+      "loss": 0.471,
+      "step": 2058
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.446956550756286,
+      "learning_rate": 9.070920195832528e-06,
+      "loss": 0.479,
+      "step": 2059
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.841060311861039,
+      "learning_rate": 9.069909069910817e-06,
+      "loss": 0.5658,
+      "step": 2060
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 14.508131558293494,
+      "learning_rate": 9.068897450498602e-06,
+      "loss": 0.526,
+      "step": 2061
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6529428159582187,
+      "learning_rate": 9.067885337718542e-06,
+      "loss": 0.5063,
+      "step": 2062
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6409406189457743,
+      "learning_rate": 9.066872731693362e-06,
+      "loss": 0.5199,
+      "step": 2063
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.62292495684176,
+      "learning_rate": 9.065859632545842e-06,
+      "loss": 0.5073,
+      "step": 2064
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.349125448037169,
+      "learning_rate": 9.064846040398822e-06,
+      "loss": 0.5266,
+      "step": 2065
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.763339202929344,
+      "learning_rate": 9.06383195537521e-06,
+      "loss": 0.52,
+      "step": 2066
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.054399671266357,
+      "learning_rate": 9.062817377597961e-06,
+      "loss": 0.5084,
+      "step": 2067
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.928155574843515,
+      "learning_rate": 9.061802307190099e-06,
+      "loss": 0.477,
+      "step": 2068
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5403302362039244,
+      "learning_rate": 9.060786744274703e-06,
+      "loss": 0.5069,
+      "step": 2069
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8656992488935042,
+      "learning_rate": 9.059770688974916e-06,
+      "loss": 0.5234,
+      "step": 2070
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.745120031201491,
+      "learning_rate": 9.058754141413936e-06,
+      "loss": 0.5385,
+      "step": 2071
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.076835108455287,
+      "learning_rate": 9.057737101715024e-06,
+      "loss": 0.5228,
+      "step": 2072
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.36699756111236,
+      "learning_rate": 9.056719570001499e-06,
+      "loss": 0.538,
+      "step": 2073
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.065186402123986,
+      "learning_rate": 9.05570154639674e-06,
+      "loss": 0.4948,
+      "step": 2074
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0886676149418717,
+      "learning_rate": 9.054683031024188e-06,
+      "loss": 0.5419,
+      "step": 2075
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.715178345503503,
+      "learning_rate": 9.053664024007339e-06,
+      "loss": 0.5146,
+      "step": 2076
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8787922959935033,
+      "learning_rate": 9.052644525469751e-06,
+      "loss": 0.5264,
+      "step": 2077
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6095849467094605,
+      "learning_rate": 9.051624535535043e-06,
+      "loss": 0.5616,
+      "step": 2078
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.756323951110512,
+      "learning_rate": 9.050604054326893e-06,
+      "loss": 0.4888,
+      "step": 2079
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.75025738890459,
+      "learning_rate": 9.049583081969038e-06,
+      "loss": 0.5468,
+      "step": 2080
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3684717704886005,
+      "learning_rate": 9.04856161858527e-06,
+      "loss": 0.5229,
+      "step": 2081
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4637960884605303,
+      "learning_rate": 9.04753966429945e-06,
+      "loss": 0.4679,
+      "step": 2082
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3834259424766355,
+      "learning_rate": 9.046517219235492e-06,
+      "loss": 0.3046,
+      "step": 2083
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7236969654117194,
+      "learning_rate": 9.045494283517371e-06,
+      "loss": 0.4959,
+      "step": 2084
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.076170779573408,
+      "learning_rate": 9.044470857269121e-06,
+      "loss": 0.5423,
+      "step": 2085
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0155810673721155,
+      "learning_rate": 9.043446940614834e-06,
+      "loss": 0.5031,
+      "step": 2086
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.426342823598323,
+      "learning_rate": 9.042422533678667e-06,
+      "loss": 0.4813,
+      "step": 2087
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.102291799004299,
+      "learning_rate": 9.041397636584833e-06,
+      "loss": 0.543,
+      "step": 2088
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0537398549449106,
+      "learning_rate": 9.0403722494576e-06,
+      "loss": 0.539,
+      "step": 2089
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.760955248799296,
+      "learning_rate": 9.039346372421304e-06,
+      "loss": 0.5552,
+      "step": 2090
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.44827295463299693,
+      "learning_rate": 9.038320005600335e-06,
+      "loss": 0.3064,
+      "step": 2091
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.017991494577602,
+      "learning_rate": 9.037293149119144e-06,
+      "loss": 0.527,
+      "step": 2092
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9892170684685038,
+      "learning_rate": 9.036265803102238e-06,
+      "loss": 0.5234,
+      "step": 2093
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.113153456788908,
+      "learning_rate": 9.035237967674189e-06,
+      "loss": 0.5288,
+      "step": 2094
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.784319509540268,
+      "learning_rate": 9.034209642959624e-06,
+      "loss": 0.5327,
+      "step": 2095
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.854674272203744,
+      "learning_rate": 9.033180829083232e-06,
+      "loss": 0.513,
+      "step": 2096
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.569850839098888,
+      "learning_rate": 9.032151526169761e-06,
+      "loss": 0.522,
+      "step": 2097
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.51055094006708,
+      "learning_rate": 9.031121734344016e-06,
+      "loss": 0.476,
+      "step": 2098
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.434257855492977,
+      "learning_rate": 9.030091453730863e-06,
+      "loss": 0.5501,
+      "step": 2099
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.42471265937716607,
+      "learning_rate": 9.02906068445523e-06,
+      "loss": 0.3114,
+      "step": 2100
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.38135624193651335,
+      "learning_rate": 9.028029426642095e-06,
+      "loss": 0.3035,
+      "step": 2101
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.069687661567821,
+      "learning_rate": 9.026997680416507e-06,
+      "loss": 0.5716,
+      "step": 2102
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.9598630428904866,
+      "learning_rate": 9.025965445903567e-06,
+      "loss": 0.5255,
+      "step": 2103
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2929117991784627,
+      "learning_rate": 9.024932723228436e-06,
+      "loss": 0.4608,
+      "step": 2104
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.897377611944763,
+      "learning_rate": 9.023899512516337e-06,
+      "loss": 0.5286,
+      "step": 2105
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9238109985186993,
+      "learning_rate": 9.02286581389255e-06,
+      "loss": 0.6044,
+      "step": 2106
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7922132936894233,
+      "learning_rate": 9.021831627482414e-06,
+      "loss": 0.5244,
+      "step": 2107
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.609674408404345,
+      "learning_rate": 9.020796953411328e-06,
+      "loss": 0.5356,
+      "step": 2108
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.68303102982411,
+      "learning_rate": 9.019761791804749e-06,
+      "loss": 0.5119,
+      "step": 2109
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.131214956469587,
+      "learning_rate": 9.018726142788195e-06,
+      "loss": 0.5363,
+      "step": 2110
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.382094256374575,
+      "learning_rate": 9.01769000648724e-06,
+      "loss": 0.5585,
+      "step": 2111
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.307324404400855,
+      "learning_rate": 9.016653383027522e-06,
+      "loss": 0.527,
+      "step": 2112
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.735034347622448,
+      "learning_rate": 9.015616272534734e-06,
+      "loss": 0.5598,
+      "step": 2113
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 45.034123779289956,
+      "learning_rate": 9.014578675134629e-06,
+      "loss": 0.5565,
+      "step": 2114
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.9509164607282403,
+      "learning_rate": 9.013540590953019e-06,
+      "loss": 0.5097,
+      "step": 2115
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0626544931026687,
+      "learning_rate": 9.012502020115777e-06,
+      "loss": 0.5468,
+      "step": 2116
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.380074659275236,
+      "learning_rate": 9.01146296274883e-06,
+      "loss": 0.4831,
+      "step": 2117
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5297471802139562,
+      "learning_rate": 9.01042341897817e-06,
+      "loss": 0.3166,
+      "step": 2118
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.405482550476925,
+      "learning_rate": 9.009383388929842e-06,
+      "loss": 0.5218,
+      "step": 2119
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.013691788335571,
+      "learning_rate": 9.008342872729958e-06,
+      "loss": 0.5186,
+      "step": 2120
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2400700413185786,
+      "learning_rate": 9.007301870504681e-06,
+      "loss": 0.5791,
+      "step": 2121
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.612286571365062,
+      "learning_rate": 9.006260382380238e-06,
+      "loss": 0.5027,
+      "step": 2122
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.822186654907425,
+      "learning_rate": 9.005218408482912e-06,
+      "loss": 0.5253,
+      "step": 2123
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.448118507211581,
+      "learning_rate": 9.004175948939045e-06,
+      "loss": 0.5163,
+      "step": 2124
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3592709299787433,
+      "learning_rate": 9.003133003875039e-06,
+      "loss": 0.5451,
+      "step": 2125
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.500709957967074,
+      "learning_rate": 9.002089573417356e-06,
+      "loss": 0.4859,
+      "step": 2126
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 41.97830400119511,
+      "learning_rate": 9.001045657692517e-06,
+      "loss": 0.5191,
+      "step": 2127
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9384958776998626,
+      "learning_rate": 9.000001256827096e-06,
+      "loss": 0.4896,
+      "step": 2128
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.4714654308585295,
+      "learning_rate": 8.998956370947733e-06,
+      "loss": 0.5496,
+      "step": 2129
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.438976838980377,
+      "learning_rate": 8.997911000181124e-06,
+      "loss": 0.4958,
+      "step": 2130
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3329922182444127,
+      "learning_rate": 8.996865144654023e-06,
+      "loss": 0.5417,
+      "step": 2131
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.31255967387725,
+      "learning_rate": 8.995818804493244e-06,
+      "loss": 0.5014,
+      "step": 2132
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.4196516728079205,
+      "learning_rate": 8.994771979825658e-06,
+      "loss": 0.4884,
+      "step": 2133
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.426794807300998,
+      "learning_rate": 8.993724670778199e-06,
+      "loss": 0.5268,
+      "step": 2134
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6617661751049266,
+      "learning_rate": 8.992676877477855e-06,
+      "loss": 0.4726,
+      "step": 2135
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3344102852204127,
+      "learning_rate": 8.991628600051674e-06,
+      "loss": 0.4892,
+      "step": 2136
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5955686106503233,
+      "learning_rate": 8.990579838626764e-06,
+      "loss": 0.5186,
+      "step": 2137
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2673862749641716,
+      "learning_rate": 8.989530593330292e-06,
+      "loss": 0.5019,
+      "step": 2138
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.193976314553291,
+      "learning_rate": 8.988480864289482e-06,
+      "loss": 0.5451,
+      "step": 2139
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.0763990832289725,
+      "learning_rate": 8.987430651631615e-06,
+      "loss": 0.4797,
+      "step": 2140
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1230707456697644,
+      "learning_rate": 8.986379955484036e-06,
+      "loss": 0.4995,
+      "step": 2141
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.875433486342317,
+      "learning_rate": 8.985328775974142e-06,
+      "loss": 0.5085,
+      "step": 2142
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2404722951025158,
+      "learning_rate": 8.984277113229398e-06,
+      "loss": 0.5049,
+      "step": 2143
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6161566940975266,
+      "learning_rate": 8.983224967377314e-06,
+      "loss": 0.3257,
+      "step": 2144
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2021245012544526,
+      "learning_rate": 8.982172338545474e-06,
+      "loss": 0.529,
+      "step": 2145
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5331885621059587,
+      "learning_rate": 8.981119226861508e-06,
+      "loss": 0.5188,
+      "step": 2146
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9929329330617558,
+      "learning_rate": 8.980065632453111e-06,
+      "loss": 0.5553,
+      "step": 2147
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8533906521727554,
+      "learning_rate": 8.979011555448035e-06,
+      "loss": 0.5483,
+      "step": 2148
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.41491898102923624,
+      "learning_rate": 8.97795699597409e-06,
+      "loss": 0.3055,
+      "step": 2149
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3773891492862886,
+      "learning_rate": 8.976901954159144e-06,
+      "loss": 0.5364,
+      "step": 2150
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8703775978436834,
+      "learning_rate": 8.975846430131127e-06,
+      "loss": 0.5517,
+      "step": 2151
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.344559648825846,
+      "learning_rate": 8.974790424018022e-06,
+      "loss": 0.5338,
+      "step": 2152
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1034443592111844,
+      "learning_rate": 8.973733935947877e-06,
+      "loss": 0.5702,
+      "step": 2153
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.275729755126836,
+      "learning_rate": 8.97267696604879e-06,
+      "loss": 0.5546,
+      "step": 2154
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.239336338131404,
+      "learning_rate": 8.971619514448928e-06,
+      "loss": 0.5582,
+      "step": 2155
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.638026722627972,
+      "learning_rate": 8.970561581276506e-06,
+      "loss": 0.5706,
+      "step": 2156
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.328447795756258,
+      "learning_rate": 8.969503166659803e-06,
+      "loss": 0.5622,
+      "step": 2157
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2651117158259786,
+      "learning_rate": 8.968444270727157e-06,
+      "loss": 0.5019,
+      "step": 2158
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1467831409828584,
+      "learning_rate": 8.967384893606962e-06,
+      "loss": 0.4982,
+      "step": 2159
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.447202113847692,
+      "learning_rate": 8.96632503542767e-06,
+      "loss": 0.5513,
+      "step": 2160
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.666126072846837,
+      "learning_rate": 8.965264696317795e-06,
+      "loss": 0.5491,
+      "step": 2161
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.059450035421164,
+      "learning_rate": 8.964203876405903e-06,
+      "loss": 0.5506,
+      "step": 2162
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.753358557117852,
+      "learning_rate": 8.963142575820626e-06,
+      "loss": 0.5415,
+      "step": 2163
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5194677289761314,
+      "learning_rate": 8.962080794690648e-06,
+      "loss": 0.3234,
+      "step": 2164
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.437298637409712,
+      "learning_rate": 8.961018533144716e-06,
+      "loss": 0.5458,
+      "step": 2165
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5782524668939133,
+      "learning_rate": 8.95995579131163e-06,
+      "loss": 0.5335,
+      "step": 2166
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4052444913524179,
+      "learning_rate": 8.95889256932025e-06,
+      "loss": 0.2911,
+      "step": 2167
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5925143654122778,
+      "learning_rate": 8.9578288672995e-06,
+      "loss": 0.4926,
+      "step": 2168
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2217991445901064,
+      "learning_rate": 8.956764685378356e-06,
+      "loss": 0.5344,
+      "step": 2169
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.414486878634304,
+      "learning_rate": 8.955700023685851e-06,
+      "loss": 0.4866,
+      "step": 2170
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.814352560603357,
+      "learning_rate": 8.95463488235108e-06,
+      "loss": 0.5403,
+      "step": 2171
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7039709982248596,
+      "learning_rate": 8.953569261503198e-06,
+      "loss": 0.4736,
+      "step": 2172
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.45631430846115434,
+      "learning_rate": 8.952503161271413e-06,
+      "loss": 0.2997,
+      "step": 2173
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2069085043551833,
+      "learning_rate": 8.951436581784992e-06,
+      "loss": 0.5013,
+      "step": 2174
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1134231403625447,
+      "learning_rate": 8.950369523173263e-06,
+      "loss": 0.5367,
+      "step": 2175
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.8568552417702455,
+      "learning_rate": 8.949301985565611e-06,
+      "loss": 0.5439,
+      "step": 2176
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.079551760994727,
+      "learning_rate": 8.948233969091478e-06,
+      "loss": 0.5274,
+      "step": 2177
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1631944036880184,
+      "learning_rate": 8.947165473880364e-06,
+      "loss": 0.5187,
+      "step": 2178
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.386966280186438,
+      "learning_rate": 8.946096500061828e-06,
+      "loss": 0.5283,
+      "step": 2179
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.1529841839765,
+      "learning_rate": 8.94502704776549e-06,
+      "loss": 0.5546,
+      "step": 2180
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8927192750492616,
+      "learning_rate": 8.943957117121018e-06,
+      "loss": 0.4794,
+      "step": 2181
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.802367305198346,
+      "learning_rate": 8.94288670825815e-06,
+      "loss": 0.4871,
+      "step": 2182
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.401073227799703,
+      "learning_rate": 8.941815821306675e-06,
+      "loss": 0.569,
+      "step": 2183
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2393333995763327,
+      "learning_rate": 8.940744456396444e-06,
+      "loss": 0.5024,
+      "step": 2184
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0373821100512126,
+      "learning_rate": 8.93967261365736e-06,
+      "loss": 0.5179,
+      "step": 2185
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.996668491598162,
+      "learning_rate": 8.93860029321939e-06,
+      "loss": 0.5463,
+      "step": 2186
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.799037953281814,
+      "learning_rate": 8.937527495212555e-06,
+      "loss": 0.5265,
+      "step": 2187
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.111174427653147,
+      "learning_rate": 8.93645421976694e-06,
+      "loss": 0.5067,
+      "step": 2188
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4640885482508726,
+      "learning_rate": 8.935380467012675e-06,
+      "loss": 0.3003,
+      "step": 2189
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1839235857400743,
+      "learning_rate": 8.934306237079963e-06,
+      "loss": 0.5937,
+      "step": 2190
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2621432548632843,
+      "learning_rate": 8.933231530099057e-06,
+      "loss": 0.5024,
+      "step": 2191
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.130338330026273,
+      "learning_rate": 8.932156346200269e-06,
+      "loss": 0.4806,
+      "step": 2192
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9359927024222587,
+      "learning_rate": 8.931080685513966e-06,
+      "loss": 0.4866,
+      "step": 2193
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.679761017260991,
+      "learning_rate": 8.930004548170577e-06,
+      "loss": 0.5287,
+      "step": 2194
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.645332514099281,
+      "learning_rate": 8.92892793430059e-06,
+      "loss": 0.5502,
+      "step": 2195
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2156275187064765,
+      "learning_rate": 8.927850844034545e-06,
+      "loss": 0.5358,
+      "step": 2196
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0488492666704055,
+      "learning_rate": 8.926773277503042e-06,
+      "loss": 0.5544,
+      "step": 2197
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.455707162028217,
+      "learning_rate": 8.925695234836743e-06,
+      "loss": 0.4973,
+      "step": 2198
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.038566007712987,
+      "learning_rate": 8.924616716166363e-06,
+      "loss": 0.5081,
+      "step": 2199
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9997769461835193,
+      "learning_rate": 8.923537721622675e-06,
+      "loss": 0.5341,
+      "step": 2200
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.096619247063633,
+      "learning_rate": 8.922458251336512e-06,
+      "loss": 0.4974,
+      "step": 2201
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.315760674702479,
+      "learning_rate": 8.921378305438763e-06,
+      "loss": 0.5815,
+      "step": 2202
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.156530408230073,
+      "learning_rate": 8.920297884060378e-06,
+      "loss": 0.5657,
+      "step": 2203
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4222578758976683,
+      "learning_rate": 8.919216987332357e-06,
+      "loss": 0.2967,
+      "step": 2204
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.868590727186708,
+      "learning_rate": 8.918135615385763e-06,
+      "loss": 0.5267,
+      "step": 2205
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.076770366496617,
+      "learning_rate": 8.917053768351719e-06,
+      "loss": 0.5623,
+      "step": 2206
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.375809705893274,
+      "learning_rate": 8.915971446361404e-06,
+      "loss": 0.5008,
+      "step": 2207
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9678401619095083,
+      "learning_rate": 8.914888649546049e-06,
+      "loss": 0.504,
+      "step": 2208
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.704506756818689,
+      "learning_rate": 8.913805378036948e-06,
+      "loss": 0.5574,
+      "step": 2209
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.910136174855623,
+      "learning_rate": 8.912721631965454e-06,
+      "loss": 0.5273,
+      "step": 2210
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7981107404309977,
+      "learning_rate": 8.91163741146297e-06,
+      "loss": 0.5158,
+      "step": 2211
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3873476689429993,
+      "learning_rate": 8.910552716660966e-06,
+      "loss": 0.4837,
+      "step": 2212
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.342185031467999,
+      "learning_rate": 8.909467547690961e-06,
+      "loss": 0.5126,
+      "step": 2213
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6492002311053646,
+      "learning_rate": 8.908381904684543e-06,
+      "loss": 0.5059,
+      "step": 2214
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3916679697044056,
+      "learning_rate": 8.90729578777334e-06,
+      "loss": 0.5462,
+      "step": 2215
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.469541659770103,
+      "learning_rate": 8.906209197089055e-06,
+      "loss": 0.514,
+      "step": 2216
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.834083327307304,
+      "learning_rate": 8.905122132763437e-06,
+      "loss": 0.5015,
+      "step": 2217
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.907968018504484,
+      "learning_rate": 8.904034594928296e-06,
+      "loss": 0.4913,
+      "step": 2218
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2003196041855797,
+      "learning_rate": 8.902946583715503e-06,
+      "loss": 0.5471,
+      "step": 2219
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8062186299555154,
+      "learning_rate": 8.901858099256981e-06,
+      "loss": 0.5118,
+      "step": 2220
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6479794311307834,
+      "learning_rate": 8.900769141684714e-06,
+      "loss": 0.5157,
+      "step": 2221
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9181977130310295,
+      "learning_rate": 8.899679711130737e-06,
+      "loss": 0.5171,
+      "step": 2222
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3634710079209733,
+      "learning_rate": 8.898589807727154e-06,
+      "loss": 0.5327,
+      "step": 2223
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4570604647815797,
+      "learning_rate": 8.897499431606116e-06,
+      "loss": 0.5559,
+      "step": 2224
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.361880377809965,
+      "learning_rate": 8.896408582899835e-06,
+      "loss": 0.5544,
+      "step": 2225
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.518122847351446,
+      "learning_rate": 8.89531726174058e-06,
+      "loss": 0.5636,
+      "step": 2226
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9719146939032988,
+      "learning_rate": 8.894225468260676e-06,
+      "loss": 0.527,
+      "step": 2227
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8054783837183543,
+      "learning_rate": 8.89313320259251e-06,
+      "loss": 0.5192,
+      "step": 2228
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3880181985005877,
+      "learning_rate": 8.89204046486852e-06,
+      "loss": 0.5069,
+      "step": 2229
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8746230664639105,
+      "learning_rate": 8.89094725522121e-06,
+      "loss": 0.547,
+      "step": 2230
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.809278206837011,
+      "learning_rate": 8.889853573783128e-06,
+      "loss": 0.539,
+      "step": 2231
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4554428146386284,
+      "learning_rate": 8.88875942068689e-06,
+      "loss": 0.3153,
+      "step": 2232
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.613653625528596,
+      "learning_rate": 8.887664796065167e-06,
+      "loss": 0.5511,
+      "step": 2233
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.258448233118173,
+      "learning_rate": 8.886569700050682e-06,
+      "loss": 0.5433,
+      "step": 2234
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9904483572399783,
+      "learning_rate": 8.885474132776223e-06,
+      "loss": 0.5639,
+      "step": 2235
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.438407385309893,
+      "learning_rate": 8.884378094374633e-06,
+      "loss": 0.5256,
+      "step": 2236
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9347634758980736,
+      "learning_rate": 8.883281584978805e-06,
+      "loss": 0.5043,
+      "step": 2237
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.849430682014274,
+      "learning_rate": 8.882184604721697e-06,
+      "loss": 0.4698,
+      "step": 2238
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.545611638814667,
+      "learning_rate": 8.88108715373632e-06,
+      "loss": 0.5308,
+      "step": 2239
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.468132221370889,
+      "learning_rate": 8.879989232155749e-06,
+      "loss": 0.4953,
+      "step": 2240
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 25.566321259321672,
+      "learning_rate": 8.878890840113106e-06,
+      "loss": 0.5179,
+      "step": 2241
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.583951115217915,
+      "learning_rate": 8.877791977741575e-06,
+      "loss": 0.509,
+      "step": 2242
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7278213576110923,
+      "learning_rate": 8.8766926451744e-06,
+      "loss": 0.5221,
+      "step": 2243
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.468055728054183,
+      "learning_rate": 8.875592842544875e-06,
+      "loss": 0.5656,
+      "step": 2244
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.686544251259313,
+      "learning_rate": 8.874492569986357e-06,
+      "loss": 0.4857,
+      "step": 2245
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.40781474996359,
+      "learning_rate": 8.873391827632258e-06,
+      "loss": 0.528,
+      "step": 2246
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.659592947141773,
+      "learning_rate": 8.872290615616045e-06,
+      "loss": 0.5215,
+      "step": 2247
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.542705602062972,
+      "learning_rate": 8.871188934071247e-06,
+      "loss": 0.5545,
+      "step": 2248
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8273279482265457,
+      "learning_rate": 8.870086783131443e-06,
+      "loss": 0.5163,
+      "step": 2249
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6399835964988507,
+      "learning_rate": 8.868984162930276e-06,
+      "loss": 0.4872,
+      "step": 2250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.284452747611832,
+      "learning_rate": 8.86788107360144e-06,
+      "loss": 0.5094,
+      "step": 2251
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.2718141390230135,
+      "learning_rate": 8.866777515278688e-06,
+      "loss": 0.5178,
+      "step": 2252
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.304964614260463,
+      "learning_rate": 8.865673488095833e-06,
+      "loss": 0.5613,
+      "step": 2253
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.060912615092749,
+      "learning_rate": 8.86456899218674e-06,
+      "loss": 0.5147,
+      "step": 2254
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7547591645779694,
+      "learning_rate": 8.863464027685333e-06,
+      "loss": 0.4963,
+      "step": 2255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.69984400118174,
+      "learning_rate": 8.862358594725595e-06,
+      "loss": 0.5666,
+      "step": 2256
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.309691039185772,
+      "learning_rate": 8.86125269344156e-06,
+      "loss": 0.471,
+      "step": 2257
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.315501624862645,
+      "learning_rate": 8.860146323967325e-06,
+      "loss": 0.5267,
+      "step": 2258
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.452944287783698,
+      "learning_rate": 8.85903948643704e-06,
+      "loss": 0.5551,
+      "step": 2259
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4847225275618956,
+      "learning_rate": 8.857932180984914e-06,
+      "loss": 0.5264,
+      "step": 2260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.073570056268865,
+      "learning_rate": 8.85682440774521e-06,
+      "loss": 0.491,
+      "step": 2261
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6456061120916283,
+      "learning_rate": 8.855716166852249e-06,
+      "loss": 0.5077,
+      "step": 2262
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8710163126493757,
+      "learning_rate": 8.854607458440412e-06,
+      "loss": 0.4841,
+      "step": 2263
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4380365548888374,
+      "learning_rate": 8.85349828264413e-06,
+      "loss": 0.4953,
+      "step": 2264
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.828733427885762,
+      "learning_rate": 8.852388639597898e-06,
+      "loss": 0.5455,
+      "step": 2265
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.452839625900196,
+      "learning_rate": 8.85127852943626e-06,
+      "loss": 0.5412,
+      "step": 2266
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2131649072617217,
+      "learning_rate": 8.850167952293825e-06,
+      "loss": 0.5392,
+      "step": 2267
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3694296644751294,
+      "learning_rate": 8.849056908305252e-06,
+      "loss": 0.5175,
+      "step": 2268
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3891080571601293,
+      "learning_rate": 8.847945397605258e-06,
+      "loss": 0.5458,
+      "step": 2269
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.854271467016337,
+      "learning_rate": 8.84683342032862e-06,
+      "loss": 0.4912,
+      "step": 2270
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.003987595630331,
+      "learning_rate": 8.845720976610168e-06,
+      "loss": 0.4857,
+      "step": 2271
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 64.9183359804107,
+      "learning_rate": 8.844608066584788e-06,
+      "loss": 0.5067,
+      "step": 2272
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4170560889219552,
+      "learning_rate": 8.843494690387427e-06,
+      "loss": 0.5321,
+      "step": 2273
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2120578247240483,
+      "learning_rate": 8.842380848153083e-06,
+      "loss": 0.5273,
+      "step": 2274
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.97743444216561,
+      "learning_rate": 8.841266540016814e-06,
+      "loss": 0.5548,
+      "step": 2275
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 25.089284719020334,
+      "learning_rate": 8.840151766113735e-06,
+      "loss": 0.5474,
+      "step": 2276
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.900554839890724,
+      "learning_rate": 8.839036526579014e-06,
+      "loss": 0.5336,
+      "step": 2277
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.951172985937207,
+      "learning_rate": 8.837920821547881e-06,
+      "loss": 0.4941,
+      "step": 2278
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9071382950909608,
+      "learning_rate": 8.836804651155617e-06,
+      "loss": 0.5107,
+      "step": 2279
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.917662849190273,
+      "learning_rate": 8.83568801553756e-06,
+      "loss": 0.4615,
+      "step": 2280
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.676412754694638,
+      "learning_rate": 8.834570914829108e-06,
+      "loss": 0.5171,
+      "step": 2281
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.657930085147365,
+      "learning_rate": 8.833453349165713e-06,
+      "loss": 0.5283,
+      "step": 2282
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6644187788928706,
+      "learning_rate": 8.832335318682884e-06,
+      "loss": 0.5145,
+      "step": 2283
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.443696647154578,
+      "learning_rate": 8.831216823516185e-06,
+      "loss": 0.3029,
+      "step": 2284
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.040799989266873,
+      "learning_rate": 8.830097863801239e-06,
+      "loss": 0.5433,
+      "step": 2285
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.628455785126419,
+      "learning_rate": 8.828978439673721e-06,
+      "loss": 0.5307,
+      "step": 2286
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6166580546258174,
+      "learning_rate": 8.827858551269368e-06,
+      "loss": 0.5293,
+      "step": 2287
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0963036849370207,
+      "learning_rate": 8.826738198723967e-06,
+      "loss": 0.5398,
+      "step": 2288
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9620856619917895,
+      "learning_rate": 8.825617382173369e-06,
+      "loss": 0.5371,
+      "step": 2289
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.30443943865154,
+      "learning_rate": 8.824496101753474e-06,
+      "loss": 0.4793,
+      "step": 2290
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.839549627548044,
+      "learning_rate": 8.82337435760024e-06,
+      "loss": 0.5209,
+      "step": 2291
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7323247124511214,
+      "learning_rate": 8.822252149849687e-06,
+      "loss": 0.5207,
+      "step": 2292
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.123239632024538,
+      "learning_rate": 8.821129478637882e-06,
+      "loss": 0.5186,
+      "step": 2293
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.751945550419488,
+      "learning_rate": 8.820006344100953e-06,
+      "loss": 0.524,
+      "step": 2294
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.537438825125903,
+      "learning_rate": 8.818882746375086e-06,
+      "loss": 0.497,
+      "step": 2295
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6987780114012403,
+      "learning_rate": 8.81775868559652e-06,
+      "loss": 0.4924,
+      "step": 2296
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.561257020058054,
+      "learning_rate": 8.816634161901553e-06,
+      "loss": 0.4823,
+      "step": 2297
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2963388340546524,
+      "learning_rate": 8.815509175426536e-06,
+      "loss": 0.5101,
+      "step": 2298
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.501184744745184,
+      "learning_rate": 8.814383726307876e-06,
+      "loss": 0.5525,
+      "step": 2299
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2067507507255946,
+      "learning_rate": 8.81325781468204e-06,
+      "loss": 0.4964,
+      "step": 2300
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.066725320124624,
+      "learning_rate": 8.812131440685545e-06,
+      "loss": 0.5339,
+      "step": 2301
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.25762235844573,
+      "learning_rate": 8.811004604454973e-06,
+      "loss": 0.5141,
+      "step": 2302
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.160645363955025,
+      "learning_rate": 8.809877306126954e-06,
+      "loss": 0.5351,
+      "step": 2303
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.0300754416316025,
+      "learning_rate": 8.808749545838177e-06,
+      "loss": 0.5475,
+      "step": 2304
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.595051749292006,
+      "learning_rate": 8.807621323725387e-06,
+      "loss": 0.4886,
+      "step": 2305
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.783440217123251,
+      "learning_rate": 8.806492639925383e-06,
+      "loss": 0.4886,
+      "step": 2306
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.816162983163063,
+      "learning_rate": 8.805363494575026e-06,
+      "loss": 0.504,
+      "step": 2307
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9296793653561113,
+      "learning_rate": 8.804233887811224e-06,
+      "loss": 0.5742,
+      "step": 2308
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.091304382360398,
+      "learning_rate": 8.803103819770948e-06,
+      "loss": 0.5757,
+      "step": 2309
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.685430652731865,
+      "learning_rate": 8.801973290591223e-06,
+      "loss": 0.547,
+      "step": 2310
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4407999569441014,
+      "learning_rate": 8.800842300409129e-06,
+      "loss": 0.2931,
+      "step": 2311
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.44382819093854,
+      "learning_rate": 8.799710849361804e-06,
+      "loss": 0.3159,
+      "step": 2312
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.519247015392766,
+      "learning_rate": 8.798578937586437e-06,
+      "loss": 0.5137,
+      "step": 2313
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8994570523161407,
+      "learning_rate": 8.797446565220279e-06,
+      "loss": 0.508,
+      "step": 2314
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4230492159363205,
+      "learning_rate": 8.796313732400634e-06,
+      "loss": 0.3005,
+      "step": 2315
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.724979794991754,
+      "learning_rate": 8.79518043926486e-06,
+      "loss": 0.5364,
+      "step": 2316
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.375532671283306,
+      "learning_rate": 8.794046685950374e-06,
+      "loss": 0.5314,
+      "step": 2317
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2891771351373436,
+      "learning_rate": 8.792912472594646e-06,
+      "loss": 0.4745,
+      "step": 2318
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9838432081007915,
+      "learning_rate": 8.791777799335206e-06,
+      "loss": 0.5227,
+      "step": 2319
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.208454369333823,
+      "learning_rate": 8.790642666309637e-06,
+      "loss": 0.5062,
+      "step": 2320
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.136192535839593,
+      "learning_rate": 8.789507073655575e-06,
+      "loss": 0.4749,
+      "step": 2321
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.32404472381063,
+      "learning_rate": 8.788371021510713e-06,
+      "loss": 0.5111,
+      "step": 2322
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.375602071653173,
+      "learning_rate": 8.787234510012808e-06,
+      "loss": 0.5406,
+      "step": 2323
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.302405647090641,
+      "learning_rate": 8.786097539299661e-06,
+      "loss": 0.5277,
+      "step": 2324
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.741945384471089,
+      "learning_rate": 8.784960109509134e-06,
+      "loss": 0.5174,
+      "step": 2325
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5224894793553915,
+      "learning_rate": 8.783822220779145e-06,
+      "loss": 0.332,
+      "step": 2326
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2882659736749162,
+      "learning_rate": 8.782683873247668e-06,
+      "loss": 0.5242,
+      "step": 2327
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9269784755208224,
+      "learning_rate": 8.781545067052731e-06,
+      "loss": 0.5523,
+      "step": 2328
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3908857963364627,
+      "learning_rate": 8.780405802332415e-06,
+      "loss": 0.3002,
+      "step": 2329
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.909471214175972,
+      "learning_rate": 8.779266079224865e-06,
+      "loss": 0.4778,
+      "step": 2330
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.804218199787873,
+      "learning_rate": 8.778125897868273e-06,
+      "loss": 0.5493,
+      "step": 2331
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.998600872164356,
+      "learning_rate": 8.77698525840089e-06,
+      "loss": 0.5501,
+      "step": 2332
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.396572106105795,
+      "learning_rate": 8.775844160961024e-06,
+      "loss": 0.573,
+      "step": 2333
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.5419022667638345,
+      "learning_rate": 8.774702605687036e-06,
+      "loss": 0.5211,
+      "step": 2334
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6462308874517597,
+      "learning_rate": 8.773560592717343e-06,
+      "loss": 0.5407,
+      "step": 2335
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2616318790049412,
+      "learning_rate": 8.772418122190418e-06,
+      "loss": 0.5502,
+      "step": 2336
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.632205508370949,
+      "learning_rate": 8.771275194244792e-06,
+      "loss": 0.5178,
+      "step": 2337
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.486140152008987,
+      "learning_rate": 8.770131809019046e-06,
+      "loss": 0.5491,
+      "step": 2338
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4470237805640003,
+      "learning_rate": 8.768987966651822e-06,
+      "loss": 0.5352,
+      "step": 2339
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.01115987906634,
+      "learning_rate": 8.767843667281812e-06,
+      "loss": 0.5447,
+      "step": 2340
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.404651946545198,
+      "learning_rate": 8.766698911047769e-06,
+      "loss": 0.5112,
+      "step": 2341
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.897135883695028,
+      "learning_rate": 8.765553698088496e-06,
+      "loss": 0.5304,
+      "step": 2342
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4209439058708253,
+      "learning_rate": 8.764408028542854e-06,
+      "loss": 0.483,
+      "step": 2343
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.299424115566661,
+      "learning_rate": 8.763261902549763e-06,
+      "loss": 0.5187,
+      "step": 2344
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.588522519592556,
+      "learning_rate": 8.762115320248192e-06,
+      "loss": 0.5575,
+      "step": 2345
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.499425620447818,
+      "learning_rate": 8.760968281777167e-06,
+      "loss": 0.4932,
+      "step": 2346
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.677580380630724,
+      "learning_rate": 8.759820787275774e-06,
+      "loss": 0.5655,
+      "step": 2347
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.575978536192906,
+      "learning_rate": 8.758672836883146e-06,
+      "loss": 0.5645,
+      "step": 2348
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9069313424820344,
+      "learning_rate": 8.75752443073848e-06,
+      "loss": 0.5037,
+      "step": 2349
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9542017990167104,
+      "learning_rate": 8.756375568981023e-06,
+      "loss": 0.5456,
+      "step": 2350
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4213822262812,
+      "learning_rate": 8.755226251750077e-06,
+      "loss": 0.5068,
+      "step": 2351
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.483492665295017,
+      "learning_rate": 8.754076479185003e-06,
+      "loss": 0.5602,
+      "step": 2352
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5884132466950796,
+      "learning_rate": 8.752926251425212e-06,
+      "loss": 0.4859,
+      "step": 2353
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.230487527824679,
+      "learning_rate": 8.751775568610175e-06,
+      "loss": 0.4859,
+      "step": 2354
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3300868834049595,
+      "learning_rate": 8.750624430879417e-06,
+      "loss": 0.5633,
+      "step": 2355
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4961611859540755,
+      "learning_rate": 8.749472838372515e-06,
+      "loss": 0.5135,
+      "step": 2356
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.119591972574581,
+      "learning_rate": 8.748320791229106e-06,
+      "loss": 0.4976,
+      "step": 2357
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5498904199701578,
+      "learning_rate": 8.747168289588878e-06,
+      "loss": 0.5647,
+      "step": 2358
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.652617709494551,
+      "learning_rate": 8.746015333591578e-06,
+      "loss": 0.4996,
+      "step": 2359
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3307537525011206,
+      "learning_rate": 8.744861923377e-06,
+      "loss": 0.4965,
+      "step": 2360
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1344982684056784,
+      "learning_rate": 8.74370805908501e-06,
+      "loss": 0.5358,
+      "step": 2361
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.371962888378493,
+      "learning_rate": 8.742553740855507e-06,
+      "loss": 0.5387,
+      "step": 2362
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.221606394103036,
+      "learning_rate": 8.74139896882846e-06,
+      "loss": 0.515,
+      "step": 2363
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4007476577361864,
+      "learning_rate": 8.74024374314389e-06,
+      "loss": 0.4986,
+      "step": 2364
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.511389892517968,
+      "learning_rate": 8.739088063941875e-06,
+      "loss": 0.5368,
+      "step": 2365
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6310608336379397,
+      "learning_rate": 8.737931931362537e-06,
+      "loss": 0.4612,
+      "step": 2366
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7730183088492604,
+      "learning_rate": 8.736775345546067e-06,
+      "loss": 0.5585,
+      "step": 2367
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5664066493434596,
+      "learning_rate": 8.735618306632706e-06,
+      "loss": 0.5014,
+      "step": 2368
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.035778197952887,
+      "learning_rate": 8.734460814762743e-06,
+      "loss": 0.5034,
+      "step": 2369
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6977715429806968,
+      "learning_rate": 8.733302870076535e-06,
+      "loss": 0.5367,
+      "step": 2370
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.946789597580481,
+      "learning_rate": 8.732144472714481e-06,
+      "loss": 0.5412,
+      "step": 2371
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6944304605151035,
+      "learning_rate": 8.730985622817043e-06,
+      "loss": 0.4519,
+      "step": 2372
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5673995460080753,
+      "learning_rate": 8.729826320524737e-06,
+      "loss": 0.308,
+      "step": 2373
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.689988757427988,
+      "learning_rate": 8.72866656597813e-06,
+      "loss": 0.572,
+      "step": 2374
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.025017863966893,
+      "learning_rate": 8.727506359317848e-06,
+      "loss": 0.5161,
+      "step": 2375
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3759007596197894,
+      "learning_rate": 8.72634570068457e-06,
+      "loss": 0.5342,
+      "step": 2376
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.394840871169317,
+      "learning_rate": 8.725184590219028e-06,
+      "loss": 0.4938,
+      "step": 2377
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4054131846440607,
+      "learning_rate": 8.724023028062011e-06,
+      "loss": 0.522,
+      "step": 2378
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.158822434057617,
+      "learning_rate": 8.722861014354363e-06,
+      "loss": 0.525,
+      "step": 2379
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6480089220130085,
+      "learning_rate": 8.721698549236982e-06,
+      "loss": 0.4613,
+      "step": 2380
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.855807882920706,
+      "learning_rate": 8.720535632850824e-06,
+      "loss": 0.4822,
+      "step": 2381
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.56711948218967,
+      "learning_rate": 8.719372265336893e-06,
+      "loss": 0.5616,
+      "step": 2382
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2988968335932167,
+      "learning_rate": 8.718208446836252e-06,
+      "loss": 0.5391,
+      "step": 2383
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.149990855910253,
+      "learning_rate": 8.717044177490016e-06,
+      "loss": 0.4954,
+      "step": 2384
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.186914593264069,
+      "learning_rate": 8.715879457439362e-06,
+      "loss": 0.5019,
+      "step": 2385
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5106724018920805,
+      "learning_rate": 8.714714286825512e-06,
+      "loss": 0.4872,
+      "step": 2386
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6139456790510067,
+      "learning_rate": 8.713548665789748e-06,
+      "loss": 0.5737,
+      "step": 2387
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.522152746474053,
+      "learning_rate": 8.712382594473405e-06,
+      "loss": 0.4424,
+      "step": 2388
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8039545815537283,
+      "learning_rate": 8.711216073017875e-06,
+      "loss": 0.5185,
+      "step": 2389
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.555949881286514,
+      "learning_rate": 8.710049101564602e-06,
+      "loss": 0.5269,
+      "step": 2390
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.367986461788422,
+      "learning_rate": 8.708881680255084e-06,
+      "loss": 0.5047,
+      "step": 2391
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4832839131528837,
+      "learning_rate": 8.707713809230875e-06,
+      "loss": 0.5269,
+      "step": 2392
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5588960748718517,
+      "learning_rate": 8.706545488633587e-06,
+      "loss": 0.3253,
+      "step": 2393
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.532162857678642,
+      "learning_rate": 8.705376718604877e-06,
+      "loss": 0.5664,
+      "step": 2394
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8245564233789886,
+      "learning_rate": 8.704207499286467e-06,
+      "loss": 0.5041,
+      "step": 2395
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.9186174518060066,
+      "learning_rate": 8.703037830820128e-06,
+      "loss": 0.4934,
+      "step": 2396
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2284721753395216,
+      "learning_rate": 8.701867713347686e-06,
+      "loss": 0.4897,
+      "step": 2397
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9332769615845655,
+      "learning_rate": 8.70069714701102e-06,
+      "loss": 0.5058,
+      "step": 2398
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5119146728628943,
+      "learning_rate": 8.699526131952065e-06,
+      "loss": 0.5091,
+      "step": 2399
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.41228278116355666,
+      "learning_rate": 8.698354668312816e-06,
+      "loss": 0.2829,
+      "step": 2400
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.686426458485488,
+      "learning_rate": 8.697182756235311e-06,
+      "loss": 0.5408,
+      "step": 2401
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3945013042870147,
+      "learning_rate": 8.696010395861651e-06,
+      "loss": 0.5527,
+      "step": 2402
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.981948886114994,
+      "learning_rate": 8.694837587333989e-06,
+      "loss": 0.4837,
+      "step": 2403
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3421174764217274,
+      "learning_rate": 8.69366433079453e-06,
+      "loss": 0.5395,
+      "step": 2404
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2896105461335003,
+      "learning_rate": 8.69249062638554e-06,
+      "loss": 0.5386,
+      "step": 2405
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3697335690807666,
+      "learning_rate": 8.691316474249329e-06,
+      "loss": 0.505,
+      "step": 2406
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.362181423814941,
+      "learning_rate": 8.69014187452827e-06,
+      "loss": 0.497,
+      "step": 2407
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.094884709195693,
+      "learning_rate": 8.688966827364788e-06,
+      "loss": 0.4973,
+      "step": 2408
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.9990521218740502,
+      "learning_rate": 8.68779133290136e-06,
+      "loss": 0.5141,
+      "step": 2409
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0527721772916876,
+      "learning_rate": 8.686615391280519e-06,
+      "loss": 0.5071,
+      "step": 2410
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4153932387265332,
+      "learning_rate": 8.68543900264485e-06,
+      "loss": 0.3046,
+      "step": 2411
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7278014267933557,
+      "learning_rate": 8.684262167136999e-06,
+      "loss": 0.5548,
+      "step": 2412
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.458443910876518,
+      "learning_rate": 8.683084884899658e-06,
+      "loss": 0.4956,
+      "step": 2413
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.499424898870345,
+      "learning_rate": 8.681907156075578e-06,
+      "loss": 0.5378,
+      "step": 2414
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.656198154209604,
+      "learning_rate": 8.68072898080756e-06,
+      "loss": 0.5709,
+      "step": 2415
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6347573604537278,
+      "learning_rate": 8.679550359238464e-06,
+      "loss": 0.4751,
+      "step": 2416
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.222946842657881,
+      "learning_rate": 8.678371291511202e-06,
+      "loss": 0.5022,
+      "step": 2417
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.653337217843201,
+      "learning_rate": 8.67719177776874e-06,
+      "loss": 0.4705,
+      "step": 2418
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.9887072190809785,
+      "learning_rate": 8.676011818154097e-06,
+      "loss": 0.524,
+      "step": 2419
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7502568427488128,
+      "learning_rate": 8.67483141281035e-06,
+      "loss": 0.5381,
+      "step": 2420
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.329662649640144,
+      "learning_rate": 8.673650561880621e-06,
+      "loss": 0.5432,
+      "step": 2421
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5351505740490787,
+      "learning_rate": 8.6724692655081e-06,
+      "loss": 0.4916,
+      "step": 2422
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 147.33021692071142,
+      "learning_rate": 8.671287523836019e-06,
+      "loss": 0.4933,
+      "step": 2423
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.638973856041268,
+      "learning_rate": 8.670105337007668e-06,
+      "loss": 0.5207,
+      "step": 2424
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5549371082030325,
+      "learning_rate": 8.668922705166392e-06,
+      "loss": 0.5195,
+      "step": 2425
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.679753171774019,
+      "learning_rate": 8.667739628455592e-06,
+      "loss": 0.4891,
+      "step": 2426
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7196290687501463,
+      "learning_rate": 8.666556107018714e-06,
+      "loss": 0.4599,
+      "step": 2427
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4449967834666153,
+      "learning_rate": 8.665372140999268e-06,
+      "loss": 0.548,
+      "step": 2428
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.553476484086879,
+      "learning_rate": 8.664187730540815e-06,
+      "loss": 0.4975,
+      "step": 2429
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6378768130010553,
+      "learning_rate": 8.663002875786965e-06,
+      "loss": 0.5088,
+      "step": 2430
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0116449422125857,
+      "learning_rate": 8.661817576881391e-06,
+      "loss": 0.5575,
+      "step": 2431
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.235620227800026,
+      "learning_rate": 8.660631833967809e-06,
+      "loss": 0.5252,
+      "step": 2432
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5665972532885952,
+      "learning_rate": 8.65944564719e-06,
+      "loss": 0.4861,
+      "step": 2433
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.798947379253605,
+      "learning_rate": 8.658259016691786e-06,
+      "loss": 0.5267,
+      "step": 2434
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4279396501072672,
+      "learning_rate": 8.657071942617057e-06,
+      "loss": 0.2929,
+      "step": 2435
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.9412216734629935,
+      "learning_rate": 8.655884425109747e-06,
+      "loss": 0.5302,
+      "step": 2436
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.60003512976642,
+      "learning_rate": 8.654696464313847e-06,
+      "loss": 0.5188,
+      "step": 2437
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.239939433683566,
+      "learning_rate": 8.6535080603734e-06,
+      "loss": 0.5023,
+      "step": 2438
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0764999134887674,
+      "learning_rate": 8.652319213432505e-06,
+      "loss": 0.5299,
+      "step": 2439
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5937300455976753,
+      "learning_rate": 8.651129923635316e-06,
+      "loss": 0.4955,
+      "step": 2440
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2020275654397667,
+      "learning_rate": 8.649940191126033e-06,
+      "loss": 0.5302,
+      "step": 2441
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.341090755263974,
+      "learning_rate": 8.64875001604892e-06,
+      "loss": 0.4929,
+      "step": 2442
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.815687469130631,
+      "learning_rate": 8.647559398548292e-06,
+      "loss": 0.4932,
+      "step": 2443
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.220570349240297,
+      "learning_rate": 8.646368338768508e-06,
+      "loss": 0.5404,
+      "step": 2444
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.620318477456658,
+      "learning_rate": 8.645176836853992e-06,
+      "loss": 0.5731,
+      "step": 2445
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.726018940488314,
+      "learning_rate": 8.643984892949218e-06,
+      "loss": 0.5309,
+      "step": 2446
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7782732803132935,
+      "learning_rate": 8.642792507198714e-06,
+      "loss": 0.5439,
+      "step": 2447
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7791390279333803,
+      "learning_rate": 8.64159967974706e-06,
+      "loss": 0.5259,
+      "step": 2448
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0812530918865844,
+      "learning_rate": 8.64040641073889e-06,
+      "loss": 0.526,
+      "step": 2449
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.003784704246589,
+      "learning_rate": 8.63921270031889e-06,
+      "loss": 0.5177,
+      "step": 2450
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.455716484117922,
+      "learning_rate": 8.638018548631805e-06,
+      "loss": 0.5115,
+      "step": 2451
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.276462167837128,
+      "learning_rate": 8.636823955822432e-06,
+      "loss": 0.5017,
+      "step": 2452
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0814695339957794,
+      "learning_rate": 8.635628922035613e-06,
+      "loss": 0.5059,
+      "step": 2453
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8513408098630646,
+      "learning_rate": 8.634433447416254e-06,
+      "loss": 0.5193,
+      "step": 2454
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7787168854848643,
+      "learning_rate": 8.63323753210931e-06,
+      "loss": 0.5441,
+      "step": 2455
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1675517306451733,
+      "learning_rate": 8.63204117625979e-06,
+      "loss": 0.5582,
+      "step": 2456
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8773616326988263,
+      "learning_rate": 8.630844380012756e-06,
+      "loss": 0.4856,
+      "step": 2457
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3043966378097815,
+      "learning_rate": 8.629647143513322e-06,
+      "loss": 0.4833,
+      "step": 2458
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.166014284564437,
+      "learning_rate": 8.628449466906659e-06,
+      "loss": 0.4701,
+      "step": 2459
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.271956933772301,
+      "learning_rate": 8.627251350337988e-06,
+      "loss": 0.5109,
+      "step": 2460
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6274637894573396,
+      "learning_rate": 8.626052793952588e-06,
+      "loss": 0.5423,
+      "step": 2461
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4183799348692902,
+      "learning_rate": 8.624853797895784e-06,
+      "loss": 0.3031,
+      "step": 2462
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3273273680717135,
+      "learning_rate": 8.623654362312962e-06,
+      "loss": 0.5063,
+      "step": 2463
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.907175841486467,
+      "learning_rate": 8.622454487349556e-06,
+      "loss": 0.4869,
+      "step": 2464
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8781316392768477,
+      "learning_rate": 8.621254173151052e-06,
+      "loss": 0.5253,
+      "step": 2465
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6811173970494626,
+      "learning_rate": 8.620053419862997e-06,
+      "loss": 0.5068,
+      "step": 2466
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.638298581926339,
+      "learning_rate": 8.618852227630986e-06,
+      "loss": 0.5357,
+      "step": 2467
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7367888435930725,
+      "learning_rate": 8.617650596600665e-06,
+      "loss": 0.5163,
+      "step": 2468
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.480199460715218,
+      "learning_rate": 8.616448526917737e-06,
+      "loss": 0.5465,
+      "step": 2469
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2536591122803555,
+      "learning_rate": 8.615246018727957e-06,
+      "loss": 0.5441,
+      "step": 2470
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.5652739962026345,
+      "learning_rate": 8.614043072177134e-06,
+      "loss": 0.5057,
+      "step": 2471
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.574243717115615,
+      "learning_rate": 8.61283968741113e-06,
+      "loss": 0.5214,
+      "step": 2472
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4757883270842918,
+      "learning_rate": 8.611635864575858e-06,
+      "loss": 0.5057,
+      "step": 2473
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.804495533665407,
+      "learning_rate": 8.610431603817287e-06,
+      "loss": 0.55,
+      "step": 2474
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.906890608552861,
+      "learning_rate": 8.609226905281434e-06,
+      "loss": 0.5271,
+      "step": 2475
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.8518183749790476,
+      "learning_rate": 8.608021769114378e-06,
+      "loss": 0.5625,
+      "step": 2476
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4673469643524387,
+      "learning_rate": 8.606816195462244e-06,
+      "loss": 0.328,
+      "step": 2477
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.180376640357439,
+      "learning_rate": 8.605610184471211e-06,
+      "loss": 0.4838,
+      "step": 2478
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.679914285985613,
+      "learning_rate": 8.604403736287513e-06,
+      "loss": 0.5248,
+      "step": 2479
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.448599562305511,
+      "learning_rate": 8.603196851057435e-06,
+      "loss": 0.5069,
+      "step": 2480
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.230062927757448,
+      "learning_rate": 8.601989528927317e-06,
+      "loss": 0.5315,
+      "step": 2481
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9021777896371486,
+      "learning_rate": 8.600781770043552e-06,
+      "loss": 0.5649,
+      "step": 2482
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.681773575394761,
+      "learning_rate": 8.599573574552581e-06,
+      "loss": 0.5364,
+      "step": 2483
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.11793701268591,
+      "learning_rate": 8.598364942600907e-06,
+      "loss": 0.5059,
+      "step": 2484
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.083439453693574,
+      "learning_rate": 8.597155874335077e-06,
+      "loss": 0.5343,
+      "step": 2485
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9930436554820967,
+      "learning_rate": 8.595946369901696e-06,
+      "loss": 0.5682,
+      "step": 2486
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.816839120312609,
+      "learning_rate": 8.594736429447421e-06,
+      "loss": 0.5414,
+      "step": 2487
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7088079863042025,
+      "learning_rate": 8.593526053118962e-06,
+      "loss": 0.4634,
+      "step": 2488
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4300608083157456,
+      "learning_rate": 8.592315241063078e-06,
+      "loss": 0.5245,
+      "step": 2489
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9497199909149856,
+      "learning_rate": 8.591103993426588e-06,
+      "loss": 0.4895,
+      "step": 2490
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.119769345676543,
+      "learning_rate": 8.589892310356359e-06,
+      "loss": 0.5455,
+      "step": 2491
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2401870091461817,
+      "learning_rate": 8.588680191999308e-06,
+      "loss": 0.5177,
+      "step": 2492
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3101452549066757,
+      "learning_rate": 8.587467638502414e-06,
+      "loss": 0.5436,
+      "step": 2493
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9044991592398266,
+      "learning_rate": 8.5862546500127e-06,
+      "loss": 0.5033,
+      "step": 2494
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.496897247706123,
+      "learning_rate": 8.585041226677247e-06,
+      "loss": 0.5361,
+      "step": 2495
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.145004442142484,
+      "learning_rate": 8.583827368643185e-06,
+      "loss": 0.5215,
+      "step": 2496
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3108707960983432,
+      "learning_rate": 8.5826130760577e-06,
+      "loss": 0.526,
+      "step": 2497
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.234243575741267,
+      "learning_rate": 8.581398349068029e-06,
+      "loss": 0.5288,
+      "step": 2498
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.984974923682512,
+      "learning_rate": 8.58018318782146e-06,
+      "loss": 0.4987,
+      "step": 2499
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2811151191130223,
+      "learning_rate": 8.578967592465337e-06,
+      "loss": 0.4988,
+      "step": 2500
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5095922506956253,
+      "learning_rate": 8.577751563147055e-06,
+      "loss": 0.3079,
+      "step": 2501
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4506729264547559,
+      "learning_rate": 8.576535100014061e-06,
+      "loss": 0.3169,
+      "step": 2502
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6831130498604066,
+      "learning_rate": 8.575318203213856e-06,
+      "loss": 0.5002,
+      "step": 2503
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1839334348940342,
+      "learning_rate": 8.574100872893993e-06,
+      "loss": 0.5812,
+      "step": 2504
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1361361037093176,
+      "learning_rate": 8.572883109202077e-06,
+      "loss": 0.5757,
+      "step": 2505
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5048510748470264,
+      "learning_rate": 8.571664912285766e-06,
+      "loss": 0.3157,
+      "step": 2506
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8373759495158666,
+      "learning_rate": 8.570446282292772e-06,
+      "loss": 0.5229,
+      "step": 2507
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4921046050516344,
+      "learning_rate": 8.569227219370857e-06,
+      "loss": 0.5071,
+      "step": 2508
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.791291498837957,
+      "learning_rate": 8.568007723667837e-06,
+      "loss": 0.4915,
+      "step": 2509
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.951338388347533,
+      "learning_rate": 8.56678779533158e-06,
+      "loss": 0.499,
+      "step": 2510
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.210694059471011,
+      "learning_rate": 8.565567434510005e-06,
+      "loss": 0.5113,
+      "step": 2511
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6649148324618115,
+      "learning_rate": 8.564346641351087e-06,
+      "loss": 0.4928,
+      "step": 2512
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8603049875359217,
+      "learning_rate": 8.56312541600285e-06,
+      "loss": 0.3054,
+      "step": 2513
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.662407771499525,
+      "learning_rate": 8.561903758613372e-06,
+      "loss": 0.5352,
+      "step": 2514
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.984982206145066,
+      "learning_rate": 8.560681669330784e-06,
+      "loss": 0.4973,
+      "step": 2515
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.881663263150538,
+      "learning_rate": 8.559459148303268e-06,
+      "loss": 0.5453,
+      "step": 2516
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.3079229983515575,
+      "learning_rate": 8.55823619567906e-06,
+      "loss": 0.5479,
+      "step": 2517
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.569125095711984,
+      "learning_rate": 8.557012811606444e-06,
+      "loss": 0.4988,
+      "step": 2518
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7053128731692306,
+      "learning_rate": 8.555788996233765e-06,
+      "loss": 0.5018,
+      "step": 2519
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4365860443144516,
+      "learning_rate": 8.554564749709408e-06,
+      "loss": 0.3296,
+      "step": 2520
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.579777652329288,
+      "learning_rate": 8.553340072181823e-06,
+      "loss": 0.5306,
+      "step": 2521
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0996194779366117,
+      "learning_rate": 8.552114963799503e-06,
+      "loss": 0.4786,
+      "step": 2522
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8274867725361186,
+      "learning_rate": 8.550889424710998e-06,
+      "loss": 0.4946,
+      "step": 2523
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.684463471365382,
+      "learning_rate": 8.549663455064908e-06,
+      "loss": 0.5354,
+      "step": 2524
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2367017255574297,
+      "learning_rate": 8.548437055009886e-06,
+      "loss": 0.5226,
+      "step": 2525
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.702831718412543,
+      "learning_rate": 8.54721022469464e-06,
+      "loss": 0.494,
+      "step": 2526
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6214217804029025,
+      "learning_rate": 8.545982964267923e-06,
+      "loss": 0.5339,
+      "step": 2527
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.301623723220784,
+      "learning_rate": 8.544755273878547e-06,
+      "loss": 0.5056,
+      "step": 2528
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6273820056774437,
+      "learning_rate": 8.543527153675374e-06,
+      "loss": 0.5478,
+      "step": 2529
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.497784236355288,
+      "learning_rate": 8.542298603807317e-06,
+      "loss": 0.5706,
+      "step": 2530
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9581858614786167,
+      "learning_rate": 8.541069624423343e-06,
+      "loss": 0.535,
+      "step": 2531
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.17505308065528,
+      "learning_rate": 8.539840215672468e-06,
+      "loss": 0.5461,
+      "step": 2532
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8931217920704582,
+      "learning_rate": 8.538610377703764e-06,
+      "loss": 0.5093,
+      "step": 2533
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.305295688893197,
+      "learning_rate": 8.537380110666351e-06,
+      "loss": 0.526,
+      "step": 2534
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.031034557784237,
+      "learning_rate": 8.536149414709405e-06,
+      "loss": 0.534,
+      "step": 2535
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.8406955467824435,
+      "learning_rate": 8.534918289982152e-06,
+      "loss": 0.5455,
+      "step": 2536
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0475020905190346,
+      "learning_rate": 8.533686736633869e-06,
+      "loss": 0.5231,
+      "step": 2537
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.087451672929721,
+      "learning_rate": 8.532454754813888e-06,
+      "loss": 0.5576,
+      "step": 2538
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.43688483874793366,
+      "learning_rate": 8.531222344671588e-06,
+      "loss": 0.3022,
+      "step": 2539
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9418470469721845,
+      "learning_rate": 8.529989506356407e-06,
+      "loss": 0.552,
+      "step": 2540
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2532678043674212,
+      "learning_rate": 8.528756240017826e-06,
+      "loss": 0.5292,
+      "step": 2541
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4034332711345634,
+      "learning_rate": 8.527522545805386e-06,
+      "loss": 0.4867,
+      "step": 2542
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7758246612863346,
+      "learning_rate": 8.526288423868676e-06,
+      "loss": 0.5416,
+      "step": 2543
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6519383419375,
+      "learning_rate": 8.525053874357338e-06,
+      "loss": 0.553,
+      "step": 2544
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.969546324757399,
+      "learning_rate": 8.523818897421066e-06,
+      "loss": 0.5813,
+      "step": 2545
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8413797505926377,
+      "learning_rate": 8.522583493209605e-06,
+      "loss": 0.5669,
+      "step": 2546
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0516700077649945,
+      "learning_rate": 8.521347661872749e-06,
+      "loss": 0.4895,
+      "step": 2547
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.994734765963836,
+      "learning_rate": 8.520111403560349e-06,
+      "loss": 0.493,
+      "step": 2548
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.510587073522671,
+      "learning_rate": 8.518874718422307e-06,
+      "loss": 0.4814,
+      "step": 2549
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.984504019897253,
+      "learning_rate": 8.517637606608574e-06,
+      "loss": 0.492,
+      "step": 2550
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4158025677618681,
+      "learning_rate": 8.516400068269153e-06,
+      "loss": 0.3158,
+      "step": 2551
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.444090127390603,
+      "learning_rate": 8.515162103554102e-06,
+      "loss": 0.5156,
+      "step": 2552
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.061191752701168,
+      "learning_rate": 8.513923712613526e-06,
+      "loss": 0.5899,
+      "step": 2553
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3770191258825951,
+      "learning_rate": 8.512684895597587e-06,
+      "loss": 0.3037,
+      "step": 2554
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.057717068787728,
+      "learning_rate": 8.511445652656493e-06,
+      "loss": 0.5,
+      "step": 2555
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9457447652133166,
+      "learning_rate": 8.510205983940508e-06,
+      "loss": 0.5137,
+      "step": 2556
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1527240087075232,
+      "learning_rate": 8.508965889599948e-06,
+      "loss": 0.5027,
+      "step": 2557
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2200839663896987,
+      "learning_rate": 8.507725369785174e-06,
+      "loss": 0.5294,
+      "step": 2558
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2049608583642955,
+      "learning_rate": 8.506484424646606e-06,
+      "loss": 0.5686,
+      "step": 2559
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3215660080157634,
+      "learning_rate": 8.505243054334714e-06,
+      "loss": 0.5051,
+      "step": 2560
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.469221694700205,
+      "learning_rate": 8.504001259000017e-06,
+      "loss": 0.485,
+      "step": 2561
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.999295751158659,
+      "learning_rate": 8.502759038793086e-06,
+      "loss": 0.5598,
+      "step": 2562
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.500576794633765,
+      "learning_rate": 8.501516393864549e-06,
+      "loss": 0.4714,
+      "step": 2563
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.826484037188617,
+      "learning_rate": 8.500273324365074e-06,
+      "loss": 0.5058,
+      "step": 2564
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.17395208927558,
+      "learning_rate": 8.499029830445393e-06,
+      "loss": 0.5752,
+      "step": 2565
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2955384221925255,
+      "learning_rate": 8.497785912256282e-06,
+      "loss": 0.5029,
+      "step": 2566
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.277001737924126,
+      "learning_rate": 8.496541569948571e-06,
+      "loss": 0.5253,
+      "step": 2567
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.343754376772084,
+      "learning_rate": 8.495296803673138e-06,
+      "loss": 0.4882,
+      "step": 2568
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0188497662921328,
+      "learning_rate": 8.49405161358092e-06,
+      "loss": 0.5188,
+      "step": 2569
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.471513577020677,
+      "learning_rate": 8.492805999822898e-06,
+      "loss": 0.5658,
+      "step": 2570
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.586461005288051,
+      "learning_rate": 8.491559962550104e-06,
+      "loss": 0.5367,
+      "step": 2571
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.534518096524652,
+      "learning_rate": 8.49031350191363e-06,
+      "loss": 0.5599,
+      "step": 2572
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7811529946647884,
+      "learning_rate": 8.48906661806461e-06,
+      "loss": 0.5406,
+      "step": 2573
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9133331805344955,
+      "learning_rate": 8.487819311154233e-06,
+      "loss": 0.5304,
+      "step": 2574
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.970070693875789,
+      "learning_rate": 8.486571581333743e-06,
+      "loss": 0.5019,
+      "step": 2575
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4987821818798315,
+      "learning_rate": 8.485323428754427e-06,
+      "loss": 0.5169,
+      "step": 2576
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0135288875777007,
+      "learning_rate": 8.48407485356763e-06,
+      "loss": 0.5085,
+      "step": 2577
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.275540047355684,
+      "learning_rate": 8.482825855924743e-06,
+      "loss": 0.4777,
+      "step": 2578
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.65913170122677,
+      "learning_rate": 8.481576435977217e-06,
+      "loss": 0.5349,
+      "step": 2579
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4665334662580856,
+      "learning_rate": 8.480326593876545e-06,
+      "loss": 0.5183,
+      "step": 2580
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7311578063628112,
+      "learning_rate": 8.479076329774274e-06,
+      "loss": 0.3106,
+      "step": 2581
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.351754060578061,
+      "learning_rate": 8.477825643822003e-06,
+      "loss": 0.5333,
+      "step": 2582
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.950570655063019,
+      "learning_rate": 8.476574536171385e-06,
+      "loss": 0.5107,
+      "step": 2583
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.701279471814038,
+      "learning_rate": 8.475323006974116e-06,
+      "loss": 0.5469,
+      "step": 2584
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.246341412184899,
+      "learning_rate": 8.474071056381953e-06,
+      "loss": 0.4773,
+      "step": 2585
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6818458637257616,
+      "learning_rate": 8.472818684546697e-06,
+      "loss": 0.5465,
+      "step": 2586
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4239835082640813,
+      "learning_rate": 8.471565891620204e-06,
+      "loss": 0.527,
+      "step": 2587
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6744288217158023,
+      "learning_rate": 8.470312677754377e-06,
+      "loss": 0.4759,
+      "step": 2588
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.1061600826823925,
+      "learning_rate": 8.469059043101176e-06,
+      "loss": 0.5044,
+      "step": 2589
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.375456662694391,
+      "learning_rate": 8.467804987812603e-06,
+      "loss": 0.4793,
+      "step": 2590
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0234579412979183,
+      "learning_rate": 8.466550512040722e-06,
+      "loss": 0.4966,
+      "step": 2591
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.516570053075619,
+      "learning_rate": 8.465295615937641e-06,
+      "loss": 0.4794,
+      "step": 2592
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.229777188149992,
+      "learning_rate": 8.464040299655518e-06,
+      "loss": 0.4529,
+      "step": 2593
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3360145058006907,
+      "learning_rate": 8.462784563346567e-06,
+      "loss": 0.4799,
+      "step": 2594
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.688452548325376,
+      "learning_rate": 8.461528407163051e-06,
+      "loss": 0.5621,
+      "step": 2595
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.47516758955484084,
+      "learning_rate": 8.46027183125728e-06,
+      "loss": 0.321,
+      "step": 2596
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9161358892402864,
+      "learning_rate": 8.459014835781622e-06,
+      "loss": 0.5134,
+      "step": 2597
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.050325147251919,
+      "learning_rate": 8.457757420888487e-06,
+      "loss": 0.4534,
+      "step": 2598
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2905255794486923,
+      "learning_rate": 8.456499586730346e-06,
+      "loss": 0.5212,
+      "step": 2599
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1657900130167866,
+      "learning_rate": 8.455241333459714e-06,
+      "loss": 0.5217,
+      "step": 2600
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.5135217134024765,
+      "learning_rate": 8.453982661229158e-06,
+      "loss": 0.5448,
+      "step": 2601
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.091310075195334,
+      "learning_rate": 8.452723570191298e-06,
+      "loss": 0.5194,
+      "step": 2602
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.509325040831739,
+      "learning_rate": 8.4514640604988e-06,
+      "loss": 0.5344,
+      "step": 2603
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.456342148350291,
+      "learning_rate": 8.450204132304387e-06,
+      "loss": 0.4635,
+      "step": 2604
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 29.125862537263423,
+      "learning_rate": 8.448943785760827e-06,
+      "loss": 0.4989,
+      "step": 2605
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7915134525996894,
+      "learning_rate": 8.447683021020943e-06,
+      "loss": 0.5287,
+      "step": 2606
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2657381280906406,
+      "learning_rate": 8.446421838237606e-06,
+      "loss": 0.5078,
+      "step": 2607
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.488641713276766,
+      "learning_rate": 8.445160237563742e-06,
+      "loss": 0.5444,
+      "step": 2608
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2575921171983406,
+      "learning_rate": 8.44389821915232e-06,
+      "loss": 0.5439,
+      "step": 2609
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.827277673501762,
+      "learning_rate": 8.442635783156365e-06,
+      "loss": 0.4764,
+      "step": 2610
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 15.949275166921552,
+      "learning_rate": 8.441372929728955e-06,
+      "loss": 0.4972,
+      "step": 2611
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.44770385689576575,
+      "learning_rate": 8.44010965902321e-06,
+      "loss": 0.3214,
+      "step": 2612
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.577526618365235,
+      "learning_rate": 8.438845971192313e-06,
+      "loss": 0.5382,
+      "step": 2613
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7257978354252423,
+      "learning_rate": 8.437581866389483e-06,
+      "loss": 0.5586,
+      "step": 2614
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.817095802828004,
+      "learning_rate": 8.436317344768004e-06,
+      "loss": 0.544,
+      "step": 2615
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.998215186568939,
+      "learning_rate": 8.435052406481196e-06,
+      "loss": 0.5183,
+      "step": 2616
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.631805534613986,
+      "learning_rate": 8.433787051682444e-06,
+      "loss": 0.5362,
+      "step": 2617
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9871332170934064,
+      "learning_rate": 8.432521280525174e-06,
+      "loss": 0.5214,
+      "step": 2618
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9315265967325894,
+      "learning_rate": 8.431255093162864e-06,
+      "loss": 0.5519,
+      "step": 2619
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.118403326020087,
+      "learning_rate": 8.429988489749045e-06,
+      "loss": 0.499,
+      "step": 2620
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2914028451723767,
+      "learning_rate": 8.428721470437296e-06,
+      "loss": 0.5405,
+      "step": 2621
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.366139826895146,
+      "learning_rate": 8.42745403538125e-06,
+      "loss": 0.4825,
+      "step": 2622
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.445402817426533,
+      "learning_rate": 8.426186184734585e-06,
+      "loss": 0.298,
+      "step": 2623
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.107681184592305,
+      "learning_rate": 8.424917918651032e-06,
+      "loss": 0.4946,
+      "step": 2624
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2693223312332402,
+      "learning_rate": 8.423649237284375e-06,
+      "loss": 0.5254,
+      "step": 2625
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.509012910784757,
+      "learning_rate": 8.422380140788446e-06,
+      "loss": 0.5574,
+      "step": 2626
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9630209122115843,
+      "learning_rate": 8.421110629317123e-06,
+      "loss": 0.5055,
+      "step": 2627
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.661396434962075,
+      "learning_rate": 8.419840703024344e-06,
+      "loss": 0.5145,
+      "step": 2628
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.321531539088039,
+      "learning_rate": 8.418570362064092e-06,
+      "loss": 0.5435,
+      "step": 2629
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.823946933568148,
+      "learning_rate": 8.417299606590394e-06,
+      "loss": 0.5403,
+      "step": 2630
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.925315743914989,
+      "learning_rate": 8.41602843675734e-06,
+      "loss": 0.5078,
+      "step": 2631
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.517751105717801,
+      "learning_rate": 8.41475685271906e-06,
+      "loss": 0.5025,
+      "step": 2632
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.311029409885593,
+      "learning_rate": 8.413484854629738e-06,
+      "loss": 0.4925,
+      "step": 2633
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.030094100707498,
+      "learning_rate": 8.412212442643611e-06,
+      "loss": 0.4968,
+      "step": 2634
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8234892286986364,
+      "learning_rate": 8.41093961691496e-06,
+      "loss": 0.5355,
+      "step": 2635
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.370246542676693,
+      "learning_rate": 8.409666377598125e-06,
+      "loss": 0.4771,
+      "step": 2636
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.572774399527857,
+      "learning_rate": 8.408392724847483e-06,
+      "loss": 0.5228,
+      "step": 2637
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.897475201913736,
+      "learning_rate": 8.407118658817475e-06,
+      "loss": 0.562,
+      "step": 2638
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9966793349079714,
+      "learning_rate": 8.405844179662582e-06,
+      "loss": 0.5667,
+      "step": 2639
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3378921963673047,
+      "learning_rate": 8.40456928753734e-06,
+      "loss": 0.515,
+      "step": 2640
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5266761037044181,
+      "learning_rate": 8.403293982596336e-06,
+      "loss": 0.3257,
+      "step": 2641
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.220325362879772,
+      "learning_rate": 8.402018264994203e-06,
+      "loss": 0.5608,
+      "step": 2642
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3663995821477166,
+      "learning_rate": 8.400742134885627e-06,
+      "loss": 0.4956,
+      "step": 2643
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1135168188246882,
+      "learning_rate": 8.399465592425342e-06,
+      "loss": 0.4996,
+      "step": 2644
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4237479408067655,
+      "learning_rate": 8.398188637768136e-06,
+      "loss": 0.473,
+      "step": 2645
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.858239688291308,
+      "learning_rate": 8.396911271068842e-06,
+      "loss": 0.5048,
+      "step": 2646
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.240558343804802,
+      "learning_rate": 8.395633492482345e-06,
+      "loss": 0.5361,
+      "step": 2647
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 16.803164690999342,
+      "learning_rate": 8.394355302163578e-06,
+      "loss": 0.5199,
+      "step": 2648
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8359433919614503,
+      "learning_rate": 8.393076700267534e-06,
+      "loss": 0.512,
+      "step": 2649
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8126235217469904,
+      "learning_rate": 8.391797686949238e-06,
+      "loss": 0.5579,
+      "step": 2650
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.42746176513512235,
+      "learning_rate": 8.390518262363781e-06,
+      "loss": 0.2854,
+      "step": 2651
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.6850545486102355,
+      "learning_rate": 8.389238426666294e-06,
+      "loss": 0.4738,
+      "step": 2652
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.397289411473436,
+      "learning_rate": 8.387958180011966e-06,
+      "loss": 0.545,
+      "step": 2653
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.387984799744047,
+      "learning_rate": 8.386677522556026e-06,
+      "loss": 0.5104,
+      "step": 2654
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.535111287090148,
+      "learning_rate": 8.385396454453763e-06,
+      "loss": 0.5319,
+      "step": 2655
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0572888947096177,
+      "learning_rate": 8.384114975860507e-06,
+      "loss": 0.4878,
+      "step": 2656
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.013257861799297,
+      "learning_rate": 8.382833086931643e-06,
+      "loss": 0.5378,
+      "step": 2657
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.119798912453484,
+      "learning_rate": 8.381550787822606e-06,
+      "loss": 0.5521,
+      "step": 2658
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.348804415737593,
+      "learning_rate": 8.380268078688877e-06,
+      "loss": 0.5323,
+      "step": 2659
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.38311343426024635,
+      "learning_rate": 8.378984959685992e-06,
+      "loss": 0.3174,
+      "step": 2660
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41052346252774985,
+      "learning_rate": 8.377701430969529e-06,
+      "loss": 0.2903,
+      "step": 2661
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1865408313755736,
+      "learning_rate": 8.376417492695123e-06,
+      "loss": 0.5366,
+      "step": 2662
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.49847452959013,
+      "learning_rate": 8.375133145018457e-06,
+      "loss": 0.5567,
+      "step": 2663
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.06910035912452,
+      "learning_rate": 8.37384838809526e-06,
+      "loss": 0.5017,
+      "step": 2664
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.514386129389434,
+      "learning_rate": 8.372563222081316e-06,
+      "loss": 0.5774,
+      "step": 2665
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.774673693995662,
+      "learning_rate": 8.371277647132454e-06,
+      "loss": 0.5285,
+      "step": 2666
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.487796373968522,
+      "learning_rate": 8.369991663404556e-06,
+      "loss": 0.5123,
+      "step": 2667
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.379416539817732,
+      "learning_rate": 8.368705271053547e-06,
+      "loss": 0.5404,
+      "step": 2668
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9514175701305496,
+      "learning_rate": 8.367418470235412e-06,
+      "loss": 0.5305,
+      "step": 2669
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.575835543727615,
+      "learning_rate": 8.366131261106179e-06,
+      "loss": 0.4359,
+      "step": 2670
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5576107962182957,
+      "learning_rate": 8.364843643821928e-06,
+      "loss": 0.5102,
+      "step": 2671
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.090162414025343,
+      "learning_rate": 8.36355561853878e-06,
+      "loss": 0.5619,
+      "step": 2672
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4676952913223653,
+      "learning_rate": 8.36226718541292e-06,
+      "loss": 0.5375,
+      "step": 2673
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.765201112254755,
+      "learning_rate": 8.360978344600572e-06,
+      "loss": 0.537,
+      "step": 2674
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 16.069041548330958,
+      "learning_rate": 8.35968909625801e-06,
+      "loss": 0.4953,
+      "step": 2675
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8357767494262944,
+      "learning_rate": 8.358399440541567e-06,
+      "loss": 0.4851,
+      "step": 2676
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4960666566819258,
+      "learning_rate": 8.357109377607611e-06,
+      "loss": 0.3107,
+      "step": 2677
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.854038896032166,
+      "learning_rate": 8.355818907612569e-06,
+      "loss": 0.5318,
+      "step": 2678
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.564168156203698,
+      "learning_rate": 8.354528030712916e-06,
+      "loss": 0.5185,
+      "step": 2679
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.052173426509063,
+      "learning_rate": 8.353236747065175e-06,
+      "loss": 0.5231,
+      "step": 2680
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.260254207757985,
+      "learning_rate": 8.351945056825918e-06,
+      "loss": 0.5161,
+      "step": 2681
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.475141413303512,
+      "learning_rate": 8.350652960151766e-06,
+      "loss": 0.5475,
+      "step": 2682
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 20.837299017596727,
+      "learning_rate": 8.349360457199392e-06,
+      "loss": 0.4716,
+      "step": 2683
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.140902529122034,
+      "learning_rate": 8.348067548125515e-06,
+      "loss": 0.5035,
+      "step": 2684
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8219721711603882,
+      "learning_rate": 8.346774233086906e-06,
+      "loss": 0.5118,
+      "step": 2685
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.052198394115833,
+      "learning_rate": 8.34548051224038e-06,
+      "loss": 0.5161,
+      "step": 2686
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5301679483027484,
+      "learning_rate": 8.344186385742811e-06,
+      "loss": 0.5044,
+      "step": 2687
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.224786604228409,
+      "learning_rate": 8.342891853751114e-06,
+      "loss": 0.507,
+      "step": 2688
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8798372675301214,
+      "learning_rate": 8.341596916422255e-06,
+      "loss": 0.5313,
+      "step": 2689
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8305959197263286,
+      "learning_rate": 8.34030157391325e-06,
+      "loss": 0.4919,
+      "step": 2690
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.48111571551968485,
+      "learning_rate": 8.33900582638116e-06,
+      "loss": 0.3258,
+      "step": 2691
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.8044590089143835,
+      "learning_rate": 8.337709673983106e-06,
+      "loss": 0.4826,
+      "step": 2692
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.247972020274013,
+      "learning_rate": 8.336413116876245e-06,
+      "loss": 0.4453,
+      "step": 2693
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2727169360597013,
+      "learning_rate": 8.335116155217793e-06,
+      "loss": 0.5234,
+      "step": 2694
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2134254946360965,
+      "learning_rate": 8.333818789165009e-06,
+      "loss": 0.5437,
+      "step": 2695
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.612988092130531,
+      "learning_rate": 8.332521018875202e-06,
+      "loss": 0.5632,
+      "step": 2696
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2327694595683965,
+      "learning_rate": 8.331222844505732e-06,
+      "loss": 0.5372,
+      "step": 2697
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.453026330068263,
+      "learning_rate": 8.32992426621401e-06,
+      "loss": 0.4803,
+      "step": 2698
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2114590457743413,
+      "learning_rate": 8.32862528415749e-06,
+      "loss": 0.5024,
+      "step": 2699
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.163779059950875,
+      "learning_rate": 8.327325898493677e-06,
+      "loss": 0.5261,
+      "step": 2700
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.44081393114801054,
+      "learning_rate": 8.32602610938013e-06,
+      "loss": 0.323,
+      "step": 2701
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.499445740072782,
+      "learning_rate": 8.32472591697445e-06,
+      "loss": 0.4918,
+      "step": 2702
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.333308105081914,
+      "learning_rate": 8.323425321434291e-06,
+      "loss": 0.5506,
+      "step": 2703
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2318140724179214,
+      "learning_rate": 8.322124322917354e-06,
+      "loss": 0.5204,
+      "step": 2704
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.5173589307124775,
+      "learning_rate": 8.320822921581388e-06,
+      "loss": 0.5081,
+      "step": 2705
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.718036547445374,
+      "learning_rate": 8.319521117584195e-06,
+      "loss": 0.4657,
+      "step": 2706
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5427999567584774,
+      "learning_rate": 8.318218911083622e-06,
+      "loss": 0.4898,
+      "step": 2707
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.059215765228931,
+      "learning_rate": 8.316916302237568e-06,
+      "loss": 0.5401,
+      "step": 2708
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4500069808855749,
+      "learning_rate": 8.315613291203977e-06,
+      "loss": 0.3096,
+      "step": 2709
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3354458783369845,
+      "learning_rate": 8.314309878140843e-06,
+      "loss": 0.5578,
+      "step": 2710
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.33395489344075,
+      "learning_rate": 8.31300606320621e-06,
+      "loss": 0.5422,
+      "step": 2711
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9600755413920594,
+      "learning_rate": 8.311701846558172e-06,
+      "loss": 0.5211,
+      "step": 2712
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4235849290379934,
+      "learning_rate": 8.310397228354864e-06,
+      "loss": 0.3249,
+      "step": 2713
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.249667434770252,
+      "learning_rate": 8.309092208754483e-06,
+      "loss": 0.5412,
+      "step": 2714
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.60590481692282,
+      "learning_rate": 8.307786787915262e-06,
+      "loss": 0.4906,
+      "step": 2715
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1751061153570888,
+      "learning_rate": 8.306480965995488e-06,
+      "loss": 0.5485,
+      "step": 2716
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5166103934675745,
+      "learning_rate": 8.3051747431535e-06,
+      "loss": 0.5122,
+      "step": 2717
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.753713257194252,
+      "learning_rate": 8.30386811954768e-06,
+      "loss": 0.5065,
+      "step": 2718
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6181416937865047,
+      "learning_rate": 8.30256109533646e-06,
+      "loss": 0.4976,
+      "step": 2719
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.085279599797346,
+      "learning_rate": 8.30125367067832e-06,
+      "loss": 0.572,
+      "step": 2720
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.939377649036369,
+      "learning_rate": 8.29994584573179e-06,
+      "loss": 0.5404,
+      "step": 2721
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.056625080048355,
+      "learning_rate": 8.298637620655454e-06,
+      "loss": 0.5204,
+      "step": 2722
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.675451424624829,
+      "learning_rate": 8.297328995607931e-06,
+      "loss": 0.5122,
+      "step": 2723
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.7995804961382027,
+      "learning_rate": 8.296019970747901e-06,
+      "loss": 0.5282,
+      "step": 2724
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.534385408588305,
+      "learning_rate": 8.294710546234086e-06,
+      "loss": 0.4866,
+      "step": 2725
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.311466618216869,
+      "learning_rate": 8.29340072222526e-06,
+      "loss": 0.4918,
+      "step": 2726
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.43417086509562913,
+      "learning_rate": 8.292090498880242e-06,
+      "loss": 0.3093,
+      "step": 2727
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6195138371058935,
+      "learning_rate": 8.2907798763579e-06,
+      "loss": 0.4635,
+      "step": 2728
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.7672888748647813,
+      "learning_rate": 8.289468854817153e-06,
+      "loss": 0.565,
+      "step": 2729
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4301694795611026,
+      "learning_rate": 8.288157434416966e-06,
+      "loss": 0.3111,
+      "step": 2730
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.918308941895195,
+      "learning_rate": 8.286845615316357e-06,
+      "loss": 0.5137,
+      "step": 2731
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.04441427022253,
+      "learning_rate": 8.285533397674384e-06,
+      "loss": 0.4671,
+      "step": 2732
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.388864511068113,
+      "learning_rate": 8.284220781650158e-06,
+      "loss": 0.5322,
+      "step": 2733
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.745412097518173,
+      "learning_rate": 8.282907767402842e-06,
+      "loss": 0.524,
+      "step": 2734
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.936903795138083,
+      "learning_rate": 8.281594355091642e-06,
+      "loss": 0.5208,
+      "step": 2735
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4528967765544048,
+      "learning_rate": 8.280280544875812e-06,
+      "loss": 0.3062,
+      "step": 2736
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.152540426291802,
+      "learning_rate": 8.278966336914656e-06,
+      "loss": 0.5046,
+      "step": 2737
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4848188635837105,
+      "learning_rate": 8.277651731367528e-06,
+      "loss": 0.5203,
+      "step": 2738
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.546631421625592,
+      "learning_rate": 8.276336728393827e-06,
+      "loss": 0.5353,
+      "step": 2739
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7352224634183515,
+      "learning_rate": 8.275021328153005e-06,
+      "loss": 0.5796,
+      "step": 2740
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.042839425041572,
+      "learning_rate": 8.273705530804554e-06,
+      "loss": 0.5564,
+      "step": 2741
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.107439450989317,
+      "learning_rate": 8.272389336508022e-06,
+      "loss": 0.5488,
+      "step": 2742
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9974873632238324,
+      "learning_rate": 8.271072745423001e-06,
+      "loss": 0.5452,
+      "step": 2743
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3881048642787808,
+      "learning_rate": 8.269755757709132e-06,
+      "loss": 0.5361,
+      "step": 2744
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.39217759696760474,
+      "learning_rate": 8.268438373526106e-06,
+      "loss": 0.2901,
+      "step": 2745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4258663566226484,
+      "learning_rate": 8.267120593033659e-06,
+      "loss": 0.4859,
+      "step": 2746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.169544072142267,
+      "learning_rate": 8.265802416391577e-06,
+      "loss": 0.4981,
+      "step": 2747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8866360191202682,
+      "learning_rate": 8.264483843759692e-06,
+      "loss": 0.5506,
+      "step": 2748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.42804002362977045,
+      "learning_rate": 8.263164875297888e-06,
+      "loss": 0.3169,
+      "step": 2749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2101901687384684,
+      "learning_rate": 8.261845511166092e-06,
+      "loss": 0.541,
+      "step": 2750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.727425415860383,
+      "learning_rate": 8.260525751524282e-06,
+      "loss": 0.4755,
+      "step": 2751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.124944978667238,
+      "learning_rate": 8.259205596532485e-06,
+      "loss": 0.5138,
+      "step": 2752
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.226824851262858,
+      "learning_rate": 8.257885046350773e-06,
+      "loss": 0.4939,
+      "step": 2753
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.734771875903018,
+      "learning_rate": 8.256564101139266e-06,
+      "loss": 0.5144,
+      "step": 2754
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8006484341169213,
+      "learning_rate": 8.255242761058136e-06,
+      "loss": 0.5325,
+      "step": 2755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.455035860234755,
+      "learning_rate": 8.2539210262676e-06,
+      "loss": 0.5458,
+      "step": 2756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2296912330923147,
+      "learning_rate": 8.252598896927918e-06,
+      "loss": 0.5007,
+      "step": 2757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.060237919961863,
+      "learning_rate": 8.25127637319941e-06,
+      "loss": 0.5348,
+      "step": 2758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.079517372861008,
+      "learning_rate": 8.24995345524243e-06,
+      "loss": 0.5443,
+      "step": 2759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.463779247531833,
+      "learning_rate": 8.248630143217391e-06,
+      "loss": 0.5178,
+      "step": 2760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2399483039517754,
+      "learning_rate": 8.247306437284748e-06,
+      "loss": 0.6229,
+      "step": 2761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4122379596665154,
+      "learning_rate": 8.245982337605003e-06,
+      "loss": 0.3021,
+      "step": 2762
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.1993867415051085,
+      "learning_rate": 8.244657844338708e-06,
+      "loss": 0.5086,
+      "step": 2763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3861184141155394,
+      "learning_rate": 8.243332957646465e-06,
+      "loss": 0.483,
+      "step": 2764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.849037033331813,
+      "learning_rate": 8.242007677688919e-06,
+      "loss": 0.5516,
+      "step": 2765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.082743434865341,
+      "learning_rate": 8.240682004626765e-06,
+      "loss": 0.5302,
+      "step": 2766
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 14.713257163411297,
+      "learning_rate": 8.239355938620746e-06,
+      "loss": 0.5375,
+      "step": 2767
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.126940389496077,
+      "learning_rate": 8.238029479831651e-06,
+      "loss": 0.504,
+      "step": 2768
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.820299315827853,
+      "learning_rate": 8.236702628420319e-06,
+      "loss": 0.5111,
+      "step": 2769
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1134196106998226,
+      "learning_rate": 8.235375384547634e-06,
+      "loss": 0.4812,
+      "step": 2770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.929362652626849,
+      "learning_rate": 8.23404774837453e-06,
+      "loss": 0.4765,
+      "step": 2771
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.437047784540283,
+      "learning_rate": 8.232719720061989e-06,
+      "loss": 0.502,
+      "step": 2772
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.39098939315950365,
+      "learning_rate": 8.231391299771034e-06,
+      "loss": 0.2841,
+      "step": 2773
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.890554402783044,
+      "learning_rate": 8.230062487662745e-06,
+      "loss": 0.5094,
+      "step": 2774
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.370154172655842,
+      "learning_rate": 8.228733283898244e-06,
+      "loss": 0.5082,
+      "step": 2775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.44475722367629,
+      "learning_rate": 8.2274036886387e-06,
+      "loss": 0.5323,
+      "step": 2776
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.551734985731164,
+      "learning_rate": 8.226073702045334e-06,
+      "loss": 0.4663,
+      "step": 2777
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.42609756391576764,
+      "learning_rate": 8.224743324279407e-06,
+      "loss": 0.3185,
+      "step": 2778
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.283347376887616,
+      "learning_rate": 8.223412555502238e-06,
+      "loss": 0.5706,
+      "step": 2779
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.7370188060395146,
+      "learning_rate": 8.222081395875182e-06,
+      "loss": 0.4964,
+      "step": 2780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5835633411661845,
+      "learning_rate": 8.220749845559648e-06,
+      "loss": 0.5285,
+      "step": 2781
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6822469572169614,
+      "learning_rate": 8.219417904717091e-06,
+      "loss": 0.5358,
+      "step": 2782
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.479801552159427,
+      "learning_rate": 8.218085573509016e-06,
+      "loss": 0.5856,
+      "step": 2783
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.861238914885516,
+      "learning_rate": 8.21675285209697e-06,
+      "loss": 0.5344,
+      "step": 2784
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.467186395000102,
+      "learning_rate": 8.21541974064255e-06,
+      "loss": 0.5142,
+      "step": 2785
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.044983936618752,
+      "learning_rate": 8.214086239307402e-06,
+      "loss": 0.5385,
+      "step": 2786
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8050052065091453,
+      "learning_rate": 8.212752348253216e-06,
+      "loss": 0.5656,
+      "step": 2787
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.376452470234331,
+      "learning_rate": 8.211418067641734e-06,
+      "loss": 0.5171,
+      "step": 2788
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8641771034413344,
+      "learning_rate": 8.210083397634738e-06,
+      "loss": 0.5509,
+      "step": 2789
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.898341415071076,
+      "learning_rate": 8.208748338394063e-06,
+      "loss": 0.4985,
+      "step": 2790
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5944472045082763,
+      "learning_rate": 8.20741289008159e-06,
+      "loss": 0.5395,
+      "step": 2791
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2437441539769636,
+      "learning_rate": 8.206077052859247e-06,
+      "loss": 0.5213,
+      "step": 2792
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8445115787171296,
+      "learning_rate": 8.204740826889009e-06,
+      "loss": 0.5309,
+      "step": 2793
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6843940345311634,
+      "learning_rate": 8.203404212332897e-06,
+      "loss": 0.5295,
+      "step": 2794
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4015983813274479,
+      "learning_rate": 8.20206720935298e-06,
+      "loss": 0.3085,
+      "step": 2795
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.244466433860968,
+      "learning_rate": 8.200729818111372e-06,
+      "loss": 0.4804,
+      "step": 2796
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.445695871794228,
+      "learning_rate": 8.199392038770243e-06,
+      "loss": 0.5161,
+      "step": 2797
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5313009571421774,
+      "learning_rate": 8.198053871491798e-06,
+      "loss": 0.505,
+      "step": 2798
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3414309732777023,
+      "learning_rate": 8.196715316438295e-06,
+      "loss": 0.4639,
+      "step": 2799
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.842143530092985,
+      "learning_rate": 8.19537637377204e-06,
+      "loss": 0.5387,
+      "step": 2800
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.154881746491387,
+      "learning_rate": 8.194037043655383e-06,
+      "loss": 0.5087,
+      "step": 2801
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.659913998904033,
+      "learning_rate": 8.192697326250722e-06,
+      "loss": 0.4877,
+      "step": 2802
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.666908532225186,
+      "learning_rate": 8.191357221720506e-06,
+      "loss": 0.541,
+      "step": 2803
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5938683922842465,
+      "learning_rate": 8.190016730227224e-06,
+      "loss": 0.5422,
+      "step": 2804
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.012902408532952,
+      "learning_rate": 8.188675851933414e-06,
+      "loss": 0.5586,
+      "step": 2805
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.373191699353745,
+      "learning_rate": 8.187334587001664e-06,
+      "loss": 0.535,
+      "step": 2806
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3966086503873595,
+      "learning_rate": 8.185992935594608e-06,
+      "loss": 0.3349,
+      "step": 2807
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.359538112049303,
+      "learning_rate": 8.184650897874924e-06,
+      "loss": 0.5308,
+      "step": 2808
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.342265693556316,
+      "learning_rate": 8.18330847400534e-06,
+      "loss": 0.5204,
+      "step": 2809
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.802576745410127,
+      "learning_rate": 8.181965664148629e-06,
+      "loss": 0.5304,
+      "step": 2810
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9469221089155138,
+      "learning_rate": 8.180622468467612e-06,
+      "loss": 0.4961,
+      "step": 2811
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7725898409097556,
+      "learning_rate": 8.179278887125152e-06,
+      "loss": 0.5302,
+      "step": 2812
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.40069033065455284,
+      "learning_rate": 8.17793492028417e-06,
+      "loss": 0.2962,
+      "step": 2813
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8665725727161133,
+      "learning_rate": 8.176590568107623e-06,
+      "loss": 0.4784,
+      "step": 2814
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.552868667997811,
+      "learning_rate": 8.175245830758516e-06,
+      "loss": 0.5212,
+      "step": 2815
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4357247915508466,
+      "learning_rate": 8.173900708399907e-06,
+      "loss": 0.4863,
+      "step": 2816
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0773430691040073,
+      "learning_rate": 8.172555201194895e-06,
+      "loss": 0.5054,
+      "step": 2817
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2842489473237406,
+      "learning_rate": 8.171209309306626e-06,
+      "loss": 0.4649,
+      "step": 2818
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.160699628446329,
+      "learning_rate": 8.169863032898297e-06,
+      "loss": 0.5208,
+      "step": 2819
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.451497620663555,
+      "learning_rate": 8.168516372133145e-06,
+      "loss": 0.5266,
+      "step": 2820
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.7111847947061394,
+      "learning_rate": 8.16716932717446e-06,
+      "loss": 0.5177,
+      "step": 2821
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.372195237433274,
+      "learning_rate": 8.165821898185577e-06,
+      "loss": 0.5782,
+      "step": 2822
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4468779430770726,
+      "learning_rate": 8.164474085329872e-06,
+      "loss": 0.4881,
+      "step": 2823
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4215241490859605,
+      "learning_rate": 8.163125888770777e-06,
+      "loss": 0.3089,
+      "step": 2824
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.840683544301137,
+      "learning_rate": 8.161777308671763e-06,
+      "loss": 0.5141,
+      "step": 2825
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.90779238293943,
+      "learning_rate": 8.160428345196347e-06,
+      "loss": 0.5436,
+      "step": 2826
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9079031142022025,
+      "learning_rate": 8.159078998508101e-06,
+      "loss": 0.5,
+      "step": 2827
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 24.60881664192593,
+      "learning_rate": 8.157729268770636e-06,
+      "loss": 0.5437,
+      "step": 2828
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.412349669406183,
+      "learning_rate": 8.156379156147608e-06,
+      "loss": 0.5572,
+      "step": 2829
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.407325108941307,
+      "learning_rate": 8.155028660802729e-06,
+      "loss": 0.5361,
+      "step": 2830
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.8115858799959295,
+      "learning_rate": 8.153677782899744e-06,
+      "loss": 0.5322,
+      "step": 2831
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.411797244824187,
+      "learning_rate": 8.152326522602458e-06,
+      "loss": 0.5457,
+      "step": 2832
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3840055934828646,
+      "learning_rate": 8.150974880074714e-06,
+      "loss": 0.291,
+      "step": 2833
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.794726293678207,
+      "learning_rate": 8.149622855480401e-06,
+      "loss": 0.5226,
+      "step": 2834
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.7273484693499683,
+      "learning_rate": 8.14827044898346e-06,
+      "loss": 0.5263,
+      "step": 2835
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.988512250496304,
+      "learning_rate": 8.146917660747872e-06,
+      "loss": 0.5404,
+      "step": 2836
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.193393160371053,
+      "learning_rate": 8.145564490937669e-06,
+      "loss": 0.4785,
+      "step": 2837
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.6850145816296043,
+      "learning_rate": 8.144210939716929e-06,
+      "loss": 0.5069,
+      "step": 2838
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.793450376363878,
+      "learning_rate": 8.142857007249769e-06,
+      "loss": 0.4718,
+      "step": 2839
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4215833763917427,
+      "learning_rate": 8.141502693700364e-06,
+      "loss": 0.57,
+      "step": 2840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5245066019645788,
+      "learning_rate": 8.140147999232925e-06,
+      "loss": 0.4974,
+      "step": 2841
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.156394379162245,
+      "learning_rate": 8.138792924011719e-06,
+      "loss": 0.5642,
+      "step": 2842
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.6088108226921345,
+      "learning_rate": 8.137437468201047e-06,
+      "loss": 0.5094,
+      "step": 2843
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.42340529793793874,
+      "learning_rate": 8.136081631965268e-06,
+      "loss": 0.3168,
+      "step": 2844
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.649915823026669,
+      "learning_rate": 8.134725415468777e-06,
+      "loss": 0.4542,
+      "step": 2845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.414290605927806,
+      "learning_rate": 8.133368818876024e-06,
+      "loss": 0.5066,
+      "step": 2846
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4066290084888817,
+      "learning_rate": 8.132011842351497e-06,
+      "loss": 0.306,
+      "step": 2847
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.761236435195732,
+      "learning_rate": 8.130654486059739e-06,
+      "loss": 0.4991,
+      "step": 2848
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.6594608002777385,
+      "learning_rate": 8.12929675016533e-06,
+      "loss": 0.5345,
+      "step": 2849
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.26697343435991,
+      "learning_rate": 8.127938634832901e-06,
+      "loss": 0.5204,
+      "step": 2850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.180804475198937,
+      "learning_rate": 8.12658014022713e-06,
+      "loss": 0.5811,
+      "step": 2851
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.879150518277116,
+      "learning_rate": 8.125221266512739e-06,
+      "loss": 0.531,
+      "step": 2852
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.55884590379252,
+      "learning_rate": 8.123862013854496e-06,
+      "loss": 0.5285,
+      "step": 2853
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.655468963783866,
+      "learning_rate": 8.122502382417211e-06,
+      "loss": 0.4925,
+      "step": 2854
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.327498712739696,
+      "learning_rate": 8.121142372365749e-06,
+      "loss": 0.5519,
+      "step": 2855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.468005254753047,
+      "learning_rate": 8.119781983865014e-06,
+      "loss": 0.4918,
+      "step": 2856
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.733668877398715,
+      "learning_rate": 8.118421217079958e-06,
+      "loss": 0.4712,
+      "step": 2857
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.360534244856767,
+      "learning_rate": 8.117060072175577e-06,
+      "loss": 0.5368,
+      "step": 2858
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5254512973053326,
+      "learning_rate": 8.11569854931692e-06,
+      "loss": 0.3102,
+      "step": 2859
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.845367795087082,
+      "learning_rate": 8.114336648669069e-06,
+      "loss": 0.5333,
+      "step": 2860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.561690275617472,
+      "learning_rate": 8.112974370397163e-06,
+      "loss": 0.5754,
+      "step": 2861
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.63640916691978,
+      "learning_rate": 8.111611714666382e-06,
+      "loss": 0.4935,
+      "step": 2862
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.490679922906,
+      "learning_rate": 8.110248681641957e-06,
+      "loss": 0.5173,
+      "step": 2863
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.91281344182185,
+      "learning_rate": 8.108885271489152e-06,
+      "loss": 0.5268,
+      "step": 2864
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.903050207838458,
+      "learning_rate": 8.107521484373293e-06,
+      "loss": 0.4528,
+      "step": 2865
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7738186504411586,
+      "learning_rate": 8.10615732045974e-06,
+      "loss": 0.4809,
+      "step": 2866
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.946188500907695,
+      "learning_rate": 8.104792779913903e-06,
+      "loss": 0.5318,
+      "step": 2867
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.976003648256952,
+      "learning_rate": 8.103427862901237e-06,
+      "loss": 0.5026,
+      "step": 2868
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.40645622320733327,
+      "learning_rate": 8.102062569587245e-06,
+      "loss": 0.3115,
+      "step": 2869
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 33.20346051324591,
+      "learning_rate": 8.10069690013747e-06,
+      "loss": 0.5281,
+      "step": 2870
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.198240662073213,
+      "learning_rate": 8.099330854717507e-06,
+      "loss": 0.5208,
+      "step": 2871
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.40642246947305855,
+      "learning_rate": 8.097964433492995e-06,
+      "loss": 0.3027,
+      "step": 2872
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.443448887767153,
+      "learning_rate": 8.096597636629612e-06,
+      "loss": 0.4806,
+      "step": 2873
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7265288506101557,
+      "learning_rate": 8.09523046429309e-06,
+      "loss": 0.4832,
+      "step": 2874
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4051250743407618,
+      "learning_rate": 8.093862916649206e-06,
+      "loss": 0.3191,
+      "step": 2875
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.083526158204868,
+      "learning_rate": 8.092494993863775e-06,
+      "loss": 0.5128,
+      "step": 2876
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.35213410485461,
+      "learning_rate": 8.091126696102666e-06,
+      "loss": 0.5301,
+      "step": 2877
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4431390148774264,
+      "learning_rate": 8.089758023531788e-06,
+      "loss": 0.4974,
+      "step": 2878
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.587027889823297,
+      "learning_rate": 8.088388976317096e-06,
+      "loss": 0.5039,
+      "step": 2879
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.44298301753653235,
+      "learning_rate": 8.087019554624595e-06,
+      "loss": 0.3011,
+      "step": 2880
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.282598071171227,
+      "learning_rate": 8.085649758620331e-06,
+      "loss": 0.5403,
+      "step": 2881
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.8607787699132103,
+      "learning_rate": 8.084279588470395e-06,
+      "loss": 0.4871,
+      "step": 2882
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.948607445963684,
+      "learning_rate": 8.082909044340926e-06,
+      "loss": 0.4756,
+      "step": 2883
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.134538315564137,
+      "learning_rate": 8.081538126398105e-06,
+      "loss": 0.5029,
+      "step": 2884
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7680777515546886,
+      "learning_rate": 8.080166834808166e-06,
+      "loss": 0.5278,
+      "step": 2885
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.779931897464195,
+      "learning_rate": 8.078795169737377e-06,
+      "loss": 0.5258,
+      "step": 2886
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.372899366829096,
+      "learning_rate": 8.077423131352058e-06,
+      "loss": 0.4903,
+      "step": 2887
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.76726466514698,
+      "learning_rate": 8.076050719818577e-06,
+      "loss": 0.5087,
+      "step": 2888
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.664732254677657,
+      "learning_rate": 8.074677935303343e-06,
+      "loss": 0.5017,
+      "step": 2889
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.8878834582523085,
+      "learning_rate": 8.073304777972806e-06,
+      "loss": 0.497,
+      "step": 2890
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.585360784640874,
+      "learning_rate": 8.071931247993472e-06,
+      "loss": 0.4702,
+      "step": 2891
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.9453701740128775,
+      "learning_rate": 8.070557345531883e-06,
+      "loss": 0.5512,
+      "step": 2892
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1099791042938887,
+      "learning_rate": 8.06918307075463e-06,
+      "loss": 0.5424,
+      "step": 2893
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3852687594350885,
+      "learning_rate": 8.067808423828347e-06,
+      "loss": 0.5278,
+      "step": 2894
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.730411187079852,
+      "learning_rate": 8.066433404919718e-06,
+      "loss": 0.4885,
+      "step": 2895
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4170552203217373,
+      "learning_rate": 8.065058014195466e-06,
+      "loss": 0.5414,
+      "step": 2896
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0027918232726867,
+      "learning_rate": 8.063682251822363e-06,
+      "loss": 0.4963,
+      "step": 2897
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3808657813264014,
+      "learning_rate": 8.062306117967224e-06,
+      "loss": 0.485,
+      "step": 2898
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.971778004557632,
+      "learning_rate": 8.060929612796914e-06,
+      "loss": 0.4758,
+      "step": 2899
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.669840726305179,
+      "learning_rate": 8.059552736478333e-06,
+      "loss": 0.5314,
+      "step": 2900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.2623621554685265,
+      "learning_rate": 8.058175489178437e-06,
+      "loss": 0.5451,
+      "step": 2901
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.388877358888651,
+      "learning_rate": 8.056797871064216e-06,
+      "loss": 0.4995,
+      "step": 2902
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0534872300234017,
+      "learning_rate": 8.05541988230272e-06,
+      "loss": 0.5118,
+      "step": 2903
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.723004153649712,
+      "learning_rate": 8.054041523061027e-06,
+      "loss": 0.486,
+      "step": 2904
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.9026813382914294,
+      "learning_rate": 8.05266279350627e-06,
+      "loss": 0.569,
+      "step": 2905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3432025880995173,
+      "learning_rate": 8.051283693805624e-06,
+      "loss": 0.444,
+      "step": 2906
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.757093478772514,
+      "learning_rate": 8.049904224126312e-06,
+      "loss": 0.5015,
+      "step": 2907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2048594480708563,
+      "learning_rate": 8.048524384635598e-06,
+      "loss": 0.5104,
+      "step": 2908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.567472536350001,
+      "learning_rate": 8.047144175500793e-06,
+      "loss": 0.5186,
+      "step": 2909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.45000161918229403,
+      "learning_rate": 8.045763596889253e-06,
+      "loss": 0.3025,
+      "step": 2910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4261545863271896,
+      "learning_rate": 8.044382648968375e-06,
+      "loss": 0.31,
+      "step": 2911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8264342472108757,
+      "learning_rate": 8.043001331905605e-06,
+      "loss": 0.4731,
+      "step": 2912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.721834697658449,
+      "learning_rate": 8.041619645868434e-06,
+      "loss": 0.5122,
+      "step": 2913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.918812153250377,
+      "learning_rate": 8.040237591024394e-06,
+      "loss": 0.4635,
+      "step": 2914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.515993799850581,
+      "learning_rate": 8.038855167541064e-06,
+      "loss": 0.4967,
+      "step": 2915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.0991637732080575,
+      "learning_rate": 8.037472375586068e-06,
+      "loss": 0.5053,
+      "step": 2916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4492360542842055,
+      "learning_rate": 8.036089215327077e-06,
+      "loss": 0.557,
+      "step": 2917
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.1828362162117045,
+      "learning_rate": 8.0347056869318e-06,
+      "loss": 0.5144,
+      "step": 2918
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.06470951456924,
+      "learning_rate": 8.033321790567997e-06,
+      "loss": 0.4579,
+      "step": 2919
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.278225972234292,
+      "learning_rate": 8.03193752640347e-06,
+      "loss": 0.4933,
+      "step": 2920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5327511802242667,
+      "learning_rate": 8.030552894606064e-06,
+      "loss": 0.3153,
+      "step": 2921
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3521246787957923,
+      "learning_rate": 8.029167895343672e-06,
+      "loss": 0.5332,
+      "step": 2922
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.0073805610243936,
+      "learning_rate": 8.02778252878423e-06,
+      "loss": 0.5229,
+      "step": 2923
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.976271297537824,
+      "learning_rate": 8.026396795095717e-06,
+      "loss": 0.5335,
+      "step": 2924
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.762550252832468,
+      "learning_rate": 8.02501069444616e-06,
+      "loss": 0.5527,
+      "step": 2925
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.208963952425751,
+      "learning_rate": 8.023624227003627e-06,
+      "loss": 0.4764,
+      "step": 2926
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.841852828635381,
+      "learning_rate": 8.022237392936232e-06,
+      "loss": 0.5513,
+      "step": 2927
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8178294899407894,
+      "learning_rate": 8.020850192412134e-06,
+      "loss": 0.4965,
+      "step": 2928
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.24751953471896,
+      "learning_rate": 8.019462625599537e-06,
+      "loss": 0.529,
+      "step": 2929
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.214803674558355,
+      "learning_rate": 8.018074692666686e-06,
+      "loss": 0.4949,
+      "step": 2930
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9296506259617288,
+      "learning_rate": 8.016686393781874e-06,
+      "loss": 0.545,
+      "step": 2931
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.427966809780375,
+      "learning_rate": 8.015297729113436e-06,
+      "loss": 0.5533,
+      "step": 2932
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.756698020158709,
+      "learning_rate": 8.013908698829754e-06,
+      "loss": 0.513,
+      "step": 2933
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.651880477410838,
+      "learning_rate": 8.012519303099251e-06,
+      "loss": 0.5384,
+      "step": 2934
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3803406773106848,
+      "learning_rate": 8.011129542090396e-06,
+      "loss": 0.508,
+      "step": 2935
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.074193239524767,
+      "learning_rate": 8.009739415971704e-06,
+      "loss": 0.5082,
+      "step": 2936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.464924417651583,
+      "learning_rate": 8.008348924911734e-06,
+      "loss": 0.5398,
+      "step": 2937
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.45932371579500203,
+      "learning_rate": 8.006958069079082e-06,
+      "loss": 0.3069,
+      "step": 2938
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.41384470803658263,
+      "learning_rate": 8.005566848642398e-06,
+      "loss": 0.2988,
+      "step": 2939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2021879077144373,
+      "learning_rate": 8.004175263770373e-06,
+      "loss": 0.4803,
+      "step": 2940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3907522485853317,
+      "learning_rate": 8.00278331463174e-06,
+      "loss": 0.298,
+      "step": 2941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7197911857192367,
+      "learning_rate": 8.001391001395278e-06,
+      "loss": 0.5292,
+      "step": 2942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.744363868855232,
+      "learning_rate": 7.99999832422981e-06,
+      "loss": 0.514,
+      "step": 2943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.7515465881131185,
+      "learning_rate": 7.998605283304201e-06,
+      "loss": 0.5143,
+      "step": 2944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.796427034371325,
+      "learning_rate": 7.997211878787366e-06,
+      "loss": 0.5791,
+      "step": 2945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4594894895795636,
+      "learning_rate": 7.995818110848256e-06,
+      "loss": 0.4741,
+      "step": 2946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8195000278333207,
+      "learning_rate": 7.994423979655873e-06,
+      "loss": 0.4857,
+      "step": 2947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2553388710303146,
+      "learning_rate": 7.993029485379258e-06,
+      "loss": 0.507,
+      "step": 2948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.932123976164065,
+      "learning_rate": 7.991634628187499e-06,
+      "loss": 0.5284,
+      "step": 2949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.087032368088779,
+      "learning_rate": 7.99023940824973e-06,
+      "loss": 0.4939,
+      "step": 2950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.192557957703767,
+      "learning_rate": 7.98884382573512e-06,
+      "loss": 0.5207,
+      "step": 2951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.074546195352778,
+      "learning_rate": 7.987447880812895e-06,
+      "loss": 0.5087,
+      "step": 2952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5458508890045577,
+      "learning_rate": 7.986051573652315e-06,
+      "loss": 0.5238,
+      "step": 2953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5908766001407697,
+      "learning_rate": 7.984654904422686e-06,
+      "loss": 0.4835,
+      "step": 2954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.447123826597102,
+      "learning_rate": 7.983257873293362e-06,
+      "loss": 0.5016,
+      "step": 2955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.687810780377215,
+      "learning_rate": 7.981860480433735e-06,
+      "loss": 0.4353,
+      "step": 2956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.836561392136581,
+      "learning_rate": 7.980462726013246e-06,
+      "loss": 0.5019,
+      "step": 2957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 28.258921021219887,
+      "learning_rate": 7.979064610201372e-06,
+      "loss": 0.4861,
+      "step": 2958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.756097086192861,
+      "learning_rate": 7.977666133167647e-06,
+      "loss": 0.5924,
+      "step": 2959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.5451565734887085,
+      "learning_rate": 7.976267295081637e-06,
+      "loss": 0.5201,
+      "step": 2960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.771602131957479,
+      "learning_rate": 7.974868096112957e-06,
+      "loss": 0.5173,
+      "step": 2961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9846653329430817,
+      "learning_rate": 7.973468536431266e-06,
+      "loss": 0.5079,
+      "step": 2962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.444698269773155,
+      "learning_rate": 7.972068616206261e-06,
+      "loss": 0.4891,
+      "step": 2963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.376698319477656,
+      "learning_rate": 7.970668335607692e-06,
+      "loss": 0.4985,
+      "step": 2964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7081313250545698,
+      "learning_rate": 7.969267694805345e-06,
+      "loss": 0.3159,
+      "step": 2965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.8826857803729844,
+      "learning_rate": 7.967866693969053e-06,
+      "loss": 0.5182,
+      "step": 2966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.188612963701901,
+      "learning_rate": 7.966465333268693e-06,
+      "loss": 0.5301,
+      "step": 2967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.559122514187511,
+      "learning_rate": 7.965063612874184e-06,
+      "loss": 0.5308,
+      "step": 2968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.415828080533882,
+      "learning_rate": 7.963661532955492e-06,
+      "loss": 0.4882,
+      "step": 2969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.890269565853594,
+      "learning_rate": 7.962259093682618e-06,
+      "loss": 0.5133,
+      "step": 2970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.52672251702571,
+      "learning_rate": 7.960856295225619e-06,
+      "loss": 0.5153,
+      "step": 2971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.24132578606781,
+      "learning_rate": 7.959453137754586e-06,
+      "loss": 0.4733,
+      "step": 2972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.496236492438045,
+      "learning_rate": 7.958049621439659e-06,
+      "loss": 0.4876,
+      "step": 2973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.510515703414671,
+      "learning_rate": 7.956645746451014e-06,
+      "loss": 0.3183,
+      "step": 2974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0360040440284184,
+      "learning_rate": 7.95524151295888e-06,
+      "loss": 0.4979,
+      "step": 2975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.920935253585619,
+      "learning_rate": 7.953836921133526e-06,
+      "loss": 0.4655,
+      "step": 2976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.105636820938825,
+      "learning_rate": 7.952431971145261e-06,
+      "loss": 0.4913,
+      "step": 2977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.077370675711057,
+      "learning_rate": 7.951026663164441e-06,
+      "loss": 0.4901,
+      "step": 2978
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.0733314818126045,
+      "learning_rate": 7.949620997361465e-06,
+      "loss": 0.532,
+      "step": 2979
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.434045989269188,
+      "learning_rate": 7.948214973906775e-06,
+      "loss": 0.4602,
+      "step": 2980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.271491791575562,
+      "learning_rate": 7.946808592970851e-06,
+      "loss": 0.5478,
+      "step": 2981
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.733511173159183,
+      "learning_rate": 7.945401854724231e-06,
+      "loss": 0.5071,
+      "step": 2982
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.785020640044596,
+      "learning_rate": 7.943994759337479e-06,
+      "loss": 0.5772,
+      "step": 2983
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.381955927968526,
+      "learning_rate": 7.942587306981213e-06,
+      "loss": 0.5207,
+      "step": 2984
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.1594993828572555,
+      "learning_rate": 7.941179497826093e-06,
+      "loss": 0.5393,
+      "step": 2985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.286264058261939,
+      "learning_rate": 7.939771332042818e-06,
+      "loss": 0.5241,
+      "step": 2986
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1001884925409464,
+      "learning_rate": 7.938362809802134e-06,
+      "loss": 0.5554,
+      "step": 2987
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.396784854317284,
+      "learning_rate": 7.936953931274828e-06,
+      "loss": 0.5103,
+      "step": 2988
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.171092930481997,
+      "learning_rate": 7.935544696631733e-06,
+      "loss": 0.5135,
+      "step": 2989
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.9133708850573194,
+      "learning_rate": 7.934135106043726e-06,
+      "loss": 0.4927,
+      "step": 2990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.005023651278931,
+      "learning_rate": 7.93272515968172e-06,
+      "loss": 0.5699,
+      "step": 2991
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.474714804787338,
+      "learning_rate": 7.931314857716677e-06,
+      "loss": 0.4858,
+      "step": 2992
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.254247126179624,
+      "learning_rate": 7.929904200319602e-06,
+      "loss": 0.4435,
+      "step": 2993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.192523695714699,
+      "learning_rate": 7.928493187661543e-06,
+      "loss": 0.5464,
+      "step": 2994
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5043099660108044,
+      "learning_rate": 7.927081819913589e-06,
+      "loss": 0.559,
+      "step": 2995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.484121566373756,
+      "learning_rate": 7.92567009724687e-06,
+      "loss": 0.4871,
+      "step": 2996
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.848502217517758,
+      "learning_rate": 7.92425801983257e-06,
+      "loss": 0.4793,
+      "step": 2997
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.833960411805838,
+      "learning_rate": 7.9228455878419e-06,
+      "loss": 0.5226,
+      "step": 2998
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.326436131719012,
+      "learning_rate": 7.921432801446128e-06,
+      "loss": 0.5285,
+      "step": 2999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.605388832287848,
+      "learning_rate": 7.920019660816556e-06,
+      "loss": 0.4804,
+      "step": 3000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.555909983894319,
+      "learning_rate": 7.918606166124534e-06,
+      "loss": 0.5082,
+      "step": 3001
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.359446590289393,
+      "learning_rate": 7.917192317541452e-06,
+      "loss": 0.5186,
+      "step": 3002
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7776542680531304,
+      "learning_rate": 7.915778115238744e-06,
+      "loss": 0.526,
+      "step": 3003
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.914237407472874,
+      "learning_rate": 7.914363559387887e-06,
+      "loss": 0.492,
+      "step": 3004
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.144145663307124,
+      "learning_rate": 7.912948650160403e-06,
+      "loss": 0.5112,
+      "step": 3005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.064473944961379,
+      "learning_rate": 7.911533387727852e-06,
+      "loss": 0.4947,
+      "step": 3006
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.007451396189253,
+      "learning_rate": 7.91011777226184e-06,
+      "loss": 0.4908,
+      "step": 3007
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.236996412961245,
+      "learning_rate": 7.908701803934014e-06,
+      "loss": 0.4936,
+      "step": 3008
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.853733846823748,
+      "learning_rate": 7.907285482916067e-06,
+      "loss": 0.5409,
+      "step": 3009
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5453417366671625,
+      "learning_rate": 7.905868809379735e-06,
+      "loss": 0.5189,
+      "step": 3010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.743556607674698,
+      "learning_rate": 7.904451783496789e-06,
+      "loss": 0.5478,
+      "step": 3011
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.247094688301214,
+      "learning_rate": 7.903034405439052e-06,
+      "loss": 0.4988,
+      "step": 3012
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.017061038408459,
+      "learning_rate": 7.901616675378384e-06,
+      "loss": 0.5295,
+      "step": 3013
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.192992302140329,
+      "learning_rate": 7.900198593486691e-06,
+      "loss": 0.5581,
+      "step": 3014
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.443016877986307,
+      "learning_rate": 7.89878015993592e-06,
+      "loss": 0.4906,
+      "step": 3015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.960745902058861,
+      "learning_rate": 7.897361374898063e-06,
+      "loss": 0.5352,
+      "step": 3016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.883663013391382,
+      "learning_rate": 7.895942238545149e-06,
+      "loss": 0.5505,
+      "step": 3017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.536468044952548,
+      "learning_rate": 7.894522751049255e-06,
+      "loss": 0.5492,
+      "step": 3018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8405271857023697,
+      "learning_rate": 7.893102912582497e-06,
+      "loss": 0.5076,
+      "step": 3019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.479316498066618,
+      "learning_rate": 7.89168272331704e-06,
+      "loss": 0.5117,
+      "step": 3020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.227078994392305,
+      "learning_rate": 7.890262183425082e-06,
+      "loss": 0.5009,
+      "step": 3021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.989943421408555,
+      "learning_rate": 7.888841293078868e-06,
+      "loss": 0.5536,
+      "step": 3022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.676021257520741,
+      "learning_rate": 7.887420052450689e-06,
+      "loss": 0.5522,
+      "step": 3023
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.286820467612126,
+      "learning_rate": 7.885998461712875e-06,
+      "loss": 0.5256,
+      "step": 3024
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.382941288360941,
+      "learning_rate": 7.8845765210378e-06,
+      "loss": 0.5312,
+      "step": 3025
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.610794889811128,
+      "learning_rate": 7.883154230597874e-06,
+      "loss": 0.4744,
+      "step": 3026
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.356901212326293,
+      "learning_rate": 7.88173159056556e-06,
+      "loss": 0.5245,
+      "step": 3027
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.58612496632678,
+      "learning_rate": 7.880308601113353e-06,
+      "loss": 0.5309,
+      "step": 3028
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.239450655455563,
+      "learning_rate": 7.8788852624138e-06,
+      "loss": 0.5549,
+      "step": 3029
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.972240160826587,
+      "learning_rate": 7.877461574639485e-06,
+      "loss": 0.5379,
+      "step": 3030
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.8021843138443834,
+      "learning_rate": 7.876037537963032e-06,
+      "loss": 0.5113,
+      "step": 3031
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.42766477450070883,
+      "learning_rate": 7.874613152557113e-06,
+      "loss": 0.309,
+      "step": 3032
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.174374060689435,
+      "learning_rate": 7.873188418594438e-06,
+      "loss": 0.527,
+      "step": 3033
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2186606839954144,
+      "learning_rate": 7.871763336247764e-06,
+      "loss": 0.5036,
+      "step": 3034
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.448454779622366,
+      "learning_rate": 7.870337905689882e-06,
+      "loss": 0.5283,
+      "step": 3035
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.881402266468378,
+      "learning_rate": 7.868912127093638e-06,
+      "loss": 0.5728,
+      "step": 3036
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3072077315954127,
+      "learning_rate": 7.867486000631902e-06,
+      "loss": 0.5326,
+      "step": 3037
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.071427576249954,
+      "learning_rate": 7.866059526477608e-06,
+      "loss": 0.5598,
+      "step": 3038
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.889951686564925,
+      "learning_rate": 7.864632704803712e-06,
+      "loss": 0.4797,
+      "step": 3039
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.349587733392355,
+      "learning_rate": 7.863205535783227e-06,
+      "loss": 0.533,
+      "step": 3040
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.177994241793068,
+      "learning_rate": 7.861778019589198e-06,
+      "loss": 0.5486,
+      "step": 3041
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.083177358700826,
+      "learning_rate": 7.86035015639472e-06,
+      "loss": 0.5362,
+      "step": 3042
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.416058345096872,
+      "learning_rate": 7.858921946372923e-06,
+      "loss": 0.5048,
+      "step": 3043
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.32204812336995,
+      "learning_rate": 7.857493389696984e-06,
+      "loss": 0.5146,
+      "step": 3044
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8883926830259443,
+      "learning_rate": 7.85606448654012e-06,
+      "loss": 0.534,
+      "step": 3045
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2004522320116795,
+      "learning_rate": 7.854635237075593e-06,
+      "loss": 0.4647,
+      "step": 3046
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.642226898681909,
+      "learning_rate": 7.8532056414767e-06,
+      "loss": 0.5536,
+      "step": 3047
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.387671047064352,
+      "learning_rate": 7.851775699916785e-06,
+      "loss": 0.5359,
+      "step": 3048
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.80784425691458,
+      "learning_rate": 7.850345412569237e-06,
+      "loss": 0.5311,
+      "step": 3049
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.011886071820627,
+      "learning_rate": 7.84891477960748e-06,
+      "loss": 0.5052,
+      "step": 3050
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.077124474004653,
+      "learning_rate": 7.847483801204984e-06,
+      "loss": 0.5033,
+      "step": 3051
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7319947560959665,
+      "learning_rate": 7.846052477535263e-06,
+      "loss": 0.59,
+      "step": 3052
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0113660630919132,
+      "learning_rate": 7.844620808771865e-06,
+      "loss": 0.4659,
+      "step": 3053
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.039978070744434,
+      "learning_rate": 7.843188795088386e-06,
+      "loss": 0.5576,
+      "step": 3054
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.5223850408021855,
+      "learning_rate": 7.841756436658466e-06,
+      "loss": 0.4711,
+      "step": 3055
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.106959568723476,
+      "learning_rate": 7.84032373365578e-06,
+      "loss": 0.5494,
+      "step": 3056
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.828419459230732,
+      "learning_rate": 7.838890686254049e-06,
+      "loss": 0.5325,
+      "step": 3057
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4071328825404668,
+      "learning_rate": 7.837457294627036e-06,
+      "loss": 0.5021,
+      "step": 3058
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5830539580296747,
+      "learning_rate": 7.836023558948542e-06,
+      "loss": 0.5241,
+      "step": 3059
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.998170962142801,
+      "learning_rate": 7.834589479392416e-06,
+      "loss": 0.5336,
+      "step": 3060
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4650661478143297,
+      "learning_rate": 7.833155056132543e-06,
+      "loss": 0.3197,
+      "step": 3061
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4412222386756952,
+      "learning_rate": 7.831720289342853e-06,
+      "loss": 0.2993,
+      "step": 3062
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.30496615854281,
+      "learning_rate": 7.830285179197314e-06,
+      "loss": 0.514,
+      "step": 3063
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5435939268135206,
+      "learning_rate": 7.828849725869939e-06,
+      "loss": 0.4809,
+      "step": 3064
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 25.358919363258135,
+      "learning_rate": 7.827413929534784e-06,
+      "loss": 0.481,
+      "step": 3065
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.480108347337476,
+      "learning_rate": 7.825977790365941e-06,
+      "loss": 0.5782,
+      "step": 3066
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.533174281667678,
+      "learning_rate": 7.824541308537548e-06,
+      "loss": 0.5088,
+      "step": 3067
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5053903434073127,
+      "learning_rate": 7.823104484223785e-06,
+      "loss": 0.3153,
+      "step": 3068
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.414011241174012,
+      "learning_rate": 7.821667317598871e-06,
+      "loss": 0.5408,
+      "step": 3069
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.894580132582315,
+      "learning_rate": 7.820229808837065e-06,
+      "loss": 0.4859,
+      "step": 3070
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7412146358122786,
+      "learning_rate": 7.818791958112673e-06,
+      "loss": 0.4642,
+      "step": 3071
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.477639626600864,
+      "learning_rate": 7.817353765600037e-06,
+      "loss": 0.521,
+      "step": 3072
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.132119703927092,
+      "learning_rate": 7.815915231473547e-06,
+      "loss": 0.572,
+      "step": 3073
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.399409224062869,
+      "learning_rate": 7.814476355907626e-06,
+      "loss": 0.5215,
+      "step": 3074
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.165549119096656,
+      "learning_rate": 7.813037139076743e-06,
+      "loss": 0.4922,
+      "step": 3075
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.15364205308413,
+      "learning_rate": 7.811597581155407e-06,
+      "loss": 0.5238,
+      "step": 3076
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.733068850995514,
+      "learning_rate": 7.810157682318174e-06,
+      "loss": 0.5252,
+      "step": 3077
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4228507297971102,
+      "learning_rate": 7.808717442739634e-06,
+      "loss": 0.5183,
+      "step": 3078
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4250841742577339,
+      "learning_rate": 7.807276862594421e-06,
+      "loss": 0.2914,
+      "step": 3079
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.2224135691260924,
+      "learning_rate": 7.80583594205721e-06,
+      "loss": 0.5597,
+      "step": 3080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.403947649859394,
+      "learning_rate": 7.804394681302716e-06,
+      "loss": 0.5047,
+      "step": 3081
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.4446513666572365,
+      "learning_rate": 7.8029530805057e-06,
+      "loss": 0.5224,
+      "step": 3082
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.4283105945646355,
+      "learning_rate": 7.80151113984096e-06,
+      "loss": 0.5373,
+      "step": 3083
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3950489064168391,
+      "learning_rate": 7.800068859483336e-06,
+      "loss": 0.3065,
+      "step": 3084
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.41241911541350734,
+      "learning_rate": 7.798626239607709e-06,
+      "loss": 0.3084,
+      "step": 3085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.353873563253003,
+      "learning_rate": 7.797183280389002e-06,
+      "loss": 0.5423,
+      "step": 3086
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3828280479934258,
+      "learning_rate": 7.79573998200218e-06,
+      "loss": 0.3093,
+      "step": 3087
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 15.398227442430615,
+      "learning_rate": 7.794296344622246e-06,
+      "loss": 0.4955,
+      "step": 3088
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.42668881934957914,
+      "learning_rate": 7.792852368424246e-06,
+      "loss": 0.3229,
+      "step": 3089
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7556096205265024,
+      "learning_rate": 7.791408053583269e-06,
+      "loss": 0.4674,
+      "step": 3090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.835462026845272,
+      "learning_rate": 7.789963400274442e-06,
+      "loss": 0.5181,
+      "step": 3091
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.467236503949054,
+      "learning_rate": 7.788518408672935e-06,
+      "loss": 0.5263,
+      "step": 3092
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.708479793361411,
+      "learning_rate": 7.787073078953955e-06,
+      "loss": 0.5008,
+      "step": 3093
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.8924983508511124,
+      "learning_rate": 7.785627411292758e-06,
+      "loss": 0.4742,
+      "step": 3094
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.005174610416288,
+      "learning_rate": 7.784181405864634e-06,
+      "loss": 0.5043,
+      "step": 3095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.285535356186545,
+      "learning_rate": 7.782735062844915e-06,
+      "loss": 0.5084,
+      "step": 3096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.260816578486147,
+      "learning_rate": 7.781288382408975e-06,
+      "loss": 0.4603,
+      "step": 3097
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.46300657144647067,
+      "learning_rate": 7.779841364732233e-06,
+      "loss": 0.3162,
+      "step": 3098
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.946387368163577,
+      "learning_rate": 7.778394009990139e-06,
+      "loss": 0.4903,
+      "step": 3099
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.627337511449446,
+      "learning_rate": 7.776946318358195e-06,
+      "loss": 0.491,
+      "step": 3100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.12215089257672,
+      "learning_rate": 7.775498290011935e-06,
+      "loss": 0.5123,
+      "step": 3101
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.957970386267614,
+      "learning_rate": 7.774049925126938e-06,
+      "loss": 0.4842,
+      "step": 3102
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.01805693572948,
+      "learning_rate": 7.772601223878824e-06,
+      "loss": 0.5068,
+      "step": 3103
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.928465847205783,
+      "learning_rate": 7.771152186443254e-06,
+      "loss": 0.523,
+      "step": 3104
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7778213963149914,
+      "learning_rate": 7.76970281299593e-06,
+      "loss": 0.4921,
+      "step": 3105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.4446106911140415,
+      "learning_rate": 7.768253103712589e-06,
+      "loss": 0.5009,
+      "step": 3106
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.816345463005742,
+      "learning_rate": 7.766803058769014e-06,
+      "loss": 0.5103,
+      "step": 3107
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.393252919785072,
+      "learning_rate": 7.765352678341032e-06,
+      "loss": 0.4662,
+      "step": 3108
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.208145240326901,
+      "learning_rate": 7.763901962604502e-06,
+      "loss": 0.5216,
+      "step": 3109
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.455790229081902,
+      "learning_rate": 7.762450911735333e-06,
+      "loss": 0.5222,
+      "step": 3110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4349996524663667,
+      "learning_rate": 7.760999525909465e-06,
+      "loss": 0.3096,
+      "step": 3111
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.191394754762834,
+      "learning_rate": 7.759547805302888e-06,
+      "loss": 0.4861,
+      "step": 3112
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.927605719701791,
+      "learning_rate": 7.758095750091625e-06,
+      "loss": 0.4983,
+      "step": 3113
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5780479264663634,
+      "learning_rate": 7.756643360451744e-06,
+      "loss": 0.5569,
+      "step": 3114
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.303114153410021,
+      "learning_rate": 7.75519063655935e-06,
+      "loss": 0.508,
+      "step": 3115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.964570381230671,
+      "learning_rate": 7.753737578590594e-06,
+      "loss": 0.5364,
+      "step": 3116
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5400600933811686,
+      "learning_rate": 7.752284186721664e-06,
+      "loss": 0.5266,
+      "step": 3117
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.4548910239829445,
+      "learning_rate": 7.750830461128786e-06,
+      "loss": 0.5031,
+      "step": 3118
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.61275222992356,
+      "learning_rate": 7.749376401988232e-06,
+      "loss": 0.5311,
+      "step": 3119
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8949264227743523,
+      "learning_rate": 7.747922009476313e-06,
+      "loss": 0.5543,
+      "step": 3120
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.926309950884724,
+      "learning_rate": 7.746467283769374e-06,
+      "loss": 0.4844,
+      "step": 3121
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1242527898806443,
+      "learning_rate": 7.745012225043809e-06,
+      "loss": 0.5467,
+      "step": 3122
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.48818915274825,
+      "learning_rate": 7.743556833476047e-06,
+      "loss": 0.5298,
+      "step": 3123
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3301450230243375,
+      "learning_rate": 7.742101109242562e-06,
+      "loss": 0.5282,
+      "step": 3124
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 20.496304108123248,
+      "learning_rate": 7.740645052519863e-06,
+      "loss": 0.5423,
+      "step": 3125
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.710715128889198,
+      "learning_rate": 7.739188663484505e-06,
+      "loss": 0.5489,
+      "step": 3126
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.3703766242471564,
+      "learning_rate": 7.737731942313077e-06,
+      "loss": 0.507,
+      "step": 3127
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.673830157375672,
+      "learning_rate": 7.736274889182211e-06,
+      "loss": 0.5546,
+      "step": 3128
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43472896408587924,
+      "learning_rate": 7.734817504268582e-06,
+      "loss": 0.318,
+      "step": 3129
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.35707636965223494,
+      "learning_rate": 7.733359787748904e-06,
+      "loss": 0.2848,
+      "step": 3130
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.019016630667951,
+      "learning_rate": 7.731901739799927e-06,
+      "loss": 0.4689,
+      "step": 3131
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.625078775021546,
+      "learning_rate": 7.730443360598447e-06,
+      "loss": 0.5524,
+      "step": 3132
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3963492028273267,
+      "learning_rate": 7.728984650321293e-06,
+      "loss": 0.3047,
+      "step": 3133
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.714301905410556,
+      "learning_rate": 7.727525609145345e-06,
+      "loss": 0.4895,
+      "step": 3134
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.328297214783978,
+      "learning_rate": 7.726066237247514e-06,
+      "loss": 0.5139,
+      "step": 3135
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.582069907446968,
+      "learning_rate": 7.72460653480475e-06,
+      "loss": 0.5344,
+      "step": 3136
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.068952625902977,
+      "learning_rate": 7.723146501994054e-06,
+      "loss": 0.5292,
+      "step": 3137
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.470247817499898,
+      "learning_rate": 7.721686138992456e-06,
+      "loss": 0.5014,
+      "step": 3138
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7844063746909353,
+      "learning_rate": 7.72022544597703e-06,
+      "loss": 0.5562,
+      "step": 3139
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.035445085811969,
+      "learning_rate": 7.718764423124892e-06,
+      "loss": 0.5319,
+      "step": 3140
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.981069213881948,
+      "learning_rate": 7.717303070613193e-06,
+      "loss": 0.5376,
+      "step": 3141
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.452071048311323,
+      "learning_rate": 7.71584138861913e-06,
+      "loss": 0.5144,
+      "step": 3142
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6390810055751674,
+      "learning_rate": 7.714379377319933e-06,
+      "loss": 0.4942,
+      "step": 3143
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5444706721986523,
+      "learning_rate": 7.712917036892881e-06,
+      "loss": 0.4596,
+      "step": 3144
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.425783687438419,
+      "learning_rate": 7.711454367515285e-06,
+      "loss": 0.2996,
+      "step": 3145
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.76691894884903,
+      "learning_rate": 7.709991369364498e-06,
+      "loss": 0.5346,
+      "step": 3146
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.202116739704948,
+      "learning_rate": 7.708528042617915e-06,
+      "loss": 0.5236,
+      "step": 3147
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.624767518390892,
+      "learning_rate": 7.707064387452971e-06,
+      "loss": 0.5121,
+      "step": 3148
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9655384477210096,
+      "learning_rate": 7.705600404047136e-06,
+      "loss": 0.4799,
+      "step": 3149
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.534364239459561,
+      "learning_rate": 7.704136092577924e-06,
+      "loss": 0.5324,
+      "step": 3150
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.3079996502160505,
+      "learning_rate": 7.70267145322289e-06,
+      "loss": 0.5075,
+      "step": 3151
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.364128464893169,
+      "learning_rate": 7.701206486159622e-06,
+      "loss": 0.5292,
+      "step": 3152
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8311825152111703,
+      "learning_rate": 7.699741191565758e-06,
+      "loss": 0.5244,
+      "step": 3153
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.026842562041294,
+      "learning_rate": 7.698275569618965e-06,
+      "loss": 0.5026,
+      "step": 3154
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.4633461397178005,
+      "learning_rate": 7.696809620496957e-06,
+      "loss": 0.5261,
+      "step": 3155
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.396258057401936,
+      "learning_rate": 7.695343344377486e-06,
+      "loss": 0.5493,
+      "step": 3156
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.815443591973296,
+      "learning_rate": 7.693876741438342e-06,
+      "loss": 0.4847,
+      "step": 3157
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.0958414629284245,
+      "learning_rate": 7.692409811857357e-06,
+      "loss": 0.5371,
+      "step": 3158
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43866083069167405,
+      "learning_rate": 7.690942555812398e-06,
+      "loss": 0.3096,
+      "step": 3159
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.694710564901458,
+      "learning_rate": 7.689474973481379e-06,
+      "loss": 0.5358,
+      "step": 3160
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.395149197644388,
+      "learning_rate": 7.688007065042246e-06,
+      "loss": 0.5384,
+      "step": 3161
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.7903908524454435,
+      "learning_rate": 7.686538830672988e-06,
+      "loss": 0.5029,
+      "step": 3162
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3254290121518624,
+      "learning_rate": 7.685070270551637e-06,
+      "loss": 0.5269,
+      "step": 3163
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0325104617051624,
+      "learning_rate": 7.68360138485626e-06,
+      "loss": 0.4705,
+      "step": 3164
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.896650299755458,
+      "learning_rate": 7.682132173764962e-06,
+      "loss": 0.4764,
+      "step": 3165
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.147097619213676,
+      "learning_rate": 7.68066263745589e-06,
+      "loss": 0.5266,
+      "step": 3166
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.582102647744946,
+      "learning_rate": 7.679192776107232e-06,
+      "loss": 0.5107,
+      "step": 3167
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.327150966203888,
+      "learning_rate": 7.677722589897214e-06,
+      "loss": 0.4602,
+      "step": 3168
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4374904257926815,
+      "learning_rate": 7.676252079004101e-06,
+      "loss": 0.5183,
+      "step": 3169
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.5457206651623374,
+      "learning_rate": 7.674781243606197e-06,
+      "loss": 0.5333,
+      "step": 3170
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.513786190328183,
+      "learning_rate": 7.673310083881844e-06,
+      "loss": 0.5199,
+      "step": 3171
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2893380613156733,
+      "learning_rate": 7.67183860000943e-06,
+      "loss": 0.5165,
+      "step": 3172
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8273988477260144,
+      "learning_rate": 7.670366792167371e-06,
+      "loss": 0.4987,
+      "step": 3173
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.178082114577306,
+      "learning_rate": 7.668894660534135e-06,
+      "loss": 0.5446,
+      "step": 3174
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3368862290529924,
+      "learning_rate": 7.66742220528822e-06,
+      "loss": 0.4519,
+      "step": 3175
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.101478422216226,
+      "learning_rate": 7.665949426608164e-06,
+      "loss": 0.5035,
+      "step": 3176
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.185912768227416,
+      "learning_rate": 7.664476324672552e-06,
+      "loss": 0.5164,
+      "step": 3177
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.186746934877164,
+      "learning_rate": 7.663002899659999e-06,
+      "loss": 0.5472,
+      "step": 3178
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.490392886645507,
+      "learning_rate": 7.661529151749163e-06,
+      "loss": 0.4711,
+      "step": 3179
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.032650389347831,
+      "learning_rate": 7.660055081118743e-06,
+      "loss": 0.517,
+      "step": 3180
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.871055233860458,
+      "learning_rate": 7.658580687947473e-06,
+      "loss": 0.5013,
+      "step": 3181
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4377296499380803,
+      "learning_rate": 7.657105972414129e-06,
+      "loss": 0.4833,
+      "step": 3182
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.466620853083225,
+      "learning_rate": 7.655630934697524e-06,
+      "loss": 0.5425,
+      "step": 3183
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.886849948169463,
+      "learning_rate": 7.654155574976516e-06,
+      "loss": 0.532,
+      "step": 3184
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.585419656552582,
+      "learning_rate": 7.652679893429993e-06,
+      "loss": 0.5005,
+      "step": 3185
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.857218982069756,
+      "learning_rate": 7.651203890236886e-06,
+      "loss": 0.5675,
+      "step": 3186
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.973300436751853,
+      "learning_rate": 7.64972756557617e-06,
+      "loss": 0.4897,
+      "step": 3187
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.440265422177732,
+      "learning_rate": 7.64825091962685e-06,
+      "loss": 0.4934,
+      "step": 3188
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3372535526475082,
+      "learning_rate": 7.646773952567976e-06,
+      "loss": 0.5116,
+      "step": 3189
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.070090221309353,
+      "learning_rate": 7.645296664578635e-06,
+      "loss": 0.5001,
+      "step": 3190
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.570341365879789,
+      "learning_rate": 7.643819055837955e-06,
+      "loss": 0.4931,
+      "step": 3191
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.689078327288583,
+      "learning_rate": 7.6423411265251e-06,
+      "loss": 0.4886,
+      "step": 3192
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.690074292058851,
+      "learning_rate": 7.640862876819272e-06,
+      "loss": 0.4708,
+      "step": 3193
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.259596497615795,
+      "learning_rate": 7.639384306899716e-06,
+      "loss": 0.4983,
+      "step": 3194
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4239290033649917,
+      "learning_rate": 7.637905416945716e-06,
+      "loss": 0.2862,
+      "step": 3195
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.952585784072994,
+      "learning_rate": 7.636426207136588e-06,
+      "loss": 0.546,
+      "step": 3196
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.042826620038685,
+      "learning_rate": 7.634946677651694e-06,
+      "loss": 0.4769,
+      "step": 3197
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.60655432383599,
+      "learning_rate": 7.633466828670431e-06,
+      "loss": 0.5193,
+      "step": 3198
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.609502122111287,
+      "learning_rate": 7.631986660372235e-06,
+      "loss": 0.5159,
+      "step": 3199
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.36495438054691215,
+      "learning_rate": 7.630506172936583e-06,
+      "loss": 0.3069,
+      "step": 3200
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.662035661508304,
+      "learning_rate": 7.6290253665429905e-06,
+      "loss": 0.5409,
+      "step": 3201
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4123991915271572,
+      "learning_rate": 7.627544241371006e-06,
+      "loss": 0.2952,
+      "step": 3202
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.012169627950289,
+      "learning_rate": 7.6260627976002255e-06,
+      "loss": 0.4701,
+      "step": 3203
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.88860611907311,
+      "learning_rate": 7.624581035410277e-06,
+      "loss": 0.4843,
+      "step": 3204
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.9820640371375795,
+      "learning_rate": 7.6230989549808296e-06,
+      "loss": 0.5348,
+      "step": 3205
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.9427620826493,
+      "learning_rate": 7.621616556491591e-06,
+      "loss": 0.5228,
+      "step": 3206
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4195872844518798,
+      "learning_rate": 7.620133840122306e-06,
+      "loss": 0.5132,
+      "step": 3207
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9500075296988117,
+      "learning_rate": 7.618650806052761e-06,
+      "loss": 0.5278,
+      "step": 3208
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.858977769065344,
+      "learning_rate": 7.617167454462777e-06,
+      "loss": 0.54,
+      "step": 3209
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.368377570890082,
+      "learning_rate": 7.615683785532218e-06,
+      "loss": 0.5487,
+      "step": 3210
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.6347091792478565,
+      "learning_rate": 7.614199799440981e-06,
+      "loss": 0.5391,
+      "step": 3211
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.41418368299739805,
+      "learning_rate": 7.612715496369007e-06,
+      "loss": 0.3029,
+      "step": 3212
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.888439465349945,
+      "learning_rate": 7.61123087649627e-06,
+      "loss": 0.5565,
+      "step": 3213
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.732670221722984,
+      "learning_rate": 7.609745940002786e-06,
+      "loss": 0.4189,
+      "step": 3214
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.6530440251080725,
+      "learning_rate": 7.608260687068611e-06,
+      "loss": 0.4924,
+      "step": 3215
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.796260625318673,
+      "learning_rate": 7.606775117873837e-06,
+      "loss": 0.5171,
+      "step": 3216
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.5761360186647595,
+      "learning_rate": 7.605289232598592e-06,
+      "loss": 0.5167,
+      "step": 3217
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.127123116451232,
+      "learning_rate": 7.603803031423046e-06,
+      "loss": 0.5508,
+      "step": 3218
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.40912252499993,
+      "learning_rate": 7.602316514527405e-06,
+      "loss": 0.5281,
+      "step": 3219
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.691588789213138,
+      "learning_rate": 7.600829682091915e-06,
+      "loss": 0.5247,
+      "step": 3220
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1455071673150634,
+      "learning_rate": 7.599342534296861e-06,
+      "loss": 0.5087,
+      "step": 3221
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.289361558986917,
+      "learning_rate": 7.597855071322562e-06,
+      "loss": 0.5288,
+      "step": 3222
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.170168154016953,
+      "learning_rate": 7.59636729334938e-06,
+      "loss": 0.5249,
+      "step": 3223
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.782489489460207,
+      "learning_rate": 7.594879200557712e-06,
+      "loss": 0.4976,
+      "step": 3224
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.640198347265379,
+      "learning_rate": 7.593390793127997e-06,
+      "loss": 0.4955,
+      "step": 3225
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.437445796304998,
+      "learning_rate": 7.591902071240705e-06,
+      "loss": 0.4886,
+      "step": 3226
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.5318453000413115,
+      "learning_rate": 7.590413035076354e-06,
+      "loss": 0.5078,
+      "step": 3227
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.003715768823673,
+      "learning_rate": 7.5889236848154905e-06,
+      "loss": 0.5003,
+      "step": 3228
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.5025238902170495,
+      "learning_rate": 7.587434020638704e-06,
+      "loss": 0.5024,
+      "step": 3229
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.708762317600356,
+      "learning_rate": 7.585944042726623e-06,
+      "loss": 0.4558,
+      "step": 3230
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.485493175660174,
+      "learning_rate": 7.584453751259912e-06,
+      "loss": 0.5293,
+      "step": 3231
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.155381094951946,
+      "learning_rate": 7.582963146419273e-06,
+      "loss": 0.522,
+      "step": 3232
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7289446920164684,
+      "learning_rate": 7.581472228385448e-06,
+      "loss": 0.5179,
+      "step": 3233
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9792496390834033,
+      "learning_rate": 7.579980997339215e-06,
+      "loss": 0.5766,
+      "step": 3234
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.42713291146349947,
+      "learning_rate": 7.5784894534613925e-06,
+      "loss": 0.3321,
+      "step": 3235
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 19.104858127938986,
+      "learning_rate": 7.576997596932834e-06,
+      "loss": 0.4698,
+      "step": 3236
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.9760240383112997,
+      "learning_rate": 7.575505427934433e-06,
+      "loss": 0.5211,
+      "step": 3237
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5403168488452903,
+      "learning_rate": 7.5740129466471204e-06,
+      "loss": 0.5305,
+      "step": 3238
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.691329635582272,
+      "learning_rate": 7.572520153251862e-06,
+      "loss": 0.5068,
+      "step": 3239
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.107679244848098,
+      "learning_rate": 7.571027047929667e-06,
+      "loss": 0.5533,
+      "step": 3240
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.206914906848164,
+      "learning_rate": 7.5695336308615785e-06,
+      "loss": 0.5385,
+      "step": 3241
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.21252811470978,
+      "learning_rate": 7.56803990222868e-06,
+      "loss": 0.4742,
+      "step": 3242
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2015828709544514,
+      "learning_rate": 7.56654586221209e-06,
+      "loss": 0.5072,
+      "step": 3243
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7706439187046055,
+      "learning_rate": 7.565051510992964e-06,
+      "loss": 0.5062,
+      "step": 3244
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.135578375177123,
+      "learning_rate": 7.563556848752502e-06,
+      "loss": 0.5062,
+      "step": 3245
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3563316684010487,
+      "learning_rate": 7.5620618756719325e-06,
+      "loss": 0.5233,
+      "step": 3246
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.40775530190457,
+      "learning_rate": 7.560566591932526e-06,
+      "loss": 0.455,
+      "step": 3247
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.827870438644589,
+      "learning_rate": 7.559070997715596e-06,
+      "loss": 0.5144,
+      "step": 3248
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4361029990052887,
+      "learning_rate": 7.557575093202484e-06,
+      "loss": 0.4802,
+      "step": 3249
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.632062424582603,
+      "learning_rate": 7.556078878574573e-06,
+      "loss": 0.5233,
+      "step": 3250
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.601522355455889,
+      "learning_rate": 7.554582354013288e-06,
+      "loss": 0.518,
+      "step": 3251
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.425334721365402,
+      "learning_rate": 7.553085519700085e-06,
+      "loss": 0.4984,
+      "step": 3252
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.636573084311702,
+      "learning_rate": 7.551588375816461e-06,
+      "loss": 0.5362,
+      "step": 3253
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.658365359508257,
+      "learning_rate": 7.550090922543949e-06,
+      "loss": 0.5069,
+      "step": 3254
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.906619045434484,
+      "learning_rate": 7.54859316006412e-06,
+      "loss": 0.5185,
+      "step": 3255
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 38.970875646629864,
+      "learning_rate": 7.547095088558586e-06,
+      "loss": 0.4325,
+      "step": 3256
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.80184811855406,
+      "learning_rate": 7.545596708208991e-06,
+      "loss": 0.5306,
+      "step": 3257
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.377441976028112,
+      "learning_rate": 7.544098019197018e-06,
+      "loss": 0.4811,
+      "step": 3258
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.634087247200744,
+      "learning_rate": 7.542599021704391e-06,
+      "loss": 0.4851,
+      "step": 3259
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5664334039361756,
+      "learning_rate": 7.541099715912867e-06,
+      "loss": 0.453,
+      "step": 3260
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.250606328928676,
+      "learning_rate": 7.539600102004241e-06,
+      "loss": 0.5524,
+      "step": 3261
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.9779461798140896,
+      "learning_rate": 7.538100180160347e-06,
+      "loss": 0.5605,
+      "step": 3262
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.888104923651797,
+      "learning_rate": 7.5365999505630575e-06,
+      "loss": 0.5094,
+      "step": 3263
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.641962801739554,
+      "learning_rate": 7.5350994133942764e-06,
+      "loss": 0.4855,
+      "step": 3264
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.1134709376888985,
+      "learning_rate": 7.533598568835954e-06,
+      "loss": 0.4917,
+      "step": 3265
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4761405685658353,
+      "learning_rate": 7.53209741707007e-06,
+      "loss": 0.5101,
+      "step": 3266
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8705862087551655,
+      "learning_rate": 7.530595958278644e-06,
+      "loss": 0.5108,
+      "step": 3267
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.725488634270594,
+      "learning_rate": 7.529094192643733e-06,
+      "loss": 0.4964,
+      "step": 3268
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7288685591135717,
+      "learning_rate": 7.527592120347433e-06,
+      "loss": 0.5126,
+      "step": 3269
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.123661116029459,
+      "learning_rate": 7.526089741571876e-06,
+      "loss": 0.5333,
+      "step": 3270
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2027803704069884,
+      "learning_rate": 7.524587056499227e-06,
+      "loss": 0.5181,
+      "step": 3271
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.054055347328473,
+      "learning_rate": 7.523084065311695e-06,
+      "loss": 0.5106,
+      "step": 3272
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6927100309394225,
+      "learning_rate": 7.521580768191521e-06,
+      "loss": 0.4969,
+      "step": 3273
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7635295015409076,
+      "learning_rate": 7.520077165320986e-06,
+      "loss": 0.4871,
+      "step": 3274
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.0220492540714226,
+      "learning_rate": 7.518573256882407e-06,
+      "loss": 0.5122,
+      "step": 3275
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.044644033947732,
+      "learning_rate": 7.517069043058136e-06,
+      "loss": 0.5066,
+      "step": 3276
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9870161435866347,
+      "learning_rate": 7.515564524030568e-06,
+      "loss": 0.5194,
+      "step": 3277
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.089074949292507,
+      "learning_rate": 7.514059699982128e-06,
+      "loss": 0.4984,
+      "step": 3278
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.204671753590511,
+      "learning_rate": 7.512554571095282e-06,
+      "loss": 0.5197,
+      "step": 3279
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4709528903797119,
+      "learning_rate": 7.511049137552533e-06,
+      "loss": 0.3177,
+      "step": 3280
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6594139494207294,
+      "learning_rate": 7.509543399536419e-06,
+      "loss": 0.4787,
+      "step": 3281
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.469346022128262,
+      "learning_rate": 7.508037357229515e-06,
+      "loss": 0.5378,
+      "step": 3282
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.753318279065782,
+      "learning_rate": 7.506531010814435e-06,
+      "loss": 0.5081,
+      "step": 3283
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.046557655477485,
+      "learning_rate": 7.505024360473829e-06,
+      "loss": 0.4733,
+      "step": 3284
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4432389453263062,
+      "learning_rate": 7.503517406390385e-06,
+      "loss": 0.321,
+      "step": 3285
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.39637034677068406,
+      "learning_rate": 7.502010148746821e-06,
+      "loss": 0.2913,
+      "step": 3286
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2983026297852063,
+      "learning_rate": 7.5005025877259045e-06,
+      "loss": 0.5272,
+      "step": 3287
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.611021565306897,
+      "learning_rate": 7.4989947235104264e-06,
+      "loss": 0.5314,
+      "step": 3288
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.325720498057133,
+      "learning_rate": 7.497486556283222e-06,
+      "loss": 0.4774,
+      "step": 3289
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.158912244431385,
+      "learning_rate": 7.495978086227164e-06,
+      "loss": 0.5521,
+      "step": 3290
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.281304242281109,
+      "learning_rate": 7.494469313525159e-06,
+      "loss": 0.4804,
+      "step": 3291
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.535634680604795,
+      "learning_rate": 7.4929602383601475e-06,
+      "loss": 0.498,
+      "step": 3292
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.322089547570366,
+      "learning_rate": 7.491450860915115e-06,
+      "loss": 0.5238,
+      "step": 3293
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.502804139019075,
+      "learning_rate": 7.489941181373075e-06,
+      "loss": 0.5736,
+      "step": 3294
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.006947974139421,
+      "learning_rate": 7.488431199917081e-06,
+      "loss": 0.5408,
+      "step": 3295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.294123382038009,
+      "learning_rate": 7.486920916730228e-06,
+      "loss": 0.5021,
+      "step": 3296
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.07878253221275,
+      "learning_rate": 7.48541033199564e-06,
+      "loss": 0.5752,
+      "step": 3297
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.661485727244736,
+      "learning_rate": 7.4838994458964785e-06,
+      "loss": 0.5052,
+      "step": 3298
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.288135287712206,
+      "learning_rate": 7.482388258615947e-06,
+      "loss": 0.5018,
+      "step": 3299
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4966042111202826,
+      "learning_rate": 7.480876770337281e-06,
+      "loss": 0.5206,
+      "step": 3300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.157447176643293,
+      "learning_rate": 7.479364981243754e-06,
+      "loss": 0.4596,
+      "step": 3301
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4023645404265377,
+      "learning_rate": 7.477852891518675e-06,
+      "loss": 0.4793,
+      "step": 3302
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.690486022223418,
+      "learning_rate": 7.4763405013453906e-06,
+      "loss": 0.5286,
+      "step": 3303
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.46221659580731955,
+      "learning_rate": 7.474827810907284e-06,
+      "loss": 0.3074,
+      "step": 3304
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.518986049373696,
+      "learning_rate": 7.473314820387772e-06,
+      "loss": 0.438,
+      "step": 3305
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8532125901063163,
+      "learning_rate": 7.471801529970312e-06,
+      "loss": 0.5216,
+      "step": 3306
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.491696940474477,
+      "learning_rate": 7.470287939838394e-06,
+      "loss": 0.5299,
+      "step": 3307
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.42827143517297,
+      "learning_rate": 7.468774050175548e-06,
+      "loss": 0.4809,
+      "step": 3308
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.7904530371677305,
+      "learning_rate": 7.467259861165335e-06,
+      "loss": 0.5307,
+      "step": 3309
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.007992935485785,
+      "learning_rate": 7.46574537299136e-06,
+      "loss": 0.4769,
+      "step": 3310
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9917973722107787,
+      "learning_rate": 7.464230585837257e-06,
+      "loss": 0.5043,
+      "step": 3311
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.359441286970144,
+      "learning_rate": 7.462715499886701e-06,
+      "loss": 0.5051,
+      "step": 3312
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.174683642910796,
+      "learning_rate": 7.4612001153234e-06,
+      "loss": 0.5072,
+      "step": 3313
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.475372550729842,
+      "learning_rate": 7.4596844323311e-06,
+      "loss": 0.4813,
+      "step": 3314
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.00832693522177,
+      "learning_rate": 7.458168451093583e-06,
+      "loss": 0.4632,
+      "step": 3315
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.596787401885228,
+      "learning_rate": 7.456652171794666e-06,
+      "loss": 0.496,
+      "step": 3316
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6442915474796114,
+      "learning_rate": 7.455135594618204e-06,
+      "loss": 0.5201,
+      "step": 3317
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.172981869261082,
+      "learning_rate": 7.453618719748087e-06,
+      "loss": 0.4794,
+      "step": 3318
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.17735355219369,
+      "learning_rate": 7.452101547368242e-06,
+      "loss": 0.5778,
+      "step": 3319
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.144312773633098,
+      "learning_rate": 7.450584077662628e-06,
+      "loss": 0.508,
+      "step": 3320
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 40.11043168698936,
+      "learning_rate": 7.449066310815249e-06,
+      "loss": 0.5375,
+      "step": 3321
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.640003584382682,
+      "learning_rate": 7.447548247010137e-06,
+      "loss": 0.517,
+      "step": 3322
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.898528433228625,
+      "learning_rate": 7.4460298864313606e-06,
+      "loss": 0.5384,
+      "step": 3323
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.062003995813404,
+      "learning_rate": 7.444511229263027e-06,
+      "loss": 0.5041,
+      "step": 3324
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9470145406787673,
+      "learning_rate": 7.442992275689281e-06,
+      "loss": 0.5358,
+      "step": 3325
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4937075890237788,
+      "learning_rate": 7.441473025894298e-06,
+      "loss": 0.5339,
+      "step": 3326
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.9108434488818937,
+      "learning_rate": 7.439953480062294e-06,
+      "loss": 0.5331,
+      "step": 3327
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6874930658556995,
+      "learning_rate": 7.43843363837752e-06,
+      "loss": 0.5626,
+      "step": 3328
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4137556139738387,
+      "learning_rate": 7.436913501024259e-06,
+      "loss": 0.3082,
+      "step": 3329
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.9492786368338955,
+      "learning_rate": 7.435393068186836e-06,
+      "loss": 0.4838,
+      "step": 3330
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.55251866305592,
+      "learning_rate": 7.4338723400496075e-06,
+      "loss": 0.5226,
+      "step": 3331
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.000038173560773,
+      "learning_rate": 7.432351316796964e-06,
+      "loss": 0.522,
+      "step": 3332
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.297365714407305,
+      "learning_rate": 7.430829998613342e-06,
+      "loss": 0.5105,
+      "step": 3333
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.091332772061091,
+      "learning_rate": 7.429308385683199e-06,
+      "loss": 0.4418,
+      "step": 3334
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4128668127959852,
+      "learning_rate": 7.427786478191041e-06,
+      "loss": 0.2981,
+      "step": 3335
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.944826758646175,
+      "learning_rate": 7.426264276321402e-06,
+      "loss": 0.5266,
+      "step": 3336
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.9308794469153985,
+      "learning_rate": 7.424741780258855e-06,
+      "loss": 0.5018,
+      "step": 3337
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.195626777556529,
+      "learning_rate": 7.423218990188008e-06,
+      "loss": 0.5357,
+      "step": 3338
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.369186334566084,
+      "learning_rate": 7.421695906293504e-06,
+      "loss": 0.4783,
+      "step": 3339
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.41105436816207785,
+      "learning_rate": 7.420172528760022e-06,
+      "loss": 0.2981,
+      "step": 3340
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.02538472339973,
+      "learning_rate": 7.418648857772279e-06,
+      "loss": 0.5431,
+      "step": 3341
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.41198000216323,
+      "learning_rate": 7.417124893515022e-06,
+      "loss": 0.5355,
+      "step": 3342
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.616643957338572,
+      "learning_rate": 7.4156006361730395e-06,
+      "loss": 0.536,
+      "step": 3343
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.163860180678707,
+      "learning_rate": 7.414076085931152e-06,
+      "loss": 0.5,
+      "step": 3344
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.968624533192347,
+      "learning_rate": 7.4125512429742166e-06,
+      "loss": 0.4835,
+      "step": 3345
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.377675871783559,
+      "learning_rate": 7.411026107487123e-06,
+      "loss": 0.4896,
+      "step": 3346
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.817423502208232,
+      "learning_rate": 7.409500679654805e-06,
+      "loss": 0.5213,
+      "step": 3347
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.8510721285100535,
+      "learning_rate": 7.407974959662222e-06,
+      "loss": 0.4642,
+      "step": 3348
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.666201171256702,
+      "learning_rate": 7.406448947694374e-06,
+      "loss": 0.5716,
+      "step": 3349
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.127971874199582,
+      "learning_rate": 7.404922643936294e-06,
+      "loss": 0.4847,
+      "step": 3350
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.27940178429152,
+      "learning_rate": 7.403396048573052e-06,
+      "loss": 0.4964,
+      "step": 3351
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.953577398990107,
+      "learning_rate": 7.4018691617897545e-06,
+      "loss": 0.5041,
+      "step": 3352
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.524232492518838,
+      "learning_rate": 7.40034198377154e-06,
+      "loss": 0.5581,
+      "step": 3353
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.087536970829396,
+      "learning_rate": 7.398814514703585e-06,
+      "loss": 0.5313,
+      "step": 3354
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.4495860506781755,
+      "learning_rate": 7.397286754771099e-06,
+      "loss": 0.5023,
+      "step": 3355
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.48796261851049,
+      "learning_rate": 7.39575870415933e-06,
+      "loss": 0.5179,
+      "step": 3356
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.218246101838255,
+      "learning_rate": 7.394230363053558e-06,
+      "loss": 0.4962,
+      "step": 3357
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.668406598683735,
+      "learning_rate": 7.3927017316391024e-06,
+      "loss": 0.5226,
+      "step": 3358
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.42989612020399903,
+      "learning_rate": 7.391172810101311e-06,
+      "loss": 0.302,
+      "step": 3359
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.428490911464109,
+      "learning_rate": 7.389643598625574e-06,
+      "loss": 0.4755,
+      "step": 3360
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.976430537431207,
+      "learning_rate": 7.388114097397312e-06,
+      "loss": 0.4908,
+      "step": 3361
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7536046485924994,
+      "learning_rate": 7.386584306601982e-06,
+      "loss": 0.4929,
+      "step": 3362
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.6048909925742,
+      "learning_rate": 7.385054226425077e-06,
+      "loss": 0.4693,
+      "step": 3363
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.050367417518781,
+      "learning_rate": 7.383523857052125e-06,
+      "loss": 0.5053,
+      "step": 3364
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.939685637042088,
+      "learning_rate": 7.381993198668689e-06,
+      "loss": 0.5298,
+      "step": 3365
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.42346362851904357,
+      "learning_rate": 7.380462251460364e-06,
+      "loss": 0.2934,
+      "step": 3366
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.619259036544197,
+      "learning_rate": 7.378931015612786e-06,
+      "loss": 0.4771,
+      "step": 3367
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.765547046142056,
+      "learning_rate": 7.377399491311619e-06,
+      "loss": 0.4933,
+      "step": 3368
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.339098580667162,
+      "learning_rate": 7.375867678742567e-06,
+      "loss": 0.4827,
+      "step": 3369
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.797468977182747,
+      "learning_rate": 7.374335578091372e-06,
+      "loss": 0.5115,
+      "step": 3370
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3993589669647666,
+      "learning_rate": 7.372803189543799e-06,
+      "loss": 0.2991,
+      "step": 3371
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.786920756289684,
+      "learning_rate": 7.37127051328566e-06,
+      "loss": 0.5122,
+      "step": 3372
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.963857644279904,
+      "learning_rate": 7.369737549502797e-06,
+      "loss": 0.549,
+      "step": 3373
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.572958223564708,
+      "learning_rate": 7.368204298381085e-06,
+      "loss": 0.4948,
+      "step": 3374
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7817041571833214,
+      "learning_rate": 7.366670760106439e-06,
+      "loss": 0.4904,
+      "step": 3375
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.187679867697758,
+      "learning_rate": 7.365136934864804e-06,
+      "loss": 0.5139,
+      "step": 3376
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.973004687683625,
+      "learning_rate": 7.3636028228421595e-06,
+      "loss": 0.5351,
+      "step": 3377
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.173082949337776,
+      "learning_rate": 7.3620684242245264e-06,
+      "loss": 0.5602,
+      "step": 3378
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.5503602687385,
+      "learning_rate": 7.360533739197952e-06,
+      "loss": 0.5053,
+      "step": 3379
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.6248891355853745,
+      "learning_rate": 7.358998767948525e-06,
+      "loss": 0.5425,
+      "step": 3380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.281482127848753,
+      "learning_rate": 7.357463510662365e-06,
+      "loss": 0.4986,
+      "step": 3381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.101108165520043,
+      "learning_rate": 7.355927967525626e-06,
+      "loss": 0.503,
+      "step": 3382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3742638590048757,
+      "learning_rate": 7.354392138724499e-06,
+      "loss": 0.4858,
+      "step": 3383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.149488406054568,
+      "learning_rate": 7.352856024445208e-06,
+      "loss": 0.4591,
+      "step": 3384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.88396514174075,
+      "learning_rate": 7.351319624874013e-06,
+      "loss": 0.5274,
+      "step": 3385
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.323035344229328,
+      "learning_rate": 7.349782940197208e-06,
+      "loss": 0.5758,
+      "step": 3386
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.494941211526117,
+      "learning_rate": 7.34824597060112e-06,
+      "loss": 0.503,
+      "step": 3387
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.883710058877308,
+      "learning_rate": 7.34670871627211e-06,
+      "loss": 0.5418,
+      "step": 3388
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.42768912850726676,
+      "learning_rate": 7.34517117739658e-06,
+      "loss": 0.2955,
+      "step": 3389
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.531430448675565,
+      "learning_rate": 7.343633354160961e-06,
+      "loss": 0.5738,
+      "step": 3390
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.361921345928044,
+      "learning_rate": 7.342095246751717e-06,
+      "loss": 0.5405,
+      "step": 3391
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.847257731253822,
+      "learning_rate": 7.340556855355349e-06,
+      "loss": 0.5263,
+      "step": 3392
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.726919070605619,
+      "learning_rate": 7.339018180158395e-06,
+      "loss": 0.5201,
+      "step": 3393
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.379805675355239,
+      "learning_rate": 7.33747922134742e-06,
+      "loss": 0.511,
+      "step": 3394
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.921670789276575,
+      "learning_rate": 7.335939979109032e-06,
+      "loss": 0.4569,
+      "step": 3395
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.029574805250892,
+      "learning_rate": 7.334400453629869e-06,
+      "loss": 0.5104,
+      "step": 3396
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.379238377035355,
+      "learning_rate": 7.332860645096604e-06,
+      "loss": 0.4449,
+      "step": 3397
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.40302062315442466,
+      "learning_rate": 7.331320553695942e-06,
+      "loss": 0.2978,
+      "step": 3398
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.4408889515338625,
+      "learning_rate": 7.329780179614625e-06,
+      "loss": 0.4981,
+      "step": 3399
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 105.80604396777599,
+      "learning_rate": 7.328239523039431e-06,
+      "loss": 0.5423,
+      "step": 3400
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 14.817280372051039,
+      "learning_rate": 7.3266985841571666e-06,
+      "loss": 0.4746,
+      "step": 3401
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.655391170563532,
+      "learning_rate": 7.325157363154678e-06,
+      "loss": 0.4837,
+      "step": 3402
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.781080914485068,
+      "learning_rate": 7.323615860218844e-06,
+      "loss": 0.5359,
+      "step": 3403
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.478120467248916,
+      "learning_rate": 7.322074075536575e-06,
+      "loss": 0.4601,
+      "step": 3404
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.953330907247552,
+      "learning_rate": 7.320532009294818e-06,
+      "loss": 0.4614,
+      "step": 3405
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.69716851221163,
+      "learning_rate": 7.3189896616805565e-06,
+      "loss": 0.516,
+      "step": 3406
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 14.224920541511205,
+      "learning_rate": 7.317447032880804e-06,
+      "loss": 0.5516,
+      "step": 3407
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.478507424021773,
+      "learning_rate": 7.315904123082609e-06,
+      "loss": 0.5138,
+      "step": 3408
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.353480871546853,
+      "learning_rate": 7.314360932473054e-06,
+      "loss": 0.5256,
+      "step": 3409
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.36019122439259,
+      "learning_rate": 7.312817461239259e-06,
+      "loss": 0.5382,
+      "step": 3410
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.1499350015197045,
+      "learning_rate": 7.311273709568371e-06,
+      "loss": 0.4827,
+      "step": 3411
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.681131079464099,
+      "learning_rate": 7.309729677647579e-06,
+      "loss": 0.5082,
+      "step": 3412
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4284017088164907,
+      "learning_rate": 7.3081853656641e-06,
+      "loss": 0.2873,
+      "step": 3413
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3853791917196803,
+      "learning_rate": 7.306640773805188e-06,
+      "loss": 0.2952,
+      "step": 3414
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.896351651629313,
+      "learning_rate": 7.30509590225813e-06,
+      "loss": 0.5329,
+      "step": 3415
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4334645538487782,
+      "learning_rate": 7.303550751210247e-06,
+      "loss": 0.478,
+      "step": 3416
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.9029041711731223,
+      "learning_rate": 7.302005320848894e-06,
+      "loss": 0.5498,
+      "step": 3417
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.94786330293285,
+      "learning_rate": 7.300459611361461e-06,
+      "loss": 0.4858,
+      "step": 3418
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4441337567857859,
+      "learning_rate": 7.298913622935366e-06,
+      "loss": 0.3164,
+      "step": 3419
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.9230000915329915,
+      "learning_rate": 7.297367355758072e-06,
+      "loss": 0.5018,
+      "step": 3420
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.1962638917894655,
+      "learning_rate": 7.295820810017065e-06,
+      "loss": 0.4875,
+      "step": 3421
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.775144306183487,
+      "learning_rate": 7.294273985899868e-06,
+      "loss": 0.5096,
+      "step": 3422
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3974558223477525,
+      "learning_rate": 7.292726883594043e-06,
+      "loss": 0.5136,
+      "step": 3423
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.369191896274544,
+      "learning_rate": 7.291179503287178e-06,
+      "loss": 0.5115,
+      "step": 3424
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.0846966666469795,
+      "learning_rate": 7.2896318451668975e-06,
+      "loss": 0.5082,
+      "step": 3425
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.014616469015428,
+      "learning_rate": 7.288083909420866e-06,
+      "loss": 0.5628,
+      "step": 3426
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38672015651568403,
+      "learning_rate": 7.28653569623677e-06,
+      "loss": 0.3081,
+      "step": 3427
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.641516932585064,
+      "learning_rate": 7.284987205802338e-06,
+      "loss": 0.4724,
+      "step": 3428
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.8014327965146872,
+      "learning_rate": 7.28343843830533e-06,
+      "loss": 0.5335,
+      "step": 3429
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.839561777189393,
+      "learning_rate": 7.281889393933539e-06,
+      "loss": 0.5121,
+      "step": 3430
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.909852009903813,
+      "learning_rate": 7.280340072874792e-06,
+      "loss": 0.5418,
+      "step": 3431
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.081266013283753,
+      "learning_rate": 7.27879047531695e-06,
+      "loss": 0.5366,
+      "step": 3432
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.46985183542515,
+      "learning_rate": 7.277240601447908e-06,
+      "loss": 0.497,
+      "step": 3433
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.2251587089161164,
+      "learning_rate": 7.27569045145559e-06,
+      "loss": 0.4761,
+      "step": 3434
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.394145611259116,
+      "learning_rate": 7.27414002552796e-06,
+      "loss": 0.5039,
+      "step": 3435
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.166587019723552,
+      "learning_rate": 7.272589323853012e-06,
+      "loss": 0.5211,
+      "step": 3436
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5204562338387895,
+      "learning_rate": 7.271038346618774e-06,
+      "loss": 0.5263,
+      "step": 3437
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.223714833422772,
+      "learning_rate": 7.2694870940133065e-06,
+      "loss": 0.5058,
+      "step": 3438
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 29.436285273469306,
+      "learning_rate": 7.267935566224705e-06,
+      "loss": 0.4941,
+      "step": 3439
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.167646233986877,
+      "learning_rate": 7.2663837634410976e-06,
+      "loss": 0.4765,
+      "step": 3440
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.129636741274262,
+      "learning_rate": 7.2648316858506455e-06,
+      "loss": 0.5445,
+      "step": 3441
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.378583523899565,
+      "learning_rate": 7.263279333641542e-06,
+      "loss": 0.4665,
+      "step": 3442
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5263663065316546,
+      "learning_rate": 7.261726707002017e-06,
+      "loss": 0.5073,
+      "step": 3443
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5448342232241763,
+      "learning_rate": 7.260173806120332e-06,
+      "loss": 0.5184,
+      "step": 3444
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.592089515544115,
+      "learning_rate": 7.258620631184782e-06,
+      "loss": 0.526,
+      "step": 3445
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.7792593262706355,
+      "learning_rate": 7.2570671823836925e-06,
+      "loss": 0.4887,
+      "step": 3446
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.903345515787338,
+      "learning_rate": 7.255513459905426e-06,
+      "loss": 0.5023,
+      "step": 3447
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.904978271519345,
+      "learning_rate": 7.253959463938376e-06,
+      "loss": 0.523,
+      "step": 3448
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.367983843783808,
+      "learning_rate": 7.252405194670973e-06,
+      "loss": 0.4779,
+      "step": 3449
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.167787253194065,
+      "learning_rate": 7.250850652291671e-06,
+      "loss": 0.4942,
+      "step": 3450
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.41110920933084,
+      "learning_rate": 7.24929583698897e-06,
+      "loss": 0.5045,
+      "step": 3451
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.008543800579565,
+      "learning_rate": 7.247740748951394e-06,
+      "loss": 0.4755,
+      "step": 3452
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0104874049986816,
+      "learning_rate": 7.246185388367502e-06,
+      "loss": 0.5406,
+      "step": 3453
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7095330188578877,
+      "learning_rate": 7.2446297554258895e-06,
+      "loss": 0.4492,
+      "step": 3454
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2922635199272023,
+      "learning_rate": 7.24307385031518e-06,
+      "loss": 0.4961,
+      "step": 3455
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.811234330758058,
+      "learning_rate": 7.2415176732240335e-06,
+      "loss": 0.5161,
+      "step": 3456
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.152714002743044,
+      "learning_rate": 7.239961224341141e-06,
+      "loss": 0.5845,
+      "step": 3457
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.392026037614858,
+      "learning_rate": 7.238404503855228e-06,
+      "loss": 0.5021,
+      "step": 3458
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1379818842572256,
+      "learning_rate": 7.236847511955051e-06,
+      "loss": 0.5133,
+      "step": 3459
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.280379918263227,
+      "learning_rate": 7.235290248829403e-06,
+      "loss": 0.5288,
+      "step": 3460
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4370685784428253,
+      "learning_rate": 7.233732714667104e-06,
+      "loss": 0.3033,
+      "step": 3461
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.575826859637568,
+      "learning_rate": 7.232174909657015e-06,
+      "loss": 0.4964,
+      "step": 3462
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.381508709191943,
+      "learning_rate": 7.230616833988021e-06,
+      "loss": 0.4785,
+      "step": 3463
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.943785882304751,
+      "learning_rate": 7.229058487849045e-06,
+      "loss": 0.4969,
+      "step": 3464
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1265724764883034,
+      "learning_rate": 7.227499871429047e-06,
+      "loss": 0.521,
+      "step": 3465
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.843557840921077,
+      "learning_rate": 7.225940984917007e-06,
+      "loss": 0.4709,
+      "step": 3466
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.36974999246031487,
+      "learning_rate": 7.224381828501947e-06,
+      "loss": 0.2915,
+      "step": 3467
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.534299588192421,
+      "learning_rate": 7.222822402372924e-06,
+      "loss": 0.5708,
+      "step": 3468
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.177553584797923,
+      "learning_rate": 7.221262706719021e-06,
+      "loss": 0.5311,
+      "step": 3469
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.438310627150371,
+      "learning_rate": 7.219702741729357e-06,
+      "loss": 0.4738,
+      "step": 3470
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8670136246527957,
+      "learning_rate": 7.2181425075930845e-06,
+      "loss": 0.5093,
+      "step": 3471
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9742402984107104,
+      "learning_rate": 7.2165820044993835e-06,
+      "loss": 0.5111,
+      "step": 3472
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.0395316918612085,
+      "learning_rate": 7.215021232637474e-06,
+      "loss": 0.4503,
+      "step": 3473
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.166744229160644,
+      "learning_rate": 7.213460192196603e-06,
+      "loss": 0.5559,
+      "step": 3474
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.911094920601515,
+      "learning_rate": 7.211898883366053e-06,
+      "loss": 0.536,
+      "step": 3475
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2588955428960733,
+      "learning_rate": 7.210337306335138e-06,
+      "loss": 0.5182,
+      "step": 3476
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.8420617646163,
+      "learning_rate": 7.208775461293205e-06,
+      "loss": 0.4856,
+      "step": 3477
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.8404641080170516,
+      "learning_rate": 7.20721334842963e-06,
+      "loss": 0.5067,
+      "step": 3478
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.625873922053,
+      "learning_rate": 7.205650967933829e-06,
+      "loss": 0.5181,
+      "step": 3479
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.424647600263937,
+      "learning_rate": 7.204088319995245e-06,
+      "loss": 0.3048,
+      "step": 3480
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.158740037679686,
+      "learning_rate": 7.202525404803352e-06,
+      "loss": 0.516,
+      "step": 3481
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8917437321604127,
+      "learning_rate": 7.2009622225476616e-06,
+      "loss": 0.4728,
+      "step": 3482
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1721753577582588,
+      "learning_rate": 7.199398773417714e-06,
+      "loss": 0.4709,
+      "step": 3483
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.903302530885904,
+      "learning_rate": 7.197835057603079e-06,
+      "loss": 0.5067,
+      "step": 3484
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.761942605172721,
+      "learning_rate": 7.196271075293371e-06,
+      "loss": 0.5638,
+      "step": 3485
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3906955532294063,
+      "learning_rate": 7.194706826678222e-06,
+      "loss": 0.5179,
+      "step": 3486
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.068397083235496,
+      "learning_rate": 7.1931423119473035e-06,
+      "loss": 0.4892,
+      "step": 3487
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.257632334713975,
+      "learning_rate": 7.191577531290318e-06,
+      "loss": 0.4614,
+      "step": 3488
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.420396131468128,
+      "learning_rate": 7.190012484897002e-06,
+      "loss": 0.5329,
+      "step": 3489
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.890481260795104,
+      "learning_rate": 7.188447172957121e-06,
+      "loss": 0.5075,
+      "step": 3490
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.4555582821086315,
+      "learning_rate": 7.186881595660477e-06,
+      "loss": 0.5201,
+      "step": 3491
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.036948605719086,
+      "learning_rate": 7.1853157531969e-06,
+      "loss": 0.519,
+      "step": 3492
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0560489196147977,
+      "learning_rate": 7.183749645756254e-06,
+      "loss": 0.5057,
+      "step": 3493
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.682610697069687,
+      "learning_rate": 7.182183273528436e-06,
+      "loss": 0.5378,
+      "step": 3494
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6581298556002073,
+      "learning_rate": 7.18061663670337e-06,
+      "loss": 0.5508,
+      "step": 3495
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.633466949294051,
+      "learning_rate": 7.1790497354710205e-06,
+      "loss": 0.4769,
+      "step": 3496
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 31.79975678187181,
+      "learning_rate": 7.17748257002138e-06,
+      "loss": 0.5205,
+      "step": 3497
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.07138416789549,
+      "learning_rate": 7.175915140544469e-06,
+      "loss": 0.4959,
+      "step": 3498
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.358348254681237,
+      "learning_rate": 7.174347447230346e-06,
+      "loss": 0.5343,
+      "step": 3499
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.1830798566808145,
+      "learning_rate": 7.172779490269099e-06,
+      "loss": 0.4967,
+      "step": 3500
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.112678023625594,
+      "learning_rate": 7.171211269850847e-06,
+      "loss": 0.5319,
+      "step": 3501
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.516599980403883,
+      "learning_rate": 7.169642786165747e-06,
+      "loss": 0.4773,
+      "step": 3502
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.272748295187081,
+      "learning_rate": 7.168074039403975e-06,
+      "loss": 0.4742,
+      "step": 3503
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3966319952007207,
+      "learning_rate": 7.166505029755753e-06,
+      "loss": 0.4787,
+      "step": 3504
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.6078145493253455,
+      "learning_rate": 7.1649357574113275e-06,
+      "loss": 0.5064,
+      "step": 3505
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.915724666998266,
+      "learning_rate": 7.163366222560977e-06,
+      "loss": 0.4884,
+      "step": 3506
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.977365853171872,
+      "learning_rate": 7.161796425395014e-06,
+      "loss": 0.4789,
+      "step": 3507
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4233535797929824,
+      "learning_rate": 7.160226366103782e-06,
+      "loss": 0.3038,
+      "step": 3508
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.368950771716921,
+      "learning_rate": 7.158656044877654e-06,
+      "loss": 0.4757,
+      "step": 3509
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.671472111712637,
+      "learning_rate": 7.15708546190704e-06,
+      "loss": 0.5483,
+      "step": 3510
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.952931848421995,
+      "learning_rate": 7.155514617382377e-06,
+      "loss": 0.4748,
+      "step": 3511
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.335131310802297,
+      "learning_rate": 7.153943511494135e-06,
+      "loss": 0.525,
+      "step": 3512
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.658245196436563,
+      "learning_rate": 7.152372144432818e-06,
+      "loss": 0.5587,
+      "step": 3513
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.029423646259287,
+      "learning_rate": 7.1508005163889575e-06,
+      "loss": 0.5279,
+      "step": 3514
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.271682676813936,
+      "learning_rate": 7.1492286275531175e-06,
+      "loss": 0.5366,
+      "step": 3515
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.157709415949006,
+      "learning_rate": 7.147656478115899e-06,
+      "loss": 0.5281,
+      "step": 3516
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.470394085456436,
+      "learning_rate": 7.146084068267928e-06,
+      "loss": 0.5519,
+      "step": 3517
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.8830540550323835,
+      "learning_rate": 7.144511398199865e-06,
+      "loss": 0.5078,
+      "step": 3518
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.128261084694632,
+      "learning_rate": 7.1429384681024e-06,
+      "loss": 0.4554,
+      "step": 3519
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.539266169926404,
+      "learning_rate": 7.141365278166261e-06,
+      "loss": 0.5045,
+      "step": 3520
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.9955554759038217,
+      "learning_rate": 7.1397918285821965e-06,
+      "loss": 0.5048,
+      "step": 3521
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.475317837082806,
+      "learning_rate": 7.138218119540998e-06,
+      "loss": 0.4887,
+      "step": 3522
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.765725930744946,
+      "learning_rate": 7.1366441512334806e-06,
+      "loss": 0.5562,
+      "step": 3523
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 13.239841641823796,
+      "learning_rate": 7.135069923850493e-06,
+      "loss": 0.5343,
+      "step": 3524
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.108769123358908,
+      "learning_rate": 7.133495437582917e-06,
+      "loss": 0.5112,
+      "step": 3525
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6359836426770595,
+      "learning_rate": 7.131920692621663e-06,
+      "loss": 0.5051,
+      "step": 3526
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.77619094962636,
+      "learning_rate": 7.130345689157676e-06,
+      "loss": 0.518,
+      "step": 3527
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.347379649362326,
+      "learning_rate": 7.12877042738193e-06,
+      "loss": 0.4795,
+      "step": 3528
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.8612020170920385,
+      "learning_rate": 7.127194907485431e-06,
+      "loss": 0.4946,
+      "step": 3529
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.989892636593509,
+      "learning_rate": 7.125619129659215e-06,
+      "loss": 0.4813,
+      "step": 3530
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.556035629604659,
+      "learning_rate": 7.1240430940943516e-06,
+      "loss": 0.4741,
+      "step": 3531
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.826185001833184,
+      "learning_rate": 7.12246680098194e-06,
+      "loss": 0.5476,
+      "step": 3532
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.4240319696805885,
+      "learning_rate": 7.120890250513111e-06,
+      "loss": 0.5118,
+      "step": 3533
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.5369825631562355,
+      "learning_rate": 7.119313442879029e-06,
+      "loss": 0.497,
+      "step": 3534
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.154139773000001,
+      "learning_rate": 7.1177363782708855e-06,
+      "loss": 0.5142,
+      "step": 3535
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 164.90485449941195,
+      "learning_rate": 7.116159056879904e-06,
+      "loss": 0.4958,
+      "step": 3536
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.531430019121945,
+      "learning_rate": 7.114581478897342e-06,
+      "loss": 0.558,
+      "step": 3537
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.801741936065927,
+      "learning_rate": 7.113003644514486e-06,
+      "loss": 0.5039,
+      "step": 3538
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.85108686602525,
+      "learning_rate": 7.111425553922653e-06,
+      "loss": 0.4963,
+      "step": 3539
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.334049929582696,
+      "learning_rate": 7.109847207313192e-06,
+      "loss": 0.561,
+      "step": 3540
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.899001296930127,
+      "learning_rate": 7.108268604877484e-06,
+      "loss": 0.5117,
+      "step": 3541
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.9927279460654166,
+      "learning_rate": 7.1066897468069394e-06,
+      "loss": 0.5734,
+      "step": 3542
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.813497907655462,
+      "learning_rate": 7.105110633292999e-06,
+      "loss": 0.478,
+      "step": 3543
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4754694629862224,
+      "learning_rate": 7.103531264527138e-06,
+      "loss": 0.2894,
+      "step": 3544
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4589748835016198,
+      "learning_rate": 7.10195164070086e-06,
+      "loss": 0.2918,
+      "step": 3545
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.16348926794708,
+      "learning_rate": 7.100371762005698e-06,
+      "loss": 0.4739,
+      "step": 3546
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.030825845862654,
+      "learning_rate": 7.098791628633218e-06,
+      "loss": 0.5317,
+      "step": 3547
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.599566756117023,
+      "learning_rate": 7.097211240775018e-06,
+      "loss": 0.4856,
+      "step": 3548
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3098695084300296,
+      "learning_rate": 7.095630598622725e-06,
+      "loss": 0.5259,
+      "step": 3549
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.953507087843633,
+      "learning_rate": 7.094049702367998e-06,
+      "loss": 0.5012,
+      "step": 3550
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6058245399266746,
+      "learning_rate": 7.092468552202524e-06,
+      "loss": 0.4861,
+      "step": 3551
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.9343176214275255,
+      "learning_rate": 7.090887148318024e-06,
+      "loss": 0.4961,
+      "step": 3552
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.417316304573027,
+      "learning_rate": 7.089305490906251e-06,
+      "loss": 0.5391,
+      "step": 3553
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.649175871620648,
+      "learning_rate": 7.0877235801589825e-06,
+      "loss": 0.5076,
+      "step": 3554
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.844122940347858,
+      "learning_rate": 7.086141416268033e-06,
+      "loss": 0.5284,
+      "step": 3555
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7605269588325425,
+      "learning_rate": 7.084558999425245e-06,
+      "loss": 0.3251,
+      "step": 3556
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.930730001239317,
+      "learning_rate": 7.082976329822491e-06,
+      "loss": 0.5062,
+      "step": 3557
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.276047923988863,
+      "learning_rate": 7.081393407651676e-06,
+      "loss": 0.5261,
+      "step": 3558
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.075205423868818,
+      "learning_rate": 7.0798102331047345e-06,
+      "loss": 0.5054,
+      "step": 3559
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.054494126862988,
+      "learning_rate": 7.078226806373632e-06,
+      "loss": 0.5374,
+      "step": 3560
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.765411649329209,
+      "learning_rate": 7.076643127650367e-06,
+      "loss": 0.496,
+      "step": 3561
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.367690285908298,
+      "learning_rate": 7.075059197126961e-06,
+      "loss": 0.4917,
+      "step": 3562
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.3025477904912295,
+      "learning_rate": 7.073475014995472e-06,
+      "loss": 0.5033,
+      "step": 3563
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5147743754810318,
+      "learning_rate": 7.071890581447992e-06,
+      "loss": 0.3121,
+      "step": 3564
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.3641676554740885,
+      "learning_rate": 7.070305896676635e-06,
+      "loss": 0.4996,
+      "step": 3565
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.469494910232441,
+      "learning_rate": 7.068720960873552e-06,
+      "loss": 0.5363,
+      "step": 3566
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.66862590370772,
+      "learning_rate": 7.067135774230919e-06,
+      "loss": 0.5005,
+      "step": 3567
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.232167418204643,
+      "learning_rate": 7.065550336940947e-06,
+      "loss": 0.4979,
+      "step": 3568
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.0798378524879935,
+      "learning_rate": 7.063964649195874e-06,
+      "loss": 0.4751,
+      "step": 3569
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.804111423768178,
+      "learning_rate": 7.0623787111879734e-06,
+      "loss": 0.5563,
+      "step": 3570
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.080872811986546,
+      "learning_rate": 7.060792523109545e-06,
+      "loss": 0.4917,
+      "step": 3571
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.098126378074902,
+      "learning_rate": 7.059206085152918e-06,
+      "loss": 0.4768,
+      "step": 3572
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.37334877385613,
+      "learning_rate": 7.057619397510453e-06,
+      "loss": 0.5459,
+      "step": 3573
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.821965017254686,
+      "learning_rate": 7.056032460374542e-06,
+      "loss": 0.4812,
+      "step": 3574
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4026128956083457,
+      "learning_rate": 7.054445273937609e-06,
+      "loss": 0.4689,
+      "step": 3575
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.516324705146607,
+      "learning_rate": 7.052857838392104e-06,
+      "loss": 0.5808,
+      "step": 3576
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5675062235690715,
+      "learning_rate": 7.0512701539305065e-06,
+      "loss": 0.4625,
+      "step": 3577
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2299773926048942,
+      "learning_rate": 7.049682220745332e-06,
+      "loss": 0.5018,
+      "step": 3578
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.204745186786549,
+      "learning_rate": 7.048094039029123e-06,
+      "loss": 0.5333,
+      "step": 3579
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.096573899519898,
+      "learning_rate": 7.046505608974448e-06,
+      "loss": 0.552,
+      "step": 3580
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.154378528170588,
+      "learning_rate": 7.044916930773915e-06,
+      "loss": 0.4959,
+      "step": 3581
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.554927177084635,
+      "learning_rate": 7.043328004620154e-06,
+      "loss": 0.5629,
+      "step": 3582
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.96845928723012,
+      "learning_rate": 7.041738830705828e-06,
+      "loss": 0.5162,
+      "step": 3583
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.43759714433948815,
+      "learning_rate": 7.040149409223629e-06,
+      "loss": 0.3069,
+      "step": 3584
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.603521472124815,
+      "learning_rate": 7.038559740366282e-06,
+      "loss": 0.5261,
+      "step": 3585
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5754060509725005,
+      "learning_rate": 7.036969824326536e-06,
+      "loss": 0.4717,
+      "step": 3586
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.748634110158527,
+      "learning_rate": 7.035379661297179e-06,
+      "loss": 0.5226,
+      "step": 3587
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.956676724701758,
+      "learning_rate": 7.033789251471019e-06,
+      "loss": 0.4963,
+      "step": 3588
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.38516171222131224,
+      "learning_rate": 7.032198595040902e-06,
+      "loss": 0.2899,
+      "step": 3589
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.481509698540348,
+      "learning_rate": 7.0306076921997e-06,
+      "loss": 0.5076,
+      "step": 3590
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8914273609087053,
+      "learning_rate": 7.029016543140312e-06,
+      "loss": 0.4813,
+      "step": 3591
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.918882140454263,
+      "learning_rate": 7.027425148055676e-06,
+      "loss": 0.5751,
+      "step": 3592
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.186479757487995,
+      "learning_rate": 7.025833507138751e-06,
+      "loss": 0.5576,
+      "step": 3593
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 14.584156061522641,
+      "learning_rate": 7.024241620582527e-06,
+      "loss": 0.4769,
+      "step": 3594
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8117460148515985,
+      "learning_rate": 7.022649488580029e-06,
+      "loss": 0.4752,
+      "step": 3595
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.83232722341304,
+      "learning_rate": 7.021057111324307e-06,
+      "loss": 0.4962,
+      "step": 3596
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9979918559205654,
+      "learning_rate": 7.019464489008443e-06,
+      "loss": 0.4368,
+      "step": 3597
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.800330712009503,
+      "learning_rate": 7.017871621825549e-06,
+      "loss": 0.4718,
+      "step": 3598
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7096013688886447,
+      "learning_rate": 7.016278509968762e-06,
+      "loss": 0.5202,
+      "step": 3599
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.237840701022627,
+      "learning_rate": 7.014685153631255e-06,
+      "loss": 0.5099,
+      "step": 3600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.008598428961999,
+      "learning_rate": 7.013091553006228e-06,
+      "loss": 0.5132,
+      "step": 3601
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.468808398253617,
+      "learning_rate": 7.01149770828691e-06,
+      "loss": 0.5561,
+      "step": 3602
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.956019340614767,
+      "learning_rate": 7.009903619666561e-06,
+      "loss": 0.4911,
+      "step": 3603
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.64845782957505,
+      "learning_rate": 7.008309287338467e-06,
+      "loss": 0.4919,
+      "step": 3604
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4595493271650435,
+      "learning_rate": 7.006714711495949e-06,
+      "loss": 0.3046,
+      "step": 3605
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6382326616202274,
+      "learning_rate": 7.0051198923323535e-06,
+      "loss": 0.4808,
+      "step": 3606
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.436271672071668,
+      "learning_rate": 7.0035248300410595e-06,
+      "loss": 0.486,
+      "step": 3607
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.291250399588223,
+      "learning_rate": 7.0019295248154714e-06,
+      "loss": 0.4646,
+      "step": 3608
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.056380595069087,
+      "learning_rate": 7.000333976849029e-06,
+      "loss": 0.4931,
+      "step": 3609
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.7703726219157545,
+      "learning_rate": 6.998738186335194e-06,
+      "loss": 0.4833,
+      "step": 3610
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.353500041361122,
+      "learning_rate": 6.9971421534674625e-06,
+      "loss": 0.5842,
+      "step": 3611
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.775890718627159,
+      "learning_rate": 6.99554587843936e-06,
+      "loss": 0.5427,
+      "step": 3612
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.2489451453568945,
+      "learning_rate": 6.993949361444441e-06,
+      "loss": 0.5056,
+      "step": 3613
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7806712358317953,
+      "learning_rate": 6.992352602676288e-06,
+      "loss": 0.5303,
+      "step": 3614
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.526959819672607,
+      "learning_rate": 6.990755602328511e-06,
+      "loss": 0.5162,
+      "step": 3615
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.596954639650535,
+      "learning_rate": 6.9891583605947565e-06,
+      "loss": 0.4925,
+      "step": 3616
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.83383204660192,
+      "learning_rate": 6.987560877668692e-06,
+      "loss": 0.5028,
+      "step": 3617
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.394416957650362,
+      "learning_rate": 6.9859631537440195e-06,
+      "loss": 0.5332,
+      "step": 3618
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.831886333589942,
+      "learning_rate": 6.984365189014468e-06,
+      "loss": 0.5107,
+      "step": 3619
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.7648622678538315,
+      "learning_rate": 6.982766983673796e-06,
+      "loss": 0.5446,
+      "step": 3620
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.157205881577877,
+      "learning_rate": 6.9811685379157916e-06,
+      "loss": 0.5162,
+      "step": 3621
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4356864381807791,
+      "learning_rate": 6.979569851934271e-06,
+      "loss": 0.3051,
+      "step": 3622
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.009642390797965,
+      "learning_rate": 6.977970925923081e-06,
+      "loss": 0.5053,
+      "step": 3623
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.44556717071008,
+      "learning_rate": 6.976371760076099e-06,
+      "loss": 0.5073,
+      "step": 3624
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.4268597875544,
+      "learning_rate": 6.974772354587226e-06,
+      "loss": 0.5147,
+      "step": 3625
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.27689339789211,
+      "learning_rate": 6.973172709650397e-06,
+      "loss": 0.5166,
+      "step": 3626
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.383887344636901,
+      "learning_rate": 6.971572825459576e-06,
+      "loss": 0.4724,
+      "step": 3627
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.361263125453349,
+      "learning_rate": 6.96997270220875e-06,
+      "loss": 0.4772,
+      "step": 3628
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5077166161521647,
+      "learning_rate": 6.9683723400919455e-06,
+      "loss": 0.5351,
+      "step": 3629
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.667417256291665,
+      "learning_rate": 6.9667717393032065e-06,
+      "loss": 0.5809,
+      "step": 3630
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3920918561959245,
+      "learning_rate": 6.965170900036614e-06,
+      "loss": 0.3004,
+      "step": 3631
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.864529480622879,
+      "learning_rate": 6.963569822486275e-06,
+      "loss": 0.4695,
+      "step": 3632
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.488732239770935,
+      "learning_rate": 6.961968506846327e-06,
+      "loss": 0.485,
+      "step": 3633
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.002120877871582,
+      "learning_rate": 6.960366953310931e-06,
+      "loss": 0.5041,
+      "step": 3634
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.957937064100483,
+      "learning_rate": 6.958765162074287e-06,
+      "loss": 0.4852,
+      "step": 3635
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.176680847427811,
+      "learning_rate": 6.957163133330611e-06,
+      "loss": 0.5378,
+      "step": 3636
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 18.158754509460824,
+      "learning_rate": 6.9555608672741596e-06,
+      "loss": 0.5098,
+      "step": 3637
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4122198464980203,
+      "learning_rate": 6.9539583640992096e-06,
+      "loss": 0.3164,
+      "step": 3638
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.581524593876136,
+      "learning_rate": 6.9523556240000725e-06,
+      "loss": 0.5037,
+      "step": 3639
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.168187007798482,
+      "learning_rate": 6.9507526471710865e-06,
+      "loss": 0.4785,
+      "step": 3640
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.089982215778436,
+      "learning_rate": 6.949149433806614e-06,
+      "loss": 0.5043,
+      "step": 3641
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.740816587254558,
+      "learning_rate": 6.947545984101054e-06,
+      "loss": 0.526,
+      "step": 3642
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4132598040784692,
+      "learning_rate": 6.94594229824883e-06,
+      "loss": 0.2919,
+      "step": 3643
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.37872235125756604,
+      "learning_rate": 6.944338376444394e-06,
+      "loss": 0.2923,
+      "step": 3644
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.915321492726467,
+      "learning_rate": 6.942734218882225e-06,
+      "loss": 0.5153,
+      "step": 3645
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1230303811120637,
+      "learning_rate": 6.941129825756836e-06,
+      "loss": 0.5355,
+      "step": 3646
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.280619008780194,
+      "learning_rate": 6.939525197262762e-06,
+      "loss": 0.4926,
+      "step": 3647
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.01530345680956,
+      "learning_rate": 6.937920333594572e-06,
+      "loss": 0.504,
+      "step": 3648
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.805045644777998,
+      "learning_rate": 6.936315234946861e-06,
+      "loss": 0.4813,
+      "step": 3649
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5471256476190782,
+      "learning_rate": 6.934709901514252e-06,
+      "loss": 0.5355,
+      "step": 3650
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.856242886912343,
+      "learning_rate": 6.933104333491399e-06,
+      "loss": 0.5135,
+      "step": 3651
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.18967939005855,
+      "learning_rate": 6.931498531072977e-06,
+      "loss": 0.5106,
+      "step": 3652
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.754560641143927,
+      "learning_rate": 6.929892494453703e-06,
+      "loss": 0.4937,
+      "step": 3653
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.80237953612873,
+      "learning_rate": 6.928286223828309e-06,
+      "loss": 0.5125,
+      "step": 3654
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.04075913661502,
+      "learning_rate": 6.926679719391562e-06,
+      "loss": 0.5406,
+      "step": 3655
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.914295692448582,
+      "learning_rate": 6.9250729813382585e-06,
+      "loss": 0.5383,
+      "step": 3656
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.169675158719112,
+      "learning_rate": 6.923466009863219e-06,
+      "loss": 0.5814,
+      "step": 3657
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.421382956565939,
+      "learning_rate": 6.921858805161295e-06,
+      "loss": 0.4973,
+      "step": 3658
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.162523893904018,
+      "learning_rate": 6.920251367427362e-06,
+      "loss": 0.5661,
+      "step": 3659
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.746942932829003,
+      "learning_rate": 6.918643696856333e-06,
+      "loss": 0.4851,
+      "step": 3660
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.226013687075737,
+      "learning_rate": 6.917035793643142e-06,
+      "loss": 0.4729,
+      "step": 3661
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9897985964397478,
+      "learning_rate": 6.915427657982752e-06,
+      "loss": 0.4805,
+      "step": 3662
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.471225021109507,
+      "learning_rate": 6.913819290070154e-06,
+      "loss": 0.5005,
+      "step": 3663
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 44.035930285928764,
+      "learning_rate": 6.912210690100369e-06,
+      "loss": 0.5434,
+      "step": 3664
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.311946242120669,
+      "learning_rate": 6.910601858268444e-06,
+      "loss": 0.486,
+      "step": 3665
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.93069709169062,
+      "learning_rate": 6.90899279476946e-06,
+      "loss": 0.5198,
+      "step": 3666
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.646583251885376,
+      "learning_rate": 6.907383499798517e-06,
+      "loss": 0.4655,
+      "step": 3667
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7054389494483444,
+      "learning_rate": 6.905773973550748e-06,
+      "loss": 0.49,
+      "step": 3668
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.784844905714255,
+      "learning_rate": 6.904164216221315e-06,
+      "loss": 0.5152,
+      "step": 3669
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.8981268601795005,
+      "learning_rate": 6.902554228005406e-06,
+      "loss": 0.5087,
+      "step": 3670
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.003780032950736,
+      "learning_rate": 6.900944009098238e-06,
+      "loss": 0.5166,
+      "step": 3671
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.988572858696553,
+      "learning_rate": 6.899333559695056e-06,
+      "loss": 0.4799,
+      "step": 3672
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.045195719798457,
+      "learning_rate": 6.897722879991131e-06,
+      "loss": 0.5134,
+      "step": 3673
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.612945419714044,
+      "learning_rate": 6.896111970181765e-06,
+      "loss": 0.5126,
+      "step": 3674
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.361137135003,
+      "learning_rate": 6.894500830462285e-06,
+      "loss": 0.5154,
+      "step": 3675
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1634604103075348,
+      "learning_rate": 6.892889461028048e-06,
+      "loss": 0.5137,
+      "step": 3676
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.009326172647919,
+      "learning_rate": 6.891277862074439e-06,
+      "loss": 0.4976,
+      "step": 3677
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47075867185647746,
+      "learning_rate": 6.889666033796869e-06,
+      "loss": 0.2912,
+      "step": 3678
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 17.310657635433845,
+      "learning_rate": 6.888053976390776e-06,
+      "loss": 0.5075,
+      "step": 3679
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.089944434350161,
+      "learning_rate": 6.886441690051631e-06,
+      "loss": 0.5274,
+      "step": 3680
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.639179280007177,
+      "learning_rate": 6.884829174974927e-06,
+      "loss": 0.4512,
+      "step": 3681
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.619284443764174,
+      "learning_rate": 6.883216431356188e-06,
+      "loss": 0.575,
+      "step": 3682
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.840840683674156,
+      "learning_rate": 6.881603459390964e-06,
+      "loss": 0.5341,
+      "step": 3683
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.8946551900072293,
+      "learning_rate": 6.879990259274832e-06,
+      "loss": 0.5204,
+      "step": 3684
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4008557804231846,
+      "learning_rate": 6.878376831203402e-06,
+      "loss": 0.3102,
+      "step": 3685
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4482492921982923,
+      "learning_rate": 6.876763175372306e-06,
+      "loss": 0.2923,
+      "step": 3686
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.778068069268274,
+      "learning_rate": 6.875149291977201e-06,
+      "loss": 0.4852,
+      "step": 3687
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.498873819407465,
+      "learning_rate": 6.873535181213784e-06,
+      "loss": 0.4966,
+      "step": 3688
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7738005266571983,
+      "learning_rate": 6.871920843277765e-06,
+      "loss": 0.4621,
+      "step": 3689
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4030093194959966,
+      "learning_rate": 6.87030627836489e-06,
+      "loss": 0.3049,
+      "step": 3690
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 15.20394825308246,
+      "learning_rate": 6.868691486670932e-06,
+      "loss": 0.4904,
+      "step": 3691
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.782348126956424,
+      "learning_rate": 6.867076468391688e-06,
+      "loss": 0.5544,
+      "step": 3692
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.904556213060868,
+      "learning_rate": 6.865461223722987e-06,
+      "loss": 0.4921,
+      "step": 3693
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.065100095153117,
+      "learning_rate": 6.86384575286068e-06,
+      "loss": 0.4981,
+      "step": 3694
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.312529636712508,
+      "learning_rate": 6.862230056000648e-06,
+      "loss": 0.5419,
+      "step": 3695
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.360745194267187,
+      "learning_rate": 6.860614133338804e-06,
+      "loss": 0.4882,
+      "step": 3696
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.266867452170587,
+      "learning_rate": 6.858997985071082e-06,
+      "loss": 0.5978,
+      "step": 3697
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4094164415964493,
+      "learning_rate": 6.857381611393445e-06,
+      "loss": 0.3021,
+      "step": 3698
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.274697253702903,
+      "learning_rate": 6.855765012501884e-06,
+      "loss": 0.5169,
+      "step": 3699
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.064041571851097,
+      "learning_rate": 6.8541481885924174e-06,
+      "loss": 0.5236,
+      "step": 3700
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.226780538378126,
+      "learning_rate": 6.852531139861091e-06,
+      "loss": 0.5363,
+      "step": 3701
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.39666442289259607,
+      "learning_rate": 6.850913866503976e-06,
+      "loss": 0.3192,
+      "step": 3702
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.938249331250595,
+      "learning_rate": 6.849296368717176e-06,
+      "loss": 0.467,
+      "step": 3703
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.65482282123723,
+      "learning_rate": 6.847678646696814e-06,
+      "loss": 0.4525,
+      "step": 3704
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.764256727534747,
+      "learning_rate": 6.846060700639047e-06,
+      "loss": 0.5216,
+      "step": 3705
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2819131012649705,
+      "learning_rate": 6.844442530740055e-06,
+      "loss": 0.4789,
+      "step": 3706
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1246016150352065,
+      "learning_rate": 6.842824137196047e-06,
+      "loss": 0.5364,
+      "step": 3707
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.23054858126686,
+      "learning_rate": 6.84120552020326e-06,
+      "loss": 0.516,
+      "step": 3708
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.42795778011857744,
+      "learning_rate": 6.839586679957957e-06,
+      "loss": 0.2977,
+      "step": 3709
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5637748510500984,
+      "learning_rate": 6.837967616656426e-06,
+      "loss": 0.5096,
+      "step": 3710
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.63683412235899,
+      "learning_rate": 6.836348330494985e-06,
+      "loss": 0.5329,
+      "step": 3711
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.784750815200717,
+      "learning_rate": 6.834728821669978e-06,
+      "loss": 0.5144,
+      "step": 3712
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.44322458657390734,
+      "learning_rate": 6.8331090903777754e-06,
+      "loss": 0.3133,
+      "step": 3713
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.141632223021734,
+      "learning_rate": 6.831489136814777e-06,
+      "loss": 0.5054,
+      "step": 3714
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.393041428546813,
+      "learning_rate": 6.829868961177406e-06,
+      "loss": 0.4751,
+      "step": 3715
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.136319539295085,
+      "learning_rate": 6.8282485636621165e-06,
+      "loss": 0.4572,
+      "step": 3716
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.326585968024174,
+      "learning_rate": 6.826627944465384e-06,
+      "loss": 0.506,
+      "step": 3717
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.512875416932784,
+      "learning_rate": 6.825007103783716e-06,
+      "loss": 0.5074,
+      "step": 3718
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.525480671854357,
+      "learning_rate": 6.8233860418136466e-06,
+      "loss": 0.5009,
+      "step": 3719
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 21.84790861912036,
+      "learning_rate": 6.821764758751732e-06,
+      "loss": 0.4964,
+      "step": 3720
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.440147643794619,
+      "learning_rate": 6.820143254794559e-06,
+      "loss": 0.4731,
+      "step": 3721
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.42460393725440887,
+      "learning_rate": 6.818521530138743e-06,
+      "loss": 0.3013,
+      "step": 3722
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.6578659139265226,
+      "learning_rate": 6.816899584980922e-06,
+      "loss": 0.4638,
+      "step": 3723
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.040634975119878,
+      "learning_rate": 6.815277419517761e-06,
+      "loss": 0.5275,
+      "step": 3724
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.247023005641972,
+      "learning_rate": 6.8136550339459566e-06,
+      "loss": 0.4996,
+      "step": 3725
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.419419748512544,
+      "learning_rate": 6.8120324284622255e-06,
+      "loss": 0.5431,
+      "step": 3726
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.797106100558889,
+      "learning_rate": 6.810409603263315e-06,
+      "loss": 0.473,
+      "step": 3727
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.6465986510674733,
+      "learning_rate": 6.808786558546e-06,
+      "loss": 0.5436,
+      "step": 3728
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7130467144327723,
+      "learning_rate": 6.807163294507078e-06,
+      "loss": 0.4817,
+      "step": 3729
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.6761381352702514,
+      "learning_rate": 6.805539811343377e-06,
+      "loss": 0.4454,
+      "step": 3730
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.019268491943481,
+      "learning_rate": 6.803916109251749e-06,
+      "loss": 0.5465,
+      "step": 3731
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.239650639366883,
+      "learning_rate": 6.802292188429072e-06,
+      "loss": 0.5136,
+      "step": 3732
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.33914275431427,
+      "learning_rate": 6.8006680490722566e-06,
+      "loss": 0.5301,
+      "step": 3733
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.531419648943567,
+      "learning_rate": 6.799043691378231e-06,
+      "loss": 0.4762,
+      "step": 3734
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.128934634139693,
+      "learning_rate": 6.797419115543955e-06,
+      "loss": 0.4754,
+      "step": 3735
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.773265695654252,
+      "learning_rate": 6.795794321766416e-06,
+      "loss": 0.4948,
+      "step": 3736
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.651486284081198,
+      "learning_rate": 6.794169310242624e-06,
+      "loss": 0.5014,
+      "step": 3737
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2902987590129578,
+      "learning_rate": 6.7925440811696165e-06,
+      "loss": 0.5081,
+      "step": 3738
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.989368471007804,
+      "learning_rate": 6.79091863474446e-06,
+      "loss": 0.5082,
+      "step": 3739
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.766098608210801,
+      "learning_rate": 6.789292971164245e-06,
+      "loss": 0.5059,
+      "step": 3740
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.147820046886001,
+      "learning_rate": 6.787667090626089e-06,
+      "loss": 0.567,
+      "step": 3741
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.294377805206786,
+      "learning_rate": 6.786040993327135e-06,
+      "loss": 0.4646,
+      "step": 3742
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.754996211172444,
+      "learning_rate": 6.784414679464553e-06,
+      "loss": 0.4781,
+      "step": 3743
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5454762847183696,
+      "learning_rate": 6.7827881492355385e-06,
+      "loss": 0.4947,
+      "step": 3744
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4273297810971787,
+      "learning_rate": 6.7811614028373166e-06,
+      "loss": 0.3135,
+      "step": 3745
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.095530316160898,
+      "learning_rate": 6.779534440467132e-06,
+      "loss": 0.5212,
+      "step": 3746
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.403239500322361,
+      "learning_rate": 6.777907262322263e-06,
+      "loss": 0.5241,
+      "step": 3747
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.890211316122613,
+      "learning_rate": 6.7762798686000086e-06,
+      "loss": 0.5397,
+      "step": 3748
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.542632429395796,
+      "learning_rate": 6.774652259497696e-06,
+      "loss": 0.4975,
+      "step": 3749
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.100170651155246,
+      "learning_rate": 6.773024435212678e-06,
+      "loss": 0.5294,
+      "step": 3750
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.43684552588164915,
+      "learning_rate": 6.7713963959423375e-06,
+      "loss": 0.2972,
+      "step": 3751
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2973318290839635,
+      "learning_rate": 6.769768141884074e-06,
+      "loss": 0.4929,
+      "step": 3752
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.45142167492594665,
+      "learning_rate": 6.768139673235323e-06,
+      "loss": 0.2917,
+      "step": 3753
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.8847548797438325,
+      "learning_rate": 6.7665109901935426e-06,
+      "loss": 0.4972,
+      "step": 3754
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9085830786867026,
+      "learning_rate": 6.764882092956212e-06,
+      "loss": 0.5122,
+      "step": 3755
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.274774133671179,
+      "learning_rate": 6.763252981720845e-06,
+      "loss": 0.5258,
+      "step": 3756
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7929456027016437,
+      "learning_rate": 6.7616236566849736e-06,
+      "loss": 0.5046,
+      "step": 3757
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.51350681121126,
+      "learning_rate": 6.759994118046161e-06,
+      "loss": 0.54,
+      "step": 3758
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.289828499372812,
+      "learning_rate": 6.758364366001995e-06,
+      "loss": 0.4911,
+      "step": 3759
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.39036658978942074,
+      "learning_rate": 6.756734400750087e-06,
+      "loss": 0.304,
+      "step": 3760
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.2197338833231814,
+      "learning_rate": 6.755104222488076e-06,
+      "loss": 0.5038,
+      "step": 3761
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.7760292845672065,
+      "learning_rate": 6.7534738314136285e-06,
+      "loss": 0.5075,
+      "step": 3762
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.0658183303432605,
+      "learning_rate": 6.751843227724433e-06,
+      "loss": 0.4675,
+      "step": 3763
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 25.63532682126861,
+      "learning_rate": 6.7502124116182066e-06,
+      "loss": 0.4976,
+      "step": 3764
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.37743568223542456,
+      "learning_rate": 6.7485813832926905e-06,
+      "loss": 0.3037,
+      "step": 3765
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5646275194068107,
+      "learning_rate": 6.746950142945654e-06,
+      "loss": 0.528,
+      "step": 3766
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.725785530219732,
+      "learning_rate": 6.7453186907748914e-06,
+      "loss": 0.4822,
+      "step": 3767
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.15461638525408,
+      "learning_rate": 6.743687026978219e-06,
+      "loss": 0.5391,
+      "step": 3768
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.0316638960274735,
+      "learning_rate": 6.742055151753484e-06,
+      "loss": 0.4985,
+      "step": 3769
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0868533383021344,
+      "learning_rate": 6.740423065298556e-06,
+      "loss": 0.4759,
+      "step": 3770
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.612494627653258,
+      "learning_rate": 6.7387907678113295e-06,
+      "loss": 0.5839,
+      "step": 3771
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.7988783249416644,
+      "learning_rate": 6.73715825948973e-06,
+      "loss": 0.4696,
+      "step": 3772
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.464863378590616,
+      "learning_rate": 6.735525540531702e-06,
+      "loss": 0.5321,
+      "step": 3773
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.508336749179028,
+      "learning_rate": 6.733892611135217e-06,
+      "loss": 0.4476,
+      "step": 3774
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0799581052026346,
+      "learning_rate": 6.732259471498278e-06,
+      "loss": 0.5378,
+      "step": 3775
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.8558959220492133,
+      "learning_rate": 6.730626121818906e-06,
+      "loss": 0.4843,
+      "step": 3776
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.569279572558163,
+      "learning_rate": 6.728992562295149e-06,
+      "loss": 0.4927,
+      "step": 3777
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.064084600303104,
+      "learning_rate": 6.727358793125084e-06,
+      "loss": 0.5286,
+      "step": 3778
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.811312787937602,
+      "learning_rate": 6.72572481450681e-06,
+      "loss": 0.5338,
+      "step": 3779
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.31347504412208,
+      "learning_rate": 6.724090626638452e-06,
+      "loss": 0.5146,
+      "step": 3780
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.664774483482,
+      "learning_rate": 6.722456229718163e-06,
+      "loss": 0.52,
+      "step": 3781
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.8544752361497685,
+      "learning_rate": 6.720821623944117e-06,
+      "loss": 0.4594,
+      "step": 3782
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.05859481594685,
+      "learning_rate": 6.7191868095145165e-06,
+      "loss": 0.4812,
+      "step": 3783
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.170238789685284,
+      "learning_rate": 6.71755178662759e-06,
+      "loss": 0.4784,
+      "step": 3784
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.145247289694925,
+      "learning_rate": 6.715916555481585e-06,
+      "loss": 0.5022,
+      "step": 3785
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.751771523829993,
+      "learning_rate": 6.714281116274783e-06,
+      "loss": 0.4878,
+      "step": 3786
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.7563726048641697,
+      "learning_rate": 6.712645469205488e-06,
+      "loss": 0.5125,
+      "step": 3787
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.519927206864728,
+      "learning_rate": 6.711009614472023e-06,
+      "loss": 0.4988,
+      "step": 3788
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.242265971239576,
+      "learning_rate": 6.709373552272744e-06,
+      "loss": 0.5642,
+      "step": 3789
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.0863530948253945,
+      "learning_rate": 6.7077372828060294e-06,
+      "loss": 0.5004,
+      "step": 3790
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.8234495975043825,
+      "learning_rate": 6.70610080627028e-06,
+      "loss": 0.4931,
+      "step": 3791
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.891968141344607,
+      "learning_rate": 6.704464122863929e-06,
+      "loss": 0.5604,
+      "step": 3792
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.40822804646311595,
+      "learning_rate": 6.702827232785426e-06,
+      "loss": 0.3121,
+      "step": 3793
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.217487137663486,
+      "learning_rate": 6.701190136233248e-06,
+      "loss": 0.4708,
+      "step": 3794
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.412602460923627,
+      "learning_rate": 6.699552833405904e-06,
+      "loss": 0.4656,
+      "step": 3795
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.8219244125329643,
+      "learning_rate": 6.697915324501918e-06,
+      "loss": 0.4915,
+      "step": 3796
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.417136691358511,
+      "learning_rate": 6.696277609719846e-06,
+      "loss": 0.2981,
+      "step": 3797
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.41725014168886,
+      "learning_rate": 6.694639689258265e-06,
+      "loss": 0.4724,
+      "step": 3798
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.190576145504185,
+      "learning_rate": 6.6930015633157815e-06,
+      "loss": 0.537,
+      "step": 3799
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.35808312998213,
+      "learning_rate": 6.691363232091019e-06,
+      "loss": 0.5496,
+      "step": 3800
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.587676388341575,
+      "learning_rate": 6.6897246957826356e-06,
+      "loss": 0.4902,
+      "step": 3801
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.153306552821862,
+      "learning_rate": 6.688085954589307e-06,
+      "loss": 0.5222,
+      "step": 3802
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.38000794043182,
+      "learning_rate": 6.686447008709735e-06,
+      "loss": 0.5281,
+      "step": 3803
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4017875172261917,
+      "learning_rate": 6.684807858342652e-06,
+      "loss": 0.2885,
+      "step": 3804
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.70570453743161,
+      "learning_rate": 6.683168503686807e-06,
+      "loss": 0.4801,
+      "step": 3805
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.41728512222486225,
+      "learning_rate": 6.681528944940977e-06,
+      "loss": 0.2985,
+      "step": 3806
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 14.281749712785976,
+      "learning_rate": 6.6798891823039655e-06,
+      "loss": 0.5334,
+      "step": 3807
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4075772506274308,
+      "learning_rate": 6.6782492159746005e-06,
+      "loss": 0.311,
+      "step": 3808
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.686371307018165,
+      "learning_rate": 6.676609046151732e-06,
+      "loss": 0.4858,
+      "step": 3809
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.156495517643448,
+      "learning_rate": 6.674968673034235e-06,
+      "loss": 0.4998,
+      "step": 3810
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4121187521348877,
+      "learning_rate": 6.6733280968210125e-06,
+      "loss": 0.4756,
+      "step": 3811
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.595771425316766,
+      "learning_rate": 6.67168731771099e-06,
+      "loss": 0.4775,
+      "step": 3812
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.398936762358265,
+      "learning_rate": 6.670046335903116e-06,
+      "loss": 0.4882,
+      "step": 3813
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.748548788012498,
+      "learning_rate": 6.668405151596366e-06,
+      "loss": 0.5733,
+      "step": 3814
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.939326720994806,
+      "learning_rate": 6.6667637649897424e-06,
+      "loss": 0.5646,
+      "step": 3815
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.613742814004435,
+      "learning_rate": 6.6651221762822635e-06,
+      "loss": 0.4635,
+      "step": 3816
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.431693144046184,
+      "learning_rate": 6.6634803856729805e-06,
+      "loss": 0.4796,
+      "step": 3817
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.333204965111059,
+      "learning_rate": 6.661838393360967e-06,
+      "loss": 0.57,
+      "step": 3818
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 22.557978637603743,
+      "learning_rate": 6.660196199545318e-06,
+      "loss": 0.5219,
+      "step": 3819
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.445464932248461,
+      "learning_rate": 6.658553804425156e-06,
+      "loss": 0.5193,
+      "step": 3820
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.473307815508696,
+      "learning_rate": 6.656911208199628e-06,
+      "loss": 0.5351,
+      "step": 3821
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.45461522801027726,
+      "learning_rate": 6.655268411067904e-06,
+      "loss": 0.3001,
+      "step": 3822
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.365374190310279,
+      "learning_rate": 6.6536254132291765e-06,
+      "loss": 0.4896,
+      "step": 3823
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.836380598386166,
+      "learning_rate": 6.651982214882669e-06,
+      "loss": 0.4908,
+      "step": 3824
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.266020045280473,
+      "learning_rate": 6.650338816227623e-06,
+      "loss": 0.5108,
+      "step": 3825
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.5635743833003035,
+      "learning_rate": 6.6486952174633045e-06,
+      "loss": 0.467,
+      "step": 3826
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.92045496848478,
+      "learning_rate": 6.647051418789008e-06,
+      "loss": 0.4643,
+      "step": 3827
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4137834873516245,
+      "learning_rate": 6.645407420404048e-06,
+      "loss": 0.2862,
+      "step": 3828
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.247140458061164,
+      "learning_rate": 6.643763222507765e-06,
+      "loss": 0.5275,
+      "step": 3829
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.218464951338543,
+      "learning_rate": 6.642118825299526e-06,
+      "loss": 0.5078,
+      "step": 3830
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.298962879441392,
+      "learning_rate": 6.640474228978717e-06,
+      "loss": 0.5132,
+      "step": 3831
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4306881705549348,
+      "learning_rate": 6.638829433744753e-06,
+      "loss": 0.3133,
+      "step": 3832
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.53585701290351,
+      "learning_rate": 6.63718443979707e-06,
+      "loss": 0.4533,
+      "step": 3833
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.467471774380599,
+      "learning_rate": 6.635539247335129e-06,
+      "loss": 0.5161,
+      "step": 3834
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.300081796025761,
+      "learning_rate": 6.633893856558416e-06,
+      "loss": 0.5128,
+      "step": 3835
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.409789773192067,
+      "learning_rate": 6.63224826766644e-06,
+      "loss": 0.5201,
+      "step": 3836
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.724166022916394,
+      "learning_rate": 6.630602480858734e-06,
+      "loss": 0.5,
+      "step": 3837
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.096274944770546,
+      "learning_rate": 6.628956496334856e-06,
+      "loss": 0.5015,
+      "step": 3838
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.559050781706807,
+      "learning_rate": 6.627310314294386e-06,
+      "loss": 0.5556,
+      "step": 3839
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.164453304323496,
+      "learning_rate": 6.6256639349369326e-06,
+      "loss": 0.4534,
+      "step": 3840
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.449448185440188,
+      "learning_rate": 6.624017358462122e-06,
+      "loss": 0.5014,
+      "step": 3841
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.351472992035255,
+      "learning_rate": 6.622370585069605e-06,
+      "loss": 0.4924,
+      "step": 3842
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.44197633858335206,
+      "learning_rate": 6.6207236149590635e-06,
+      "loss": 0.2988,
+      "step": 3843
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.480119383405883,
+      "learning_rate": 6.619076448330197e-06,
+      "loss": 0.4831,
+      "step": 3844
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.074134119266324,
+      "learning_rate": 6.6174290853827275e-06,
+      "loss": 0.4533,
+      "step": 3845
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.245376139363979,
+      "learning_rate": 6.615781526316407e-06,
+      "loss": 0.5293,
+      "step": 3846
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4010740137712842,
+      "learning_rate": 6.614133771331006e-06,
+      "loss": 0.2808,
+      "step": 3847
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.492335150691726,
+      "learning_rate": 6.612485820626318e-06,
+      "loss": 0.5288,
+      "step": 3848
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.024222927571268,
+      "learning_rate": 6.610837674402167e-06,
+      "loss": 0.4892,
+      "step": 3849
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.964314068786558,
+      "learning_rate": 6.609189332858394e-06,
+      "loss": 0.4705,
+      "step": 3850
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1610137068470916,
+      "learning_rate": 6.607540796194867e-06,
+      "loss": 0.4469,
+      "step": 3851
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.736755151588019,
+      "learning_rate": 6.605892064611478e-06,
+      "loss": 0.5055,
+      "step": 3852
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.988879374031688,
+      "learning_rate": 6.604243138308138e-06,
+      "loss": 0.4883,
+      "step": 3853
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.141404388205631,
+      "learning_rate": 6.602594017484786e-06,
+      "loss": 0.4865,
+      "step": 3854
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.818265951261711,
+      "learning_rate": 6.600944702341385e-06,
+      "loss": 0.52,
+      "step": 3855
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.837953355710897,
+      "learning_rate": 6.599295193077921e-06,
+      "loss": 0.5077,
+      "step": 3856
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.43932977824010655,
+      "learning_rate": 6.5976454898944e-06,
+      "loss": 0.2942,
+      "step": 3857
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.370156769725326,
+      "learning_rate": 6.595995592990856e-06,
+      "loss": 0.5079,
+      "step": 3858
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.882487672903146,
+      "learning_rate": 6.594345502567343e-06,
+      "loss": 0.5276,
+      "step": 3859
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.364124624904627,
+      "learning_rate": 6.592695218823943e-06,
+      "loss": 0.5172,
+      "step": 3860
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.75099434528002,
+      "learning_rate": 6.591044741960755e-06,
+      "loss": 0.4952,
+      "step": 3861
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.9961256392515585,
+      "learning_rate": 6.589394072177908e-06,
+      "loss": 0.4569,
+      "step": 3862
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.574639257135775,
+      "learning_rate": 6.58774320967555e-06,
+      "loss": 0.4941,
+      "step": 3863
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.949566840633263,
+      "learning_rate": 6.586092154653855e-06,
+      "loss": 0.5623,
+      "step": 3864
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.441444913371467,
+      "learning_rate": 6.584440907313017e-06,
+      "loss": 0.5083,
+      "step": 3865
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.446405076680615,
+      "learning_rate": 6.582789467853257e-06,
+      "loss": 0.5096,
+      "step": 3866
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.529977478129513,
+      "learning_rate": 6.581137836474817e-06,
+      "loss": 0.5097,
+      "step": 3867
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.40367138012939,
+      "learning_rate": 6.579486013377963e-06,
+      "loss": 0.5317,
+      "step": 3868
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.842925842463216,
+      "learning_rate": 6.577833998762986e-06,
+      "loss": 0.4888,
+      "step": 3869
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.289295030328228,
+      "learning_rate": 6.576181792830194e-06,
+      "loss": 0.4585,
+      "step": 3870
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.90041634810382,
+      "learning_rate": 6.574529395779928e-06,
+      "loss": 0.5599,
+      "step": 3871
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.72350957516245,
+      "learning_rate": 6.572876807812545e-06,
+      "loss": 0.5605,
+      "step": 3872
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3553023678246494,
+      "learning_rate": 6.571224029128426e-06,
+      "loss": 0.5358,
+      "step": 3873
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.0666448697189,
+      "learning_rate": 6.569571059927977e-06,
+      "loss": 0.4834,
+      "step": 3874
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7764790666783914,
+      "learning_rate": 6.567917900411625e-06,
+      "loss": 0.5022,
+      "step": 3875
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.637897716292941,
+      "learning_rate": 6.566264550779821e-06,
+      "loss": 0.4955,
+      "step": 3876
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.566082078711867,
+      "learning_rate": 6.564611011233043e-06,
+      "loss": 0.4737,
+      "step": 3877
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.720761346100291,
+      "learning_rate": 6.562957281971786e-06,
+      "loss": 0.5038,
+      "step": 3878
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3140507786690456,
+      "learning_rate": 6.561303363196568e-06,
+      "loss": 0.473,
+      "step": 3879
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.4571367913078666,
+      "learning_rate": 6.559649255107936e-06,
+      "loss": 0.5487,
+      "step": 3880
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.979702406666144,
+      "learning_rate": 6.557994957906455e-06,
+      "loss": 0.4827,
+      "step": 3881
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.260068836413299,
+      "learning_rate": 6.556340471792713e-06,
+      "loss": 0.5038,
+      "step": 3882
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.659877407597438,
+      "learning_rate": 6.554685796967324e-06,
+      "loss": 0.5257,
+      "step": 3883
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4148194626640244,
+      "learning_rate": 6.553030933630922e-06,
+      "loss": 0.2957,
+      "step": 3884
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.982121193770625,
+      "learning_rate": 6.551375881984166e-06,
+      "loss": 0.5018,
+      "step": 3885
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.7807221313820305,
+      "learning_rate": 6.549720642227735e-06,
+      "loss": 0.5027,
+      "step": 3886
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.138601305869855,
+      "learning_rate": 6.548065214562333e-06,
+      "loss": 0.4967,
+      "step": 3887
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.420996734984906,
+      "learning_rate": 6.546409599188687e-06,
+      "loss": 0.4946,
+      "step": 3888
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.182439067026761,
+      "learning_rate": 6.5447537963075485e-06,
+      "loss": 0.522,
+      "step": 3889
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.3922884380582445,
+      "learning_rate": 6.543097806119682e-06,
+      "loss": 0.4908,
+      "step": 3890
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.409554519209648,
+      "learning_rate": 6.54144162882589e-06,
+      "loss": 0.4832,
+      "step": 3891
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.557159277302413,
+      "learning_rate": 6.539785264626985e-06,
+      "loss": 0.4996,
+      "step": 3892
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.8926508428163022,
+      "learning_rate": 6.538128713723808e-06,
+      "loss": 0.5139,
+      "step": 3893
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.9173125981055374,
+      "learning_rate": 6.536471976317223e-06,
+      "loss": 0.4681,
+      "step": 3894
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.1943452676883535,
+      "learning_rate": 6.534815052608114e-06,
+      "loss": 0.5012,
+      "step": 3895
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.813332273807043,
+      "learning_rate": 6.533157942797387e-06,
+      "loss": 0.5065,
+      "step": 3896
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.634766852158943,
+      "learning_rate": 6.531500647085974e-06,
+      "loss": 0.5002,
+      "step": 3897
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.098196631311209,
+      "learning_rate": 6.529843165674828e-06,
+      "loss": 0.4635,
+      "step": 3898
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.468543291343753,
+      "learning_rate": 6.528185498764925e-06,
+      "loss": 0.5138,
+      "step": 3899
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3729935212270603,
+      "learning_rate": 6.526527646557261e-06,
+      "loss": 0.5415,
+      "step": 3900
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.29881737202273,
+      "learning_rate": 6.5248696092528565e-06,
+      "loss": 0.4883,
+      "step": 3901
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.391153257260267,
+      "learning_rate": 6.523211387052756e-06,
+      "loss": 0.4757,
+      "step": 3902
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.625078463578658,
+      "learning_rate": 6.521552980158024e-06,
+      "loss": 0.4895,
+      "step": 3903
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.382625323798777,
+      "learning_rate": 6.519894388769745e-06,
+      "loss": 0.5523,
+      "step": 3904
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.437162263737679,
+      "learning_rate": 6.518235613089034e-06,
+      "loss": 0.3001,
+      "step": 3905
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8726044443910634,
+      "learning_rate": 6.516576653317019e-06,
+      "loss": 0.5288,
+      "step": 3906
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.55493345227098,
+      "learning_rate": 6.514917509654856e-06,
+      "loss": 0.5944,
+      "step": 3907
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.166954392048629,
+      "learning_rate": 6.513258182303724e-06,
+      "loss": 0.4623,
+      "step": 3908
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.350611390671484,
+      "learning_rate": 6.511598671464821e-06,
+      "loss": 0.5238,
+      "step": 3909
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.448777962428998,
+      "learning_rate": 6.509938977339366e-06,
+      "loss": 0.5216,
+      "step": 3910
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.59045328255715,
+      "learning_rate": 6.508279100128605e-06,
+      "loss": 0.4809,
+      "step": 3911
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.157277155391788,
+      "learning_rate": 6.5066190400338035e-06,
+      "loss": 0.442,
+      "step": 3912
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.411869777760511,
+      "learning_rate": 6.5049587972562485e-06,
+      "loss": 0.4896,
+      "step": 3913
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5917209816040536,
+      "learning_rate": 6.503298371997252e-06,
+      "loss": 0.4413,
+      "step": 3914
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.388264846120092,
+      "learning_rate": 6.501637764458145e-06,
+      "loss": 0.5352,
+      "step": 3915
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.904988790141371,
+      "learning_rate": 6.499976974840282e-06,
+      "loss": 0.4929,
+      "step": 3916
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7734071916782588,
+      "learning_rate": 6.4983160033450395e-06,
+      "loss": 0.4636,
+      "step": 3917
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.654436574336724,
+      "learning_rate": 6.496654850173816e-06,
+      "loss": 0.5163,
+      "step": 3918
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.353946578767231,
+      "learning_rate": 6.494993515528032e-06,
+      "loss": 0.4973,
+      "step": 3919
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.630583245769177,
+      "learning_rate": 6.493331999609132e-06,
+      "loss": 0.549,
+      "step": 3920
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.0775227971767825,
+      "learning_rate": 6.4916703026185755e-06,
+      "loss": 0.5302,
+      "step": 3921
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4331316125490863,
+      "learning_rate": 6.490008424757855e-06,
+      "loss": 0.3001,
+      "step": 3922
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4196140469985593,
+      "learning_rate": 6.488346366228475e-06,
+      "loss": 0.3006,
+      "step": 3923
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.188956799290901,
+      "learning_rate": 6.4866841272319665e-06,
+      "loss": 0.5178,
+      "step": 3924
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.991743570953788,
+      "learning_rate": 6.485021707969883e-06,
+      "loss": 0.5231,
+      "step": 3925
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.256977029241211,
+      "learning_rate": 6.483359108643798e-06,
+      "loss": 0.4746,
+      "step": 3926
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.962385131848286,
+      "learning_rate": 6.481696329455307e-06,
+      "loss": 0.5267,
+      "step": 3927
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.254024556251789,
+      "learning_rate": 6.480033370606027e-06,
+      "loss": 0.4934,
+      "step": 3928
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.371636748896005,
+      "learning_rate": 6.478370232297599e-06,
+      "loss": 0.4676,
+      "step": 3929
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5793042466643814,
+      "learning_rate": 6.476706914731683e-06,
+      "loss": 0.4624,
+      "step": 3930
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.316080802864057,
+      "learning_rate": 6.475043418109966e-06,
+      "loss": 0.5194,
+      "step": 3931
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.029154430699935,
+      "learning_rate": 6.473379742634145e-06,
+      "loss": 0.556,
+      "step": 3932
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.016461987824266,
+      "learning_rate": 6.471715888505952e-06,
+      "loss": 0.5298,
+      "step": 3933
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.89163235554973,
+      "learning_rate": 6.470051855927134e-06,
+      "loss": 0.5192,
+      "step": 3934
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.312834292831638,
+      "learning_rate": 6.46838764509946e-06,
+      "loss": 0.5227,
+      "step": 3935
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.634902777363971,
+      "learning_rate": 6.466723256224723e-06,
+      "loss": 0.5147,
+      "step": 3936
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.095648544789993,
+      "learning_rate": 6.465058689504734e-06,
+      "loss": 0.4575,
+      "step": 3937
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.918209377853925,
+      "learning_rate": 6.463393945141328e-06,
+      "loss": 0.5301,
+      "step": 3938
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.021628454439125,
+      "learning_rate": 6.4617290233363604e-06,
+      "loss": 0.4896,
+      "step": 3939
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.499463947333945,
+      "learning_rate": 6.46006392429171e-06,
+      "loss": 0.5173,
+      "step": 3940
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.678286722211551,
+      "learning_rate": 6.458398648209274e-06,
+      "loss": 0.3252,
+      "step": 3941
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5513014050071656,
+      "learning_rate": 6.456733195290976e-06,
+      "loss": 0.5082,
+      "step": 3942
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.20671560304617,
+      "learning_rate": 6.455067565738755e-06,
+      "loss": 0.5037,
+      "step": 3943
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.5177746075519,
+      "learning_rate": 6.453401759754574e-06,
+      "loss": 0.506,
+      "step": 3944
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.181722276244913,
+      "learning_rate": 6.451735777540421e-06,
+      "loss": 0.5075,
+      "step": 3945
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.124813857524007,
+      "learning_rate": 6.450069619298299e-06,
+      "loss": 0.5129,
+      "step": 3946
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.5863648846392495,
+      "learning_rate": 6.448403285230238e-06,
+      "loss": 0.5787,
+      "step": 3947
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.656100730581425,
+      "learning_rate": 6.446736775538284e-06,
+      "loss": 0.4963,
+      "step": 3948
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.6236285012044744,
+      "learning_rate": 6.4450700904245076e-06,
+      "loss": 0.5166,
+      "step": 3949
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.755136560374504,
+      "learning_rate": 6.443403230091002e-06,
+      "loss": 0.4769,
+      "step": 3950
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.369698462078864,
+      "learning_rate": 6.4417361947398785e-06,
+      "loss": 0.4787,
+      "step": 3951
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.186561443010901,
+      "learning_rate": 6.440068984573271e-06,
+      "loss": 0.5309,
+      "step": 3952
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.764574131318307,
+      "learning_rate": 6.438401599793333e-06,
+      "loss": 0.4938,
+      "step": 3953
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.746143385926865,
+      "learning_rate": 6.436734040602244e-06,
+      "loss": 0.5724,
+      "step": 3954
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.527560480636529,
+      "learning_rate": 6.435066307202198e-06,
+      "loss": 0.4967,
+      "step": 3955
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.662149603519029,
+      "learning_rate": 6.4333983997954155e-06,
+      "loss": 0.5081,
+      "step": 3956
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6351267742907921,
+      "learning_rate": 6.431730318584135e-06,
+      "loss": 0.305,
+      "step": 3957
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.130562655776222,
+      "learning_rate": 6.430062063770619e-06,
+      "loss": 0.5396,
+      "step": 3958
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4421057919065405,
+      "learning_rate": 6.428393635557146e-06,
+      "loss": 0.4873,
+      "step": 3959
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1653831996754125,
+      "learning_rate": 6.426725034146021e-06,
+      "loss": 0.446,
+      "step": 3960
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.9986750587534794,
+      "learning_rate": 6.425056259739566e-06,
+      "loss": 0.5395,
+      "step": 3961
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.8364962739437716,
+      "learning_rate": 6.4233873125401266e-06,
+      "loss": 0.4789,
+      "step": 3962
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.569971327190316,
+      "learning_rate": 6.4217181927500695e-06,
+      "loss": 0.4681,
+      "step": 3963
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4116262439273237,
+      "learning_rate": 6.4200489005717806e-06,
+      "loss": 0.4492,
+      "step": 3964
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 21.479490971001674,
+      "learning_rate": 6.418379436207664e-06,
+      "loss": 0.5019,
+      "step": 3965
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.309970448550176,
+      "learning_rate": 6.416709799860152e-06,
+      "loss": 0.4914,
+      "step": 3966
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.381392805207714,
+      "learning_rate": 6.4150399917316945e-06,
+      "loss": 0.5157,
+      "step": 3967
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4766968032907963,
+      "learning_rate": 6.41337001202476e-06,
+      "loss": 0.2842,
+      "step": 3968
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.461046954504248,
+      "learning_rate": 6.411699860941836e-06,
+      "loss": 0.3081,
+      "step": 3969
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.819353696781099,
+      "learning_rate": 6.410029538685438e-06,
+      "loss": 0.4957,
+      "step": 3970
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.1499025558662685,
+      "learning_rate": 6.408359045458099e-06,
+      "loss": 0.5199,
+      "step": 3971
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3859517736208127,
+      "learning_rate": 6.4066883814623674e-06,
+      "loss": 0.2873,
+      "step": 3972
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2740689427866494,
+      "learning_rate": 6.4050175469008225e-06,
+      "loss": 0.5448,
+      "step": 3973
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.072421527869849,
+      "learning_rate": 6.403346541976057e-06,
+      "loss": 0.492,
+      "step": 3974
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.00452775845017,
+      "learning_rate": 6.401675366890682e-06,
+      "loss": 0.483,
+      "step": 3975
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.037262813475974,
+      "learning_rate": 6.400004021847338e-06,
+      "loss": 0.507,
+      "step": 3976
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.57290087563283,
+      "learning_rate": 6.39833250704868e-06,
+      "loss": 0.4925,
+      "step": 3977
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3045293317037054,
+      "learning_rate": 6.396660822697384e-06,
+      "loss": 0.5045,
+      "step": 3978
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0969727197709433,
+      "learning_rate": 6.39498896899615e-06,
+      "loss": 0.4959,
+      "step": 3979
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.451209319712195,
+      "learning_rate": 6.393316946147692e-06,
+      "loss": 0.5011,
+      "step": 3980
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.467031836913279,
+      "learning_rate": 6.391644754354751e-06,
+      "loss": 0.471,
+      "step": 3981
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.909686144692193,
+      "learning_rate": 6.389972393820088e-06,
+      "loss": 0.5241,
+      "step": 3982
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.875470460393872,
+      "learning_rate": 6.388299864746476e-06,
+      "loss": 0.5008,
+      "step": 3983
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.304789301161786,
+      "learning_rate": 6.386627167336724e-06,
+      "loss": 0.5494,
+      "step": 3984
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.52460284015295,
+      "learning_rate": 6.384954301793647e-06,
+      "loss": 0.5115,
+      "step": 3985
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6255584644467755,
+      "learning_rate": 6.3832812683200825e-06,
+      "loss": 0.5844,
+      "step": 3986
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.54859869499836,
+      "learning_rate": 6.381608067118899e-06,
+      "loss": 0.502,
+      "step": 3987
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6305029394242397,
+      "learning_rate": 6.379934698392973e-06,
+      "loss": 0.3008,
+      "step": 3988
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.780383796435878,
+      "learning_rate": 6.378261162345207e-06,
+      "loss": 0.4824,
+      "step": 3989
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.44974045987717476,
+      "learning_rate": 6.3765874591785245e-06,
+      "loss": 0.2932,
+      "step": 3990
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5946520113929,
+      "learning_rate": 6.374913589095866e-06,
+      "loss": 0.5058,
+      "step": 3991
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.0095555452453935,
+      "learning_rate": 6.373239552300194e-06,
+      "loss": 0.4529,
+      "step": 3992
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.8129498916606535,
+      "learning_rate": 6.371565348994493e-06,
+      "loss": 0.5594,
+      "step": 3993
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2858548255414934,
+      "learning_rate": 6.3698909793817656e-06,
+      "loss": 0.5538,
+      "step": 3994
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.246583723834495,
+      "learning_rate": 6.3682164436650326e-06,
+      "loss": 0.4699,
+      "step": 3995
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2993141972259616,
+      "learning_rate": 6.366541742047341e-06,
+      "loss": 0.4931,
+      "step": 3996
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.289219766300349,
+      "learning_rate": 6.364866874731749e-06,
+      "loss": 0.5243,
+      "step": 3997
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.701748705155616,
+      "learning_rate": 6.363191841921345e-06,
+      "loss": 0.5371,
+      "step": 3998
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.091317581618426,
+      "learning_rate": 6.36151664381923e-06,
+      "loss": 0.5197,
+      "step": 3999
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.795839714173064,
+      "learning_rate": 6.359841280628529e-06,
+      "loss": 0.466,
+      "step": 4000
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.125833015729276,
+      "learning_rate": 6.358165752552383e-06,
+      "loss": 0.4833,
+      "step": 4001
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.75944013979236,
+      "learning_rate": 6.356490059793959e-06,
+      "loss": 0.4943,
+      "step": 4002
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.731619474770425,
+      "learning_rate": 6.354814202556437e-06,
+      "loss": 0.4914,
+      "step": 4003
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.728668351828812,
+      "learning_rate": 6.353138181043024e-06,
+      "loss": 0.5105,
+      "step": 4004
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.124133925003753,
+      "learning_rate": 6.351461995456941e-06,
+      "loss": 0.5128,
+      "step": 4005
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.9601887915579237,
+      "learning_rate": 6.349785646001431e-06,
+      "loss": 0.5149,
+      "step": 4006
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.478282572354148,
+      "learning_rate": 6.348109132879758e-06,
+      "loss": 0.5141,
+      "step": 4007
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.226981840830098,
+      "learning_rate": 6.346432456295206e-06,
+      "loss": 0.5306,
+      "step": 4008
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7277695774769386,
+      "learning_rate": 6.344755616451075e-06,
+      "loss": 0.4926,
+      "step": 4009
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.417359489272913,
+      "learning_rate": 6.343078613550691e-06,
+      "loss": 0.4913,
+      "step": 4010
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.281153495685797,
+      "learning_rate": 6.341401447797395e-06,
+      "loss": 0.5405,
+      "step": 4011
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.079190535635813,
+      "learning_rate": 6.339724119394548e-06,
+      "loss": 0.4968,
+      "step": 4012
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8128476162642946,
+      "learning_rate": 6.3380466285455336e-06,
+      "loss": 0.3366,
+      "step": 4013
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.879586140692613,
+      "learning_rate": 6.336368975453752e-06,
+      "loss": 0.5041,
+      "step": 4014
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.191525725802229,
+      "learning_rate": 6.3346911603226245e-06,
+      "loss": 0.5053,
+      "step": 4015
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 16.225756079363915,
+      "learning_rate": 6.333013183355594e-06,
+      "loss": 0.4873,
+      "step": 4016
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.621622868975093,
+      "learning_rate": 6.331335044756119e-06,
+      "loss": 0.5375,
+      "step": 4017
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.117576119955022,
+      "learning_rate": 6.329656744727679e-06,
+      "loss": 0.5235,
+      "step": 4018
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.71757931619707,
+      "learning_rate": 6.3279782834737755e-06,
+      "loss": 0.5249,
+      "step": 4019
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.9525301938574025,
+      "learning_rate": 6.326299661197926e-06,
+      "loss": 0.4689,
+      "step": 4020
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.249553387626549,
+      "learning_rate": 6.324620878103671e-06,
+      "loss": 0.5051,
+      "step": 4021
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.809543127483784,
+      "learning_rate": 6.322941934394568e-06,
+      "loss": 0.5594,
+      "step": 4022
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.267794839092637,
+      "learning_rate": 6.321262830274193e-06,
+      "loss": 0.4606,
+      "step": 4023
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.009481767840635,
+      "learning_rate": 6.319583565946147e-06,
+      "loss": 0.4771,
+      "step": 4024
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.730577752249115,
+      "learning_rate": 6.317904141614043e-06,
+      "loss": 0.5013,
+      "step": 4025
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.9724687183340945,
+      "learning_rate": 6.3162245574815186e-06,
+      "loss": 0.5421,
+      "step": 4026
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.145868632139356,
+      "learning_rate": 6.314544813752229e-06,
+      "loss": 0.4528,
+      "step": 4027
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.039371625572302,
+      "learning_rate": 6.312864910629848e-06,
+      "loss": 0.5185,
+      "step": 4028
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.473730032804855,
+      "learning_rate": 6.311184848318072e-06,
+      "loss": 0.482,
+      "step": 4029
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.517436855902935,
+      "learning_rate": 6.309504627020612e-06,
+      "loss": 0.3118,
+      "step": 4030
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.250118308818624,
+      "learning_rate": 6.3078242469412e-06,
+      "loss": 0.5037,
+      "step": 4031
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.688955733029054,
+      "learning_rate": 6.306143708283592e-06,
+      "loss": 0.487,
+      "step": 4032
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.005487913382792,
+      "learning_rate": 6.3044630112515545e-06,
+      "loss": 0.5308,
+      "step": 4033
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.004680010873418,
+      "learning_rate": 6.30278215604888e-06,
+      "loss": 0.4883,
+      "step": 4034
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7651599568347702,
+      "learning_rate": 6.3011011428793776e-06,
+      "loss": 0.5273,
+      "step": 4035
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.299279782433623,
+      "learning_rate": 6.299419971946876e-06,
+      "loss": 0.4989,
+      "step": 4036
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.395678187340109,
+      "learning_rate": 6.297738643455225e-06,
+      "loss": 0.4747,
+      "step": 4037
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4204566021831678,
+      "learning_rate": 6.2960571576082875e-06,
+      "loss": 0.3041,
+      "step": 4038
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.74034162989668,
+      "learning_rate": 6.294375514609952e-06,
+      "loss": 0.471,
+      "step": 4039
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.870071446020838,
+      "learning_rate": 6.292693714664122e-06,
+      "loss": 0.4621,
+      "step": 4040
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.213233990831375,
+      "learning_rate": 6.291011757974723e-06,
+      "loss": 0.5325,
+      "step": 4041
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.914345294133987,
+      "learning_rate": 6.289329644745699e-06,
+      "loss": 0.5266,
+      "step": 4042
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.59024837540751,
+      "learning_rate": 6.287647375181009e-06,
+      "loss": 0.4605,
+      "step": 4043
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0963069057617254,
+      "learning_rate": 6.285964949484636e-06,
+      "loss": 0.5013,
+      "step": 4044
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6627552998144295,
+      "learning_rate": 6.284282367860579e-06,
+      "loss": 0.5185,
+      "step": 4045
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 67.21636787483072,
+      "learning_rate": 6.282599630512858e-06,
+      "loss": 0.5314,
+      "step": 4046
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 19.2179038334971,
+      "learning_rate": 6.280916737645511e-06,
+      "loss": 0.5092,
+      "step": 4047
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.2798650171715105,
+      "learning_rate": 6.2792336894625915e-06,
+      "loss": 0.5142,
+      "step": 4048
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1538945997000676,
+      "learning_rate": 6.2775504861681775e-06,
+      "loss": 0.4821,
+      "step": 4049
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7296826183999245,
+      "learning_rate": 6.275867127966364e-06,
+      "loss": 0.5413,
+      "step": 4050
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.582874519820803,
+      "learning_rate": 6.27418361506126e-06,
+      "loss": 0.4863,
+      "step": 4051
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.288773298670132,
+      "learning_rate": 6.272499947657002e-06,
+      "loss": 0.5244,
+      "step": 4052
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4393063292044724,
+      "learning_rate": 6.270816125957739e-06,
+      "loss": 0.3111,
+      "step": 4053
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3903166164493045,
+      "learning_rate": 6.269132150167639e-06,
+      "loss": 0.3004,
+      "step": 4054
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.149358612254402,
+      "learning_rate": 6.267448020490889e-06,
+      "loss": 0.4768,
+      "step": 4055
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.235550972881352,
+      "learning_rate": 6.265763737131698e-06,
+      "loss": 0.5089,
+      "step": 4056
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.056083757750029,
+      "learning_rate": 6.26407930029429e-06,
+      "loss": 0.5082,
+      "step": 4057
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3950622169539966,
+      "learning_rate": 6.26239471018291e-06,
+      "loss": 0.2977,
+      "step": 4058
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.68475453252424,
+      "learning_rate": 6.260709967001816e-06,
+      "loss": 0.497,
+      "step": 4059
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4033295543455543,
+      "learning_rate": 6.259025070955296e-06,
+      "loss": 0.2991,
+      "step": 4060
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.369380117607907,
+      "learning_rate": 6.2573400222476435e-06,
+      "loss": 0.4732,
+      "step": 4061
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.642684459335446,
+      "learning_rate": 6.255654821083178e-06,
+      "loss": 0.4767,
+      "step": 4062
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.559627096875906,
+      "learning_rate": 6.253969467666239e-06,
+      "loss": 0.5176,
+      "step": 4063
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.361208332908745,
+      "learning_rate": 6.252283962201177e-06,
+      "loss": 0.5256,
+      "step": 4064
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4094336645503541,
+      "learning_rate": 6.2505983048923675e-06,
+      "loss": 0.3012,
+      "step": 4065
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9457717492223954,
+      "learning_rate": 6.2489124959442026e-06,
+      "loss": 0.4716,
+      "step": 4066
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.647943174097425,
+      "learning_rate": 6.247226535561092e-06,
+      "loss": 0.4727,
+      "step": 4067
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.904634177709741,
+      "learning_rate": 6.245540423947463e-06,
+      "loss": 0.5233,
+      "step": 4068
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.723577879766844,
+      "learning_rate": 6.243854161307765e-06,
+      "loss": 0.5102,
+      "step": 4069
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.387181504400332,
+      "learning_rate": 6.242167747846461e-06,
+      "loss": 0.2973,
+      "step": 4070
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.400873837203459,
+      "learning_rate": 6.240481183768036e-06,
+      "loss": 0.5258,
+      "step": 4071
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.046857530652091,
+      "learning_rate": 6.2387944692769896e-06,
+      "loss": 0.4994,
+      "step": 4072
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3858900453482482,
+      "learning_rate": 6.237107604577843e-06,
+      "loss": 0.2997,
+      "step": 4073
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 36.24879809759056,
+      "learning_rate": 6.235420589875136e-06,
+      "loss": 0.4291,
+      "step": 4074
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1851125357943455,
+      "learning_rate": 6.233733425373423e-06,
+      "loss": 0.4725,
+      "step": 4075
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.851023985108144,
+      "learning_rate": 6.232046111277277e-06,
+      "loss": 0.4934,
+      "step": 4076
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6422843651293344,
+      "learning_rate": 6.230358647791293e-06,
+      "loss": 0.4913,
+      "step": 4077
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.682377653324191,
+      "learning_rate": 6.228671035120083e-06,
+      "loss": 0.5188,
+      "step": 4078
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.008178101580401,
+      "learning_rate": 6.226983273468273e-06,
+      "loss": 0.5749,
+      "step": 4079
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.171921865101341,
+      "learning_rate": 6.225295363040511e-06,
+      "loss": 0.494,
+      "step": 4080
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2804711349700293,
+      "learning_rate": 6.2236073040414625e-06,
+      "loss": 0.5707,
+      "step": 4081
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.2366692332069,
+      "learning_rate": 6.221919096675808e-06,
+      "loss": 0.533,
+      "step": 4082
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8394341845159965,
+      "learning_rate": 6.220230741148253e-06,
+      "loss": 0.4585,
+      "step": 4083
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8669011139124856,
+      "learning_rate": 6.218542237663513e-06,
+      "loss": 0.5107,
+      "step": 4084
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.334272579211447,
+      "learning_rate": 6.216853586426326e-06,
+      "loss": 0.5122,
+      "step": 4085
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.499582401291445,
+      "learning_rate": 6.215164787641446e-06,
+      "loss": 0.4832,
+      "step": 4086
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.603873139883521,
+      "learning_rate": 6.2134758415136465e-06,
+      "loss": 0.4894,
+      "step": 4087
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.112422604356654,
+      "learning_rate": 6.211786748247717e-06,
+      "loss": 0.476,
+      "step": 4088
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.218006495712872,
+      "learning_rate": 6.210097508048467e-06,
+      "loss": 0.5373,
+      "step": 4089
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.34632552546709,
+      "learning_rate": 6.208408121120722e-06,
+      "loss": 0.5444,
+      "step": 4090
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.833270763183586,
+      "learning_rate": 6.206718587669327e-06,
+      "loss": 0.4375,
+      "step": 4091
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.241697452616535,
+      "learning_rate": 6.205028907899143e-06,
+      "loss": 0.4952,
+      "step": 4092
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.386150180716263,
+      "learning_rate": 6.203339082015049e-06,
+      "loss": 0.5627,
+      "step": 4093
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4343227171305837,
+      "learning_rate": 6.2016491102219435e-06,
+      "loss": 0.5315,
+      "step": 4094
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4062132543790287,
+      "learning_rate": 6.19995899272474e-06,
+      "loss": 0.3033,
+      "step": 4095
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.868456300992819,
+      "learning_rate": 6.198268729728372e-06,
+      "loss": 0.444,
+      "step": 4096
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.332810724880862,
+      "learning_rate": 6.1965783214377895e-06,
+      "loss": 0.5326,
+      "step": 4097
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.894277377182052,
+      "learning_rate": 6.19488776805796e-06,
+      "loss": 0.4644,
+      "step": 4098
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.297580921946911,
+      "learning_rate": 6.1931970697938695e-06,
+      "loss": 0.4232,
+      "step": 4099
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7869649951007984,
+      "learning_rate": 6.191506226850521e-06,
+      "loss": 0.5212,
+      "step": 4100
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.04049085232926,
+      "learning_rate": 6.189815239432935e-06,
+      "loss": 0.5041,
+      "step": 4101
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 17.152072356246915,
+      "learning_rate": 6.188124107746148e-06,
+      "loss": 0.4632,
+      "step": 4102
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.617447761934923,
+      "learning_rate": 6.186432831995218e-06,
+      "loss": 0.4744,
+      "step": 4103
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.025877548347526,
+      "learning_rate": 6.184741412385217e-06,
+      "loss": 0.4758,
+      "step": 4104
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.557701686318651,
+      "learning_rate": 6.183049849121233e-06,
+      "loss": 0.529,
+      "step": 4105
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.571707795939485,
+      "learning_rate": 6.18135814240838e-06,
+      "loss": 0.5216,
+      "step": 4106
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.846031308507201,
+      "learning_rate": 6.179666292451776e-06,
+      "loss": 0.5242,
+      "step": 4107
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.475801295981737,
+      "learning_rate": 6.1779742994565686e-06,
+      "loss": 0.5326,
+      "step": 4108
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.42895038025355237,
+      "learning_rate": 6.176282163627917e-06,
+      "loss": 0.2992,
+      "step": 4109
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.563074156694209,
+      "learning_rate": 6.174589885170996e-06,
+      "loss": 0.4736,
+      "step": 4110
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.5614162915118746,
+      "learning_rate": 6.172897464291005e-06,
+      "loss": 0.5144,
+      "step": 4111
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.925176171338453,
+      "learning_rate": 6.171204901193151e-06,
+      "loss": 0.5114,
+      "step": 4112
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.44713235240107324,
+      "learning_rate": 6.169512196082663e-06,
+      "loss": 0.321,
+      "step": 4113
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0077078135698763,
+      "learning_rate": 6.167819349164792e-06,
+      "loss": 0.5407,
+      "step": 4114
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.355206822096643,
+      "learning_rate": 6.166126360644797e-06,
+      "loss": 0.4872,
+      "step": 4115
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.063042933685018,
+      "learning_rate": 6.1644332307279616e-06,
+      "loss": 0.472,
+      "step": 4116
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.745136099156893,
+      "learning_rate": 6.1627399596195835e-06,
+      "loss": 0.5179,
+      "step": 4117
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6977679817183335,
+      "learning_rate": 6.161046547524976e-06,
+      "loss": 0.5183,
+      "step": 4118
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5311859143058766,
+      "learning_rate": 6.15935299464947e-06,
+      "loss": 0.4804,
+      "step": 4119
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7470613672187016,
+      "learning_rate": 6.157659301198418e-06,
+      "loss": 0.5625,
+      "step": 4120
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9260516435023782,
+      "learning_rate": 6.1559654673771865e-06,
+      "loss": 0.4857,
+      "step": 4121
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.702098918420559,
+      "learning_rate": 6.154271493391155e-06,
+      "loss": 0.512,
+      "step": 4122
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.282292880023393,
+      "learning_rate": 6.152577379445725e-06,
+      "loss": 0.4846,
+      "step": 4123
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.865024158677332,
+      "learning_rate": 6.150883125746314e-06,
+      "loss": 0.5359,
+      "step": 4124
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.581720193751394,
+      "learning_rate": 6.149188732498358e-06,
+      "loss": 0.5259,
+      "step": 4125
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.690887111239698,
+      "learning_rate": 6.147494199907306e-06,
+      "loss": 0.4872,
+      "step": 4126
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.933635237451068,
+      "learning_rate": 6.145799528178625e-06,
+      "loss": 0.4894,
+      "step": 4127
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7239155860328514,
+      "learning_rate": 6.1441047175178025e-06,
+      "loss": 0.4741,
+      "step": 4128
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.45852145649044584,
+      "learning_rate": 6.142409768130339e-06,
+      "loss": 0.2964,
+      "step": 4129
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.43676343079053,
+      "learning_rate": 6.14071468022175e-06,
+      "loss": 0.5148,
+      "step": 4130
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.28047741576359,
+      "learning_rate": 6.1390194539975754e-06,
+      "loss": 0.4892,
+      "step": 4131
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.396916418185684,
+      "learning_rate": 6.137324089663365e-06,
+      "loss": 0.5442,
+      "step": 4132
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1216276930133153,
+      "learning_rate": 6.135628587424688e-06,
+      "loss": 0.5051,
+      "step": 4133
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.310735886368675,
+      "learning_rate": 6.133932947487129e-06,
+      "loss": 0.5534,
+      "step": 4134
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.003177788006538,
+      "learning_rate": 6.132237170056291e-06,
+      "loss": 0.5103,
+      "step": 4135
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8083787039480628,
+      "learning_rate": 6.130541255337792e-06,
+      "loss": 0.4463,
+      "step": 4136
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9168718927287003,
+      "learning_rate": 6.128845203537269e-06,
+      "loss": 0.5447,
+      "step": 4137
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.587063070777114,
+      "learning_rate": 6.127149014860374e-06,
+      "loss": 0.5226,
+      "step": 4138
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.445063885415015,
+      "learning_rate": 6.125452689512774e-06,
+      "loss": 0.441,
+      "step": 4139
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.717434626257729,
+      "learning_rate": 6.123756227700156e-06,
+      "loss": 0.5224,
+      "step": 4140
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.40814061764066545,
+      "learning_rate": 6.122059629628221e-06,
+      "loss": 0.3067,
+      "step": 4141
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1749958134608836,
+      "learning_rate": 6.1203628955026874e-06,
+      "loss": 0.5188,
+      "step": 4142
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.300355371069774,
+      "learning_rate": 6.118666025529292e-06,
+      "loss": 0.4937,
+      "step": 4143
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7505804765966713,
+      "learning_rate": 6.116969019913781e-06,
+      "loss": 0.4839,
+      "step": 4144
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0483021224296527,
+      "learning_rate": 6.115271878861928e-06,
+      "loss": 0.4862,
+      "step": 4145
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1477912971627373,
+      "learning_rate": 6.113574602579515e-06,
+      "loss": 0.5224,
+      "step": 4146
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.290965818384541,
+      "learning_rate": 6.1118771912723396e-06,
+      "loss": 0.497,
+      "step": 4147
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.137339079130657,
+      "learning_rate": 6.110179645146225e-06,
+      "loss": 0.5365,
+      "step": 4148
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.037304760065889,
+      "learning_rate": 6.1084819644070004e-06,
+      "loss": 0.4723,
+      "step": 4149
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.581166821356093,
+      "learning_rate": 6.106784149260514e-06,
+      "loss": 0.5222,
+      "step": 4150
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.165533026243112,
+      "learning_rate": 6.1050861999126355e-06,
+      "loss": 0.5192,
+      "step": 4151
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.538772296828655,
+      "learning_rate": 6.103388116569245e-06,
+      "loss": 0.4751,
+      "step": 4152
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4687895501516586,
+      "learning_rate": 6.101689899436241e-06,
+      "loss": 0.5191,
+      "step": 4153
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.468774019273712,
+      "learning_rate": 6.0999915487195395e-06,
+      "loss": 0.4869,
+      "step": 4154
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5651371449785962,
+      "learning_rate": 6.09829306462507e-06,
+      "loss": 0.5493,
+      "step": 4155
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.752268652916431,
+      "learning_rate": 6.096594447358779e-06,
+      "loss": 0.4842,
+      "step": 4156
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.576422989019938,
+      "learning_rate": 6.094895697126631e-06,
+      "loss": 0.5132,
+      "step": 4157
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2741439056064223,
+      "learning_rate": 6.0931968141346054e-06,
+      "loss": 0.4998,
+      "step": 4158
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9224894997159034,
+      "learning_rate": 6.091497798588699e-06,
+      "loss": 0.534,
+      "step": 4159
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2810835626709753,
+      "learning_rate": 6.08979865069492e-06,
+      "loss": 0.4431,
+      "step": 4160
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4208623552654601,
+      "learning_rate": 6.088099370659297e-06,
+      "loss": 0.2737,
+      "step": 4161
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.6534389007186565,
+      "learning_rate": 6.086399958687875e-06,
+      "loss": 0.5556,
+      "step": 4162
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.576801465063578,
+      "learning_rate": 6.0847004149867125e-06,
+      "loss": 0.5369,
+      "step": 4163
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.39466704368727623,
+      "learning_rate": 6.083000739761885e-06,
+      "loss": 0.2929,
+      "step": 4164
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.1797277793455825,
+      "learning_rate": 6.081300933219485e-06,
+      "loss": 0.531,
+      "step": 4165
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.30196499645032,
+      "learning_rate": 6.079600995565618e-06,
+      "loss": 0.5151,
+      "step": 4166
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.40776891098362006,
+      "learning_rate": 6.077900927006408e-06,
+      "loss": 0.3077,
+      "step": 4167
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.214851857660111,
+      "learning_rate": 6.0762007277479955e-06,
+      "loss": 0.5718,
+      "step": 4168
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.522586224744275,
+      "learning_rate": 6.074500397996533e-06,
+      "loss": 0.4951,
+      "step": 4169
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.846689129363745,
+      "learning_rate": 6.072799937958195e-06,
+      "loss": 0.5391,
+      "step": 4170
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.345531942267421,
+      "learning_rate": 6.0710993478391645e-06,
+      "loss": 0.5159,
+      "step": 4171
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7559216401740056,
+      "learning_rate": 6.0693986278456456e-06,
+      "loss": 0.4938,
+      "step": 4172
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.411288243521473,
+      "learning_rate": 6.067697778183857e-06,
+      "loss": 0.5526,
+      "step": 4173
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7291983815860306,
+      "learning_rate": 6.065996799060031e-06,
+      "loss": 0.5068,
+      "step": 4174
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.87401819069274,
+      "learning_rate": 6.064295690680418e-06,
+      "loss": 0.5071,
+      "step": 4175
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.924142422309146,
+      "learning_rate": 6.0625944532512845e-06,
+      "loss": 0.5107,
+      "step": 4176
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7204674432914864,
+      "learning_rate": 6.06089308697891e-06,
+      "loss": 0.4737,
+      "step": 4177
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.497186124837891,
+      "learning_rate": 6.05919159206959e-06,
+      "loss": 0.4923,
+      "step": 4178
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8532883705550858,
+      "learning_rate": 6.0574899687296385e-06,
+      "loss": 0.5539,
+      "step": 4179
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.312647819695276,
+      "learning_rate": 6.055788217165384e-06,
+      "loss": 0.4936,
+      "step": 4180
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.313848757074314,
+      "learning_rate": 6.054086337583166e-06,
+      "loss": 0.5009,
+      "step": 4181
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.788094578834226,
+      "learning_rate": 6.052384330189347e-06,
+      "loss": 0.5127,
+      "step": 4182
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.679294513585138,
+      "learning_rate": 6.0506821951903e-06,
+      "loss": 0.4923,
+      "step": 4183
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.47571327943759667,
+      "learning_rate": 6.0489799327924125e-06,
+      "loss": 0.3152,
+      "step": 4184
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4563571978244456,
+      "learning_rate": 6.0472775432020945e-06,
+      "loss": 0.3012,
+      "step": 4185
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 21.80912673778816,
+      "learning_rate": 6.0455750266257625e-06,
+      "loss": 0.5235,
+      "step": 4186
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.9611971608367105,
+      "learning_rate": 6.043872383269855e-06,
+      "loss": 0.5142,
+      "step": 4187
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.944605368139933,
+      "learning_rate": 6.04216961334082e-06,
+      "loss": 0.468,
+      "step": 4188
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.807796356626068,
+      "learning_rate": 6.040466717045127e-06,
+      "loss": 0.4777,
+      "step": 4189
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.363604226397761,
+      "learning_rate": 6.038763694589259e-06,
+      "loss": 0.5104,
+      "step": 4190
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.7402468073262956,
+      "learning_rate": 6.0370605461797124e-06,
+      "loss": 0.5248,
+      "step": 4191
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.870642921416966,
+      "learning_rate": 6.035357272022997e-06,
+      "loss": 0.4803,
+      "step": 4192
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.4386944753341675,
+      "learning_rate": 6.033653872325644e-06,
+      "loss": 0.4887,
+      "step": 4193
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.1298150935509135,
+      "learning_rate": 6.031950347294196e-06,
+      "loss": 0.5169,
+      "step": 4194
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.3200111033332504,
+      "learning_rate": 6.030246697135209e-06,
+      "loss": 0.5151,
+      "step": 4195
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.9576764218819953,
+      "learning_rate": 6.028542922055261e-06,
+      "loss": 0.488,
+      "step": 4196
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.43369898755686653,
+      "learning_rate": 6.026839022260936e-06,
+      "loss": 0.3067,
+      "step": 4197
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.620775404671606,
+      "learning_rate": 6.0251349979588395e-06,
+      "loss": 0.4734,
+      "step": 4198
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 16.553969439419234,
+      "learning_rate": 6.0234308493555906e-06,
+      "loss": 0.4671,
+      "step": 4199
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.352697218754615,
+      "learning_rate": 6.021726576657823e-06,
+      "loss": 0.4782,
+      "step": 4200
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.38916911369406154,
+      "learning_rate": 6.0200221800721845e-06,
+      "loss": 0.2995,
+      "step": 4201
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1410862266307276,
+      "learning_rate": 6.018317659805341e-06,
+      "loss": 0.5233,
+      "step": 4202
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.40864158422256,
+      "learning_rate": 6.01661301606397e-06,
+      "loss": 0.528,
+      "step": 4203
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.305136308327676,
+      "learning_rate": 6.014908249054767e-06,
+      "loss": 0.4882,
+      "step": 4204
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.43280758947919,
+      "learning_rate": 6.0132033589844395e-06,
+      "loss": 0.4857,
+      "step": 4205
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.235612198771758,
+      "learning_rate": 6.011498346059712e-06,
+      "loss": 0.5351,
+      "step": 4206
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.595264987350088,
+      "learning_rate": 6.009793210487323e-06,
+      "loss": 0.5271,
+      "step": 4207
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.792490478296635,
+      "learning_rate": 6.008087952474025e-06,
+      "loss": 0.5473,
+      "step": 4208
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.9684840611355066,
+      "learning_rate": 6.006382572226587e-06,
+      "loss": 0.4928,
+      "step": 4209
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.987073472244828,
+      "learning_rate": 6.004677069951793e-06,
+      "loss": 0.4863,
+      "step": 4210
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9515040004291344,
+      "learning_rate": 6.002971445856441e-06,
+      "loss": 0.5184,
+      "step": 4211
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.055564566996582,
+      "learning_rate": 6.001265700147344e-06,
+      "loss": 0.4946,
+      "step": 4212
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.522906462493639,
+      "learning_rate": 5.999559833031328e-06,
+      "loss": 0.4714,
+      "step": 4213
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 32.95647788037811,
+      "learning_rate": 5.997853844715237e-06,
+      "loss": 0.5211,
+      "step": 4214
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.9272855897462575,
+      "learning_rate": 5.996147735405926e-06,
+      "loss": 0.536,
+      "step": 4215
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.9119900873864286,
+      "learning_rate": 5.994441505310269e-06,
+      "loss": 0.4807,
+      "step": 4216
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.0775504779051905,
+      "learning_rate": 5.992735154635151e-06,
+      "loss": 0.4981,
+      "step": 4217
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.258390795398319,
+      "learning_rate": 5.991028683587472e-06,
+      "loss": 0.4661,
+      "step": 4218
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.249035396765323,
+      "learning_rate": 5.98932209237415e-06,
+      "loss": 0.5417,
+      "step": 4219
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.040469335740253,
+      "learning_rate": 5.987615381202112e-06,
+      "loss": 0.4959,
+      "step": 4220
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.175361550884258,
+      "learning_rate": 5.985908550278306e-06,
+      "loss": 0.4327,
+      "step": 4221
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.152107361535952,
+      "learning_rate": 5.98420159980969e-06,
+      "loss": 0.5271,
+      "step": 4222
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4225036446480916,
+      "learning_rate": 5.982494530003233e-06,
+      "loss": 0.4597,
+      "step": 4223
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.654583003689705,
+      "learning_rate": 5.98078734106593e-06,
+      "loss": 0.5145,
+      "step": 4224
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8061151322048983,
+      "learning_rate": 5.97908003320478e-06,
+      "loss": 0.491,
+      "step": 4225
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.123428848457462,
+      "learning_rate": 5.9773726066268e-06,
+      "loss": 0.5206,
+      "step": 4226
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.510126762438169,
+      "learning_rate": 5.9756650615390226e-06,
+      "loss": 0.4968,
+      "step": 4227
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4219107657768812,
+      "learning_rate": 5.973957398148493e-06,
+      "loss": 0.2969,
+      "step": 4228
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.1230815511876635,
+      "learning_rate": 5.972249616662271e-06,
+      "loss": 0.5266,
+      "step": 4229
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.100850041693068,
+      "learning_rate": 5.970541717287432e-06,
+      "loss": 0.4997,
+      "step": 4230
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.4264453875346765,
+      "learning_rate": 5.968833700231062e-06,
+      "loss": 0.5187,
+      "step": 4231
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8595653497210933,
+      "learning_rate": 5.967125565700266e-06,
+      "loss": 0.492,
+      "step": 4232
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.033408828979947,
+      "learning_rate": 5.965417313902162e-06,
+      "loss": 0.5424,
+      "step": 4233
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.225167216471399,
+      "learning_rate": 5.9637089450438805e-06,
+      "loss": 0.5254,
+      "step": 4234
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.993813054811557,
+      "learning_rate": 5.962000459332566e-06,
+      "loss": 0.5023,
+      "step": 4235
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.5219196308787515,
+      "learning_rate": 5.96029185697538e-06,
+      "loss": 0.5069,
+      "step": 4236
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.40469275282385797,
+      "learning_rate": 5.958583138179494e-06,
+      "loss": 0.3054,
+      "step": 4237
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.870502721042856,
+      "learning_rate": 5.9568743031521e-06,
+      "loss": 0.4854,
+      "step": 4238
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.694793724205383,
+      "learning_rate": 5.955165352100399e-06,
+      "loss": 0.4822,
+      "step": 4239
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.503484204235054,
+      "learning_rate": 5.953456285231602e-06,
+      "loss": 0.4884,
+      "step": 4240
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.4485658146486164,
+      "learning_rate": 5.951747102752946e-06,
+      "loss": 0.4793,
+      "step": 4241
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.363142467032272,
+      "learning_rate": 5.950037804871673e-06,
+      "loss": 0.4792,
+      "step": 4242
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.435930273435102,
+      "learning_rate": 5.948328391795038e-06,
+      "loss": 0.536,
+      "step": 4243
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.694679455937676,
+      "learning_rate": 5.9466188637303206e-06,
+      "loss": 0.5578,
+      "step": 4244
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.947233489602711,
+      "learning_rate": 5.9449092208848e-06,
+      "loss": 0.4926,
+      "step": 4245
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4189040650788347,
+      "learning_rate": 5.943199463465779e-06,
+      "loss": 0.5158,
+      "step": 4246
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.636468015080374,
+      "learning_rate": 5.9414895916805705e-06,
+      "loss": 0.4839,
+      "step": 4247
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.066189420698882,
+      "learning_rate": 5.939779605736504e-06,
+      "loss": 0.4742,
+      "step": 4248
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.721043706179373,
+      "learning_rate": 5.93806950584092e-06,
+      "loss": 0.4714,
+      "step": 4249
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.89776683061868,
+      "learning_rate": 5.936359292201175e-06,
+      "loss": 0.4989,
+      "step": 4250
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.017036618792244,
+      "learning_rate": 5.934648965024636e-06,
+      "loss": 0.4557,
+      "step": 4251
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9568735235184223,
+      "learning_rate": 5.93293852451869e-06,
+      "loss": 0.5109,
+      "step": 4252
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.5463677316390765,
+      "learning_rate": 5.931227970890732e-06,
+      "loss": 0.5195,
+      "step": 4253
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 22.935119403943176,
+      "learning_rate": 5.92951730434817e-06,
+      "loss": 0.4888,
+      "step": 4254
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.2112020651430075,
+      "learning_rate": 5.927806525098432e-06,
+      "loss": 0.478,
+      "step": 4255
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.108359591021292,
+      "learning_rate": 5.926095633348953e-06,
+      "loss": 0.4993,
+      "step": 4256
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.312615629267165,
+      "learning_rate": 5.9243846293071846e-06,
+      "loss": 0.5158,
+      "step": 4257
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.74300163971227,
+      "learning_rate": 5.922673513180596e-06,
+      "loss": 0.5421,
+      "step": 4258
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.910094024283261,
+      "learning_rate": 5.9209622851766615e-06,
+      "loss": 0.4903,
+      "step": 4259
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.4118304431244955,
+      "learning_rate": 5.919250945502875e-06,
+      "loss": 0.4609,
+      "step": 4260
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.453964905748554,
+      "learning_rate": 5.9175394943667415e-06,
+      "loss": 0.4871,
+      "step": 4261
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5394298607504444,
+      "learning_rate": 5.915827931975782e-06,
+      "loss": 0.4688,
+      "step": 4262
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4286238157140144,
+      "learning_rate": 5.914116258537528e-06,
+      "loss": 0.3079,
+      "step": 4263
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.40726820649448797,
+      "learning_rate": 5.912404474259525e-06,
+      "loss": 0.3009,
+      "step": 4264
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.640111359487022,
+      "learning_rate": 5.910692579349336e-06,
+      "loss": 0.466,
+      "step": 4265
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.985548852089222,
+      "learning_rate": 5.908980574014532e-06,
+      "loss": 0.494,
+      "step": 4266
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8930879273499115,
+      "learning_rate": 5.907268458462699e-06,
+      "loss": 0.5123,
+      "step": 4267
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.873656281415922,
+      "learning_rate": 5.905556232901437e-06,
+      "loss": 0.5034,
+      "step": 4268
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4088436663924379,
+      "learning_rate": 5.90384389753836e-06,
+      "loss": 0.2929,
+      "step": 4269
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.936129499207117,
+      "learning_rate": 5.902131452581097e-06,
+      "loss": 0.5071,
+      "step": 4270
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.313514096772514,
+      "learning_rate": 5.900418898237281e-06,
+      "loss": 0.5603,
+      "step": 4271
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.497302339587927,
+      "learning_rate": 5.8987062347145726e-06,
+      "loss": 0.5104,
+      "step": 4272
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.955254004606895,
+      "learning_rate": 5.896993462220634e-06,
+      "loss": 0.5122,
+      "step": 4273
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.558271981320908,
+      "learning_rate": 5.895280580963144e-06,
+      "loss": 0.465,
+      "step": 4274
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.460202865721472,
+      "learning_rate": 5.8935675911498e-06,
+      "loss": 0.5361,
+      "step": 4275
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.138826622221375,
+      "learning_rate": 5.891854492988303e-06,
+      "loss": 0.4586,
+      "step": 4276
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.475800046900567,
+      "learning_rate": 5.890141286686375e-06,
+      "loss": 0.4989,
+      "step": 4277
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.943693698161066,
+      "learning_rate": 5.888427972451747e-06,
+      "loss": 0.4885,
+      "step": 4278
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.88339811483524,
+      "learning_rate": 5.886714550492164e-06,
+      "loss": 0.518,
+      "step": 4279
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3422769904836884,
+      "learning_rate": 5.8850010210153844e-06,
+      "loss": 0.4744,
+      "step": 4280
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.1762396722798805,
+      "learning_rate": 5.883287384229182e-06,
+      "loss": 0.509,
+      "step": 4281
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.459353626607506,
+      "learning_rate": 5.881573640341336e-06,
+      "loss": 0.5018,
+      "step": 4282
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.264192158532596,
+      "learning_rate": 5.879859789559649e-06,
+      "loss": 0.5502,
+      "step": 4283
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.485367027740347,
+      "learning_rate": 5.878145832091929e-06,
+      "loss": 0.4807,
+      "step": 4284
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.78816226152491,
+      "learning_rate": 5.876431768145998e-06,
+      "loss": 0.503,
+      "step": 4285
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.535141325970753,
+      "learning_rate": 5.874717597929697e-06,
+      "loss": 0.5155,
+      "step": 4286
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3882523646761915,
+      "learning_rate": 5.873003321650869e-06,
+      "loss": 0.5274,
+      "step": 4287
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4226376325402661,
+      "learning_rate": 5.871288939517378e-06,
+      "loss": 0.312,
+      "step": 4288
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.495376055139042,
+      "learning_rate": 5.8695744517371e-06,
+      "loss": 0.4503,
+      "step": 4289
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.101005745895843,
+      "learning_rate": 5.8678598585179226e-06,
+      "loss": 0.4891,
+      "step": 4290
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.187833921710325,
+      "learning_rate": 5.866145160067746e-06,
+      "loss": 0.5175,
+      "step": 4291
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.570801200624612,
+      "learning_rate": 5.8644303565944805e-06,
+      "loss": 0.507,
+      "step": 4292
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.655277169965128,
+      "learning_rate": 5.862715448306056e-06,
+      "loss": 0.4501,
+      "step": 4293
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.541738841201937,
+      "learning_rate": 5.861000435410406e-06,
+      "loss": 0.5051,
+      "step": 4294
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.7631757633933365,
+      "learning_rate": 5.859285318115488e-06,
+      "loss": 0.4888,
+      "step": 4295
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.048234187208262,
+      "learning_rate": 5.857570096629263e-06,
+      "loss": 0.5047,
+      "step": 4296
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.642895947756619,
+      "learning_rate": 5.855854771159706e-06,
+      "loss": 0.5027,
+      "step": 4297
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2375244321677217,
+      "learning_rate": 5.854139341914808e-06,
+      "loss": 0.4431,
+      "step": 4298
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.535806621455097,
+      "learning_rate": 5.8524238091025684e-06,
+      "loss": 0.5143,
+      "step": 4299
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.308579379561763,
+      "learning_rate": 5.850708172931005e-06,
+      "loss": 0.4848,
+      "step": 4300
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.433682978772818,
+      "learning_rate": 5.848992433608143e-06,
+      "loss": 0.5218,
+      "step": 4301
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.7929109498048375,
+      "learning_rate": 5.847276591342021e-06,
+      "loss": 0.5002,
+      "step": 4302
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.1024589344500235,
+      "learning_rate": 5.845560646340691e-06,
+      "loss": 0.5331,
+      "step": 4303
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.594769395842418,
+      "learning_rate": 5.843844598812218e-06,
+      "loss": 0.4775,
+      "step": 4304
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4338912028112296,
+      "learning_rate": 5.842128448964677e-06,
+      "loss": 0.5021,
+      "step": 4305
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.814196444889715,
+      "learning_rate": 5.84041219700616e-06,
+      "loss": 0.5188,
+      "step": 4306
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.768557545677142,
+      "learning_rate": 5.838695843144766e-06,
+      "loss": 0.5365,
+      "step": 4307
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.562399544260929,
+      "learning_rate": 5.836979387588611e-06,
+      "loss": 0.5321,
+      "step": 4308
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4461398255240608,
+      "learning_rate": 5.835262830545817e-06,
+      "loss": 0.3052,
+      "step": 4309
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.6705148694705585,
+      "learning_rate": 5.833546172224527e-06,
+      "loss": 0.5219,
+      "step": 4310
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.86058418214563,
+      "learning_rate": 5.8318294128328885e-06,
+      "loss": 0.5353,
+      "step": 4311
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.014055940535888,
+      "learning_rate": 5.830112552579068e-06,
+      "loss": 0.4952,
+      "step": 4312
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.123464563293792,
+      "learning_rate": 5.828395591671238e-06,
+      "loss": 0.566,
+      "step": 4313
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.543434543977749,
+      "learning_rate": 5.826678530317585e-06,
+      "loss": 0.5269,
+      "step": 4314
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9240468915352587,
+      "learning_rate": 5.824961368726312e-06,
+      "loss": 0.5261,
+      "step": 4315
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.256048591015618,
+      "learning_rate": 5.823244107105628e-06,
+      "loss": 0.5042,
+      "step": 4316
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.850980692944686,
+      "learning_rate": 5.821526745663758e-06,
+      "loss": 0.5076,
+      "step": 4317
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.18127751550033,
+      "learning_rate": 5.8198092846089396e-06,
+      "loss": 0.4725,
+      "step": 4318
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.658201459440127,
+      "learning_rate": 5.818091724149417e-06,
+      "loss": 0.5285,
+      "step": 4319
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.08912799931431,
+      "learning_rate": 5.816374064493453e-06,
+      "loss": 0.4938,
+      "step": 4320
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8066897346747086,
+      "learning_rate": 5.81465630584932e-06,
+      "loss": 0.5504,
+      "step": 4321
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 22.623242963363786,
+      "learning_rate": 5.812938448425299e-06,
+      "loss": 0.5136,
+      "step": 4322
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.517278655387347,
+      "learning_rate": 5.811220492429692e-06,
+      "loss": 0.5303,
+      "step": 4323
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.0780958956981745,
+      "learning_rate": 5.809502438070801e-06,
+      "loss": 0.413,
+      "step": 4324
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9598437341707062,
+      "learning_rate": 5.80778428555695e-06,
+      "loss": 0.5333,
+      "step": 4325
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.169631385094758,
+      "learning_rate": 5.8060660350964685e-06,
+      "loss": 0.5049,
+      "step": 4326
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.172283203784086,
+      "learning_rate": 5.8043476868977025e-06,
+      "loss": 0.4408,
+      "step": 4327
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.40252366623761887,
+      "learning_rate": 5.802629241169006e-06,
+      "loss": 0.2948,
+      "step": 4328
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.158553400102274,
+      "learning_rate": 5.800910698118746e-06,
+      "loss": 0.5052,
+      "step": 4329
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.5242335257635995,
+      "learning_rate": 5.799192057955303e-06,
+      "loss": 0.4702,
+      "step": 4330
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.870844887509811,
+      "learning_rate": 5.7974733208870686e-06,
+      "loss": 0.4918,
+      "step": 4331
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.539444618373553,
+      "learning_rate": 5.795754487122444e-06,
+      "loss": 0.517,
+      "step": 4332
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.841795494909806,
+      "learning_rate": 5.794035556869843e-06,
+      "loss": 0.5365,
+      "step": 4333
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.695658718929564,
+      "learning_rate": 5.792316530337696e-06,
+      "loss": 0.5518,
+      "step": 4334
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.994963986079389,
+      "learning_rate": 5.790597407734437e-06,
+      "loss": 0.5029,
+      "step": 4335
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.882354146157046,
+      "learning_rate": 5.788878189268516e-06,
+      "loss": 0.5214,
+      "step": 4336
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4238268648922676,
+      "learning_rate": 5.787158875148396e-06,
+      "loss": 0.4983,
+      "step": 4337
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9455059048825483,
+      "learning_rate": 5.785439465582549e-06,
+      "loss": 0.5816,
+      "step": 4338
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.3820137321217825,
+      "learning_rate": 5.783719960779458e-06,
+      "loss": 0.466,
+      "step": 4339
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.813707943986676,
+      "learning_rate": 5.78200036094762e-06,
+      "loss": 0.4522,
+      "step": 4340
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.765225864912961,
+      "learning_rate": 5.780280666295544e-06,
+      "loss": 0.5084,
+      "step": 4341
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8460997395455134,
+      "learning_rate": 5.778560877031744e-06,
+      "loss": 0.4868,
+      "step": 4342
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.864807724416076,
+      "learning_rate": 5.7768409933647555e-06,
+      "loss": 0.4459,
+      "step": 4343
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.18291712664548,
+      "learning_rate": 5.775121015503118e-06,
+      "loss": 0.4787,
+      "step": 4344
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.863732733404876,
+      "learning_rate": 5.773400943655385e-06,
+      "loss": 0.5461,
+      "step": 4345
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.843036434545256,
+      "learning_rate": 5.771680778030122e-06,
+      "loss": 0.5018,
+      "step": 4346
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.871652024207888,
+      "learning_rate": 5.769960518835902e-06,
+      "loss": 0.5608,
+      "step": 4347
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.09392249804851,
+      "learning_rate": 5.768240166281317e-06,
+      "loss": 0.5271,
+      "step": 4348
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.39080800912982,
+      "learning_rate": 5.7665197205749636e-06,
+      "loss": 0.5035,
+      "step": 4349
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 15.512514881958678,
+      "learning_rate": 5.764799181925449e-06,
+      "loss": 0.4893,
+      "step": 4350
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.850116374710508,
+      "learning_rate": 5.763078550541399e-06,
+      "loss": 0.5177,
+      "step": 4351
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.495926468353721,
+      "learning_rate": 5.761357826631444e-06,
+      "loss": 0.4919,
+      "step": 4352
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.0524288887052,
+      "learning_rate": 5.759637010404226e-06,
+      "loss": 0.57,
+      "step": 4353
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.877522480746724,
+      "learning_rate": 5.757916102068403e-06,
+      "loss": 0.4827,
+      "step": 4354
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.168679844181967,
+      "learning_rate": 5.75619510183264e-06,
+      "loss": 0.4708,
+      "step": 4355
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.406645470309423,
+      "learning_rate": 5.754474009905613e-06,
+      "loss": 0.5375,
+      "step": 4356
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.149897357924335,
+      "learning_rate": 5.752752826496012e-06,
+      "loss": 0.484,
+      "step": 4357
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.036347970604309,
+      "learning_rate": 5.751031551812536e-06,
+      "loss": 0.5402,
+      "step": 4358
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.915545855537348,
+      "learning_rate": 5.749310186063892e-06,
+      "loss": 0.5444,
+      "step": 4359
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.626701776393109,
+      "learning_rate": 5.74758872945881e-06,
+      "loss": 0.4679,
+      "step": 4360
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9783670782543785,
+      "learning_rate": 5.745867182206013e-06,
+      "loss": 0.5433,
+      "step": 4361
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.115124279000797,
+      "learning_rate": 5.7441455445142505e-06,
+      "loss": 0.4864,
+      "step": 4362
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.937918598382472,
+      "learning_rate": 5.7424238165922745e-06,
+      "loss": 0.5416,
+      "step": 4363
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.420034506806576,
+      "learning_rate": 5.74070199864885e-06,
+      "loss": 0.5245,
+      "step": 4364
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4021218982159454,
+      "learning_rate": 5.7389800908927575e-06,
+      "loss": 0.5097,
+      "step": 4365
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.382682385021985,
+      "learning_rate": 5.737258093532781e-06,
+      "loss": 0.5063,
+      "step": 4366
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.6168119461096437,
+      "learning_rate": 5.735536006777717e-06,
+      "loss": 0.5209,
+      "step": 4367
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.6708164983063973,
+      "learning_rate": 5.733813830836377e-06,
+      "loss": 0.4981,
+      "step": 4368
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.506237037413279,
+      "learning_rate": 5.73209156591758e-06,
+      "loss": 0.4823,
+      "step": 4369
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7917043039342317,
+      "learning_rate": 5.7303692122301566e-06,
+      "loss": 0.4697,
+      "step": 4370
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.395850212588713,
+      "learning_rate": 5.728646769982951e-06,
+      "loss": 0.5308,
+      "step": 4371
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4094404589044873,
+      "learning_rate": 5.726924239384809e-06,
+      "loss": 0.3034,
+      "step": 4372
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.89355033493539,
+      "learning_rate": 5.725201620644598e-06,
+      "loss": 0.5072,
+      "step": 4373
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4063953065374332,
+      "learning_rate": 5.7234789139711914e-06,
+      "loss": 0.2994,
+      "step": 4374
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.623813822899887,
+      "learning_rate": 5.721756119573471e-06,
+      "loss": 0.5223,
+      "step": 4375
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7794375173581045,
+      "learning_rate": 5.720033237660334e-06,
+      "loss": 0.4959,
+      "step": 4376
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.619982498270374,
+      "learning_rate": 5.718310268440684e-06,
+      "loss": 0.5185,
+      "step": 4377
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.14363272909695,
+      "learning_rate": 5.716587212123436e-06,
+      "loss": 0.5043,
+      "step": 4378
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.372494277590319,
+      "learning_rate": 5.7148640689175194e-06,
+      "loss": 0.4892,
+      "step": 4379
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.140981527099241,
+      "learning_rate": 5.713140839031871e-06,
+      "loss": 0.4595,
+      "step": 4380
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.097064578418535,
+      "learning_rate": 5.711417522675437e-06,
+      "loss": 0.5534,
+      "step": 4381
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.741612914158712,
+      "learning_rate": 5.709694120057174e-06,
+      "loss": 0.5159,
+      "step": 4382
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.111056984429118,
+      "learning_rate": 5.707970631386054e-06,
+      "loss": 0.5307,
+      "step": 4383
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.925459117271098,
+      "learning_rate": 5.706247056871052e-06,
+      "loss": 0.462,
+      "step": 4384
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.564633519044099,
+      "learning_rate": 5.704523396721162e-06,
+      "loss": 0.5229,
+      "step": 4385
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9030666205551,
+      "learning_rate": 5.702799651145381e-06,
+      "loss": 0.4682,
+      "step": 4386
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.623947923103586,
+      "learning_rate": 5.701075820352718e-06,
+      "loss": 0.5347,
+      "step": 4387
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.1895878244421,
+      "learning_rate": 5.699351904552196e-06,
+      "loss": 0.5024,
+      "step": 4388
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.834025088751581,
+      "learning_rate": 5.697627903952845e-06,
+      "loss": 0.468,
+      "step": 4389
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.5567796506094504,
+      "learning_rate": 5.695903818763703e-06,
+      "loss": 0.4794,
+      "step": 4390
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.522782043016068,
+      "learning_rate": 5.694179649193826e-06,
+      "loss": 0.4253,
+      "step": 4391
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.580133666785593,
+      "learning_rate": 5.692455395452272e-06,
+      "loss": 0.498,
+      "step": 4392
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.862842594919066,
+      "learning_rate": 5.6907310577481155e-06,
+      "loss": 0.4799,
+      "step": 4393
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.386161811285565,
+      "learning_rate": 5.6890066362904374e-06,
+      "loss": 0.4995,
+      "step": 4394
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.85023622105803,
+      "learning_rate": 5.687282131288326e-06,
+      "loss": 0.523,
+      "step": 4395
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5726198812415766,
+      "learning_rate": 5.68555754295089e-06,
+      "loss": 0.464,
+      "step": 4396
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.265033475404112,
+      "learning_rate": 5.683832871487238e-06,
+      "loss": 0.4887,
+      "step": 4397
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.372724621883629,
+      "learning_rate": 5.682108117106491e-06,
+      "loss": 0.4503,
+      "step": 4398
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.857134961704146,
+      "learning_rate": 5.680383280017785e-06,
+      "loss": 0.4946,
+      "step": 4399
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.44795755770750556,
+      "learning_rate": 5.678658360430261e-06,
+      "loss": 0.2953,
+      "step": 4400
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.754517244237935,
+      "learning_rate": 5.676933358553068e-06,
+      "loss": 0.5425,
+      "step": 4401
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.6342812159997098,
+      "learning_rate": 5.6752082745953765e-06,
+      "loss": 0.4802,
+      "step": 4402
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.31258766599549,
+      "learning_rate": 5.673483108766349e-06,
+      "loss": 0.5436,
+      "step": 4403
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.754896081477901,
+      "learning_rate": 5.671757861275176e-06,
+      "loss": 0.5303,
+      "step": 4404
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.750429316130947,
+      "learning_rate": 5.670032532331045e-06,
+      "loss": 0.4866,
+      "step": 4405
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.365932482107271,
+      "learning_rate": 5.668307122143159e-06,
+      "loss": 0.52,
+      "step": 4406
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1912086290566224,
+      "learning_rate": 5.666581630920731e-06,
+      "loss": 0.4557,
+      "step": 4407
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.40940246025715393,
+      "learning_rate": 5.664856058872984e-06,
+      "loss": 0.317,
+      "step": 4408
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.953591478487465,
+      "learning_rate": 5.6631304062091456e-06,
+      "loss": 0.5683,
+      "step": 4409
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.757442061648173,
+      "learning_rate": 5.661404673138459e-06,
+      "loss": 0.5031,
+      "step": 4410
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.664625060530396,
+      "learning_rate": 5.659678859870177e-06,
+      "loss": 0.4844,
+      "step": 4411
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.468103733500771,
+      "learning_rate": 5.657952966613558e-06,
+      "loss": 0.4808,
+      "step": 4412
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4913744636276996,
+      "learning_rate": 5.656226993577875e-06,
+      "loss": 0.3038,
+      "step": 4413
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.406842924949899,
+      "learning_rate": 5.654500940972405e-06,
+      "loss": 0.5365,
+      "step": 4414
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6393864098088198,
+      "learning_rate": 5.652774809006439e-06,
+      "loss": 0.511,
+      "step": 4415
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9188873984398933,
+      "learning_rate": 5.6510485978892785e-06,
+      "loss": 0.5318,
+      "step": 4416
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5804597658903687,
+      "learning_rate": 5.64932230783023e-06,
+      "loss": 0.4672,
+      "step": 4417
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.686081556701269,
+      "learning_rate": 5.647595939038615e-06,
+      "loss": 0.4822,
+      "step": 4418
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.714569000715928,
+      "learning_rate": 5.6458694917237575e-06,
+      "loss": 0.4818,
+      "step": 4419
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.282180853366825,
+      "learning_rate": 5.644142966094997e-06,
+      "loss": 0.4926,
+      "step": 4420
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.86660971942284,
+      "learning_rate": 5.642416362361683e-06,
+      "loss": 0.48,
+      "step": 4421
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.319019347023991,
+      "learning_rate": 5.640689680733171e-06,
+      "loss": 0.4557,
+      "step": 4422
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.40175222737234434,
+      "learning_rate": 5.638962921418825e-06,
+      "loss": 0.2927,
+      "step": 4423
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.6396156620170705,
+      "learning_rate": 5.6372360846280224e-06,
+      "loss": 0.5146,
+      "step": 4424
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4076259025264878,
+      "learning_rate": 5.635509170570148e-06,
+      "loss": 0.2956,
+      "step": 4425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1313361789395007,
+      "learning_rate": 5.633782179454594e-06,
+      "loss": 0.5099,
+      "step": 4426
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.649028489205156,
+      "learning_rate": 5.632055111490768e-06,
+      "loss": 0.513,
+      "step": 4427
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.38921574554550586,
+      "learning_rate": 5.6303279668880794e-06,
+      "loss": 0.2979,
+      "step": 4428
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.500316222606366,
+      "learning_rate": 5.6286007458559525e-06,
+      "loss": 0.5256,
+      "step": 4429
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8496300426366963,
+      "learning_rate": 5.6268734486038164e-06,
+      "loss": 0.5006,
+      "step": 4430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1504356983051047,
+      "learning_rate": 5.6251460753411155e-06,
+      "loss": 0.507,
+      "step": 4431
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.7095375749220003,
+      "learning_rate": 5.623418626277296e-06,
+      "loss": 0.5081,
+      "step": 4432
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.9255527453332673,
+      "learning_rate": 5.62169110162182e-06,
+      "loss": 0.5232,
+      "step": 4433
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8272009391250235,
+      "learning_rate": 5.619963501584155e-06,
+      "loss": 0.483,
+      "step": 4434
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.585215158270238,
+      "learning_rate": 5.618235826373777e-06,
+      "loss": 0.5171,
+      "step": 4435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.656534928775183,
+      "learning_rate": 5.616508076200175e-06,
+      "loss": 0.4643,
+      "step": 4436
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4494165129591714,
+      "learning_rate": 5.6147802512728425e-06,
+      "loss": 0.5295,
+      "step": 4437
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.41270094730200935,
+      "learning_rate": 5.613052351801284e-06,
+      "loss": 0.3092,
+      "step": 4438
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.839482811865374,
+      "learning_rate": 5.611324377995017e-06,
+      "loss": 0.5151,
+      "step": 4439
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3929985216107586,
+      "learning_rate": 5.6095963300635585e-06,
+      "loss": 0.2817,
+      "step": 4440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.40652333694286,
+      "learning_rate": 5.607868208216445e-06,
+      "loss": 0.536,
+      "step": 4441
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.160512979891745,
+      "learning_rate": 5.606140012663215e-06,
+      "loss": 0.4836,
+      "step": 4442
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.738432697587726,
+      "learning_rate": 5.604411743613418e-06,
+      "loss": 0.5372,
+      "step": 4443
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.845561451384982,
+      "learning_rate": 5.6026834012766155e-06,
+      "loss": 0.4615,
+      "step": 4444
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3570196628245963,
+      "learning_rate": 5.6009549858623735e-06,
+      "loss": 0.4824,
+      "step": 4445
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.130980965596341,
+      "learning_rate": 5.5992264975802646e-06,
+      "loss": 0.5203,
+      "step": 4446
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.43312940350351053,
+      "learning_rate": 5.597497936639879e-06,
+      "loss": 0.2866,
+      "step": 4447
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4101590398710025,
+      "learning_rate": 5.595769303250809e-06,
+      "loss": 0.294,
+      "step": 4448
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.134679065095982,
+      "learning_rate": 5.594040597622654e-06,
+      "loss": 0.4813,
+      "step": 4449
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2876613963078953,
+      "learning_rate": 5.592311819965033e-06,
+      "loss": 0.5039,
+      "step": 4450
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3751875090813795,
+      "learning_rate": 5.590582970487558e-06,
+      "loss": 0.2914,
+      "step": 4451
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.512428863774006,
+      "learning_rate": 5.588854049399864e-06,
+      "loss": 0.5513,
+      "step": 4452
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.632195295175215,
+      "learning_rate": 5.587125056911587e-06,
+      "loss": 0.4807,
+      "step": 4453
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.7682114738960593,
+      "learning_rate": 5.585395993232371e-06,
+      "loss": 0.5517,
+      "step": 4454
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4275555894556664,
+      "learning_rate": 5.583666858571874e-06,
+      "loss": 0.4664,
+      "step": 4455
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.706682401084143,
+      "learning_rate": 5.581937653139757e-06,
+      "loss": 0.5183,
+      "step": 4456
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.83458671479752,
+      "learning_rate": 5.580208377145693e-06,
+      "loss": 0.4671,
+      "step": 4457
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.0110861618526865,
+      "learning_rate": 5.578479030799363e-06,
+      "loss": 0.5645,
+      "step": 4458
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.325648259149116,
+      "learning_rate": 5.576749614310457e-06,
+      "loss": 0.4512,
+      "step": 4459
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8888605204397386,
+      "learning_rate": 5.575020127888672e-06,
+      "loss": 0.4906,
+      "step": 4460
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.745258878853858,
+      "learning_rate": 5.573290571743715e-06,
+      "loss": 0.5278,
+      "step": 4461
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4655638079696381,
+      "learning_rate": 5.5715609460852985e-06,
+      "loss": 0.2914,
+      "step": 4462
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.45283863287011866,
+      "learning_rate": 5.5698312511231454e-06,
+      "loss": 0.3001,
+      "step": 4463
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 41.32795416488486,
+      "learning_rate": 5.568101487066991e-06,
+      "loss": 0.466,
+      "step": 4464
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.162410310696307,
+      "learning_rate": 5.566371654126573e-06,
+      "loss": 0.4718,
+      "step": 4465
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.491303662462943,
+      "learning_rate": 5.564641752511637e-06,
+      "loss": 0.5592,
+      "step": 4466
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.38747888459471524,
+      "learning_rate": 5.562911782431943e-06,
+      "loss": 0.2781,
+      "step": 4467
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.946027164438863,
+      "learning_rate": 5.561181744097255e-06,
+      "loss": 0.4977,
+      "step": 4468
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.987369744508452,
+      "learning_rate": 5.559451637717346e-06,
+      "loss": 0.5682,
+      "step": 4469
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4513549443700812,
+      "learning_rate": 5.557721463501997e-06,
+      "loss": 0.2983,
+      "step": 4470
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.379347451608599,
+      "learning_rate": 5.555991221660998e-06,
+      "loss": 0.4848,
+      "step": 4471
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4250753369072286,
+      "learning_rate": 5.554260912404146e-06,
+      "loss": 0.3069,
+      "step": 4472
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.469056266919626,
+      "learning_rate": 5.5525305359412485e-06,
+      "loss": 0.529,
+      "step": 4473
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6152966165317917,
+      "learning_rate": 5.550800092482117e-06,
+      "loss": 0.5154,
+      "step": 4474
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.24125126658158,
+      "learning_rate": 5.549069582236577e-06,
+      "loss": 0.4571,
+      "step": 4475
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.4680442999078736,
+      "learning_rate": 5.547339005414457e-06,
+      "loss": 0.5143,
+      "step": 4476
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.506594066497685,
+      "learning_rate": 5.545608362225594e-06,
+      "loss": 0.485,
+      "step": 4477
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.7211587143206137,
+      "learning_rate": 5.543877652879838e-06,
+      "loss": 0.5318,
+      "step": 4478
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.521968122599781,
+      "learning_rate": 5.542146877587042e-06,
+      "loss": 0.4684,
+      "step": 4479
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4358606347742382,
+      "learning_rate": 5.540416036557064e-06,
+      "loss": 0.2973,
+      "step": 4480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3845527903060852,
+      "learning_rate": 5.538685129999782e-06,
+      "loss": 0.4629,
+      "step": 4481
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.116962171309215,
+      "learning_rate": 5.536954158125071e-06,
+      "loss": 0.501,
+      "step": 4482
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4832295338650123,
+      "learning_rate": 5.535223121142817e-06,
+      "loss": 0.481,
+      "step": 4483
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.062874229113113,
+      "learning_rate": 5.5334920192629135e-06,
+      "loss": 0.5357,
+      "step": 4484
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.052193256584299,
+      "learning_rate": 5.531760852695264e-06,
+      "loss": 0.5233,
+      "step": 4485
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.920066980611309,
+      "learning_rate": 5.530029621649777e-06,
+      "loss": 0.4723,
+      "step": 4486
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5979669242317165,
+      "learning_rate": 5.528298326336372e-06,
+      "loss": 0.5534,
+      "step": 4487
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.804511940304774,
+      "learning_rate": 5.5265669669649725e-06,
+      "loss": 0.506,
+      "step": 4488
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.23649978941679,
+      "learning_rate": 5.524835543745515e-06,
+      "loss": 0.5051,
+      "step": 4489
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 15.296150284617124,
+      "learning_rate": 5.523104056887937e-06,
+      "loss": 0.4673,
+      "step": 4490
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6461199963473425,
+      "learning_rate": 5.521372506602187e-06,
+      "loss": 0.4893,
+      "step": 4491
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.4030320557233775,
+      "learning_rate": 5.519640893098227e-06,
+      "loss": 0.5521,
+      "step": 4492
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6727735974497135,
+      "learning_rate": 5.517909216586015e-06,
+      "loss": 0.51,
+      "step": 4493
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.0978528610782785,
+      "learning_rate": 5.516177477275524e-06,
+      "loss": 0.4954,
+      "step": 4494
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.280100111249574,
+      "learning_rate": 5.514445675376736e-06,
+      "loss": 0.4705,
+      "step": 4495
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 17.94301101171106,
+      "learning_rate": 5.512713811099636e-06,
+      "loss": 0.4814,
+      "step": 4496
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.699088150211978,
+      "learning_rate": 5.510981884654217e-06,
+      "loss": 0.5047,
+      "step": 4497
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.152950492872421,
+      "learning_rate": 5.509249896250486e-06,
+      "loss": 0.4645,
+      "step": 4498
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.198119290560598,
+      "learning_rate": 5.507517846098447e-06,
+      "loss": 0.5031,
+      "step": 4499
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.212034108510188,
+      "learning_rate": 5.505785734408121e-06,
+      "loss": 0.5078,
+      "step": 4500
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.518928454831877,
+      "learning_rate": 5.504053561389531e-06,
+      "loss": 0.471,
+      "step": 4501
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4115161795656745,
+      "learning_rate": 5.502321327252708e-06,
+      "loss": 0.2931,
+      "step": 4502
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.474372197050642,
+      "learning_rate": 5.500589032207693e-06,
+      "loss": 0.5064,
+      "step": 4503
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.63750419230446,
+      "learning_rate": 5.4988566764645315e-06,
+      "loss": 0.5353,
+      "step": 4504
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.50464049660943,
+      "learning_rate": 5.497124260233278e-06,
+      "loss": 0.5295,
+      "step": 4505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.789446976494075,
+      "learning_rate": 5.495391783723993e-06,
+      "loss": 0.4977,
+      "step": 4506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.871461850646656,
+      "learning_rate": 5.493659247146749e-06,
+      "loss": 0.5307,
+      "step": 4507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.7965263888483864,
+      "learning_rate": 5.491926650711618e-06,
+      "loss": 0.5063,
+      "step": 4508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6887886130609053,
+      "learning_rate": 5.490193994628685e-06,
+      "loss": 0.5234,
+      "step": 4509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.876201263482505,
+      "learning_rate": 5.488461279108041e-06,
+      "loss": 0.486,
+      "step": 4510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.2986596801398855,
+      "learning_rate": 5.486728504359781e-06,
+      "loss": 0.4596,
+      "step": 4511
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4133079070222929,
+      "learning_rate": 5.484995670594012e-06,
+      "loss": 0.2903,
+      "step": 4512
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.989377067545663,
+      "learning_rate": 5.483262778020847e-06,
+      "loss": 0.5022,
+      "step": 4513
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.5262061411695,
+      "learning_rate": 5.481529826850404e-06,
+      "loss": 0.567,
+      "step": 4514
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.410320298313644,
+      "learning_rate": 5.479796817292809e-06,
+      "loss": 0.5261,
+      "step": 4515
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.250036359854238,
+      "learning_rate": 5.478063749558195e-06,
+      "loss": 0.499,
+      "step": 4516
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6199643953596836,
+      "learning_rate": 5.476330623856705e-06,
+      "loss": 0.4537,
+      "step": 4517
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.134305989292847,
+      "learning_rate": 5.474597440398483e-06,
+      "loss": 0.4996,
+      "step": 4518
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3855637878248297,
+      "learning_rate": 5.472864199393687e-06,
+      "loss": 0.5534,
+      "step": 4519
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.598129468301732,
+      "learning_rate": 5.4711309010524764e-06,
+      "loss": 0.4949,
+      "step": 4520
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.534239662300832,
+      "learning_rate": 5.469397545585019e-06,
+      "loss": 0.5411,
+      "step": 4521
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.98020570531071,
+      "learning_rate": 5.46766413320149e-06,
+      "loss": 0.5016,
+      "step": 4522
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.922702921602212,
+      "learning_rate": 5.465930664112073e-06,
+      "loss": 0.5521,
+      "step": 4523
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.704909520411667,
+      "learning_rate": 5.46419713852696e-06,
+      "loss": 0.5308,
+      "step": 4524
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.862846954923357,
+      "learning_rate": 5.46246355665634e-06,
+      "loss": 0.498,
+      "step": 4525
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.255373497390539,
+      "learning_rate": 5.4607299187104205e-06,
+      "loss": 0.5082,
+      "step": 4526
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.741626696313572,
+      "learning_rate": 5.458996224899409e-06,
+      "loss": 0.5247,
+      "step": 4527
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.359616498647196,
+      "learning_rate": 5.457262475433523e-06,
+      "loss": 0.5761,
+      "step": 4528
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1800717967167262,
+      "learning_rate": 5.455528670522987e-06,
+      "loss": 0.4833,
+      "step": 4529
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.682770309230754,
+      "learning_rate": 5.453794810378028e-06,
+      "loss": 0.5381,
+      "step": 4530
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.269346075602417,
+      "learning_rate": 5.452060895208883e-06,
+      "loss": 0.5211,
+      "step": 4531
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.713316683576639,
+      "learning_rate": 5.450326925225798e-06,
+      "loss": 0.4736,
+      "step": 4532
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3950956784415751,
+      "learning_rate": 5.44859290063902e-06,
+      "loss": 0.2951,
+      "step": 4533
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.370893156362593,
+      "learning_rate": 5.446858821658806e-06,
+      "loss": 0.4462,
+      "step": 4534
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.40733316647582757,
+      "learning_rate": 5.44512468849542e-06,
+      "loss": 0.3074,
+      "step": 4535
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.11155946143057,
+      "learning_rate": 5.44339050135913e-06,
+      "loss": 0.4741,
+      "step": 4536
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.97780823222654,
+      "learning_rate": 5.441656260460213e-06,
+      "loss": 0.5095,
+      "step": 4537
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.256743224944103,
+      "learning_rate": 5.439921966008953e-06,
+      "loss": 0.5179,
+      "step": 4538
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.363069664758185,
+      "learning_rate": 5.438187618215636e-06,
+      "loss": 0.4802,
+      "step": 4539
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.598149902666416,
+      "learning_rate": 5.436453217290562e-06,
+      "loss": 0.4771,
+      "step": 4540
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.719986314755784,
+      "learning_rate": 5.43471876344403e-06,
+      "loss": 0.5063,
+      "step": 4541
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.066562033139768,
+      "learning_rate": 5.432984256886348e-06,
+      "loss": 0.5224,
+      "step": 4542
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6016568842860974,
+      "learning_rate": 5.431249697827833e-06,
+      "loss": 0.4913,
+      "step": 4543
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3803719306589954,
+      "learning_rate": 5.429515086478805e-06,
+      "loss": 0.4833,
+      "step": 4544
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.867777721067538,
+      "learning_rate": 5.4277804230495935e-06,
+      "loss": 0.4832,
+      "step": 4545
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.317453658178561,
+      "learning_rate": 5.426045707750529e-06,
+      "loss": 0.5517,
+      "step": 4546
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.56570476676683,
+      "learning_rate": 5.424310940791954e-06,
+      "loss": 0.5159,
+      "step": 4547
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.062939513780345,
+      "learning_rate": 5.422576122384216e-06,
+      "loss": 0.5124,
+      "step": 4548
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.09032322781076,
+      "learning_rate": 5.420841252737664e-06,
+      "loss": 0.498,
+      "step": 4549
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.043033451117925,
+      "learning_rate": 5.419106332062661e-06,
+      "loss": 0.526,
+      "step": 4550
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.993041097672667,
+      "learning_rate": 5.41737136056957e-06,
+      "loss": 0.4813,
+      "step": 4551
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6543443658259034,
+      "learning_rate": 5.415636338468763e-06,
+      "loss": 0.5036,
+      "step": 4552
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.326727657748518,
+      "learning_rate": 5.413901265970616e-06,
+      "loss": 0.5246,
+      "step": 4553
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0746649312898295,
+      "learning_rate": 5.412166143285515e-06,
+      "loss": 0.4724,
+      "step": 4554
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9012519650049002,
+      "learning_rate": 5.410430970623847e-06,
+      "loss": 0.5007,
+      "step": 4555
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9066815521938523,
+      "learning_rate": 5.408695748196009e-06,
+      "loss": 0.5283,
+      "step": 4556
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.706357221514703,
+      "learning_rate": 5.406960476212403e-06,
+      "loss": 0.4723,
+      "step": 4557
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.99399712405928,
+      "learning_rate": 5.405225154883436e-06,
+      "loss": 0.4971,
+      "step": 4558
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.238969651827891,
+      "learning_rate": 5.403489784419521e-06,
+      "loss": 0.486,
+      "step": 4559
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.0479092794061335,
+      "learning_rate": 5.40175436503108e-06,
+      "loss": 0.5075,
+      "step": 4560
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.40885681356214326,
+      "learning_rate": 5.400018896928537e-06,
+      "loss": 0.2922,
+      "step": 4561
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.567289101209348,
+      "learning_rate": 5.398283380322323e-06,
+      "loss": 0.4952,
+      "step": 4562
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.167467790370543,
+      "learning_rate": 5.396547815422876e-06,
+      "loss": 0.5421,
+      "step": 4563
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4940154823862453,
+      "learning_rate": 5.3948122024406405e-06,
+      "loss": 0.4618,
+      "step": 4564
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6334425495210216,
+      "learning_rate": 5.393076541586063e-06,
+      "loss": 0.5526,
+      "step": 4565
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 27.166859475234773,
+      "learning_rate": 5.391340833069601e-06,
+      "loss": 0.4795,
+      "step": 4566
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.55272061583898,
+      "learning_rate": 5.389605077101713e-06,
+      "loss": 0.5507,
+      "step": 4567
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4102373235908168,
+      "learning_rate": 5.387869273892866e-06,
+      "loss": 0.5378,
+      "step": 4568
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.841627875410967,
+      "learning_rate": 5.386133423653532e-06,
+      "loss": 0.5146,
+      "step": 4569
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.422310324478609,
+      "learning_rate": 5.3843975265941896e-06,
+      "loss": 0.5384,
+      "step": 4570
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.135748305822562,
+      "learning_rate": 5.382661582925322e-06,
+      "loss": 0.5185,
+      "step": 4571
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.836137702813622,
+      "learning_rate": 5.38092559285742e-06,
+      "loss": 0.5053,
+      "step": 4572
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.691751102793737,
+      "learning_rate": 5.379189556600974e-06,
+      "loss": 0.4799,
+      "step": 4573
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2723372943797653,
+      "learning_rate": 5.3774534743664885e-06,
+      "loss": 0.494,
+      "step": 4574
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.511648988789479,
+      "learning_rate": 5.375717346364467e-06,
+      "loss": 0.4942,
+      "step": 4575
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.719599821681616,
+      "learning_rate": 5.373981172805421e-06,
+      "loss": 0.4997,
+      "step": 4576
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 15.447465684064463,
+      "learning_rate": 5.3722449538998725e-06,
+      "loss": 0.471,
+      "step": 4577
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.293326355091955,
+      "learning_rate": 5.370508689858336e-06,
+      "loss": 0.4945,
+      "step": 4578
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.33088188084676,
+      "learning_rate": 5.368772380891345e-06,
+      "loss": 0.5204,
+      "step": 4579
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.591020627247747,
+      "learning_rate": 5.367036027209431e-06,
+      "loss": 0.5544,
+      "step": 4580
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0493527954465294,
+      "learning_rate": 5.365299629023134e-06,
+      "loss": 0.4985,
+      "step": 4581
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.330075539177581,
+      "learning_rate": 5.363563186542997e-06,
+      "loss": 0.4366,
+      "step": 4582
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.544567985798193,
+      "learning_rate": 5.361826699979571e-06,
+      "loss": 0.5281,
+      "step": 4583
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.84835628984391,
+      "learning_rate": 5.36009016954341e-06,
+      "loss": 0.533,
+      "step": 4584
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.84796403730724,
+      "learning_rate": 5.358353595445074e-06,
+      "loss": 0.4886,
+      "step": 4585
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.308468347484837,
+      "learning_rate": 5.35661697789513e-06,
+      "loss": 0.5062,
+      "step": 4586
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.665180942960016,
+      "learning_rate": 5.354880317104145e-06,
+      "loss": 0.5285,
+      "step": 4587
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.216766912950157,
+      "learning_rate": 5.353143613282702e-06,
+      "loss": 0.4765,
+      "step": 4588
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8348502009362444,
+      "learning_rate": 5.351406866641378e-06,
+      "loss": 0.4776,
+      "step": 4589
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4057685528438898,
+      "learning_rate": 5.349670077390757e-06,
+      "loss": 0.2917,
+      "step": 4590
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.149578837844703,
+      "learning_rate": 5.347933245741435e-06,
+      "loss": 0.5097,
+      "step": 4591
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.460340180520887,
+      "learning_rate": 5.346196371904009e-06,
+      "loss": 0.5084,
+      "step": 4592
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.914354009068969,
+      "learning_rate": 5.344459456089078e-06,
+      "loss": 0.4946,
+      "step": 4593
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.850146563479278,
+      "learning_rate": 5.342722498507251e-06,
+      "loss": 0.4592,
+      "step": 4594
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.185477742984284,
+      "learning_rate": 5.340985499369138e-06,
+      "loss": 0.5206,
+      "step": 4595
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.957271946523696,
+      "learning_rate": 5.33924845888536e-06,
+      "loss": 0.5546,
+      "step": 4596
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.56505179309784,
+      "learning_rate": 5.337511377266535e-06,
+      "loss": 0.489,
+      "step": 4597
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.6693139797253105,
+      "learning_rate": 5.335774254723293e-06,
+      "loss": 0.5138,
+      "step": 4598
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.402602275876383,
+      "learning_rate": 5.334037091466265e-06,
+      "loss": 0.5209,
+      "step": 4599
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.2165181762450015,
+      "learning_rate": 5.332299887706087e-06,
+      "loss": 0.4928,
+      "step": 4600
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.974142094161453,
+      "learning_rate": 5.330562643653402e-06,
+      "loss": 0.5318,
+      "step": 4601
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2812319063915902,
+      "learning_rate": 5.3288253595188575e-06,
+      "loss": 0.4934,
+      "step": 4602
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.444057891906714,
+      "learning_rate": 5.327088035513105e-06,
+      "loss": 0.5103,
+      "step": 4603
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.417857338012199,
+      "learning_rate": 5.325350671846801e-06,
+      "loss": 0.5315,
+      "step": 4604
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.832196381900148,
+      "learning_rate": 5.323613268730605e-06,
+      "loss": 0.5155,
+      "step": 4605
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.260507603803636,
+      "learning_rate": 5.321875826375185e-06,
+      "loss": 0.4437,
+      "step": 4606
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.581715625256831,
+      "learning_rate": 5.32013834499121e-06,
+      "loss": 0.556,
+      "step": 4607
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.43439315267799733,
+      "learning_rate": 5.3184008247893595e-06,
+      "loss": 0.2838,
+      "step": 4608
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.939574520506094,
+      "learning_rate": 5.31666326598031e-06,
+      "loss": 0.4522,
+      "step": 4609
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.501806288739884,
+      "learning_rate": 5.314925668774747e-06,
+      "loss": 0.5495,
+      "step": 4610
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.186368132351436,
+      "learning_rate": 5.313188033383363e-06,
+      "loss": 0.5626,
+      "step": 4611
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1604933548580223,
+      "learning_rate": 5.311450360016848e-06,
+      "loss": 0.4518,
+      "step": 4612
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.336473438081651,
+      "learning_rate": 5.3097126488859044e-06,
+      "loss": 0.5166,
+      "step": 4613
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.877449515027714,
+      "learning_rate": 5.3079749002012345e-06,
+      "loss": 0.4928,
+      "step": 4614
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.30639788731358,
+      "learning_rate": 5.306237114173546e-06,
+      "loss": 0.481,
+      "step": 4615
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.842125375763373,
+      "learning_rate": 5.3044992910135515e-06,
+      "loss": 0.4962,
+      "step": 4616
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.802306308057851,
+      "learning_rate": 5.302761430931969e-06,
+      "loss": 0.4846,
+      "step": 4617
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.229625938973382,
+      "learning_rate": 5.301023534139516e-06,
+      "loss": 0.4924,
+      "step": 4618
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.266374185503815,
+      "learning_rate": 5.299285600846926e-06,
+      "loss": 0.4626,
+      "step": 4619
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7262722163696473,
+      "learning_rate": 5.297547631264923e-06,
+      "loss": 0.5197,
+      "step": 4620
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9632921543094555,
+      "learning_rate": 5.2958096256042435e-06,
+      "loss": 0.5211,
+      "step": 4621
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.40718587366392733,
+      "learning_rate": 5.294071584075628e-06,
+      "loss": 0.3014,
+      "step": 4622
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.398878730548879,
+      "learning_rate": 5.2923335068898195e-06,
+      "loss": 0.5317,
+      "step": 4623
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5254345156673357,
+      "learning_rate": 5.290595394257564e-06,
+      "loss": 0.4616,
+      "step": 4624
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7638167942408813,
+      "learning_rate": 5.288857246389617e-06,
+      "loss": 0.4721,
+      "step": 4625
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.32711050650719,
+      "learning_rate": 5.287119063496731e-06,
+      "loss": 0.4466,
+      "step": 4626
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.183870196351034,
+      "learning_rate": 5.2853808457896695e-06,
+      "loss": 0.4621,
+      "step": 4627
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.698292100146581,
+      "learning_rate": 5.2836425934791965e-06,
+      "loss": 0.4967,
+      "step": 4628
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4054051518096644,
+      "learning_rate": 5.2819043067760824e-06,
+      "loss": 0.4929,
+      "step": 4629
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.973093052590071,
+      "learning_rate": 5.280165985891098e-06,
+      "loss": 0.5063,
+      "step": 4630
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.8379911556076958,
+      "learning_rate": 5.278427631035022e-06,
+      "loss": 0.4509,
+      "step": 4631
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.864706412279725,
+      "learning_rate": 5.276689242418635e-06,
+      "loss": 0.516,
+      "step": 4632
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.087282811527672,
+      "learning_rate": 5.274950820252725e-06,
+      "loss": 0.4856,
+      "step": 4633
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7827423588131923,
+      "learning_rate": 5.27321236474808e-06,
+      "loss": 0.5511,
+      "step": 4634
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.707082149799519,
+      "learning_rate": 5.271473876115494e-06,
+      "loss": 0.4542,
+      "step": 4635
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7583627471284085,
+      "learning_rate": 5.269735354565764e-06,
+      "loss": 0.4685,
+      "step": 4636
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.751533675265402,
+      "learning_rate": 5.267996800309693e-06,
+      "loss": 0.5018,
+      "step": 4637
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9884159008709594,
+      "learning_rate": 5.266258213558084e-06,
+      "loss": 0.5274,
+      "step": 4638
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.215676613578224,
+      "learning_rate": 5.264519594521751e-06,
+      "loss": 0.4771,
+      "step": 4639
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.072580124691432,
+      "learning_rate": 5.262780943411504e-06,
+      "loss": 0.4862,
+      "step": 4640
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.487393613520695,
+      "learning_rate": 5.261042260438163e-06,
+      "loss": 0.492,
+      "step": 4641
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.231761894212566,
+      "learning_rate": 5.259303545812546e-06,
+      "loss": 0.4821,
+      "step": 4642
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.647461988952553,
+      "learning_rate": 5.257564799745481e-06,
+      "loss": 0.55,
+      "step": 4643
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.081293498536146,
+      "learning_rate": 5.255826022447796e-06,
+      "loss": 0.5259,
+      "step": 4644
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.1692087260141335,
+      "learning_rate": 5.254087214130324e-06,
+      "loss": 0.5005,
+      "step": 4645
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.126157551431122,
+      "learning_rate": 5.252348375003902e-06,
+      "loss": 0.5242,
+      "step": 4646
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3719642301377672,
+      "learning_rate": 5.2506095052793695e-06,
+      "loss": 0.4593,
+      "step": 4647
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.598576167785402,
+      "learning_rate": 5.24887060516757e-06,
+      "loss": 0.4902,
+      "step": 4648
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4017017066179695,
+      "learning_rate": 5.24713167487935e-06,
+      "loss": 0.2907,
+      "step": 4649
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.13689070647237,
+      "learning_rate": 5.2453927146255646e-06,
+      "loss": 0.5211,
+      "step": 4650
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.264454789848069,
+      "learning_rate": 5.2436537246170675e-06,
+      "loss": 0.4915,
+      "step": 4651
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.137686749873323,
+      "learning_rate": 5.241914705064713e-06,
+      "loss": 0.5161,
+      "step": 4652
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3959851810066143,
+      "learning_rate": 5.2401756561793686e-06,
+      "loss": 0.2934,
+      "step": 4653
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.507922625378454,
+      "learning_rate": 5.238436578171899e-06,
+      "loss": 0.4674,
+      "step": 4654
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.8387645574158387,
+      "learning_rate": 5.236697471253168e-06,
+      "loss": 0.5112,
+      "step": 4655
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3666730421514272,
+      "learning_rate": 5.2349583356340575e-06,
+      "loss": 0.2932,
+      "step": 4656
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.592651616934946,
+      "learning_rate": 5.233219171525435e-06,
+      "loss": 0.5154,
+      "step": 4657
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.176355812543826,
+      "learning_rate": 5.231479979138187e-06,
+      "loss": 0.4502,
+      "step": 4658
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.896675198495905,
+      "learning_rate": 5.229740758683191e-06,
+      "loss": 0.4845,
+      "step": 4659
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4706129837262063,
+      "learning_rate": 5.228001510371336e-06,
+      "loss": 0.5146,
+      "step": 4660
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.05360534698578,
+      "learning_rate": 5.226262234413514e-06,
+      "loss": 0.4298,
+      "step": 4661
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.195824517830949,
+      "learning_rate": 5.224522931020616e-06,
+      "loss": 0.4426,
+      "step": 4662
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6671351519371638,
+      "learning_rate": 5.222783600403536e-06,
+      "loss": 0.5593,
+      "step": 4663
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.838081504984221,
+      "learning_rate": 5.2210442427731774e-06,
+      "loss": 0.53,
+      "step": 4664
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1004258660634982,
+      "learning_rate": 5.2193048583404435e-06,
+      "loss": 0.4454,
+      "step": 4665
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.654251743708246,
+      "learning_rate": 5.217565447316238e-06,
+      "loss": 0.4825,
+      "step": 4666
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.86803374837713,
+      "learning_rate": 5.215826009911474e-06,
+      "loss": 0.5343,
+      "step": 4667
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0468929494411037,
+      "learning_rate": 5.214086546337061e-06,
+      "loss": 0.5217,
+      "step": 4668
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.7627857296889875,
+      "learning_rate": 5.212347056803916e-06,
+      "loss": 0.5153,
+      "step": 4669
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.65994406939705,
+      "learning_rate": 5.210607541522959e-06,
+      "loss": 0.5433,
+      "step": 4670
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.550930736751321,
+      "learning_rate": 5.208868000705109e-06,
+      "loss": 0.5165,
+      "step": 4671
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.682824889186032,
+      "learning_rate": 5.2071284345612975e-06,
+      "loss": 0.4816,
+      "step": 4672
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.404574974388358,
+      "learning_rate": 5.205388843302447e-06,
+      "loss": 0.5035,
+      "step": 4673
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.592164134863301,
+      "learning_rate": 5.2036492271394915e-06,
+      "loss": 0.4397,
+      "step": 4674
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.522648852109246,
+      "learning_rate": 5.201909586283365e-06,
+      "loss": 0.4634,
+      "step": 4675
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.89156871015695,
+      "learning_rate": 5.200169920945005e-06,
+      "loss": 0.4657,
+      "step": 4676
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2268882357446875,
+      "learning_rate": 5.198430231335353e-06,
+      "loss": 0.5383,
+      "step": 4677
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.582613581005698,
+      "learning_rate": 5.19669051766535e-06,
+      "loss": 0.4817,
+      "step": 4678
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.236144588819692,
+      "learning_rate": 5.1949507801459455e-06,
+      "loss": 0.4852,
+      "step": 4679
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.585313169403181,
+      "learning_rate": 5.193211018988085e-06,
+      "loss": 0.4663,
+      "step": 4680
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.522520794995867,
+      "learning_rate": 5.191471234402723e-06,
+      "loss": 0.5137,
+      "step": 4681
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.435782100707419,
+      "learning_rate": 5.189731426600814e-06,
+      "loss": 0.4794,
+      "step": 4682
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0740595384230383,
+      "learning_rate": 5.187991595793314e-06,
+      "loss": 0.4445,
+      "step": 4683
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.8903644994818696,
+      "learning_rate": 5.1862517421911875e-06,
+      "loss": 0.462,
+      "step": 4684
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6604148209841294,
+      "learning_rate": 5.184511866005393e-06,
+      "loss": 0.4908,
+      "step": 4685
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.673809082140862,
+      "learning_rate": 5.182771967446899e-06,
+      "loss": 0.5228,
+      "step": 4686
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5586161411514317,
+      "learning_rate": 5.181032046726674e-06,
+      "loss": 0.5143,
+      "step": 4687
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.240420512883406,
+      "learning_rate": 5.17929210405569e-06,
+      "loss": 0.5153,
+      "step": 4688
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.01501746180332,
+      "learning_rate": 5.177552139644919e-06,
+      "loss": 0.5383,
+      "step": 4689
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.116218258574809,
+      "learning_rate": 5.17581215370534e-06,
+      "loss": 0.5126,
+      "step": 4690
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.221176631932174,
+      "learning_rate": 5.17407214644793e-06,
+      "loss": 0.4802,
+      "step": 4691
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7935171576314812,
+      "learning_rate": 5.172332118083673e-06,
+      "loss": 0.529,
+      "step": 4692
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5495606792428243,
+      "learning_rate": 5.170592068823554e-06,
+      "loss": 0.5221,
+      "step": 4693
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.774203818671671,
+      "learning_rate": 5.168851998878556e-06,
+      "loss": 0.4812,
+      "step": 4694
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.218614392539019,
+      "learning_rate": 5.1671119084596725e-06,
+      "loss": 0.5213,
+      "step": 4695
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4047699834221342,
+      "learning_rate": 5.165371797777894e-06,
+      "loss": 0.2964,
+      "step": 4696
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.765639643822658,
+      "learning_rate": 5.163631667044213e-06,
+      "loss": 0.4916,
+      "step": 4697
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7200860126694604,
+      "learning_rate": 5.1618915164696304e-06,
+      "loss": 0.4897,
+      "step": 4698
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.918616696862313,
+      "learning_rate": 5.160151346265143e-06,
+      "loss": 0.4647,
+      "step": 4699
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.694908141241071,
+      "learning_rate": 5.158411156641752e-06,
+      "loss": 0.4755,
+      "step": 4700
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.394240634760606,
+      "learning_rate": 5.156670947810463e-06,
+      "loss": 0.4636,
+      "step": 4701
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.795076721078398,
+      "learning_rate": 5.154930719982281e-06,
+      "loss": 0.4975,
+      "step": 4702
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.018492667593874,
+      "learning_rate": 5.153190473368213e-06,
+      "loss": 0.4872,
+      "step": 4703
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.588799805805137,
+      "learning_rate": 5.151450208179276e-06,
+      "loss": 0.4615,
+      "step": 4704
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.07927638010551,
+      "learning_rate": 5.149709924626476e-06,
+      "loss": 0.4485,
+      "step": 4705
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.161420588278182,
+      "learning_rate": 5.147969622920833e-06,
+      "loss": 0.4739,
+      "step": 4706
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0251344277725374,
+      "learning_rate": 5.146229303273364e-06,
+      "loss": 0.487,
+      "step": 4707
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.7200287008077866,
+      "learning_rate": 5.144488965895084e-06,
+      "loss": 0.5313,
+      "step": 4708
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9426569768086,
+      "learning_rate": 5.142748610997024e-06,
+      "loss": 0.4908,
+      "step": 4709
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1079814413789397,
+      "learning_rate": 5.1410082387902e-06,
+      "loss": 0.4753,
+      "step": 4710
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.021269560622619,
+      "learning_rate": 5.139267849485639e-06,
+      "loss": 0.4763,
+      "step": 4711
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.853021007234958,
+      "learning_rate": 5.1375274432943745e-06,
+      "loss": 0.572,
+      "step": 4712
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.079198228108806,
+      "learning_rate": 5.135787020427431e-06,
+      "loss": 0.4605,
+      "step": 4713
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.93558150847388,
+      "learning_rate": 5.134046581095844e-06,
+      "loss": 0.4761,
+      "step": 4714
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.924034323816166,
+      "learning_rate": 5.132306125510649e-06,
+      "loss": 0.4643,
+      "step": 4715
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.799692114614864,
+      "learning_rate": 5.130565653882877e-06,
+      "loss": 0.4946,
+      "step": 4716
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.305947012206886,
+      "learning_rate": 5.12882516642357e-06,
+      "loss": 0.5164,
+      "step": 4717
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.062245903817741,
+      "learning_rate": 5.127084663343769e-06,
+      "loss": 0.4984,
+      "step": 4718
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2864332381876986,
+      "learning_rate": 5.125344144854514e-06,
+      "loss": 0.4997,
+      "step": 4719
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.747629656125985,
+      "learning_rate": 5.123603611166849e-06,
+      "loss": 0.487,
+      "step": 4720
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.41321311779743036,
+      "learning_rate": 5.121863062491819e-06,
+      "loss": 0.3126,
+      "step": 4721
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.709663486456733,
+      "learning_rate": 5.120122499040473e-06,
+      "loss": 0.4385,
+      "step": 4722
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 21.65718614128039,
+      "learning_rate": 5.118381921023859e-06,
+      "loss": 0.5018,
+      "step": 4723
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 32.96628164653601,
+      "learning_rate": 5.116641328653031e-06,
+      "loss": 0.5034,
+      "step": 4724
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4030455403807594,
+      "learning_rate": 5.1149007221390386e-06,
+      "loss": 0.2844,
+      "step": 4725
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.9339274308882484,
+      "learning_rate": 5.113160101692939e-06,
+      "loss": 0.484,
+      "step": 4726
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.446350434228706,
+      "learning_rate": 5.111419467525786e-06,
+      "loss": 0.4783,
+      "step": 4727
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.596783918554954,
+      "learning_rate": 5.109678819848637e-06,
+      "loss": 0.4341,
+      "step": 4728
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1618679287538365,
+      "learning_rate": 5.1079381588725544e-06,
+      "loss": 0.5429,
+      "step": 4729
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.252620613675395,
+      "learning_rate": 5.106197484808598e-06,
+      "loss": 0.4972,
+      "step": 4730
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.579893191018884,
+      "learning_rate": 5.104456797867831e-06,
+      "loss": 0.5169,
+      "step": 4731
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.944981186606525,
+      "learning_rate": 5.102716098261316e-06,
+      "loss": 0.5166,
+      "step": 4732
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4141647734618723,
+      "learning_rate": 5.10097538620012e-06,
+      "loss": 0.5533,
+      "step": 4733
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.129532803800301,
+      "learning_rate": 5.09923466189531e-06,
+      "loss": 0.5864,
+      "step": 4734
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.424380072541432,
+      "learning_rate": 5.097493925557956e-06,
+      "loss": 0.4943,
+      "step": 4735
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9601912577646106,
+      "learning_rate": 5.095753177399127e-06,
+      "loss": 0.4992,
+      "step": 4736
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.405490204041079,
+      "learning_rate": 5.0940124176298955e-06,
+      "loss": 0.4828,
+      "step": 4737
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.920513366856633,
+      "learning_rate": 5.092271646461334e-06,
+      "loss": 0.5191,
+      "step": 4738
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.80609564949806,
+      "learning_rate": 5.090530864104517e-06,
+      "loss": 0.4927,
+      "step": 4739
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.092136265044978,
+      "learning_rate": 5.0887900707705205e-06,
+      "loss": 0.5137,
+      "step": 4740
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.962419518449081,
+      "learning_rate": 5.087049266670424e-06,
+      "loss": 0.54,
+      "step": 4741
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9961524757645996,
+      "learning_rate": 5.085308452015301e-06,
+      "loss": 0.5083,
+      "step": 4742
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.87079212514761,
+      "learning_rate": 5.0835676270162346e-06,
+      "loss": 0.5414,
+      "step": 4743
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5862574311507074,
+      "learning_rate": 5.081826791884307e-06,
+      "loss": 0.4438,
+      "step": 4744
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.01552293129829,
+      "learning_rate": 5.080085946830597e-06,
+      "loss": 0.4639,
+      "step": 4745
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.728044248570752,
+      "learning_rate": 5.078345092066191e-06,
+      "loss": 0.5464,
+      "step": 4746
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.426352441134003,
+      "learning_rate": 5.076604227802172e-06,
+      "loss": 0.4614,
+      "step": 4747
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.813700152425339,
+      "learning_rate": 5.074863354249625e-06,
+      "loss": 0.4855,
+      "step": 4748
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.492100172346991,
+      "learning_rate": 5.073122471619641e-06,
+      "loss": 0.5223,
+      "step": 4749
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.38726674653970106,
+      "learning_rate": 5.0713815801233025e-06,
+      "loss": 0.3055,
+      "step": 4750
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.961228249132467,
+      "learning_rate": 5.069640679971702e-06,
+      "loss": 0.5249,
+      "step": 4751
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.144937468046842,
+      "learning_rate": 5.0678997713759305e-06,
+      "loss": 0.5219,
+      "step": 4752
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.226817452572389,
+      "learning_rate": 5.066158854547076e-06,
+      "loss": 0.5095,
+      "step": 4753
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.785025416721538,
+      "learning_rate": 5.064417929696233e-06,
+      "loss": 0.5355,
+      "step": 4754
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5251633451194384,
+      "learning_rate": 5.062676997034494e-06,
+      "loss": 0.485,
+      "step": 4755
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6917017505709726,
+      "learning_rate": 5.060936056772952e-06,
+      "loss": 0.4786,
+      "step": 4756
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.15997160416424,
+      "learning_rate": 5.059195109122705e-06,
+      "loss": 0.5,
+      "step": 4757
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.099381119748845,
+      "learning_rate": 5.057454154294846e-06,
+      "loss": 0.4969,
+      "step": 4758
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3947945239029496,
+      "learning_rate": 5.055713192500472e-06,
+      "loss": 0.5019,
+      "step": 4759
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.596923359926309,
+      "learning_rate": 5.053972223950682e-06,
+      "loss": 0.4957,
+      "step": 4760
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.006864954315962,
+      "learning_rate": 5.052231248856574e-06,
+      "loss": 0.4871,
+      "step": 4761
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.773877394986138,
+      "learning_rate": 5.050490267429246e-06,
+      "loss": 0.4877,
+      "step": 4762
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.8300334660130675,
+      "learning_rate": 5.0487492798798e-06,
+      "loss": 0.5082,
+      "step": 4763
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.718524029843667,
+      "learning_rate": 5.047008286419336e-06,
+      "loss": 0.5096,
+      "step": 4764
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.260700665937473,
+      "learning_rate": 5.0452672872589535e-06,
+      "loss": 0.5318,
+      "step": 4765
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6380064264524576,
+      "learning_rate": 5.043526282609757e-06,
+      "loss": 0.5035,
+      "step": 4766
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.373589900277393,
+      "learning_rate": 5.04178527268285e-06,
+      "loss": 0.4766,
+      "step": 4767
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4010727464812,
+      "learning_rate": 5.040044257689334e-06,
+      "loss": 0.299,
+      "step": 4768
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.037410361007008,
+      "learning_rate": 5.038303237840314e-06,
+      "loss": 0.494,
+      "step": 4769
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.447929145160349,
+      "learning_rate": 5.036562213346892e-06,
+      "loss": 0.4924,
+      "step": 4770
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6541118309845966,
+      "learning_rate": 5.0348211844201786e-06,
+      "loss": 0.5018,
+      "step": 4771
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.162532099969498,
+      "learning_rate": 5.033080151271276e-06,
+      "loss": 0.4868,
+      "step": 4772
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.335339950161004,
+      "learning_rate": 5.03133911411129e-06,
+      "loss": 0.5165,
+      "step": 4773
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.570689143356796,
+      "learning_rate": 5.029598073151329e-06,
+      "loss": 0.4592,
+      "step": 4774
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.280968586794217,
+      "learning_rate": 5.0278570286025e-06,
+      "loss": 0.531,
+      "step": 4775
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.703274939325296,
+      "learning_rate": 5.0261159806759086e-06,
+      "loss": 0.5338,
+      "step": 4776
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4539650747531523,
+      "learning_rate": 5.024374929582665e-06,
+      "loss": 0.4816,
+      "step": 4777
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5173168346825374,
+      "learning_rate": 5.022633875533879e-06,
+      "loss": 0.5551,
+      "step": 4778
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4291206328835004,
+      "learning_rate": 5.020892818740655e-06,
+      "loss": 0.5102,
+      "step": 4779
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.9273161720194634,
+      "learning_rate": 5.019151759414107e-06,
+      "loss": 0.5151,
+      "step": 4780
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.00985803086776,
+      "learning_rate": 5.0174106977653415e-06,
+      "loss": 0.5196,
+      "step": 4781
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.151035535120201,
+      "learning_rate": 5.015669634005467e-06,
+      "loss": 0.4935,
+      "step": 4782
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.9325826360472496,
+      "learning_rate": 5.013928568345598e-06,
+      "loss": 0.4669,
+      "step": 4783
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.41001845367631207,
+      "learning_rate": 5.01218750099684e-06,
+      "loss": 0.2935,
+      "step": 4784
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.420582372173561,
+      "learning_rate": 5.010446432170306e-06,
+      "loss": 0.5163,
+      "step": 4785
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7447559499198633,
+      "learning_rate": 5.008705362077108e-06,
+      "loss": 0.5366,
+      "step": 4786
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.521563699377046,
+      "learning_rate": 5.006964290928351e-06,
+      "loss": 0.4546,
+      "step": 4787
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.188406657410041,
+      "learning_rate": 5.0052232189351524e-06,
+      "loss": 0.5439,
+      "step": 4788
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4372315411793237,
+      "learning_rate": 5.003482146308621e-06,
+      "loss": 0.4852,
+      "step": 4789
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.336376948992106,
+      "learning_rate": 5.001741073259866e-06,
+      "loss": 0.5001,
+      "step": 4790
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5266843007086806,
+      "learning_rate": 5e-06,
+      "loss": 0.4492,
+      "step": 4791
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.605359665961064,
+      "learning_rate": 4.998258926740135e-06,
+      "loss": 0.5064,
+      "step": 4792
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7340392245338156,
+      "learning_rate": 4.99651785369138e-06,
+      "loss": 0.4989,
+      "step": 4793
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1487417512320226,
+      "learning_rate": 4.9947767810648475e-06,
+      "loss": 0.4672,
+      "step": 4794
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.061777052504663,
+      "learning_rate": 4.993035709071649e-06,
+      "loss": 0.5102,
+      "step": 4795
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.74226052245792,
+      "learning_rate": 4.991294637922893e-06,
+      "loss": 0.5221,
+      "step": 4796
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.989818499879878,
+      "learning_rate": 4.989553567829696e-06,
+      "loss": 0.541,
+      "step": 4797
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.0825308443368975,
+      "learning_rate": 4.987812499003162e-06,
+      "loss": 0.4738,
+      "step": 4798
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.242018337952088,
+      "learning_rate": 4.986071431654405e-06,
+      "loss": 0.4813,
+      "step": 4799
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6907140560946483,
+      "learning_rate": 4.984330365994535e-06,
+      "loss": 0.5063,
+      "step": 4800
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.100822446443358,
+      "learning_rate": 4.98258930223466e-06,
+      "loss": 0.4767,
+      "step": 4801
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.0331026417822615,
+      "learning_rate": 4.980848240585895e-06,
+      "loss": 0.4831,
+      "step": 4802
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.632126812488932,
+      "learning_rate": 4.979107181259346e-06,
+      "loss": 0.5432,
+      "step": 4803
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.797701310884147,
+      "learning_rate": 4.977366124466122e-06,
+      "loss": 0.5169,
+      "step": 4804
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.013171262281086,
+      "learning_rate": 4.975625070417336e-06,
+      "loss": 0.498,
+      "step": 4805
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1643418380653974,
+      "learning_rate": 4.973884019324092e-06,
+      "loss": 0.5029,
+      "step": 4806
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3514133136187887,
+      "learning_rate": 4.972142971397504e-06,
+      "loss": 0.4649,
+      "step": 4807
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.41982597496069696,
+      "learning_rate": 4.970401926848674e-06,
+      "loss": 0.3071,
+      "step": 4808
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.605639959162163,
+      "learning_rate": 4.968660885888711e-06,
+      "loss": 0.4677,
+      "step": 4809
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.400010841284177,
+      "learning_rate": 4.966919848728726e-06,
+      "loss": 0.4835,
+      "step": 4810
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.772124082691743,
+      "learning_rate": 4.965178815579823e-06,
+      "loss": 0.5185,
+      "step": 4811
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.37024433098213166,
+      "learning_rate": 4.963437786653108e-06,
+      "loss": 0.2943,
+      "step": 4812
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.813865257961038,
+      "learning_rate": 4.961696762159688e-06,
+      "loss": 0.5333,
+      "step": 4813
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.400842342599736,
+      "learning_rate": 4.959955742310668e-06,
+      "loss": 0.4646,
+      "step": 4814
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0381554346487136,
+      "learning_rate": 4.958214727317152e-06,
+      "loss": 0.5424,
+      "step": 4815
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6738993039036276,
+      "learning_rate": 4.956473717390243e-06,
+      "loss": 0.5208,
+      "step": 4816
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.38441209552950156,
+      "learning_rate": 4.9547327127410465e-06,
+      "loss": 0.2891,
+      "step": 4817
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4099171757406357,
+      "learning_rate": 4.952991713580666e-06,
+      "loss": 0.2968,
+      "step": 4818
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.0575041072296125,
+      "learning_rate": 4.951250720120203e-06,
+      "loss": 0.5606,
+      "step": 4819
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.141961437650632,
+      "learning_rate": 4.949509732570756e-06,
+      "loss": 0.454,
+      "step": 4820
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3814310442984044,
+      "learning_rate": 4.947768751143428e-06,
+      "loss": 0.2927,
+      "step": 4821
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.747544228067299,
+      "learning_rate": 4.94602777604932e-06,
+      "loss": 0.513,
+      "step": 4822
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.832513649628059,
+      "learning_rate": 4.9442868074995296e-06,
+      "loss": 0.4561,
+      "step": 4823
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8536132229512403,
+      "learning_rate": 4.942545845705156e-06,
+      "loss": 0.4469,
+      "step": 4824
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3108353973774522,
+      "learning_rate": 4.9408048908772965e-06,
+      "loss": 0.4906,
+      "step": 4825
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.305601019711942,
+      "learning_rate": 4.939063943227049e-06,
+      "loss": 0.4693,
+      "step": 4826
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.725911199791081,
+      "learning_rate": 4.937323002965507e-06,
+      "loss": 0.4778,
+      "step": 4827
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.842084257601222,
+      "learning_rate": 4.935582070303767e-06,
+      "loss": 0.4469,
+      "step": 4828
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.823161005099549,
+      "learning_rate": 4.933841145452925e-06,
+      "loss": 0.5315,
+      "step": 4829
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5415782263261306,
+      "learning_rate": 4.932100228624072e-06,
+      "loss": 0.5143,
+      "step": 4830
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5963876453142394,
+      "learning_rate": 4.9303593200282995e-06,
+      "loss": 0.5009,
+      "step": 4831
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0034948348395583,
+      "learning_rate": 4.928618419876698e-06,
+      "loss": 0.5702,
+      "step": 4832
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.121547785085847,
+      "learning_rate": 4.926877528380362e-06,
+      "loss": 0.4795,
+      "step": 4833
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6500644015097023,
+      "learning_rate": 4.925136645750377e-06,
+      "loss": 0.4867,
+      "step": 4834
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.317939059215237,
+      "learning_rate": 4.923395772197829e-06,
+      "loss": 0.533,
+      "step": 4835
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.06647922694297,
+      "learning_rate": 4.9216549079338105e-06,
+      "loss": 0.5227,
+      "step": 4836
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.4518747354162125,
+      "learning_rate": 4.919914053169404e-06,
+      "loss": 0.418,
+      "step": 4837
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8091446591005838,
+      "learning_rate": 4.918173208115695e-06,
+      "loss": 0.5194,
+      "step": 4838
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3635619455338364,
+      "learning_rate": 4.916432372983768e-06,
+      "loss": 0.4654,
+      "step": 4839
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7688478681812905,
+      "learning_rate": 4.914691547984701e-06,
+      "loss": 0.5099,
+      "step": 4840
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.401396577196492,
+      "learning_rate": 4.912950733329579e-06,
+      "loss": 0.2747,
+      "step": 4841
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.061974680577852,
+      "learning_rate": 4.91120992922948e-06,
+      "loss": 0.5527,
+      "step": 4842
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.41961367032056734,
+      "learning_rate": 4.909469135895485e-06,
+      "loss": 0.2733,
+      "step": 4843
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0176214295261534,
+      "learning_rate": 4.9077283535386675e-06,
+      "loss": 0.4887,
+      "step": 4844
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.042184027269588,
+      "learning_rate": 4.905987582370106e-06,
+      "loss": 0.4909,
+      "step": 4845
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.571127946152801,
+      "learning_rate": 4.904246822600874e-06,
+      "loss": 0.5137,
+      "step": 4846
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4759629147604305,
+      "learning_rate": 4.902506074442044e-06,
+      "loss": 0.4771,
+      "step": 4847
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8421658160850836,
+      "learning_rate": 4.900765338104691e-06,
+      "loss": 0.5252,
+      "step": 4848
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.3882107979453835,
+      "learning_rate": 4.899024613799881e-06,
+      "loss": 0.4854,
+      "step": 4849
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.8466051840826765,
+      "learning_rate": 4.897283901738687e-06,
+      "loss": 0.5454,
+      "step": 4850
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.723457416797036,
+      "learning_rate": 4.8955432021321725e-06,
+      "loss": 0.4915,
+      "step": 4851
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.217524930301587,
+      "learning_rate": 4.893802515191404e-06,
+      "loss": 0.4755,
+      "step": 4852
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.610416914277798,
+      "learning_rate": 4.892061841127446e-06,
+      "loss": 0.5547,
+      "step": 4853
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9015190457086715,
+      "learning_rate": 4.8903211801513645e-06,
+      "loss": 0.4718,
+      "step": 4854
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.42132987155323653,
+      "learning_rate": 4.888580532474216e-06,
+      "loss": 0.2915,
+      "step": 4855
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.203607884521121,
+      "learning_rate": 4.886839898307062e-06,
+      "loss": 0.5762,
+      "step": 4856
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.992418472009365,
+      "learning_rate": 4.885099277860962e-06,
+      "loss": 0.4697,
+      "step": 4857
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9030604884515143,
+      "learning_rate": 4.883358671346969e-06,
+      "loss": 0.491,
+      "step": 4858
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5891705554166893,
+      "learning_rate": 4.88161807897614e-06,
+      "loss": 0.4769,
+      "step": 4859
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.50389840193508,
+      "learning_rate": 4.879877500959529e-06,
+      "loss": 0.5468,
+      "step": 4860
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7218169253151645,
+      "learning_rate": 4.878136937508183e-06,
+      "loss": 0.4743,
+      "step": 4861
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1640368180223652,
+      "learning_rate": 4.876396388833155e-06,
+      "loss": 0.4987,
+      "step": 4862
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.167024304877624,
+      "learning_rate": 4.874655855145488e-06,
+      "loss": 0.5362,
+      "step": 4863
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.9615729652221563,
+      "learning_rate": 4.872915336656233e-06,
+      "loss": 0.5048,
+      "step": 4864
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8394262613255186,
+      "learning_rate": 4.871174833576431e-06,
+      "loss": 0.538,
+      "step": 4865
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.4918427365143865,
+      "learning_rate": 4.869434346117124e-06,
+      "loss": 0.4737,
+      "step": 4866
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.95325885680188,
+      "learning_rate": 4.867693874489353e-06,
+      "loss": 0.438,
+      "step": 4867
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.500318963804961,
+      "learning_rate": 4.865953418904156e-06,
+      "loss": 0.5002,
+      "step": 4868
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.10994614884598,
+      "learning_rate": 4.864212979572569e-06,
+      "loss": 0.5001,
+      "step": 4869
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.777232549312106,
+      "learning_rate": 4.862472556705626e-06,
+      "loss": 0.5466,
+      "step": 4870
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5331573900464956,
+      "learning_rate": 4.8607321505143614e-06,
+      "loss": 0.4849,
+      "step": 4871
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.741251860477148,
+      "learning_rate": 4.858991761209803e-06,
+      "loss": 0.5296,
+      "step": 4872
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.076244243886272,
+      "learning_rate": 4.85725138900298e-06,
+      "loss": 0.5046,
+      "step": 4873
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1092695328053037,
+      "learning_rate": 4.855511034104916e-06,
+      "loss": 0.4836,
+      "step": 4874
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.241368065828665,
+      "learning_rate": 4.853770696726639e-06,
+      "loss": 0.5512,
+      "step": 4875
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.897362123707919,
+      "learning_rate": 4.852030377079169e-06,
+      "loss": 0.4991,
+      "step": 4876
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4301043833641474,
+      "learning_rate": 4.850290075373525e-06,
+      "loss": 0.2678,
+      "step": 4877
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.149055811837718,
+      "learning_rate": 4.848549791820725e-06,
+      "loss": 0.5001,
+      "step": 4878
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.7451772658519644,
+      "learning_rate": 4.846809526631787e-06,
+      "loss": 0.5099,
+      "step": 4879
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.72045889150617,
+      "learning_rate": 4.845069280017721e-06,
+      "loss": 0.5106,
+      "step": 4880
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.897643687263991,
+      "learning_rate": 4.843329052189538e-06,
+      "loss": 0.4965,
+      "step": 4881
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.460285971980085,
+      "learning_rate": 4.841588843358251e-06,
+      "loss": 0.5095,
+      "step": 4882
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.68544808622292,
+      "learning_rate": 4.839848653734859e-06,
+      "loss": 0.5396,
+      "step": 4883
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8504872589718095,
+      "learning_rate": 4.838108483530371e-06,
+      "loss": 0.5112,
+      "step": 4884
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.003176614045116,
+      "learning_rate": 4.8363683329557885e-06,
+      "loss": 0.4885,
+      "step": 4885
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.921374813069386,
+      "learning_rate": 4.834628202222107e-06,
+      "loss": 0.4576,
+      "step": 4886
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.714320912591947,
+      "learning_rate": 4.832888091540328e-06,
+      "loss": 0.4986,
+      "step": 4887
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.090880494006239,
+      "learning_rate": 4.831148001121445e-06,
+      "loss": 0.5058,
+      "step": 4888
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.834701164060588,
+      "learning_rate": 4.829407931176447e-06,
+      "loss": 0.4903,
+      "step": 4889
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.96210203473085,
+      "learning_rate": 4.827667881916327e-06,
+      "loss": 0.5364,
+      "step": 4890
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0408935504882613,
+      "learning_rate": 4.825927853552071e-06,
+      "loss": 0.4885,
+      "step": 4891
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.036703654155283,
+      "learning_rate": 4.824187846294662e-06,
+      "loss": 0.4789,
+      "step": 4892
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6111039144403705,
+      "learning_rate": 4.822447860355083e-06,
+      "loss": 0.4978,
+      "step": 4893
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7980057833268797,
+      "learning_rate": 4.820707895944312e-06,
+      "loss": 0.5058,
+      "step": 4894
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.586587906260997,
+      "learning_rate": 4.818967953273327e-06,
+      "loss": 0.4997,
+      "step": 4895
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1824349789660276,
+      "learning_rate": 4.817228032553103e-06,
+      "loss": 0.5052,
+      "step": 4896
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0967268489719872,
+      "learning_rate": 4.815488133994608e-06,
+      "loss": 0.4871,
+      "step": 4897
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.653591399502859,
+      "learning_rate": 4.813748257808814e-06,
+      "loss": 0.4585,
+      "step": 4898
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.73778175666179,
+      "learning_rate": 4.8120084042066865e-06,
+      "loss": 0.5056,
+      "step": 4899
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6348522467415396,
+      "learning_rate": 4.810268573399187e-06,
+      "loss": 0.4772,
+      "step": 4900
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1113157189991587,
+      "learning_rate": 4.8085287655972775e-06,
+      "loss": 0.4811,
+      "step": 4901
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3119355638413763,
+      "learning_rate": 4.806788981011916e-06,
+      "loss": 0.4895,
+      "step": 4902
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.39800825199728196,
+      "learning_rate": 4.805049219854058e-06,
+      "loss": 0.2967,
+      "step": 4903
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.50197798521418,
+      "learning_rate": 4.803309482334652e-06,
+      "loss": 0.52,
+      "step": 4904
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0148261951768096,
+      "learning_rate": 4.801569768664649e-06,
+      "loss": 0.4585,
+      "step": 4905
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6821059839921872,
+      "learning_rate": 4.799830079054996e-06,
+      "loss": 0.5062,
+      "step": 4906
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.927652310990686,
+      "learning_rate": 4.798090413716637e-06,
+      "loss": 0.5317,
+      "step": 4907
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.151559732279038,
+      "learning_rate": 4.796350772860511e-06,
+      "loss": 0.5267,
+      "step": 4908
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.08862616307918,
+      "learning_rate": 4.7946111566975545e-06,
+      "loss": 0.542,
+      "step": 4909
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.885820838068632,
+      "learning_rate": 4.792871565438705e-06,
+      "loss": 0.5166,
+      "step": 4910
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5757665325351984,
+      "learning_rate": 4.7911319992948914e-06,
+      "loss": 0.4586,
+      "step": 4911
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.633789985575929,
+      "learning_rate": 4.789392458477042e-06,
+      "loss": 0.5081,
+      "step": 4912
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.7558843415090735,
+      "learning_rate": 4.787652943196087e-06,
+      "loss": 0.4626,
+      "step": 4913
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9139301991364692,
+      "learning_rate": 4.785913453662941e-06,
+      "loss": 0.4887,
+      "step": 4914
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2493759086733127,
+      "learning_rate": 4.784173990088529e-06,
+      "loss": 0.4601,
+      "step": 4915
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.713078577588637,
+      "learning_rate": 4.782434552683763e-06,
+      "loss": 0.4671,
+      "step": 4916
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.288134769107871,
+      "learning_rate": 4.780695141659557e-06,
+      "loss": 0.5302,
+      "step": 4917
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.099672340288758,
+      "learning_rate": 4.778955757226823e-06,
+      "loss": 0.5109,
+      "step": 4918
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.0101116752004335,
+      "learning_rate": 4.777216399596466e-06,
+      "loss": 0.5295,
+      "step": 4919
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.037830613613973,
+      "learning_rate": 4.775477068979386e-06,
+      "loss": 0.4931,
+      "step": 4920
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.39353868231399985,
+      "learning_rate": 4.773737765586487e-06,
+      "loss": 0.278,
+      "step": 4921
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5203732711450297,
+      "learning_rate": 4.771998489628664e-06,
+      "loss": 0.4773,
+      "step": 4922
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.70548482288412,
+      "learning_rate": 4.7702592413168095e-06,
+      "loss": 0.5036,
+      "step": 4923
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.381135705092849,
+      "learning_rate": 4.768520020861817e-06,
+      "loss": 0.5549,
+      "step": 4924
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3584556579352713,
+      "learning_rate": 4.766780828474566e-06,
+      "loss": 0.2746,
+      "step": 4925
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.546392466460975,
+      "learning_rate": 4.765041664365945e-06,
+      "loss": 0.482,
+      "step": 4926
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4598859660469711,
+      "learning_rate": 4.763302528746833e-06,
+      "loss": 0.3038,
+      "step": 4927
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.048771596030454,
+      "learning_rate": 4.761563421828104e-06,
+      "loss": 0.4836,
+      "step": 4928
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3635408906959006,
+      "learning_rate": 4.759824343820632e-06,
+      "loss": 0.5544,
+      "step": 4929
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.360489982532305,
+      "learning_rate": 4.758085294935288e-06,
+      "loss": 0.4843,
+      "step": 4930
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.435056648515827,
+      "learning_rate": 4.756346275382934e-06,
+      "loss": 0.5471,
+      "step": 4931
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.269711199358357,
+      "learning_rate": 4.754607285374435e-06,
+      "loss": 0.5124,
+      "step": 4932
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.947876169240694,
+      "learning_rate": 4.75286832512065e-06,
+      "loss": 0.4995,
+      "step": 4933
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.448620634341447,
+      "learning_rate": 4.7511293948324325e-06,
+      "loss": 0.5031,
+      "step": 4934
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2525034022374406,
+      "learning_rate": 4.749390494720633e-06,
+      "loss": 0.5113,
+      "step": 4935
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.199863349418056,
+      "learning_rate": 4.7476516249961e-06,
+      "loss": 0.4723,
+      "step": 4936
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4054597476752855,
+      "learning_rate": 4.7459127858696765e-06,
+      "loss": 0.2847,
+      "step": 4937
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.96582917101424,
+      "learning_rate": 4.744173977552205e-06,
+      "loss": 0.5088,
+      "step": 4938
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.539170882843725,
+      "learning_rate": 4.74243520025452e-06,
+      "loss": 0.5275,
+      "step": 4939
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.583768331243209,
+      "learning_rate": 4.7406964541874546e-06,
+      "loss": 0.4774,
+      "step": 4940
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4722967321053835,
+      "learning_rate": 4.7389577395618395e-06,
+      "loss": 0.4938,
+      "step": 4941
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.470944415417504,
+      "learning_rate": 4.737219056588497e-06,
+      "loss": 0.4908,
+      "step": 4942
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.898269569103496,
+      "learning_rate": 4.735480405478249e-06,
+      "loss": 0.528,
+      "step": 4943
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.224588903386021,
+      "learning_rate": 4.733741786441916e-06,
+      "loss": 0.5238,
+      "step": 4944
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0134917694961523,
+      "learning_rate": 4.73200319969031e-06,
+      "loss": 0.5286,
+      "step": 4945
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.701719821185734,
+      "learning_rate": 4.730264645434238e-06,
+      "loss": 0.4743,
+      "step": 4946
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.019840035016248,
+      "learning_rate": 4.728526123884509e-06,
+      "loss": 0.5118,
+      "step": 4947
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.624191546383508,
+      "learning_rate": 4.7267876352519215e-06,
+      "loss": 0.5082,
+      "step": 4948
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.812816659890102,
+      "learning_rate": 4.725049179747276e-06,
+      "loss": 0.5107,
+      "step": 4949
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.81686811028975,
+      "learning_rate": 4.723310757581365e-06,
+      "loss": 0.4771,
+      "step": 4950
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.40972420413648636,
+      "learning_rate": 4.721572368964979e-06,
+      "loss": 0.2941,
+      "step": 4951
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.885764793532447,
+      "learning_rate": 4.719834014108903e-06,
+      "loss": 0.5006,
+      "step": 4952
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6195431834616425,
+      "learning_rate": 4.718095693223919e-06,
+      "loss": 0.4803,
+      "step": 4953
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.63416461009706,
+      "learning_rate": 4.7163574065208034e-06,
+      "loss": 0.4558,
+      "step": 4954
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.422534402066523,
+      "learning_rate": 4.7146191542103305e-06,
+      "loss": 0.4953,
+      "step": 4955
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1161483418036497,
+      "learning_rate": 4.712880936503271e-06,
+      "loss": 0.522,
+      "step": 4956
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.851038080850313,
+      "learning_rate": 4.7111427536103856e-06,
+      "loss": 0.5066,
+      "step": 4957
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3838980642300952,
+      "learning_rate": 4.709404605742438e-06,
+      "loss": 0.2843,
+      "step": 4958
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.145307548847913,
+      "learning_rate": 4.707666493110182e-06,
+      "loss": 0.5351,
+      "step": 4959
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.254533714173447,
+      "learning_rate": 4.7059284159243725e-06,
+      "loss": 0.4758,
+      "step": 4960
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.162185779407232,
+      "learning_rate": 4.704190374395757e-06,
+      "loss": 0.4777,
+      "step": 4961
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.767952456383832,
+      "learning_rate": 4.7024523687350775e-06,
+      "loss": 0.5339,
+      "step": 4962
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4807186108570103,
+      "learning_rate": 4.700714399153075e-06,
+      "loss": 0.4692,
+      "step": 4963
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3341361560230776,
+      "learning_rate": 4.698976465860484e-06,
+      "loss": 0.4766,
+      "step": 4964
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4592477784478883,
+      "learning_rate": 4.697238569068033e-06,
+      "loss": 0.5024,
+      "step": 4965
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6349144530289568,
+      "learning_rate": 4.695500708986451e-06,
+      "loss": 0.4499,
+      "step": 4966
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7696462016270766,
+      "learning_rate": 4.693762885826456e-06,
+      "loss": 0.5291,
+      "step": 4967
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1953639170196633,
+      "learning_rate": 4.692025099798767e-06,
+      "loss": 0.4663,
+      "step": 4968
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.976398346145285,
+      "learning_rate": 4.690287351114097e-06,
+      "loss": 0.5334,
+      "step": 4969
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.39371569736558654,
+      "learning_rate": 4.688549639983154e-06,
+      "loss": 0.2928,
+      "step": 4970
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.120212211349781,
+      "learning_rate": 4.686811966616639e-06,
+      "loss": 0.5186,
+      "step": 4971
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6114246269863135,
+      "learning_rate": 4.685074331225254e-06,
+      "loss": 0.4651,
+      "step": 4972
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.032149378832427,
+      "learning_rate": 4.683336734019692e-06,
+      "loss": 0.4983,
+      "step": 4973
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7807483472920667,
+      "learning_rate": 4.681599175210641e-06,
+      "loss": 0.5514,
+      "step": 4974
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 26.91613788237094,
+      "learning_rate": 4.67986165500879e-06,
+      "loss": 0.4556,
+      "step": 4975
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.797408379109964,
+      "learning_rate": 4.678124173624816e-06,
+      "loss": 0.4795,
+      "step": 4976
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.8654428362911983,
+      "learning_rate": 4.676386731269398e-06,
+      "loss": 0.5138,
+      "step": 4977
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.862755423228541,
+      "learning_rate": 4.674649328153202e-06,
+      "loss": 0.5011,
+      "step": 4978
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2934869031655993,
+      "learning_rate": 4.672911964486896e-06,
+      "loss": 0.4846,
+      "step": 4979
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.162659860473806,
+      "learning_rate": 4.671174640481143e-06,
+      "loss": 0.5124,
+      "step": 4980
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 25.66884969851148,
+      "learning_rate": 4.669437356346599e-06,
+      "loss": 0.5034,
+      "step": 4981
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0909435303028614,
+      "learning_rate": 4.6677001122939134e-06,
+      "loss": 0.4936,
+      "step": 4982
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.808002953137469,
+      "learning_rate": 4.665962908533736e-06,
+      "loss": 0.4795,
+      "step": 4983
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.573844891382278,
+      "learning_rate": 4.664225745276709e-06,
+      "loss": 0.4704,
+      "step": 4984
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.4038844874985905,
+      "learning_rate": 4.662488622733466e-06,
+      "loss": 0.4663,
+      "step": 4985
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5961288905869013,
+      "learning_rate": 4.660751541114641e-06,
+      "loss": 0.5595,
+      "step": 4986
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.168308669585231,
+      "learning_rate": 4.659014500630863e-06,
+      "loss": 0.4753,
+      "step": 4987
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.559839477380135,
+      "learning_rate": 4.657277501492751e-06,
+      "loss": 0.5238,
+      "step": 4988
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5654850430911864,
+      "learning_rate": 4.655540543910924e-06,
+      "loss": 0.4699,
+      "step": 4989
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.026659420568941,
+      "learning_rate": 4.653803628095993e-06,
+      "loss": 0.5139,
+      "step": 4990
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0828465574412336,
+      "learning_rate": 4.6520667542585655e-06,
+      "loss": 0.477,
+      "step": 4991
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.39838064729522,
+      "learning_rate": 4.650329922609244e-06,
+      "loss": 0.4899,
+      "step": 4992
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.58783306391612,
+      "learning_rate": 4.648593133358625e-06,
+      "loss": 0.4843,
+      "step": 4993
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5871214844401567,
+      "learning_rate": 4.646856386717299e-06,
+      "loss": 0.4943,
+      "step": 4994
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.426415071421818,
+      "learning_rate": 4.645119682895855e-06,
+      "loss": 0.473,
+      "step": 4995
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2340000735328447,
+      "learning_rate": 4.643383022104871e-06,
+      "loss": 0.4426,
+      "step": 4996
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.098333109928701,
+      "learning_rate": 4.641646404554927e-06,
+      "loss": 0.4688,
+      "step": 4997
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.66252645533234,
+      "learning_rate": 4.639909830456593e-06,
+      "loss": 0.5224,
+      "step": 4998
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.27676460027856,
+      "learning_rate": 4.638173300020431e-06,
+      "loss": 0.4894,
+      "step": 4999
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0563484181418357,
+      "learning_rate": 4.6364368134570046e-06,
+      "loss": 0.4985,
+      "step": 5000
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6022887464535183,
+      "learning_rate": 4.634700370976867e-06,
+      "loss": 0.4958,
+      "step": 5001
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.477094947308327,
+      "learning_rate": 4.63296397279057e-06,
+      "loss": 0.4627,
+      "step": 5002
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.124793595051956,
+      "learning_rate": 4.631227619108657e-06,
+      "loss": 0.5025,
+      "step": 5003
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.669476797308278,
+      "learning_rate": 4.629491310141666e-06,
+      "loss": 0.5143,
+      "step": 5004
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6737998683068755,
+      "learning_rate": 4.62775504610013e-06,
+      "loss": 0.5182,
+      "step": 5005
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8983908208566476,
+      "learning_rate": 4.626018827194579e-06,
+      "loss": 0.4909,
+      "step": 5006
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.14691542356229,
+      "learning_rate": 4.624282653635534e-06,
+      "loss": 0.4782,
+      "step": 5007
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2021410614689083,
+      "learning_rate": 4.622546525633512e-06,
+      "loss": 0.484,
+      "step": 5008
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.639542454460927,
+      "learning_rate": 4.6208104433990285e-06,
+      "loss": 0.4602,
+      "step": 5009
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8631253332872997,
+      "learning_rate": 4.619074407142582e-06,
+      "loss": 0.5027,
+      "step": 5010
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8596358024137456,
+      "learning_rate": 4.617338417074679e-06,
+      "loss": 0.4875,
+      "step": 5011
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3955992348529873,
+      "learning_rate": 4.615602473405812e-06,
+      "loss": 0.2991,
+      "step": 5012
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.436494475214874,
+      "learning_rate": 4.6138665763464685e-06,
+      "loss": 0.4764,
+      "step": 5013
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.149904524021902,
+      "learning_rate": 4.612130726107135e-06,
+      "loss": 0.4953,
+      "step": 5014
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.37923830742291975,
+      "learning_rate": 4.610394922898289e-06,
+      "loss": 0.3075,
+      "step": 5015
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4783311700584187,
+      "learning_rate": 4.6086591669303995e-06,
+      "loss": 0.4938,
+      "step": 5016
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7526024192191296,
+      "learning_rate": 4.6069234584139374e-06,
+      "loss": 0.5274,
+      "step": 5017
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.793010055725422,
+      "learning_rate": 4.60518779755936e-06,
+      "loss": 0.4617,
+      "step": 5018
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.577841266440298,
+      "learning_rate": 4.603452184577126e-06,
+      "loss": 0.4388,
+      "step": 5019
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.7440079548245504,
+      "learning_rate": 4.6017166196776795e-06,
+      "loss": 0.5274,
+      "step": 5020
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.719326027122517,
+      "learning_rate": 4.599981103071465e-06,
+      "loss": 0.4895,
+      "step": 5021
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.800661011177367,
+      "learning_rate": 4.5982456349689214e-06,
+      "loss": 0.5161,
+      "step": 5022
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.043341963959627,
+      "learning_rate": 4.59651021558048e-06,
+      "loss": 0.4667,
+      "step": 5023
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5158271488718693,
+      "learning_rate": 4.594774845116565e-06,
+      "loss": 0.487,
+      "step": 5024
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.1706700944238815,
+      "learning_rate": 4.5930395237875986e-06,
+      "loss": 0.5152,
+      "step": 5025
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.794549479202592,
+      "learning_rate": 4.5913042518039925e-06,
+      "loss": 0.5237,
+      "step": 5026
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7326931545386053,
+      "learning_rate": 4.589569029376153e-06,
+      "loss": 0.5267,
+      "step": 5027
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.910166501115604,
+      "learning_rate": 4.587833856714486e-06,
+      "loss": 0.4701,
+      "step": 5028
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.12397370289469,
+      "learning_rate": 4.586098734029384e-06,
+      "loss": 0.4932,
+      "step": 5029
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.014449296976718,
+      "learning_rate": 4.5843636615312395e-06,
+      "loss": 0.467,
+      "step": 5030
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.202501034216289,
+      "learning_rate": 4.582628639430432e-06,
+      "loss": 0.4726,
+      "step": 5031
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.705159019532914,
+      "learning_rate": 4.58089366793734e-06,
+      "loss": 0.5784,
+      "step": 5032
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4713077080009442,
+      "learning_rate": 4.579158747262337e-06,
+      "loss": 0.2934,
+      "step": 5033
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.141956451778381,
+      "learning_rate": 4.577423877615787e-06,
+      "loss": 0.5282,
+      "step": 5034
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.75312659798355,
+      "learning_rate": 4.575689059208048e-06,
+      "loss": 0.4895,
+      "step": 5035
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5633490913503394,
+      "learning_rate": 4.573954292249472e-06,
+      "loss": 0.5199,
+      "step": 5036
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.693500914592683,
+      "learning_rate": 4.572219576950408e-06,
+      "loss": 0.5216,
+      "step": 5037
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.093475051726135,
+      "learning_rate": 4.570484913521196e-06,
+      "loss": 0.4747,
+      "step": 5038
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8594881625232915,
+      "learning_rate": 4.568750302172168e-06,
+      "loss": 0.5051,
+      "step": 5039
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.40398266959932755,
+      "learning_rate": 4.5670157431136546e-06,
+      "loss": 0.2849,
+      "step": 5040
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.1835420536785755,
+      "learning_rate": 4.565281236555973e-06,
+      "loss": 0.4899,
+      "step": 5041
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.482778388635551,
+      "learning_rate": 4.56354678270944e-06,
+      "loss": 0.4497,
+      "step": 5042
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7711175154399883,
+      "learning_rate": 4.561812381784366e-06,
+      "loss": 0.5194,
+      "step": 5043
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0941290876635934,
+      "learning_rate": 4.560078033991049e-06,
+      "loss": 0.4584,
+      "step": 5044
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9086766320096578,
+      "learning_rate": 4.558343739539788e-06,
+      "loss": 0.5077,
+      "step": 5045
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3069152144968026,
+      "learning_rate": 4.556609498640872e-06,
+      "loss": 0.4829,
+      "step": 5046
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.334220710406296,
+      "learning_rate": 4.554875311504581e-06,
+      "loss": 0.4818,
+      "step": 5047
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.751501845098622,
+      "learning_rate": 4.553141178341195e-06,
+      "loss": 0.5406,
+      "step": 5048
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3380440703385412,
+      "learning_rate": 4.551407099360981e-06,
+      "loss": 0.5628,
+      "step": 5049
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8558664656202803,
+      "learning_rate": 4.549673074774203e-06,
+      "loss": 0.5205,
+      "step": 5050
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.41937484663264907,
+      "learning_rate": 4.547939104791118e-06,
+      "loss": 0.3038,
+      "step": 5051
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.8712248774576365,
+      "learning_rate": 4.546205189621974e-06,
+      "loss": 0.5029,
+      "step": 5052
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.974935074593861,
+      "learning_rate": 4.544471329477015e-06,
+      "loss": 0.5163,
+      "step": 5053
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.071255814711369,
+      "learning_rate": 4.542737524566479e-06,
+      "loss": 0.4743,
+      "step": 5054
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0185743496526545,
+      "learning_rate": 4.541003775100592e-06,
+      "loss": 0.4655,
+      "step": 5055
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.559405171924411,
+      "learning_rate": 4.539270081289581e-06,
+      "loss": 0.4623,
+      "step": 5056
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4610841724910397,
+      "learning_rate": 4.537536443343662e-06,
+      "loss": 0.5287,
+      "step": 5057
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3862235916493495,
+      "learning_rate": 4.5358028614730424e-06,
+      "loss": 0.2896,
+      "step": 5058
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3476879903716994,
+      "learning_rate": 4.534069335887927e-06,
+      "loss": 0.4828,
+      "step": 5059
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.687373363293724,
+      "learning_rate": 4.53233586679851e-06,
+      "loss": 0.477,
+      "step": 5060
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.591220438763525,
+      "learning_rate": 4.530602454414983e-06,
+      "loss": 0.4783,
+      "step": 5061
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.406277210563149,
+      "learning_rate": 4.528869098947526e-06,
+      "loss": 0.5194,
+      "step": 5062
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.9330238037450256,
+      "learning_rate": 4.527135800606314e-06,
+      "loss": 0.5167,
+      "step": 5063
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.3793860389019335,
+      "learning_rate": 4.5254025596015175e-06,
+      "loss": 0.5324,
+      "step": 5064
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.452871721215655,
+      "learning_rate": 4.523669376143296e-06,
+      "loss": 0.4911,
+      "step": 5065
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.658703245706444,
+      "learning_rate": 4.5219362504418055e-06,
+      "loss": 0.4898,
+      "step": 5066
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.7970097240691936,
+      "learning_rate": 4.5202031827071915e-06,
+      "loss": 0.47,
+      "step": 5067
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3383020065094007,
+      "learning_rate": 4.518470173149597e-06,
+      "loss": 0.4649,
+      "step": 5068
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5914482029605757,
+      "learning_rate": 4.5167372219791545e-06,
+      "loss": 0.4927,
+      "step": 5069
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.02484988106579,
+      "learning_rate": 4.515004329405988e-06,
+      "loss": 0.5092,
+      "step": 5070
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.894245724953331,
+      "learning_rate": 4.51327149564022e-06,
+      "loss": 0.5094,
+      "step": 5071
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.249234736893511,
+      "learning_rate": 4.511538720891963e-06,
+      "loss": 0.4885,
+      "step": 5072
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.40186483142570145,
+      "learning_rate": 4.509806005371317e-06,
+      "loss": 0.2836,
+      "step": 5073
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.162826708452556,
+      "learning_rate": 4.508073349288384e-06,
+      "loss": 0.5127,
+      "step": 5074
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.234368146303347,
+      "learning_rate": 4.506340752853253e-06,
+      "loss": 0.4789,
+      "step": 5075
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8573890466715164,
+      "learning_rate": 4.5046082162760076e-06,
+      "loss": 0.5396,
+      "step": 5076
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6696336356346433,
+      "learning_rate": 4.502875739766725e-06,
+      "loss": 0.5065,
+      "step": 5077
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9628011295668415,
+      "learning_rate": 4.50114332353547e-06,
+      "loss": 0.5202,
+      "step": 5078
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6066184499986953,
+      "learning_rate": 4.499410967792309e-06,
+      "loss": 0.4981,
+      "step": 5079
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8896740984818616,
+      "learning_rate": 4.497678672747294e-06,
+      "loss": 0.4987,
+      "step": 5080
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6606871307816333,
+      "learning_rate": 4.49594643861047e-06,
+      "loss": 0.4806,
+      "step": 5081
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8690955977155785,
+      "learning_rate": 4.49421426559188e-06,
+      "loss": 0.4323,
+      "step": 5082
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.73072896318558,
+      "learning_rate": 4.492482153901554e-06,
+      "loss": 0.4845,
+      "step": 5083
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9680788355216796,
+      "learning_rate": 4.490750103749516e-06,
+      "loss": 0.4924,
+      "step": 5084
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.994551088421462,
+      "learning_rate": 4.489018115345784e-06,
+      "loss": 0.5019,
+      "step": 5085
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3950397934547674,
+      "learning_rate": 4.487286188900365e-06,
+      "loss": 0.3033,
+      "step": 5086
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.741062724722784,
+      "learning_rate": 4.485554324623266e-06,
+      "loss": 0.4757,
+      "step": 5087
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.236438755961788,
+      "learning_rate": 4.483822522724477e-06,
+      "loss": 0.4959,
+      "step": 5088
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4413073730847703,
+      "learning_rate": 4.4820907834139865e-06,
+      "loss": 0.4825,
+      "step": 5089
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.40965055261379807,
+      "learning_rate": 4.480359106901775e-06,
+      "loss": 0.2866,
+      "step": 5090
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4434840665113104,
+      "learning_rate": 4.478627493397813e-06,
+      "loss": 0.5109,
+      "step": 5091
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8442548118800834,
+      "learning_rate": 4.476895943112064e-06,
+      "loss": 0.5007,
+      "step": 5092
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.381767659165573,
+      "learning_rate": 4.475164456254488e-06,
+      "loss": 0.4907,
+      "step": 5093
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3682508738099978,
+      "learning_rate": 4.473433033035028e-06,
+      "loss": 0.2838,
+      "step": 5094
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.843084827167982,
+      "learning_rate": 4.4717016736636295e-06,
+      "loss": 0.5462,
+      "step": 5095
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.367448980604217,
+      "learning_rate": 4.469970378350225e-06,
+      "loss": 0.5671,
+      "step": 5096
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4057948811963882,
+      "learning_rate": 4.468239147304737e-06,
+      "loss": 0.2787,
+      "step": 5097
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9288101027737934,
+      "learning_rate": 4.466507980737087e-06,
+      "loss": 0.5082,
+      "step": 5098
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.011980943835008,
+      "learning_rate": 4.464776878857184e-06,
+      "loss": 0.5354,
+      "step": 5099
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6581281686544203,
+      "learning_rate": 4.46304584187493e-06,
+      "loss": 0.4826,
+      "step": 5100
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.148263776541828,
+      "learning_rate": 4.461314870000218e-06,
+      "loss": 0.5127,
+      "step": 5101
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1200925620444595,
+      "learning_rate": 4.459583963442935e-06,
+      "loss": 0.5085,
+      "step": 5102
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6065429226055543,
+      "learning_rate": 4.45785312241296e-06,
+      "loss": 0.5083,
+      "step": 5103
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 19.68106016900626,
+      "learning_rate": 4.456122347120165e-06,
+      "loss": 0.4435,
+      "step": 5104
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.425816934685371,
+      "learning_rate": 4.454391637774408e-06,
+      "loss": 0.4549,
+      "step": 5105
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.163029003771076,
+      "learning_rate": 4.452660994585546e-06,
+      "loss": 0.4591,
+      "step": 5106
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.578978353605464,
+      "learning_rate": 4.450930417763425e-06,
+      "loss": 0.4798,
+      "step": 5107
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.323409516307366,
+      "learning_rate": 4.449199907517885e-06,
+      "loss": 0.4666,
+      "step": 5108
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2487432035370816,
+      "learning_rate": 4.447469464058753e-06,
+      "loss": 0.472,
+      "step": 5109
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.313566767706938,
+      "learning_rate": 4.445739087595855e-06,
+      "loss": 0.4565,
+      "step": 5110
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.231203011755798,
+      "learning_rate": 4.444008778339003e-06,
+      "loss": 0.5623,
+      "step": 5111
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5987460916314293,
+      "learning_rate": 4.4422785364980036e-06,
+      "loss": 0.5535,
+      "step": 5112
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.30554195185206,
+      "learning_rate": 4.440548362282655e-06,
+      "loss": 0.4779,
+      "step": 5113
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.663561978303976,
+      "learning_rate": 4.438818255902746e-06,
+      "loss": 0.4662,
+      "step": 5114
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8173364907326106,
+      "learning_rate": 4.437088217568059e-06,
+      "loss": 0.5025,
+      "step": 5115
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.060358774153265,
+      "learning_rate": 4.435358247488365e-06,
+      "loss": 0.4725,
+      "step": 5116
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.014788817966356,
+      "learning_rate": 4.433628345873429e-06,
+      "loss": 0.5149,
+      "step": 5117
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.093527903203639,
+      "learning_rate": 4.431898512933011e-06,
+      "loss": 0.5063,
+      "step": 5118
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8759292850097387,
+      "learning_rate": 4.430168748876855e-06,
+      "loss": 0.5154,
+      "step": 5119
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3206541345915634,
+      "learning_rate": 4.428439053914702e-06,
+      "loss": 0.5072,
+      "step": 5120
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7581682643358363,
+      "learning_rate": 4.426709428256287e-06,
+      "loss": 0.5103,
+      "step": 5121
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.168449813027306,
+      "learning_rate": 4.424979872111329e-06,
+      "loss": 0.4467,
+      "step": 5122
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.762492765698035,
+      "learning_rate": 4.423250385689543e-06,
+      "loss": 0.5594,
+      "step": 5123
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7586174982430283,
+      "learning_rate": 4.421520969200637e-06,
+      "loss": 0.4982,
+      "step": 5124
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.995755162747381,
+      "learning_rate": 4.419791622854309e-06,
+      "loss": 0.5022,
+      "step": 5125
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3876737438525044,
+      "learning_rate": 4.4180623468602455e-06,
+      "loss": 0.492,
+      "step": 5126
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.412965697705843,
+      "learning_rate": 4.416333141428129e-06,
+      "loss": 0.5204,
+      "step": 5127
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.714506575233435,
+      "learning_rate": 4.414604006767631e-06,
+      "loss": 0.4906,
+      "step": 5128
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.399684552490458,
+      "learning_rate": 4.412874943088416e-06,
+      "loss": 0.5051,
+      "step": 5129
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.830608273947725,
+      "learning_rate": 4.411145950600137e-06,
+      "loss": 0.4941,
+      "step": 5130
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.294137220569014,
+      "learning_rate": 4.4094170295124426e-06,
+      "loss": 0.4996,
+      "step": 5131
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.039412809284853,
+      "learning_rate": 4.407688180034969e-06,
+      "loss": 0.5289,
+      "step": 5132
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7040979077104503,
+      "learning_rate": 4.405959402377345e-06,
+      "loss": 0.4822,
+      "step": 5133
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.594200925831913,
+      "learning_rate": 4.4042306967491935e-06,
+      "loss": 0.5178,
+      "step": 5134
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0408746710470904,
+      "learning_rate": 4.402502063360121e-06,
+      "loss": 0.5142,
+      "step": 5135
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.338577938856459,
+      "learning_rate": 4.400773502419738e-06,
+      "loss": 0.4938,
+      "step": 5136
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8508792461982138,
+      "learning_rate": 4.39904501413763e-06,
+      "loss": 0.5383,
+      "step": 5137
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.606309750322229,
+      "learning_rate": 4.397316598723385e-06,
+      "loss": 0.5278,
+      "step": 5138
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3936159341744191,
+      "learning_rate": 4.395588256386583e-06,
+      "loss": 0.296,
+      "step": 5139
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6232375047173204,
+      "learning_rate": 4.393859987336786e-06,
+      "loss": 0.5056,
+      "step": 5140
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9043399139522736,
+      "learning_rate": 4.392131791783557e-06,
+      "loss": 0.5054,
+      "step": 5141
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3642527383668037,
+      "learning_rate": 4.390403669936443e-06,
+      "loss": 0.4962,
+      "step": 5142
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.268932159783979,
+      "learning_rate": 4.3886756220049855e-06,
+      "loss": 0.4583,
+      "step": 5143
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.721308695134451,
+      "learning_rate": 4.386947648198717e-06,
+      "loss": 0.5005,
+      "step": 5144
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0499442124878064,
+      "learning_rate": 4.38521974872716e-06,
+      "loss": 0.4832,
+      "step": 5145
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6753991473083008,
+      "learning_rate": 4.383491923799828e-06,
+      "loss": 0.5335,
+      "step": 5146
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3410230765807447,
+      "learning_rate": 4.3817641736262255e-06,
+      "loss": 0.5168,
+      "step": 5147
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8403751947650595,
+      "learning_rate": 4.380036498415847e-06,
+      "loss": 0.5231,
+      "step": 5148
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.245991061418102,
+      "learning_rate": 4.3783088983781815e-06,
+      "loss": 0.5018,
+      "step": 5149
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3794053678804308,
+      "learning_rate": 4.376581373722705e-06,
+      "loss": 0.5138,
+      "step": 5150
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.415558252056151,
+      "learning_rate": 4.374853924658886e-06,
+      "loss": 0.5321,
+      "step": 5151
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.714951989946059,
+      "learning_rate": 4.373126551396184e-06,
+      "loss": 0.5428,
+      "step": 5152
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9796060358038177,
+      "learning_rate": 4.37139925414405e-06,
+      "loss": 0.4926,
+      "step": 5153
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.066433418377666,
+      "learning_rate": 4.369672033111921e-06,
+      "loss": 0.4834,
+      "step": 5154
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3722103573663992,
+      "learning_rate": 4.367944888509233e-06,
+      "loss": 0.2999,
+      "step": 5155
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.877229384650831,
+      "learning_rate": 4.366217820545407e-06,
+      "loss": 0.5083,
+      "step": 5156
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.001368929022195,
+      "learning_rate": 4.364490829429855e-06,
+      "loss": 0.4611,
+      "step": 5157
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.314064340943968,
+      "learning_rate": 4.36276391537198e-06,
+      "loss": 0.4638,
+      "step": 5158
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.07492278185023,
+      "learning_rate": 4.361037078581176e-06,
+      "loss": 0.542,
+      "step": 5159
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.25436192186618,
+      "learning_rate": 4.35931031926683e-06,
+      "loss": 0.4631,
+      "step": 5160
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.60989774192595,
+      "learning_rate": 4.3575836376383175e-06,
+      "loss": 0.5205,
+      "step": 5161
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.617387035495487,
+      "learning_rate": 4.3558570339050034e-06,
+      "loss": 0.5016,
+      "step": 5162
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2612845901494314,
+      "learning_rate": 4.354130508276243e-06,
+      "loss": 0.4792,
+      "step": 5163
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3633812267023244,
+      "learning_rate": 4.352404060961387e-06,
+      "loss": 0.4586,
+      "step": 5164
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.40229498576440625,
+      "learning_rate": 4.3506776921697705e-06,
+      "loss": 0.3003,
+      "step": 5165
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.081420074227284,
+      "learning_rate": 4.3489514021107215e-06,
+      "loss": 0.545,
+      "step": 5166
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5736932699607347,
+      "learning_rate": 4.3472251909935635e-06,
+      "loss": 0.5095,
+      "step": 5167
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2018399614350406,
+      "learning_rate": 4.3454990590275966e-06,
+      "loss": 0.464,
+      "step": 5168
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0877737503640637,
+      "learning_rate": 4.343773006422128e-06,
+      "loss": 0.4751,
+      "step": 5169
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.39199548984135996,
+      "learning_rate": 4.342047033386444e-06,
+      "loss": 0.2906,
+      "step": 5170
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1184005789735183,
+      "learning_rate": 4.3403211401298245e-06,
+      "loss": 0.5095,
+      "step": 5171
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6969445267386867,
+      "learning_rate": 4.338595326861542e-06,
+      "loss": 0.5439,
+      "step": 5172
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.7296097133067505,
+      "learning_rate": 4.336869593790857e-06,
+      "loss": 0.5314,
+      "step": 5173
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1100091848836158,
+      "learning_rate": 4.335143941127018e-06,
+      "loss": 0.511,
+      "step": 5174
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5644576710390097,
+      "learning_rate": 4.3334183690792695e-06,
+      "loss": 0.4888,
+      "step": 5175
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7346700216330246,
+      "learning_rate": 4.3316928778568414e-06,
+      "loss": 0.5453,
+      "step": 5176
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.979062647389978,
+      "learning_rate": 4.329967467668955e-06,
+      "loss": 0.4666,
+      "step": 5177
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.02397488650552,
+      "learning_rate": 4.328242138724827e-06,
+      "loss": 0.5193,
+      "step": 5178
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8567049264775,
+      "learning_rate": 4.326516891233652e-06,
+      "loss": 0.4668,
+      "step": 5179
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.072814175594654,
+      "learning_rate": 4.324791725404627e-06,
+      "loss": 0.4931,
+      "step": 5180
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.618262559561906,
+      "learning_rate": 4.3230666414469325e-06,
+      "loss": 0.5168,
+      "step": 5181
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.39753744195901786,
+      "learning_rate": 4.321341639569741e-06,
+      "loss": 0.2851,
+      "step": 5182
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1122918995180657,
+      "learning_rate": 4.3196167199822155e-06,
+      "loss": 0.4645,
+      "step": 5183
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1664713979052377,
+      "learning_rate": 4.317891882893509e-06,
+      "loss": 0.4769,
+      "step": 5184
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.867084636779875,
+      "learning_rate": 4.316167128512763e-06,
+      "loss": 0.493,
+      "step": 5185
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.4420737793637795,
+      "learning_rate": 4.314442457049111e-06,
+      "loss": 0.4812,
+      "step": 5186
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.317611973227783,
+      "learning_rate": 4.312717868711674e-06,
+      "loss": 0.4828,
+      "step": 5187
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.87813880828753,
+      "learning_rate": 4.310993363709563e-06,
+      "loss": 0.5083,
+      "step": 5188
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9013946040084955,
+      "learning_rate": 4.309268942251887e-06,
+      "loss": 0.5055,
+      "step": 5189
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.531874433561587,
+      "learning_rate": 4.307544604547729e-06,
+      "loss": 0.4939,
+      "step": 5190
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4766565795680036,
+      "learning_rate": 4.305820350806176e-06,
+      "loss": 0.4938,
+      "step": 5191
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.41825878975197317,
+      "learning_rate": 4.304096181236299e-06,
+      "loss": 0.2788,
+      "step": 5192
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.001270115498302,
+      "learning_rate": 4.302372096047157e-06,
+      "loss": 0.4691,
+      "step": 5193
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.051067998992695,
+      "learning_rate": 4.300648095447806e-06,
+      "loss": 0.5334,
+      "step": 5194
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.853665704485969,
+      "learning_rate": 4.298924179647283e-06,
+      "loss": 0.5369,
+      "step": 5195
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0347094887388337,
+      "learning_rate": 4.297200348854621e-06,
+      "loss": 0.4998,
+      "step": 5196
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.274227108110564,
+      "learning_rate": 4.295476603278839e-06,
+      "loss": 0.548,
+      "step": 5197
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9077997096723145,
+      "learning_rate": 4.293752943128948e-06,
+      "loss": 0.4123,
+      "step": 5198
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 23.88082331659904,
+      "learning_rate": 4.2920293686139485e-06,
+      "loss": 0.4516,
+      "step": 5199
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5037528742510355,
+      "learning_rate": 4.290305879942828e-06,
+      "loss": 0.4705,
+      "step": 5200
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.355297517225834,
+      "learning_rate": 4.288582477324567e-06,
+      "loss": 0.4813,
+      "step": 5201
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8402104636279923,
+      "learning_rate": 4.286859160968131e-06,
+      "loss": 0.5204,
+      "step": 5202
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.457295647654533,
+      "learning_rate": 4.285135931082481e-06,
+      "loss": 0.4567,
+      "step": 5203
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.685165365013145,
+      "learning_rate": 4.283412787876565e-06,
+      "loss": 0.5098,
+      "step": 5204
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.017564771293986,
+      "learning_rate": 4.281689731559318e-06,
+      "loss": 0.4923,
+      "step": 5205
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9902985951778773,
+      "learning_rate": 4.279966762339668e-06,
+      "loss": 0.4823,
+      "step": 5206
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4038715441092065,
+      "learning_rate": 4.27824388042653e-06,
+      "loss": 0.2925,
+      "step": 5207
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4972849267477524,
+      "learning_rate": 4.276521086028809e-06,
+      "loss": 0.4905,
+      "step": 5208
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.541252662737706,
+      "learning_rate": 4.274798379355403e-06,
+      "loss": 0.5099,
+      "step": 5209
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.477810956643992,
+      "learning_rate": 4.273075760615193e-06,
+      "loss": 0.4967,
+      "step": 5210
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.204341412757974,
+      "learning_rate": 4.271353230017053e-06,
+      "loss": 0.4878,
+      "step": 5211
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.273121701784035,
+      "learning_rate": 4.269630787769845e-06,
+      "loss": 0.4733,
+      "step": 5212
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.1771966757746375,
+      "learning_rate": 4.2679084340824215e-06,
+      "loss": 0.5482,
+      "step": 5213
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.383076382961576,
+      "learning_rate": 4.266186169163625e-06,
+      "loss": 0.5218,
+      "step": 5214
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.108350214248773,
+      "learning_rate": 4.264463993222286e-06,
+      "loss": 0.5111,
+      "step": 5215
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3917604689791676,
+      "learning_rate": 4.262741906467221e-06,
+      "loss": 0.2918,
+      "step": 5216
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.96039590965726,
+      "learning_rate": 4.261019909107243e-06,
+      "loss": 0.496,
+      "step": 5217
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0216353916180676,
+      "learning_rate": 4.25929800135115e-06,
+      "loss": 0.4409,
+      "step": 5218
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.474798255348862,
+      "learning_rate": 4.257576183407726e-06,
+      "loss": 0.4404,
+      "step": 5219
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.182438045779249,
+      "learning_rate": 4.255854455485753e-06,
+      "loss": 0.5051,
+      "step": 5220
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.0280583816069475,
+      "learning_rate": 4.254132817793989e-06,
+      "loss": 0.4988,
+      "step": 5221
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.341418853482199,
+      "learning_rate": 4.252411270541194e-06,
+      "loss": 0.4869,
+      "step": 5222
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.380582176084413,
+      "learning_rate": 4.250689813936109e-06,
+      "loss": 0.5243,
+      "step": 5223
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9375782422997636,
+      "learning_rate": 4.248968448187466e-06,
+      "loss": 0.5248,
+      "step": 5224
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.452240176212645,
+      "learning_rate": 4.247247173503989e-06,
+      "loss": 0.505,
+      "step": 5225
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2357808762767655,
+      "learning_rate": 4.245525990094388e-06,
+      "loss": 0.4928,
+      "step": 5226
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.849619720869286,
+      "learning_rate": 4.2438048981673615e-06,
+      "loss": 0.441,
+      "step": 5227
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3979767184487755,
+      "learning_rate": 4.242083897931597e-06,
+      "loss": 0.503,
+      "step": 5228
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.364961866638325,
+      "learning_rate": 4.240362989595775e-06,
+      "loss": 0.4199,
+      "step": 5229
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4021232404761964,
+      "learning_rate": 4.2386421733685575e-06,
+      "loss": 0.2802,
+      "step": 5230
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4427626650450027,
+      "learning_rate": 4.236921449458603e-06,
+      "loss": 0.5106,
+      "step": 5231
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.665064848128177,
+      "learning_rate": 4.235200818074553e-06,
+      "loss": 0.4853,
+      "step": 5232
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.173075090069957,
+      "learning_rate": 4.233480279425039e-06,
+      "loss": 0.5432,
+      "step": 5233
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.312485705231297,
+      "learning_rate": 4.231759833718685e-06,
+      "loss": 0.5042,
+      "step": 5234
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.138360525606304,
+      "learning_rate": 4.2300394811641e-06,
+      "loss": 0.5359,
+      "step": 5235
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0056056530270143,
+      "learning_rate": 4.22831922196988e-06,
+      "loss": 0.5159,
+      "step": 5236
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.635817741821625,
+      "learning_rate": 4.2265990563446165e-06,
+      "loss": 0.4637,
+      "step": 5237
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.84099318217268,
+      "learning_rate": 4.224878984496884e-06,
+      "loss": 0.5093,
+      "step": 5238
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7490698596720584,
+      "learning_rate": 4.223159006635245e-06,
+      "loss": 0.4813,
+      "step": 5239
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 12.587008350683789,
+      "learning_rate": 4.221439122968256e-06,
+      "loss": 0.5008,
+      "step": 5240
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.267139414231485,
+      "learning_rate": 4.219719333704459e-06,
+      "loss": 0.4658,
+      "step": 5241
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8276060468490125,
+      "learning_rate": 4.217999639052382e-06,
+      "loss": 0.5124,
+      "step": 5242
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.361128380364196,
+      "learning_rate": 4.216280039220544e-06,
+      "loss": 0.503,
+      "step": 5243
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2049273427984675,
+      "learning_rate": 4.214560534417453e-06,
+      "loss": 0.4862,
+      "step": 5244
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3968627280933373,
+      "learning_rate": 4.2128411248516055e-06,
+      "loss": 0.2903,
+      "step": 5245
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.876778383811551,
+      "learning_rate": 4.211121810731485e-06,
+      "loss": 0.5615,
+      "step": 5246
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8556627728961095,
+      "learning_rate": 4.209402592265564e-06,
+      "loss": 0.487,
+      "step": 5247
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.811289399349098,
+      "learning_rate": 4.207683469662306e-06,
+      "loss": 0.4928,
+      "step": 5248
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.23174179647011,
+      "learning_rate": 4.2059644431301575e-06,
+      "loss": 0.4961,
+      "step": 5249
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.263004618985025,
+      "learning_rate": 4.204245512877557e-06,
+      "loss": 0.4917,
+      "step": 5250
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5173644478224646,
+      "learning_rate": 4.202526679112932e-06,
+      "loss": 0.5632,
+      "step": 5251
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.39754805647740676,
+      "learning_rate": 4.200807942044699e-06,
+      "loss": 0.2861,
+      "step": 5252
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.28102853395238,
+      "learning_rate": 4.199089301881256e-06,
+      "loss": 0.5167,
+      "step": 5253
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.38776567513767224,
+      "learning_rate": 4.197370758830998e-06,
+      "loss": 0.2931,
+      "step": 5254
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2433330087446244,
+      "learning_rate": 4.1956523131023e-06,
+      "loss": 0.508,
+      "step": 5255
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.939314690204752,
+      "learning_rate": 4.193933964903532e-06,
+      "loss": 0.5424,
+      "step": 5256
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.585322471038182,
+      "learning_rate": 4.192215714443052e-06,
+      "loss": 0.468,
+      "step": 5257
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.38640301400845506,
+      "learning_rate": 4.1904975619292e-06,
+      "loss": 0.3139,
+      "step": 5258
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5704073946979604,
+      "learning_rate": 4.18877950757031e-06,
+      "loss": 0.5325,
+      "step": 5259
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.71008258157925,
+      "learning_rate": 4.187061551574701e-06,
+      "loss": 0.5089,
+      "step": 5260
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0065665783002893,
+      "learning_rate": 4.185343694150682e-06,
+      "loss": 0.5003,
+      "step": 5261
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.996997957479146,
+      "learning_rate": 4.1836259355065475e-06,
+      "loss": 0.5059,
+      "step": 5262
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.750984543229581,
+      "learning_rate": 4.181908275850586e-06,
+      "loss": 0.4866,
+      "step": 5263
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8715247820393563,
+      "learning_rate": 4.180190715391063e-06,
+      "loss": 0.5095,
+      "step": 5264
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.960005317519349,
+      "learning_rate": 4.178473254336243e-06,
+      "loss": 0.5187,
+      "step": 5265
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 273.09643175266507,
+      "learning_rate": 4.176755892894374e-06,
+      "loss": 0.5129,
+      "step": 5266
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.234099809001462,
+      "learning_rate": 4.175038631273689e-06,
+      "loss": 0.4832,
+      "step": 5267
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8658713579809625,
+      "learning_rate": 4.173321469682416e-06,
+      "loss": 0.4951,
+      "step": 5268
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.328072892780463,
+      "learning_rate": 4.171604408328765e-06,
+      "loss": 0.4857,
+      "step": 5269
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.174644928146073,
+      "learning_rate": 4.1698874474209326e-06,
+      "loss": 0.473,
+      "step": 5270
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9227059218509126,
+      "learning_rate": 4.1681705871671114e-06,
+      "loss": 0.5129,
+      "step": 5271
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5852356376313326,
+      "learning_rate": 4.166453827775474e-06,
+      "loss": 0.4464,
+      "step": 5272
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.048390804470224,
+      "learning_rate": 4.164737169454185e-06,
+      "loss": 0.4946,
+      "step": 5273
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9966930367443245,
+      "learning_rate": 4.1630206124113925e-06,
+      "loss": 0.5164,
+      "step": 5274
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.693375122378021,
+      "learning_rate": 4.161304156855235e-06,
+      "loss": 0.4466,
+      "step": 5275
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3796966207737432,
+      "learning_rate": 4.159587802993842e-06,
+      "loss": 0.3004,
+      "step": 5276
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.441320416245484,
+      "learning_rate": 4.157871551035324e-06,
+      "loss": 0.4858,
+      "step": 5277
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 12.757747675061534,
+      "learning_rate": 4.156155401187783e-06,
+      "loss": 0.4574,
+      "step": 5278
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.836908642366653,
+      "learning_rate": 4.15443935365931e-06,
+      "loss": 0.5822,
+      "step": 5279
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.554131670499752,
+      "learning_rate": 4.1527234086579806e-06,
+      "loss": 0.4661,
+      "step": 5280
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7893041244944414,
+      "learning_rate": 4.151007566391857e-06,
+      "loss": 0.5046,
+      "step": 5281
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0070812650394627,
+      "learning_rate": 4.149291827068995e-06,
+      "loss": 0.4493,
+      "step": 5282
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3592440124308087,
+      "learning_rate": 4.1475761908974315e-06,
+      "loss": 0.5024,
+      "step": 5283
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.539905422618504,
+      "learning_rate": 4.145860658085194e-06,
+      "loss": 0.4586,
+      "step": 5284
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4129660051317448,
+      "learning_rate": 4.1441452288402965e-06,
+      "loss": 0.5001,
+      "step": 5285
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.454341512165117,
+      "learning_rate": 4.142429903370739e-06,
+      "loss": 0.469,
+      "step": 5286
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.37554067350191384,
+      "learning_rate": 4.140714681884513e-06,
+      "loss": 0.2861,
+      "step": 5287
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.093478821993771,
+      "learning_rate": 4.1389995645895944e-06,
+      "loss": 0.4829,
+      "step": 5288
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.161008162360307,
+      "learning_rate": 4.137284551693946e-06,
+      "loss": 0.4717,
+      "step": 5289
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.96526866942232,
+      "learning_rate": 4.13556964340552e-06,
+      "loss": 0.4835,
+      "step": 5290
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9679600729350164,
+      "learning_rate": 4.133854839932257e-06,
+      "loss": 0.453,
+      "step": 5291
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.679010778104957,
+      "learning_rate": 4.132140141482079e-06,
+      "loss": 0.4965,
+      "step": 5292
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7145936917297786,
+      "learning_rate": 4.1304255482629e-06,
+      "loss": 0.5252,
+      "step": 5293
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9260430847625947,
+      "learning_rate": 4.128711060482624e-06,
+      "loss": 0.4627,
+      "step": 5294
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.574855804231774,
+      "learning_rate": 4.126996678349133e-06,
+      "loss": 0.5311,
+      "step": 5295
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8678574169375617,
+      "learning_rate": 4.125282402070307e-06,
+      "loss": 0.4576,
+      "step": 5296
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5598285424943574,
+      "learning_rate": 4.1235682318540035e-06,
+      "loss": 0.5372,
+      "step": 5297
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.36389811794954,
+      "learning_rate": 4.121854167908072e-06,
+      "loss": 0.5,
+      "step": 5298
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.94993962581014,
+      "learning_rate": 4.120140210440352e-06,
+      "loss": 0.4507,
+      "step": 5299
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.30170593166142,
+      "learning_rate": 4.1184263596586645e-06,
+      "loss": 0.4831,
+      "step": 5300
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.6783363738694637,
+      "learning_rate": 4.11671261577082e-06,
+      "loss": 0.5213,
+      "step": 5301
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.378032736880876,
+      "learning_rate": 4.114998978984616e-06,
+      "loss": 0.5582,
+      "step": 5302
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.510535524217773,
+      "learning_rate": 4.113285449507837e-06,
+      "loss": 0.5371,
+      "step": 5303
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.208753449927913,
+      "learning_rate": 4.111572027548253e-06,
+      "loss": 0.5241,
+      "step": 5304
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.815990204007157,
+      "learning_rate": 4.109858713313628e-06,
+      "loss": 0.4794,
+      "step": 5305
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.37625469760331,
+      "learning_rate": 4.108145507011698e-06,
+      "loss": 0.2857,
+      "step": 5306
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.545244548896631,
+      "learning_rate": 4.1064324088502026e-06,
+      "loss": 0.4456,
+      "step": 5307
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0276720428106505,
+      "learning_rate": 4.104719419036857e-06,
+      "loss": 0.4875,
+      "step": 5308
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.921298962665014,
+      "learning_rate": 4.1030065377793675e-06,
+      "loss": 0.4681,
+      "step": 5309
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4503689173992007,
+      "learning_rate": 4.101293765285429e-06,
+      "loss": 0.4779,
+      "step": 5310
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.0558585170382475,
+      "learning_rate": 4.09958110176272e-06,
+      "loss": 0.5571,
+      "step": 5311
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.38041096914186817,
+      "learning_rate": 4.097868547418905e-06,
+      "loss": 0.2856,
+      "step": 5312
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.359553343233173,
+      "learning_rate": 4.09615610246164e-06,
+      "loss": 0.4972,
+      "step": 5313
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.599024330765357,
+      "learning_rate": 4.0944437670985635e-06,
+      "loss": 0.5137,
+      "step": 5314
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.264463064935826,
+      "learning_rate": 4.0927315415373016e-06,
+      "loss": 0.5086,
+      "step": 5315
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.29860791332592,
+      "learning_rate": 4.091019425985471e-06,
+      "loss": 0.4875,
+      "step": 5316
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.736653969306901,
+      "learning_rate": 4.089307420650665e-06,
+      "loss": 0.5134,
+      "step": 5317
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.035538938667559,
+      "learning_rate": 4.087595525740476e-06,
+      "loss": 0.4846,
+      "step": 5318
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9874233044144884,
+      "learning_rate": 4.085883741462474e-06,
+      "loss": 0.4832,
+      "step": 5319
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.961204524607424,
+      "learning_rate": 4.084172068024219e-06,
+      "loss": 0.5132,
+      "step": 5320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.954122390185929,
+      "learning_rate": 4.082460505633259e-06,
+      "loss": 0.4545,
+      "step": 5321
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.226875712537303,
+      "learning_rate": 4.080749054497127e-06,
+      "loss": 0.4677,
+      "step": 5322
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5378958587766283,
+      "learning_rate": 4.079037714823341e-06,
+      "loss": 0.485,
+      "step": 5323
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.740521020694054,
+      "learning_rate": 4.077326486819405e-06,
+      "loss": 0.4718,
+      "step": 5324
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 19.14006705326584,
+      "learning_rate": 4.075615370692815e-06,
+      "loss": 0.5249,
+      "step": 5325
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.466004508017082,
+      "learning_rate": 4.073904366651049e-06,
+      "loss": 0.4773,
+      "step": 5326
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40287916112573013,
+      "learning_rate": 4.072193474901571e-06,
+      "loss": 0.2863,
+      "step": 5327
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.3495074382933385,
+      "learning_rate": 4.070482695651832e-06,
+      "loss": 0.517,
+      "step": 5328
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6277128303633983,
+      "learning_rate": 4.068772029109271e-06,
+      "loss": 0.4684,
+      "step": 5329
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.269882367659584,
+      "learning_rate": 4.067061475481311e-06,
+      "loss": 0.478,
+      "step": 5330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1389602091656683,
+      "learning_rate": 4.0653510349753645e-06,
+      "loss": 0.4985,
+      "step": 5331
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.793407421235822,
+      "learning_rate": 4.063640707798826e-06,
+      "loss": 0.5638,
+      "step": 5332
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.804482001503223,
+      "learning_rate": 4.061930494159081e-06,
+      "loss": 0.4753,
+      "step": 5333
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.050190419615653,
+      "learning_rate": 4.060220394263497e-06,
+      "loss": 0.5231,
+      "step": 5334
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5827485422297065,
+      "learning_rate": 4.0585104083194295e-06,
+      "loss": 0.5033,
+      "step": 5335
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.163776521289463,
+      "learning_rate": 4.056800536534222e-06,
+      "loss": 0.5328,
+      "step": 5336
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6268231122479224,
+      "learning_rate": 4.0550907791152025e-06,
+      "loss": 0.4865,
+      "step": 5337
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.263257657804248,
+      "learning_rate": 4.053381136269682e-06,
+      "loss": 0.4796,
+      "step": 5338
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.992759701523924,
+      "learning_rate": 4.0516716082049625e-06,
+      "loss": 0.5242,
+      "step": 5339
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.761569299805236,
+      "learning_rate": 4.049962195128328e-06,
+      "loss": 0.4892,
+      "step": 5340
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.444048056647661,
+      "learning_rate": 4.0482528972470545e-06,
+      "loss": 0.4833,
+      "step": 5341
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.537544266187836,
+      "learning_rate": 4.046543714768399e-06,
+      "loss": 0.4907,
+      "step": 5342
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5838135158105016,
+      "learning_rate": 4.044834647899603e-06,
+      "loss": 0.497,
+      "step": 5343
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40054905865469853,
+      "learning_rate": 4.043125696847901e-06,
+      "loss": 0.2953,
+      "step": 5344
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.41043370825886,
+      "learning_rate": 4.041416861820506e-06,
+      "loss": 0.4783,
+      "step": 5345
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.858606860796346,
+      "learning_rate": 4.039708143024621e-06,
+      "loss": 0.5027,
+      "step": 5346
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.493974114462822,
+      "learning_rate": 4.037999540667436e-06,
+      "loss": 0.4803,
+      "step": 5347
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.40413510876322906,
+      "learning_rate": 4.036291054956122e-06,
+      "loss": 0.2919,
+      "step": 5348
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7441901309762597,
+      "learning_rate": 4.0345826860978395e-06,
+      "loss": 0.4853,
+      "step": 5349
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.393611042407267,
+      "learning_rate": 4.0328744342997355e-06,
+      "loss": 0.4821,
+      "step": 5350
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3648549064683482,
+      "learning_rate": 4.031166299768939e-06,
+      "loss": 0.2923,
+      "step": 5351
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0391757893244864,
+      "learning_rate": 4.029458282712571e-06,
+      "loss": 0.4994,
+      "step": 5352
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.146525032142957,
+      "learning_rate": 4.0277503833377306e-06,
+      "loss": 0.4745,
+      "step": 5353
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.786737150013939,
+      "learning_rate": 4.026042601851509e-06,
+      "loss": 0.5105,
+      "step": 5354
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3724650450241077,
+      "learning_rate": 4.024334938460978e-06,
+      "loss": 0.2862,
+      "step": 5355
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9636288810362155,
+      "learning_rate": 4.022627393373201e-06,
+      "loss": 0.4533,
+      "step": 5356
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.28405655815301,
+      "learning_rate": 4.020919966795221e-06,
+      "loss": 0.4798,
+      "step": 5357
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.7582375853065795,
+      "learning_rate": 4.019212658934073e-06,
+      "loss": 0.5582,
+      "step": 5358
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.137949257610928,
+      "learning_rate": 4.01750546999677e-06,
+      "loss": 0.5281,
+      "step": 5359
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.1175232367746695,
+      "learning_rate": 4.015798400190314e-06,
+      "loss": 0.5196,
+      "step": 5360
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0619038785287467,
+      "learning_rate": 4.014091449721697e-06,
+      "loss": 0.5242,
+      "step": 5361
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.932656265007809,
+      "learning_rate": 4.01238461879789e-06,
+      "loss": 0.4921,
+      "step": 5362
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7379567476605486,
+      "learning_rate": 4.0106779076258515e-06,
+      "loss": 0.4899,
+      "step": 5363
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.314171295013133,
+      "learning_rate": 4.0089713164125285e-06,
+      "loss": 0.4611,
+      "step": 5364
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.246469375342958,
+      "learning_rate": 4.007264845364851e-06,
+      "loss": 0.5089,
+      "step": 5365
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.68807759469003,
+      "learning_rate": 4.005558494689732e-06,
+      "loss": 0.4753,
+      "step": 5366
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.748879432478926,
+      "learning_rate": 4.0038522645940746e-06,
+      "loss": 0.5702,
+      "step": 5367
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4516902919488004,
+      "learning_rate": 4.002146155284765e-06,
+      "loss": 0.4781,
+      "step": 5368
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.972688490148699,
+      "learning_rate": 4.0004401669686746e-06,
+      "loss": 0.4879,
+      "step": 5369
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.41936745908274,
+      "learning_rate": 3.998734299852659e-06,
+      "loss": 0.4653,
+      "step": 5370
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8757222354577245,
+      "learning_rate": 3.99702855414356e-06,
+      "loss": 0.4414,
+      "step": 5371
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.552776826259869,
+      "learning_rate": 3.9953229300482085e-06,
+      "loss": 0.4687,
+      "step": 5372
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.834833861548482,
+      "learning_rate": 3.993617427773416e-06,
+      "loss": 0.4553,
+      "step": 5373
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.692953809406797,
+      "learning_rate": 3.991912047525976e-06,
+      "loss": 0.447,
+      "step": 5374
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.48092915630988,
+      "learning_rate": 3.99020678951268e-06,
+      "loss": 0.513,
+      "step": 5375
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.463295274789725,
+      "learning_rate": 3.9885016539402896e-06,
+      "loss": 0.487,
+      "step": 5376
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.546874441253759,
+      "learning_rate": 3.986796641015561e-06,
+      "loss": 0.4826,
+      "step": 5377
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.480412057131895,
+      "learning_rate": 3.985091750945234e-06,
+      "loss": 0.5019,
+      "step": 5378
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.704571757248637,
+      "learning_rate": 3.983386983936031e-06,
+      "loss": 0.4839,
+      "step": 5379
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0505740478775807,
+      "learning_rate": 3.981682340194661e-06,
+      "loss": 0.4919,
+      "step": 5380
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.848516063104505,
+      "learning_rate": 3.979977819927818e-06,
+      "loss": 0.4666,
+      "step": 5381
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8264616780574046,
+      "learning_rate": 3.97827342334218e-06,
+      "loss": 0.4888,
+      "step": 5382
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.181534510765595,
+      "learning_rate": 3.976569150644411e-06,
+      "loss": 0.5116,
+      "step": 5383
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 13.91180686821159,
+      "learning_rate": 3.974865002041163e-06,
+      "loss": 0.4926,
+      "step": 5384
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.6824050498430605,
+      "learning_rate": 3.973160977739065e-06,
+      "loss": 0.5069,
+      "step": 5385
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.3607894108270795,
+      "learning_rate": 3.97145707794474e-06,
+      "loss": 0.48,
+      "step": 5386
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.8985084283312865,
+      "learning_rate": 3.969753302864791e-06,
+      "loss": 0.4983,
+      "step": 5387
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.555146353403277,
+      "learning_rate": 3.9680496527058055e-06,
+      "loss": 0.5067,
+      "step": 5388
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.5851133349605995,
+      "learning_rate": 3.966346127674356e-06,
+      "loss": 0.4901,
+      "step": 5389
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.0136739231919645,
+      "learning_rate": 3.964642727977004e-06,
+      "loss": 0.4799,
+      "step": 5390
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.860334072531309,
+      "learning_rate": 3.96293945382029e-06,
+      "loss": 0.4847,
+      "step": 5391
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.334238850716513,
+      "learning_rate": 3.9612363054107425e-06,
+      "loss": 0.4948,
+      "step": 5392
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.313744216853986,
+      "learning_rate": 3.959533282954874e-06,
+      "loss": 0.4593,
+      "step": 5393
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.294542063616271,
+      "learning_rate": 3.957830386659181e-06,
+      "loss": 0.4647,
+      "step": 5394
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1569278699092833,
+      "learning_rate": 3.956127616730148e-06,
+      "loss": 0.502,
+      "step": 5395
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9093462700787542,
+      "learning_rate": 3.954424973374239e-06,
+      "loss": 0.4863,
+      "step": 5396
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.39683271749851917,
+      "learning_rate": 3.952722456797906e-06,
+      "loss": 0.2835,
+      "step": 5397
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.581321340432739,
+      "learning_rate": 3.9510200672075874e-06,
+      "loss": 0.4872,
+      "step": 5398
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.417055150174939,
+      "learning_rate": 3.949317804809701e-06,
+      "loss": 0.4769,
+      "step": 5399
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3093269285245057,
+      "learning_rate": 3.9476156698106555e-06,
+      "loss": 0.5635,
+      "step": 5400
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.107131314059395,
+      "learning_rate": 3.945913662416836e-06,
+      "loss": 0.5088,
+      "step": 5401
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0824905568928433,
+      "learning_rate": 3.944211782834618e-06,
+      "loss": 0.4992,
+      "step": 5402
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8498639203759275,
+      "learning_rate": 3.942510031270363e-06,
+      "loss": 0.4806,
+      "step": 5403
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4600160166753136,
+      "learning_rate": 3.940808407930412e-06,
+      "loss": 0.4852,
+      "step": 5404
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.970092202249619,
+      "learning_rate": 3.939106913021091e-06,
+      "loss": 0.4635,
+      "step": 5405
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.267477649842354,
+      "learning_rate": 3.937405546748716e-06,
+      "loss": 0.4533,
+      "step": 5406
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.037950933488053,
+      "learning_rate": 3.935704309319583e-06,
+      "loss": 0.5211,
+      "step": 5407
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.178983545758624,
+      "learning_rate": 3.934003200939969e-06,
+      "loss": 0.5,
+      "step": 5408
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.42082431638775075,
+      "learning_rate": 3.932302221816144e-06,
+      "loss": 0.296,
+      "step": 5409
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.42227762586702255,
+      "learning_rate": 3.930601372154355e-06,
+      "loss": 0.2617,
+      "step": 5410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.188626762551968,
+      "learning_rate": 3.928900652160837e-06,
+      "loss": 0.5117,
+      "step": 5411
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3233785664603848,
+      "learning_rate": 3.927200062041808e-06,
+      "loss": 0.4914,
+      "step": 5412
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3627799804577374,
+      "learning_rate": 3.925499602003468e-06,
+      "loss": 0.5603,
+      "step": 5413
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.26985352289672,
+      "learning_rate": 3.923799272252007e-06,
+      "loss": 0.4424,
+      "step": 5414
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5703170984603596,
+      "learning_rate": 3.9220990729935935e-06,
+      "loss": 0.4918,
+      "step": 5415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.358656050952053,
+      "learning_rate": 3.920399004434383e-06,
+      "loss": 0.4886,
+      "step": 5416
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.266022970309969,
+      "learning_rate": 3.918699066780517e-06,
+      "loss": 0.4565,
+      "step": 5417
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3269747835675396,
+      "learning_rate": 3.916999260238116e-06,
+      "loss": 0.4884,
+      "step": 5418
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.726046185980277,
+      "learning_rate": 3.915299585013288e-06,
+      "loss": 0.522,
+      "step": 5419
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8371829131223905,
+      "learning_rate": 3.913600041312125e-06,
+      "loss": 0.5279,
+      "step": 5420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.180807666959125,
+      "learning_rate": 3.911900629340703e-06,
+      "loss": 0.479,
+      "step": 5421
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8887683885756466,
+      "learning_rate": 3.910201349305082e-06,
+      "loss": 0.4659,
+      "step": 5422
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.229191183182075,
+      "learning_rate": 3.908502201411304e-06,
+      "loss": 0.5006,
+      "step": 5423
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4147580526641507,
+      "learning_rate": 3.906803185865395e-06,
+      "loss": 0.29,
+      "step": 5424
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.39237442951714,
+      "learning_rate": 3.9051043028733696e-06,
+      "loss": 0.4625,
+      "step": 5425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4039388583539827,
+      "learning_rate": 3.903405552641222e-06,
+      "loss": 0.2926,
+      "step": 5426
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.933805372418194,
+      "learning_rate": 3.901706935374933e-06,
+      "loss": 0.5084,
+      "step": 5427
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.289777523706703,
+      "learning_rate": 3.900008451280462e-06,
+      "loss": 0.5036,
+      "step": 5428
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2923497569472397,
+      "learning_rate": 3.89831010056376e-06,
+      "loss": 0.4655,
+      "step": 5429
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.983986293687524,
+      "learning_rate": 3.896611883430756e-06,
+      "loss": 0.5384,
+      "step": 5430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3868510496645703,
+      "learning_rate": 3.894913800087365e-06,
+      "loss": 0.5251,
+      "step": 5431
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.56039954371811,
+      "learning_rate": 3.8932158507394885e-06,
+      "loss": 0.5211,
+      "step": 5432
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.495915834342415,
+      "learning_rate": 3.891518035593003e-06,
+      "loss": 0.5058,
+      "step": 5433
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.953101972548483,
+      "learning_rate": 3.889820354853777e-06,
+      "loss": 0.4685,
+      "step": 5434
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.6698517229737435,
+      "learning_rate": 3.888122808727661e-06,
+      "loss": 0.465,
+      "step": 5435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.685527993210346,
+      "learning_rate": 3.886425397420487e-06,
+      "loss": 0.4797,
+      "step": 5436
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7496085342469065,
+      "learning_rate": 3.884728121138073e-06,
+      "loss": 0.5018,
+      "step": 5437
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.1470112141624815,
+      "learning_rate": 3.8830309800862196e-06,
+      "loss": 0.5042,
+      "step": 5438
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0699567075836316,
+      "learning_rate": 3.88133397447071e-06,
+      "loss": 0.4822,
+      "step": 5439
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0229102391654674,
+      "learning_rate": 3.879637104497313e-06,
+      "loss": 0.5721,
+      "step": 5440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.111025215015094,
+      "learning_rate": 3.87794037037178e-06,
+      "loss": 0.5021,
+      "step": 5441
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.343657785097313,
+      "learning_rate": 3.876243772299844e-06,
+      "loss": 0.4839,
+      "step": 5442
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.136690427882092,
+      "learning_rate": 3.8745473104872276e-06,
+      "loss": 0.5071,
+      "step": 5443
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.100007896817971,
+      "learning_rate": 3.872850985139628e-06,
+      "loss": 0.5276,
+      "step": 5444
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.475082936006984,
+      "learning_rate": 3.871154796462732e-06,
+      "loss": 0.4689,
+      "step": 5445
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2768025363435798,
+      "learning_rate": 3.86945874466221e-06,
+      "loss": 0.4798,
+      "step": 5446
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0906002195652063,
+      "learning_rate": 3.86776282994371e-06,
+      "loss": 0.558,
+      "step": 5447
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.020844737772027,
+      "learning_rate": 3.866067052512872e-06,
+      "loss": 0.5233,
+      "step": 5448
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.680092736447733,
+      "learning_rate": 3.864371412575314e-06,
+      "loss": 0.4777,
+      "step": 5449
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.07836373750058,
+      "learning_rate": 3.862675910336637e-06,
+      "loss": 0.5365,
+      "step": 5450
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.235735128095259,
+      "learning_rate": 3.860980546002425e-06,
+      "loss": 0.4764,
+      "step": 5451
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.485376040192694,
+      "learning_rate": 3.859285319778251e-06,
+      "loss": 0.4525,
+      "step": 5452
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.272828362348663,
+      "learning_rate": 3.857590231869665e-06,
+      "loss": 0.4807,
+      "step": 5453
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5627423560137417,
+      "learning_rate": 3.8558952824822e-06,
+      "loss": 0.5455,
+      "step": 5454
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.479519155188008,
+      "learning_rate": 3.854200471821376e-06,
+      "loss": 0.4861,
+      "step": 5455
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5216980830406013,
+      "learning_rate": 3.852505800092696e-06,
+      "loss": 0.5155,
+      "step": 5456
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.347050216045169,
+      "learning_rate": 3.850811267501643e-06,
+      "loss": 0.4771,
+      "step": 5457
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1715462701065924,
+      "learning_rate": 3.8491168742536865e-06,
+      "loss": 0.4983,
+      "step": 5458
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.22054513787475,
+      "learning_rate": 3.847422620554276e-06,
+      "loss": 0.5021,
+      "step": 5459
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.257672172480012,
+      "learning_rate": 3.845728506608847e-06,
+      "loss": 0.4896,
+      "step": 5460
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.288771013656835,
+      "learning_rate": 3.844034532622816e-06,
+      "loss": 0.477,
+      "step": 5461
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.387779501786191,
+      "learning_rate": 3.842340698801581e-06,
+      "loss": 0.5259,
+      "step": 5462
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.127323827987824,
+      "learning_rate": 3.84064700535053e-06,
+      "loss": 0.4787,
+      "step": 5463
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6397693388195362,
+      "learning_rate": 3.838953452475026e-06,
+      "loss": 0.4927,
+      "step": 5464
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.020891981355561,
+      "learning_rate": 3.837260040380419e-06,
+      "loss": 0.498,
+      "step": 5465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.480480769584782,
+      "learning_rate": 3.83556676927204e-06,
+      "loss": 0.5116,
+      "step": 5466
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9445421047799054,
+      "learning_rate": 3.833873639355204e-06,
+      "loss": 0.4905,
+      "step": 5467
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.717414176680201,
+      "learning_rate": 3.83218065083521e-06,
+      "loss": 0.4976,
+      "step": 5468
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.991147886741841,
+      "learning_rate": 3.830487803917338e-06,
+      "loss": 0.4428,
+      "step": 5469
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.725811985810711,
+      "learning_rate": 3.828795098806851e-06,
+      "loss": 0.4734,
+      "step": 5470
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1851652764423735,
+      "learning_rate": 3.827102535708998e-06,
+      "loss": 0.4697,
+      "step": 5471
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7033208930729864,
+      "learning_rate": 3.825410114829005e-06,
+      "loss": 0.4969,
+      "step": 5472
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.42359422305032157,
+      "learning_rate": 3.823717836372084e-06,
+      "loss": 0.2952,
+      "step": 5473
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6587898293763996,
+      "learning_rate": 3.822025700543431e-06,
+      "loss": 0.5083,
+      "step": 5474
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.918849616497741,
+      "learning_rate": 3.820333707548225e-06,
+      "loss": 0.5091,
+      "step": 5475
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6351940318369893,
+      "learning_rate": 3.818641857591623e-06,
+      "loss": 0.5149,
+      "step": 5476
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.804252455306913,
+      "learning_rate": 3.8169501508787686e-06,
+      "loss": 0.4761,
+      "step": 5477
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6058122769739374,
+      "learning_rate": 3.815258587614785e-06,
+      "loss": 0.508,
+      "step": 5478
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.309681413930542,
+      "learning_rate": 3.8135671680047837e-06,
+      "loss": 0.4945,
+      "step": 5479
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.171166643992821,
+      "learning_rate": 3.8118758922538533e-06,
+      "loss": 0.46,
+      "step": 5480
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.714658390445693,
+      "learning_rate": 3.8101847605670663e-06,
+      "loss": 0.5436,
+      "step": 5481
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2716706783678426,
+      "learning_rate": 3.80849377314948e-06,
+      "loss": 0.5614,
+      "step": 5482
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.420444120354342,
+      "learning_rate": 3.806802930206131e-06,
+      "loss": 0.5181,
+      "step": 5483
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.999977105723907,
+      "learning_rate": 3.8051122319420406e-06,
+      "loss": 0.4799,
+      "step": 5484
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5998307359846127,
+      "learning_rate": 3.803421678562213e-06,
+      "loss": 0.4682,
+      "step": 5485
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.76705898019624,
+      "learning_rate": 3.801731270271629e-06,
+      "loss": 0.475,
+      "step": 5486
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.219929330189998,
+      "learning_rate": 3.8000410072752614e-06,
+      "loss": 0.601,
+      "step": 5487
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.584200855021406,
+      "learning_rate": 3.7983508897780586e-06,
+      "loss": 0.508,
+      "step": 5488
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2723189512322497,
+      "learning_rate": 3.7966609179849528e-06,
+      "loss": 0.4854,
+      "step": 5489
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9372739836349044,
+      "learning_rate": 3.794971092100858e-06,
+      "loss": 0.4668,
+      "step": 5490
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.351492565889659,
+      "learning_rate": 3.7932814123306737e-06,
+      "loss": 0.4933,
+      "step": 5491
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1282544035985533,
+      "learning_rate": 3.7915918788792793e-06,
+      "loss": 0.4835,
+      "step": 5492
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.41636228522703045,
+      "learning_rate": 3.7899024919515337e-06,
+      "loss": 0.3022,
+      "step": 5493
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.693864252873787,
+      "learning_rate": 3.788213251752284e-06,
+      "loss": 0.4928,
+      "step": 5494
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.1644542067973624,
+      "learning_rate": 3.7865241584863547e-06,
+      "loss": 0.5271,
+      "step": 5495
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.544255974581878,
+      "learning_rate": 3.7848352123585562e-06,
+      "loss": 0.5015,
+      "step": 5496
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.075794843717481,
+      "learning_rate": 3.7831464135736767e-06,
+      "loss": 0.4746,
+      "step": 5497
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8164552682663286,
+      "learning_rate": 3.7814577623364885e-06,
+      "loss": 0.5066,
+      "step": 5498
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.799383681880823,
+      "learning_rate": 3.7797692588517487e-06,
+      "loss": 0.547,
+      "step": 5499
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.017278090454311,
+      "learning_rate": 3.7780809033241927e-06,
+      "loss": 0.4885,
+      "step": 5500
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.039166141847858,
+      "learning_rate": 3.776392695958539e-06,
+      "loss": 0.5023,
+      "step": 5501
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3034920276244213,
+      "learning_rate": 3.7747046369594905e-06,
+      "loss": 0.4831,
+      "step": 5502
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.096538692301829,
+      "learning_rate": 3.7730167265317288e-06,
+      "loss": 0.5034,
+      "step": 5503
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0172686816886025,
+      "learning_rate": 3.771328964879918e-06,
+      "loss": 0.4358,
+      "step": 5504
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.37945497949304696,
+      "learning_rate": 3.7696413522087068e-06,
+      "loss": 0.2936,
+      "step": 5505
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.963246671074834,
+      "learning_rate": 3.7679538887227247e-06,
+      "loss": 0.4821,
+      "step": 5506
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6331846362865643,
+      "learning_rate": 3.7662665746265803e-06,
+      "loss": 0.5154,
+      "step": 5507
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.030705197170781,
+      "learning_rate": 3.7645794101248666e-06,
+      "loss": 0.5249,
+      "step": 5508
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3343070046864933,
+      "learning_rate": 3.7628923954221575e-06,
+      "loss": 0.5001,
+      "step": 5509
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4660902063072996,
+      "learning_rate": 3.7612055307230117e-06,
+      "loss": 0.4834,
+      "step": 5510
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6285543814841374,
+      "learning_rate": 3.759518816231966e-06,
+      "loss": 0.51,
+      "step": 5511
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.279716534134998,
+      "learning_rate": 3.7578322521535395e-06,
+      "loss": 0.5255,
+      "step": 5512
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.62739420352473,
+      "learning_rate": 3.7561458386922355e-06,
+      "loss": 0.4806,
+      "step": 5513
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.289952232880056,
+      "learning_rate": 3.754459576052537e-06,
+      "loss": 0.4754,
+      "step": 5514
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3121679997458764,
+      "learning_rate": 3.752773464438909e-06,
+      "loss": 0.5303,
+      "step": 5515
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0504613977038946,
+      "learning_rate": 3.7510875040557974e-06,
+      "loss": 0.4904,
+      "step": 5516
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.249133700949703,
+      "learning_rate": 3.749401695107634e-06,
+      "loss": 0.4919,
+      "step": 5517
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3983298188437385,
+      "learning_rate": 3.7477160377988247e-06,
+      "loss": 0.2898,
+      "step": 5518
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6945683518443198,
+      "learning_rate": 3.746030532333763e-06,
+      "loss": 0.479,
+      "step": 5519
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.697049079258694,
+      "learning_rate": 3.7443451789168235e-06,
+      "loss": 0.4951,
+      "step": 5520
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0770825695081974,
+      "learning_rate": 3.7426599777523577e-06,
+      "loss": 0.5099,
+      "step": 5521
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.559100435485521,
+      "learning_rate": 3.7409749290447062e-06,
+      "loss": 0.5051,
+      "step": 5522
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.6250425082189155,
+      "learning_rate": 3.7392900329981845e-06,
+      "loss": 0.4818,
+      "step": 5523
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.705463956119469,
+      "learning_rate": 3.7376052898170916e-06,
+      "loss": 0.4984,
+      "step": 5524
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.547889052823693,
+      "learning_rate": 3.735920699705711e-06,
+      "loss": 0.4767,
+      "step": 5525
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.818884099700887,
+      "learning_rate": 3.734236262868303e-06,
+      "loss": 0.5659,
+      "step": 5526
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.758401725611337,
+      "learning_rate": 3.732551979509113e-06,
+      "loss": 0.465,
+      "step": 5527
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9377414581858208,
+      "learning_rate": 3.7308678498323648e-06,
+      "loss": 0.4936,
+      "step": 5528
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.895434656259651,
+      "learning_rate": 3.7291838740422627e-06,
+      "loss": 0.5291,
+      "step": 5529
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.359082051045564,
+      "learning_rate": 3.7275000523429993e-06,
+      "loss": 0.2923,
+      "step": 5530
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.055205529278035,
+      "learning_rate": 3.725816384938741e-06,
+      "loss": 0.5213,
+      "step": 5531
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4365191989265083,
+      "learning_rate": 3.7241328720336377e-06,
+      "loss": 0.4617,
+      "step": 5532
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.450132867642605,
+      "learning_rate": 3.722449513831823e-06,
+      "loss": 0.4718,
+      "step": 5533
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.325993725913382,
+      "learning_rate": 3.7207663105374098e-06,
+      "loss": 0.4864,
+      "step": 5534
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.809899281192372,
+      "learning_rate": 3.7190832623544902e-06,
+      "loss": 0.49,
+      "step": 5535
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.991451260819146,
+      "learning_rate": 3.717400369487142e-06,
+      "loss": 0.5706,
+      "step": 5536
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.42042224524300925,
+      "learning_rate": 3.715717632139421e-06,
+      "loss": 0.2844,
+      "step": 5537
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.42647790615561304,
+      "learning_rate": 3.714035050515366e-06,
+      "loss": 0.2937,
+      "step": 5538
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.42397573711446546,
+      "learning_rate": 3.7123526248189935e-06,
+      "loss": 0.3132,
+      "step": 5539
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.152644120522345,
+      "learning_rate": 3.7106703552543026e-06,
+      "loss": 0.5195,
+      "step": 5540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9592378442169522,
+      "learning_rate": 3.7089882420252775e-06,
+      "loss": 0.4231,
+      "step": 5541
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7049558795687445,
+      "learning_rate": 3.7073062853358794e-06,
+      "loss": 0.5281,
+      "step": 5542
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.907764615572916,
+      "learning_rate": 3.7056244853900493e-06,
+      "loss": 0.4581,
+      "step": 5543
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1866251067435956,
+      "learning_rate": 3.7039428423917138e-06,
+      "loss": 0.5303,
+      "step": 5544
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0507021164045693,
+      "learning_rate": 3.7022613565447774e-06,
+      "loss": 0.4643,
+      "step": 5545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.67065302808788,
+      "learning_rate": 3.7005800280531244e-06,
+      "loss": 0.4977,
+      "step": 5546
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.2985624242694005,
+      "learning_rate": 3.6988988571206224e-06,
+      "loss": 0.4768,
+      "step": 5547
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3823886064530038,
+      "learning_rate": 3.697217843951121e-06,
+      "loss": 0.2842,
+      "step": 5548
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.840419068897809,
+      "learning_rate": 3.6955369887484476e-06,
+      "loss": 0.4897,
+      "step": 5549
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.109883830640453,
+      "learning_rate": 3.693856291716411e-06,
+      "loss": 0.5089,
+      "step": 5550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.2522428601722835,
+      "learning_rate": 3.692175753058802e-06,
+      "loss": 0.4587,
+      "step": 5551
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.681369681393401,
+      "learning_rate": 3.69049537297939e-06,
+      "loss": 0.4747,
+      "step": 5552
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.992760288143961,
+      "learning_rate": 3.6888151516819303e-06,
+      "loss": 0.5991,
+      "step": 5553
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.411707978638974,
+      "learning_rate": 3.6871350893701533e-06,
+      "loss": 0.5142,
+      "step": 5554
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.829057210017366,
+      "learning_rate": 3.6854551862477724e-06,
+      "loss": 0.4498,
+      "step": 5555
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7630372627661055,
+      "learning_rate": 3.6837754425184823e-06,
+      "loss": 0.511,
+      "step": 5556
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.935703799868562,
+      "learning_rate": 3.6820958583859585e-06,
+      "loss": 0.5162,
+      "step": 5557
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.509313279777048,
+      "learning_rate": 3.680416434053854e-06,
+      "loss": 0.4528,
+      "step": 5558
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.582250362937724,
+      "learning_rate": 3.678737169725809e-06,
+      "loss": 0.4597,
+      "step": 5559
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4042121873001418,
+      "learning_rate": 3.6770580656054343e-06,
+      "loss": 0.2968,
+      "step": 5560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9763104382148815,
+      "learning_rate": 3.6753791218963306e-06,
+      "loss": 0.4885,
+      "step": 5561
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.772146276183636,
+      "learning_rate": 3.673700338802076e-06,
+      "loss": 0.4815,
+      "step": 5562
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.34857990466704,
+      "learning_rate": 3.672021716526226e-06,
+      "loss": 0.4787,
+      "step": 5563
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.23478360675451,
+      "learning_rate": 3.670343255272322e-06,
+      "loss": 0.5264,
+      "step": 5564
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.39802815893155336,
+      "learning_rate": 3.668664955243883e-06,
+      "loss": 0.2828,
+      "step": 5565
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.713103154191567,
+      "learning_rate": 3.666986816644407e-06,
+      "loss": 0.4823,
+      "step": 5566
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6625065990630388,
+      "learning_rate": 3.665308839677375e-06,
+      "loss": 0.4631,
+      "step": 5567
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8305634103198365,
+      "learning_rate": 3.663631024546249e-06,
+      "loss": 0.4522,
+      "step": 5568
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6925072507705177,
+      "learning_rate": 3.6619533714544664e-06,
+      "loss": 0.519,
+      "step": 5569
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.060736677758069,
+      "learning_rate": 3.6602758806054537e-06,
+      "loss": 0.533,
+      "step": 5570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3750216456143893,
+      "learning_rate": 3.6585985522026067e-06,
+      "loss": 0.287,
+      "step": 5571
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0587446936741407,
+      "learning_rate": 3.6569213864493103e-06,
+      "loss": 0.4599,
+      "step": 5572
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.55237577162907,
+      "learning_rate": 3.655244383548926e-06,
+      "loss": 0.4894,
+      "step": 5573
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.38485480119253207,
+      "learning_rate": 3.6535675437047955e-06,
+      "loss": 0.2765,
+      "step": 5574
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.012824440866243,
+      "learning_rate": 3.651890867120243e-06,
+      "loss": 0.5008,
+      "step": 5575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.128873296073289,
+      "learning_rate": 3.6502143539985706e-06,
+      "loss": 0.539,
+      "step": 5576
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.1334860986816455,
+      "learning_rate": 3.64853800454306e-06,
+      "loss": 0.4876,
+      "step": 5577
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4719657998657976,
+      "learning_rate": 3.646861818956977e-06,
+      "loss": 0.53,
+      "step": 5578
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.730711120872087,
+      "learning_rate": 3.6451857974435635e-06,
+      "loss": 0.516,
+      "step": 5579
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4172081771251312,
+      "learning_rate": 3.6435099402060437e-06,
+      "loss": 0.2869,
+      "step": 5580
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.39826067553752015,
+      "learning_rate": 3.6418342474476186e-06,
+      "loss": 0.2652,
+      "step": 5581
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.225752726506289,
+      "learning_rate": 3.640158719371473e-06,
+      "loss": 0.5029,
+      "step": 5582
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4122277969497217,
+      "learning_rate": 3.638483356180771e-06,
+      "loss": 0.2695,
+      "step": 5583
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.220027602880508,
+      "learning_rate": 3.636808158078656e-06,
+      "loss": 0.5334,
+      "step": 5584
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7920569495232868,
+      "learning_rate": 3.635133125268252e-06,
+      "loss": 0.4534,
+      "step": 5585
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.38288994952100336,
+      "learning_rate": 3.633458257952661e-06,
+      "loss": 0.2867,
+      "step": 5586
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5315092207933914,
+      "learning_rate": 3.631783556334968e-06,
+      "loss": 0.4323,
+      "step": 5587
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2896241035796816,
+      "learning_rate": 3.6301090206182365e-06,
+      "loss": 0.4763,
+      "step": 5588
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.257143995634231,
+      "learning_rate": 3.628434651005507e-06,
+      "loss": 0.5255,
+      "step": 5589
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.497530314393498,
+      "learning_rate": 3.6267604476998063e-06,
+      "loss": 0.4739,
+      "step": 5590
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.309002068307091,
+      "learning_rate": 3.625086410904136e-06,
+      "loss": 0.4926,
+      "step": 5591
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8944233040918426,
+      "learning_rate": 3.623412540821478e-06,
+      "loss": 0.4764,
+      "step": 5592
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.111500527082817,
+      "learning_rate": 3.621738837654795e-06,
+      "loss": 0.4713,
+      "step": 5593
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 14.939274997365146,
+      "learning_rate": 3.620065301607029e-06,
+      "loss": 0.5289,
+      "step": 5594
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.723705232969561,
+      "learning_rate": 3.618391932881102e-06,
+      "loss": 0.5368,
+      "step": 5595
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.40071249759653493,
+      "learning_rate": 3.616718731679918e-06,
+      "loss": 0.2978,
+      "step": 5596
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3665854521944847,
+      "learning_rate": 3.6150456982063555e-06,
+      "loss": 0.4964,
+      "step": 5597
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4859605998588044,
+      "learning_rate": 3.6133728326632765e-06,
+      "loss": 0.5184,
+      "step": 5598
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.41927880957342406,
+      "learning_rate": 3.611700135253523e-06,
+      "loss": 0.2846,
+      "step": 5599
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.868354770984522,
+      "learning_rate": 3.6100276061799133e-06,
+      "loss": 0.5271,
+      "step": 5600
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0207998385003116,
+      "learning_rate": 3.6083552456452487e-06,
+      "loss": 0.4977,
+      "step": 5601
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.3672253657123,
+      "learning_rate": 3.60668305385231e-06,
+      "loss": 0.4675,
+      "step": 5602
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7159186252313536,
+      "learning_rate": 3.6050110310038533e-06,
+      "loss": 0.5074,
+      "step": 5603
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9148628717878733,
+      "learning_rate": 3.603339177302618e-06,
+      "loss": 0.4895,
+      "step": 5604
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.470055167442333,
+      "learning_rate": 3.6016674929513218e-06,
+      "loss": 0.5041,
+      "step": 5605
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.689871398239656,
+      "learning_rate": 3.599995978152664e-06,
+      "loss": 0.4611,
+      "step": 5606
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.6703183138247213,
+      "learning_rate": 3.5983246331093196e-06,
+      "loss": 0.4845,
+      "step": 5607
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4247508613468942,
+      "learning_rate": 3.5966534580239454e-06,
+      "loss": 0.2809,
+      "step": 5608
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.7384191325364675,
+      "learning_rate": 3.5949824530991783e-06,
+      "loss": 0.4744,
+      "step": 5609
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.007623971366464,
+      "learning_rate": 3.5933116185376325e-06,
+      "loss": 0.4939,
+      "step": 5610
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1912930576322185,
+      "learning_rate": 3.591640954541903e-06,
+      "loss": 0.5436,
+      "step": 5611
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.0203018188732695,
+      "learning_rate": 3.5899704613145637e-06,
+      "loss": 0.4875,
+      "step": 5612
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.993053778561335,
+      "learning_rate": 3.588300139058165e-06,
+      "loss": 0.4355,
+      "step": 5613
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.226441241672384,
+      "learning_rate": 3.586629987975243e-06,
+      "loss": 0.5005,
+      "step": 5614
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5906832959323216,
+      "learning_rate": 3.5849600082683068e-06,
+      "loss": 0.4641,
+      "step": 5615
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.39289115991247675,
+      "learning_rate": 3.5832902001398486e-06,
+      "loss": 0.281,
+      "step": 5616
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.6389279985829606,
+      "learning_rate": 3.581620563792336e-06,
+      "loss": 0.4828,
+      "step": 5617
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.490963422263885,
+      "learning_rate": 3.5799510994282215e-06,
+      "loss": 0.4937,
+      "step": 5618
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3917118096260863,
+      "learning_rate": 3.5782818072499313e-06,
+      "loss": 0.2914,
+      "step": 5619
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 83.11524992172643,
+      "learning_rate": 3.5766126874598734e-06,
+      "loss": 0.4588,
+      "step": 5620
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.309017592004988,
+      "learning_rate": 3.574943740260435e-06,
+      "loss": 0.5085,
+      "step": 5621
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3849290806854575,
+      "learning_rate": 3.57327496585398e-06,
+      "loss": 0.2832,
+      "step": 5622
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.420816831266826,
+      "learning_rate": 3.571606364442857e-06,
+      "loss": 0.504,
+      "step": 5623
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.13189908928867,
+      "learning_rate": 3.569937936229384e-06,
+      "loss": 0.4721,
+      "step": 5624
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.628643670331295,
+      "learning_rate": 3.568269681415866e-06,
+      "loss": 0.5087,
+      "step": 5625
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.651919886827424,
+      "learning_rate": 3.5666016002045857e-06,
+      "loss": 0.4572,
+      "step": 5626
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.840044790643522,
+      "learning_rate": 3.5649336927978036e-06,
+      "loss": 0.4968,
+      "step": 5627
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.184400831616412,
+      "learning_rate": 3.563265959397757e-06,
+      "loss": 0.5014,
+      "step": 5628
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3291117835184703,
+      "learning_rate": 3.561598400206667e-06,
+      "loss": 0.4572,
+      "step": 5629
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.580697659160208,
+      "learning_rate": 3.559931015426731e-06,
+      "loss": 0.4582,
+      "step": 5630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.315669693800664,
+      "learning_rate": 3.5582638052601227e-06,
+      "loss": 0.5052,
+      "step": 5631
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.109597638156066,
+      "learning_rate": 3.556596769908999e-06,
+      "loss": 0.5162,
+      "step": 5632
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.296065933775735,
+      "learning_rate": 3.554929909575494e-06,
+      "loss": 0.47,
+      "step": 5633
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.485393285130322,
+      "learning_rate": 3.5532632244617184e-06,
+      "loss": 0.5044,
+      "step": 5634
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.219900038241693,
+      "learning_rate": 3.551596714769765e-06,
+      "loss": 0.4359,
+      "step": 5635
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.733259673152027,
+      "learning_rate": 3.5499303807017018e-06,
+      "loss": 0.4514,
+      "step": 5636
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5191860492418496,
+      "learning_rate": 3.54826422245958e-06,
+      "loss": 0.4723,
+      "step": 5637
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6967454853832447,
+      "learning_rate": 3.546598240245427e-06,
+      "loss": 0.5174,
+      "step": 5638
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.40010205474711,
+      "learning_rate": 3.5449324342612464e-06,
+      "loss": 0.4267,
+      "step": 5639
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.857900753733776,
+      "learning_rate": 3.5432668047090258e-06,
+      "loss": 0.5,
+      "step": 5640
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.6933941460012925,
+      "learning_rate": 3.5416013517907265e-06,
+      "loss": 0.4971,
+      "step": 5641
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.3175412872425545,
+      "learning_rate": 3.5399360757082918e-06,
+      "loss": 0.4558,
+      "step": 5642
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3831776851517688,
+      "learning_rate": 3.5382709766636404e-06,
+      "loss": 0.2699,
+      "step": 5643
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.616179158681995,
+      "learning_rate": 3.5366060548586745e-06,
+      "loss": 0.4875,
+      "step": 5644
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.928807627584806,
+      "learning_rate": 3.5349413104952686e-06,
+      "loss": 0.4733,
+      "step": 5645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.812575794704909,
+      "learning_rate": 3.5332767437752792e-06,
+      "loss": 0.4617,
+      "step": 5646
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.078397503895924,
+      "learning_rate": 3.531612354900542e-06,
+      "loss": 0.466,
+      "step": 5647
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.533029490586084,
+      "learning_rate": 3.5299481440728673e-06,
+      "loss": 0.5193,
+      "step": 5648
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3789534573451653,
+      "learning_rate": 3.528284111494049e-06,
+      "loss": 0.2747,
+      "step": 5649
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.10608232384316,
+      "learning_rate": 3.526620257365857e-06,
+      "loss": 0.5256,
+      "step": 5650
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3769954611081766,
+      "learning_rate": 3.5249565818900365e-06,
+      "loss": 0.2947,
+      "step": 5651
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.4926409396849065,
+      "learning_rate": 3.5232930852683168e-06,
+      "loss": 0.4825,
+      "step": 5652
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2581919937361543,
+      "learning_rate": 3.521629767702401e-06,
+      "loss": 0.524,
+      "step": 5653
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.356286686602352,
+      "learning_rate": 3.5199666293939725e-06,
+      "loss": 0.5017,
+      "step": 5654
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.747744784420382,
+      "learning_rate": 3.518303670544696e-06,
+      "loss": 0.527,
+      "step": 5655
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.39511962513983384,
+      "learning_rate": 3.5166408913562033e-06,
+      "loss": 0.2992,
+      "step": 5656
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4059012167037531,
+      "learning_rate": 3.5149782920301178e-06,
+      "loss": 0.2873,
+      "step": 5657
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6182092901377394,
+      "learning_rate": 3.513315872768035e-06,
+      "loss": 0.4735,
+      "step": 5658
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.277622065786922,
+      "learning_rate": 3.5116536337715255e-06,
+      "loss": 0.5231,
+      "step": 5659
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.251439288226134,
+      "learning_rate": 3.5099915752421467e-06,
+      "loss": 0.4885,
+      "step": 5660
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.418178299969354,
+      "learning_rate": 3.5083296973814253e-06,
+      "loss": 0.4602,
+      "step": 5661
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.231294799926804,
+      "learning_rate": 3.5066680003908695e-06,
+      "loss": 0.4587,
+      "step": 5662
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.161038914581696,
+      "learning_rate": 3.5050064844719677e-06,
+      "loss": 0.4677,
+      "step": 5663
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3309733072460905,
+      "learning_rate": 3.503345149826185e-06,
+      "loss": 0.4648,
+      "step": 5664
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.923981023192934,
+      "learning_rate": 3.5016839966549626e-06,
+      "loss": 0.5283,
+      "step": 5665
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.043018573542302,
+      "learning_rate": 3.50002302515972e-06,
+      "loss": 0.5206,
+      "step": 5666
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.5422317434628585,
+      "learning_rate": 3.498362235541856e-06,
+      "loss": 0.4595,
+      "step": 5667
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.170400137356699,
+      "learning_rate": 3.4967016280027494e-06,
+      "loss": 0.4615,
+      "step": 5668
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.41695640627488,
+      "learning_rate": 3.4950412027437523e-06,
+      "loss": 0.4825,
+      "step": 5669
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.37105106583730113,
+      "learning_rate": 3.4933809599661973e-06,
+      "loss": 0.2758,
+      "step": 5670
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.38634190522914835,
+      "learning_rate": 3.491720899871396e-06,
+      "loss": 0.2841,
+      "step": 5671
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.692297441489497,
+      "learning_rate": 3.4900610226606346e-06,
+      "loss": 0.4724,
+      "step": 5672
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.558356077646476,
+      "learning_rate": 3.4884013285351796e-06,
+      "loss": 0.4785,
+      "step": 5673
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.893745392642539,
+      "learning_rate": 3.4867418176962755e-06,
+      "loss": 0.4865,
+      "step": 5674
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.96755533811278,
+      "learning_rate": 3.485082490345143e-06,
+      "loss": 0.5032,
+      "step": 5675
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.798644751679194,
+      "learning_rate": 3.4834233466829827e-06,
+      "loss": 0.5005,
+      "step": 5676
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.750780284947853,
+      "learning_rate": 3.4817643869109683e-06,
+      "loss": 0.5077,
+      "step": 5677
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9825884130946836,
+      "learning_rate": 3.4801056112302554e-06,
+      "loss": 0.5048,
+      "step": 5678
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.026962780667075,
+      "learning_rate": 3.4784470198419783e-06,
+      "loss": 0.466,
+      "step": 5679
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.749576377548013,
+      "learning_rate": 3.4767886129472455e-06,
+      "loss": 0.4934,
+      "step": 5680
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 16.30670780001252,
+      "learning_rate": 3.4751303907471444e-06,
+      "loss": 0.5012,
+      "step": 5681
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.080805954936274,
+      "learning_rate": 3.47347235344274e-06,
+      "loss": 0.5379,
+      "step": 5682
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.306274624608302,
+      "learning_rate": 3.471814501235076e-06,
+      "loss": 0.4396,
+      "step": 5683
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.0269813846112195,
+      "learning_rate": 3.4701568343251723e-06,
+      "loss": 0.4585,
+      "step": 5684
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.801138968950147,
+      "learning_rate": 3.468499352914026e-06,
+      "loss": 0.4786,
+      "step": 5685
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.021542629280829,
+      "learning_rate": 3.4668420572026158e-06,
+      "loss": 0.5059,
+      "step": 5686
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.0591021436924475,
+      "learning_rate": 3.4651849473918887e-06,
+      "loss": 0.4762,
+      "step": 5687
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.924464296860959,
+      "learning_rate": 3.463528023682779e-06,
+      "loss": 0.5036,
+      "step": 5688
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.341474363871247,
+      "learning_rate": 3.461871286276194e-06,
+      "loss": 0.4905,
+      "step": 5689
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.286766815647873,
+      "learning_rate": 3.4602147353730165e-06,
+      "loss": 0.542,
+      "step": 5690
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.17855908177564,
+      "learning_rate": 3.4585583711741115e-06,
+      "loss": 0.5009,
+      "step": 5691
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.078888623029554,
+      "learning_rate": 3.456902193880319e-06,
+      "loss": 0.5249,
+      "step": 5692
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.002505023922944,
+      "learning_rate": 3.455246203692454e-06,
+      "loss": 0.4746,
+      "step": 5693
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.971311498588449,
+      "learning_rate": 3.453590400811313e-06,
+      "loss": 0.5207,
+      "step": 5694
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8106877010971276,
+      "learning_rate": 3.4519347854376677e-06,
+      "loss": 0.487,
+      "step": 5695
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.731940408869642,
+      "learning_rate": 3.450279357772266e-06,
+      "loss": 0.5236,
+      "step": 5696
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7071201153824607,
+      "learning_rate": 3.4486241180158373e-06,
+      "loss": 0.5501,
+      "step": 5697
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7913954668693965,
+      "learning_rate": 3.44696906636908e-06,
+      "loss": 0.4999,
+      "step": 5698
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.492880362927079,
+      "learning_rate": 3.445314203032678e-06,
+      "loss": 0.5358,
+      "step": 5699
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.324045157952538,
+      "learning_rate": 3.443659528207289e-06,
+      "loss": 0.4537,
+      "step": 5700
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4083940339924218,
+      "learning_rate": 3.442005042093547e-06,
+      "loss": 0.2892,
+      "step": 5701
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0376744943016964,
+      "learning_rate": 3.440350744892066e-06,
+      "loss": 0.5375,
+      "step": 5702
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5939609530773833,
+      "learning_rate": 3.4386966368034335e-06,
+      "loss": 0.4812,
+      "step": 5703
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.402486842935717,
+      "learning_rate": 3.437042718028215e-06,
+      "loss": 0.5108,
+      "step": 5704
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.983148518247467,
+      "learning_rate": 3.435388988766958e-06,
+      "loss": 0.4824,
+      "step": 5705
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.09848787445105,
+      "learning_rate": 3.4337354492201786e-06,
+      "loss": 0.4696,
+      "step": 5706
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8924158141899037,
+      "learning_rate": 3.4320820995883776e-06,
+      "loss": 0.5619,
+      "step": 5707
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.838739656134047,
+      "learning_rate": 3.430428940072026e-06,
+      "loss": 0.4911,
+      "step": 5708
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.98072679592024,
+      "learning_rate": 3.4287759708715753e-06,
+      "loss": 0.5223,
+      "step": 5709
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.584674550111695,
+      "learning_rate": 3.427123192187456e-06,
+      "loss": 0.5102,
+      "step": 5710
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6983213439933937,
+      "learning_rate": 3.4254706042200724e-06,
+      "loss": 0.4767,
+      "step": 5711
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0530211345731657,
+      "learning_rate": 3.4238182071698066e-06,
+      "loss": 0.5059,
+      "step": 5712
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.947436827118052,
+      "learning_rate": 3.422166001237016e-06,
+      "loss": 0.4699,
+      "step": 5713
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5907636031312453,
+      "learning_rate": 3.4205139866220384e-06,
+      "loss": 0.4929,
+      "step": 5714
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.346342571547737,
+      "learning_rate": 3.418862163525185e-06,
+      "loss": 0.496,
+      "step": 5715
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.0145509787537845,
+      "learning_rate": 3.417210532146744e-06,
+      "loss": 0.4914,
+      "step": 5716
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.873769033754022,
+      "learning_rate": 3.415559092686984e-06,
+      "loss": 0.5227,
+      "step": 5717
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0213824327395384,
+      "learning_rate": 3.4139078453461472e-06,
+      "loss": 0.5009,
+      "step": 5718
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4547470583373743,
+      "learning_rate": 3.412256790324452e-06,
+      "loss": 0.5141,
+      "step": 5719
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.461122807992848,
+      "learning_rate": 3.4106059278220938e-06,
+      "loss": 0.458,
+      "step": 5720
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9701186281894354,
+      "learning_rate": 3.408955258039246e-06,
+      "loss": 0.4818,
+      "step": 5721
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.347522672521174,
+      "learning_rate": 3.407304781176059e-06,
+      "loss": 0.4643,
+      "step": 5722
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.965443939004649,
+      "learning_rate": 3.4056544974326584e-06,
+      "loss": 0.5436,
+      "step": 5723
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.990419032868023,
+      "learning_rate": 3.404004407009145e-06,
+      "loss": 0.4485,
+      "step": 5724
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8736257498233404,
+      "learning_rate": 3.402354510105601e-06,
+      "loss": 0.4861,
+      "step": 5725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.781820194640273,
+      "learning_rate": 3.4007048069220807e-06,
+      "loss": 0.504,
+      "step": 5726
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.9237645399914625,
+      "learning_rate": 3.3990552976586144e-06,
+      "loss": 0.4373,
+      "step": 5727
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.036450887097535,
+      "learning_rate": 3.397405982515214e-06,
+      "loss": 0.4809,
+      "step": 5728
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4017445164853799,
+      "learning_rate": 3.3957568616918644e-06,
+      "loss": 0.29,
+      "step": 5729
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.745263107438877,
+      "learning_rate": 3.3941079353885255e-06,
+      "loss": 0.4783,
+      "step": 5730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.276644879433859,
+      "learning_rate": 3.3924592038051345e-06,
+      "loss": 0.5109,
+      "step": 5731
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.319511869697513,
+      "learning_rate": 3.390810667141606e-06,
+      "loss": 0.4886,
+      "step": 5732
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8452870595889386,
+      "learning_rate": 3.3891623255978347e-06,
+      "loss": 0.5124,
+      "step": 5733
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.163680282807494,
+      "learning_rate": 3.387514179373683e-06,
+      "loss": 0.4975,
+      "step": 5734
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.9128167128482225,
+      "learning_rate": 3.385866228668996e-06,
+      "loss": 0.5174,
+      "step": 5735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.25393196931734,
+      "learning_rate": 3.3842184736835943e-06,
+      "loss": 0.4968,
+      "step": 5736
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.531168241314873,
+      "learning_rate": 3.382570914617273e-06,
+      "loss": 0.5371,
+      "step": 5737
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9516661139128444,
+      "learning_rate": 3.3809235516698045e-06,
+      "loss": 0.5183,
+      "step": 5738
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.511973290065537,
+      "learning_rate": 3.379276385040938e-06,
+      "loss": 0.5348,
+      "step": 5739
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.348934117535263,
+      "learning_rate": 3.3776294149303956e-06,
+      "loss": 0.5526,
+      "step": 5740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6866796755862707,
+      "learning_rate": 3.375982641537881e-06,
+      "loss": 0.4779,
+      "step": 5741
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3811923458502794,
+      "learning_rate": 3.3743360650630695e-06,
+      "loss": 0.2842,
+      "step": 5742
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3923740874817835,
+      "learning_rate": 3.3726896857056145e-06,
+      "loss": 0.2912,
+      "step": 5743
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.328070623503195,
+      "learning_rate": 3.3710435036651453e-06,
+      "loss": 0.4985,
+      "step": 5744
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.942519124367301,
+      "learning_rate": 3.369397519141267e-06,
+      "loss": 0.4738,
+      "step": 5745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.111718568959279,
+      "learning_rate": 3.3677517323335616e-06,
+      "loss": 0.4741,
+      "step": 5746
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.441493293936847,
+      "learning_rate": 3.3661061434415847e-06,
+      "loss": 0.5485,
+      "step": 5747
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.891888567001333,
+      "learning_rate": 3.3644607526648722e-06,
+      "loss": 0.4902,
+      "step": 5748
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5955247757885727,
+      "learning_rate": 3.3628155602029312e-06,
+      "loss": 0.5029,
+      "step": 5749
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.00881540242329,
+      "learning_rate": 3.3611705662552494e-06,
+      "loss": 0.5175,
+      "step": 5750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.855857144970629,
+      "learning_rate": 3.3595257710212853e-06,
+      "loss": 0.4806,
+      "step": 5751
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.315089068942223,
+      "learning_rate": 3.3578811747004757e-06,
+      "loss": 0.4949,
+      "step": 5752
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.517279344490831,
+      "learning_rate": 3.356236777492236e-06,
+      "loss": 0.4539,
+      "step": 5753
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.130531474777287,
+      "learning_rate": 3.3545925795959545e-06,
+      "loss": 0.4737,
+      "step": 5754
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.993035298199807,
+      "learning_rate": 3.3529485812109934e-06,
+      "loss": 0.5369,
+      "step": 5755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.518482538778584,
+      "learning_rate": 3.3513047825366968e-06,
+      "loss": 0.4669,
+      "step": 5756
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.195799813281315,
+      "learning_rate": 3.349661183772379e-06,
+      "loss": 0.4971,
+      "step": 5757
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4894048402957174,
+      "learning_rate": 3.3480177851173314e-06,
+      "loss": 0.4966,
+      "step": 5758
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.162624286216601,
+      "learning_rate": 3.346374586770823e-06,
+      "loss": 0.4553,
+      "step": 5759
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.230693276897137,
+      "learning_rate": 3.3447315889320985e-06,
+      "loss": 0.5178,
+      "step": 5760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.466678653293106,
+      "learning_rate": 3.343088791800374e-06,
+      "loss": 0.4738,
+      "step": 5761
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.404240835798909,
+      "learning_rate": 3.341446195574846e-06,
+      "loss": 0.5042,
+      "step": 5762
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.454159927989077,
+      "learning_rate": 3.339803800454684e-06,
+      "loss": 0.5671,
+      "step": 5763
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1615729141889557,
+      "learning_rate": 3.3381616066390355e-06,
+      "loss": 0.4897,
+      "step": 5764
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.458752473271114,
+      "learning_rate": 3.3365196143270207e-06,
+      "loss": 0.4944,
+      "step": 5765
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.784467878594411,
+      "learning_rate": 3.334877823717737e-06,
+      "loss": 0.4894,
+      "step": 5766
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.653176374125501,
+      "learning_rate": 3.3332362350102592e-06,
+      "loss": 0.4815,
+      "step": 5767
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5150471152370253,
+      "learning_rate": 3.3315948484036344e-06,
+      "loss": 0.5081,
+      "step": 5768
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.226999449853298,
+      "learning_rate": 3.3299536640968844e-06,
+      "loss": 0.5027,
+      "step": 5769
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.819665587290788,
+      "learning_rate": 3.328312682289011e-06,
+      "loss": 0.5164,
+      "step": 5770
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1429014424231463,
+      "learning_rate": 3.3266719031789896e-06,
+      "loss": 0.4817,
+      "step": 5771
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6984066742741346,
+      "learning_rate": 3.3250313269657673e-06,
+      "loss": 0.4675,
+      "step": 5772
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.1740044366471345,
+      "learning_rate": 3.3233909538482712e-06,
+      "loss": 0.4867,
+      "step": 5773
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.040125947849242,
+      "learning_rate": 3.321750784025401e-06,
+      "loss": 0.4785,
+      "step": 5774
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.224770016421799,
+      "learning_rate": 3.320110817696035e-06,
+      "loss": 0.4999,
+      "step": 5775
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.522351303924487,
+      "learning_rate": 3.3184710550590244e-06,
+      "loss": 0.5278,
+      "step": 5776
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.664204171677671,
+      "learning_rate": 3.316831496313195e-06,
+      "loss": 0.4917,
+      "step": 5777
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.611233360656905,
+      "learning_rate": 3.3151921416573486e-06,
+      "loss": 0.5184,
+      "step": 5778
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4113808291281571,
+      "learning_rate": 3.3135529912902644e-06,
+      "loss": 0.2765,
+      "step": 5779
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.467496740904918,
+      "learning_rate": 3.3119140454106945e-06,
+      "loss": 0.4559,
+      "step": 5780
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.208762724858853,
+      "learning_rate": 3.310275304217365e-06,
+      "loss": 0.4467,
+      "step": 5781
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.047903902691503,
+      "learning_rate": 3.3086367679089824e-06,
+      "loss": 0.5195,
+      "step": 5782
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.912566876706618,
+      "learning_rate": 3.306998436684221e-06,
+      "loss": 0.4921,
+      "step": 5783
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.941129915457737,
+      "learning_rate": 3.3053603107417365e-06,
+      "loss": 0.506,
+      "step": 5784
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.5452817565769905,
+      "learning_rate": 3.3037223902801565e-06,
+      "loss": 0.4324,
+      "step": 5785
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.617788073889263,
+      "learning_rate": 3.302084675498083e-06,
+      "loss": 0.5076,
+      "step": 5786
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.286139999811088,
+      "learning_rate": 3.3004471665940975e-06,
+      "loss": 0.4785,
+      "step": 5787
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.126980826982476,
+      "learning_rate": 3.2988098637667524e-06,
+      "loss": 0.4504,
+      "step": 5788
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9416404917286494,
+      "learning_rate": 3.297172767214576e-06,
+      "loss": 0.5128,
+      "step": 5789
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.914885698060897,
+      "learning_rate": 3.2955358771360725e-06,
+      "loss": 0.5125,
+      "step": 5790
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.806026106122407,
+      "learning_rate": 3.293899193729719e-06,
+      "loss": 0.4699,
+      "step": 5791
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8581540448236047,
+      "learning_rate": 3.2922627171939726e-06,
+      "loss": 0.4974,
+      "step": 5792
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3333795573933283,
+      "learning_rate": 3.2906264477272576e-06,
+      "loss": 0.4842,
+      "step": 5793
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.026445441704174,
+      "learning_rate": 3.288990385527978e-06,
+      "loss": 0.4696,
+      "step": 5794
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5231208276159562,
+      "learning_rate": 3.2873545307945144e-06,
+      "loss": 0.4894,
+      "step": 5795
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.059086261609559,
+      "learning_rate": 3.2857188837252175e-06,
+      "loss": 0.4868,
+      "step": 5796
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.411382541134298,
+      "learning_rate": 3.2840834445184156e-06,
+      "loss": 0.4873,
+      "step": 5797
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.837305640320435,
+      "learning_rate": 3.2824482133724124e-06,
+      "loss": 0.4815,
+      "step": 5798
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.587500652442501,
+      "learning_rate": 3.2808131904854843e-06,
+      "loss": 0.4878,
+      "step": 5799
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.040564223699885,
+      "learning_rate": 3.2791783760558836e-06,
+      "loss": 0.5406,
+      "step": 5800
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.40760458434000857,
+      "learning_rate": 3.277543770281838e-06,
+      "loss": 0.2968,
+      "step": 5801
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9421425380736252,
+      "learning_rate": 3.2759093733615482e-06,
+      "loss": 0.4815,
+      "step": 5802
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3782010686921698,
+      "learning_rate": 3.274275185493192e-06,
+      "loss": 0.2905,
+      "step": 5803
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6281068942772374,
+      "learning_rate": 3.272641206874918e-06,
+      "loss": 0.4981,
+      "step": 5804
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.383591486887252,
+      "learning_rate": 3.2710074377048516e-06,
+      "loss": 0.4723,
+      "step": 5805
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.470583400087601,
+      "learning_rate": 3.269373878181096e-06,
+      "loss": 0.5025,
+      "step": 5806
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 16.242346738488614,
+      "learning_rate": 3.2677405285017227e-06,
+      "loss": 0.5172,
+      "step": 5807
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.4401611527129985,
+      "learning_rate": 3.266107388864783e-06,
+      "loss": 0.5237,
+      "step": 5808
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3654648981369495,
+      "learning_rate": 3.264474459468299e-06,
+      "loss": 0.5524,
+      "step": 5809
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5404038970235523,
+      "learning_rate": 3.262841740510271e-06,
+      "loss": 0.4649,
+      "step": 5810
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.615532637197802,
+      "learning_rate": 3.2612092321886713e-06,
+      "loss": 0.4987,
+      "step": 5811
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3910887360995674,
+      "learning_rate": 3.259576934701445e-06,
+      "loss": 0.4641,
+      "step": 5812
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.41910559401946584,
+      "learning_rate": 3.2579448482465194e-06,
+      "loss": 0.2888,
+      "step": 5813
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.027265644642556,
+      "learning_rate": 3.256312973021783e-06,
+      "loss": 0.4824,
+      "step": 5814
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8649125242088647,
+      "learning_rate": 3.254681309225111e-06,
+      "loss": 0.4823,
+      "step": 5815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4243139345385372,
+      "learning_rate": 3.2530498570543477e-06,
+      "loss": 0.2968,
+      "step": 5816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.282419883015666,
+      "learning_rate": 3.2514186167073103e-06,
+      "loss": 0.5017,
+      "step": 5817
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5521586063784376,
+      "learning_rate": 3.2497875883817955e-06,
+      "loss": 0.4683,
+      "step": 5818
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6752551391453676,
+      "learning_rate": 3.248156772275569e-06,
+      "loss": 0.5077,
+      "step": 5819
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 14.787352210821135,
+      "learning_rate": 3.2465261685863723e-06,
+      "loss": 0.5115,
+      "step": 5820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.289935242139524,
+      "learning_rate": 3.244895777511925e-06,
+      "loss": 0.4472,
+      "step": 5821
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.0145249725220475,
+      "learning_rate": 3.243265599249914e-06,
+      "loss": 0.5011,
+      "step": 5822
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.118014055621992,
+      "learning_rate": 3.2416356339980056e-06,
+      "loss": 0.4877,
+      "step": 5823
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.543442171626445,
+      "learning_rate": 3.240005881953841e-06,
+      "loss": 0.4949,
+      "step": 5824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.72798579129477,
+      "learning_rate": 3.2383763433150277e-06,
+      "loss": 0.5232,
+      "step": 5825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.91223272097914,
+      "learning_rate": 3.236747018279157e-06,
+      "loss": 0.4371,
+      "step": 5826
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.855260489966852,
+      "learning_rate": 3.23511790704379e-06,
+      "loss": 0.5032,
+      "step": 5827
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.627737682161596,
+      "learning_rate": 3.23348900980646e-06,
+      "loss": 0.4893,
+      "step": 5828
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9820334233779864,
+      "learning_rate": 3.2318603267646774e-06,
+      "loss": 0.4919,
+      "step": 5829
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5356980909113314,
+      "learning_rate": 3.2302318581159268e-06,
+      "loss": 0.4626,
+      "step": 5830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.0327890800411526,
+      "learning_rate": 3.228603604057664e-06,
+      "loss": 0.5061,
+      "step": 5831
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.178321804664838,
+      "learning_rate": 3.226975564787322e-06,
+      "loss": 0.5041,
+      "step": 5832
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3965174482748247,
+      "learning_rate": 3.2253477405023047e-06,
+      "loss": 0.2886,
+      "step": 5833
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3500818338619878,
+      "learning_rate": 3.2237201313999927e-06,
+      "loss": 0.4925,
+      "step": 5834
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.315730945941062,
+      "learning_rate": 3.2220927376777393e-06,
+      "loss": 0.5045,
+      "step": 5835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.095987220775791,
+      "learning_rate": 3.2204655595328693e-06,
+      "loss": 0.4764,
+      "step": 5836
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.454525978856075,
+      "learning_rate": 3.2188385971626855e-06,
+      "loss": 0.5096,
+      "step": 5837
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3012301745709625,
+      "learning_rate": 3.2172118507644624e-06,
+      "loss": 0.5357,
+      "step": 5838
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.621460873838232,
+      "learning_rate": 3.215585320535449e-06,
+      "loss": 0.5243,
+      "step": 5839
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1310851551649606,
+      "learning_rate": 3.2139590066728667e-06,
+      "loss": 0.4733,
+      "step": 5840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.748241689424012,
+      "learning_rate": 3.212332909373912e-06,
+      "loss": 0.4999,
+      "step": 5841
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.122560192857993,
+      "learning_rate": 3.2107070288357557e-06,
+      "loss": 0.473,
+      "step": 5842
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.184225832354248,
+      "learning_rate": 3.2090813652555396e-06,
+      "loss": 0.4807,
+      "step": 5843
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.481400093280752,
+      "learning_rate": 3.207455918830384e-06,
+      "loss": 0.5192,
+      "step": 5844
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.569875046850805,
+      "learning_rate": 3.205830689757379e-06,
+      "loss": 0.4877,
+      "step": 5845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.039389057412207,
+      "learning_rate": 3.204205678233586e-06,
+      "loss": 0.5082,
+      "step": 5846
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6034711210863772,
+      "learning_rate": 3.202580884456047e-06,
+      "loss": 0.4927,
+      "step": 5847
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.077567668342028,
+      "learning_rate": 3.2009563086217703e-06,
+      "loss": 0.5244,
+      "step": 5848
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.32031538941273,
+      "learning_rate": 3.1993319509277455e-06,
+      "loss": 0.4964,
+      "step": 5849
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.009747770293923,
+      "learning_rate": 3.1977078115709285e-06,
+      "loss": 0.5029,
+      "step": 5850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.627468868896176,
+      "learning_rate": 3.1960838907482524e-06,
+      "loss": 0.5211,
+      "step": 5851
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.480657163986763,
+      "learning_rate": 3.194460188656624e-06,
+      "loss": 0.4872,
+      "step": 5852
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.054838711348235,
+      "learning_rate": 3.1928367054929234e-06,
+      "loss": 0.499,
+      "step": 5853
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.366409499328319,
+      "learning_rate": 3.1912134414540008e-06,
+      "loss": 0.5127,
+      "step": 5854
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7000098288704937,
+      "learning_rate": 3.1895903967366847e-06,
+      "loss": 0.4571,
+      "step": 5855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.638152667383693,
+      "learning_rate": 3.1879675715377766e-06,
+      "loss": 0.5112,
+      "step": 5856
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.475062692610507,
+      "learning_rate": 3.186344966054046e-06,
+      "loss": 0.4361,
+      "step": 5857
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8322180693288064,
+      "learning_rate": 3.1847225804822408e-06,
+      "loss": 0.4865,
+      "step": 5858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0547513926558842,
+      "learning_rate": 3.18310041501908e-06,
+      "loss": 0.4742,
+      "step": 5859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.411572722049329,
+      "learning_rate": 3.181478469861259e-06,
+      "loss": 0.4983,
+      "step": 5860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.119349799854223,
+      "learning_rate": 3.179856745205442e-06,
+      "loss": 0.4525,
+      "step": 5861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.027417267908709,
+      "learning_rate": 3.1782352412482693e-06,
+      "loss": 0.4658,
+      "step": 5862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.22032940200824,
+      "learning_rate": 3.176613958186355e-06,
+      "loss": 0.4849,
+      "step": 5863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.762456895260865,
+      "learning_rate": 3.1749928962162845e-06,
+      "loss": 0.4496,
+      "step": 5864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6468323342998072,
+      "learning_rate": 3.1733720555346158e-06,
+      "loss": 0.53,
+      "step": 5865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.675949877467838,
+      "learning_rate": 3.1717514363378864e-06,
+      "loss": 0.4937,
+      "step": 5866
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4214560714487188,
+      "learning_rate": 3.170131038822595e-06,
+      "loss": 0.293,
+      "step": 5867
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.728903925060021,
+      "learning_rate": 3.1685108631852244e-06,
+      "loss": 0.5101,
+      "step": 5868
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3174141594613515,
+      "learning_rate": 3.1668909096222254e-06,
+      "loss": 0.45,
+      "step": 5869
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.2170857485778095,
+      "learning_rate": 3.1652711783300234e-06,
+      "loss": 0.5523,
+      "step": 5870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.613004251610108,
+      "learning_rate": 3.163651669505017e-06,
+      "loss": 0.5053,
+      "step": 5871
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1897588276945816,
+      "learning_rate": 3.1620323833435763e-06,
+      "loss": 0.5605,
+      "step": 5872
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.660678039179464,
+      "learning_rate": 3.160413320042045e-06,
+      "loss": 0.4271,
+      "step": 5873
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8650068378357867,
+      "learning_rate": 3.15879447979674e-06,
+      "loss": 0.5103,
+      "step": 5874
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.210026762356835,
+      "learning_rate": 3.157175862803953e-06,
+      "loss": 0.4934,
+      "step": 5875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5659234382846132,
+      "learning_rate": 3.155557469259946e-06,
+      "loss": 0.4482,
+      "step": 5876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9457995710322593,
+      "learning_rate": 3.153939299360956e-06,
+      "loss": 0.568,
+      "step": 5877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0927286990856673,
+      "learning_rate": 3.1523213533031884e-06,
+      "loss": 0.482,
+      "step": 5878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.367533024394704,
+      "learning_rate": 3.150703631282826e-06,
+      "loss": 0.4648,
+      "step": 5879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8236212404973227,
+      "learning_rate": 3.149086133496025e-06,
+      "loss": 0.5195,
+      "step": 5880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.879365090078015,
+      "learning_rate": 3.1474688601389113e-06,
+      "loss": 0.4949,
+      "step": 5881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.642351075022468,
+      "learning_rate": 3.145851811407584e-06,
+      "loss": 0.4668,
+      "step": 5882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.53063098899685,
+      "learning_rate": 3.144234987498117e-06,
+      "loss": 0.5384,
+      "step": 5883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 19.852270201622822,
+      "learning_rate": 3.1426183886065565e-06,
+      "loss": 0.5035,
+      "step": 5884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.684801751405249,
+      "learning_rate": 3.141002014928918e-06,
+      "loss": 0.4575,
+      "step": 5885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2428829331769697,
+      "learning_rate": 3.1393858666611958e-06,
+      "loss": 0.4527,
+      "step": 5886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3335472969492295,
+      "learning_rate": 3.1377699439993525e-06,
+      "loss": 0.5153,
+      "step": 5887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 15.646790163926905,
+      "learning_rate": 3.1361542471393232e-06,
+      "loss": 0.4531,
+      "step": 5888
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7794518227834963,
+      "learning_rate": 3.1345387762770165e-06,
+      "loss": 0.497,
+      "step": 5889
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.851970948413009,
+      "learning_rate": 3.132923531608313e-06,
+      "loss": 0.4838,
+      "step": 5890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3915518159368867,
+      "learning_rate": 3.13130851332907e-06,
+      "loss": 0.294,
+      "step": 5891
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.138547638147508,
+      "learning_rate": 3.1296937216351115e-06,
+      "loss": 0.5063,
+      "step": 5892
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.148546626135569,
+      "learning_rate": 3.128079156722236e-06,
+      "loss": 0.544,
+      "step": 5893
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.754861611765595,
+      "learning_rate": 3.126464818786218e-06,
+      "loss": 0.4821,
+      "step": 5894
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.304304127552133,
+      "learning_rate": 3.124850708022799e-06,
+      "loss": 0.5194,
+      "step": 5895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.9405848653246225,
+      "learning_rate": 3.1232368246276956e-06,
+      "loss": 0.4621,
+      "step": 5896
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.7943099598869505,
+      "learning_rate": 3.121623168796598e-06,
+      "loss": 0.4902,
+      "step": 5897
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.209681099716648,
+      "learning_rate": 3.120009740725169e-06,
+      "loss": 0.4997,
+      "step": 5898
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 115.25956142837472,
+      "learning_rate": 3.118396540609038e-06,
+      "loss": 0.5121,
+      "step": 5899
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.642289389037368,
+      "learning_rate": 3.116783568643814e-06,
+      "loss": 0.4926,
+      "step": 5900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.487087680167752,
+      "learning_rate": 3.115170825025074e-06,
+      "loss": 0.4785,
+      "step": 5901
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1071752493094893,
+      "learning_rate": 3.1135583099483703e-06,
+      "loss": 0.4572,
+      "step": 5902
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.491703467988325,
+      "learning_rate": 3.1119460236092246e-06,
+      "loss": 0.4775,
+      "step": 5903
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4362048185657465,
+      "learning_rate": 3.1103339662031328e-06,
+      "loss": 0.2922,
+      "step": 5904
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.897915987664322,
+      "learning_rate": 3.1087221379255615e-06,
+      "loss": 0.5118,
+      "step": 5905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.366856491691781,
+      "learning_rate": 3.1071105389719523e-06,
+      "loss": 0.5136,
+      "step": 5906
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9026094287174806,
+      "learning_rate": 3.1054991695377156e-06,
+      "loss": 0.4835,
+      "step": 5907
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.087316388982521,
+      "learning_rate": 3.103888029818235e-06,
+      "loss": 0.4832,
+      "step": 5908
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.138005930932698,
+      "learning_rate": 3.102277120008871e-06,
+      "loss": 0.5118,
+      "step": 5909
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.256229732624393,
+      "learning_rate": 3.100666440304946e-06,
+      "loss": 0.5332,
+      "step": 5910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.132312909505318,
+      "learning_rate": 3.0990559909017635e-06,
+      "loss": 0.4817,
+      "step": 5911
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4009297866325381,
+      "learning_rate": 3.0974457719945953e-06,
+      "loss": 0.291,
+      "step": 5912
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.062000328791398,
+      "learning_rate": 3.095835783778686e-06,
+      "loss": 0.5341,
+      "step": 5913
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.645754147757515,
+      "learning_rate": 3.0942260264492537e-06,
+      "loss": 0.4923,
+      "step": 5914
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4052085810036208,
+      "learning_rate": 3.0926165002014854e-06,
+      "loss": 0.2985,
+      "step": 5915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0173109285750033,
+      "learning_rate": 3.091007205230541e-06,
+      "loss": 0.4372,
+      "step": 5916
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.232447695812431,
+      "learning_rate": 3.0893981417315556e-06,
+      "loss": 0.5369,
+      "step": 5917
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2494970690899896,
+      "learning_rate": 3.0877893098996324e-06,
+      "loss": 0.4937,
+      "step": 5918
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3941378119988227,
+      "learning_rate": 3.086180709929849e-06,
+      "loss": 0.2647,
+      "step": 5919
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.542869822577555,
+      "learning_rate": 3.0845723420172514e-06,
+      "loss": 0.5504,
+      "step": 5920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.283542424616454,
+      "learning_rate": 3.08296420635686e-06,
+      "loss": 0.5395,
+      "step": 5921
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8733012254916015,
+      "learning_rate": 3.0813563031436676e-06,
+      "loss": 0.4992,
+      "step": 5922
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8439007259318516,
+      "learning_rate": 3.079748632572639e-06,
+      "loss": 0.4686,
+      "step": 5923
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.161383656861599,
+      "learning_rate": 3.0781411948387074e-06,
+      "loss": 0.4957,
+      "step": 5924
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.8231598455806415,
+      "learning_rate": 3.0765339901367826e-06,
+      "loss": 0.484,
+      "step": 5925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.307056795661094,
+      "learning_rate": 3.074927018661743e-06,
+      "loss": 0.4784,
+      "step": 5926
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.885364852291333,
+      "learning_rate": 3.073320280608437e-06,
+      "loss": 0.5191,
+      "step": 5927
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.930417254148933,
+      "learning_rate": 3.071713776171692e-06,
+      "loss": 0.4715,
+      "step": 5928
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.385005083878641,
+      "learning_rate": 3.070107505546298e-06,
+      "loss": 0.5412,
+      "step": 5929
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.493023149684985,
+      "learning_rate": 3.0685014689270244e-06,
+      "loss": 0.4764,
+      "step": 5930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5028541725950846,
+      "learning_rate": 3.0668956665086047e-06,
+      "loss": 0.4825,
+      "step": 5931
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6089092352855308,
+      "learning_rate": 3.0652900984857492e-06,
+      "loss": 0.5529,
+      "step": 5932
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.623020596837828,
+      "learning_rate": 3.0636847650531404e-06,
+      "loss": 0.4937,
+      "step": 5933
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8616170247541035,
+      "learning_rate": 3.0620796664054294e-06,
+      "loss": 0.4513,
+      "step": 5934
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.398081896822031,
+      "learning_rate": 3.0604748027372393e-06,
+      "loss": 0.5231,
+      "step": 5935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.303786954679718,
+      "learning_rate": 3.0588701742431654e-06,
+      "loss": 0.4687,
+      "step": 5936
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.000360629351401,
+      "learning_rate": 3.057265781117776e-06,
+      "loss": 0.5021,
+      "step": 5937
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.350879771011658,
+      "learning_rate": 3.055661623555608e-06,
+      "loss": 0.479,
+      "step": 5938
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.634564096773649,
+      "learning_rate": 3.0540577017511703e-06,
+      "loss": 0.5135,
+      "step": 5939
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.40518168022832707,
+      "learning_rate": 3.0524540158989478e-06,
+      "loss": 0.277,
+      "step": 5940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.683932444199837,
+      "learning_rate": 3.0508505661933873e-06,
+      "loss": 0.5223,
+      "step": 5941
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.279109343066478,
+      "learning_rate": 3.049247352828917e-06,
+      "loss": 0.4903,
+      "step": 5942
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.647221335368545,
+      "learning_rate": 3.0476443759999296e-06,
+      "loss": 0.499,
+      "step": 5943
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.447147068601716,
+      "learning_rate": 3.0460416359007917e-06,
+      "loss": 0.5653,
+      "step": 5944
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.472603309242972,
+      "learning_rate": 3.0444391327258425e-06,
+      "loss": 0.4909,
+      "step": 5945
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4006705319274804,
+      "learning_rate": 3.0428368666693907e-06,
+      "loss": 0.2841,
+      "step": 5946
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2090842063183174,
+      "learning_rate": 3.041234837925715e-06,
+      "loss": 0.5074,
+      "step": 5947
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5421723434832884,
+      "learning_rate": 3.039633046689069e-06,
+      "loss": 0.466,
+      "step": 5948
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.874370253538772,
+      "learning_rate": 3.038031493153675e-06,
+      "loss": 0.5091,
+      "step": 5949
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3936428804586366,
+      "learning_rate": 3.0364301775137246e-06,
+      "loss": 0.2988,
+      "step": 5950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.288995253826903,
+      "learning_rate": 3.0348290999633877e-06,
+      "loss": 0.5243,
+      "step": 5951
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3548447629292433,
+      "learning_rate": 3.033228260696795e-06,
+      "loss": 0.5231,
+      "step": 5952
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.787780176169132,
+      "learning_rate": 3.031627659908057e-06,
+      "loss": 0.4828,
+      "step": 5953
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.39337669381781953,
+      "learning_rate": 3.0300272977912516e-06,
+      "loss": 0.2969,
+      "step": 5954
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.742980235845788,
+      "learning_rate": 3.028427174540426e-06,
+      "loss": 0.426,
+      "step": 5955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.316101258628836,
+      "learning_rate": 3.0268272903496036e-06,
+      "loss": 0.5564,
+      "step": 5956
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.404912750908363,
+      "learning_rate": 3.0252276454127753e-06,
+      "loss": 0.4736,
+      "step": 5957
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3218030979960536,
+      "learning_rate": 3.023628239923902e-06,
+      "loss": 0.4168,
+      "step": 5958
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.40693894664879143,
+      "learning_rate": 3.0220290740769193e-06,
+      "loss": 0.2724,
+      "step": 5959
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.022640212925349,
+      "learning_rate": 3.0204301480657306e-06,
+      "loss": 0.4605,
+      "step": 5960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8074829105400347,
+      "learning_rate": 3.0188314620842097e-06,
+      "loss": 0.5044,
+      "step": 5961
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.389856581372792,
+      "learning_rate": 3.0172330163262072e-06,
+      "loss": 0.2733,
+      "step": 5962
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9571885169700742,
+      "learning_rate": 3.0156348109855337e-06,
+      "loss": 0.4867,
+      "step": 5963
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.393015644647228,
+      "learning_rate": 3.014036846255982e-06,
+      "loss": 0.4943,
+      "step": 5964
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.176973230489788,
+      "learning_rate": 3.0124391223313093e-06,
+      "loss": 0.4984,
+      "step": 5965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.060822397453157,
+      "learning_rate": 3.0108416394052443e-06,
+      "loss": 0.5201,
+      "step": 5966
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1322834574974037,
+      "learning_rate": 3.009244397671489e-06,
+      "loss": 0.477,
+      "step": 5967
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4087188904186856,
+      "learning_rate": 3.007647397323714e-06,
+      "loss": 0.2834,
+      "step": 5968
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.026054170155776,
+      "learning_rate": 3.00605063855556e-06,
+      "loss": 0.5049,
+      "step": 5969
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.238170316122692,
+      "learning_rate": 3.0044541215606398e-06,
+      "loss": 0.4519,
+      "step": 5970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.42391078924816583,
+      "learning_rate": 3.002857846532538e-06,
+      "loss": 0.2937,
+      "step": 5971
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 18.351785137400196,
+      "learning_rate": 3.0012618136648085e-06,
+      "loss": 0.5178,
+      "step": 5972
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.425239682680217,
+      "learning_rate": 2.9996660231509742e-06,
+      "loss": 0.4699,
+      "step": 5973
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.925532028283065,
+      "learning_rate": 2.9980704751845302e-06,
+      "loss": 0.4846,
+      "step": 5974
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.09312346791429,
+      "learning_rate": 2.996475169958942e-06,
+      "loss": 0.5468,
+      "step": 5975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.118641204620175,
+      "learning_rate": 2.9948801076676477e-06,
+      "loss": 0.5057,
+      "step": 5976
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9372516833291087,
+      "learning_rate": 2.993285288504053e-06,
+      "loss": 0.5052,
+      "step": 5977
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.12156069173019,
+      "learning_rate": 2.9916907126615343e-06,
+      "loss": 0.5046,
+      "step": 5978
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5655082182856526,
+      "learning_rate": 2.9900963803334417e-06,
+      "loss": 0.4664,
+      "step": 5979
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.77310013817278,
+      "learning_rate": 2.9885022917130914e-06,
+      "loss": 0.4944,
+      "step": 5980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.954736303843322,
+      "learning_rate": 2.986908446993772e-06,
+      "loss": 0.4757,
+      "step": 5981
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.535445626885491,
+      "learning_rate": 2.9853148463687455e-06,
+      "loss": 0.4966,
+      "step": 5982
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7685333276373916,
+      "learning_rate": 2.9837214900312396e-06,
+      "loss": 0.4858,
+      "step": 5983
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3833883869373951,
+      "learning_rate": 2.9821283781744537e-06,
+      "loss": 0.2767,
+      "step": 5984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.396732067317747,
+      "learning_rate": 2.980535510991558e-06,
+      "loss": 0.5157,
+      "step": 5985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.260820914277313,
+      "learning_rate": 2.978942888675693e-06,
+      "loss": 0.4925,
+      "step": 5986
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6144911837454643,
+      "learning_rate": 2.977350511419972e-06,
+      "loss": 0.4898,
+      "step": 5987
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.46931615426001,
+      "learning_rate": 2.9757583794174744e-06,
+      "loss": 0.4614,
+      "step": 5988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.827700211646617,
+      "learning_rate": 2.9741664928612506e-06,
+      "loss": 0.462,
+      "step": 5989
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.793066175326051,
+      "learning_rate": 2.972574851944325e-06,
+      "loss": 0.517,
+      "step": 5990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.33253615832244,
+      "learning_rate": 2.9709834568596873e-06,
+      "loss": 0.4494,
+      "step": 5991
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.555349445687083,
+      "learning_rate": 2.9693923078003005e-06,
+      "loss": 0.4653,
+      "step": 5992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7835183851376186,
+      "learning_rate": 2.9678014049591e-06,
+      "loss": 0.5027,
+      "step": 5993
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.821872949453576,
+      "learning_rate": 2.9662107485289815e-06,
+      "loss": 0.4577,
+      "step": 5994
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.130640235808893,
+      "learning_rate": 2.964620338702823e-06,
+      "loss": 0.5133,
+      "step": 5995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.4045299973958505,
+      "learning_rate": 2.9630301756734646e-06,
+      "loss": 0.5143,
+      "step": 5996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.22126632486183,
+      "learning_rate": 2.96144025963372e-06,
+      "loss": 0.4635,
+      "step": 5997
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.896262859476694,
+      "learning_rate": 2.959850590776372e-06,
+      "loss": 0.4728,
+      "step": 5998
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.21209770715149,
+      "learning_rate": 2.958261169294174e-06,
+      "loss": 0.4713,
+      "step": 5999
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.350689362624568,
+      "learning_rate": 2.9566719953798474e-06,
+      "loss": 0.4468,
+      "step": 6000
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.67678343579036,
+      "learning_rate": 2.9550830692260856e-06,
+      "loss": 0.4645,
+      "step": 6001
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.517692251363455,
+      "learning_rate": 2.953494391025552e-06,
+      "loss": 0.5457,
+      "step": 6002
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9955564458614155,
+      "learning_rate": 2.951905960970879e-06,
+      "loss": 0.5505,
+      "step": 6003
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.74427606993088,
+      "learning_rate": 2.95031777925467e-06,
+      "loss": 0.4836,
+      "step": 6004
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.816959545505191,
+      "learning_rate": 2.9487298460694955e-06,
+      "loss": 0.498,
+      "step": 6005
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.271943277672497,
+      "learning_rate": 2.9471421616078987e-06,
+      "loss": 0.4723,
+      "step": 6006
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.796807173909205,
+      "learning_rate": 2.9455547260623926e-06,
+      "loss": 0.4722,
+      "step": 6007
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1101964295810047,
+      "learning_rate": 2.9439675396254586e-06,
+      "loss": 0.5151,
+      "step": 6008
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.009579160713503,
+      "learning_rate": 2.9423806024895475e-06,
+      "loss": 0.5069,
+      "step": 6009
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.887063336792003,
+      "learning_rate": 2.9407939148470834e-06,
+      "loss": 0.479,
+      "step": 6010
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4035259244236742,
+      "learning_rate": 2.939207476890456e-06,
+      "loss": 0.2814,
+      "step": 6011
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.964185835627431,
+      "learning_rate": 2.937621288812026e-06,
+      "loss": 0.5171,
+      "step": 6012
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.223715473386748,
+      "learning_rate": 2.9360353508041257e-06,
+      "loss": 0.4739,
+      "step": 6013
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.68686198276826,
+      "learning_rate": 2.9344496630590548e-06,
+      "loss": 0.5066,
+      "step": 6014
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.553955761920698,
+      "learning_rate": 2.9328642257690835e-06,
+      "loss": 0.5026,
+      "step": 6015
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.028496722634013,
+      "learning_rate": 2.931279039126451e-06,
+      "loss": 0.5154,
+      "step": 6016
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3962324601396348,
+      "learning_rate": 2.9296941033233662e-06,
+      "loss": 0.4832,
+      "step": 6017
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.674811966998368,
+      "learning_rate": 2.9281094185520097e-06,
+      "loss": 0.4822,
+      "step": 6018
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9796677047983446,
+      "learning_rate": 2.9265249850045284e-06,
+      "loss": 0.513,
+      "step": 6019
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.509446195080455,
+      "learning_rate": 2.924940802873041e-06,
+      "loss": 0.4587,
+      "step": 6020
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.184147592666508,
+      "learning_rate": 2.9233568723496354e-06,
+      "loss": 0.5038,
+      "step": 6021
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.487595694660274,
+      "learning_rate": 2.9217731936263683e-06,
+      "loss": 0.4359,
+      "step": 6022
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.050876322318151,
+      "learning_rate": 2.9201897668952655e-06,
+      "loss": 0.4784,
+      "step": 6023
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.41438432641257084,
+      "learning_rate": 2.918606592348324e-06,
+      "loss": 0.2838,
+      "step": 6024
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4085018780458124,
+      "learning_rate": 2.9170236701775113e-06,
+      "loss": 0.2773,
+      "step": 6025
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.004031030561373,
+      "learning_rate": 2.9154410005747586e-06,
+      "loss": 0.4534,
+      "step": 6026
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5401947751621616,
+      "learning_rate": 2.9138585837319694e-06,
+      "loss": 0.4743,
+      "step": 6027
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.521658510429005,
+      "learning_rate": 2.912276419841019e-06,
+      "loss": 0.5138,
+      "step": 6028
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5371997442072454,
+      "learning_rate": 2.9106945090937524e-06,
+      "loss": 0.516,
+      "step": 6029
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.425192925156674,
+      "learning_rate": 2.9091128516819767e-06,
+      "loss": 0.5024,
+      "step": 6030
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7977668906491973,
+      "learning_rate": 2.9075314477974764e-06,
+      "loss": 0.516,
+      "step": 6031
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.93869584486166,
+      "learning_rate": 2.905950297632004e-06,
+      "loss": 0.4875,
+      "step": 6032
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.0303605293227225,
+      "learning_rate": 2.9043694013772756e-06,
+      "loss": 0.4423,
+      "step": 6033
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.393675648864306,
+      "learning_rate": 2.9027887592249817e-06,
+      "loss": 0.5293,
+      "step": 6034
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8937948731020904,
+      "learning_rate": 2.9012083713667833e-06,
+      "loss": 0.4832,
+      "step": 6035
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.800594509853555,
+      "learning_rate": 2.8996282379943053e-06,
+      "loss": 0.5141,
+      "step": 6036
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3478374061042326,
+      "learning_rate": 2.8980483592991426e-06,
+      "loss": 0.4838,
+      "step": 6037
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1221401256490298,
+      "learning_rate": 2.896468735472863e-06,
+      "loss": 0.492,
+      "step": 6038
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.134799217120893,
+      "learning_rate": 2.894889366707002e-06,
+      "loss": 0.4963,
+      "step": 6039
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4180019859996644,
+      "learning_rate": 2.8933102531930622e-06,
+      "loss": 0.5492,
+      "step": 6040
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.135717503845526,
+      "learning_rate": 2.8917313951225164e-06,
+      "loss": 0.4674,
+      "step": 6041
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.355342512275825,
+      "learning_rate": 2.8901527926868088e-06,
+      "loss": 0.4901,
+      "step": 6042
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6421451924071917,
+      "learning_rate": 2.888574446077348e-06,
+      "loss": 0.4832,
+      "step": 6043
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.655815051015024,
+      "learning_rate": 2.886996355485514e-06,
+      "loss": 0.5235,
+      "step": 6044
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.021643228695073,
+      "learning_rate": 2.885418521102659e-06,
+      "loss": 0.4988,
+      "step": 6045
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.30292034370486,
+      "learning_rate": 2.8838409431200976e-06,
+      "loss": 0.4718,
+      "step": 6046
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.873006450760716,
+      "learning_rate": 2.882263621729116e-06,
+      "loss": 0.5005,
+      "step": 6047
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5488397318645855,
+      "learning_rate": 2.880686557120973e-06,
+      "loss": 0.4578,
+      "step": 6048
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7785073800234703,
+      "learning_rate": 2.8791097494868896e-06,
+      "loss": 0.4665,
+      "step": 6049
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.676561961336602,
+      "learning_rate": 2.877533199018061e-06,
+      "loss": 0.4798,
+      "step": 6050
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.515416689994906,
+      "learning_rate": 2.875956905905651e-06,
+      "loss": 0.4767,
+      "step": 6051
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.947947804477801,
+      "learning_rate": 2.8743808703407866e-06,
+      "loss": 0.5202,
+      "step": 6052
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3920184319140358,
+      "learning_rate": 2.87280509251457e-06,
+      "loss": 0.2553,
+      "step": 6053
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.525923085297504,
+      "learning_rate": 2.8712295726180717e-06,
+      "loss": 0.4971,
+      "step": 6054
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.617060502273378,
+      "learning_rate": 2.8696543108423246e-06,
+      "loss": 0.4941,
+      "step": 6055
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3978271923892748,
+      "learning_rate": 2.8680793073783363e-06,
+      "loss": 0.2772,
+      "step": 6056
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6188035882597624,
+      "learning_rate": 2.866504562417086e-06,
+      "loss": 0.5041,
+      "step": 6057
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3865316090663398,
+      "learning_rate": 2.864930076149509e-06,
+      "loss": 0.4978,
+      "step": 6058
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.106236164278532,
+      "learning_rate": 2.8633558487665215e-06,
+      "loss": 0.4666,
+      "step": 6059
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1501183490290496,
+      "learning_rate": 2.861781880459005e-06,
+      "loss": 0.4517,
+      "step": 6060
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 16.355878978726988,
+      "learning_rate": 2.8602081714178043e-06,
+      "loss": 0.4617,
+      "step": 6061
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.187014921247909,
+      "learning_rate": 2.8586347218337406e-06,
+      "loss": 0.5361,
+      "step": 6062
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.309055416434983,
+      "learning_rate": 2.8570615318976013e-06,
+      "loss": 0.4756,
+      "step": 6063
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8658345898591286,
+      "learning_rate": 2.855488601800137e-06,
+      "loss": 0.4628,
+      "step": 6064
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.088051792218449,
+      "learning_rate": 2.853915931732073e-06,
+      "loss": 0.4841,
+      "step": 6065
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8594351551796438,
+      "learning_rate": 2.852343521884103e-06,
+      "loss": 0.5343,
+      "step": 6066
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.008546052713982,
+      "learning_rate": 2.850771372446884e-06,
+      "loss": 0.5366,
+      "step": 6067
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.836400001644415,
+      "learning_rate": 2.8491994836110454e-06,
+      "loss": 0.4991,
+      "step": 6068
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.38937936435769993,
+      "learning_rate": 2.8476278555671834e-06,
+      "loss": 0.2771,
+      "step": 6069
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.407942048100618,
+      "learning_rate": 2.846056488505866e-06,
+      "loss": 0.4875,
+      "step": 6070
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.111257361543985,
+      "learning_rate": 2.844485382617624e-06,
+      "loss": 0.4821,
+      "step": 6071
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.40893976976913,
+      "learning_rate": 2.84291453809296e-06,
+      "loss": 0.5049,
+      "step": 6072
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.888169721128463,
+      "learning_rate": 2.8413439551223472e-06,
+      "loss": 0.4508,
+      "step": 6073
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.846546133518774,
+      "learning_rate": 2.839773633896219e-06,
+      "loss": 0.506,
+      "step": 6074
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3748978750719445,
+      "learning_rate": 2.8382035746049863e-06,
+      "loss": 0.4625,
+      "step": 6075
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6159704517165636,
+      "learning_rate": 2.836633777439024e-06,
+      "loss": 0.502,
+      "step": 6076
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2490721526244313,
+      "learning_rate": 2.8350642425886734e-06,
+      "loss": 0.532,
+      "step": 6077
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4389598055412063,
+      "learning_rate": 2.833494970244248e-06,
+      "loss": 0.4769,
+      "step": 6078
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7115969920590213,
+      "learning_rate": 2.831925960596027e-06,
+      "loss": 0.512,
+      "step": 6079
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.400699875354969,
+      "learning_rate": 2.830357213834256e-06,
+      "loss": 0.4878,
+      "step": 6080
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.800320927307934,
+      "learning_rate": 2.8287887301491522e-06,
+      "loss": 0.4977,
+      "step": 6081
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.606046600526665,
+      "learning_rate": 2.827220509730903e-06,
+      "loss": 0.4998,
+      "step": 6082
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 18.3988795392786,
+      "learning_rate": 2.8256525527696554e-06,
+      "loss": 0.5102,
+      "step": 6083
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.110328630315307,
+      "learning_rate": 2.8240848594555314e-06,
+      "loss": 0.5059,
+      "step": 6084
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.327828184497507,
+      "learning_rate": 2.822517429978622e-06,
+      "loss": 0.4899,
+      "step": 6085
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.670815642303389,
+      "learning_rate": 2.8209502645289795e-06,
+      "loss": 0.4568,
+      "step": 6086
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9786969074866834,
+      "learning_rate": 2.8193833632966296e-06,
+      "loss": 0.4574,
+      "step": 6087
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 25.544720341917113,
+      "learning_rate": 2.817816726471566e-06,
+      "loss": 0.5255,
+      "step": 6088
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3991277185916078,
+      "learning_rate": 2.816250354243748e-06,
+      "loss": 0.2857,
+      "step": 6089
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.290121809338533,
+      "learning_rate": 2.814684246803101e-06,
+      "loss": 0.4492,
+      "step": 6090
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6465000460099968,
+      "learning_rate": 2.8131184043395256e-06,
+      "loss": 0.4628,
+      "step": 6091
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.190365894892974,
+      "learning_rate": 2.8115528270428798e-06,
+      "loss": 0.508,
+      "step": 6092
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.952507313014563,
+      "learning_rate": 2.8099875151029987e-06,
+      "loss": 0.4428,
+      "step": 6093
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2980808568280557,
+      "learning_rate": 2.808422468709684e-06,
+      "loss": 0.4907,
+      "step": 6094
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8225128570411475,
+      "learning_rate": 2.8068576880526986e-06,
+      "loss": 0.4811,
+      "step": 6095
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.400754792147991,
+      "learning_rate": 2.805293173321779e-06,
+      "loss": 0.5226,
+      "step": 6096
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.643420146495444,
+      "learning_rate": 2.8037289247066307e-06,
+      "loss": 0.5554,
+      "step": 6097
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.05130966235618,
+      "learning_rate": 2.8021649423969198e-06,
+      "loss": 0.5006,
+      "step": 6098
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7967242745769383,
+      "learning_rate": 2.800601226582289e-06,
+      "loss": 0.5174,
+      "step": 6099
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9532654181881624,
+      "learning_rate": 2.7990377774523397e-06,
+      "loss": 0.4378,
+      "step": 6100
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.823091934915752,
+      "learning_rate": 2.79747459519665e-06,
+      "loss": 0.4639,
+      "step": 6101
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.262640377502838,
+      "learning_rate": 2.795911680004757e-06,
+      "loss": 0.4593,
+      "step": 6102
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.398093902670561,
+      "learning_rate": 2.7943490320661703e-06,
+      "loss": 0.4863,
+      "step": 6103
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7196069387193003,
+      "learning_rate": 2.7927866515703705e-06,
+      "loss": 0.4488,
+      "step": 6104
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.269729674811963,
+      "learning_rate": 2.7912245387067972e-06,
+      "loss": 0.5163,
+      "step": 6105
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8771229886347096,
+      "learning_rate": 2.7896626936648622e-06,
+      "loss": 0.4854,
+      "step": 6106
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0135307809922964,
+      "learning_rate": 2.7881011166339485e-06,
+      "loss": 0.4853,
+      "step": 6107
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 18.596706232014096,
+      "learning_rate": 2.7865398078033976e-06,
+      "loss": 0.5249,
+      "step": 6108
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.88594490150374,
+      "learning_rate": 2.784978767362528e-06,
+      "loss": 0.5098,
+      "step": 6109
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.970975429013034,
+      "learning_rate": 2.783417995500619e-06,
+      "loss": 0.5277,
+      "step": 6110
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.828104586988451,
+      "learning_rate": 2.7818574924069184e-06,
+      "loss": 0.451,
+      "step": 6111
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.440551191471079,
+      "learning_rate": 2.7802972582706434e-06,
+      "loss": 0.5178,
+      "step": 6112
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.40129874496941387,
+      "learning_rate": 2.778737293280981e-06,
+      "loss": 0.2843,
+      "step": 6113
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.057698118037187,
+      "learning_rate": 2.777177597627077e-06,
+      "loss": 0.5026,
+      "step": 6114
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.39609970987366455,
+      "learning_rate": 2.7756181714980525e-06,
+      "loss": 0.2694,
+      "step": 6115
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.402463304322569,
+      "learning_rate": 2.774059015082996e-06,
+      "loss": 0.468,
+      "step": 6116
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.057270286359847,
+      "learning_rate": 2.772500128570955e-06,
+      "loss": 0.4909,
+      "step": 6117
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5091614787922616,
+      "learning_rate": 2.770941512150953e-06,
+      "loss": 0.4905,
+      "step": 6118
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.763849019004615,
+      "learning_rate": 2.76938316601198e-06,
+      "loss": 0.4548,
+      "step": 6119
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.80520595245157,
+      "learning_rate": 2.7678250903429875e-06,
+      "loss": 0.5096,
+      "step": 6120
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.741595858035602,
+      "learning_rate": 2.7662672853328966e-06,
+      "loss": 0.458,
+      "step": 6121
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.451725138948839,
+      "learning_rate": 2.7647097511706e-06,
+      "loss": 0.4897,
+      "step": 6122
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8133400886273927,
+      "learning_rate": 2.7631524880449513e-06,
+      "loss": 0.5502,
+      "step": 6123
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.151562914212076,
+      "learning_rate": 2.7615954961447733e-06,
+      "loss": 0.4718,
+      "step": 6124
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5235633304046146,
+      "learning_rate": 2.7600387756588616e-06,
+      "loss": 0.5277,
+      "step": 6125
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.233346906938439,
+      "learning_rate": 2.7584823267759686e-06,
+      "loss": 0.4427,
+      "step": 6126
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6351345313447796,
+      "learning_rate": 2.7569261496848203e-06,
+      "loss": 0.5363,
+      "step": 6127
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.39508591907325497,
+      "learning_rate": 2.7553702445741125e-06,
+      "loss": 0.2785,
+      "step": 6128
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4175458379881087,
+      "learning_rate": 2.753814611632498e-06,
+      "loss": 0.4408,
+      "step": 6129
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.974874980438178,
+      "learning_rate": 2.752259251048606e-06,
+      "loss": 0.4931,
+      "step": 6130
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2977265448224182,
+      "learning_rate": 2.750704163011031e-06,
+      "loss": 0.4807,
+      "step": 6131
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.040276535080275,
+      "learning_rate": 2.749149347708331e-06,
+      "loss": 0.504,
+      "step": 6132
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.088023401000323,
+      "learning_rate": 2.7475948053290302e-06,
+      "loss": 0.4944,
+      "step": 6133
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.12309542813999,
+      "learning_rate": 2.7460405360616243e-06,
+      "loss": 0.5047,
+      "step": 6134
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.537700978548845,
+      "learning_rate": 2.7444865400945763e-06,
+      "loss": 0.4895,
+      "step": 6135
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6498229663803703,
+      "learning_rate": 2.7429328176163088e-06,
+      "loss": 0.4928,
+      "step": 6136
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.638081535671068,
+      "learning_rate": 2.741379368815219e-06,
+      "loss": 0.468,
+      "step": 6137
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3559311717652456,
+      "learning_rate": 2.7398261938796688e-06,
+      "loss": 0.4654,
+      "step": 6138
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.279835490547733,
+      "learning_rate": 2.7382732929979833e-06,
+      "loss": 0.4701,
+      "step": 6139
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.91925321842847,
+      "learning_rate": 2.7367206663584578e-06,
+      "loss": 0.5143,
+      "step": 6140
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.824221328155054,
+      "learning_rate": 2.7351683141493566e-06,
+      "loss": 0.4564,
+      "step": 6141
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.696592887307569,
+      "learning_rate": 2.733616236558905e-06,
+      "loss": 0.5046,
+      "step": 6142
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 30.35857167870842,
+      "learning_rate": 2.7320644337752966e-06,
+      "loss": 0.4627,
+      "step": 6143
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.084338191212948,
+      "learning_rate": 2.7305129059866956e-06,
+      "loss": 0.437,
+      "step": 6144
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.046224406752896,
+      "learning_rate": 2.728961653381228e-06,
+      "loss": 0.5054,
+      "step": 6145
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.713671532893246,
+      "learning_rate": 2.7274106761469887e-06,
+      "loss": 0.4978,
+      "step": 6146
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.548898564078269,
+      "learning_rate": 2.7258599744720416e-06,
+      "loss": 0.5241,
+      "step": 6147
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 17.36660076490788,
+      "learning_rate": 2.7243095485444103e-06,
+      "loss": 0.4522,
+      "step": 6148
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.223507276771732,
+      "learning_rate": 2.722759398552093e-06,
+      "loss": 0.4701,
+      "step": 6149
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.242565185113706,
+      "learning_rate": 2.7212095246830505e-06,
+      "loss": 0.49,
+      "step": 6150
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.351600741851818,
+      "learning_rate": 2.719659927125208e-06,
+      "loss": 0.4889,
+      "step": 6151
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8627951031173717,
+      "learning_rate": 2.718110606066462e-06,
+      "loss": 0.4834,
+      "step": 6152
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5442835793840977,
+      "learning_rate": 2.7165615616946726e-06,
+      "loss": 0.4964,
+      "step": 6153
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.309379968975917,
+      "learning_rate": 2.715012794197664e-06,
+      "loss": 0.4725,
+      "step": 6154
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.873450867189514,
+      "learning_rate": 2.7134643037632315e-06,
+      "loss": 0.4536,
+      "step": 6155
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.2608400944414395,
+      "learning_rate": 2.711916090579137e-06,
+      "loss": 0.4508,
+      "step": 6156
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.336565945264602,
+      "learning_rate": 2.710368154833103e-06,
+      "loss": 0.4841,
+      "step": 6157
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.28009377017396,
+      "learning_rate": 2.7088204967128236e-06,
+      "loss": 0.4574,
+      "step": 6158
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.005328075000617,
+      "learning_rate": 2.7072731164059594e-06,
+      "loss": 0.5175,
+      "step": 6159
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.053609991876095,
+      "learning_rate": 2.7057260141001328e-06,
+      "loss": 0.515,
+      "step": 6160
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8158805822700708,
+      "learning_rate": 2.704179189982936e-06,
+      "loss": 0.506,
+      "step": 6161
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.788195261031548,
+      "learning_rate": 2.70263264424193e-06,
+      "loss": 0.5075,
+      "step": 6162
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.435743508208828,
+      "learning_rate": 2.7010863770646357e-06,
+      "loss": 0.4688,
+      "step": 6163
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.490893847630234,
+      "learning_rate": 2.699540388638542e-06,
+      "loss": 0.4951,
+      "step": 6164
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.494110922184743,
+      "learning_rate": 2.6979946791511066e-06,
+      "loss": 0.4595,
+      "step": 6165
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.436926580250283,
+      "learning_rate": 2.696449248789754e-06,
+      "loss": 0.5106,
+      "step": 6166
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.602543263612595,
+      "learning_rate": 2.6949040977418706e-06,
+      "loss": 0.5158,
+      "step": 6167
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.241359067562818,
+      "learning_rate": 2.693359226194812e-06,
+      "loss": 0.5093,
+      "step": 6168
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.501130746276081,
+      "learning_rate": 2.691814634335902e-06,
+      "loss": 0.478,
+      "step": 6169
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.982193964353605,
+      "learning_rate": 2.6902703223524216e-06,
+      "loss": 0.4698,
+      "step": 6170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 20.486577204647087,
+      "learning_rate": 2.688726290431629e-06,
+      "loss": 0.4767,
+      "step": 6171
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.212740005651337,
+      "learning_rate": 2.687182538760743e-06,
+      "loss": 0.4566,
+      "step": 6172
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.395886298459823,
+      "learning_rate": 2.6856390675269478e-06,
+      "loss": 0.4501,
+      "step": 6173
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.002560319828364,
+      "learning_rate": 2.684095876917393e-06,
+      "loss": 0.5083,
+      "step": 6174
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.751429563468531,
+      "learning_rate": 2.6825529671191984e-06,
+      "loss": 0.5026,
+      "step": 6175
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8047864902186492,
+      "learning_rate": 2.681010338319444e-06,
+      "loss": 0.4799,
+      "step": 6176
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5299755517705984,
+      "learning_rate": 2.6794679907051817e-06,
+      "loss": 0.5153,
+      "step": 6177
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 14.251230144982326,
+      "learning_rate": 2.6779259244634272e-06,
+      "loss": 0.4474,
+      "step": 6178
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5714574918153965,
+      "learning_rate": 2.6763841397811576e-06,
+      "loss": 0.5389,
+      "step": 6179
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.718742293807596,
+      "learning_rate": 2.674842636845322e-06,
+      "loss": 0.5109,
+      "step": 6180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.39197448638199356,
+      "learning_rate": 2.673301415842835e-06,
+      "loss": 0.2727,
+      "step": 6181
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.989594262896392,
+      "learning_rate": 2.67176047696057e-06,
+      "loss": 0.4925,
+      "step": 6182
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.367201330048803,
+      "learning_rate": 2.670219820385374e-06,
+      "loss": 0.4421,
+      "step": 6183
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.412875484844189,
+      "learning_rate": 2.6686794463040613e-06,
+      "loss": 0.5557,
+      "step": 6184
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.410139828102311,
+      "learning_rate": 2.6671393549033984e-06,
+      "loss": 0.4789,
+      "step": 6185
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.689561215624285,
+      "learning_rate": 2.665599546370131e-06,
+      "loss": 0.4715,
+      "step": 6186
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.660113317058864,
+      "learning_rate": 2.6640600208909697e-06,
+      "loss": 0.5276,
+      "step": 6187
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.338867400243075,
+      "learning_rate": 2.662520778652581e-06,
+      "loss": 0.4384,
+      "step": 6188
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.732006039895668,
+      "learning_rate": 2.660981819841607e-06,
+      "loss": 0.4545,
+      "step": 6189
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.0996331073445065,
+      "learning_rate": 2.6594431446446528e-06,
+      "loss": 0.478,
+      "step": 6190
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.972570853768847,
+      "learning_rate": 2.657904753248285e-06,
+      "loss": 0.4509,
+      "step": 6191
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.570972078204064,
+      "learning_rate": 2.6563666458390396e-06,
+      "loss": 0.5248,
+      "step": 6192
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.460891303598941,
+      "learning_rate": 2.65482882260342e-06,
+      "loss": 0.519,
+      "step": 6193
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.7856433580754487,
+      "learning_rate": 2.6532912837278896e-06,
+      "loss": 0.5043,
+      "step": 6194
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3704745100612529,
+      "learning_rate": 2.6517540293988832e-06,
+      "loss": 0.2726,
+      "step": 6195
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.39886046270146197,
+      "learning_rate": 2.650217059802794e-06,
+      "loss": 0.283,
+      "step": 6196
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.62228514351116,
+      "learning_rate": 2.648680375125989e-06,
+      "loss": 0.4641,
+      "step": 6197
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.954074657411158,
+      "learning_rate": 2.6471439755547934e-06,
+      "loss": 0.4314,
+      "step": 6198
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.671869082160599,
+      "learning_rate": 2.645607861275502e-06,
+      "loss": 0.4791,
+      "step": 6199
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.233846430553387,
+      "learning_rate": 2.644072032474376e-06,
+      "loss": 0.518,
+      "step": 6200
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.716221684800135,
+      "learning_rate": 2.642536489337636e-06,
+      "loss": 0.5262,
+      "step": 6201
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4915579535042296,
+      "learning_rate": 2.6410012320514742e-06,
+      "loss": 0.4761,
+      "step": 6202
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.3480045910608744,
+      "learning_rate": 2.639466260802048e-06,
+      "loss": 0.5093,
+      "step": 6203
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.259813503296676,
+      "learning_rate": 2.6379315757754744e-06,
+      "loss": 0.5167,
+      "step": 6204
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4016742878679795,
+      "learning_rate": 2.6363971771578417e-06,
+      "loss": 0.5137,
+      "step": 6205
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.4289059855940005,
+      "learning_rate": 2.6348630651351997e-06,
+      "loss": 0.4521,
+      "step": 6206
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9860847682649534,
+      "learning_rate": 2.6333292398935636e-06,
+      "loss": 0.508,
+      "step": 6207
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3331880570416073,
+      "learning_rate": 2.6317957016189155e-06,
+      "loss": 0.5207,
+      "step": 6208
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0058385709373994,
+      "learning_rate": 2.6302624504972054e-06,
+      "loss": 0.4895,
+      "step": 6209
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.933121865492769,
+      "learning_rate": 2.628729486714341e-06,
+      "loss": 0.4936,
+      "step": 6210
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.446683820691992,
+      "learning_rate": 2.627196810456201e-06,
+      "loss": 0.5135,
+      "step": 6211
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4042867877746863,
+      "learning_rate": 2.62566442190863e-06,
+      "loss": 0.2783,
+      "step": 6212
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1836574640199977,
+      "learning_rate": 2.624132321257432e-06,
+      "loss": 0.5134,
+      "step": 6213
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.347009388207431,
+      "learning_rate": 2.622600508688381e-06,
+      "loss": 0.5034,
+      "step": 6214
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 17.874091933343443,
+      "learning_rate": 2.621068984387216e-06,
+      "loss": 0.489,
+      "step": 6215
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.360760674131823,
+      "learning_rate": 2.6195377485396377e-06,
+      "loss": 0.4958,
+      "step": 6216
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.885458619521748,
+      "learning_rate": 2.6180068013313133e-06,
+      "loss": 0.4978,
+      "step": 6217
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.40165888249620246,
+      "learning_rate": 2.616476142947877e-06,
+      "loss": 0.2737,
+      "step": 6218
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.3273089857240015,
+      "learning_rate": 2.614945773574924e-06,
+      "loss": 0.4651,
+      "step": 6219
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0471740300468197,
+      "learning_rate": 2.6134156933980186e-06,
+      "loss": 0.4522,
+      "step": 6220
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.373060913154184,
+      "learning_rate": 2.61188590260269e-06,
+      "loss": 0.5188,
+      "step": 6221
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9559551666175112,
+      "learning_rate": 2.610356401374427e-06,
+      "loss": 0.4461,
+      "step": 6222
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.171374633001337,
+      "learning_rate": 2.608827189898688e-06,
+      "loss": 0.4955,
+      "step": 6223
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9200165725861726,
+      "learning_rate": 2.6072982683608992e-06,
+      "loss": 0.4751,
+      "step": 6224
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1559867625895768,
+      "learning_rate": 2.6057696369464412e-06,
+      "loss": 0.466,
+      "step": 6225
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.536026960919498,
+      "learning_rate": 2.6042412958406717e-06,
+      "loss": 0.4456,
+      "step": 6226
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.373751610314056,
+      "learning_rate": 2.6027132452289015e-06,
+      "loss": 0.4796,
+      "step": 6227
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.620601870018376,
+      "learning_rate": 2.601185485296418e-06,
+      "loss": 0.5221,
+      "step": 6228
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.664931234631929,
+      "learning_rate": 2.5996580162284618e-06,
+      "loss": 0.4494,
+      "step": 6229
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7267182962661582,
+      "learning_rate": 2.5981308382102467e-06,
+      "loss": 0.5089,
+      "step": 6230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.767931529748696,
+      "learning_rate": 2.5966039514269494e-06,
+      "loss": 0.5578,
+      "step": 6231
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2352725944529768,
+      "learning_rate": 2.595077356063708e-06,
+      "loss": 0.4958,
+      "step": 6232
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.564479741795582,
+      "learning_rate": 2.5935510523056268e-06,
+      "loss": 0.4985,
+      "step": 6233
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.30134662536825,
+      "learning_rate": 2.592025040337779e-06,
+      "loss": 0.4796,
+      "step": 6234
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.869868331525056,
+      "learning_rate": 2.5904993203451957e-06,
+      "loss": 0.4782,
+      "step": 6235
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.379802372772823,
+      "learning_rate": 2.5889738925128753e-06,
+      "loss": 0.5092,
+      "step": 6236
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.487623413587782,
+      "learning_rate": 2.587448757025787e-06,
+      "loss": 0.5073,
+      "step": 6237
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8230729863788966,
+      "learning_rate": 2.585923914068851e-06,
+      "loss": 0.4711,
+      "step": 6238
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.433529976005897,
+      "learning_rate": 2.5843993638269618e-06,
+      "loss": 0.4737,
+      "step": 6239
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.678351676808199,
+      "learning_rate": 2.58287510648498e-06,
+      "loss": 0.4553,
+      "step": 6240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4146480598610733,
+      "learning_rate": 2.5813511422277227e-06,
+      "loss": 0.2717,
+      "step": 6241
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.872647952679751,
+      "learning_rate": 2.579827471239978e-06,
+      "loss": 0.4716,
+      "step": 6242
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3872988887112229,
+      "learning_rate": 2.5783040937064973e-06,
+      "loss": 0.2755,
+      "step": 6243
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.624366492841229,
+      "learning_rate": 2.5767810098119925e-06,
+      "loss": 0.4984,
+      "step": 6244
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1743635159322645,
+      "learning_rate": 2.575258219741145e-06,
+      "loss": 0.4745,
+      "step": 6245
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3899639332836905,
+      "learning_rate": 2.573735723678599e-06,
+      "loss": 0.2726,
+      "step": 6246
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 17.58369385357364,
+      "learning_rate": 2.5722135218089596e-06,
+      "loss": 0.4444,
+      "step": 6247
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6846286849675995,
+      "learning_rate": 2.570691614316802e-06,
+      "loss": 0.5136,
+      "step": 6248
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.98421573101494,
+      "learning_rate": 2.5691700013866615e-06,
+      "loss": 0.4782,
+      "step": 6249
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.43471417541428,
+      "learning_rate": 2.5676486832030366e-06,
+      "loss": 0.4659,
+      "step": 6250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.9144661891924075,
+      "learning_rate": 2.5661276599503946e-06,
+      "loss": 0.4869,
+      "step": 6251
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6460216731117425,
+      "learning_rate": 2.5646069318131663e-06,
+      "loss": 0.4717,
+      "step": 6252
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4016448517921363,
+      "learning_rate": 2.563086498975742e-06,
+      "loss": 0.2886,
+      "step": 6253
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.798689852754895,
+      "learning_rate": 2.5615663616224807e-06,
+      "loss": 0.4602,
+      "step": 6254
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2574889223067247,
+      "learning_rate": 2.560046519937707e-06,
+      "loss": 0.5173,
+      "step": 6255
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.412029453698146,
+      "learning_rate": 2.558526974105703e-06,
+      "loss": 0.4712,
+      "step": 6256
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9769521620810018,
+      "learning_rate": 2.5570077243107195e-06,
+      "loss": 0.5127,
+      "step": 6257
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.950125029937368,
+      "learning_rate": 2.5554887707369734e-06,
+      "loss": 0.4837,
+      "step": 6258
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.024560237990457,
+      "learning_rate": 2.5539701135686424e-06,
+      "loss": 0.4913,
+      "step": 6259
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.653961541411252,
+      "learning_rate": 2.552451752989865e-06,
+      "loss": 0.5073,
+      "step": 6260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.19165934042678,
+      "learning_rate": 2.5509336891847516e-06,
+      "loss": 0.5327,
+      "step": 6261
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.562912535534653,
+      "learning_rate": 2.549415922337373e-06,
+      "loss": 0.4843,
+      "step": 6262
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.132353258283455,
+      "learning_rate": 2.5478984526317596e-06,
+      "loss": 0.4954,
+      "step": 6263
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8758309052837254,
+      "learning_rate": 2.5463812802519133e-06,
+      "loss": 0.4943,
+      "step": 6264
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3944215132491357,
+      "learning_rate": 2.544864405381797e-06,
+      "loss": 0.282,
+      "step": 6265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.0006228522373135,
+      "learning_rate": 2.543347828205335e-06,
+      "loss": 0.4708,
+      "step": 6266
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.4815129575985955,
+      "learning_rate": 2.541831548906417e-06,
+      "loss": 0.4952,
+      "step": 6267
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.1962939363636,
+      "learning_rate": 2.5403155676689005e-06,
+      "loss": 0.5126,
+      "step": 6268
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.890637671160937,
+      "learning_rate": 2.5387998846766016e-06,
+      "loss": 0.5675,
+      "step": 6269
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.2414716413473155,
+      "learning_rate": 2.5372845001132995e-06,
+      "loss": 0.531,
+      "step": 6270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 25.552095437724017,
+      "learning_rate": 2.5357694141627444e-06,
+      "loss": 0.5007,
+      "step": 6271
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0084449166777776,
+      "learning_rate": 2.534254627008641e-06,
+      "loss": 0.4982,
+      "step": 6272
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.57224122797247,
+      "learning_rate": 2.532740138834665e-06,
+      "loss": 0.4849,
+      "step": 6273
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5128003198819315,
+      "learning_rate": 2.5312259498244544e-06,
+      "loss": 0.468,
+      "step": 6274
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4878582399122684,
+      "learning_rate": 2.5297120601616078e-06,
+      "loss": 0.4945,
+      "step": 6275
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.7559588448473455,
+      "learning_rate": 2.5281984700296893e-06,
+      "loss": 0.4822,
+      "step": 6276
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.664853966568525,
+      "learning_rate": 2.52668517961223e-06,
+      "loss": 0.4608,
+      "step": 6277
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9569333911561517,
+      "learning_rate": 2.5251721890927173e-06,
+      "loss": 0.4917,
+      "step": 6278
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.361330656298769,
+      "learning_rate": 2.5236594986546115e-06,
+      "loss": 0.4653,
+      "step": 6279
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3586698785101069,
+      "learning_rate": 2.5221471084813275e-06,
+      "loss": 0.2901,
+      "step": 6280
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6433427701710834,
+      "learning_rate": 2.5206350187562478e-06,
+      "loss": 0.4169,
+      "step": 6281
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5482122745931544,
+      "learning_rate": 2.51912322966272e-06,
+      "loss": 0.472,
+      "step": 6282
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.8512945054810848,
+      "learning_rate": 2.517611741384055e-06,
+      "loss": 0.4498,
+      "step": 6283
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6190395976332232,
+      "learning_rate": 2.516100554103523e-06,
+      "loss": 0.4603,
+      "step": 6284
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.000558807903653,
+      "learning_rate": 2.514589668004362e-06,
+      "loss": 0.5157,
+      "step": 6285
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.547165606286531,
+      "learning_rate": 2.513079083269774e-06,
+      "loss": 0.4657,
+      "step": 6286
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.99583254771144,
+      "learning_rate": 2.511568800082919e-06,
+      "loss": 0.4577,
+      "step": 6287
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9092130181377307,
+      "learning_rate": 2.5100588186269257e-06,
+      "loss": 0.5122,
+      "step": 6288
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.760873059866888,
+      "learning_rate": 2.508549139084887e-06,
+      "loss": 0.4838,
+      "step": 6289
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.107922755190473,
+      "learning_rate": 2.5070397616398546e-06,
+      "loss": 0.5158,
+      "step": 6290
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7369108040791463,
+      "learning_rate": 2.5055306864748434e-06,
+      "loss": 0.4361,
+      "step": 6291
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3945890363860827,
+      "learning_rate": 2.504021913772836e-06,
+      "loss": 0.2944,
+      "step": 6292
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.120872058904016,
+      "learning_rate": 2.502513443716779e-06,
+      "loss": 0.4611,
+      "step": 6293
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.551351783783852,
+      "learning_rate": 2.5010052764895752e-06,
+      "loss": 0.4793,
+      "step": 6294
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.649248145249195,
+      "learning_rate": 2.4994974122740968e-06,
+      "loss": 0.5383,
+      "step": 6295
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.671278226929113,
+      "learning_rate": 2.4979898512531796e-06,
+      "loss": 0.5133,
+      "step": 6296
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.802989781071768,
+      "learning_rate": 2.4964825936096164e-06,
+      "loss": 0.4666,
+      "step": 6297
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.865103304103332,
+      "learning_rate": 2.4949756395261706e-06,
+      "loss": 0.4574,
+      "step": 6298
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.622503014311542,
+      "learning_rate": 2.493468989185566e-06,
+      "loss": 0.515,
+      "step": 6299
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0536702062787304,
+      "learning_rate": 2.4919626427704873e-06,
+      "loss": 0.4734,
+      "step": 6300
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.127416455136559,
+      "learning_rate": 2.4904566004635835e-06,
+      "loss": 0.4753,
+      "step": 6301
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.131388755946532,
+      "learning_rate": 2.4889508624474694e-06,
+      "loss": 0.5066,
+      "step": 6302
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.98174968034551,
+      "learning_rate": 2.4874454289047194e-06,
+      "loss": 0.5162,
+      "step": 6303
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.548614345364991,
+      "learning_rate": 2.4859403000178727e-06,
+      "loss": 0.4937,
+      "step": 6304
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.37117623641191694,
+      "learning_rate": 2.484435475969434e-06,
+      "loss": 0.2679,
+      "step": 6305
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7076431514043873,
+      "learning_rate": 2.4829309569418647e-06,
+      "loss": 0.4724,
+      "step": 6306
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8324072605022264,
+      "learning_rate": 2.4814267431175942e-06,
+      "loss": 0.4942,
+      "step": 6307
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.186081351086941,
+      "learning_rate": 2.4799228346790157e-06,
+      "loss": 0.4766,
+      "step": 6308
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.44001593121678917,
+      "learning_rate": 2.47841923180848e-06,
+      "loss": 0.2785,
+      "step": 6309
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.511947005249711,
+      "learning_rate": 2.4769159346883053e-06,
+      "loss": 0.5112,
+      "step": 6310
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9307698779241877,
+      "learning_rate": 2.4754129435007757e-06,
+      "loss": 0.4646,
+      "step": 6311
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.204099509163726,
+      "learning_rate": 2.4739102584281268e-06,
+      "loss": 0.4537,
+      "step": 6312
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.710276739731352,
+      "learning_rate": 2.4724078796525674e-06,
+      "loss": 0.4644,
+      "step": 6313
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.389472406308016,
+      "learning_rate": 2.4709058073562688e-06,
+      "loss": 0.4869,
+      "step": 6314
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.349415154030884,
+      "learning_rate": 2.4694040417213584e-06,
+      "loss": 0.4649,
+      "step": 6315
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.852258812634446,
+      "learning_rate": 2.467902582929931e-06,
+      "loss": 0.4683,
+      "step": 6316
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4069358264075642,
+      "learning_rate": 2.4664014311640483e-06,
+      "loss": 0.2789,
+      "step": 6317
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.85736401367726,
+      "learning_rate": 2.464900586605724e-06,
+      "loss": 0.4735,
+      "step": 6318
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.54097076210243,
+      "learning_rate": 2.463400049436944e-06,
+      "loss": 0.4686,
+      "step": 6319
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.866927885698881,
+      "learning_rate": 2.4618998198396544e-06,
+      "loss": 0.4697,
+      "step": 6320
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.619523892445509,
+      "learning_rate": 2.46039989799576e-06,
+      "loss": 0.4892,
+      "step": 6321
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.30733775287188,
+      "learning_rate": 2.4589002840871356e-06,
+      "loss": 0.5074,
+      "step": 6322
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.3298614677538785,
+      "learning_rate": 2.4574009782956098e-06,
+      "loss": 0.4765,
+      "step": 6323
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.506327658489976,
+      "learning_rate": 2.455901980802983e-06,
+      "loss": 0.5032,
+      "step": 6324
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3946207604587824,
+      "learning_rate": 2.4544032917910105e-06,
+      "loss": 0.4306,
+      "step": 6325
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.166630811600924,
+      "learning_rate": 2.4529049114414143e-06,
+      "loss": 0.4787,
+      "step": 6326
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.314979355793733,
+      "learning_rate": 2.451406839935881e-06,
+      "loss": 0.4834,
+      "step": 6327
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.263327230503886,
+      "learning_rate": 2.4499090774560525e-06,
+      "loss": 0.4843,
+      "step": 6328
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.36881734903830743,
+      "learning_rate": 2.4484116241835406e-06,
+      "loss": 0.2824,
+      "step": 6329
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.618800431648303,
+      "learning_rate": 2.446914480299917e-06,
+      "loss": 0.4771,
+      "step": 6330
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.543143649196179,
+      "learning_rate": 2.4454176459867128e-06,
+      "loss": 0.4733,
+      "step": 6331
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.7205546225909005,
+      "learning_rate": 2.4439211214254277e-06,
+      "loss": 0.4699,
+      "step": 6332
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.39199181681300166,
+      "learning_rate": 2.4424249067975193e-06,
+      "loss": 0.2808,
+      "step": 6333
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5184279310059186,
+      "learning_rate": 2.4409290022844063e-06,
+      "loss": 0.5131,
+      "step": 6334
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4334952917376387,
+      "learning_rate": 2.4394334080674737e-06,
+      "loss": 0.4382,
+      "step": 6335
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.864285455449995,
+      "learning_rate": 2.4379381243280705e-06,
+      "loss": 0.4845,
+      "step": 6336
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 15.421091238727566,
+      "learning_rate": 2.4364431512475e-06,
+      "loss": 0.5085,
+      "step": 6337
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7419208070361853,
+      "learning_rate": 2.4349484890070357e-06,
+      "loss": 0.5063,
+      "step": 6338
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.046306931289578,
+      "learning_rate": 2.433454137787912e-06,
+      "loss": 0.5378,
+      "step": 6339
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7447880997391048,
+      "learning_rate": 2.4319600977713202e-06,
+      "loss": 0.4876,
+      "step": 6340
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.44807530924707,
+      "learning_rate": 2.4304663691384206e-06,
+      "loss": 0.495,
+      "step": 6341
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.938664467991565,
+      "learning_rate": 2.4289729520703335e-06,
+      "loss": 0.431,
+      "step": 6342
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9993414093086015,
+      "learning_rate": 2.4274798467481397e-06,
+      "loss": 0.479,
+      "step": 6343
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.057855327501566,
+      "learning_rate": 2.4259870533528816e-06,
+      "loss": 0.4514,
+      "step": 6344
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.477936547611664,
+      "learning_rate": 2.4244945720655693e-06,
+      "loss": 0.4451,
+      "step": 6345
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.464466451572523,
+      "learning_rate": 2.4230024030671674e-06,
+      "loss": 0.5513,
+      "step": 6346
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.41239328384531027,
+      "learning_rate": 2.421510546538608e-06,
+      "loss": 0.2829,
+      "step": 6347
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.369570568696745,
+      "learning_rate": 2.4200190026607863e-06,
+      "loss": 0.2682,
+      "step": 6348
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.885858007708947,
+      "learning_rate": 2.418527771614553e-06,
+      "loss": 0.5109,
+      "step": 6349
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.274525619758466,
+      "learning_rate": 2.4170368535807276e-06,
+      "loss": 0.4722,
+      "step": 6350
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.092886609847444,
+      "learning_rate": 2.41554624874009e-06,
+      "loss": 0.4937,
+      "step": 6351
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.2428851933900225,
+      "learning_rate": 2.414055957273378e-06,
+      "loss": 0.5087,
+      "step": 6352
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.979211191714314,
+      "learning_rate": 2.412565979361298e-06,
+      "loss": 0.4844,
+      "step": 6353
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.37498699655444506,
+      "learning_rate": 2.411076315184511e-06,
+      "loss": 0.2843,
+      "step": 6354
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7275389465920115,
+      "learning_rate": 2.409586964923649e-06,
+      "loss": 0.4696,
+      "step": 6355
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1836503466732107,
+      "learning_rate": 2.4080979287592956e-06,
+      "loss": 0.5076,
+      "step": 6356
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.083747846828803,
+      "learning_rate": 2.4066092068720044e-06,
+      "loss": 0.5417,
+      "step": 6357
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.688434345859118,
+      "learning_rate": 2.4051207994422886e-06,
+      "loss": 0.5433,
+      "step": 6358
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.776584483832037,
+      "learning_rate": 2.4036327066506204e-06,
+      "loss": 0.4793,
+      "step": 6359
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.1192755585049925,
+      "learning_rate": 2.402144928677438e-06,
+      "loss": 0.5387,
+      "step": 6360
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.454200170086892,
+      "learning_rate": 2.4006574657031406e-06,
+      "loss": 0.5263,
+      "step": 6361
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7540878963691413,
+      "learning_rate": 2.399170317908085e-06,
+      "loss": 0.493,
+      "step": 6362
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.332408235135705,
+      "learning_rate": 2.3976834854725945e-06,
+      "loss": 0.4675,
+      "step": 6363
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6506740774620385,
+      "learning_rate": 2.396196968576957e-06,
+      "loss": 0.4719,
+      "step": 6364
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.140021610989488,
+      "learning_rate": 2.3947107674014097e-06,
+      "loss": 0.4568,
+      "step": 6365
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.38169406074049667,
+      "learning_rate": 2.393224882126164e-06,
+      "loss": 0.263,
+      "step": 6366
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.183097333106695,
+      "learning_rate": 2.39173931293139e-06,
+      "loss": 0.4761,
+      "step": 6367
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.877196222581997,
+      "learning_rate": 2.3902540599972145e-06,
+      "loss": 0.4979,
+      "step": 6368
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.421896841741886,
+      "learning_rate": 2.388769123503731e-06,
+      "loss": 0.5404,
+      "step": 6369
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7033521227695156,
+      "learning_rate": 2.387284503630996e-06,
+      "loss": 0.4666,
+      "step": 6370
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6013319969248254,
+      "learning_rate": 2.38580020055902e-06,
+      "loss": 0.473,
+      "step": 6371
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.271054755839478,
+      "learning_rate": 2.3843162144677828e-06,
+      "loss": 0.4596,
+      "step": 6372
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8470604017990886,
+      "learning_rate": 2.3828325455372235e-06,
+      "loss": 0.4877,
+      "step": 6373
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.217031656318485,
+      "learning_rate": 2.3813491939472395e-06,
+      "loss": 0.4883,
+      "step": 6374
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.287733538252977,
+      "learning_rate": 2.379866159877695e-06,
+      "loss": 0.4956,
+      "step": 6375
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.417916058760398,
+      "learning_rate": 2.3783834435084118e-06,
+      "loss": 0.5254,
+      "step": 6376
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.925870321528973,
+      "learning_rate": 2.376901045019172e-06,
+      "loss": 0.4985,
+      "step": 6377
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.039880080492773,
+      "learning_rate": 2.375418964589724e-06,
+      "loss": 0.4496,
+      "step": 6378
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.110025495468874,
+      "learning_rate": 2.3739372023997766e-06,
+      "loss": 0.5426,
+      "step": 6379
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5744765441628585,
+      "learning_rate": 2.3724557586289948e-06,
+      "loss": 0.5211,
+      "step": 6380
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.89628252775071,
+      "learning_rate": 2.370974633457011e-06,
+      "loss": 0.5055,
+      "step": 6381
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1706926742581842,
+      "learning_rate": 2.3694938270634187e-06,
+      "loss": 0.4728,
+      "step": 6382
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.74285118490453,
+      "learning_rate": 2.368013339627766e-06,
+      "loss": 0.5238,
+      "step": 6383
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2855747061657854,
+      "learning_rate": 2.3665331713295702e-06,
+      "loss": 0.493,
+      "step": 6384
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.6768970300189325,
+      "learning_rate": 2.3650533223483087e-06,
+      "loss": 0.4775,
+      "step": 6385
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2615975547002845,
+      "learning_rate": 2.363573792863415e-06,
+      "loss": 0.4462,
+      "step": 6386
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.41416402333384034,
+      "learning_rate": 2.3620945830542865e-06,
+      "loss": 0.2876,
+      "step": 6387
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.5958824393592765,
+      "learning_rate": 2.360615693100284e-06,
+      "loss": 0.4787,
+      "step": 6388
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.330843642260771,
+      "learning_rate": 2.35913712318073e-06,
+      "loss": 0.5106,
+      "step": 6389
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.0739870288565365,
+      "learning_rate": 2.3576588734749022e-06,
+      "loss": 0.482,
+      "step": 6390
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.837787174386992,
+      "learning_rate": 2.3561809441620458e-06,
+      "loss": 0.5331,
+      "step": 6391
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.226040397804297,
+      "learning_rate": 2.3547033354213663e-06,
+      "loss": 0.4582,
+      "step": 6392
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.146576370946879,
+      "learning_rate": 2.3532260474320253e-06,
+      "loss": 0.4785,
+      "step": 6393
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7876590825715244,
+      "learning_rate": 2.3517490803731506e-06,
+      "loss": 0.4965,
+      "step": 6394
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0441688067823263,
+      "learning_rate": 2.350272434423832e-06,
+      "loss": 0.466,
+      "step": 6395
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4679862443881535,
+      "learning_rate": 2.3487961097631157e-06,
+      "loss": 0.4894,
+      "step": 6396
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.601616525313084,
+      "learning_rate": 2.347320106570009e-06,
+      "loss": 0.4865,
+      "step": 6397
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.312227502047483,
+      "learning_rate": 2.3458444250234867e-06,
+      "loss": 0.482,
+      "step": 6398
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.560424658323853,
+      "learning_rate": 2.3443690653024766e-06,
+      "loss": 0.4726,
+      "step": 6399
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.625031994685323,
+      "learning_rate": 2.342894027585872e-06,
+      "loss": 0.4582,
+      "step": 6400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3358784718871863,
+      "learning_rate": 2.341419312052529e-06,
+      "loss": 0.4511,
+      "step": 6401
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.424507787221964,
+      "learning_rate": 2.3399449188812585e-06,
+      "loss": 0.5032,
+      "step": 6402
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 13.645828841430923,
+      "learning_rate": 2.338470848250837e-06,
+      "loss": 0.5334,
+      "step": 6403
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.546005785891959,
+      "learning_rate": 2.3369971003400023e-06,
+      "loss": 0.4975,
+      "step": 6404
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.302850499989943,
+      "learning_rate": 2.3355236753274485e-06,
+      "loss": 0.4979,
+      "step": 6405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.082129087284372,
+      "learning_rate": 2.3340505733918366e-06,
+      "loss": 0.5027,
+      "step": 6406
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6964506967769304,
+      "learning_rate": 2.3325777947117834e-06,
+      "loss": 0.4784,
+      "step": 6407
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.231762225098031,
+      "learning_rate": 2.3311053394658673e-06,
+      "loss": 0.4779,
+      "step": 6408
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.709377477580394,
+      "learning_rate": 2.3296332078326295e-06,
+      "loss": 0.4771,
+      "step": 6409
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.110916406124316,
+      "learning_rate": 2.3281613999905735e-06,
+      "loss": 0.4357,
+      "step": 6410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.166907211926423,
+      "learning_rate": 2.326689916118157e-06,
+      "loss": 0.5225,
+      "step": 6411
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8069371033766495,
+      "learning_rate": 2.3252187563938043e-06,
+      "loss": 0.4374,
+      "step": 6412
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6309926121077827,
+      "learning_rate": 2.3237479209959007e-06,
+      "loss": 0.4793,
+      "step": 6413
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.209647660759158,
+      "learning_rate": 2.3222774101027864e-06,
+      "loss": 0.4372,
+      "step": 6414
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.052418144323378,
+      "learning_rate": 2.3208072238927676e-06,
+      "loss": 0.5269,
+      "step": 6415
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.374926832730677,
+      "learning_rate": 2.3193373625441113e-06,
+      "loss": 0.502,
+      "step": 6416
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0482211514588933,
+      "learning_rate": 2.3178678262350413e-06,
+      "loss": 0.514,
+      "step": 6417
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.909646584119572,
+      "learning_rate": 2.3163986151437425e-06,
+      "loss": 0.4998,
+      "step": 6418
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3784128722683685,
+      "learning_rate": 2.314929729448363e-06,
+      "loss": 0.2761,
+      "step": 6419
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.761483195578955,
+      "learning_rate": 2.313461169327013e-06,
+      "loss": 0.4639,
+      "step": 6420
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.542109207510116,
+      "learning_rate": 2.311992934957756e-06,
+      "loss": 0.4895,
+      "step": 6421
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.901677425173536,
+      "learning_rate": 2.3105250265186223e-06,
+      "loss": 0.499,
+      "step": 6422
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3312101030145675,
+      "learning_rate": 2.3090574441876037e-06,
+      "loss": 0.4451,
+      "step": 6423
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.5999180243238555,
+      "learning_rate": 2.3075901881426447e-06,
+      "loss": 0.5313,
+      "step": 6424
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.941675213202291,
+      "learning_rate": 2.306123258561658e-06,
+      "loss": 0.4597,
+      "step": 6425
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.639782209280648,
+      "learning_rate": 2.3046566556225147e-06,
+      "loss": 0.4935,
+      "step": 6426
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.94891406963964,
+      "learning_rate": 2.303190379503043e-06,
+      "loss": 0.4579,
+      "step": 6427
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.47085617681544456,
+      "learning_rate": 2.3017244303810367e-06,
+      "loss": 0.2721,
+      "step": 6428
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4697421659436767,
+      "learning_rate": 2.3002588084342453e-06,
+      "loss": 0.5152,
+      "step": 6429
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4227000391990527,
+      "learning_rate": 2.2987935138403786e-06,
+      "loss": 0.4852,
+      "step": 6430
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8334501179198894,
+      "learning_rate": 2.2973285467771117e-06,
+      "loss": 0.4662,
+      "step": 6431
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3712417022440637,
+      "learning_rate": 2.2958639074220773e-06,
+      "loss": 0.5149,
+      "step": 6432
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.0868443394878025,
+      "learning_rate": 2.2943995959528654e-06,
+      "loss": 0.5017,
+      "step": 6433
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.478381104355501,
+      "learning_rate": 2.2929356125470297e-06,
+      "loss": 0.4903,
+      "step": 6434
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.710583286113085,
+      "learning_rate": 2.2914719573820854e-06,
+      "loss": 0.5357,
+      "step": 6435
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.611607875971611,
+      "learning_rate": 2.2900086306355024e-06,
+      "loss": 0.5197,
+      "step": 6436
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.486406939424266,
+      "learning_rate": 2.2885456324847156e-06,
+      "loss": 0.5322,
+      "step": 6437
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7519027472756474,
+      "learning_rate": 2.2870829631071207e-06,
+      "loss": 0.4279,
+      "step": 6438
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5478560643441974,
+      "learning_rate": 2.2856206226800686e-06,
+      "loss": 0.5067,
+      "step": 6439
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3049605439212932,
+      "learning_rate": 2.284158611380873e-06,
+      "loss": 0.4636,
+      "step": 6440
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.036549355145706,
+      "learning_rate": 2.2826969293868103e-06,
+      "loss": 0.4641,
+      "step": 6441
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.926160147742191,
+      "learning_rate": 2.2812355768751106e-06,
+      "loss": 0.4587,
+      "step": 6442
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.594001208999805,
+      "learning_rate": 2.2797745540229703e-06,
+      "loss": 0.5546,
+      "step": 6443
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.801929072393963,
+      "learning_rate": 2.2783138610075454e-06,
+      "loss": 0.5115,
+      "step": 6444
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1245888434480475,
+      "learning_rate": 2.2768534980059464e-06,
+      "loss": 0.4774,
+      "step": 6445
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.749164045488934,
+      "learning_rate": 2.2753934651952485e-06,
+      "loss": 0.5048,
+      "step": 6446
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9780255636526975,
+      "learning_rate": 2.273933762752488e-06,
+      "loss": 0.4486,
+      "step": 6447
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7144843902941878,
+      "learning_rate": 2.2724743908546553e-06,
+      "loss": 0.4946,
+      "step": 6448
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.536355432617286,
+      "learning_rate": 2.271015349678707e-06,
+      "loss": 0.4666,
+      "step": 6449
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6393649569017126,
+      "learning_rate": 2.2695566394015552e-06,
+      "loss": 0.4646,
+      "step": 6450
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.036097540826213,
+      "learning_rate": 2.268098260200075e-06,
+      "loss": 0.4796,
+      "step": 6451
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.9489726380383896,
+      "learning_rate": 2.2666402122510976e-06,
+      "loss": 0.457,
+      "step": 6452
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8572506300617375,
+      "learning_rate": 2.2651824957314177e-06,
+      "loss": 0.4875,
+      "step": 6453
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.127780648395724,
+      "learning_rate": 2.2637251108177903e-06,
+      "loss": 0.5007,
+      "step": 6454
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.333705679115578,
+      "learning_rate": 2.262268057686925e-06,
+      "loss": 0.49,
+      "step": 6455
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4662787734791207,
+      "learning_rate": 2.260811336515496e-06,
+      "loss": 0.4819,
+      "step": 6456
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9796770364926606,
+      "learning_rate": 2.259354947480138e-06,
+      "loss": 0.5049,
+      "step": 6457
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1130438278595154,
+      "learning_rate": 2.257898890757439e-06,
+      "loss": 0.504,
+      "step": 6458
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0480540633786486,
+      "learning_rate": 2.2564431665239545e-06,
+      "loss": 0.5131,
+      "step": 6459
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4485756108391294,
+      "learning_rate": 2.2549877749561942e-06,
+      "loss": 0.4515,
+      "step": 6460
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.101885268532178,
+      "learning_rate": 2.2535327162306286e-06,
+      "loss": 0.5109,
+      "step": 6461
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.229278410787904,
+      "learning_rate": 2.2520779905236895e-06,
+      "loss": 0.5268,
+      "step": 6462
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.522812422346412,
+      "learning_rate": 2.2506235980117695e-06,
+      "loss": 0.436,
+      "step": 6463
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.26594084593186,
+      "learning_rate": 2.2491695388712147e-06,
+      "loss": 0.4773,
+      "step": 6464
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.638469414610764,
+      "learning_rate": 2.2477158132783373e-06,
+      "loss": 0.4332,
+      "step": 6465
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2178515849392126,
+      "learning_rate": 2.2462624214094074e-06,
+      "loss": 0.4481,
+      "step": 6466
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.81011843864766,
+      "learning_rate": 2.244809363440651e-06,
+      "loss": 0.5077,
+      "step": 6467
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.40660433163110393,
+      "learning_rate": 2.2433566395482577e-06,
+      "loss": 0.2988,
+      "step": 6468
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0928451241645325,
+      "learning_rate": 2.241904249908377e-06,
+      "loss": 0.5207,
+      "step": 6469
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.365420650828266,
+      "learning_rate": 2.2404521946971153e-06,
+      "loss": 0.5165,
+      "step": 6470
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.301067430757302,
+      "learning_rate": 2.2390004740905363e-06,
+      "loss": 0.505,
+      "step": 6471
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3770258600193093,
+      "learning_rate": 2.2375490882646698e-06,
+      "loss": 0.5033,
+      "step": 6472
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0880054288512038,
+      "learning_rate": 2.2360980373954988e-06,
+      "loss": 0.4836,
+      "step": 6473
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9109515787442386,
+      "learning_rate": 2.2346473216589694e-06,
+      "loss": 0.4939,
+      "step": 6474
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.597204779944459,
+      "learning_rate": 2.233196941230988e-06,
+      "loss": 0.5041,
+      "step": 6475
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.717672437463567,
+      "learning_rate": 2.2317468962874136e-06,
+      "loss": 0.4723,
+      "step": 6476
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.355355513281137,
+      "learning_rate": 2.230297187004072e-06,
+      "loss": 0.5054,
+      "step": 6477
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.907143108947738,
+      "learning_rate": 2.2288478135567466e-06,
+      "loss": 0.4792,
+      "step": 6478
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3991618480974418,
+      "learning_rate": 2.2273987761211755e-06,
+      "loss": 0.2732,
+      "step": 6479
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8548959927942064,
+      "learning_rate": 2.225950074873064e-06,
+      "loss": 0.5133,
+      "step": 6480
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5177192964946573,
+      "learning_rate": 2.2245017099880667e-06,
+      "loss": 0.509,
+      "step": 6481
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.016466912262018,
+      "learning_rate": 2.223053681641808e-06,
+      "loss": 0.5216,
+      "step": 6482
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.498612054716943,
+      "learning_rate": 2.2216059900098624e-06,
+      "loss": 0.4912,
+      "step": 6483
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.616844393239889,
+      "learning_rate": 2.220158635267769e-06,
+      "loss": 0.433,
+      "step": 6484
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0512376448148724,
+      "learning_rate": 2.2187116175910263e-06,
+      "loss": 0.2852,
+      "step": 6485
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0372352050453797,
+      "learning_rate": 2.2172649371550863e-06,
+      "loss": 0.452,
+      "step": 6486
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1426134599855797,
+      "learning_rate": 2.2158185941353672e-06,
+      "loss": 0.4648,
+      "step": 6487
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9564403140865214,
+      "learning_rate": 2.2143725887072433e-06,
+      "loss": 0.4539,
+      "step": 6488
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 34.072866134185496,
+      "learning_rate": 2.212926921046045e-06,
+      "loss": 0.535,
+      "step": 6489
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.38102634367616656,
+      "learning_rate": 2.2114815913270654e-06,
+      "loss": 0.2657,
+      "step": 6490
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.389235527224049,
+      "learning_rate": 2.2100365997255603e-06,
+      "loss": 0.4552,
+      "step": 6491
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.491143530971966,
+      "learning_rate": 2.2085919464167327e-06,
+      "loss": 0.485,
+      "step": 6492
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0342514837930707,
+      "learning_rate": 2.2071476315757547e-06,
+      "loss": 0.5228,
+      "step": 6493
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.223364424589803,
+      "learning_rate": 2.2057036553777565e-06,
+      "loss": 0.487,
+      "step": 6494
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.560189268405409,
+      "learning_rate": 2.2042600179978217e-06,
+      "loss": 0.4605,
+      "step": 6495
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.351389943504937,
+      "learning_rate": 2.2028167196109984e-06,
+      "loss": 0.5003,
+      "step": 6496
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.376249818883973,
+      "learning_rate": 2.201373760392293e-06,
+      "loss": 0.4755,
+      "step": 6497
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.439258909922372,
+      "learning_rate": 2.1999311405166653e-06,
+      "loss": 0.5427,
+      "step": 6498
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.755062318984289,
+      "learning_rate": 2.1984888601590402e-06,
+      "loss": 0.4781,
+      "step": 6499
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.921949171183079,
+      "learning_rate": 2.1970469194943013e-06,
+      "loss": 0.4875,
+      "step": 6500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7018755707837023,
+      "learning_rate": 2.195605318697284e-06,
+      "loss": 0.4693,
+      "step": 6501
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.82385704350503,
+      "learning_rate": 2.194164057942793e-06,
+      "loss": 0.4924,
+      "step": 6502
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.257643402045547,
+      "learning_rate": 2.1927231374055824e-06,
+      "loss": 0.4715,
+      "step": 6503
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6236029015123603,
+      "learning_rate": 2.191282557260368e-06,
+      "loss": 0.507,
+      "step": 6504
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4173375672537263,
+      "learning_rate": 2.1898423176818266e-06,
+      "loss": 0.2781,
+      "step": 6505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4313013178295546,
+      "learning_rate": 2.188402418844594e-06,
+      "loss": 0.524,
+      "step": 6506
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.657653476048108,
+      "learning_rate": 2.1869628609232595e-06,
+      "loss": 0.5225,
+      "step": 6507
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.238908692011099,
+      "learning_rate": 2.185523644092376e-06,
+      "loss": 0.4742,
+      "step": 6508
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.980411757773278,
+      "learning_rate": 2.1840847685264553e-06,
+      "loss": 0.4809,
+      "step": 6509
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5239284551149863,
+      "learning_rate": 2.182646234399963e-06,
+      "loss": 0.4739,
+      "step": 6510
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.563626712075365,
+      "learning_rate": 2.181208041887327e-06,
+      "loss": 0.4571,
+      "step": 6511
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4780242836837956,
+      "learning_rate": 2.179770191162936e-06,
+      "loss": 0.4831,
+      "step": 6512
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.436323901984778,
+      "learning_rate": 2.1783326824011325e-06,
+      "loss": 0.4959,
+      "step": 6513
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.4950778933302145,
+      "learning_rate": 2.1768955157762165e-06,
+      "loss": 0.4675,
+      "step": 6514
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4979752404460003,
+      "learning_rate": 2.1754586914624524e-06,
+      "loss": 0.521,
+      "step": 6515
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.386143491716298,
+      "learning_rate": 2.174022209634061e-06,
+      "loss": 0.5107,
+      "step": 6516
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4377035374140701,
+      "learning_rate": 2.172586070465218e-06,
+      "loss": 0.3023,
+      "step": 6517
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4976039100297465,
+      "learning_rate": 2.171150274130061e-06,
+      "loss": 0.5108,
+      "step": 6518
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2196694327444084,
+      "learning_rate": 2.169714820802688e-06,
+      "loss": 0.5424,
+      "step": 6519
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8302057409775583,
+      "learning_rate": 2.168279710657149e-06,
+      "loss": 0.4703,
+      "step": 6520
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.038887888405881,
+      "learning_rate": 2.1668449438674573e-06,
+      "loss": 0.4729,
+      "step": 6521
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.133247355994509,
+      "learning_rate": 2.165410520607585e-06,
+      "loss": 0.4512,
+      "step": 6522
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1787010980307087,
+      "learning_rate": 2.1639764410514593e-06,
+      "loss": 0.4989,
+      "step": 6523
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9940474114835585,
+      "learning_rate": 2.162542705372966e-06,
+      "loss": 0.5111,
+      "step": 6524
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.844618303545147,
+      "learning_rate": 2.1611093137459533e-06,
+      "loss": 0.4651,
+      "step": 6525
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.583929710607457,
+      "learning_rate": 2.159676266344222e-06,
+      "loss": 0.5215,
+      "step": 6526
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.819844612276832,
+      "learning_rate": 2.158243563341535e-06,
+      "loss": 0.4907,
+      "step": 6527
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.948870000653969,
+      "learning_rate": 2.1568112049116152e-06,
+      "loss": 0.485,
+      "step": 6528
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 16.187923455515065,
+      "learning_rate": 2.155379191228137e-06,
+      "loss": 0.4617,
+      "step": 6529
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.374536763313081,
+      "learning_rate": 2.1539475224647382e-06,
+      "loss": 0.2883,
+      "step": 6530
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.541234310187119,
+      "learning_rate": 2.1525161987950165e-06,
+      "loss": 0.4455,
+      "step": 6531
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3959615274496793,
+      "learning_rate": 2.1510852203925208e-06,
+      "loss": 0.5333,
+      "step": 6532
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.86919511429457,
+      "learning_rate": 2.149654587430765e-06,
+      "loss": 0.4814,
+      "step": 6533
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3764085954969163,
+      "learning_rate": 2.148224300083216e-06,
+      "loss": 0.4434,
+      "step": 6534
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.349987484406894,
+      "learning_rate": 2.1467943585233037e-06,
+      "loss": 0.451,
+      "step": 6535
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.9126180058123268,
+      "learning_rate": 2.14536476292441e-06,
+      "loss": 0.4652,
+      "step": 6536
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.36580817620812695,
+      "learning_rate": 2.143935513459882e-06,
+      "loss": 0.2811,
+      "step": 6537
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.890855766185078,
+      "learning_rate": 2.1425066103030174e-06,
+      "loss": 0.4826,
+      "step": 6538
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3677649149574853,
+      "learning_rate": 2.141078053627078e-06,
+      "loss": 0.2845,
+      "step": 6539
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.144012759359797,
+      "learning_rate": 2.1396498436052824e-06,
+      "loss": 0.4752,
+      "step": 6540
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.146209440382103,
+      "learning_rate": 2.1382219804108023e-06,
+      "loss": 0.5238,
+      "step": 6541
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.955011752293364,
+      "learning_rate": 2.136794464216774e-06,
+      "loss": 0.4502,
+      "step": 6542
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.15659969168003,
+      "learning_rate": 2.135367295196289e-06,
+      "loss": 0.5104,
+      "step": 6543
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.823211828121459,
+      "learning_rate": 2.133940473522395e-06,
+      "loss": 0.5579,
+      "step": 6544
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.086997525379617,
+      "learning_rate": 2.132513999368098e-06,
+      "loss": 0.4869,
+      "step": 6545
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3695406048218519,
+      "learning_rate": 2.1310878729063645e-06,
+      "loss": 0.2755,
+      "step": 6546
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.905984691196227,
+      "learning_rate": 2.1296620943101187e-06,
+      "loss": 0.4535,
+      "step": 6547
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.831709599882994,
+      "learning_rate": 2.128236663752238e-06,
+      "loss": 0.491,
+      "step": 6548
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.699931476419991,
+      "learning_rate": 2.1268115814055616e-06,
+      "loss": 0.463,
+      "step": 6549
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.107696943828327,
+      "learning_rate": 2.1253868474428892e-06,
+      "loss": 0.4846,
+      "step": 6550
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.749737393396485,
+      "learning_rate": 2.123962462036969e-06,
+      "loss": 0.4939,
+      "step": 6551
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.275360483537335,
+      "learning_rate": 2.122538425360516e-06,
+      "loss": 0.5251,
+      "step": 6552
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5886243777896363,
+      "learning_rate": 2.1211147375862005e-06,
+      "loss": 0.4617,
+      "step": 6553
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.339097840364311,
+      "learning_rate": 2.1196913988866467e-06,
+      "loss": 0.4547,
+      "step": 6554
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.879486978219213,
+      "learning_rate": 2.1182684094344425e-06,
+      "loss": 0.4899,
+      "step": 6555
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.082356090715204,
+      "learning_rate": 2.116845769402128e-06,
+      "loss": 0.4818,
+      "step": 6556
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.83967856738305,
+      "learning_rate": 2.1154234789622023e-06,
+      "loss": 0.495,
+      "step": 6557
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.544075765334444,
+      "learning_rate": 2.114001538287125e-06,
+      "loss": 0.4529,
+      "step": 6558
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.5675983361800228,
+      "learning_rate": 2.1125799475493116e-06,
+      "loss": 0.5039,
+      "step": 6559
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.912912184301816,
+      "learning_rate": 2.1111587069211324e-06,
+      "loss": 0.5059,
+      "step": 6560
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.604457729551446,
+      "learning_rate": 2.1097378165749192e-06,
+      "loss": 0.4774,
+      "step": 6561
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.40651351020640003,
+      "learning_rate": 2.1083172766829624e-06,
+      "loss": 0.2915,
+      "step": 6562
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.523897095412887,
+      "learning_rate": 2.1068970874175028e-06,
+      "loss": 0.5303,
+      "step": 6563
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.509944549093908,
+      "learning_rate": 2.1054772489507454e-06,
+      "loss": 0.4563,
+      "step": 6564
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.910481010310416,
+      "learning_rate": 2.1040577614548523e-06,
+      "loss": 0.485,
+      "step": 6565
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.03983876405138,
+      "learning_rate": 2.1026386251019394e-06,
+      "loss": 0.4764,
+      "step": 6566
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3991083800078537,
+      "learning_rate": 2.1012198400640804e-06,
+      "loss": 0.2836,
+      "step": 6567
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0418554408785226,
+      "learning_rate": 2.099801406513311e-06,
+      "loss": 0.4627,
+      "step": 6568
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.502146205429372,
+      "learning_rate": 2.0983833246216177e-06,
+      "loss": 0.5161,
+      "step": 6569
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.997432743929544,
+      "learning_rate": 2.0969655945609496e-06,
+      "loss": 0.5328,
+      "step": 6570
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6387358889865427,
+      "learning_rate": 2.095548216503214e-06,
+      "loss": 0.5036,
+      "step": 6571
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.788589989589009,
+      "learning_rate": 2.0941311906202672e-06,
+      "loss": 0.5107,
+      "step": 6572
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.800668762098273,
+      "learning_rate": 2.0927145170839324e-06,
+      "loss": 0.4724,
+      "step": 6573
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.747349384787291,
+      "learning_rate": 2.0912981960659873e-06,
+      "loss": 0.4638,
+      "step": 6574
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1187830793462723,
+      "learning_rate": 2.0898822277381614e-06,
+      "loss": 0.4648,
+      "step": 6575
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0786815438806316,
+      "learning_rate": 2.0884666122721504e-06,
+      "loss": 0.4743,
+      "step": 6576
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9354528582009056,
+      "learning_rate": 2.0870513498395976e-06,
+      "loss": 0.4376,
+      "step": 6577
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.991535776007897,
+      "learning_rate": 2.085636440612114e-06,
+      "loss": 0.4976,
+      "step": 6578
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.453547719703615,
+      "learning_rate": 2.0842218847612574e-06,
+      "loss": 0.5203,
+      "step": 6579
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.407378878463731,
+      "learning_rate": 2.0828076824585484e-06,
+      "loss": 0.4975,
+      "step": 6580
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.141417754731103,
+      "learning_rate": 2.081393833875468e-06,
+      "loss": 0.5047,
+      "step": 6581
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.992020008480937,
+      "learning_rate": 2.0799803391834444e-06,
+      "loss": 0.5028,
+      "step": 6582
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.126547522680152,
+      "learning_rate": 2.0785671985538724e-06,
+      "loss": 0.5126,
+      "step": 6583
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0263965878360786,
+      "learning_rate": 2.0771544121581004e-06,
+      "loss": 0.4739,
+      "step": 6584
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.254397517185784,
+      "learning_rate": 2.075741980167431e-06,
+      "loss": 0.5284,
+      "step": 6585
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7803712195506054,
+      "learning_rate": 2.07432990275313e-06,
+      "loss": 0.4611,
+      "step": 6586
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1838639154025254,
+      "learning_rate": 2.0729181800864147e-06,
+      "loss": 0.49,
+      "step": 6587
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.169306219733252,
+      "learning_rate": 2.0715068123384593e-06,
+      "loss": 0.4864,
+      "step": 6588
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.838348616190677,
+      "learning_rate": 2.0700957996803983e-06,
+      "loss": 0.4897,
+      "step": 6589
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.210737141201784,
+      "learning_rate": 2.0686851422833248e-06,
+      "loss": 0.5405,
+      "step": 6590
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.632163019053023,
+      "learning_rate": 2.067274840318282e-06,
+      "loss": 0.4853,
+      "step": 6591
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.582790437213192,
+      "learning_rate": 2.0658648939562752e-06,
+      "loss": 0.4985,
+      "step": 6592
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.004360755924753,
+      "learning_rate": 2.0644553033682675e-06,
+      "loss": 0.5006,
+      "step": 6593
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.090019472671015,
+      "learning_rate": 2.0630460687251722e-06,
+      "loss": 0.466,
+      "step": 6594
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.763604576628799,
+      "learning_rate": 2.0616371901978672e-06,
+      "loss": 0.4647,
+      "step": 6595
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.722306950739988,
+      "learning_rate": 2.060228667957184e-06,
+      "loss": 0.5069,
+      "step": 6596
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.5843195202212645,
+      "learning_rate": 2.05882050217391e-06,
+      "loss": 0.5112,
+      "step": 6597
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3963317040659893,
+      "learning_rate": 2.0574126930187882e-06,
+      "loss": 0.2813,
+      "step": 6598
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.889286098580364,
+      "learning_rate": 2.0560052406625236e-06,
+      "loss": 0.4422,
+      "step": 6599
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8383231981813983,
+      "learning_rate": 2.054598145275772e-06,
+      "loss": 0.4659,
+      "step": 6600
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4893712137593123,
+      "learning_rate": 2.0531914070291485e-06,
+      "loss": 0.4821,
+      "step": 6601
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.4416594524383175,
+      "learning_rate": 2.0517850260932288e-06,
+      "loss": 0.4838,
+      "step": 6602
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.5103879953854715,
+      "learning_rate": 2.050379002638537e-06,
+      "loss": 0.5081,
+      "step": 6603
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.414748545570361,
+      "learning_rate": 2.048973336835559e-06,
+      "loss": 0.4715,
+      "step": 6604
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.037863457468075,
+      "learning_rate": 2.04756802885474e-06,
+      "loss": 0.5239,
+      "step": 6605
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.202537802363849,
+      "learning_rate": 2.0461630788664746e-06,
+      "loss": 0.4966,
+      "step": 6606
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.39207883798742454,
+      "learning_rate": 2.044758487041119e-06,
+      "loss": 0.2603,
+      "step": 6607
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3750383426824774,
+      "learning_rate": 2.0433542535489868e-06,
+      "loss": 0.2788,
+      "step": 6608
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.3609098084235,
+      "learning_rate": 2.0419503785603444e-06,
+      "loss": 0.4158,
+      "step": 6609
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3932252686104325,
+      "learning_rate": 2.0405468622454157e-06,
+      "loss": 0.2978,
+      "step": 6610
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.39242888720791136,
+      "learning_rate": 2.039143704774382e-06,
+      "loss": 0.2783,
+      "step": 6611
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3979657573972493,
+      "learning_rate": 2.037740906317383e-06,
+      "loss": 0.2867,
+      "step": 6612
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.783992357496204,
+      "learning_rate": 2.03633846704451e-06,
+      "loss": 0.515,
+      "step": 6613
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.401496984521365,
+      "learning_rate": 2.034936387125816e-06,
+      "loss": 0.4962,
+      "step": 6614
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.797459314556303,
+      "learning_rate": 2.033534666731308e-06,
+      "loss": 0.5245,
+      "step": 6615
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.617188938314723,
+      "learning_rate": 2.032133306030948e-06,
+      "loss": 0.4741,
+      "step": 6616
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2939092215337524,
+      "learning_rate": 2.0307323051946554e-06,
+      "loss": 0.4698,
+      "step": 6617
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.086564547891484,
+      "learning_rate": 2.0293316643923112e-06,
+      "loss": 0.4575,
+      "step": 6618
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1935054169178483,
+      "learning_rate": 2.0279313837937407e-06,
+      "loss": 0.4831,
+      "step": 6619
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.387596152790208,
+      "learning_rate": 2.026531463568736e-06,
+      "loss": 0.4654,
+      "step": 6620
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.439045452931724,
+      "learning_rate": 2.0251319038870448e-06,
+      "loss": 0.4508,
+      "step": 6621
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4105288014682924,
+      "learning_rate": 2.023732704918364e-06,
+      "loss": 0.4616,
+      "step": 6622
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4016102984795774,
+      "learning_rate": 2.0223338668323534e-06,
+      "loss": 0.2632,
+      "step": 6623
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.480244758146933,
+      "learning_rate": 2.0209353897986288e-06,
+      "loss": 0.4747,
+      "step": 6624
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8636458055513505,
+      "learning_rate": 2.019537273986757e-06,
+      "loss": 0.4501,
+      "step": 6625
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.495533453796403,
+      "learning_rate": 2.0181395195662658e-06,
+      "loss": 0.4848,
+      "step": 6626
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.993271613261735,
+      "learning_rate": 2.0167421267066396e-06,
+      "loss": 0.4708,
+      "step": 6627
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.391600102682154,
+      "learning_rate": 2.015345095577314e-06,
+      "loss": 0.5144,
+      "step": 6628
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5861156279490998,
+      "learning_rate": 2.0139484263476867e-06,
+      "loss": 0.4986,
+      "step": 6629
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.620843403759987,
+      "learning_rate": 2.0125521191871054e-06,
+      "loss": 0.4707,
+      "step": 6630
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.136360094831676,
+      "learning_rate": 2.0111561742648807e-06,
+      "loss": 0.496,
+      "step": 6631
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.725527873264446,
+      "learning_rate": 2.0097605917502722e-06,
+      "loss": 0.5048,
+      "step": 6632
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.794086766903132,
+      "learning_rate": 2.0083653718125028e-06,
+      "loss": 0.4558,
+      "step": 6633
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.698767258674007,
+      "learning_rate": 2.0069705146207443e-06,
+      "loss": 0.4791,
+      "step": 6634
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.763652977087611,
+      "learning_rate": 2.005576020344129e-06,
+      "loss": 0.5379,
+      "step": 6635
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.038832867645454,
+      "learning_rate": 2.004181889151746e-06,
+      "loss": 0.4897,
+      "step": 6636
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.37803969961557427,
+      "learning_rate": 2.002788121212636e-06,
+      "loss": 0.2782,
+      "step": 6637
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3700705622284172,
+      "learning_rate": 2.0013947166957994e-06,
+      "loss": 0.2825,
+      "step": 6638
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7200268481146903,
+      "learning_rate": 2.0000016757701924e-06,
+      "loss": 0.4679,
+      "step": 6639
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.5204479692996546,
+      "learning_rate": 1.998608998604725e-06,
+      "loss": 0.5069,
+      "step": 6640
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9713867156067164,
+      "learning_rate": 1.997216685368262e-06,
+      "loss": 0.4528,
+      "step": 6641
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4042562863108529,
+      "learning_rate": 1.995824736229628e-06,
+      "loss": 0.2913,
+      "step": 6642
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.705800784747902,
+      "learning_rate": 1.9944331513576038e-06,
+      "loss": 0.5297,
+      "step": 6643
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.160862441874734,
+      "learning_rate": 1.9930419309209197e-06,
+      "loss": 0.5003,
+      "step": 6644
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7937778249846206,
+      "learning_rate": 1.9916510750882685e-06,
+      "loss": 0.4678,
+      "step": 6645
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.986113569396944,
+      "learning_rate": 1.990260584028297e-06,
+      "loss": 0.4542,
+      "step": 6646
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7622577898329816,
+      "learning_rate": 1.988870457909604e-06,
+      "loss": 0.5647,
+      "step": 6647
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.763162076800695,
+      "learning_rate": 1.987480696900749e-06,
+      "loss": 0.5019,
+      "step": 6648
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1085688026374583,
+      "learning_rate": 1.9860913011702477e-06,
+      "loss": 0.4876,
+      "step": 6649
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8658503186229956,
+      "learning_rate": 1.984702270886566e-06,
+      "loss": 0.5039,
+      "step": 6650
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6939060812339557,
+      "learning_rate": 1.983313606218128e-06,
+      "loss": 0.4578,
+      "step": 6651
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.924540972425472,
+      "learning_rate": 1.981925307333317e-06,
+      "loss": 0.433,
+      "step": 6652
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.225631219131147,
+      "learning_rate": 1.980537374400465e-06,
+      "loss": 0.4852,
+      "step": 6653
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.611582748594615,
+      "learning_rate": 1.979149807587866e-06,
+      "loss": 0.4558,
+      "step": 6654
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.6445651081261765,
+      "learning_rate": 1.97776260706377e-06,
+      "loss": 0.4415,
+      "step": 6655
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.7074262714928796,
+      "learning_rate": 1.9763757729963746e-06,
+      "loss": 0.4599,
+      "step": 6656
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 15.74724078743025,
+      "learning_rate": 1.974989305553841e-06,
+      "loss": 0.4807,
+      "step": 6657
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0945635315526103,
+      "learning_rate": 1.9736032049042846e-06,
+      "loss": 0.4744,
+      "step": 6658
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.8798250880537513,
+      "learning_rate": 1.9722174712157714e-06,
+      "loss": 0.5031,
+      "step": 6659
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.496708983021654,
+      "learning_rate": 1.97083210465633e-06,
+      "loss": 0.5225,
+      "step": 6660
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3927439188890934,
+      "learning_rate": 1.9694471053939375e-06,
+      "loss": 0.2858,
+      "step": 6661
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.329195781009061,
+      "learning_rate": 1.9680624735965327e-06,
+      "loss": 0.4717,
+      "step": 6662
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.052902935758603,
+      "learning_rate": 1.966678209432004e-06,
+      "loss": 0.5455,
+      "step": 6663
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.135880840180775,
+      "learning_rate": 1.965294313068202e-06,
+      "loss": 0.4829,
+      "step": 6664
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9972107405149027,
+      "learning_rate": 1.9639107846729246e-06,
+      "loss": 0.4954,
+      "step": 6665
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.741981570521192,
+      "learning_rate": 1.962527624413932e-06,
+      "loss": 0.4888,
+      "step": 6666
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3476563986460453,
+      "learning_rate": 1.961144832458938e-06,
+      "loss": 0.4801,
+      "step": 6667
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2704385382685035,
+      "learning_rate": 1.9597624089756078e-06,
+      "loss": 0.4493,
+      "step": 6668
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.22179246096472,
+      "learning_rate": 1.958380354131567e-06,
+      "loss": 0.4326,
+      "step": 6669
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.569162426746741,
+      "learning_rate": 1.9569986680943958e-06,
+      "loss": 0.4879,
+      "step": 6670
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2997814633085314,
+      "learning_rate": 1.9556173510316273e-06,
+      "loss": 0.4856,
+      "step": 6671
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.165196389519241,
+      "learning_rate": 1.9542364031107486e-06,
+      "loss": 0.5116,
+      "step": 6672
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.687272427117456,
+      "learning_rate": 1.9528558244992063e-06,
+      "loss": 0.4683,
+      "step": 6673
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.774022489030407,
+      "learning_rate": 1.9514756153644025e-06,
+      "loss": 0.4839,
+      "step": 6674
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.2645501448431755,
+      "learning_rate": 1.9500957758736884e-06,
+      "loss": 0.5382,
+      "step": 6675
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.08973949868471,
+      "learning_rate": 1.948716306194376e-06,
+      "loss": 0.4313,
+      "step": 6676
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 12.252981066796574,
+      "learning_rate": 1.947337206493732e-06,
+      "loss": 0.5196,
+      "step": 6677
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.406298772617568,
+      "learning_rate": 1.945958476938975e-06,
+      "loss": 0.4826,
+      "step": 6678
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.00589596108456,
+      "learning_rate": 1.944580117697281e-06,
+      "loss": 0.4756,
+      "step": 6679
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.165176282341102,
+      "learning_rate": 1.9432021289357833e-06,
+      "loss": 0.4429,
+      "step": 6680
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3888964459777289,
+      "learning_rate": 1.941824510821564e-06,
+      "loss": 0.2883,
+      "step": 6681
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.436015501562162,
+      "learning_rate": 1.9404472635216674e-06,
+      "loss": 0.4873,
+      "step": 6682
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3966412423544443,
+      "learning_rate": 1.9390703872030885e-06,
+      "loss": 0.483,
+      "step": 6683
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2871734944659776,
+      "learning_rate": 1.937693882032776e-06,
+      "loss": 0.464,
+      "step": 6684
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.4062553537747124,
+      "learning_rate": 1.936317748177637e-06,
+      "loss": 0.4799,
+      "step": 6685
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.928624243091815,
+      "learning_rate": 1.934941985804536e-06,
+      "loss": 0.495,
+      "step": 6686
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1248495697574983,
+      "learning_rate": 1.933566595080284e-06,
+      "loss": 0.5198,
+      "step": 6687
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0072320422555947,
+      "learning_rate": 1.9321915761716536e-06,
+      "loss": 0.4662,
+      "step": 6688
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0632712659668186,
+      "learning_rate": 1.930816929245373e-06,
+      "loss": 0.4706,
+      "step": 6689
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.435831383034639,
+      "learning_rate": 1.9294426544681183e-06,
+      "loss": 0.5322,
+      "step": 6690
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4595182921872782,
+      "learning_rate": 1.9280687520065285e-06,
+      "loss": 0.2822,
+      "step": 6691
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9618996779537,
+      "learning_rate": 1.926695222027194e-06,
+      "loss": 0.4906,
+      "step": 6692
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3954144625680373,
+      "learning_rate": 1.9253220646966596e-06,
+      "loss": 0.2832,
+      "step": 6693
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.129239534546395,
+      "learning_rate": 1.923949280181423e-06,
+      "loss": 0.5024,
+      "step": 6694
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9768948135832054,
+      "learning_rate": 1.922576868647943e-06,
+      "loss": 0.4534,
+      "step": 6695
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6986016762814344,
+      "learning_rate": 1.921204830262625e-06,
+      "loss": 0.4397,
+      "step": 6696
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4400351017674673,
+      "learning_rate": 1.919833165191836e-06,
+      "loss": 0.467,
+      "step": 6697
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 15.162821010941455,
+      "learning_rate": 1.918461873601896e-06,
+      "loss": 0.5221,
+      "step": 6698
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 24.127703172874547,
+      "learning_rate": 1.917090955659076e-06,
+      "loss": 0.4531,
+      "step": 6699
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.119285893941949,
+      "learning_rate": 1.9157204115296063e-06,
+      "loss": 0.4923,
+      "step": 6700
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3851673680878158,
+      "learning_rate": 1.914350241379671e-06,
+      "loss": 0.2584,
+      "step": 6701
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.5363290646287995,
+      "learning_rate": 1.9129804453754053e-06,
+      "loss": 0.4822,
+      "step": 6702
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0516577558027076,
+      "learning_rate": 1.911611023682905e-06,
+      "loss": 0.4871,
+      "step": 6703
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.626083733102078,
+      "learning_rate": 1.9102419764682133e-06,
+      "loss": 0.4897,
+      "step": 6704
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.002842970153063,
+      "learning_rate": 1.908873303897336e-06,
+      "loss": 0.4654,
+      "step": 6705
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.859699816820726,
+      "learning_rate": 1.9075050061362255e-06,
+      "loss": 0.4815,
+      "step": 6706
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.064702216551047,
+      "learning_rate": 1.9061370833507947e-06,
+      "loss": 0.5091,
+      "step": 6707
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.382828775057162,
+      "learning_rate": 1.90476953570691e-06,
+      "loss": 0.5012,
+      "step": 6708
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.38579925124572495,
+      "learning_rate": 1.9034023633703885e-06,
+      "loss": 0.2687,
+      "step": 6709
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3957270128835037,
+      "learning_rate": 1.9020355665070062e-06,
+      "loss": 0.286,
+      "step": 6710
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.6782594030185125,
+      "learning_rate": 1.9006691452824932e-06,
+      "loss": 0.4905,
+      "step": 6711
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.819445666455717,
+      "learning_rate": 1.8993030998625295e-06,
+      "loss": 0.5167,
+      "step": 6712
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2416050227957447,
+      "learning_rate": 1.8979374304127567e-06,
+      "loss": 0.5129,
+      "step": 6713
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.8408328656018567,
+      "learning_rate": 1.896572137098765e-06,
+      "loss": 0.4764,
+      "step": 6714
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3743431214698526,
+      "learning_rate": 1.8952072200860988e-06,
+      "loss": 0.2773,
+      "step": 6715
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.89968966044454,
+      "learning_rate": 1.8938426795402614e-06,
+      "loss": 0.4688,
+      "step": 6716
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.876643808475535,
+      "learning_rate": 1.8924785156267089e-06,
+      "loss": 0.4594,
+      "step": 6717
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.884220792524193,
+      "learning_rate": 1.8911147285108484e-06,
+      "loss": 0.5011,
+      "step": 6718
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4948449915361963,
+      "learning_rate": 1.8897513183580445e-06,
+      "loss": 0.495,
+      "step": 6719
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.751736579598498,
+      "learning_rate": 1.8883882853336183e-06,
+      "loss": 0.4499,
+      "step": 6720
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.397300422362987,
+      "learning_rate": 1.8870256296028378e-06,
+      "loss": 0.5065,
+      "step": 6721
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 34.64207749468025,
+      "learning_rate": 1.8856633513309313e-06,
+      "loss": 0.4947,
+      "step": 6722
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.35581013427038,
+      "learning_rate": 1.8843014506830826e-06,
+      "loss": 0.511,
+      "step": 6723
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.671892609334541,
+      "learning_rate": 1.8829399278244243e-06,
+      "loss": 0.4853,
+      "step": 6724
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.833914564160339,
+      "learning_rate": 1.8815787829200438e-06,
+      "loss": 0.4701,
+      "step": 6725
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6633696964123352,
+      "learning_rate": 1.880218016134987e-06,
+      "loss": 0.4824,
+      "step": 6726
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.0174235315308895,
+      "learning_rate": 1.878857627634253e-06,
+      "loss": 0.4811,
+      "step": 6727
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.0569707071416765,
+      "learning_rate": 1.8774976175827898e-06,
+      "loss": 0.4703,
+      "step": 6728
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.144200069113256,
+      "learning_rate": 1.8761379861455076e-06,
+      "loss": 0.4715,
+      "step": 6729
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.694799017159299,
+      "learning_rate": 1.874778733487262e-06,
+      "loss": 0.463,
+      "step": 6730
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.544318263532223,
+      "learning_rate": 1.8734198597728698e-06,
+      "loss": 0.5256,
+      "step": 6731
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.41070761755345,
+      "learning_rate": 1.8720613651670999e-06,
+      "loss": 0.4487,
+      "step": 6732
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.679870211466289,
+      "learning_rate": 1.8707032498346711e-06,
+      "loss": 0.5075,
+      "step": 6733
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.155330872148872,
+      "learning_rate": 1.8693455139402621e-06,
+      "loss": 0.4773,
+      "step": 6734
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.491590374960669,
+      "learning_rate": 1.8679881576485042e-06,
+      "loss": 0.4891,
+      "step": 6735
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.93835884101288,
+      "learning_rate": 1.8666311811239796e-06,
+      "loss": 0.5552,
+      "step": 6736
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.384879777179308,
+      "learning_rate": 1.8652745845312253e-06,
+      "loss": 0.428,
+      "step": 6737
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.4630108654882985,
+      "learning_rate": 1.8639183680347345e-06,
+      "loss": 0.4877,
+      "step": 6738
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.414748702346305,
+      "learning_rate": 1.8625625317989548e-06,
+      "loss": 0.5022,
+      "step": 6739
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.363931749627895,
+      "learning_rate": 1.8612070759882827e-06,
+      "loss": 0.4271,
+      "step": 6740
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.143637538640232,
+      "learning_rate": 1.8598520007670744e-06,
+      "loss": 0.5254,
+      "step": 6741
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6045899404369868,
+      "learning_rate": 1.8584973062996382e-06,
+      "loss": 0.486,
+      "step": 6742
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.847727794644387,
+      "learning_rate": 1.8571429927502322e-06,
+      "loss": 0.4913,
+      "step": 6743
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.907661904706295,
+      "learning_rate": 1.8557890602830731e-06,
+      "loss": 0.4724,
+      "step": 6744
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.39536058406617874,
+      "learning_rate": 1.854435509062334e-06,
+      "loss": 0.2892,
+      "step": 6745
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9798794040979955,
+      "learning_rate": 1.8530823392521302e-06,
+      "loss": 0.5372,
+      "step": 6746
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.625491229133372,
+      "learning_rate": 1.8517295510165416e-06,
+      "loss": 0.5368,
+      "step": 6747
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.744839264908361,
+      "learning_rate": 1.8503771445196005e-06,
+      "loss": 0.4597,
+      "step": 6748
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7623243303189424,
+      "learning_rate": 1.849025119925288e-06,
+      "loss": 0.5443,
+      "step": 6749
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.020189421234123,
+      "learning_rate": 1.847673477397542e-06,
+      "loss": 0.4938,
+      "step": 6750
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.116777516951569,
+      "learning_rate": 1.8463222171002565e-06,
+      "loss": 0.5142,
+      "step": 6751
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.84348163487612,
+      "learning_rate": 1.8449713391972734e-06,
+      "loss": 0.4822,
+      "step": 6752
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.8342850642947064,
+      "learning_rate": 1.8436208438523916e-06,
+      "loss": 0.5351,
+      "step": 6753
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.482850725133086,
+      "learning_rate": 1.8422707312293663e-06,
+      "loss": 0.5164,
+      "step": 6754
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.354460182422603,
+      "learning_rate": 1.8409210014918994e-06,
+      "loss": 0.5116,
+      "step": 6755
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.815516795500333,
+      "learning_rate": 1.839571654803654e-06,
+      "loss": 0.5097,
+      "step": 6756
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.965720239353216,
+      "learning_rate": 1.8382226913282391e-06,
+      "loss": 0.5313,
+      "step": 6757
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.6318429171830084,
+      "learning_rate": 1.8368741112292255e-06,
+      "loss": 0.5096,
+      "step": 6758
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.123682531228758,
+      "learning_rate": 1.8355259146701282e-06,
+      "loss": 0.5232,
+      "step": 6759
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7925000904749333,
+      "learning_rate": 1.8341781018144256e-06,
+      "loss": 0.4738,
+      "step": 6760
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5821346227869375,
+      "learning_rate": 1.8328306728255407e-06,
+      "loss": 0.474,
+      "step": 6761
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.3451224443383,
+      "learning_rate": 1.8314836278668558e-06,
+      "loss": 0.5089,
+      "step": 6762
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.183188045191418,
+      "learning_rate": 1.830136967101706e-06,
+      "loss": 0.4933,
+      "step": 6763
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.954388377241453,
+      "learning_rate": 1.8287906906933755e-06,
+      "loss": 0.4842,
+      "step": 6764
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.837913428854258,
+      "learning_rate": 1.8274447988051065e-06,
+      "loss": 0.5187,
+      "step": 6765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.870237706529287,
+      "learning_rate": 1.8260992916000947e-06,
+      "loss": 0.4789,
+      "step": 6766
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 14.286871145534304,
+      "learning_rate": 1.824754169241486e-06,
+      "loss": 0.5723,
+      "step": 6767
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.669385883053252,
+      "learning_rate": 1.8234094318923795e-06,
+      "loss": 0.467,
+      "step": 6768
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.330303993204561,
+      "learning_rate": 1.82206507971583e-06,
+      "loss": 0.4429,
+      "step": 6769
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.38190069639599716,
+      "learning_rate": 1.8207211128748482e-06,
+      "loss": 0.2681,
+      "step": 6770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.2734448879363525,
+      "learning_rate": 1.8193775315323903e-06,
+      "loss": 0.4723,
+      "step": 6771
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.6125868229440408,
+      "learning_rate": 1.8180343358513714e-06,
+      "loss": 0.5002,
+      "step": 6772
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.918100258030074,
+      "learning_rate": 1.8166915259946617e-06,
+      "loss": 0.5166,
+      "step": 6773
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.183954358761095,
+      "learning_rate": 1.8153491021250763e-06,
+      "loss": 0.4857,
+      "step": 6774
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.518876331858195,
+      "learning_rate": 1.814007064405392e-06,
+      "loss": 0.4778,
+      "step": 6775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.38607927985236684,
+      "learning_rate": 1.8126654129983367e-06,
+      "loss": 0.2636,
+      "step": 6776
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.37047319278765517,
+      "learning_rate": 1.8113241480665883e-06,
+      "loss": 0.2908,
+      "step": 6777
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.868117585639275,
+      "learning_rate": 1.8099832697727787e-06,
+      "loss": 0.4432,
+      "step": 6778
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3798798274561524,
+      "learning_rate": 1.8086427782794964e-06,
+      "loss": 0.2746,
+      "step": 6779
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.731622298667673,
+      "learning_rate": 1.8073026737492783e-06,
+      "loss": 0.488,
+      "step": 6780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 21.49066895067963,
+      "learning_rate": 1.8059629563446173e-06,
+      "loss": 0.4761,
+      "step": 6781
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.746802737919897,
+      "learning_rate": 1.8046236262279615e-06,
+      "loss": 0.4696,
+      "step": 6782
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.237476791881364,
+      "learning_rate": 1.8032846835617056e-06,
+      "loss": 0.4216,
+      "step": 6783
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1537013378485605,
+      "learning_rate": 1.8019461285082023e-06,
+      "loss": 0.45,
+      "step": 6784
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4003440825223187,
+      "learning_rate": 1.8006079612297583e-06,
+      "loss": 0.2848,
+      "step": 6785
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.792785320186379,
+      "learning_rate": 1.7992701818886271e-06,
+      "loss": 0.4956,
+      "step": 6786
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.636414101512541,
+      "learning_rate": 1.7979327906470206e-06,
+      "loss": 0.4649,
+      "step": 6787
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.518942493402015,
+      "learning_rate": 1.7965957876671047e-06,
+      "loss": 0.5026,
+      "step": 6788
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.067159349390195,
+      "learning_rate": 1.795259173110993e-06,
+      "loss": 0.4754,
+      "step": 6789
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.728366698157248,
+      "learning_rate": 1.7939229471407532e-06,
+      "loss": 0.5054,
+      "step": 6790
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7777027109008525,
+      "learning_rate": 1.792587109918411e-06,
+      "loss": 0.478,
+      "step": 6791
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.550880841382673,
+      "learning_rate": 1.7912516616059377e-06,
+      "loss": 0.475,
+      "step": 6792
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9876850501057506,
+      "learning_rate": 1.7899166023652625e-06,
+      "loss": 0.5458,
+      "step": 6793
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.751635729441481,
+      "learning_rate": 1.788581932358268e-06,
+      "loss": 0.4743,
+      "step": 6794
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.080447893178521,
+      "learning_rate": 1.787247651746784e-06,
+      "loss": 0.4704,
+      "step": 6795
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9723450098476296,
+      "learning_rate": 1.785913760692598e-06,
+      "loss": 0.474,
+      "step": 6796
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.024245783778389,
+      "learning_rate": 1.7845802593574512e-06,
+      "loss": 0.4752,
+      "step": 6797
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0163183763347092,
+      "learning_rate": 1.7832471479030328e-06,
+      "loss": 0.4825,
+      "step": 6798
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.360316876884474,
+      "learning_rate": 1.781914426490986e-06,
+      "loss": 0.5077,
+      "step": 6799
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4530036217567934,
+      "learning_rate": 1.7805820952829095e-06,
+      "loss": 0.4345,
+      "step": 6800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7017319512556823,
+      "learning_rate": 1.7792501544403545e-06,
+      "loss": 0.501,
+      "step": 6801
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.256773524445571,
+      "learning_rate": 1.7779186041248204e-06,
+      "loss": 0.4869,
+      "step": 6802
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.016404842644574,
+      "learning_rate": 1.7765874444977638e-06,
+      "loss": 0.4642,
+      "step": 6803
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4586257275888848,
+      "learning_rate": 1.7752566757205936e-06,
+      "loss": 0.2993,
+      "step": 6804
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.100695977815987,
+      "learning_rate": 1.7739262979546674e-06,
+      "loss": 0.46,
+      "step": 6805
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.302531317320699,
+      "learning_rate": 1.7725963113612998e-06,
+      "loss": 0.4826,
+      "step": 6806
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.231628516267893,
+      "learning_rate": 1.771266716101757e-06,
+      "loss": 0.4612,
+      "step": 6807
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.117366040692751,
+      "learning_rate": 1.7699375123372554e-06,
+      "loss": 0.5028,
+      "step": 6808
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.7862642829441855,
+      "learning_rate": 1.768608700228967e-06,
+      "loss": 0.4983,
+      "step": 6809
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.6483290470974024,
+      "learning_rate": 1.767280279938014e-06,
+      "loss": 0.4523,
+      "step": 6810
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.74530328586815,
+      "learning_rate": 1.7659522516254707e-06,
+      "loss": 0.4825,
+      "step": 6811
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.743289498929049,
+      "learning_rate": 1.764624615452366e-06,
+      "loss": 0.463,
+      "step": 6812
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9086954565998666,
+      "learning_rate": 1.7632973715796825e-06,
+      "loss": 0.4276,
+      "step": 6813
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7355616448116025,
+      "learning_rate": 1.7619705201683495e-06,
+      "loss": 0.4548,
+      "step": 6814
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5211090433321095,
+      "learning_rate": 1.7606440613792547e-06,
+      "loss": 0.4725,
+      "step": 6815
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.766897883067708,
+      "learning_rate": 1.7593179953732364e-06,
+      "loss": 0.485,
+      "step": 6816
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 16.973927751110317,
+      "learning_rate": 1.7579923223110818e-06,
+      "loss": 0.5198,
+      "step": 6817
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.915470273295993,
+      "learning_rate": 1.7566670423535348e-06,
+      "loss": 0.4646,
+      "step": 6818
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.116145832999429,
+      "learning_rate": 1.7553421556612927e-06,
+      "loss": 0.5165,
+      "step": 6819
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.330045056626534,
+      "learning_rate": 1.754017662395e-06,
+      "loss": 0.4953,
+      "step": 6820
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.747982892124763,
+      "learning_rate": 1.7526935627152542e-06,
+      "loss": 0.5053,
+      "step": 6821
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.65118752651015,
+      "learning_rate": 1.75136985678261e-06,
+      "loss": 0.4922,
+      "step": 6822
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.2268641548811554,
+      "learning_rate": 1.750046544757571e-06,
+      "loss": 0.5158,
+      "step": 6823
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3896664537125374,
+      "learning_rate": 1.7487236268005919e-06,
+      "loss": 0.4488,
+      "step": 6824
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.202876966644492,
+      "learning_rate": 1.7474011030720834e-06,
+      "loss": 0.4978,
+      "step": 6825
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.2258439178562535,
+      "learning_rate": 1.7460789737324025e-06,
+      "loss": 0.4734,
+      "step": 6826
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.712705951725461,
+      "learning_rate": 1.7447572389418643e-06,
+      "loss": 0.4742,
+      "step": 6827
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9058134379154983,
+      "learning_rate": 1.7434358988607353e-06,
+      "loss": 0.5057,
+      "step": 6828
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1100502479443364,
+      "learning_rate": 1.7421149536492282e-06,
+      "loss": 0.4464,
+      "step": 6829
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.548947918449908,
+      "learning_rate": 1.7407944034675173e-06,
+      "loss": 0.4704,
+      "step": 6830
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3833191378508897,
+      "learning_rate": 1.7394742484757187e-06,
+      "loss": 0.2785,
+      "step": 6831
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.850286689862948,
+      "learning_rate": 1.7381544888339103e-06,
+      "loss": 0.5036,
+      "step": 6832
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.6414346042689623,
+      "learning_rate": 1.7368351247021136e-06,
+      "loss": 0.4707,
+      "step": 6833
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.98545276227036,
+      "learning_rate": 1.7355161562403078e-06,
+      "loss": 0.4516,
+      "step": 6834
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.904730623139026,
+      "learning_rate": 1.7341975836084247e-06,
+      "loss": 0.474,
+      "step": 6835
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.125025640658642,
+      "learning_rate": 1.7328794069663414e-06,
+      "loss": 0.4473,
+      "step": 6836
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.902503692160915,
+      "learning_rate": 1.7315616264738933e-06,
+      "loss": 0.4737,
+      "step": 6837
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.34283942890316,
+      "learning_rate": 1.7302442422908677e-06,
+      "loss": 0.5221,
+      "step": 6838
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.116081263879703,
+      "learning_rate": 1.7289272545769987e-06,
+      "loss": 0.5134,
+      "step": 6839
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.296186433135132,
+      "learning_rate": 1.7276106634919775e-06,
+      "loss": 0.4658,
+      "step": 6840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.978323657826015,
+      "learning_rate": 1.726294469195448e-06,
+      "loss": 0.4724,
+      "step": 6841
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.620355284181277,
+      "learning_rate": 1.7249786718469968e-06,
+      "loss": 0.5465,
+      "step": 6842
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4019983797520281,
+      "learning_rate": 1.723663271606173e-06,
+      "loss": 0.2829,
+      "step": 6843
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.212085034018598,
+      "learning_rate": 1.7223482686324738e-06,
+      "loss": 0.5182,
+      "step": 6844
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.950630981076666,
+      "learning_rate": 1.7210336630853453e-06,
+      "loss": 0.4905,
+      "step": 6845
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4701958103099497,
+      "learning_rate": 1.7197194551241896e-06,
+      "loss": 0.4393,
+      "step": 6846
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 52.213954644310334,
+      "learning_rate": 1.7184056449083603e-06,
+      "loss": 0.5059,
+      "step": 6847
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.27066076794718,
+      "learning_rate": 1.7170922325971585e-06,
+      "loss": 0.4686,
+      "step": 6848
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8246069073416322,
+      "learning_rate": 1.7157792183498413e-06,
+      "loss": 0.528,
+      "step": 6849
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.001356956618978,
+      "learning_rate": 1.714466602325618e-06,
+      "loss": 0.487,
+      "step": 6850
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.260104980034376,
+      "learning_rate": 1.7131543846836457e-06,
+      "loss": 0.4339,
+      "step": 6851
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.382620471070377,
+      "learning_rate": 1.7118425655830346e-06,
+      "loss": 0.513,
+      "step": 6852
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4384886867535642,
+      "learning_rate": 1.7105311451828483e-06,
+      "loss": 0.4851,
+      "step": 6853
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4949532861746007,
+      "learning_rate": 1.709220123642103e-06,
+      "loss": 0.5106,
+      "step": 6854
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.562797737520399,
+      "learning_rate": 1.707909501119761e-06,
+      "loss": 0.4507,
+      "step": 6855
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.2225375845526765,
+      "learning_rate": 1.706599277774743e-06,
+      "loss": 0.4487,
+      "step": 6856
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.917231889413469,
+      "learning_rate": 1.705289453765915e-06,
+      "loss": 0.4978,
+      "step": 6857
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7460698017882557,
+      "learning_rate": 1.7039800292520997e-06,
+      "loss": 0.4863,
+      "step": 6858
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3942736503656814,
+      "learning_rate": 1.7026710043920702e-06,
+      "loss": 0.5001,
+      "step": 6859
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8465920677539787,
+      "learning_rate": 1.701362379344547e-06,
+      "loss": 0.4892,
+      "step": 6860
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1047667744155554,
+      "learning_rate": 1.7000541542682087e-06,
+      "loss": 0.4844,
+      "step": 6861
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4978585492576073,
+      "learning_rate": 1.6987463293216817e-06,
+      "loss": 0.5201,
+      "step": 6862
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9583944624780534,
+      "learning_rate": 1.697438904663543e-06,
+      "loss": 0.4746,
+      "step": 6863
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3744672915449849,
+      "learning_rate": 1.696131880452322e-06,
+      "loss": 0.2979,
+      "step": 6864
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9606943799204677,
+      "learning_rate": 1.6948252568465001e-06,
+      "loss": 0.433,
+      "step": 6865
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.024546887160117,
+      "learning_rate": 1.6935190340045121e-06,
+      "loss": 0.5139,
+      "step": 6866
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2931544592010726,
+      "learning_rate": 1.692213212084739e-06,
+      "loss": 0.525,
+      "step": 6867
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8658111550083105,
+      "learning_rate": 1.690907791245518e-06,
+      "loss": 0.4863,
+      "step": 6868
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2138419080979106,
+      "learning_rate": 1.6896027716451364e-06,
+      "loss": 0.51,
+      "step": 6869
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.420553101233802,
+      "learning_rate": 1.6882981534418301e-06,
+      "loss": 0.5371,
+      "step": 6870
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9450480884983063,
+      "learning_rate": 1.6869939367937899e-06,
+      "loss": 0.4617,
+      "step": 6871
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3958575763099626,
+      "learning_rate": 1.6856901218591598e-06,
+      "loss": 0.2761,
+      "step": 6872
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.498245325460224,
+      "learning_rate": 1.6843867087960252e-06,
+      "loss": 0.5734,
+      "step": 6873
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0913031999130793,
+      "learning_rate": 1.6830836977624326e-06,
+      "loss": 0.4803,
+      "step": 6874
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2671545691552293,
+      "learning_rate": 1.6817810889163788e-06,
+      "loss": 0.4827,
+      "step": 6875
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.120802081523071,
+      "learning_rate": 1.6804788824158058e-06,
+      "loss": 0.5229,
+      "step": 6876
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8788140618459814,
+      "learning_rate": 1.6791770784186128e-06,
+      "loss": 0.5226,
+      "step": 6877
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7958952994001116,
+      "learning_rate": 1.6778756770826488e-06,
+      "loss": 0.4775,
+      "step": 6878
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.66408450036149,
+      "learning_rate": 1.6765746785657104e-06,
+      "loss": 0.5005,
+      "step": 6879
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8220749882387035,
+      "learning_rate": 1.6752740830255503e-06,
+      "loss": 0.5033,
+      "step": 6880
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1510462028508663,
+      "learning_rate": 1.6739738906198711e-06,
+      "loss": 0.506,
+      "step": 6881
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.151988817116163,
+      "learning_rate": 1.6726741015063225e-06,
+      "loss": 0.5233,
+      "step": 6882
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8861338817375453,
+      "learning_rate": 1.671374715842512e-06,
+      "loss": 0.474,
+      "step": 6883
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.118802413773495,
+      "learning_rate": 1.6700757337859907e-06,
+      "loss": 0.4811,
+      "step": 6884
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8119049688562,
+      "learning_rate": 1.668777155494269e-06,
+      "loss": 0.5372,
+      "step": 6885
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9836822288106464,
+      "learning_rate": 1.6674789811247993e-06,
+      "loss": 0.443,
+      "step": 6886
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.849831308697057,
+      "learning_rate": 1.666181210834994e-06,
+      "loss": 0.4853,
+      "step": 6887
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.268983353886467,
+      "learning_rate": 1.6648838447822084e-06,
+      "loss": 0.5132,
+      "step": 6888
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3597063840243999,
+      "learning_rate": 1.6635868831237551e-06,
+      "loss": 0.2811,
+      "step": 6889
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.028306875601375,
+      "learning_rate": 1.6622903260168954e-06,
+      "loss": 0.5017,
+      "step": 6890
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9960594544849517,
+      "learning_rate": 1.6609941736188395e-06,
+      "loss": 0.4911,
+      "step": 6891
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8086112192654715,
+      "learning_rate": 1.6596984260867516e-06,
+      "loss": 0.5165,
+      "step": 6892
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.6500313769401482,
+      "learning_rate": 1.6584030835777465e-06,
+      "loss": 0.4584,
+      "step": 6893
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.3594587738283295,
+      "learning_rate": 1.6571081462488876e-06,
+      "loss": 0.4852,
+      "step": 6894
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3977013873914147,
+      "learning_rate": 1.655813614257189e-06,
+      "loss": 0.2714,
+      "step": 6895
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4691089881937915,
+      "learning_rate": 1.6545194877596194e-06,
+      "loss": 0.501,
+      "step": 6896
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.694599619576885,
+      "learning_rate": 1.6532257669130968e-06,
+      "loss": 0.4992,
+      "step": 6897
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.954857759807152,
+      "learning_rate": 1.6519324518744867e-06,
+      "loss": 0.4764,
+      "step": 6898
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.184503350558774,
+      "learning_rate": 1.650639542800609e-06,
+      "loss": 0.5034,
+      "step": 6899
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.603860720142931,
+      "learning_rate": 1.6493470398482352e-06,
+      "loss": 0.4916,
+      "step": 6900
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9840523394396095,
+      "learning_rate": 1.6480549431740834e-06,
+      "loss": 0.498,
+      "step": 6901
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.345205170954864,
+      "learning_rate": 1.646763252934825e-06,
+      "loss": 0.438,
+      "step": 6902
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.719697426371623,
+      "learning_rate": 1.645471969287084e-06,
+      "loss": 0.4936,
+      "step": 6903
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.342184409440305,
+      "learning_rate": 1.644181092387432e-06,
+      "loss": 0.4899,
+      "step": 6904
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.462972821871555,
+      "learning_rate": 1.6428906223923902e-06,
+      "loss": 0.4708,
+      "step": 6905
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.378578797146407,
+      "learning_rate": 1.6416005594584355e-06,
+      "loss": 0.5078,
+      "step": 6906
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3877026966989555,
+      "learning_rate": 1.6403109037419895e-06,
+      "loss": 0.2857,
+      "step": 6907
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3928851324364847,
+      "learning_rate": 1.6390216553994292e-06,
+      "loss": 0.2743,
+      "step": 6908
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.70617524171482,
+      "learning_rate": 1.6377328145870824e-06,
+      "loss": 0.4364,
+      "step": 6909
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.6273004819822665,
+      "learning_rate": 1.6364443814612207e-06,
+      "loss": 0.4479,
+      "step": 6910
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4012329942005683,
+      "learning_rate": 1.6351563561780742e-06,
+      "loss": 0.4943,
+      "step": 6911
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.390493831276614,
+      "learning_rate": 1.6338687388938218e-06,
+      "loss": 0.4979,
+      "step": 6912
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.678575965089306,
+      "learning_rate": 1.6325815297645875e-06,
+      "loss": 0.5201,
+      "step": 6913
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7886253238044523,
+      "learning_rate": 1.631294728946452e-06,
+      "loss": 0.5103,
+      "step": 6914
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.040797642262076,
+      "learning_rate": 1.6300083365954462e-06,
+      "loss": 0.5256,
+      "step": 6915
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9410237556508014,
+      "learning_rate": 1.6287223528675478e-06,
+      "loss": 0.4907,
+      "step": 6916
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.312660561542412,
+      "learning_rate": 1.6274367779186845e-06,
+      "loss": 0.5547,
+      "step": 6917
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.29236005709402,
+      "learning_rate": 1.6261516119047393e-06,
+      "loss": 0.5335,
+      "step": 6918
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8266116870302316,
+      "learning_rate": 1.6248668549815445e-06,
+      "loss": 0.4639,
+      "step": 6919
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.425784981580522,
+      "learning_rate": 1.6235825073048772e-06,
+      "loss": 0.4945,
+      "step": 6920
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8831633452962473,
+      "learning_rate": 1.6222985690304732e-06,
+      "loss": 0.4812,
+      "step": 6921
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.341971315283977,
+      "learning_rate": 1.62101504031401e-06,
+      "loss": 0.4952,
+      "step": 6922
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.375083234717395,
+      "learning_rate": 1.6197319213111235e-06,
+      "loss": 0.2808,
+      "step": 6923
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.661568548060713,
+      "learning_rate": 1.6184492121773958e-06,
+      "loss": 0.4733,
+      "step": 6924
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.6005941154697148,
+      "learning_rate": 1.6171669130683593e-06,
+      "loss": 0.4714,
+      "step": 6925
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5807337699845228,
+      "learning_rate": 1.6158850241394957e-06,
+      "loss": 0.4845,
+      "step": 6926
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0945393719834855,
+      "learning_rate": 1.6146035455462395e-06,
+      "loss": 0.5007,
+      "step": 6927
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.27498533064477,
+      "learning_rate": 1.613322477443976e-06,
+      "loss": 0.4679,
+      "step": 6928
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.308743422326728,
+      "learning_rate": 1.6120418199880366e-06,
+      "loss": 0.5115,
+      "step": 6929
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.280405528104632,
+      "learning_rate": 1.6107615733337061e-06,
+      "loss": 0.4774,
+      "step": 6930
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.188332721601546,
+      "learning_rate": 1.6094817376362214e-06,
+      "loss": 0.4276,
+      "step": 6931
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.409810537167152,
+      "learning_rate": 1.6082023130507628e-06,
+      "loss": 0.4661,
+      "step": 6932
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.427263295922954,
+      "learning_rate": 1.6069232997324679e-06,
+      "loss": 0.4427,
+      "step": 6933
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.356548368869052,
+      "learning_rate": 1.6056446978364215e-06,
+      "loss": 0.4321,
+      "step": 6934
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.737976057839552,
+      "learning_rate": 1.6043665075176561e-06,
+      "loss": 0.5139,
+      "step": 6935
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1375376060213114,
+      "learning_rate": 1.6030887289311604e-06,
+      "loss": 0.4912,
+      "step": 6936
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.212226538966337,
+      "learning_rate": 1.6018113622318666e-06,
+      "loss": 0.4554,
+      "step": 6937
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.983090673118139,
+      "learning_rate": 1.6005344075746586e-06,
+      "loss": 0.5335,
+      "step": 6938
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.98839868910532,
+      "learning_rate": 1.5992578651143742e-06,
+      "loss": 0.5077,
+      "step": 6939
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.844196584575487,
+      "learning_rate": 1.5979817350057991e-06,
+      "loss": 0.464,
+      "step": 6940
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4906697921243546,
+      "learning_rate": 1.5967060174036652e-06,
+      "loss": 0.5204,
+      "step": 6941
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5498726502146614,
+      "learning_rate": 1.59543071246266e-06,
+      "loss": 0.4492,
+      "step": 6942
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2878089879584924,
+      "learning_rate": 1.5941558203374198e-06,
+      "loss": 0.4819,
+      "step": 6943
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.227086814524392,
+      "learning_rate": 1.5928813411825267e-06,
+      "loss": 0.5132,
+      "step": 6944
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4355903204459422,
+      "learning_rate": 1.591607275152517e-06,
+      "loss": 0.2793,
+      "step": 6945
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.778087297843016,
+      "learning_rate": 1.5903336224018774e-06,
+      "loss": 0.4566,
+      "step": 6946
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.229988297340312,
+      "learning_rate": 1.5890603830850403e-06,
+      "loss": 0.5066,
+      "step": 6947
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.190488883159964,
+      "learning_rate": 1.58778755735639e-06,
+      "loss": 0.4964,
+      "step": 6948
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.115037419855932,
+      "learning_rate": 1.586515145370262e-06,
+      "loss": 0.5093,
+      "step": 6949
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.0433273162310535,
+      "learning_rate": 1.5852431472809426e-06,
+      "loss": 0.5467,
+      "step": 6950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.902259378038772,
+      "learning_rate": 1.5839715632426623e-06,
+      "loss": 0.4431,
+      "step": 6951
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.046881922062473,
+      "learning_rate": 1.582700393409608e-06,
+      "loss": 0.4647,
+      "step": 6952
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0551115375573485,
+      "learning_rate": 1.5814296379359107e-06,
+      "loss": 0.4749,
+      "step": 6953
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8517137295119745,
+      "learning_rate": 1.5801592969756558e-06,
+      "loss": 0.4912,
+      "step": 6954
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5416989574163282,
+      "learning_rate": 1.5788893706828773e-06,
+      "loss": 0.4851,
+      "step": 6955
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.514253539118914,
+      "learning_rate": 1.5776198592115555e-06,
+      "loss": 0.5141,
+      "step": 6956
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1818672692157275,
+      "learning_rate": 1.5763507627156265e-06,
+      "loss": 0.5166,
+      "step": 6957
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5579442560626666,
+      "learning_rate": 1.5750820813489688e-06,
+      "loss": 0.4395,
+      "step": 6958
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.37812569449369116,
+      "learning_rate": 1.5738138152654175e-06,
+      "loss": 0.284,
+      "step": 6959
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4118830517039185,
+      "learning_rate": 1.572545964618752e-06,
+      "loss": 0.4691,
+      "step": 6960
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.2367589572976145,
+      "learning_rate": 1.5712785295627037e-06,
+      "loss": 0.4877,
+      "step": 6961
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5082334610462254,
+      "learning_rate": 1.5700115102509562e-06,
+      "loss": 0.5332,
+      "step": 6962
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.620691534622045,
+      "learning_rate": 1.5687449068371368e-06,
+      "loss": 0.4843,
+      "step": 6963
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.274460226046598,
+      "learning_rate": 1.5674787194748264e-06,
+      "loss": 0.4977,
+      "step": 6964
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5945182364886015,
+      "learning_rate": 1.566212948317557e-06,
+      "loss": 0.4912,
+      "step": 6965
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4110248600553526,
+      "learning_rate": 1.5649475935188034e-06,
+      "loss": 0.2785,
+      "step": 6966
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.856660539894233,
+      "learning_rate": 1.5636826552319973e-06,
+      "loss": 0.4567,
+      "step": 6967
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.098960386509065,
+      "learning_rate": 1.5624181336105188e-06,
+      "loss": 0.4948,
+      "step": 6968
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.538901537522811,
+      "learning_rate": 1.5611540288076892e-06,
+      "loss": 0.5533,
+      "step": 6969
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.981947379593503,
+      "learning_rate": 1.5598903409767895e-06,
+      "loss": 0.4924,
+      "step": 6970
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.5770111348353915,
+      "learning_rate": 1.5586270702710477e-06,
+      "loss": 0.5037,
+      "step": 6971
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.691300053396237,
+      "learning_rate": 1.5573642168436358e-06,
+      "loss": 0.4495,
+      "step": 6972
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.814086365595459,
+      "learning_rate": 1.5561017808476815e-06,
+      "loss": 0.5256,
+      "step": 6973
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.445765972495321,
+      "learning_rate": 1.5548397624362605e-06,
+      "loss": 0.4677,
+      "step": 6974
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2550694011100423,
+      "learning_rate": 1.5535781617623942e-06,
+      "loss": 0.4531,
+      "step": 6975
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0259580443339718,
+      "learning_rate": 1.5523169789790576e-06,
+      "loss": 0.4683,
+      "step": 6976
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6115300467604454,
+      "learning_rate": 1.5510562142391744e-06,
+      "loss": 0.5,
+      "step": 6977
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7753667600013774,
+      "learning_rate": 1.549795867695616e-06,
+      "loss": 0.5011,
+      "step": 6978
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.36634506445307446,
+      "learning_rate": 1.5485359395012012e-06,
+      "loss": 0.2798,
+      "step": 6979
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2216267839110824,
+      "learning_rate": 1.547276429808703e-06,
+      "loss": 0.4853,
+      "step": 6980
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.768984255303177,
+      "learning_rate": 1.5460173387708427e-06,
+      "loss": 0.4977,
+      "step": 6981
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.984581324745464,
+      "learning_rate": 1.544758666540286e-06,
+      "loss": 0.5261,
+      "step": 6982
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1699444885516885,
+      "learning_rate": 1.5435004132696547e-06,
+      "loss": 0.4547,
+      "step": 6983
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.876707954241641,
+      "learning_rate": 1.5422425791115132e-06,
+      "loss": 0.4851,
+      "step": 6984
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3902926575291109,
+      "learning_rate": 1.5409851642183792e-06,
+      "loss": 0.2664,
+      "step": 6985
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.103098729372193,
+      "learning_rate": 1.5397281687427212e-06,
+      "loss": 0.5146,
+      "step": 6986
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3751406697205222,
+      "learning_rate": 1.5384715928369503e-06,
+      "loss": 0.2793,
+      "step": 6987
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.983441568461677,
+      "learning_rate": 1.5372154366534325e-06,
+      "loss": 0.4999,
+      "step": 6988
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.789786484857457,
+      "learning_rate": 1.5359597003444827e-06,
+      "loss": 0.5376,
+      "step": 6989
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3728577360045415,
+      "learning_rate": 1.5347043840623617e-06,
+      "loss": 0.4804,
+      "step": 6990
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.7929264764151105,
+      "learning_rate": 1.5334494879592786e-06,
+      "loss": 0.4447,
+      "step": 6991
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.906511866418492,
+      "learning_rate": 1.5321950121873968e-06,
+      "loss": 0.4623,
+      "step": 6992
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.284163753588965,
+      "learning_rate": 1.5309409568988264e-06,
+      "loss": 0.4362,
+      "step": 6993
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8621845863716313,
+      "learning_rate": 1.5296873222456232e-06,
+      "loss": 0.4807,
+      "step": 6994
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.631968430721373,
+      "learning_rate": 1.528434108379796e-06,
+      "loss": 0.4559,
+      "step": 6995
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8055420591638676,
+      "learning_rate": 1.5271813154533033e-06,
+      "loss": 0.5252,
+      "step": 6996
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.5281692961791515,
+      "learning_rate": 1.5259289436180468e-06,
+      "loss": 0.4915,
+      "step": 6997
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.9742244531639495,
+      "learning_rate": 1.524676993025883e-06,
+      "loss": 0.4822,
+      "step": 6998
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.908460268546017,
+      "learning_rate": 1.5234254638286183e-06,
+      "loss": 0.4627,
+      "step": 6999
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.496732437970542,
+      "learning_rate": 1.5221743561779988e-06,
+      "loss": 0.4722,
+      "step": 7000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9550845842750357,
+      "learning_rate": 1.520923670225728e-06,
+      "loss": 0.5032,
+      "step": 7001
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7135199157072685,
+      "learning_rate": 1.5196734061234581e-06,
+      "loss": 0.4393,
+      "step": 7002
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4091370979698703,
+      "learning_rate": 1.5184235640227846e-06,
+      "loss": 0.454,
+      "step": 7003
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.430444549153816,
+      "learning_rate": 1.5171741440752568e-06,
+      "loss": 0.5065,
+      "step": 7004
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.350318207631487,
+      "learning_rate": 1.5159251464323732e-06,
+      "loss": 0.4816,
+      "step": 7005
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8326101416589036,
+      "learning_rate": 1.5146765712455747e-06,
+      "loss": 0.4569,
+      "step": 7006
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0339291858632977,
+      "learning_rate": 1.5134284186662585e-06,
+      "loss": 0.4768,
+      "step": 7007
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.833819648548076,
+      "learning_rate": 1.5121806888457674e-06,
+      "loss": 0.4746,
+      "step": 7008
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3880714003881605,
+      "learning_rate": 1.5109333819353904e-06,
+      "loss": 0.2865,
+      "step": 7009
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.263433905364,
+      "learning_rate": 1.5096864980863718e-06,
+      "loss": 0.4943,
+      "step": 7010
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.621902164607379,
+      "learning_rate": 1.5084400374498964e-06,
+      "loss": 0.5128,
+      "step": 7011
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.39147553176709476,
+      "learning_rate": 1.507194000177105e-06,
+      "loss": 0.2797,
+      "step": 7012
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8568929155055125,
+      "learning_rate": 1.5059483864190817e-06,
+      "loss": 0.4784,
+      "step": 7013
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.1798686354803625,
+      "learning_rate": 1.5047031963268617e-06,
+      "loss": 0.52,
+      "step": 7014
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.246856155237056,
+      "learning_rate": 1.503458430051431e-06,
+      "loss": 0.4865,
+      "step": 7015
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.731692880813957,
+      "learning_rate": 1.5022140877437186e-06,
+      "loss": 0.5223,
+      "step": 7016
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1819600091226943,
+      "learning_rate": 1.500970169554608e-06,
+      "loss": 0.4324,
+      "step": 7017
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6128996216661378,
+      "learning_rate": 1.4997266756349265e-06,
+      "loss": 0.4953,
+      "step": 7018
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4256780852149443,
+      "learning_rate": 1.4984836061354524e-06,
+      "loss": 0.2779,
+      "step": 7019
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.292873032152159,
+      "learning_rate": 1.497240961206914e-06,
+      "loss": 0.505,
+      "step": 7020
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.298955644660916,
+      "learning_rate": 1.4959987409999855e-06,
+      "loss": 0.4756,
+      "step": 7021
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.4926206388243175,
+      "learning_rate": 1.4947569456652877e-06,
+      "loss": 0.511,
+      "step": 7022
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.358934866158324,
+      "learning_rate": 1.4935155753533947e-06,
+      "loss": 0.4685,
+      "step": 7023
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1065719676392014,
+      "learning_rate": 1.4922746302148283e-06,
+      "loss": 0.4772,
+      "step": 7024
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4043602973175266,
+      "learning_rate": 1.4910341104000548e-06,
+      "loss": 0.273,
+      "step": 7025
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.92625026033141,
+      "learning_rate": 1.4897940160594926e-06,
+      "loss": 0.4805,
+      "step": 7026
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.7037106572118557,
+      "learning_rate": 1.4885543473435088e-06,
+      "loss": 0.5151,
+      "step": 7027
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5426086376147627,
+      "learning_rate": 1.4873151044024149e-06,
+      "loss": 0.4532,
+      "step": 7028
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.442029644807284,
+      "learning_rate": 1.4860762873864741e-06,
+      "loss": 0.5053,
+      "step": 7029
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6010931506016424,
+      "learning_rate": 1.4848378964458997e-06,
+      "loss": 0.4221,
+      "step": 7030
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.4533082677980045,
+      "learning_rate": 1.4835999317308492e-06,
+      "loss": 0.4614,
+      "step": 7031
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9575226990453505,
+      "learning_rate": 1.4823623933914277e-06,
+      "loss": 0.4662,
+      "step": 7032
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.215115555411596,
+      "learning_rate": 1.4811252815776955e-06,
+      "loss": 0.4433,
+      "step": 7033
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.140866394020413,
+      "learning_rate": 1.479888596439652e-06,
+      "loss": 0.4574,
+      "step": 7034
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.44356004303302,
+      "learning_rate": 1.4786523381272521e-06,
+      "loss": 0.5273,
+      "step": 7035
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3679726093861517,
+      "learning_rate": 1.4774165067903984e-06,
+      "loss": 0.2821,
+      "step": 7036
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.303113529775893,
+      "learning_rate": 1.4761811025789352e-06,
+      "loss": 0.4825,
+      "step": 7037
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.002927720413696,
+      "learning_rate": 1.4749461256426617e-06,
+      "loss": 0.4819,
+      "step": 7038
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5122053550189887,
+      "learning_rate": 1.4737115761313249e-06,
+      "loss": 0.4929,
+      "step": 7039
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0905324637547538,
+      "learning_rate": 1.4724774541946145e-06,
+      "loss": 0.5025,
+      "step": 7040
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.497804590795736,
+      "learning_rate": 1.4712437599821743e-06,
+      "loss": 0.4763,
+      "step": 7041
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.39766097400581607,
+      "learning_rate": 1.4700104936435955e-06,
+      "loss": 0.2618,
+      "step": 7042
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.522476408897036,
+      "learning_rate": 1.468777655328414e-06,
+      "loss": 0.5143,
+      "step": 7043
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1650470140706926,
+      "learning_rate": 1.467545245186114e-06,
+      "loss": 0.4441,
+      "step": 7044
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.366066574113921,
+      "learning_rate": 1.4663132633661315e-06,
+      "loss": 0.5401,
+      "step": 7045
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.36639561671698573,
+      "learning_rate": 1.4650817100178494e-06,
+      "loss": 0.2778,
+      "step": 7046
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.020618375555241,
+      "learning_rate": 1.4638505852905955e-06,
+      "loss": 0.4936,
+      "step": 7047
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.689757625850739,
+      "learning_rate": 1.4626198893336507e-06,
+      "loss": 0.5092,
+      "step": 7048
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.034393400047513,
+      "learning_rate": 1.4613896222962376e-06,
+      "loss": 0.4728,
+      "step": 7049
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.4523845319763957,
+      "learning_rate": 1.4601597843275328e-06,
+      "loss": 0.489,
+      "step": 7050
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.437001409263914,
+      "learning_rate": 1.4589303755766586e-06,
+      "loss": 0.4817,
+      "step": 7051
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.707646179861104,
+      "learning_rate": 1.457701396192685e-06,
+      "loss": 0.4966,
+      "step": 7052
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.848143542883891,
+      "learning_rate": 1.4564728463246274e-06,
+      "loss": 0.4903,
+      "step": 7053
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.427403969112525,
+      "learning_rate": 1.4552447261214536e-06,
+      "loss": 0.537,
+      "step": 7054
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.101827987557004,
+      "learning_rate": 1.4540170357320788e-06,
+      "loss": 0.4692,
+      "step": 7055
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3644335561403396,
+      "learning_rate": 1.4527897753053621e-06,
+      "loss": 0.4995,
+      "step": 7056
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2887765545140883,
+      "learning_rate": 1.4515629449901142e-06,
+      "loss": 0.4961,
+      "step": 7057
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.568101598587981,
+      "learning_rate": 1.4503365449350938e-06,
+      "loss": 0.4516,
+      "step": 7058
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.990905004538226,
+      "learning_rate": 1.4491105752890034e-06,
+      "loss": 0.511,
+      "step": 7059
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.781956029210336,
+      "learning_rate": 1.4478850362004976e-06,
+      "loss": 0.4788,
+      "step": 7060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9716828341253936,
+      "learning_rate": 1.4466599278181786e-06,
+      "loss": 0.498,
+      "step": 7061
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.814485787194627,
+      "learning_rate": 1.4454352502905922e-06,
+      "loss": 0.4726,
+      "step": 7062
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.547269553317776,
+      "learning_rate": 1.4442110037662377e-06,
+      "loss": 0.4853,
+      "step": 7063
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.777853890911055,
+      "learning_rate": 1.4429871883935577e-06,
+      "loss": 0.4681,
+      "step": 7064
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3029817708243634,
+      "learning_rate": 1.441763804320942e-06,
+      "loss": 0.4723,
+      "step": 7065
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.9677888479772445,
+      "learning_rate": 1.440540851696733e-06,
+      "loss": 0.4993,
+      "step": 7066
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.056220130649491,
+      "learning_rate": 1.4393183306692176e-06,
+      "loss": 0.4637,
+      "step": 7067
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.9268857612516648,
+      "learning_rate": 1.4380962413866289e-06,
+      "loss": 0.4735,
+      "step": 7068
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8683388589077405,
+      "learning_rate": 1.4368745839971509e-06,
+      "loss": 0.4767,
+      "step": 7069
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9235828532905055,
+      "learning_rate": 1.4356533586489152e-06,
+      "loss": 0.5019,
+      "step": 7070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8533908888133097,
+      "learning_rate": 1.4344325654899965e-06,
+      "loss": 0.4838,
+      "step": 7071
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6434625027090077,
+      "learning_rate": 1.4332122046684211e-06,
+      "loss": 0.5205,
+      "step": 7072
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.505515997995035,
+      "learning_rate": 1.4319922763321641e-06,
+      "loss": 0.5143,
+      "step": 7073
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.546628092074526,
+      "learning_rate": 1.430772780629145e-06,
+      "loss": 0.5232,
+      "step": 7074
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.788616378282748,
+      "learning_rate": 1.4295537177072289e-06,
+      "loss": 0.5291,
+      "step": 7075
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.135239663494389,
+      "learning_rate": 1.4283350877142343e-06,
+      "loss": 0.4903,
+      "step": 7076
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.096368441761572,
+      "learning_rate": 1.4271168907979249e-06,
+      "loss": 0.5474,
+      "step": 7077
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8976750140890566,
+      "learning_rate": 1.4258991271060086e-06,
+      "loss": 0.4863,
+      "step": 7078
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.575454348933659,
+      "learning_rate": 1.4246817967861464e-06,
+      "loss": 0.4865,
+      "step": 7079
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 16.213146065310212,
+      "learning_rate": 1.4234648999859413e-06,
+      "loss": 0.4336,
+      "step": 7080
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.91197997565297,
+      "learning_rate": 1.4222484368529472e-06,
+      "loss": 0.4358,
+      "step": 7081
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.480376630495968,
+      "learning_rate": 1.4210324075346654e-06,
+      "loss": 0.4831,
+      "step": 7082
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.647178023901512,
+      "learning_rate": 1.4198168121785416e-06,
+      "loss": 0.4843,
+      "step": 7083
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.234458022561605,
+      "learning_rate": 1.418601650931974e-06,
+      "loss": 0.4717,
+      "step": 7084
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0433475847108866,
+      "learning_rate": 1.417386923942301e-06,
+      "loss": 0.4717,
+      "step": 7085
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.931028094113179,
+      "learning_rate": 1.4161726313568165e-06,
+      "loss": 0.4741,
+      "step": 7086
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.746918172238769,
+      "learning_rate": 1.4149587733227543e-06,
+      "loss": 0.451,
+      "step": 7087
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.37951963378101267,
+      "learning_rate": 1.4137453499872999e-06,
+      "loss": 0.2642,
+      "step": 7088
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.508601169723972,
+      "learning_rate": 1.4125323614975878e-06,
+      "loss": 0.4945,
+      "step": 7089
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8524286219303203,
+      "learning_rate": 1.4113198080006929e-06,
+      "loss": 0.4829,
+      "step": 7090
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0953409080409657,
+      "learning_rate": 1.4101076896436427e-06,
+      "loss": 0.5081,
+      "step": 7091
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.10413547253054,
+      "learning_rate": 1.4088960065734137e-06,
+      "loss": 0.4952,
+      "step": 7092
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.7504238997929704,
+      "learning_rate": 1.4076847589369225e-06,
+      "loss": 0.5731,
+      "step": 7093
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.352105071021394,
+      "learning_rate": 1.406473946881039e-06,
+      "loss": 0.5104,
+      "step": 7094
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.135393503728689,
+      "learning_rate": 1.4052635705525813e-06,
+      "loss": 0.4922,
+      "step": 7095
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.738509591077141,
+      "learning_rate": 1.4040536300983053e-06,
+      "loss": 0.5032,
+      "step": 7096
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.120201138782238,
+      "learning_rate": 1.4028441256649238e-06,
+      "loss": 0.467,
+      "step": 7097
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.317976163816844,
+      "learning_rate": 1.401635057399095e-06,
+      "loss": 0.4641,
+      "step": 7098
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3791890311091076,
+      "learning_rate": 1.4004264254474193e-06,
+      "loss": 0.2807,
+      "step": 7099
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6140614599144554,
+      "learning_rate": 1.3992182299564493e-06,
+      "loss": 0.4739,
+      "step": 7100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.313902608944506,
+      "learning_rate": 1.3980104710726843e-06,
+      "loss": 0.4841,
+      "step": 7101
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.973189061365011,
+      "learning_rate": 1.3968031489425659e-06,
+      "loss": 0.4354,
+      "step": 7102
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.816179789573375,
+      "learning_rate": 1.395596263712488e-06,
+      "loss": 0.4594,
+      "step": 7103
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.004204636945553,
+      "learning_rate": 1.3943898155287905e-06,
+      "loss": 0.4907,
+      "step": 7104
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.593035612825384,
+      "learning_rate": 1.3931838045377587e-06,
+      "loss": 0.471,
+      "step": 7105
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.102375797580305,
+      "learning_rate": 1.3919782308856233e-06,
+      "loss": 0.4995,
+      "step": 7106
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9212189706918563,
+      "learning_rate": 1.3907730947185665e-06,
+      "loss": 0.4679,
+      "step": 7107
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.28398057726666,
+      "learning_rate": 1.3895683961827166e-06,
+      "loss": 0.531,
+      "step": 7108
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.050614542961073,
+      "learning_rate": 1.3883641354241439e-06,
+      "loss": 0.4936,
+      "step": 7109
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.93967369431864,
+      "learning_rate": 1.3871603125888705e-06,
+      "loss": 0.4841,
+      "step": 7110
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 20.432138137921978,
+      "learning_rate": 1.3859569278228668e-06,
+      "loss": 0.4865,
+      "step": 7111
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.032441936403832,
+      "learning_rate": 1.3847539812720434e-06,
+      "loss": 0.4723,
+      "step": 7112
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.159212829199925,
+      "learning_rate": 1.3835514730822647e-06,
+      "loss": 0.4111,
+      "step": 7113
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7219849718549174,
+      "learning_rate": 1.3823494033993363e-06,
+      "loss": 0.4822,
+      "step": 7114
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.29464003795833,
+      "learning_rate": 1.3811477723690148e-06,
+      "loss": 0.4688,
+      "step": 7115
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9577426516996894,
+      "learning_rate": 1.3799465801370032e-06,
+      "loss": 0.4415,
+      "step": 7116
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.425909964529057,
+      "learning_rate": 1.378745826848949e-06,
+      "loss": 0.5021,
+      "step": 7117
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.829059918994497,
+      "learning_rate": 1.3775455126504466e-06,
+      "loss": 0.4748,
+      "step": 7118
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.295823031759024,
+      "learning_rate": 1.3763456376870387e-06,
+      "loss": 0.4879,
+      "step": 7119
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.016290515755181,
+      "learning_rate": 1.3751462021042167e-06,
+      "loss": 0.4245,
+      "step": 7120
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.6345285370701506,
+      "learning_rate": 1.373947206047413e-06,
+      "loss": 0.4644,
+      "step": 7121
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.238836695642783,
+      "learning_rate": 1.3727486496620113e-06,
+      "loss": 0.4974,
+      "step": 7122
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4228273369869666,
+      "learning_rate": 1.3715505330933426e-06,
+      "loss": 0.498,
+      "step": 7123
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.0626328026294605,
+      "learning_rate": 1.3703528564866792e-06,
+      "loss": 0.4608,
+      "step": 7124
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.205661472250954,
+      "learning_rate": 1.3691556199872453e-06,
+      "loss": 0.4367,
+      "step": 7125
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.891093448661352,
+      "learning_rate": 1.367958823740213e-06,
+      "loss": 0.461,
+      "step": 7126
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.706572613617276,
+      "learning_rate": 1.366762467890692e-06,
+      "loss": 0.499,
+      "step": 7127
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.320891789510844,
+      "learning_rate": 1.3655665525837475e-06,
+      "loss": 0.4582,
+      "step": 7128
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6895142562183016,
+      "learning_rate": 1.3643710779643892e-06,
+      "loss": 0.4466,
+      "step": 7129
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.174052228284837,
+      "learning_rate": 1.3631760441775704e-06,
+      "loss": 0.5142,
+      "step": 7130
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.601811764299521,
+      "learning_rate": 1.3619814513681944e-06,
+      "loss": 0.5084,
+      "step": 7131
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.424407238039677,
+      "learning_rate": 1.3607872996811112e-06,
+      "loss": 0.5117,
+      "step": 7132
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8209075935254395,
+      "learning_rate": 1.3595935892611123e-06,
+      "loss": 0.4651,
+      "step": 7133
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4190884373335712,
+      "learning_rate": 1.3584003202529417e-06,
+      "loss": 0.2974,
+      "step": 7134
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8784240040295006,
+      "learning_rate": 1.3572074928012878e-06,
+      "loss": 0.4474,
+      "step": 7135
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1851462615372297,
+      "learning_rate": 1.3560151070507827e-06,
+      "loss": 0.4991,
+      "step": 7136
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3964060752644804,
+      "learning_rate": 1.3548231631460096e-06,
+      "loss": 0.2703,
+      "step": 7137
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.42332212591953206,
+      "learning_rate": 1.3536316612314937e-06,
+      "loss": 0.292,
+      "step": 7138
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.390150611703037,
+      "learning_rate": 1.3524406014517116e-06,
+      "loss": 0.4923,
+      "step": 7139
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.08910960669878,
+      "learning_rate": 1.3512499839510795e-06,
+      "loss": 0.5427,
+      "step": 7140
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7627715202106917,
+      "learning_rate": 1.3500598088739664e-06,
+      "loss": 0.4771,
+      "step": 7141
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9620566155703134,
+      "learning_rate": 1.3488700763646862e-06,
+      "loss": 0.4776,
+      "step": 7142
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.43627013165989664,
+      "learning_rate": 1.3476807865674951e-06,
+      "loss": 0.2843,
+      "step": 7143
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.38106680954001554,
+      "learning_rate": 1.3464919396266018e-06,
+      "loss": 0.267,
+      "step": 7144
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.555315018491086,
+      "learning_rate": 1.3453035356861544e-06,
+      "loss": 0.5128,
+      "step": 7145
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.570065868792922,
+      "learning_rate": 1.3441155748902534e-06,
+      "loss": 0.4874,
+      "step": 7146
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7299963082068577,
+      "learning_rate": 1.342928057382944e-06,
+      "loss": 0.5164,
+      "step": 7147
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.939215470592042,
+      "learning_rate": 1.3417409833082157e-06,
+      "loss": 0.555,
+      "step": 7148
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.692221600770733,
+      "learning_rate": 1.340554352810003e-06,
+      "loss": 0.4668,
+      "step": 7149
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.662925656181425,
+      "learning_rate": 1.3393681660321917e-06,
+      "loss": 0.4499,
+      "step": 7150
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.0429876962906235,
+      "learning_rate": 1.3381824231186113e-06,
+      "loss": 0.4742,
+      "step": 7151
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.38879464061166663,
+      "learning_rate": 1.3369971242130354e-06,
+      "loss": 0.2785,
+      "step": 7152
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.677426918162286,
+      "learning_rate": 1.3358122694591862e-06,
+      "loss": 0.5518,
+      "step": 7153
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.393090293666266,
+      "learning_rate": 1.3346278590007334e-06,
+      "loss": 0.4851,
+      "step": 7154
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0360797286380943,
+      "learning_rate": 1.333443892981287e-06,
+      "loss": 0.4758,
+      "step": 7155
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5939908832325203,
+      "learning_rate": 1.33226037154441e-06,
+      "loss": 0.5397,
+      "step": 7156
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.37260884743677775,
+      "learning_rate": 1.3310772948336086e-06,
+      "loss": 0.2676,
+      "step": 7157
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.329832634567608,
+      "learning_rate": 1.3298946629923337e-06,
+      "loss": 0.5328,
+      "step": 7158
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.3055590958378165,
+      "learning_rate": 1.3287124761639824e-06,
+      "loss": 0.4649,
+      "step": 7159
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.19294839618934,
+      "learning_rate": 1.327530734491902e-06,
+      "loss": 0.4789,
+      "step": 7160
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.054432834808713,
+      "learning_rate": 1.3263494381193792e-06,
+      "loss": 0.4831,
+      "step": 7161
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.349684156983369,
+      "learning_rate": 1.3251685871896519e-06,
+      "loss": 0.4931,
+      "step": 7162
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.946373856909307,
+      "learning_rate": 1.3239881818459044e-06,
+      "loss": 0.5049,
+      "step": 7163
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.9261068635810816,
+      "learning_rate": 1.322808222231261e-06,
+      "loss": 0.5109,
+      "step": 7164
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3877899170220207,
+      "learning_rate": 1.3216287084887985e-06,
+      "loss": 0.2879,
+      "step": 7165
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.212646796067199,
+      "learning_rate": 1.3204496407615374e-06,
+      "loss": 0.5287,
+      "step": 7166
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3528414235368715,
+      "learning_rate": 1.3192710191924413e-06,
+      "loss": 0.476,
+      "step": 7167
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.52371960120505,
+      "learning_rate": 1.3180928439244233e-06,
+      "loss": 0.513,
+      "step": 7168
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3977407353899691,
+      "learning_rate": 1.3169151151003435e-06,
+      "loss": 0.2759,
+      "step": 7169
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.579682262963759,
+      "learning_rate": 1.3157378328630027e-06,
+      "loss": 0.4416,
+      "step": 7170
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.430856575967493,
+      "learning_rate": 1.3145609973551505e-06,
+      "loss": 0.4514,
+      "step": 7171
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9786831919327006,
+      "learning_rate": 1.3133846087194823e-06,
+      "loss": 0.4864,
+      "step": 7172
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.365101125836016,
+      "learning_rate": 1.3122086670986422e-06,
+      "loss": 0.4909,
+      "step": 7173
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.6070834745744085,
+      "learning_rate": 1.3110331726352132e-06,
+      "loss": 0.4797,
+      "step": 7174
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8810140402571656,
+      "learning_rate": 1.3098581254717313e-06,
+      "loss": 0.5113,
+      "step": 7175
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.959827560276627,
+      "learning_rate": 1.3086835257506719e-06,
+      "loss": 0.4857,
+      "step": 7176
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.988980681680758,
+      "learning_rate": 1.3075093736144612e-06,
+      "loss": 0.5108,
+      "step": 7177
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7503303159514165,
+      "learning_rate": 1.30633566920547e-06,
+      "loss": 0.4596,
+      "step": 7178
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8277129743481644,
+      "learning_rate": 1.3051624126660134e-06,
+      "loss": 0.5251,
+      "step": 7179
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.149519310623597,
+      "learning_rate": 1.3039896041383505e-06,
+      "loss": 0.4718,
+      "step": 7180
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3912747687616615,
+      "learning_rate": 1.3028172437646901e-06,
+      "loss": 0.2655,
+      "step": 7181
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.897134494456515,
+      "learning_rate": 1.3016453316871868e-06,
+      "loss": 0.4423,
+      "step": 7182
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.452471306245368,
+      "learning_rate": 1.3004738680479357e-06,
+      "loss": 0.5717,
+      "step": 7183
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5348729038551943,
+      "learning_rate": 1.2993028529889817e-06,
+      "loss": 0.4644,
+      "step": 7184
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.879925153372688,
+      "learning_rate": 1.2981322866523171e-06,
+      "loss": 0.5295,
+      "step": 7185
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.049500530408035,
+      "learning_rate": 1.2969621691798733e-06,
+      "loss": 0.4744,
+      "step": 7186
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1929358781227237,
+      "learning_rate": 1.295792500713533e-06,
+      "loss": 0.491,
+      "step": 7187
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0325245562762984,
+      "learning_rate": 1.2946232813951236e-06,
+      "loss": 0.4963,
+      "step": 7188
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.35770918969231363,
+      "learning_rate": 1.2934545113664142e-06,
+      "loss": 0.2697,
+      "step": 7189
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.4512287771240775,
+      "learning_rate": 1.2922861907691258e-06,
+      "loss": 0.5038,
+      "step": 7190
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0889759902873246,
+      "learning_rate": 1.2911183197449183e-06,
+      "loss": 0.491,
+      "step": 7191
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.021068025629,
+      "learning_rate": 1.2899508984354002e-06,
+      "loss": 0.4553,
+      "step": 7192
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8228092599358945,
+      "learning_rate": 1.2887839269821262e-06,
+      "loss": 0.4999,
+      "step": 7193
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.219842294856967,
+      "learning_rate": 1.2876174055265967e-06,
+      "loss": 0.4934,
+      "step": 7194
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.167130249039396,
+      "learning_rate": 1.2864513342102536e-06,
+      "loss": 0.4831,
+      "step": 7195
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.316981866383826,
+      "learning_rate": 1.285285713174489e-06,
+      "loss": 0.4475,
+      "step": 7196
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7243161943603487,
+      "learning_rate": 1.2841205425606396e-06,
+      "loss": 0.5461,
+      "step": 7197
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.882904294979939,
+      "learning_rate": 1.2829558225099836e-06,
+      "loss": 0.4725,
+      "step": 7198
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2850673398008556,
+      "learning_rate": 1.2817915531637492e-06,
+      "loss": 0.4646,
+      "step": 7199
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.147321971253397,
+      "learning_rate": 1.2806277346631085e-06,
+      "loss": 0.4146,
+      "step": 7200
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3940440270418767,
+      "learning_rate": 1.279464367149178e-06,
+      "loss": 0.2736,
+      "step": 7201
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.772427556173238,
+      "learning_rate": 1.2783014507630175e-06,
+      "loss": 0.432,
+      "step": 7202
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9173847721041177,
+      "learning_rate": 1.2771389856456374e-06,
+      "loss": 0.5021,
+      "step": 7203
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7094389584709844,
+      "learning_rate": 1.275976971937991e-06,
+      "loss": 0.4798,
+      "step": 7204
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.700898873455292,
+      "learning_rate": 1.2748154097809745e-06,
+      "loss": 0.4226,
+      "step": 7205
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.357154413137957,
+      "learning_rate": 1.2736542993154316e-06,
+      "loss": 0.4712,
+      "step": 7206
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2439203701824364,
+      "learning_rate": 1.2724936406821536e-06,
+      "loss": 0.5417,
+      "step": 7207
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.355319757950974,
+      "learning_rate": 1.2713334340218703e-06,
+      "loss": 0.4525,
+      "step": 7208
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.3752539858235675,
+      "learning_rate": 1.2701736794752645e-06,
+      "loss": 0.5319,
+      "step": 7209
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.124274667339558,
+      "learning_rate": 1.2690143771829572e-06,
+      "loss": 0.4874,
+      "step": 7210
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5308795259113253,
+      "learning_rate": 1.267855527285521e-06,
+      "loss": 0.5183,
+      "step": 7211
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.007030989067218,
+      "learning_rate": 1.266697129923467e-06,
+      "loss": 0.4966,
+      "step": 7212
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5032721793991577,
+      "learning_rate": 1.2655391852372583e-06,
+      "loss": 0.4705,
+      "step": 7213
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.687001848517088,
+      "learning_rate": 1.264381693367297e-06,
+      "loss": 0.4501,
+      "step": 7214
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.171404631565638,
+      "learning_rate": 1.2632246544539333e-06,
+      "loss": 0.5104,
+      "step": 7215
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.389215245760195,
+      "learning_rate": 1.2620680686374649e-06,
+      "loss": 0.4617,
+      "step": 7216
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.086604872212371,
+      "learning_rate": 1.2609119360581278e-06,
+      "loss": 0.4537,
+      "step": 7217
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.639282448830096,
+      "learning_rate": 1.2597562568561095e-06,
+      "loss": 0.507,
+      "step": 7218
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.2688426887909126,
+      "learning_rate": 1.2586010311715408e-06,
+      "loss": 0.457,
+      "step": 7219
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6057919299743872,
+      "learning_rate": 1.257446259144494e-06,
+      "loss": 0.4883,
+      "step": 7220
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.652744868108684,
+      "learning_rate": 1.2562919409149916e-06,
+      "loss": 0.5266,
+      "step": 7221
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.506487069747698,
+      "learning_rate": 1.2551380766230003e-06,
+      "loss": 0.4504,
+      "step": 7222
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.721200463180726,
+      "learning_rate": 1.253984666408425e-06,
+      "loss": 0.5391,
+      "step": 7223
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.587136751267675,
+      "learning_rate": 1.2528317104111226e-06,
+      "loss": 0.4762,
+      "step": 7224
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.60600526331206,
+      "learning_rate": 1.2516792087708963e-06,
+      "loss": 0.4778,
+      "step": 7225
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.414247262101044,
+      "learning_rate": 1.2505271616274861e-06,
+      "loss": 0.5555,
+      "step": 7226
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.825290781615607,
+      "learning_rate": 1.2493755691205845e-06,
+      "loss": 0.449,
+      "step": 7227
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6480954161786276,
+      "learning_rate": 1.2482244313898268e-06,
+      "loss": 0.4707,
+      "step": 7228
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.983201251261823,
+      "learning_rate": 1.247073748574789e-06,
+      "loss": 0.4425,
+      "step": 7229
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.23268798491683,
+      "learning_rate": 1.2459235208149984e-06,
+      "loss": 0.4863,
+      "step": 7230
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.158212893310038,
+      "learning_rate": 1.2447737482499245e-06,
+      "loss": 0.486,
+      "step": 7231
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.39411077482926765,
+      "learning_rate": 1.24362443101898e-06,
+      "loss": 0.2728,
+      "step": 7232
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.6550484762658835,
+      "learning_rate": 1.2424755692615214e-06,
+      "loss": 0.4832,
+      "step": 7233
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.411090369997105,
+      "learning_rate": 1.2413271631168544e-06,
+      "loss": 0.4958,
+      "step": 7234
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.962506836956147,
+      "learning_rate": 1.2401792127242284e-06,
+      "loss": 0.4544,
+      "step": 7235
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.20108347936592,
+      "learning_rate": 1.2390317182228334e-06,
+      "loss": 0.493,
+      "step": 7236
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.128608631032135,
+      "learning_rate": 1.2378846797518091e-06,
+      "loss": 0.5128,
+      "step": 7237
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.248431804802335,
+      "learning_rate": 1.2367380974502386e-06,
+      "loss": 0.4942,
+      "step": 7238
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3811295471377116,
+      "learning_rate": 1.2355919714571457e-06,
+      "loss": 0.489,
+      "step": 7239
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.9625382340311006,
+      "learning_rate": 1.234446301911506e-06,
+      "loss": 0.4636,
+      "step": 7240
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.724163064954356,
+      "learning_rate": 1.2333010889522328e-06,
+      "loss": 0.4445,
+      "step": 7241
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.2109693220987285,
+      "learning_rate": 1.2321563327181885e-06,
+      "loss": 0.4753,
+      "step": 7242
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.942252566387309,
+      "learning_rate": 1.2310120333481795e-06,
+      "loss": 0.4545,
+      "step": 7243
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.635965851753865,
+      "learning_rate": 1.229868190980955e-06,
+      "loss": 0.511,
+      "step": 7244
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.027267979900791,
+      "learning_rate": 1.2287248057552094e-06,
+      "loss": 0.4611,
+      "step": 7245
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.881440888654315,
+      "learning_rate": 1.2275818778095821e-06,
+      "loss": 0.4904,
+      "step": 7246
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.803408200962887,
+      "learning_rate": 1.226439407282659e-06,
+      "loss": 0.5077,
+      "step": 7247
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.905036659194412,
+      "learning_rate": 1.225297394312966e-06,
+      "loss": 0.4756,
+      "step": 7248
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.107008706627112,
+      "learning_rate": 1.224155839038977e-06,
+      "loss": 0.4555,
+      "step": 7249
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.98828524691693,
+      "learning_rate": 1.2230147415991117e-06,
+      "loss": 0.4477,
+      "step": 7250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.387847801048511,
+      "learning_rate": 1.2218741021317282e-06,
+      "loss": 0.4944,
+      "step": 7251
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.672447911018028,
+      "learning_rate": 1.2207339207751356e-06,
+      "loss": 0.4548,
+      "step": 7252
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.629341521262038,
+      "learning_rate": 1.2195941976675869e-06,
+      "loss": 0.4885,
+      "step": 7253
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6521064448020324,
+      "learning_rate": 1.2184549329472717e-06,
+      "loss": 0.5083,
+      "step": 7254
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.404437378899303,
+      "learning_rate": 1.2173161267523332e-06,
+      "loss": 0.4437,
+      "step": 7255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7167659684952206,
+      "learning_rate": 1.2161777792208563e-06,
+      "loss": 0.4894,
+      "step": 7256
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9318162001521357,
+      "learning_rate": 1.2150398904908673e-06,
+      "loss": 0.5319,
+      "step": 7257
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.680618503129098,
+      "learning_rate": 1.2139024607003401e-06,
+      "loss": 0.4757,
+      "step": 7258
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.179611536263094,
+      "learning_rate": 1.2127654899871937e-06,
+      "loss": 0.4911,
+      "step": 7259
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.998385390746837,
+      "learning_rate": 1.2116289784892872e-06,
+      "loss": 0.4877,
+      "step": 7260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.1776920403112054,
+      "learning_rate": 1.210492926344427e-06,
+      "loss": 0.4789,
+      "step": 7261
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.41013330000772297,
+      "learning_rate": 1.2093573336903652e-06,
+      "loss": 0.272,
+      "step": 7262
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.69646527149736,
+      "learning_rate": 1.2082222006647942e-06,
+      "loss": 0.4708,
+      "step": 7263
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.475164445860998,
+      "learning_rate": 1.207087527405355e-06,
+      "loss": 0.5405,
+      "step": 7264
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.669278139351816,
+      "learning_rate": 1.2059533140496277e-06,
+      "loss": 0.4823,
+      "step": 7265
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 17.35886708900088,
+      "learning_rate": 1.2048195607351421e-06,
+      "loss": 0.5242,
+      "step": 7266
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.012352235281106,
+      "learning_rate": 1.2036862675993678e-06,
+      "loss": 0.4366,
+      "step": 7267
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.888906270722023,
+      "learning_rate": 1.2025534347797214e-06,
+      "loss": 0.4743,
+      "step": 7268
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.526078939935179,
+      "learning_rate": 1.201421062413564e-06,
+      "loss": 0.504,
+      "step": 7269
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3812342046538907,
+      "learning_rate": 1.2002891506381976e-06,
+      "loss": 0.2515,
+      "step": 7270
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.4777164410069314,
+      "learning_rate": 1.199157699590872e-06,
+      "loss": 0.4572,
+      "step": 7271
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6262796286687906,
+      "learning_rate": 1.1980267094087777e-06,
+      "loss": 0.466,
+      "step": 7272
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.01806687141449,
+      "learning_rate": 1.1968961802290524e-06,
+      "loss": 0.4685,
+      "step": 7273
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.0667217023123206,
+      "learning_rate": 1.1957661121887782e-06,
+      "loss": 0.4608,
+      "step": 7274
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.20772137541221,
+      "learning_rate": 1.1946365054249775e-06,
+      "loss": 0.4534,
+      "step": 7275
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3878796884351164,
+      "learning_rate": 1.1935073600746184e-06,
+      "loss": 0.26,
+      "step": 7276
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.640473013630788,
+      "learning_rate": 1.1923786762746148e-06,
+      "loss": 0.5003,
+      "step": 7277
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.065025284316591,
+      "learning_rate": 1.1912504541618252e-06,
+      "loss": 0.4935,
+      "step": 7278
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.0136342264435525,
+      "learning_rate": 1.1901226938730471e-06,
+      "loss": 0.5299,
+      "step": 7279
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.8181317495160436,
+      "learning_rate": 1.1889953955450273e-06,
+      "loss": 0.5104,
+      "step": 7280
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.547454475473884,
+      "learning_rate": 1.1878685593144558e-06,
+      "loss": 0.4955,
+      "step": 7281
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.776300299880052,
+      "learning_rate": 1.1867421853179622e-06,
+      "loss": 0.5447,
+      "step": 7282
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.276499281585739,
+      "learning_rate": 1.185616273692125e-06,
+      "loss": 0.4347,
+      "step": 7283
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.429462297447585,
+      "learning_rate": 1.1844908245734659e-06,
+      "loss": 0.5011,
+      "step": 7284
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.486845434325268,
+      "learning_rate": 1.183365838098449e-06,
+      "loss": 0.5345,
+      "step": 7285
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.798543395394872,
+      "learning_rate": 1.1822413144034806e-06,
+      "loss": 0.4951,
+      "step": 7286
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.514280810475054,
+      "learning_rate": 1.181117253624916e-06,
+      "loss": 0.4697,
+      "step": 7287
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.39722566765164363,
+      "learning_rate": 1.1799936558990483e-06,
+      "loss": 0.285,
+      "step": 7288
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.711632657889343,
+      "learning_rate": 1.1788705213621199e-06,
+      "loss": 0.4935,
+      "step": 7289
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.744851911279111,
+      "learning_rate": 1.1777478501503154e-06,
+      "loss": 0.5177,
+      "step": 7290
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5532120781300947,
+      "learning_rate": 1.17662564239976e-06,
+      "loss": 0.5274,
+      "step": 7291
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5500768436325196,
+      "learning_rate": 1.1755038982465266e-06,
+      "loss": 0.4103,
+      "step": 7292
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.757095263366809,
+      "learning_rate": 1.174382617826632e-06,
+      "loss": 0.4886,
+      "step": 7293
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.453991950947888,
+      "learning_rate": 1.1732618012760327e-06,
+      "loss": 0.4503,
+      "step": 7294
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.768202910941129,
+      "learning_rate": 1.1721414487306326e-06,
+      "loss": 0.4896,
+      "step": 7295
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.53224838901182,
+      "learning_rate": 1.1710215603262798e-06,
+      "loss": 0.5296,
+      "step": 7296
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.38586893846897813,
+      "learning_rate": 1.1699021361987634e-06,
+      "loss": 0.2651,
+      "step": 7297
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.841703624755327,
+      "learning_rate": 1.1687831764838158e-06,
+      "loss": 0.481,
+      "step": 7298
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.414817323024341,
+      "learning_rate": 1.1676646813171166e-06,
+      "loss": 0.5004,
+      "step": 7299
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.446887312193473,
+      "learning_rate": 1.1665466508342876e-06,
+      "loss": 0.4856,
+      "step": 7300
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.277306288039563,
+      "learning_rate": 1.1654290851708921e-06,
+      "loss": 0.4929,
+      "step": 7301
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.298949028123502,
+      "learning_rate": 1.16431198446244e-06,
+      "loss": 0.4958,
+      "step": 7302
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4645827639006145,
+      "learning_rate": 1.1631953488443847e-06,
+      "loss": 0.289,
+      "step": 7303
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.165927611536244,
+      "learning_rate": 1.1620791784521196e-06,
+      "loss": 0.4807,
+      "step": 7304
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.722916492155908,
+      "learning_rate": 1.1609634734209867e-06,
+      "loss": 0.4947,
+      "step": 7305
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.711689353740015,
+      "learning_rate": 1.1598482338862676e-06,
+      "loss": 0.521,
+      "step": 7306
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3737364377787387,
+      "learning_rate": 1.158733459983188e-06,
+      "loss": 0.4833,
+      "step": 7307
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.660806035903858,
+      "learning_rate": 1.1576191518469193e-06,
+      "loss": 0.4747,
+      "step": 7308
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.174894486674276,
+      "learning_rate": 1.156505309612576e-06,
+      "loss": 0.5026,
+      "step": 7309
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.39392260686214475,
+      "learning_rate": 1.1553919334152137e-06,
+      "loss": 0.2773,
+      "step": 7310
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3808812755584693,
+      "learning_rate": 1.1542790233898333e-06,
+      "loss": 0.2861,
+      "step": 7311
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.543814082845273,
+      "learning_rate": 1.1531665796713814e-06,
+      "loss": 0.4978,
+      "step": 7312
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.549740699320028,
+      "learning_rate": 1.1520546023947421e-06,
+      "loss": 0.4782,
+      "step": 7313
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6541985380652893,
+      "learning_rate": 1.1509430916947483e-06,
+      "loss": 0.5365,
+      "step": 7314
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.130180478308134,
+      "learning_rate": 1.149832047706176e-06,
+      "loss": 0.4333,
+      "step": 7315
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8710790228995458,
+      "learning_rate": 1.1487214705637395e-06,
+      "loss": 0.3974,
+      "step": 7316
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4070497935746913,
+      "learning_rate": 1.1476113604021039e-06,
+      "loss": 0.2704,
+      "step": 7317
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.868088966864799,
+      "learning_rate": 1.1465017173558717e-06,
+      "loss": 0.5038,
+      "step": 7318
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.805418267086768,
+      "learning_rate": 1.1453925415595902e-06,
+      "loss": 0.4722,
+      "step": 7319
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.073066088563762,
+      "learning_rate": 1.1442838331477512e-06,
+      "loss": 0.4881,
+      "step": 7320
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.9303209456326575,
+      "learning_rate": 1.143175592254792e-06,
+      "loss": 0.5013,
+      "step": 7321
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 15.720495078252961,
+      "learning_rate": 1.142067819015088e-06,
+      "loss": 0.4538,
+      "step": 7322
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.371725020559998,
+      "learning_rate": 1.1409605135629603e-06,
+      "loss": 0.4859,
+      "step": 7323
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.58529593832996,
+      "learning_rate": 1.1398536760326762e-06,
+      "loss": 0.4265,
+      "step": 7324
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.77007547177991,
+      "learning_rate": 1.1387473065584404e-06,
+      "loss": 0.4393,
+      "step": 7325
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.828602904232773,
+      "learning_rate": 1.1376414052744055e-06,
+      "loss": 0.5089,
+      "step": 7326
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.817069286463349,
+      "learning_rate": 1.1365359723146673e-06,
+      "loss": 0.4591,
+      "step": 7327
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7526393924648036,
+      "learning_rate": 1.1354310078132618e-06,
+      "loss": 0.4709,
+      "step": 7328
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.872823133888001,
+      "learning_rate": 1.1343265119041685e-06,
+      "loss": 0.5045,
+      "step": 7329
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 20.489944885119275,
+      "learning_rate": 1.1332224847213125e-06,
+      "loss": 0.4859,
+      "step": 7330
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.987730280789074,
+      "learning_rate": 1.1321189263985622e-06,
+      "loss": 0.481,
+      "step": 7331
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.544963132483705,
+      "learning_rate": 1.1310158370697254e-06,
+      "loss": 0.5149,
+      "step": 7332
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.053722000887862,
+      "learning_rate": 1.1299132168685567e-06,
+      "loss": 0.4877,
+      "step": 7333
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.509158140966509,
+      "learning_rate": 1.1288110659287544e-06,
+      "loss": 0.4956,
+      "step": 7334
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.26892913314806,
+      "learning_rate": 1.127709384383955e-06,
+      "loss": 0.4622,
+      "step": 7335
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.518902282649454,
+      "learning_rate": 1.1266081723677435e-06,
+      "loss": 0.4606,
+      "step": 7336
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3974664640799346,
+      "learning_rate": 1.1255074300136437e-06,
+      "loss": 0.2845,
+      "step": 7337
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.982047483151544,
+      "learning_rate": 1.124407157455127e-06,
+      "loss": 0.5271,
+      "step": 7338
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.6354266164526585,
+      "learning_rate": 1.1233073548256019e-06,
+      "loss": 0.4927,
+      "step": 7339
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7210413172511823,
+      "learning_rate": 1.1222080222584265e-06,
+      "loss": 0.4609,
+      "step": 7340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.252668393771056,
+      "learning_rate": 1.1211091598868956e-06,
+      "loss": 0.5103,
+      "step": 7341
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 19.0350925931634,
+      "learning_rate": 1.1200107678442518e-06,
+      "loss": 0.482,
+      "step": 7342
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.445257631766087,
+      "learning_rate": 1.11891284626368e-06,
+      "loss": 0.4945,
+      "step": 7343
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.29377041786483,
+      "learning_rate": 1.1178153952783045e-06,
+      "loss": 0.4562,
+      "step": 7344
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7017348133496717,
+      "learning_rate": 1.1167184150211962e-06,
+      "loss": 0.4869,
+      "step": 7345
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.41848672716065144,
+      "learning_rate": 1.1156219056253692e-06,
+      "loss": 0.2842,
+      "step": 7346
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3612048440720343,
+      "learning_rate": 1.1145258672237764e-06,
+      "loss": 0.491,
+      "step": 7347
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7772149434124365,
+      "learning_rate": 1.1134302999493174e-06,
+      "loss": 0.4796,
+      "step": 7348
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.876571453569028,
+      "learning_rate": 1.1123352039348362e-06,
+      "loss": 0.5249,
+      "step": 7349
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.268543230605333,
+      "learning_rate": 1.1112405793131114e-06,
+      "loss": 0.4988,
+      "step": 7350
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.62763214265825,
+      "learning_rate": 1.1101464262168733e-06,
+      "loss": 0.4827,
+      "step": 7351
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.711594245885522,
+      "learning_rate": 1.1090527447787924e-06,
+      "loss": 0.5091,
+      "step": 7352
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.22113222398689,
+      "learning_rate": 1.1079595351314793e-06,
+      "loss": 0.4846,
+      "step": 7353
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.015858797547741,
+      "learning_rate": 1.1068667974074903e-06,
+      "loss": 0.492,
+      "step": 7354
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.8573846807774648,
+      "learning_rate": 1.1057745317393254e-06,
+      "loss": 0.4277,
+      "step": 7355
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.692683092951361,
+      "learning_rate": 1.1046827382594227e-06,
+      "loss": 0.4671,
+      "step": 7356
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.026873210931686,
+      "learning_rate": 1.1035914171001665e-06,
+      "loss": 0.4763,
+      "step": 7357
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.24945138542691,
+      "learning_rate": 1.1025005683938862e-06,
+      "loss": 0.4866,
+      "step": 7358
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.714536199222808,
+      "learning_rate": 1.1014101922728482e-06,
+      "loss": 0.4624,
+      "step": 7359
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.311053359389749,
+      "learning_rate": 1.1003202888692633e-06,
+      "loss": 0.4915,
+      "step": 7360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.39704227180192214,
+      "learning_rate": 1.0992308583152878e-06,
+      "loss": 0.2674,
+      "step": 7361
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.296477549369513,
+      "learning_rate": 1.09814190074302e-06,
+      "loss": 0.4673,
+      "step": 7362
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.4215984535602395,
+      "learning_rate": 1.0970534162844977e-06,
+      "loss": 0.5006,
+      "step": 7363
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9549282028587447,
+      "learning_rate": 1.0959654050717034e-06,
+      "loss": 0.4468,
+      "step": 7364
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7074000770224247,
+      "learning_rate": 1.0948778672365646e-06,
+      "loss": 0.5205,
+      "step": 7365
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4547325754274483,
+      "learning_rate": 1.0937908029109461e-06,
+      "loss": 0.4824,
+      "step": 7366
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.835727927352113,
+      "learning_rate": 1.0927042122266613e-06,
+      "loss": 0.4669,
+      "step": 7367
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0530002947932804,
+      "learning_rate": 1.0916180953154592e-06,
+      "loss": 0.5074,
+      "step": 7368
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.824800881108723,
+      "learning_rate": 1.0905324523090376e-06,
+      "loss": 0.4419,
+      "step": 7369
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.519938698347909,
+      "learning_rate": 1.0894472833390357e-06,
+      "loss": 0.5343,
+      "step": 7370
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.398272086910393,
+      "learning_rate": 1.0883625885370319e-06,
+      "loss": 0.3945,
+      "step": 7371
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.051517282079546,
+      "learning_rate": 1.0872783680345488e-06,
+      "loss": 0.4851,
+      "step": 7372
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.38093422541277494,
+      "learning_rate": 1.0861946219630527e-06,
+      "loss": 0.2838,
+      "step": 7373
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.2152962448756455,
+      "learning_rate": 1.0851113504539528e-06,
+      "loss": 0.4403,
+      "step": 7374
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.331194895731557,
+      "learning_rate": 1.084028553638597e-06,
+      "loss": 0.4511,
+      "step": 7375
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.400046066286304,
+      "learning_rate": 1.0829462316482798e-06,
+      "loss": 0.4998,
+      "step": 7376
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3687021876416501,
+      "learning_rate": 1.0818643846142373e-06,
+      "loss": 0.2682,
+      "step": 7377
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.4934187457657275,
+      "learning_rate": 1.0807830126676444e-06,
+      "loss": 0.4995,
+      "step": 7378
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4851877675016594,
+      "learning_rate": 1.0797021159396231e-06,
+      "loss": 0.4957,
+      "step": 7379
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.8229173959567433,
+      "learning_rate": 1.0786216945612371e-06,
+      "loss": 0.4803,
+      "step": 7380
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.326098644180291,
+      "learning_rate": 1.0775417486634893e-06,
+      "loss": 0.4868,
+      "step": 7381
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3124660334322016,
+      "learning_rate": 1.076462278377326e-06,
+      "loss": 0.5497,
+      "step": 7382
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.728607617690388,
+      "learning_rate": 1.0753832838336398e-06,
+      "loss": 0.4996,
+      "step": 7383
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5277366643234602,
+      "learning_rate": 1.0743047651632587e-06,
+      "loss": 0.4394,
+      "step": 7384
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.095908631634849,
+      "learning_rate": 1.073226722496959e-06,
+      "loss": 0.5042,
+      "step": 7385
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.063449492463749,
+      "learning_rate": 1.0721491559654579e-06,
+      "loss": 0.4703,
+      "step": 7386
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.641584161468228,
+      "learning_rate": 1.0710720656994116e-06,
+      "loss": 0.5435,
+      "step": 7387
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.842057925361734,
+      "learning_rate": 1.0699954518294225e-06,
+      "loss": 0.434,
+      "step": 7388
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.737125770553423,
+      "learning_rate": 1.0689193144860356e-06,
+      "loss": 0.4807,
+      "step": 7389
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5967396208199154,
+      "learning_rate": 1.0678436537997321e-06,
+      "loss": 0.4974,
+      "step": 7390
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.167592613281043,
+      "learning_rate": 1.066768469900944e-06,
+      "loss": 0.4528,
+      "step": 7391
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.884541403918665,
+      "learning_rate": 1.065693762920037e-06,
+      "loss": 0.5261,
+      "step": 7392
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.615827237311673,
+      "learning_rate": 1.064619532987326e-06,
+      "loss": 0.5183,
+      "step": 7393
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.9418914068892525,
+      "learning_rate": 1.063545780233063e-06,
+      "loss": 0.4836,
+      "step": 7394
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.260893337639337,
+      "learning_rate": 1.0624725047874452e-06,
+      "loss": 0.4659,
+      "step": 7395
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.608127259436277,
+      "learning_rate": 1.061399706780612e-06,
+      "loss": 0.4927,
+      "step": 7396
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3810356232647244,
+      "learning_rate": 1.0603273863426412e-06,
+      "loss": 0.2757,
+      "step": 7397
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.357103649528519,
+      "learning_rate": 1.0592555436035573e-06,
+      "loss": 0.5211,
+      "step": 7398
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.509429179242793,
+      "learning_rate": 1.0581841786933262e-06,
+      "loss": 0.4732,
+      "step": 7399
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.0043053297112206,
+      "learning_rate": 1.0571132917418508e-06,
+      "loss": 0.5272,
+      "step": 7400
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4071572828193608,
+      "learning_rate": 1.0560428828789837e-06,
+      "loss": 0.29,
+      "step": 7401
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.695748734744872,
+      "learning_rate": 1.0549729522345142e-06,
+      "loss": 0.4997,
+      "step": 7402
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.981284683901653,
+      "learning_rate": 1.0539034999381731e-06,
+      "loss": 0.5024,
+      "step": 7403
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.300765316247162,
+      "learning_rate": 1.052834526119637e-06,
+      "loss": 0.491,
+      "step": 7404
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.636975481305436,
+      "learning_rate": 1.0517660309085242e-06,
+      "loss": 0.4722,
+      "step": 7405
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3880795910853226,
+      "learning_rate": 1.0506980144343898e-06,
+      "loss": 0.2887,
+      "step": 7406
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.478229019518038,
+      "learning_rate": 1.0496304768267373e-06,
+      "loss": 0.4791,
+      "step": 7407
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.554171451716907,
+      "learning_rate": 1.048563418215009e-06,
+      "loss": 0.4744,
+      "step": 7408
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.878265953464856,
+      "learning_rate": 1.0474968387285884e-06,
+      "loss": 0.5026,
+      "step": 7409
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.43101956863552143,
+      "learning_rate": 1.046430738496802e-06,
+      "loss": 0.2839,
+      "step": 7410
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3959780642731425,
+      "learning_rate": 1.04536511764892e-06,
+      "loss": 0.2785,
+      "step": 7411
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.211874008856224,
+      "learning_rate": 1.044299976314151e-06,
+      "loss": 0.4814,
+      "step": 7412
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.39431639633175547,
+      "learning_rate": 1.0432353146216455e-06,
+      "loss": 0.2822,
+      "step": 7413
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.78644527335586,
+      "learning_rate": 1.0421711327005014e-06,
+      "loss": 0.5579,
+      "step": 7414
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 19.26075467373079,
+      "learning_rate": 1.0411074306797502e-06,
+      "loss": 0.5203,
+      "step": 7415
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.270339817666043,
+      "learning_rate": 1.0400442086883717e-06,
+      "loss": 0.4459,
+      "step": 7416
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.075172112288939,
+      "learning_rate": 1.038981466855286e-06,
+      "loss": 0.4868,
+      "step": 7417
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.16381593708086,
+      "learning_rate": 1.0379192053093523e-06,
+      "loss": 0.4946,
+      "step": 7418
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9959511336453235,
+      "learning_rate": 1.036857424179374e-06,
+      "loss": 0.525,
+      "step": 7419
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.914676141120922,
+      "learning_rate": 1.0357961235940978e-06,
+      "loss": 0.4736,
+      "step": 7420
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.492541715783809,
+      "learning_rate": 1.0347353036822061e-06,
+      "loss": 0.5343,
+      "step": 7421
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.582069151489943,
+      "learning_rate": 1.0336749645723298e-06,
+      "loss": 0.4509,
+      "step": 7422
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.176174368764357,
+      "learning_rate": 1.0326151063930396e-06,
+      "loss": 0.4543,
+      "step": 7423
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.508276200751804,
+      "learning_rate": 1.031555729272845e-06,
+      "loss": 0.4996,
+      "step": 7424
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.595327058375553,
+      "learning_rate": 1.0304968333401983e-06,
+      "loss": 0.4925,
+      "step": 7425
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.429187085723412,
+      "learning_rate": 1.0294384187234952e-06,
+      "loss": 0.4706,
+      "step": 7426
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.392196719915605,
+      "learning_rate": 1.0283804855510744e-06,
+      "loss": 0.2696,
+      "step": 7427
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8039605579042286,
+      "learning_rate": 1.0273230339512103e-06,
+      "loss": 0.4679,
+      "step": 7428
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3808123556999769,
+      "learning_rate": 1.0262660640521245e-06,
+      "loss": 0.2612,
+      "step": 7429
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.871597883241589,
+      "learning_rate": 1.0252095759819785e-06,
+      "loss": 0.5118,
+      "step": 7430
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.333298460353676,
+      "learning_rate": 1.0241535698688742e-06,
+      "loss": 0.4814,
+      "step": 7431
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.740581554577404,
+      "learning_rate": 1.0230980458408574e-06,
+      "loss": 0.4582,
+      "step": 7432
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.42686391059909,
+      "learning_rate": 1.0220430040259116e-06,
+      "loss": 0.496,
+      "step": 7433
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.584594354502913,
+      "learning_rate": 1.020988444551967e-06,
+      "loss": 0.4311,
+      "step": 7434
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3457118319284898,
+      "learning_rate": 1.0199343675468898e-06,
+      "loss": 0.5125,
+      "step": 7435
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.65350544357006,
+      "learning_rate": 1.0188807731384932e-06,
+      "loss": 0.4975,
+      "step": 7436
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.436338052158908,
+      "learning_rate": 1.0178276614545269e-06,
+      "loss": 0.5105,
+      "step": 7437
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9287582340423373,
+      "learning_rate": 1.016775032622685e-06,
+      "loss": 0.4883,
+      "step": 7438
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.8809776869171655,
+      "learning_rate": 1.0157228867706042e-06,
+      "loss": 0.4848,
+      "step": 7439
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.853555265761985,
+      "learning_rate": 1.0146712240258578e-06,
+      "loss": 0.4777,
+      "step": 7440
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8302546882637576,
+      "learning_rate": 1.013620044515965e-06,
+      "loss": 0.485,
+      "step": 7441
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9582380291951105,
+      "learning_rate": 1.0125693483683863e-06,
+      "loss": 0.4721,
+      "step": 7442
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.33868718039949,
+      "learning_rate": 1.0115191357105192e-06,
+      "loss": 0.4701,
+      "step": 7443
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.078475629817764,
+      "learning_rate": 1.010469406669709e-06,
+      "loss": 0.4694,
+      "step": 7444
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.162175750577533,
+      "learning_rate": 1.0094201613732374e-06,
+      "loss": 0.4778,
+      "step": 7445
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.63745145096291,
+      "learning_rate": 1.008371399948327e-06,
+      "loss": 0.4745,
+      "step": 7446
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.032175336446502,
+      "learning_rate": 1.0073231225221457e-06,
+      "loss": 0.4589,
+      "step": 7447
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.995297482675955,
+      "learning_rate": 1.0062753292218025e-06,
+      "loss": 0.4761,
+      "step": 7448
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.121902974292089,
+      "learning_rate": 1.0052280201743426e-06,
+      "loss": 0.5244,
+      "step": 7449
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.984303350130565,
+      "learning_rate": 1.004181195506757e-06,
+      "loss": 0.488,
+      "step": 7450
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.841484294403754,
+      "learning_rate": 1.0031348553459785e-06,
+      "loss": 0.4605,
+      "step": 7451
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4088020688628267,
+      "learning_rate": 1.002088999818877e-06,
+      "loss": 0.2711,
+      "step": 7452
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.817949968971835,
+      "learning_rate": 1.0010436290522674e-06,
+      "loss": 0.4923,
+      "step": 7453
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.474755027974136,
+      "learning_rate": 9.999987431729052e-07,
+      "loss": 0.4384,
+      "step": 7454
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.176460251949192,
+      "learning_rate": 9.989543423074855e-07,
+      "loss": 0.4704,
+      "step": 7455
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.020805114316525,
+      "learning_rate": 9.979104265826438e-07,
+      "loss": 0.517,
+      "step": 7456
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.77362522434404,
+      "learning_rate": 9.968669961249611e-07,
+      "loss": 0.4468,
+      "step": 7457
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.203885342755972,
+      "learning_rate": 9.95824051060957e-07,
+      "loss": 0.4775,
+      "step": 7458
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.09164589687471,
+      "learning_rate": 9.947815915170895e-07,
+      "loss": 0.4734,
+      "step": 7459
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8431398299319888,
+      "learning_rate": 9.937396176197623e-07,
+      "loss": 0.4824,
+      "step": 7460
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.832339553816365,
+      "learning_rate": 9.926981294953192e-07,
+      "loss": 0.4873,
+      "step": 7461
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3163319843771752,
+      "learning_rate": 9.916571272700421e-07,
+      "loss": 0.4858,
+      "step": 7462
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.001130935895508,
+      "learning_rate": 9.906166110701588e-07,
+      "loss": 0.4849,
+      "step": 7463
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.384647992914807,
+      "learning_rate": 9.895765810218322e-07,
+      "loss": 0.4701,
+      "step": 7464
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.452990765370961,
+      "learning_rate": 9.885370372511727e-07,
+      "loss": 0.4684,
+      "step": 7465
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9263388912614117,
+      "learning_rate": 9.874979798842254e-07,
+      "loss": 0.451,
+      "step": 7466
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.039241922947984,
+      "learning_rate": 9.86459409046983e-07,
+      "loss": 0.5256,
+      "step": 7467
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9171111452202947,
+      "learning_rate": 9.85421324865372e-07,
+      "loss": 0.5075,
+      "step": 7468
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.419993456824345,
+      "learning_rate": 9.843837274652667e-07,
+      "loss": 0.4714,
+      "step": 7469
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.626760293848431,
+      "learning_rate": 9.833466169724792e-07,
+      "loss": 0.5101,
+      "step": 7470
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.790712808270033,
+      "learning_rate": 9.823099935127605e-07,
+      "loss": 0.5263,
+      "step": 7471
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0793190545106706,
+      "learning_rate": 9.812738572118063e-07,
+      "loss": 0.5061,
+      "step": 7472
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.602919210228062,
+      "learning_rate": 9.802382081952528e-07,
+      "loss": 0.5092,
+      "step": 7473
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.451593271639395,
+      "learning_rate": 9.792030465886736e-07,
+      "loss": 0.5086,
+      "step": 7474
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.807114785433626,
+      "learning_rate": 9.781683725175867e-07,
+      "loss": 0.4992,
+      "step": 7475
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4395634683084584,
+      "learning_rate": 9.771341861074523e-07,
+      "loss": 0.5517,
+      "step": 7476
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.664345379792163,
+      "learning_rate": 9.761004874836643e-07,
+      "loss": 0.5302,
+      "step": 7477
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5132092649670663,
+      "learning_rate": 9.750672767715651e-07,
+      "loss": 0.5201,
+      "step": 7478
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.9041618119614165,
+      "learning_rate": 9.740345540964359e-07,
+      "loss": 0.5007,
+      "step": 7479
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.060115187597284,
+      "learning_rate": 9.730023195834948e-07,
+      "loss": 0.4804,
+      "step": 7480
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3275041686278275,
+      "learning_rate": 9.71970573357906e-07,
+      "loss": 0.4818,
+      "step": 7481
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9145803129722374,
+      "learning_rate": 9.709393155447734e-07,
+      "loss": 0.4577,
+      "step": 7482
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.406410642696756,
+      "learning_rate": 9.699085462691376e-07,
+      "loss": 0.4651,
+      "step": 7483
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4383280959228637,
+      "learning_rate": 9.688782656559842e-07,
+      "loss": 0.5075,
+      "step": 7484
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6089213678443857,
+      "learning_rate": 9.678484738302401e-07,
+      "loss": 0.466,
+      "step": 7485
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8363907039274414,
+      "learning_rate": 9.66819170916769e-07,
+      "loss": 0.4683,
+      "step": 7486
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.892078206227808,
+      "learning_rate": 9.65790357040377e-07,
+      "loss": 0.486,
+      "step": 7487
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.078796941851218,
+      "learning_rate": 9.647620323258123e-07,
+      "loss": 0.47,
+      "step": 7488
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3981367544463736,
+      "learning_rate": 9.637341968977638e-07,
+      "loss": 0.2773,
+      "step": 7489
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.707976622199702,
+      "learning_rate": 9.62706850880858e-07,
+      "loss": 0.4835,
+      "step": 7490
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.016690589257326,
+      "learning_rate": 9.616799943996652e-07,
+      "loss": 0.5206,
+      "step": 7491
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.363751158271156,
+      "learning_rate": 9.606536275786965e-07,
+      "loss": 0.5158,
+      "step": 7492
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 47.34109632617501,
+      "learning_rate": 9.596277505423996e-07,
+      "loss": 0.5133,
+      "step": 7493
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.514910149046119,
+      "learning_rate": 9.586023634151675e-07,
+      "loss": 0.4758,
+      "step": 7494
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5829175018101314,
+      "learning_rate": 9.575774663213327e-07,
+      "loss": 0.5121,
+      "step": 7495
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.482179087943264,
+      "learning_rate": 9.565530593851657e-07,
+      "loss": 0.5096,
+      "step": 7496
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1847206800549333,
+      "learning_rate": 9.55529142730881e-07,
+      "loss": 0.4613,
+      "step": 7497
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5781227823248436,
+      "learning_rate": 9.545057164826317e-07,
+      "loss": 0.4733,
+      "step": 7498
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.988045078814089,
+      "learning_rate": 9.534827807645091e-07,
+      "loss": 0.4632,
+      "step": 7499
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5250561585930087,
+      "learning_rate": 9.524603357005502e-07,
+      "loss": 0.4928,
+      "step": 7500
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.384210173911245,
+      "learning_rate": 9.51438381414731e-07,
+      "loss": 0.2671,
+      "step": 7501
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.43337265163589184,
+      "learning_rate": 9.504169180309641e-07,
+      "loss": 0.2788,
+      "step": 7502
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2263262461096356,
+      "learning_rate": 9.493959456731072e-07,
+      "loss": 0.4895,
+      "step": 7503
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.038084121449141,
+      "learning_rate": 9.483754644649573e-07,
+      "loss": 0.4697,
+      "step": 7504
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.009498250594309,
+      "learning_rate": 9.473554745302494e-07,
+      "loss": 0.4754,
+      "step": 7505
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.6462453945682345,
+      "learning_rate": 9.463359759926616e-07,
+      "loss": 0.4839,
+      "step": 7506
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.015906621489798,
+      "learning_rate": 9.453169689758135e-07,
+      "loss": 0.4742,
+      "step": 7507
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.366667136032155,
+      "learning_rate": 9.442984536032612e-07,
+      "loss": 0.2698,
+      "step": 7508
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.629213768902488,
+      "learning_rate": 9.432804299985021e-07,
+      "loss": 0.5114,
+      "step": 7509
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.471531353229579,
+      "learning_rate": 9.422628982849786e-07,
+      "loss": 0.5074,
+      "step": 7510
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.104011791410252,
+      "learning_rate": 9.412458585860657e-07,
+      "loss": 0.4876,
+      "step": 7511
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.6558825716834,
+      "learning_rate": 9.402293110250854e-07,
+      "loss": 0.4905,
+      "step": 7512
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7128543480506,
+      "learning_rate": 9.392132557252986e-07,
+      "loss": 0.4578,
+      "step": 7513
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.258619107767063,
+      "learning_rate": 9.381976928099029e-07,
+      "loss": 0.4685,
+      "step": 7514
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0384540496504338,
+      "learning_rate": 9.371826224020397e-07,
+      "loss": 0.5003,
+      "step": 7515
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.533262805322407,
+      "learning_rate": 9.361680446247923e-07,
+      "loss": 0.442,
+      "step": 7516
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.382570551930112,
+      "learning_rate": 9.351539596011777e-07,
+      "loss": 0.4795,
+      "step": 7517
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3673610600782667,
+      "learning_rate": 9.341403674541605e-07,
+      "loss": 0.2804,
+      "step": 7518
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.238707730698761,
+      "learning_rate": 9.3312726830664e-07,
+      "loss": 0.4869,
+      "step": 7519
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.238816479597212,
+      "learning_rate": 9.321146622814597e-07,
+      "loss": 0.5255,
+      "step": 7520
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.044878485320424,
+      "learning_rate": 9.311025495013998e-07,
+      "loss": 0.4594,
+      "step": 7521
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.103700390953911,
+      "learning_rate": 9.300909300891831e-07,
+      "loss": 0.5281,
+      "step": 7522
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0685544937068037,
+      "learning_rate": 9.290798041674737e-07,
+      "loss": 0.4859,
+      "step": 7523
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6668763799952053,
+      "learning_rate": 9.280691718588713e-07,
+      "loss": 0.5088,
+      "step": 7524
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3903347473529269,
+      "learning_rate": 9.270590332859202e-07,
+      "loss": 0.2821,
+      "step": 7525
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.223786038007647,
+      "learning_rate": 9.260493885711036e-07,
+      "loss": 0.4745,
+      "step": 7526
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.407876409228036,
+      "learning_rate": 9.250402378368428e-07,
+      "loss": 0.5037,
+      "step": 7527
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.90758661605296,
+      "learning_rate": 9.240315812055029e-07,
+      "loss": 0.509,
+      "step": 7528
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.332827176685118,
+      "learning_rate": 9.230234187993859e-07,
+      "loss": 0.5067,
+      "step": 7529
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.201195834225458,
+      "learning_rate": 9.220157507407335e-07,
+      "loss": 0.4525,
+      "step": 7530
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.182122924223946,
+      "learning_rate": 9.210085771517297e-07,
+      "loss": 0.4691,
+      "step": 7531
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.7026341244379766,
+      "learning_rate": 9.200018981545001e-07,
+      "loss": 0.4757,
+      "step": 7532
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.256141991408536,
+      "learning_rate": 9.189957138711053e-07,
+      "loss": 0.5048,
+      "step": 7533
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3164474549605365,
+      "learning_rate": 9.17990024423549e-07,
+      "loss": 0.5106,
+      "step": 7534
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3192383155822314,
+      "learning_rate": 9.169848299337763e-07,
+      "loss": 0.4791,
+      "step": 7535
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.646503453682478,
+      "learning_rate": 9.159801305236687e-07,
+      "loss": 0.4714,
+      "step": 7536
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0069394425352582,
+      "learning_rate": 9.149759263150493e-07,
+      "loss": 0.4195,
+      "step": 7537
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.766895768621995,
+      "learning_rate": 9.139722174296839e-07,
+      "loss": 0.4782,
+      "step": 7538
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.133618937887289,
+      "learning_rate": 9.129690039892736e-07,
+      "loss": 0.5151,
+      "step": 7539
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.57130888845237,
+      "learning_rate": 9.119662861154599e-07,
+      "loss": 0.4982,
+      "step": 7540
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.599047692702887,
+      "learning_rate": 9.109640639298295e-07,
+      "loss": 0.5016,
+      "step": 7541
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8733483527724215,
+      "learning_rate": 9.099623375539019e-07,
+      "loss": 0.4916,
+      "step": 7542
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.147820873371193,
+      "learning_rate": 9.089611071091414e-07,
+      "loss": 0.4993,
+      "step": 7543
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7429639555160885,
+      "learning_rate": 9.079603727169523e-07,
+      "loss": 0.4688,
+      "step": 7544
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1035240176500483,
+      "learning_rate": 9.069601344986734e-07,
+      "loss": 0.4966,
+      "step": 7545
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.68923714938122,
+      "learning_rate": 9.059603925755895e-07,
+      "loss": 0.5394,
+      "step": 7546
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.780755721023002,
+      "learning_rate": 9.049611470689235e-07,
+      "loss": 0.4895,
+      "step": 7547
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2010104698205235,
+      "learning_rate": 9.039623980998347e-07,
+      "loss": 0.478,
+      "step": 7548
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.959111547589863,
+      "learning_rate": 9.029641457894267e-07,
+      "loss": 0.4533,
+      "step": 7549
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.548302393522332,
+      "learning_rate": 9.019663902587416e-07,
+      "loss": 0.4605,
+      "step": 7550
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.9047778855189925,
+      "learning_rate": 9.009691316287599e-07,
+      "loss": 0.4899,
+      "step": 7551
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.6325464190223915,
+      "learning_rate": 8.999723700204011e-07,
+      "loss": 0.4313,
+      "step": 7552
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0131228710554008,
+      "learning_rate": 8.989761055545276e-07,
+      "loss": 0.5077,
+      "step": 7553
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4038562625296731,
+      "learning_rate": 8.979803383519409e-07,
+      "loss": 0.2706,
+      "step": 7554
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5431152360497746,
+      "learning_rate": 8.969850685333786e-07,
+      "loss": 0.5048,
+      "step": 7555
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.689755489415775,
+      "learning_rate": 8.959902962195222e-07,
+      "loss": 0.4599,
+      "step": 7556
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.6809471984102475,
+      "learning_rate": 8.949960215309922e-07,
+      "loss": 0.4827,
+      "step": 7557
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7252581652192953,
+      "learning_rate": 8.940022445883451e-07,
+      "loss": 0.457,
+      "step": 7558
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.024107817947805,
+      "learning_rate": 8.930089655120833e-07,
+      "loss": 0.5138,
+      "step": 7559
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.266694048661424,
+      "learning_rate": 8.920161844226416e-07,
+      "loss": 0.4426,
+      "step": 7560
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9443272058758145,
+      "learning_rate": 8.910239014404016e-07,
+      "loss": 0.4593,
+      "step": 7561
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.22331904050576,
+      "learning_rate": 8.900321166856774e-07,
+      "loss": 0.4787,
+      "step": 7562
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0615556618916417,
+      "learning_rate": 8.890408302787301e-07,
+      "loss": 0.4777,
+      "step": 7563
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.343207814876628,
+      "learning_rate": 8.880500423397531e-07,
+      "loss": 0.5001,
+      "step": 7564
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 18.060392238792165,
+      "learning_rate": 8.870597529888847e-07,
+      "loss": 0.4993,
+      "step": 7565
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4236761717630215,
+      "learning_rate": 8.860699623462021e-07,
+      "loss": 0.5044,
+      "step": 7566
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.076199206948909,
+      "learning_rate": 8.850806705317183e-07,
+      "loss": 0.5365,
+      "step": 7567
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.209187038117141,
+      "learning_rate": 8.84091877665389e-07,
+      "loss": 0.4646,
+      "step": 7568
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.956426866822158,
+      "learning_rate": 8.83103583867111e-07,
+      "loss": 0.4572,
+      "step": 7569
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.073340078634453,
+      "learning_rate": 8.82115789256715e-07,
+      "loss": 0.4828,
+      "step": 7570
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.841202862436687,
+      "learning_rate": 8.811284939539771e-07,
+      "loss": 0.4764,
+      "step": 7571
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.12354625595122,
+      "learning_rate": 8.801416980786099e-07,
+      "loss": 0.4899,
+      "step": 7572
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.899131511819314,
+      "learning_rate": 8.791554017502634e-07,
+      "loss": 0.5057,
+      "step": 7573
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.543293014525811,
+      "learning_rate": 8.781696050885313e-07,
+      "loss": 0.5187,
+      "step": 7574
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.822669586828627,
+      "learning_rate": 8.771843082129455e-07,
+      "loss": 0.4799,
+      "step": 7575
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.119108846844055,
+      "learning_rate": 8.76199511242975e-07,
+      "loss": 0.5056,
+      "step": 7576
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8723410053237948,
+      "learning_rate": 8.752152142980308e-07,
+      "loss": 0.5154,
+      "step": 7577
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.143377313235403,
+      "learning_rate": 8.74231417497463e-07,
+      "loss": 0.4561,
+      "step": 7578
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.25592503357773,
+      "learning_rate": 8.732481209605587e-07,
+      "loss": 0.4755,
+      "step": 7579
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 14.433350246105466,
+      "learning_rate": 8.722653248065466e-07,
+      "loss": 0.4771,
+      "step": 7580
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5476555376847188,
+      "learning_rate": 8.712830291545954e-07,
+      "loss": 0.4915,
+      "step": 7581
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.806692121147838,
+      "learning_rate": 8.70301234123811e-07,
+      "loss": 0.5077,
+      "step": 7582
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2422103109542766,
+      "learning_rate": 8.693199398332381e-07,
+      "loss": 0.5034,
+      "step": 7583
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.56915257524845,
+      "learning_rate": 8.683391464018625e-07,
+      "loss": 0.4809,
+      "step": 7584
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.755248341544295,
+      "learning_rate": 8.673588539486117e-07,
+      "loss": 0.4715,
+      "step": 7585
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.913321395788709,
+      "learning_rate": 8.663790625923451e-07,
+      "loss": 0.4731,
+      "step": 7586
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.741717726855729,
+      "learning_rate": 8.653997724518682e-07,
+      "loss": 0.4414,
+      "step": 7587
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.842435822977686,
+      "learning_rate": 8.644209836459244e-07,
+      "loss": 0.4566,
+      "step": 7588
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7887456447769483,
+      "learning_rate": 8.634426962931925e-07,
+      "loss": 0.4954,
+      "step": 7589
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.510121140740122,
+      "learning_rate": 8.62464910512294e-07,
+      "loss": 0.5012,
+      "step": 7590
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.104217929554705,
+      "learning_rate": 8.614876264217914e-07,
+      "loss": 0.4547,
+      "step": 7591
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.604937263591896,
+      "learning_rate": 8.60510844140181e-07,
+      "loss": 0.5466,
+      "step": 7592
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1973748127711796,
+      "learning_rate": 8.59534563785901e-07,
+      "loss": 0.491,
+      "step": 7593
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.044652779242993,
+      "learning_rate": 8.585587854773308e-07,
+      "loss": 0.4833,
+      "step": 7594
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3841864211818759,
+      "learning_rate": 8.575835093327845e-07,
+      "loss": 0.2708,
+      "step": 7595
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.8601313211677875,
+      "learning_rate": 8.566087354705182e-07,
+      "loss": 0.5137,
+      "step": 7596
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.957196220866098,
+      "learning_rate": 8.556344640087289e-07,
+      "loss": 0.4598,
+      "step": 7597
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.472032308211737,
+      "learning_rate": 8.546606950655467e-07,
+      "loss": 0.4396,
+      "step": 7598
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3969022746642183,
+      "learning_rate": 8.536874287590469e-07,
+      "loss": 0.4561,
+      "step": 7599
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6921088741623542,
+      "learning_rate": 8.527146652072421e-07,
+      "loss": 0.4489,
+      "step": 7600
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.711894454377692,
+      "learning_rate": 8.517424045280809e-07,
+      "loss": 0.4639,
+      "step": 7601
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.442961728228567,
+      "learning_rate": 8.507706468394544e-07,
+      "loss": 0.4982,
+      "step": 7602
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9954425761227617,
+      "learning_rate": 8.497993922591935e-07,
+      "loss": 0.4559,
+      "step": 7603
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.488836425602104,
+      "learning_rate": 8.488286409050623e-07,
+      "loss": 0.5083,
+      "step": 7604
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8289279243013867,
+      "learning_rate": 8.478583928947692e-07,
+      "loss": 0.5008,
+      "step": 7605
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.948748728541699,
+      "learning_rate": 8.46888648345962e-07,
+      "loss": 0.5094,
+      "step": 7606
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.4718988788445655,
+      "learning_rate": 8.459194073762228e-07,
+      "loss": 0.4655,
+      "step": 7607
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.509483066775838,
+      "learning_rate": 8.449506701030774e-07,
+      "loss": 0.4928,
+      "step": 7608
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2006873010216874,
+      "learning_rate": 8.439824366439886e-07,
+      "loss": 0.4697,
+      "step": 7609
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.218840094756955,
+      "learning_rate": 8.430147071163558e-07,
+      "loss": 0.4413,
+      "step": 7610
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.715743445528306,
+      "learning_rate": 8.420474816375213e-07,
+      "loss": 0.5279,
+      "step": 7611
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.712443974215908,
+      "learning_rate": 8.410807603247656e-07,
+      "loss": 0.4795,
+      "step": 7612
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.077618445608209,
+      "learning_rate": 8.401145432953045e-07,
+      "loss": 0.4845,
+      "step": 7613
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5761408165658892,
+      "learning_rate": 8.391488306662976e-07,
+      "loss": 0.5266,
+      "step": 7614
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 120.69296291317141,
+      "learning_rate": 8.381836225548384e-07,
+      "loss": 0.4617,
+      "step": 7615
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.795688968203749,
+      "learning_rate": 8.372189190779639e-07,
+      "loss": 0.4816,
+      "step": 7616
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.36276434404729324,
+      "learning_rate": 8.362547203526456e-07,
+      "loss": 0.2689,
+      "step": 7617
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.848247245851432,
+      "learning_rate": 8.35291026495797e-07,
+      "loss": 0.4663,
+      "step": 7618
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.918157710330075,
+      "learning_rate": 8.343278376242709e-07,
+      "loss": 0.5008,
+      "step": 7619
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8070469096825423,
+      "learning_rate": 8.333651538548543e-07,
+      "loss": 0.4569,
+      "step": 7620
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.39440919470203845,
+      "learning_rate": 8.324029753042773e-07,
+      "loss": 0.2671,
+      "step": 7621
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 15.105341743829195,
+      "learning_rate": 8.314413020892087e-07,
+      "loss": 0.5487,
+      "step": 7622
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3809080064622955,
+      "learning_rate": 8.304801343262525e-07,
+      "loss": 0.2834,
+      "step": 7623
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.163340434292813,
+      "learning_rate": 8.295194721319555e-07,
+      "loss": 0.441,
+      "step": 7624
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.114137880441608,
+      "learning_rate": 8.285593156228006e-07,
+      "loss": 0.458,
+      "step": 7625
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.5170322248361545,
+      "learning_rate": 8.275996649152085e-07,
+      "loss": 0.5263,
+      "step": 7626
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4551457786375366,
+      "learning_rate": 8.266405201255412e-07,
+      "loss": 0.472,
+      "step": 7627
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.856012756932305,
+      "learning_rate": 8.256818813701e-07,
+      "loss": 0.4859,
+      "step": 7628
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6477655727598406,
+      "learning_rate": 8.247237487651205e-07,
+      "loss": 0.4373,
+      "step": 7629
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.302532262207357,
+      "learning_rate": 8.237661224267807e-07,
+      "loss": 0.5059,
+      "step": 7630
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.59985806435492,
+      "learning_rate": 8.228090024711977e-07,
+      "loss": 0.4794,
+      "step": 7631
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6517383656166755,
+      "learning_rate": 8.218523890144225e-07,
+      "loss": 0.4721,
+      "step": 7632
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.518181233255869,
+      "learning_rate": 8.208962821724498e-07,
+      "loss": 0.494,
+      "step": 7633
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5841715955352464,
+      "learning_rate": 8.199406820612115e-07,
+      "loss": 0.4876,
+      "step": 7634
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.4163189025333764,
+      "learning_rate": 8.189855887965764e-07,
+      "loss": 0.462,
+      "step": 7635
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0143145574877463,
+      "learning_rate": 8.180310024943517e-07,
+      "loss": 0.4535,
+      "step": 7636
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2317848867959222,
+      "learning_rate": 8.170769232702869e-07,
+      "loss": 0.4795,
+      "step": 7637
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.879967454747519,
+      "learning_rate": 8.161233512400641e-07,
+      "loss": 0.5032,
+      "step": 7638
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.001534254799429,
+      "learning_rate": 8.15170286519309e-07,
+      "loss": 0.4975,
+      "step": 7639
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.158399012023099,
+      "learning_rate": 8.142177292235854e-07,
+      "loss": 0.4412,
+      "step": 7640
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.239420224536711,
+      "learning_rate": 8.132656794683914e-07,
+      "loss": 0.5083,
+      "step": 7641
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.39938138626280295,
+      "learning_rate": 8.123141373691678e-07,
+      "loss": 0.279,
+      "step": 7642
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.121746115568437,
+      "learning_rate": 8.113631030412932e-07,
+      "loss": 0.5228,
+      "step": 7643
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 55.40971622931736,
+      "learning_rate": 8.104125766000814e-07,
+      "loss": 0.5091,
+      "step": 7644
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.415599166686529,
+      "learning_rate": 8.094625581607901e-07,
+      "loss": 0.4286,
+      "step": 7645
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8663537123512968,
+      "learning_rate": 8.085130478386089e-07,
+      "loss": 0.4945,
+      "step": 7646
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0788541058885555,
+      "learning_rate": 8.075640457486722e-07,
+      "loss": 0.5291,
+      "step": 7647
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.759604974293848,
+      "learning_rate": 8.066155520060476e-07,
+      "loss": 0.5109,
+      "step": 7648
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.38531648619106,
+      "learning_rate": 8.056675667257441e-07,
+      "loss": 0.4863,
+      "step": 7649
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5249180583417297,
+      "learning_rate": 8.047200900227092e-07,
+      "loss": 0.5005,
+      "step": 7650
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.011770236700434,
+      "learning_rate": 8.037731220118256e-07,
+      "loss": 0.4787,
+      "step": 7651
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.08859657732275,
+      "learning_rate": 8.028266628079173e-07,
+      "loss": 0.4825,
+      "step": 7652
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1346043241758497,
+      "learning_rate": 8.018807125257478e-07,
+      "loss": 0.4613,
+      "step": 7653
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5363369790078645,
+      "learning_rate": 8.009352712800128e-07,
+      "loss": 0.4551,
+      "step": 7654
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5513281972700494,
+      "learning_rate": 7.999903391853547e-07,
+      "loss": 0.4675,
+      "step": 7655
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5994424417222914,
+      "learning_rate": 7.990459163563474e-07,
+      "loss": 0.4491,
+      "step": 7656
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.36379511627802935,
+      "learning_rate": 7.981020029075043e-07,
+      "loss": 0.2745,
+      "step": 7657
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6683907096796626,
+      "learning_rate": 7.971585989532793e-07,
+      "loss": 0.4815,
+      "step": 7658
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3073383874278184,
+      "learning_rate": 7.962157046080648e-07,
+      "loss": 0.4396,
+      "step": 7659
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.9466240934037082,
+      "learning_rate": 7.952733199861879e-07,
+      "loss": 0.4674,
+      "step": 7660
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9890100950127168,
+      "learning_rate": 7.943314452019169e-07,
+      "loss": 0.4844,
+      "step": 7661
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.286570499214542,
+      "learning_rate": 7.933900803694578e-07,
+      "loss": 0.5009,
+      "step": 7662
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 50.32249738932716,
+      "learning_rate": 7.924492256029531e-07,
+      "loss": 0.4824,
+      "step": 7663
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.391452274625463,
+      "learning_rate": 7.915088810164856e-07,
+      "loss": 0.4945,
+      "step": 7664
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8305674300084913,
+      "learning_rate": 7.905690467240762e-07,
+      "loss": 0.4956,
+      "step": 7665
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1488149726796983,
+      "learning_rate": 7.896297228396821e-07,
+      "loss": 0.4846,
+      "step": 7666
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0277983978471994,
+      "learning_rate": 7.886909094771983e-07,
+      "loss": 0.4532,
+      "step": 7667
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.4378838424397378,
+      "learning_rate": 7.877526067504615e-07,
+      "loss": 0.4769,
+      "step": 7668
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3358975463763647,
+      "learning_rate": 7.868148147732413e-07,
+      "loss": 0.501,
+      "step": 7669
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.872717299875294,
+      "learning_rate": 7.858775336592506e-07,
+      "loss": 0.4565,
+      "step": 7670
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.43835931029033026,
+      "learning_rate": 7.849407635221378e-07,
+      "loss": 0.2759,
+      "step": 7671
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.9624528251346174,
+      "learning_rate": 7.840045044754879e-07,
+      "loss": 0.474,
+      "step": 7672
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.221091368171235,
+      "learning_rate": 7.830687566328266e-07,
+      "loss": 0.4254,
+      "step": 7673
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8911818509740588,
+      "learning_rate": 7.82133520107618e-07,
+      "loss": 0.4582,
+      "step": 7674
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.834318063826199,
+      "learning_rate": 7.811987950132599e-07,
+      "loss": 0.5249,
+      "step": 7675
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.5271473095422285,
+      "learning_rate": 7.802645814630921e-07,
+      "loss": 0.4683,
+      "step": 7676
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.541313777248314,
+      "learning_rate": 7.79330879570393e-07,
+      "loss": 0.4502,
+      "step": 7677
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.8007530201532282,
+      "learning_rate": 7.783976894483752e-07,
+      "loss": 0.445,
+      "step": 7678
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.029560653683497,
+      "learning_rate": 7.774650112101911e-07,
+      "loss": 0.4724,
+      "step": 7679
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4566937590723845,
+      "learning_rate": 7.765328449689313e-07,
+      "loss": 0.4917,
+      "step": 7680
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.022567386246196,
+      "learning_rate": 7.756011908376254e-07,
+      "loss": 0.4753,
+      "step": 7681
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.328867880229343,
+      "learning_rate": 7.746700489292381e-07,
+      "loss": 0.4462,
+      "step": 7682
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.695691066055379,
+      "learning_rate": 7.737394193566744e-07,
+      "loss": 0.4756,
+      "step": 7683
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.268452032026529,
+      "learning_rate": 7.72809302232777e-07,
+      "loss": 0.5262,
+      "step": 7684
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.910985738163341,
+      "learning_rate": 7.718796976703241e-07,
+      "loss": 0.4906,
+      "step": 7685
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.342068183506587,
+      "learning_rate": 7.709506057820343e-07,
+      "loss": 0.4877,
+      "step": 7686
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.1435849261705675,
+      "learning_rate": 7.700220266805647e-07,
+      "loss": 0.4859,
+      "step": 7687
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.547041348051068,
+      "learning_rate": 7.690939604785069e-07,
+      "loss": 0.4557,
+      "step": 7688
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.996384333017095,
+      "learning_rate": 7.681664072883915e-07,
+      "loss": 0.486,
+      "step": 7689
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.254646706305946,
+      "learning_rate": 7.672393672226902e-07,
+      "loss": 0.4724,
+      "step": 7690
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.571806820739956,
+      "learning_rate": 7.663128403938063e-07,
+      "loss": 0.5198,
+      "step": 7691
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.095556908133512,
+      "learning_rate": 7.653868269140868e-07,
+      "loss": 0.4674,
+      "step": 7692
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3790866568041066,
+      "learning_rate": 7.644613268958146e-07,
+      "loss": 0.279,
+      "step": 7693
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.385454160103183,
+      "learning_rate": 7.635363404512069e-07,
+      "loss": 0.4739,
+      "step": 7694
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.090558056547504,
+      "learning_rate": 7.626118676924238e-07,
+      "loss": 0.4923,
+      "step": 7695
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.21965301635262,
+      "learning_rate": 7.616879087315615e-07,
+      "loss": 0.5342,
+      "step": 7696
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.40557454293059203,
+      "learning_rate": 7.607644636806505e-07,
+      "loss": 0.2688,
+      "step": 7697
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4265543528969675,
+      "learning_rate": 7.59841532651664e-07,
+      "loss": 0.5096,
+      "step": 7698
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.243471795948674,
+      "learning_rate": 7.589191157565102e-07,
+      "loss": 0.5268,
+      "step": 7699
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.37222802826890367,
+      "learning_rate": 7.579972131070335e-07,
+      "loss": 0.269,
+      "step": 7700
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.396101044810851,
+      "learning_rate": 7.570758248150195e-07,
+      "loss": 0.4938,
+      "step": 7701
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.648754800181969,
+      "learning_rate": 7.561549509921901e-07,
+      "loss": 0.5045,
+      "step": 7702
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 16.89502195803844,
+      "learning_rate": 7.552345917502025e-07,
+      "loss": 0.5577,
+      "step": 7703
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.106929843072961,
+      "learning_rate": 7.543147472006545e-07,
+      "loss": 0.5068,
+      "step": 7704
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.8561863955655884,
+      "learning_rate": 7.533954174550817e-07,
+      "loss": 0.5029,
+      "step": 7705
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.3592560264181035,
+      "learning_rate": 7.524766026249535e-07,
+      "loss": 0.5006,
+      "step": 7706
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3781366484185993,
+      "learning_rate": 7.515583028216811e-07,
+      "loss": 0.4441,
+      "step": 7707
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.120597492813285,
+      "learning_rate": 7.506405181566123e-07,
+      "loss": 0.5158,
+      "step": 7708
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2568489975846178,
+      "learning_rate": 7.4972324874103e-07,
+      "loss": 0.4743,
+      "step": 7709
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3693235631024071,
+      "learning_rate": 7.488064946861556e-07,
+      "loss": 0.2763,
+      "step": 7710
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.918562393040958,
+      "learning_rate": 7.478902561031504e-07,
+      "loss": 0.4476,
+      "step": 7711
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.37962309090538104,
+      "learning_rate": 7.469745331031114e-07,
+      "loss": 0.2722,
+      "step": 7712
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.7003152689634335,
+      "learning_rate": 7.460593257970716e-07,
+      "loss": 0.4497,
+      "step": 7713
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.558459345274217,
+      "learning_rate": 7.451446342960044e-07,
+      "loss": 0.513,
+      "step": 7714
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.37291271553757854,
+      "learning_rate": 7.442304587108202e-07,
+      "loss": 0.2798,
+      "step": 7715
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.026582136630567,
+      "learning_rate": 7.433167991523632e-07,
+      "loss": 0.4624,
+      "step": 7716
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2661887356247767,
+      "learning_rate": 7.424036557314196e-07,
+      "loss": 0.4373,
+      "step": 7717
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.597419577122389,
+      "learning_rate": 7.414910285587118e-07,
+      "loss": 0.5445,
+      "step": 7718
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.420853397982494,
+      "learning_rate": 7.405789177448986e-07,
+      "loss": 0.4341,
+      "step": 7719
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2659981880603697,
+      "learning_rate": 7.396673234005741e-07,
+      "loss": 0.475,
+      "step": 7720
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.212913621760533,
+      "learning_rate": 7.387562456362762e-07,
+      "loss": 0.4611,
+      "step": 7721
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8351615504254775,
+      "learning_rate": 7.378456845624726e-07,
+      "loss": 0.4721,
+      "step": 7722
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.263750709478472,
+      "learning_rate": 7.369356402895739e-07,
+      "loss": 0.4982,
+      "step": 7723
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.089951989003569,
+      "learning_rate": 7.360261129279272e-07,
+      "loss": 0.4591,
+      "step": 7724
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.64304917441362,
+      "learning_rate": 7.351171025878129e-07,
+      "loss": 0.4201,
+      "step": 7725
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.160431122860172,
+      "learning_rate": 7.342086093794532e-07,
+      "loss": 0.5234,
+      "step": 7726
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5153024597803357,
+      "learning_rate": 7.333006334130078e-07,
+      "loss": 0.52,
+      "step": 7727
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.680224487722656,
+      "learning_rate": 7.323931747985679e-07,
+      "loss": 0.4835,
+      "step": 7728
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.675148875591901,
+      "learning_rate": 7.314862336461692e-07,
+      "loss": 0.4591,
+      "step": 7729
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.095179893822223,
+      "learning_rate": 7.305798100657823e-07,
+      "loss": 0.5166,
+      "step": 7730
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.983179420888415,
+      "learning_rate": 7.2967390416731e-07,
+      "loss": 0.5126,
+      "step": 7731
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.9261463514980477,
+      "learning_rate": 7.287685160605995e-07,
+      "loss": 0.4877,
+      "step": 7732
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.111508088464377,
+      "learning_rate": 7.278636458554322e-07,
+      "loss": 0.4763,
+      "step": 7733
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.623846788262555,
+      "learning_rate": 7.269592936615249e-07,
+      "loss": 0.4524,
+      "step": 7734
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3437795143252257,
+      "learning_rate": 7.260554595885349e-07,
+      "loss": 0.2785,
+      "step": 7735
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.490175661428506,
+      "learning_rate": 7.251521437460562e-07,
+      "loss": 0.4568,
+      "step": 7736
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3203924083090044,
+      "learning_rate": 7.242493462436162e-07,
+      "loss": 0.4667,
+      "step": 7737
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.565806582553676,
+      "learning_rate": 7.233470671906845e-07,
+      "loss": 0.4863,
+      "step": 7738
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4171360234594923,
+      "learning_rate": 7.224453066966652e-07,
+      "loss": 0.4928,
+      "step": 7739
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.259652049781979,
+      "learning_rate": 7.215440648708983e-07,
+      "loss": 0.4773,
+      "step": 7740
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.953745504145203,
+      "learning_rate": 7.206433418226649e-07,
+      "loss": 0.4732,
+      "step": 7741
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.335994251829295,
+      "learning_rate": 7.197431376611785e-07,
+      "loss": 0.4567,
+      "step": 7742
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4952172139324946,
+      "learning_rate": 7.18843452495594e-07,
+      "loss": 0.4562,
+      "step": 7743
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.7780279659582594,
+      "learning_rate": 7.179442864349989e-07,
+      "loss": 0.516,
+      "step": 7744
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.7324422854503907,
+      "learning_rate": 7.170456395884218e-07,
+      "loss": 0.525,
+      "step": 7745
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.252203934482893,
+      "learning_rate": 7.161475120648276e-07,
+      "loss": 0.4716,
+      "step": 7746
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.149704815113762,
+      "learning_rate": 7.152499039731153e-07,
+      "loss": 0.4663,
+      "step": 7747
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.315943739195209,
+      "learning_rate": 7.143528154221236e-07,
+      "loss": 0.4791,
+      "step": 7748
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5553274067653065,
+      "learning_rate": 7.13456246520629e-07,
+      "loss": 0.5122,
+      "step": 7749
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.14780895688247,
+      "learning_rate": 7.125601973773416e-07,
+      "loss": 0.5444,
+      "step": 7750
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.057051297543939,
+      "learning_rate": 7.116646681009121e-07,
+      "loss": 0.4973,
+      "step": 7751
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.974147161173928,
+      "learning_rate": 7.107696587999264e-07,
+      "loss": 0.5156,
+      "step": 7752
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.825602832367516,
+      "learning_rate": 7.098751695829048e-07,
+      "loss": 0.4601,
+      "step": 7753
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.295603980206641,
+      "learning_rate": 7.089812005583097e-07,
+      "loss": 0.4562,
+      "step": 7754
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8306027569463015,
+      "learning_rate": 7.080877518345386e-07,
+      "loss": 0.4702,
+      "step": 7755
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.2860742124269375,
+      "learning_rate": 7.071948235199227e-07,
+      "loss": 0.4893,
+      "step": 7756
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.044086044366995,
+      "learning_rate": 7.063024157227343e-07,
+      "loss": 0.4558,
+      "step": 7757
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4074892139393793,
+      "learning_rate": 7.054105285511814e-07,
+      "loss": 0.2836,
+      "step": 7758
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2294707381464933,
+      "learning_rate": 7.045191621134067e-07,
+      "loss": 0.4777,
+      "step": 7759
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.367565831057372,
+      "learning_rate": 7.036283165174923e-07,
+      "loss": 0.4775,
+      "step": 7760
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.54653620483379,
+      "learning_rate": 7.027379918714578e-07,
+      "loss": 0.4646,
+      "step": 7761
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.523152006635428,
+      "learning_rate": 7.018481882832562e-07,
+      "loss": 0.5481,
+      "step": 7762
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7557547423013085,
+      "learning_rate": 7.009589058607791e-07,
+      "loss": 0.4809,
+      "step": 7763
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.270037170873265,
+      "learning_rate": 7.000701447118563e-07,
+      "loss": 0.401,
+      "step": 7764
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.8210790543827153,
+      "learning_rate": 6.991819049442516e-07,
+      "loss": 0.4411,
+      "step": 7765
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.776478875433449,
+      "learning_rate": 6.982941866656684e-07,
+      "loss": 0.5272,
+      "step": 7766
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.168448460856312,
+      "learning_rate": 6.974069899837465e-07,
+      "loss": 0.4743,
+      "step": 7767
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.370988501913208,
+      "learning_rate": 6.96520315006059e-07,
+      "loss": 0.4894,
+      "step": 7768
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7577068755904492,
+      "learning_rate": 6.956341618401197e-07,
+      "loss": 0.5079,
+      "step": 7769
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.207441382902313,
+      "learning_rate": 6.947485305933787e-07,
+      "loss": 0.5253,
+      "step": 7770
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.692690056365726,
+      "learning_rate": 6.938634213732198e-07,
+      "loss": 0.4689,
+      "step": 7771
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2202569105457575,
+      "learning_rate": 6.929788342869675e-07,
+      "loss": 0.4662,
+      "step": 7772
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9315439422280494,
+      "learning_rate": 6.92094769441879e-07,
+      "loss": 0.5275,
+      "step": 7773
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3966497881171124,
+      "learning_rate": 6.912112269451526e-07,
+      "loss": 0.2733,
+      "step": 7774
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.832075429412881,
+      "learning_rate": 6.903282069039185e-07,
+      "loss": 0.5184,
+      "step": 7775
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4002928121150306,
+      "learning_rate": 6.894457094252471e-07,
+      "loss": 0.4466,
+      "step": 7776
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.541291539819691,
+      "learning_rate": 6.88563734616145e-07,
+      "loss": 0.4344,
+      "step": 7777
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3381474823071486,
+      "learning_rate": 6.876822825835528e-07,
+      "loss": 0.5066,
+      "step": 7778
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3835502262575003,
+      "learning_rate": 6.868013534343509e-07,
+      "loss": 0.4578,
+      "step": 7779
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.40561254410912684,
+      "learning_rate": 6.859209472753558e-07,
+      "loss": 0.2713,
+      "step": 7780
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.617750222947581,
+      "learning_rate": 6.850410642133176e-07,
+      "loss": 0.4644,
+      "step": 7781
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.008157912432583,
+      "learning_rate": 6.841617043549259e-07,
+      "loss": 0.4469,
+      "step": 7782
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.275996647147198,
+      "learning_rate": 6.832828678068093e-07,
+      "loss": 0.5233,
+      "step": 7783
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4895250731546574,
+      "learning_rate": 6.824045546755243e-07,
+      "loss": 0.502,
+      "step": 7784
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4052528156589977,
+      "learning_rate": 6.815267650675722e-07,
+      "loss": 0.2787,
+      "step": 7785
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.233835903268761,
+      "learning_rate": 6.806494990893892e-07,
+      "loss": 0.5001,
+      "step": 7786
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 14.633486807534124,
+      "learning_rate": 6.797727568473444e-07,
+      "loss": 0.4997,
+      "step": 7787
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.496133212780219,
+      "learning_rate": 6.788965384477464e-07,
+      "loss": 0.4745,
+      "step": 7788
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.1875616037832355,
+      "learning_rate": 6.780208439968416e-07,
+      "loss": 0.509,
+      "step": 7789
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.41920547324712176,
+      "learning_rate": 6.771456736008086e-07,
+      "loss": 0.2884,
+      "step": 7790
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.572682735938944,
+      "learning_rate": 6.762710273657658e-07,
+      "loss": 0.47,
+      "step": 7791
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.39252028809149414,
+      "learning_rate": 6.753969053977683e-07,
+      "loss": 0.2764,
+      "step": 7792
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7608662624975437,
+      "learning_rate": 6.745233078028041e-07,
+      "loss": 0.4935,
+      "step": 7793
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.91050617318769,
+      "learning_rate": 6.736502346868018e-07,
+      "loss": 0.5363,
+      "step": 7794
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4585826175657894,
+      "learning_rate": 6.72777686155624e-07,
+      "loss": 0.4628,
+      "step": 7795
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.3030084190036755,
+      "learning_rate": 6.719056623150688e-07,
+      "loss": 0.5073,
+      "step": 7796
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7587484240944535,
+      "learning_rate": 6.710341632708734e-07,
+      "loss": 0.5179,
+      "step": 7797
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.418199469875087,
+      "learning_rate": 6.701631891287108e-07,
+      "loss": 0.4777,
+      "step": 7798
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4464666293372077,
+      "learning_rate": 6.692927399941879e-07,
+      "loss": 0.4914,
+      "step": 7799
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.104593274423523,
+      "learning_rate": 6.6842281597285e-07,
+      "loss": 0.5435,
+      "step": 7800
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.815200842130119,
+      "learning_rate": 6.675534171701803e-07,
+      "loss": 0.4769,
+      "step": 7801
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.295672120226515,
+      "learning_rate": 6.666845436915942e-07,
+      "loss": 0.4579,
+      "step": 7802
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.549335489807461,
+      "learning_rate": 6.658161956424458e-07,
+      "loss": 0.4951,
+      "step": 7803
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.0165010107712895,
+      "learning_rate": 6.64948373128027e-07,
+      "loss": 0.5146,
+      "step": 7804
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.858915032876546,
+      "learning_rate": 6.640810762535637e-07,
+      "loss": 0.4815,
+      "step": 7805
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.216091456171551,
+      "learning_rate": 6.632143051242168e-07,
+      "loss": 0.5088,
+      "step": 7806
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8356306371037894,
+      "learning_rate": 6.623480598450865e-07,
+      "loss": 0.4688,
+      "step": 7807
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.639377746597069,
+      "learning_rate": 6.61482340521209e-07,
+      "loss": 0.443,
+      "step": 7808
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.353612358307862,
+      "learning_rate": 6.606171472575539e-07,
+      "loss": 0.4957,
+      "step": 7809
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.096778303997985,
+      "learning_rate": 6.597524801590299e-07,
+      "loss": 0.459,
+      "step": 7810
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.0104905151673504,
+      "learning_rate": 6.588883393304818e-07,
+      "loss": 0.5403,
+      "step": 7811
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.479157245331361,
+      "learning_rate": 6.580247248766875e-07,
+      "loss": 0.4548,
+      "step": 7812
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4828557732464827,
+      "learning_rate": 6.571616369023642e-07,
+      "loss": 0.4931,
+      "step": 7813
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 24.744865655531754,
+      "learning_rate": 6.562990755121662e-07,
+      "loss": 0.5179,
+      "step": 7814
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.503288651380729,
+      "learning_rate": 6.554370408106797e-07,
+      "loss": 0.4143,
+      "step": 7815
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.469537601459686,
+      "learning_rate": 6.545755329024295e-07,
+      "loss": 0.4437,
+      "step": 7816
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5704549028507526,
+      "learning_rate": 6.537145518918775e-07,
+      "loss": 0.5273,
+      "step": 7817
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.384154297658447,
+      "learning_rate": 6.528540978834191e-07,
+      "loss": 0.2744,
+      "step": 7818
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.989263333342746,
+      "learning_rate": 6.519941709813881e-07,
+      "loss": 0.4595,
+      "step": 7819
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.542776608187691,
+      "learning_rate": 6.511347712900545e-07,
+      "loss": 0.4933,
+      "step": 7820
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1237741090688442,
+      "learning_rate": 6.502758989136221e-07,
+      "loss": 0.4505,
+      "step": 7821
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.38129306546451747,
+      "learning_rate": 6.494175539562326e-07,
+      "loss": 0.2779,
+      "step": 7822
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5088468273471176,
+      "learning_rate": 6.485597365219648e-07,
+      "loss": 0.4942,
+      "step": 7823
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 46.01189428106445,
+      "learning_rate": 6.47702446714829e-07,
+      "loss": 0.4837,
+      "step": 7824
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.6876146095646694,
+      "learning_rate": 6.468456846387777e-07,
+      "loss": 0.4725,
+      "step": 7825
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.36639507077570205,
+      "learning_rate": 6.459894503976943e-07,
+      "loss": 0.2711,
+      "step": 7826
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.76371037372736,
+      "learning_rate": 6.451337440953997e-07,
+      "loss": 0.4983,
+      "step": 7827
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.40867051687980477,
+      "learning_rate": 6.442785658356526e-07,
+      "loss": 0.2808,
+      "step": 7828
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.530106031771651,
+      "learning_rate": 6.434239157221467e-07,
+      "loss": 0.4975,
+      "step": 7829
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9570558325477223,
+      "learning_rate": 6.425697938585091e-07,
+      "loss": 0.4329,
+      "step": 7830
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.105299992897392,
+      "learning_rate": 6.417162003483062e-07,
+      "loss": 0.5006,
+      "step": 7831
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9917772915449192,
+      "learning_rate": 6.408631352950406e-07,
+      "loss": 0.5236,
+      "step": 7832
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7817832642958042,
+      "learning_rate": 6.400105988021471e-07,
+      "loss": 0.4968,
+      "step": 7833
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.338968122393776,
+      "learning_rate": 6.391585909729998e-07,
+      "loss": 0.4981,
+      "step": 7834
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.6342057017361205,
+      "learning_rate": 6.38307111910908e-07,
+      "loss": 0.4636,
+      "step": 7835
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.215723474152873,
+      "learning_rate": 6.374561617191166e-07,
+      "loss": 0.5347,
+      "step": 7836
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.239914205591022,
+      "learning_rate": 6.366057405008041e-07,
+      "loss": 0.4576,
+      "step": 7837
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.665508837222767,
+      "learning_rate": 6.357558483590887e-07,
+      "loss": 0.4387,
+      "step": 7838
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.939191700845622,
+      "learning_rate": 6.349064853970233e-07,
+      "loss": 0.4789,
+      "step": 7839
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.7425351215235985,
+      "learning_rate": 6.340576517175939e-07,
+      "loss": 0.4741,
+      "step": 7840
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.384080173319251,
+      "learning_rate": 6.332093474237266e-07,
+      "loss": 0.5419,
+      "step": 7841
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3786852083244783,
+      "learning_rate": 6.323615726182813e-07,
+      "loss": 0.2797,
+      "step": 7842
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5722250208228337,
+      "learning_rate": 6.31514327404052e-07,
+      "loss": 0.4996,
+      "step": 7843
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.850210566338312,
+      "learning_rate": 6.306676118837707e-07,
+      "loss": 0.4798,
+      "step": 7844
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7577545502664904,
+      "learning_rate": 6.29821426160106e-07,
+      "loss": 0.5291,
+      "step": 7845
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.960441682274992,
+      "learning_rate": 6.289757703356597e-07,
+      "loss": 0.4748,
+      "step": 7846
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.11349832670986,
+      "learning_rate": 6.281306445129698e-07,
+      "loss": 0.5296,
+      "step": 7847
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.475315513089003,
+      "learning_rate": 6.27286048794512e-07,
+      "loss": 0.4677,
+      "step": 7848
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.610851375524315,
+      "learning_rate": 6.264419832826946e-07,
+      "loss": 0.4462,
+      "step": 7849
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.183634906504074,
+      "learning_rate": 6.255984480798649e-07,
+      "loss": 0.5298,
+      "step": 7850
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.304364087284888,
+      "learning_rate": 6.247554432883052e-07,
+      "loss": 0.4751,
+      "step": 7851
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.863319493124654,
+      "learning_rate": 6.239129690102308e-07,
+      "loss": 0.4729,
+      "step": 7852
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.75689384322336,
+      "learning_rate": 6.230710253477956e-07,
+      "loss": 0.4805,
+      "step": 7853
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8428552119950417,
+      "learning_rate": 6.222296124030891e-07,
+      "loss": 0.4252,
+      "step": 7854
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8237753339359415,
+      "learning_rate": 6.213887302781335e-07,
+      "loss": 0.4886,
+      "step": 7855
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.002968506450385,
+      "learning_rate": 6.205483790748901e-07,
+      "loss": 0.5136,
+      "step": 7856
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.63043661416841,
+      "learning_rate": 6.197085588952556e-07,
+      "loss": 0.5023,
+      "step": 7857
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7840123124833625,
+      "learning_rate": 6.188692698410576e-07,
+      "loss": 0.4504,
+      "step": 7858
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.506546448795916,
+      "learning_rate": 6.180305120140645e-07,
+      "loss": 0.4604,
+      "step": 7859
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.695428104060644,
+      "learning_rate": 6.171922855159801e-07,
+      "loss": 0.4689,
+      "step": 7860
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.193281131523582,
+      "learning_rate": 6.163545904484397e-07,
+      "loss": 0.5362,
+      "step": 7861
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.2114751721838095,
+      "learning_rate": 6.155174269130182e-07,
+      "loss": 0.4814,
+      "step": 7862
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.654323820894098,
+      "learning_rate": 6.146807950112249e-07,
+      "loss": 0.5547,
+      "step": 7863
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.106652134495157,
+      "learning_rate": 6.138446948445032e-07,
+      "loss": 0.5134,
+      "step": 7864
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.183980303694253,
+      "learning_rate": 6.130091265142329e-07,
+      "loss": 0.4975,
+      "step": 7865
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.82234478915749,
+      "learning_rate": 6.121740901217322e-07,
+      "loss": 0.4748,
+      "step": 7866
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.4962716341896325,
+      "learning_rate": 6.113395857682486e-07,
+      "loss": 0.504,
+      "step": 7867
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.819509677295161,
+      "learning_rate": 6.105056135549714e-07,
+      "loss": 0.4848,
+      "step": 7868
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.551876521369376,
+      "learning_rate": 6.096721735830202e-07,
+      "loss": 0.4722,
+      "step": 7869
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.378497844363434,
+      "learning_rate": 6.088392659534547e-07,
+      "loss": 0.525,
+      "step": 7870
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.914170563146186,
+      "learning_rate": 6.080068907672659e-07,
+      "loss": 0.5481,
+      "step": 7871
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.152901478300714,
+      "learning_rate": 6.071750481253835e-07,
+      "loss": 0.4608,
+      "step": 7872
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.58805713600808,
+      "learning_rate": 6.063437381286719e-07,
+      "loss": 0.4813,
+      "step": 7873
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4777671263820675,
+      "learning_rate": 6.055129608779276e-07,
+      "loss": 0.5001,
+      "step": 7874
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.403209522021868,
+      "learning_rate": 6.046827164738872e-07,
+      "loss": 0.4767,
+      "step": 7875
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.27704605468651,
+      "learning_rate": 6.03853005017222e-07,
+      "loss": 0.5035,
+      "step": 7876
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.712226793807638,
+      "learning_rate": 6.030238266085342e-07,
+      "loss": 0.5136,
+      "step": 7877
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4158248801677303,
+      "learning_rate": 6.021951813483673e-07,
+      "loss": 0.2794,
+      "step": 7878
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.953299787543567,
+      "learning_rate": 6.013670693371959e-07,
+      "loss": 0.4736,
+      "step": 7879
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5369296896773004,
+      "learning_rate": 6.005394906754309e-07,
+      "loss": 0.4484,
+      "step": 7880
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6561443482185374,
+      "learning_rate": 5.997124454634201e-07,
+      "loss": 0.4788,
+      "step": 7881
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3067141985195647,
+      "learning_rate": 5.988859338014457e-07,
+      "loss": 0.5208,
+      "step": 7882
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.997346922109789,
+      "learning_rate": 5.980599557897243e-07,
+      "loss": 0.4745,
+      "step": 7883
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.826525204421534,
+      "learning_rate": 5.972345115284094e-07,
+      "loss": 0.4485,
+      "step": 7884
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.476614060620013,
+      "learning_rate": 5.96409601117589e-07,
+      "loss": 0.4648,
+      "step": 7885
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.418184803661519,
+      "learning_rate": 5.955852246572852e-07,
+      "loss": 0.4555,
+      "step": 7886
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.239211455610939,
+      "learning_rate": 5.947613822474574e-07,
+      "loss": 0.4475,
+      "step": 7887
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.731533101893151,
+      "learning_rate": 5.939380739880002e-07,
+      "loss": 0.5206,
+      "step": 7888
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.993740236325758,
+      "learning_rate": 5.93115299978741e-07,
+      "loss": 0.4752,
+      "step": 7889
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.68532697872025,
+      "learning_rate": 5.922930603194438e-07,
+      "loss": 0.4901,
+      "step": 7890
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.430471044675298,
+      "learning_rate": 5.9147135510981e-07,
+      "loss": 0.4907,
+      "step": 7891
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5358268506372053,
+      "learning_rate": 5.906501844494716e-07,
+      "loss": 0.4385,
+      "step": 7892
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.36856006052356327,
+      "learning_rate": 5.898295484379996e-07,
+      "loss": 0.2664,
+      "step": 7893
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.750638145437625,
+      "learning_rate": 5.890094471749003e-07,
+      "loss": 0.4481,
+      "step": 7894
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.735348174232971,
+      "learning_rate": 5.881898807596115e-07,
+      "loss": 0.4847,
+      "step": 7895
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.831588345275058,
+      "learning_rate": 5.873708492915093e-07,
+      "loss": 0.4587,
+      "step": 7896
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.478599162517511,
+      "learning_rate": 5.865523528699051e-07,
+      "loss": 0.5124,
+      "step": 7897
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.609246830916243,
+      "learning_rate": 5.857343915940434e-07,
+      "loss": 0.5026,
+      "step": 7898
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3751373046346083,
+      "learning_rate": 5.849169655631054e-07,
+      "loss": 0.5023,
+      "step": 7899
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 14.796892426828851,
+      "learning_rate": 5.841000748762055e-07,
+      "loss": 0.4701,
+      "step": 7900
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0778762492865073,
+      "learning_rate": 5.832837196323965e-07,
+      "loss": 0.4583,
+      "step": 7901
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4142140867887636,
+      "learning_rate": 5.824678999306622e-07,
+      "loss": 0.5268,
+      "step": 7902
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1168053718497313,
+      "learning_rate": 5.816526158699243e-07,
+      "loss": 0.5339,
+      "step": 7903
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.366968313955123,
+      "learning_rate": 5.8083786754904e-07,
+      "loss": 0.4476,
+      "step": 7904
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.340927135127445,
+      "learning_rate": 5.800236550667992e-07,
+      "loss": 0.4869,
+      "step": 7905
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8516879005287126,
+      "learning_rate": 5.792099785219274e-07,
+      "loss": 0.4301,
+      "step": 7906
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.534805224104772,
+      "learning_rate": 5.783968380130877e-07,
+      "loss": 0.4915,
+      "step": 7907
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.874361033732114,
+      "learning_rate": 5.775842336388742e-07,
+      "loss": 0.4753,
+      "step": 7908
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.545985749266089,
+      "learning_rate": 5.767721654978187e-07,
+      "loss": 0.4608,
+      "step": 7909
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7025446715363888,
+      "learning_rate": 5.759606336883894e-07,
+      "loss": 0.5179,
+      "step": 7910
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9796541814585096,
+      "learning_rate": 5.751496383089833e-07,
+      "loss": 0.4566,
+      "step": 7911
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1329992989867623,
+      "learning_rate": 5.743391794579379e-07,
+      "loss": 0.4885,
+      "step": 7912
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.381446276519732,
+      "learning_rate": 5.735292572335261e-07,
+      "loss": 0.4898,
+      "step": 7913
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1347780583620595,
+      "learning_rate": 5.727198717339511e-07,
+      "loss": 0.5054,
+      "step": 7914
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5736109427467286,
+      "learning_rate": 5.719110230573543e-07,
+      "loss": 0.4843,
+      "step": 7915
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5524387872845193,
+      "learning_rate": 5.711027113018136e-07,
+      "loss": 0.4596,
+      "step": 7916
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.656212797606447,
+      "learning_rate": 5.702949365653365e-07,
+      "loss": 0.4551,
+      "step": 7917
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.37864884665131415,
+      "learning_rate": 5.694876989458698e-07,
+      "loss": 0.2705,
+      "step": 7918
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.751277695889469,
+      "learning_rate": 5.68680998541295e-07,
+      "loss": 0.4917,
+      "step": 7919
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3322315559561884,
+      "learning_rate": 5.678748354494251e-07,
+      "loss": 0.4202,
+      "step": 7920
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.692815623118857,
+      "learning_rate": 5.670692097680124e-07,
+      "loss": 0.4898,
+      "step": 7921
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4013235014667316,
+      "learning_rate": 5.662641215947406e-07,
+      "loss": 0.2644,
+      "step": 7922
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.76456453079996,
+      "learning_rate": 5.654595710272287e-07,
+      "loss": 0.4564,
+      "step": 7923
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.771836644580547,
+      "learning_rate": 5.646555581630319e-07,
+      "loss": 0.4734,
+      "step": 7924
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3717993137900409,
+      "learning_rate": 5.638520830996402e-07,
+      "loss": 0.2585,
+      "step": 7925
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9490937640210353,
+      "learning_rate": 5.630491459344766e-07,
+      "loss": 0.4434,
+      "step": 7926
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0682995806470243,
+      "learning_rate": 5.622467467649007e-07,
+      "loss": 0.4652,
+      "step": 7927
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.282466566876941,
+      "learning_rate": 5.614448856882066e-07,
+      "loss": 0.4474,
+      "step": 7928
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 17.1251821701484,
+      "learning_rate": 5.60643562801621e-07,
+      "loss": 0.4664,
+      "step": 7929
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.669251984005528,
+      "learning_rate": 5.598427782023086e-07,
+      "loss": 0.4814,
+      "step": 7930
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6624321858212356,
+      "learning_rate": 5.590425319873683e-07,
+      "loss": 0.5093,
+      "step": 7931
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.863949861253507,
+      "learning_rate": 5.582428242538307e-07,
+      "loss": 0.4737,
+      "step": 7932
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7612101143689083,
+      "learning_rate": 5.574436550986628e-07,
+      "loss": 0.512,
+      "step": 7933
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4219426272277884,
+      "learning_rate": 5.566450246187672e-07,
+      "loss": 0.494,
+      "step": 7934
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.9083218511226177,
+      "learning_rate": 5.558469329109823e-07,
+      "loss": 0.4664,
+      "step": 7935
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.571697697123912,
+      "learning_rate": 5.550493800720774e-07,
+      "loss": 0.5001,
+      "step": 7936
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3570074342786795,
+      "learning_rate": 5.542523661987586e-07,
+      "loss": 0.2667,
+      "step": 7937
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.37957377504362105,
+      "learning_rate": 5.534558913876686e-07,
+      "loss": 0.2642,
+      "step": 7938
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.39449866620372054,
+      "learning_rate": 5.526599557353807e-07,
+      "loss": 0.2664,
+      "step": 7939
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6433543364839704,
+      "learning_rate": 5.518645593384047e-07,
+      "loss": 0.4896,
+      "step": 7940
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6259507622156573,
+      "learning_rate": 5.510697022931876e-07,
+      "loss": 0.5462,
+      "step": 7941
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.283413070303731,
+      "learning_rate": 5.502753846961067e-07,
+      "loss": 0.4755,
+      "step": 7942
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4022212033641749,
+      "learning_rate": 5.494816066434749e-07,
+      "loss": 0.2993,
+      "step": 7943
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.274194292852197,
+      "learning_rate": 5.486883682315425e-07,
+      "loss": 0.5375,
+      "step": 7944
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.865167875558897,
+      "learning_rate": 5.478956695564902e-07,
+      "loss": 0.4894,
+      "step": 7945
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8850479587762816,
+      "learning_rate": 5.471035107144374e-07,
+      "loss": 0.4732,
+      "step": 7946
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 27.95360386246314,
+      "learning_rate": 5.46311891801436e-07,
+      "loss": 0.49,
+      "step": 7947
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.78295914995517,
+      "learning_rate": 5.455208129134704e-07,
+      "loss": 0.4442,
+      "step": 7948
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9577343414516943,
+      "learning_rate": 5.447302741464638e-07,
+      "loss": 0.4914,
+      "step": 7949
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.7213103879718252,
+      "learning_rate": 5.439402755962719e-07,
+      "loss": 0.5129,
+      "step": 7950
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1918460395446226,
+      "learning_rate": 5.43150817358683e-07,
+      "loss": 0.4587,
+      "step": 7951
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5692218827072146,
+      "learning_rate": 5.423618995294239e-07,
+      "loss": 0.4563,
+      "step": 7952
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.94817391665543,
+      "learning_rate": 5.415735222041518e-07,
+      "loss": 0.439,
+      "step": 7953
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.298636709201372,
+      "learning_rate": 5.407856854784599e-07,
+      "loss": 0.5183,
+      "step": 7954
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8265388317180276,
+      "learning_rate": 5.399983894478772e-07,
+      "loss": 0.4606,
+      "step": 7955
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.3137310393351385,
+      "learning_rate": 5.392116342078662e-07,
+      "loss": 0.4362,
+      "step": 7956
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.716546876852122,
+      "learning_rate": 5.384254198538224e-07,
+      "loss": 0.4765,
+      "step": 7957
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9306816342519793,
+      "learning_rate": 5.376397464810779e-07,
+      "loss": 0.4512,
+      "step": 7958
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.129924508410385,
+      "learning_rate": 5.36854614184899e-07,
+      "loss": 0.4937,
+      "step": 7959
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.803701579900726,
+      "learning_rate": 5.360700230604842e-07,
+      "loss": 0.5225,
+      "step": 7960
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.799122099431319,
+      "learning_rate": 5.352859732029686e-07,
+      "loss": 0.4726,
+      "step": 7961
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.3535771871082245,
+      "learning_rate": 5.345024647074221e-07,
+      "loss": 0.4792,
+      "step": 7962
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.301194218063768,
+      "learning_rate": 5.337194976688464e-07,
+      "loss": 0.5176,
+      "step": 7963
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.141590833949314,
+      "learning_rate": 5.329370721821781e-07,
+      "loss": 0.478,
+      "step": 7964
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3339667490427733,
+      "learning_rate": 5.321551883422904e-07,
+      "loss": 0.4877,
+      "step": 7965
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.545850895291743,
+      "learning_rate": 5.3137384624399e-07,
+      "loss": 0.4783,
+      "step": 7966
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.162390090641765,
+      "learning_rate": 5.305930459820158e-07,
+      "loss": 0.4371,
+      "step": 7967
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8264841664266935,
+      "learning_rate": 5.298127876510429e-07,
+      "loss": 0.4744,
+      "step": 7968
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.2991658569170585,
+      "learning_rate": 5.29033071345681e-07,
+      "loss": 0.5198,
+      "step": 7969
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.191771927636866,
+      "learning_rate": 5.282538971604729e-07,
+      "loss": 0.4951,
+      "step": 7970
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6402800484373348,
+      "learning_rate": 5.274752651898957e-07,
+      "loss": 0.5129,
+      "step": 7971
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.654101118896311,
+      "learning_rate": 5.266971755283628e-07,
+      "loss": 0.4941,
+      "step": 7972
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.104494001733222,
+      "learning_rate": 5.259196282702178e-07,
+      "loss": 0.473,
+      "step": 7973
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.620236282535001,
+      "learning_rate": 5.25142623509744e-07,
+      "loss": 0.4665,
+      "step": 7974
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.437404142534207,
+      "learning_rate": 5.243661613411544e-07,
+      "loss": 0.4673,
+      "step": 7975
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.187234598447495,
+      "learning_rate": 5.235902418585958e-07,
+      "loss": 0.471,
+      "step": 7976
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7187975328909624,
+      "learning_rate": 5.228148651561537e-07,
+      "loss": 0.4334,
+      "step": 7977
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.587448310535323,
+      "learning_rate": 5.22040031327845e-07,
+      "loss": 0.4776,
+      "step": 7978
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.818319528644026,
+      "learning_rate": 5.212657404676192e-07,
+      "loss": 0.4893,
+      "step": 7979
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.433264674526942,
+      "learning_rate": 5.204919926693635e-07,
+      "loss": 0.4718,
+      "step": 7980
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.638378752179209,
+      "learning_rate": 5.197187880268973e-07,
+      "loss": 0.5165,
+      "step": 7981
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.614720423072915,
+      "learning_rate": 5.189461266339735e-07,
+      "loss": 0.48,
+      "step": 7982
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3824746603258258,
+      "learning_rate": 5.181740085842795e-07,
+      "loss": 0.2715,
+      "step": 7983
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2981888462536006,
+      "learning_rate": 5.174024339714401e-07,
+      "loss": 0.5221,
+      "step": 7984
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 13.737776650254071,
+      "learning_rate": 5.166314028890091e-07,
+      "loss": 0.5198,
+      "step": 7985
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9793066297972683,
+      "learning_rate": 5.158609154304761e-07,
+      "loss": 0.4112,
+      "step": 7986
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.6935341646961035,
+      "learning_rate": 5.150909716892671e-07,
+      "loss": 0.451,
+      "step": 7987
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.820320109466844,
+      "learning_rate": 5.143215717587385e-07,
+      "loss": 0.5214,
+      "step": 7988
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.922078646019367,
+      "learning_rate": 5.13552715732184e-07,
+      "loss": 0.4764,
+      "step": 7989
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.899019687536353,
+      "learning_rate": 5.127844037028313e-07,
+      "loss": 0.4793,
+      "step": 7990
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.47762604610422,
+      "learning_rate": 5.120166357638378e-07,
+      "loss": 0.4876,
+      "step": 7991
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.281978760266442,
+      "learning_rate": 5.112494120083e-07,
+      "loss": 0.4943,
+      "step": 7992
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.461606540269323,
+      "learning_rate": 5.104827325292466e-07,
+      "loss": 0.4719,
+      "step": 7993
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.440124954838877,
+      "learning_rate": 5.097165974196389e-07,
+      "loss": 0.4796,
+      "step": 7994
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4095893886430451,
+      "learning_rate": 5.089510067723752e-07,
+      "loss": 0.2837,
+      "step": 7995
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8669434211246707,
+      "learning_rate": 5.081859606802841e-07,
+      "loss": 0.4646,
+      "step": 7996
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4461938302802784,
+      "learning_rate": 5.074214592361315e-07,
+      "loss": 0.5037,
+      "step": 7997
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.17081013428014,
+      "learning_rate": 5.066575025326148e-07,
+      "loss": 0.5661,
+      "step": 7998
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.105965064248626,
+      "learning_rate": 5.05894090662366e-07,
+      "loss": 0.444,
+      "step": 7999
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3904989316073198,
+      "learning_rate": 5.051312237179539e-07,
+      "loss": 0.2819,
+      "step": 8000
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.385445418417046,
+      "learning_rate": 5.043689017918757e-07,
+      "loss": 0.4891,
+      "step": 8001
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.831519572024611,
+      "learning_rate": 5.036071249765673e-07,
+      "loss": 0.4736,
+      "step": 8002
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.46754322628502,
+      "learning_rate": 5.028458933643976e-07,
+      "loss": 0.5367,
+      "step": 8003
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.719397200350784,
+      "learning_rate": 5.020852070476657e-07,
+      "loss": 0.4539,
+      "step": 8004
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7882176167767465,
+      "learning_rate": 5.013250661186103e-07,
+      "loss": 0.4259,
+      "step": 8005
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3920714765788396,
+      "learning_rate": 5.005654706694002e-07,
+      "loss": 0.4808,
+      "step": 8006
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.712763143550038,
+      "learning_rate": 4.998064207921377e-07,
+      "loss": 0.4491,
+      "step": 8007
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.5902472788938122,
+      "learning_rate": 4.990479165788609e-07,
+      "loss": 0.512,
+      "step": 8008
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0733222419069346,
+      "learning_rate": 4.982899581215428e-07,
+      "loss": 0.4846,
+      "step": 8009
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9699176709912787,
+      "learning_rate": 4.975325455120855e-07,
+      "loss": 0.5036,
+      "step": 8010
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.132683292526016,
+      "learning_rate": 4.9677567884233e-07,
+      "loss": 0.4591,
+      "step": 8011
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0234473131692705,
+      "learning_rate": 4.96019358204049e-07,
+      "loss": 0.4377,
+      "step": 8012
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.926487163047114,
+      "learning_rate": 4.952635836889475e-07,
+      "loss": 0.4561,
+      "step": 8013
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.5260462299042707,
+      "learning_rate": 4.94508355388667e-07,
+      "loss": 0.4613,
+      "step": 8014
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.689294286128493,
+      "learning_rate": 4.937536733947817e-07,
+      "loss": 0.5071,
+      "step": 8015
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5491980027755528,
+      "learning_rate": 4.92999537798799e-07,
+      "loss": 0.4395,
+      "step": 8016
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.483884592774077,
+      "learning_rate": 4.922459486921594e-07,
+      "loss": 0.4348,
+      "step": 8017
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.443117022172283,
+      "learning_rate": 4.914929061662399e-07,
+      "loss": 0.4507,
+      "step": 8018
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3832757924357032,
+      "learning_rate": 4.907404103123481e-07,
+      "loss": 0.2761,
+      "step": 8019
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.999457667197051,
+      "learning_rate": 4.89988461221727e-07,
+      "loss": 0.4315,
+      "step": 8020
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.196152715420318,
+      "learning_rate": 4.89237058985555e-07,
+      "loss": 0.4997,
+      "step": 8021
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4836937292940546,
+      "learning_rate": 4.88486203694939e-07,
+      "loss": 0.4855,
+      "step": 8022
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.611156866969053,
+      "learning_rate": 4.877358954409245e-07,
+      "loss": 0.4732,
+      "step": 8023
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.079631635267597,
+      "learning_rate": 4.869861343144894e-07,
+      "loss": 0.4322,
+      "step": 8024
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.414155860764693,
+      "learning_rate": 4.862369204065437e-07,
+      "loss": 0.4764,
+      "step": 8025
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.998433302767149,
+      "learning_rate": 4.854882538079331e-07,
+      "loss": 0.4872,
+      "step": 8026
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7078534222344133,
+      "learning_rate": 4.847401346094361e-07,
+      "loss": 0.4575,
+      "step": 8027
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4227601801292944,
+      "learning_rate": 4.839925629017638e-07,
+      "loss": 0.2929,
+      "step": 8028
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.960160225718414,
+      "learning_rate": 4.832455387755619e-07,
+      "loss": 0.5254,
+      "step": 8029
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.3707380615641975,
+      "learning_rate": 4.824990623214099e-07,
+      "loss": 0.5163,
+      "step": 8030
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.297312080699592,
+      "learning_rate": 4.817531336298221e-07,
+      "loss": 0.4502,
+      "step": 8031
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 15.244062306858655,
+      "learning_rate": 4.810077527912422e-07,
+      "loss": 0.4703,
+      "step": 8032
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.83575012131532,
+      "learning_rate": 4.802629198960517e-07,
+      "loss": 0.5181,
+      "step": 8033
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9918538120715468,
+      "learning_rate": 4.795186350345648e-07,
+      "loss": 0.4747,
+      "step": 8034
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3678291288289631,
+      "learning_rate": 4.787748982970275e-07,
+      "loss": 0.2524,
+      "step": 8035
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 15.994846932620323,
+      "learning_rate": 4.780317097736203e-07,
+      "loss": 0.5207,
+      "step": 8036
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3797553610895816,
+      "learning_rate": 4.772890695544602e-07,
+      "loss": 0.4813,
+      "step": 8037
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7043696179865226,
+      "learning_rate": 4.7654697772959057e-07,
+      "loss": 0.525,
+      "step": 8038
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9120519189279452,
+      "learning_rate": 4.758054343889945e-07,
+      "loss": 0.4973,
+      "step": 8039
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.906659605071496,
+      "learning_rate": 4.750644396225873e-07,
+      "loss": 0.4825,
+      "step": 8040
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.936318244820105,
+      "learning_rate": 4.743239935202165e-07,
+      "loss": 0.4814,
+      "step": 8041
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.628378817087108,
+      "learning_rate": 4.735840961716631e-07,
+      "loss": 0.4535,
+      "step": 8042
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.823188430204171,
+      "learning_rate": 4.728447476666442e-07,
+      "loss": 0.5114,
+      "step": 8043
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7954262476083684,
+      "learning_rate": 4.7210594809480647e-07,
+      "loss": 0.513,
+      "step": 8044
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.556138109256956,
+      "learning_rate": 4.713676975457321e-07,
+      "loss": 0.4674,
+      "step": 8045
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.224489444423062,
+      "learning_rate": 4.7062999610893835e-07,
+      "loss": 0.4691,
+      "step": 8046
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2891948383049048,
+      "learning_rate": 4.698928438738715e-07,
+      "loss": 0.4931,
+      "step": 8047
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.055734041557022,
+      "learning_rate": 4.691562409299161e-07,
+      "loss": 0.483,
+      "step": 8048
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.41041379702256015,
+      "learning_rate": 4.6842018736638685e-07,
+      "loss": 0.2728,
+      "step": 8049
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3439019046288645,
+      "learning_rate": 4.6768468327253127e-07,
+      "loss": 0.4707,
+      "step": 8050
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.9638848125972173,
+      "learning_rate": 4.66949728737533e-07,
+      "loss": 0.4696,
+      "step": 8051
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.368153886637305,
+      "learning_rate": 4.6621532385050904e-07,
+      "loss": 0.5086,
+      "step": 8052
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2331400876359058,
+      "learning_rate": 4.654814687005066e-07,
+      "loss": 0.5225,
+      "step": 8053
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5840768203490954,
+      "learning_rate": 4.6474816337650883e-07,
+      "loss": 0.4709,
+      "step": 8054
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 23.666944082035485,
+      "learning_rate": 4.640154079674325e-07,
+      "loss": 0.4348,
+      "step": 8055
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.3064390901270535,
+      "learning_rate": 4.6328320256212534e-07,
+      "loss": 0.4713,
+      "step": 8056
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.214709580192385,
+      "learning_rate": 4.625515472493697e-07,
+      "loss": 0.4542,
+      "step": 8057
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.717480654519902,
+      "learning_rate": 4.618204421178835e-07,
+      "loss": 0.4499,
+      "step": 8058
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1979521333287266,
+      "learning_rate": 4.610898872563135e-07,
+      "loss": 0.5181,
+      "step": 8059
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.251383538509404,
+      "learning_rate": 4.603598827532424e-07,
+      "loss": 0.4711,
+      "step": 8060
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.569453367587647,
+      "learning_rate": 4.596304286971853e-07,
+      "loss": 0.4963,
+      "step": 8061
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6591684402192506,
+      "learning_rate": 4.589015251765933e-07,
+      "loss": 0.477,
+      "step": 8062
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2901149422370644,
+      "learning_rate": 4.58173172279846e-07,
+      "loss": 0.4805,
+      "step": 8063
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.787332243103131,
+      "learning_rate": 4.574453700952591e-07,
+      "loss": 0.4439,
+      "step": 8064
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.601537643143989,
+      "learning_rate": 4.567181187110831e-07,
+      "loss": 0.4614,
+      "step": 8065
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.041400542344912,
+      "learning_rate": 4.5599141821549673e-07,
+      "loss": 0.5371,
+      "step": 8066
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.209565477723137,
+      "learning_rate": 4.552652686966169e-07,
+      "loss": 0.5313,
+      "step": 8067
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.397363832023855,
+      "learning_rate": 4.545396702424926e-07,
+      "loss": 0.533,
+      "step": 8068
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5838585596220573,
+      "learning_rate": 4.5381462294110336e-07,
+      "loss": 0.4759,
+      "step": 8069
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7064601984410497,
+      "learning_rate": 4.5309012688036335e-07,
+      "loss": 0.469,
+      "step": 8070
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.1117880778762625,
+      "learning_rate": 4.5236618214812233e-07,
+      "loss": 0.4252,
+      "step": 8071
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.411375279715031,
+      "learning_rate": 4.516427888321584e-07,
+      "loss": 0.4834,
+      "step": 8072
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4747246288778553,
+      "learning_rate": 4.5091994702018695e-07,
+      "loss": 0.4614,
+      "step": 8073
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.258332255688346,
+      "learning_rate": 4.5019765679985614e-07,
+      "loss": 0.5063,
+      "step": 8074
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.106304632281978,
+      "learning_rate": 4.4947591825874383e-07,
+      "loss": 0.5316,
+      "step": 8075
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.454832750914892,
+      "learning_rate": 4.487547314843643e-07,
+      "loss": 0.4997,
+      "step": 8076
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.001269601752422,
+      "learning_rate": 4.480340965641655e-07,
+      "loss": 0.4761,
+      "step": 8077
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3153952365706485,
+      "learning_rate": 4.473140135855247e-07,
+      "loss": 0.4963,
+      "step": 8078
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.847672378852405,
+      "learning_rate": 4.4659448263575544e-07,
+      "loss": 0.4947,
+      "step": 8079
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6418802797988477,
+      "learning_rate": 4.458755038021029e-07,
+      "loss": 0.4736,
+      "step": 8080
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.712208548586137,
+      "learning_rate": 4.451570771717467e-07,
+      "loss": 0.4458,
+      "step": 8081
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.7773567018392304,
+      "learning_rate": 4.444392028317968e-07,
+      "loss": 0.4883,
+      "step": 8082
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6532691494351477,
+      "learning_rate": 4.437218808693e-07,
+      "loss": 0.5084,
+      "step": 8083
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.143502283966991,
+      "learning_rate": 4.4300511137123193e-07,
+      "loss": 0.4878,
+      "step": 8084
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.42057000894965607,
+      "learning_rate": 4.4228889442450463e-07,
+      "loss": 0.2871,
+      "step": 8085
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.402327141189362,
+      "learning_rate": 4.4157323011596254e-07,
+      "loss": 0.481,
+      "step": 8086
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.6908030964180805,
+      "learning_rate": 4.408581185323807e-07,
+      "loss": 0.4383,
+      "step": 8087
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0436816832025664,
+      "learning_rate": 4.4014355976046974e-07,
+      "loss": 0.4603,
+      "step": 8088
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.303371785350145,
+      "learning_rate": 4.394295538868731e-07,
+      "loss": 0.4825,
+      "step": 8089
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.255326655575552,
+      "learning_rate": 4.3871610099816606e-07,
+      "loss": 0.4767,
+      "step": 8090
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.837857963849661,
+      "learning_rate": 4.3800320118085604e-07,
+      "loss": 0.5001,
+      "step": 8091
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.260133515039705,
+      "learning_rate": 4.37290854521385e-07,
+      "loss": 0.4602,
+      "step": 8092
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.480967290386756,
+      "learning_rate": 4.3657906110612935e-07,
+      "loss": 0.4932,
+      "step": 8093
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.956413444925701,
+      "learning_rate": 4.358678210213935e-07,
+      "loss": 0.4786,
+      "step": 8094
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.031743399198921,
+      "learning_rate": 4.3515713435342e-07,
+      "loss": 0.4772,
+      "step": 8095
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5879705329958997,
+      "learning_rate": 4.3444700118838167e-07,
+      "loss": 0.4523,
+      "step": 8096
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.685340742079363,
+      "learning_rate": 4.337374216123841e-07,
+      "loss": 0.4743,
+      "step": 8097
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.244695296633857,
+      "learning_rate": 4.330283957114656e-07,
+      "loss": 0.4978,
+      "step": 8098
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.1648715457099375,
+      "learning_rate": 4.3231992357160026e-07,
+      "loss": 0.4169,
+      "step": 8099
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.970932414449429,
+      "learning_rate": 4.316120052786904e-07,
+      "loss": 0.4221,
+      "step": 8100
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7162128023680396,
+      "learning_rate": 4.3090464091857576e-07,
+      "loss": 0.4876,
+      "step": 8101
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.213610369849614,
+      "learning_rate": 4.3019783057702493e-07,
+      "loss": 0.4916,
+      "step": 8102
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.564105456483165,
+      "learning_rate": 4.29491574339741e-07,
+      "loss": 0.5206,
+      "step": 8103
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3509253291567695,
+      "learning_rate": 4.287858722923605e-07,
+      "loss": 0.4414,
+      "step": 8104
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.39851816884525654,
+      "learning_rate": 4.2808072452045377e-07,
+      "loss": 0.2707,
+      "step": 8105
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5513131544854804,
+      "learning_rate": 4.2737613110951924e-07,
+      "loss": 0.4286,
+      "step": 8106
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4762172111984877,
+      "learning_rate": 4.266720921449935e-07,
+      "loss": 0.482,
+      "step": 8107
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.296523627222795,
+      "learning_rate": 4.2596860771224435e-07,
+      "loss": 0.4369,
+      "step": 8108
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.992949968722491,
+      "learning_rate": 4.252656778965697e-07,
+      "loss": 0.4765,
+      "step": 8109
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3617063567500916,
+      "learning_rate": 4.2456330278320245e-07,
+      "loss": 0.4712,
+      "step": 8110
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.514434703930507,
+      "learning_rate": 4.238614824573101e-07,
+      "loss": 0.4592,
+      "step": 8111
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.030245212294359,
+      "learning_rate": 4.2316021700398945e-07,
+      "loss": 0.445,
+      "step": 8112
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.44713407359492,
+      "learning_rate": 4.2245950650827037e-07,
+      "loss": 0.5173,
+      "step": 8113
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.388114954438636,
+      "learning_rate": 4.2175935105511767e-07,
+      "loss": 0.2757,
+      "step": 8114
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.346362325402245,
+      "learning_rate": 4.2105975072942674e-07,
+      "loss": 0.4972,
+      "step": 8115
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3999582643858362,
+      "learning_rate": 4.2036070561602695e-07,
+      "loss": 0.2947,
+      "step": 8116
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.721948092221754,
+      "learning_rate": 4.1966221579968103e-07,
+      "loss": 0.5337,
+      "step": 8117
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.632783213773352,
+      "learning_rate": 4.1896428136508126e-07,
+      "loss": 0.5197,
+      "step": 8118
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.810109740532846,
+      "learning_rate": 4.182669023968561e-07,
+      "loss": 0.4952,
+      "step": 8119
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.183702049248703,
+      "learning_rate": 4.1757007897956503e-07,
+      "loss": 0.5317,
+      "step": 8120
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.37601603767014724,
+      "learning_rate": 4.1687381119769887e-07,
+      "loss": 0.263,
+      "step": 8121
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3892381181271491,
+      "learning_rate": 4.161780991356845e-07,
+      "loss": 0.2645,
+      "step": 8122
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8969118134442917,
+      "learning_rate": 4.154829428778778e-07,
+      "loss": 0.4961,
+      "step": 8123
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.4884824115564665,
+      "learning_rate": 4.1478834250857027e-07,
+      "loss": 0.4743,
+      "step": 8124
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.9927411183283077,
+      "learning_rate": 4.140942981119833e-07,
+      "loss": 0.4947,
+      "step": 8125
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8496426460523008,
+      "learning_rate": 4.134008097722725e-07,
+      "loss": 0.4147,
+      "step": 8126
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.949570483260543,
+      "learning_rate": 4.127078775735266e-07,
+      "loss": 0.4197,
+      "step": 8127
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.234643755575205,
+      "learning_rate": 4.1201550159976454e-07,
+      "loss": 0.5306,
+      "step": 8128
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.922425977541475,
+      "learning_rate": 4.113236819349403e-07,
+      "loss": 0.4971,
+      "step": 8129
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.539096022690103,
+      "learning_rate": 4.1063241866293955e-07,
+      "loss": 0.4711,
+      "step": 8130
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.322059463414458,
+      "learning_rate": 4.099417118675797e-07,
+      "loss": 0.4894,
+      "step": 8131
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.1288213465781585,
+      "learning_rate": 4.092515616326126e-07,
+      "loss": 0.5222,
+      "step": 8132
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3762439692183041,
+      "learning_rate": 4.0856196804171965e-07,
+      "loss": 0.2761,
+      "step": 8133
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.408606298458517,
+      "learning_rate": 4.078729311785168e-07,
+      "loss": 0.2833,
+      "step": 8134
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.545995544371098,
+      "learning_rate": 4.071844511265527e-07,
+      "loss": 0.4746,
+      "step": 8135
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3952701279755564,
+      "learning_rate": 4.064965279693084e-07,
+      "loss": 0.5054,
+      "step": 8136
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.79713871091845,
+      "learning_rate": 4.058091617901949e-07,
+      "loss": 0.4684,
+      "step": 8137
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.795451640021837,
+      "learning_rate": 4.0512235267256007e-07,
+      "loss": 0.5304,
+      "step": 8138
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.1943427303595335,
+      "learning_rate": 4.044361006996811e-07,
+      "loss": 0.4771,
+      "step": 8139
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.914311243849739,
+      "learning_rate": 4.0375040595476753e-07,
+      "loss": 0.5253,
+      "step": 8140
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4004749733992086,
+      "learning_rate": 4.0306526852096296e-07,
+      "loss": 0.2637,
+      "step": 8141
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8165570487439573,
+      "learning_rate": 4.023806884813436e-07,
+      "loss": 0.455,
+      "step": 8142
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3353821244299926,
+      "learning_rate": 4.016966659189159e-07,
+      "loss": 0.4956,
+      "step": 8143
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5156315618759137,
+      "learning_rate": 4.0101320091661954e-07,
+      "loss": 0.5106,
+      "step": 8144
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.819303328638301,
+      "learning_rate": 4.003302935573289e-07,
+      "loss": 0.551,
+      "step": 8145
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.010074483286804,
+      "learning_rate": 3.996479439238465e-07,
+      "loss": 0.4935,
+      "step": 8146
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.209259113513927,
+      "learning_rate": 3.989661520989102e-07,
+      "loss": 0.4308,
+      "step": 8147
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.069543173701213,
+      "learning_rate": 3.9828491816519153e-07,
+      "loss": 0.4672,
+      "step": 8148
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.36912272678185853,
+      "learning_rate": 3.9760424220529e-07,
+      "loss": 0.2722,
+      "step": 8149
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2536333492582843,
+      "learning_rate": 3.969241243017413e-07,
+      "loss": 0.5012,
+      "step": 8150
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.079034911143597,
+      "learning_rate": 3.962445645370122e-07,
+      "loss": 0.4825,
+      "step": 8151
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.358145836675113,
+      "learning_rate": 3.955655629935007e-07,
+      "loss": 0.3868,
+      "step": 8152
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.766517061382993,
+      "learning_rate": 3.9488711975353866e-07,
+      "loss": 0.4938,
+      "step": 8153
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.608873947892509,
+      "learning_rate": 3.9420923489939034e-07,
+      "loss": 0.4623,
+      "step": 8154
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9721432912495,
+      "learning_rate": 3.935319085132505e-07,
+      "loss": 0.4816,
+      "step": 8155
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.91670559928195,
+      "learning_rate": 3.928551406772468e-07,
+      "loss": 0.5018,
+      "step": 8156
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.955714727506886,
+      "learning_rate": 3.921789314734409e-07,
+      "loss": 0.4515,
+      "step": 8157
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.324921465545479,
+      "learning_rate": 3.9150328098382593e-07,
+      "loss": 0.441,
+      "step": 8158
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.458314986751764,
+      "learning_rate": 3.908281892903254e-07,
+      "loss": 0.5207,
+      "step": 8159
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 15.521431952461521,
+      "learning_rate": 3.9015365647479655e-07,
+      "loss": 0.4667,
+      "step": 8160
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 21.455553024356266,
+      "learning_rate": 3.8947968261903055e-07,
+      "loss": 0.4644,
+      "step": 8161
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3925933670022403,
+      "learning_rate": 3.8880626780474717e-07,
+      "loss": 0.4701,
+      "step": 8162
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.485905513099537,
+      "learning_rate": 3.88133412113601e-07,
+      "loss": 0.4163,
+      "step": 8163
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.046114366940229,
+      "learning_rate": 3.874611156271801e-07,
+      "loss": 0.4709,
+      "step": 8164
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.773947508125634,
+      "learning_rate": 3.8678937842699884e-07,
+      "loss": 0.4943,
+      "step": 8165
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.216864505550966,
+      "learning_rate": 3.8611820059450913e-07,
+      "loss": 0.4975,
+      "step": 8166
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4009691822516971,
+      "learning_rate": 3.8544758221109534e-07,
+      "loss": 0.2883,
+      "step": 8167
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.434369437260064,
+      "learning_rate": 3.8477752335807026e-07,
+      "loss": 0.4624,
+      "step": 8168
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.062402720678291,
+      "learning_rate": 3.841080241166811e-07,
+      "loss": 0.5299,
+      "step": 8169
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.156421822011142,
+      "learning_rate": 3.834390845681091e-07,
+      "loss": 0.472,
+      "step": 8170
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.247793330578803,
+      "learning_rate": 3.82770704793462e-07,
+      "loss": 0.4993,
+      "step": 8171
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.42613848446558,
+      "learning_rate": 3.821028848737857e-07,
+      "loss": 0.4538,
+      "step": 8172
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.286103385736838,
+      "learning_rate": 3.8143562489005525e-07,
+      "loss": 0.4874,
+      "step": 8173
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7050728071264913,
+      "learning_rate": 3.807689249231772e-07,
+      "loss": 0.475,
+      "step": 8174
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.4626240204660155,
+      "learning_rate": 3.801027850539929e-07,
+      "loss": 0.4311,
+      "step": 8175
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.092156557278627,
+      "learning_rate": 3.794372053632722e-07,
+      "loss": 0.455,
+      "step": 8176
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.850652454779329,
+      "learning_rate": 3.787721859317206e-07,
+      "loss": 0.4642,
+      "step": 8177
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.374201743689213,
+      "learning_rate": 3.7810772683997245e-07,
+      "loss": 0.286,
+      "step": 8178
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.06466823662857,
+      "learning_rate": 3.774438281685977e-07,
+      "loss": 0.4843,
+      "step": 8179
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7949685961619624,
+      "learning_rate": 3.767804899980937e-07,
+      "loss": 0.5149,
+      "step": 8180
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.9288641600294225,
+      "learning_rate": 3.761177124088944e-07,
+      "loss": 0.4973,
+      "step": 8181
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.057346379357827,
+      "learning_rate": 3.754554954813644e-07,
+      "loss": 0.4913,
+      "step": 8182
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.040819834105292,
+      "learning_rate": 3.747938392957973e-07,
+      "loss": 0.4462,
+      "step": 8183
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.289212823502151,
+      "learning_rate": 3.7413274393242327e-07,
+      "loss": 0.4788,
+      "step": 8184
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.62822506260676,
+      "learning_rate": 3.7347220947140273e-07,
+      "loss": 0.5057,
+      "step": 8185
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.444659970850248,
+      "learning_rate": 3.7281223599282656e-07,
+      "loss": 0.5053,
+      "step": 8186
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.0056249553977334,
+      "learning_rate": 3.7215282357671803e-07,
+      "loss": 0.4812,
+      "step": 8187
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.499033831941137,
+      "learning_rate": 3.7149397230303473e-07,
+      "loss": 0.5109,
+      "step": 8188
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.57994436571759,
+      "learning_rate": 3.7083568225166456e-07,
+      "loss": 0.436,
+      "step": 8189
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.58663107359806,
+      "learning_rate": 3.701779535024269e-07,
+      "loss": 0.5004,
+      "step": 8190
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9375256071544826,
+      "learning_rate": 3.6952078613507304e-07,
+      "loss": 0.484,
+      "step": 8191
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.201994503890201,
+      "learning_rate": 3.688641802292892e-07,
+      "loss": 0.4626,
+      "step": 8192
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.515078247387615,
+      "learning_rate": 3.682081358646883e-07,
+      "loss": 0.491,
+      "step": 8193
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.855915816797372,
+      "learning_rate": 3.67552653120819e-07,
+      "loss": 0.4493,
+      "step": 8194
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.637248638855968,
+      "learning_rate": 3.668977320771616e-07,
+      "loss": 0.5,
+      "step": 8195
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.8470625570908625,
+      "learning_rate": 3.6624337281312704e-07,
+      "loss": 0.5112,
+      "step": 8196
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.331525158144491,
+      "learning_rate": 3.6558957540805794e-07,
+      "loss": 0.4535,
+      "step": 8197
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.8315057561768455,
+      "learning_rate": 3.649363399412309e-07,
+      "loss": 0.5412,
+      "step": 8198
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.158985883969981,
+      "learning_rate": 3.642836664918509e-07,
+      "loss": 0.4317,
+      "step": 8199
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8392046378767852,
+      "learning_rate": 3.6363155513905847e-07,
+      "loss": 0.5039,
+      "step": 8200
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 19.640029766864913,
+      "learning_rate": 3.629800059619248e-07,
+      "loss": 0.4796,
+      "step": 8201
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5711543245602857,
+      "learning_rate": 3.6232901903945075e-07,
+      "loss": 0.4555,
+      "step": 8202
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.820586307522603,
+      "learning_rate": 3.616785944505713e-07,
+      "loss": 0.504,
+      "step": 8203
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4039181440272702,
+      "learning_rate": 3.61028732274154e-07,
+      "loss": 0.2536,
+      "step": 8204
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0457595433711466,
+      "learning_rate": 3.603794325889953e-07,
+      "loss": 0.5304,
+      "step": 8205
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.362168428165214,
+      "learning_rate": 3.5973069547382556e-07,
+      "loss": 0.4537,
+      "step": 8206
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.33855233928916,
+      "learning_rate": 3.5908252100730735e-07,
+      "loss": 0.4938,
+      "step": 8207
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.074118741821459,
+      "learning_rate": 3.584349092680328e-07,
+      "loss": 0.4588,
+      "step": 8208
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.570600683219978,
+      "learning_rate": 3.577878603345269e-07,
+      "loss": 0.527,
+      "step": 8209
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.497611964765638,
+      "learning_rate": 3.5714137428524754e-07,
+      "loss": 0.4923,
+      "step": 8210
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.230285833030696,
+      "learning_rate": 3.5649545119858244e-07,
+      "loss": 0.4846,
+      "step": 8211
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.098257326861237,
+      "learning_rate": 3.558500911528523e-07,
+      "loss": 0.5025,
+      "step": 8212
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3760349324331078,
+      "learning_rate": 3.552052942263101e-07,
+      "loss": 0.279,
+      "step": 8213
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.69801973430072,
+      "learning_rate": 3.5456106049713835e-07,
+      "loss": 0.4269,
+      "step": 8214
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3864322424161376,
+      "learning_rate": 3.5391739004345336e-07,
+      "loss": 0.2826,
+      "step": 8215
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3389253539950987,
+      "learning_rate": 3.5327428294330333e-07,
+      "loss": 0.5049,
+      "step": 8216
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.36719766467201115,
+      "learning_rate": 3.5263173927466585e-07,
+      "loss": 0.2472,
+      "step": 8217
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 16.884116177080777,
+      "learning_rate": 3.5198975911545143e-07,
+      "loss": 0.4719,
+      "step": 8218
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.032802192872389,
+      "learning_rate": 3.5134834254350213e-07,
+      "loss": 0.4805,
+      "step": 8219
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.220896243435433,
+      "learning_rate": 3.5070748963659417e-07,
+      "loss": 0.4623,
+      "step": 8220
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.705529787751602,
+      "learning_rate": 3.5006720047243037e-07,
+      "loss": 0.5127,
+      "step": 8221
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.093637184913449,
+      "learning_rate": 3.494274751286497e-07,
+      "loss": 0.4733,
+      "step": 8222
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9736280653363596,
+      "learning_rate": 3.487883136828213e-07,
+      "loss": 0.4819,
+      "step": 8223
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.177490378996627,
+      "learning_rate": 3.481497162124442e-07,
+      "loss": 0.4671,
+      "step": 8224
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.692825773217719,
+      "learning_rate": 3.4751168279495094e-07,
+      "loss": 0.5197,
+      "step": 8225
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.017121167380698,
+      "learning_rate": 3.468742135077069e-07,
+      "loss": 0.4992,
+      "step": 8226
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.83628865549746,
+      "learning_rate": 3.4623730842800573e-07,
+      "loss": 0.4422,
+      "step": 8227
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8871727839814625,
+      "learning_rate": 3.4560096763307516e-07,
+      "loss": 0.505,
+      "step": 8228
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.755310351184211,
+      "learning_rate": 3.4496519120007344e-07,
+      "loss": 0.4573,
+      "step": 8229
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5182343503200006,
+      "learning_rate": 3.4432997920609e-07,
+      "loss": 0.4342,
+      "step": 8230
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.40113438090613485,
+      "learning_rate": 3.4369533172814715e-07,
+      "loss": 0.2528,
+      "step": 8231
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.781210516939347,
+      "learning_rate": 3.4306124884319894e-07,
+      "loss": 0.5188,
+      "step": 8232
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.154666700825303,
+      "learning_rate": 3.424277306281282e-07,
+      "loss": 0.4525,
+      "step": 8233
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.7469010328285455,
+      "learning_rate": 3.417947771597524e-07,
+      "loss": 0.485,
+      "step": 8234
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.027193393423928,
+      "learning_rate": 3.411623885148202e-07,
+      "loss": 0.4223,
+      "step": 8235
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.377821706295917,
+      "learning_rate": 3.4053056477000856e-07,
+      "loss": 0.4724,
+      "step": 8236
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.57135724320625,
+      "learning_rate": 3.3989930600192953e-07,
+      "loss": 0.5248,
+      "step": 8237
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.324885539369289,
+      "learning_rate": 3.3926861228712637e-07,
+      "loss": 0.5362,
+      "step": 8238
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.927134207872489,
+      "learning_rate": 3.386384837020723e-07,
+      "loss": 0.4655,
+      "step": 8239
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.706452686974156,
+      "learning_rate": 3.3800892032317123e-07,
+      "loss": 0.4656,
+      "step": 8240
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5212232542647555,
+      "learning_rate": 3.373799222267615e-07,
+      "loss": 0.5247,
+      "step": 8241
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.697117998277992,
+      "learning_rate": 3.3675148948910996e-07,
+      "loss": 0.4738,
+      "step": 8242
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.40483799473873483,
+      "learning_rate": 3.3612362218641727e-07,
+      "loss": 0.2739,
+      "step": 8243
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1304323738491764,
+      "learning_rate": 3.354963203948147e-07,
+      "loss": 0.479,
+      "step": 8244
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2768022896116142,
+      "learning_rate": 3.348695841903637e-07,
+      "loss": 0.4901,
+      "step": 8245
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.9051461248911186,
+      "learning_rate": 3.3424341364905853e-07,
+      "loss": 0.4575,
+      "step": 8246
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.632414391746175,
+      "learning_rate": 3.3361780884682615e-07,
+      "loss": 0.4329,
+      "step": 8247
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.668651316533972,
+      "learning_rate": 3.3299276985952034e-07,
+      "loss": 0.4808,
+      "step": 8248
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6639184687612376,
+      "learning_rate": 3.323682967629321e-07,
+      "loss": 0.4987,
+      "step": 8249
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.851627579091931,
+      "learning_rate": 3.317443896327788e-07,
+      "loss": 0.459,
+      "step": 8250
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.574514074032049,
+      "learning_rate": 3.311210485447125e-07,
+      "loss": 0.5537,
+      "step": 8251
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.41478682769839575,
+      "learning_rate": 3.3049827357431506e-07,
+      "loss": 0.2581,
+      "step": 8252
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5264287627562267,
+      "learning_rate": 3.2987606479709946e-07,
+      "loss": 0.426,
+      "step": 8253
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3798857200373718,
+      "learning_rate": 3.2925442228851246e-07,
+      "loss": 0.4908,
+      "step": 8254
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.942183039412618,
+      "learning_rate": 3.286333461239288e-07,
+      "loss": 0.4723,
+      "step": 8255
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.040208875159907,
+      "learning_rate": 3.2801283637865613e-07,
+      "loss": 0.4393,
+      "step": 8256
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.044756339198047,
+      "learning_rate": 3.2739289312793467e-07,
+      "loss": 0.5159,
+      "step": 8257
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4134599275736033,
+      "learning_rate": 3.2677351644693323e-07,
+      "loss": 0.4836,
+      "step": 8258
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4629162849535073,
+      "learning_rate": 3.261547064107551e-07,
+      "loss": 0.464,
+      "step": 8259
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.201258518540733,
+      "learning_rate": 3.2553646309443133e-07,
+      "loss": 0.4694,
+      "step": 8260
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0598667691455588,
+      "learning_rate": 3.2491878657292643e-07,
+      "loss": 0.4584,
+      "step": 8261
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.0908070394130736,
+      "learning_rate": 3.24301676921136e-07,
+      "loss": 0.4897,
+      "step": 8262
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.530958136928353,
+      "learning_rate": 3.2368513421388745e-07,
+      "loss": 0.4607,
+      "step": 8263
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.1253455128968435,
+      "learning_rate": 3.230691585259371e-07,
+      "loss": 0.4964,
+      "step": 8264
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 48.71385965925272,
+      "learning_rate": 3.2245374993197577e-07,
+      "loss": 0.474,
+      "step": 8265
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3537564787594425,
+      "learning_rate": 3.218389085066237e-07,
+      "loss": 0.5009,
+      "step": 8266
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.372777936481912,
+      "learning_rate": 3.212246343244313e-07,
+      "loss": 0.4829,
+      "step": 8267
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.610429120897996,
+      "learning_rate": 3.2061092745988175e-07,
+      "loss": 0.4866,
+      "step": 8268
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.389337682868464,
+      "learning_rate": 3.1999778798739057e-07,
+      "loss": 0.2807,
+      "step": 8269
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.958397365574749,
+      "learning_rate": 3.1938521598130213e-07,
+      "loss": 0.436,
+      "step": 8270
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.13411380427378,
+      "learning_rate": 3.18773211515892e-07,
+      "loss": 0.4323,
+      "step": 8271
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.7702921498031525,
+      "learning_rate": 3.181617746653687e-07,
+      "loss": 0.5109,
+      "step": 8272
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.279587845746013,
+      "learning_rate": 3.1755090550387167e-07,
+      "loss": 0.4829,
+      "step": 8273
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.287624767209549,
+      "learning_rate": 3.169406041054696e-07,
+      "loss": 0.4699,
+      "step": 8274
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2090494290359772,
+      "learning_rate": 3.163308705441648e-07,
+      "loss": 0.4887,
+      "step": 8275
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.949832445016032,
+      "learning_rate": 3.1572170489388823e-07,
+      "loss": 0.4858,
+      "step": 8276
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.196922814630735,
+      "learning_rate": 3.151131072285041e-07,
+      "loss": 0.455,
+      "step": 8277
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.895455329508671,
+      "learning_rate": 3.145050776218078e-07,
+      "loss": 0.5175,
+      "step": 8278
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.900885386325597,
+      "learning_rate": 3.1389761614752376e-07,
+      "loss": 0.5094,
+      "step": 8279
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.948465823919465,
+      "learning_rate": 3.132907228793086e-07,
+      "loss": 0.4885,
+      "step": 8280
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.535831639286963,
+      "learning_rate": 3.126843978907518e-07,
+      "loss": 0.4947,
+      "step": 8281
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.144705146650608,
+      "learning_rate": 3.1207864125537167e-07,
+      "loss": 0.5087,
+      "step": 8282
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.723352302983914,
+      "learning_rate": 3.1147345304661736e-07,
+      "loss": 0.5135,
+      "step": 8283
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.010763453858462,
+      "learning_rate": 3.108688333378701e-07,
+      "loss": 0.4579,
+      "step": 8284
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.753043902647446,
+      "learning_rate": 3.1026478220244406e-07,
+      "loss": 0.4574,
+      "step": 8285
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.123291027576058,
+      "learning_rate": 3.0966129971358003e-07,
+      "loss": 0.4338,
+      "step": 8286
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9759511078761287,
+      "learning_rate": 3.090583859444535e-07,
+      "loss": 0.4691,
+      "step": 8287
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.2588616287777485,
+      "learning_rate": 3.084560409681703e-07,
+      "loss": 0.4696,
+      "step": 8288
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9208209444450954,
+      "learning_rate": 3.078542648577659e-07,
+      "loss": 0.4861,
+      "step": 8289
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5799867683007203,
+      "learning_rate": 3.0725305768620815e-07,
+      "loss": 0.5312,
+      "step": 8290
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.843241927070779,
+      "learning_rate": 3.0665241952639704e-07,
+      "loss": 0.4609,
+      "step": 8291
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.1547716421151195,
+      "learning_rate": 3.0605235045115867e-07,
+      "loss": 0.4482,
+      "step": 8292
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.319635499160741,
+      "learning_rate": 3.0545285053325604e-07,
+      "loss": 0.4336,
+      "step": 8293
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1391181201873315,
+      "learning_rate": 3.048539198453798e-07,
+      "loss": 0.5444,
+      "step": 8294
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.216695326734268,
+      "learning_rate": 3.04255558460152e-07,
+      "loss": 0.4314,
+      "step": 8295
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.23630668976904,
+      "learning_rate": 3.036577664501267e-07,
+      "loss": 0.4893,
+      "step": 8296
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.280385358437894,
+      "learning_rate": 3.030605438877882e-07,
+      "loss": 0.5184,
+      "step": 8297
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0617712550302336,
+      "learning_rate": 3.024638908455513e-07,
+      "loss": 0.4651,
+      "step": 8298
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.008298383381389,
+      "learning_rate": 3.0186780739576204e-07,
+      "loss": 0.4552,
+      "step": 8299
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.951772897438051,
+      "learning_rate": 3.012722936106993e-07,
+      "loss": 0.4898,
+      "step": 8300
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3131255016292083,
+      "learning_rate": 3.0067734956256867e-07,
+      "loss": 0.4806,
+      "step": 8301
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3426879238146756,
+      "learning_rate": 3.000829753235118e-07,
+      "loss": 0.4856,
+      "step": 8302
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.829826300650556,
+      "learning_rate": 2.9948917096559617e-07,
+      "loss": 0.5069,
+      "step": 8303
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.060338674340327,
+      "learning_rate": 2.9889593656082405e-07,
+      "loss": 0.546,
+      "step": 8304
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.054670027322201,
+      "learning_rate": 2.983032721811263e-07,
+      "loss": 0.5184,
+      "step": 8305
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.575654835884226,
+      "learning_rate": 2.9771117789836713e-07,
+      "loss": 0.4816,
+      "step": 8306
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.083848401692234,
+      "learning_rate": 2.971196537843379e-07,
+      "loss": 0.4899,
+      "step": 8307
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.252420145429833,
+      "learning_rate": 2.9652869991076414e-07,
+      "loss": 0.5118,
+      "step": 8308
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.125604134316206,
+      "learning_rate": 2.9593831634930127e-07,
+      "loss": 0.5038,
+      "step": 8309
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.488664093399569,
+      "learning_rate": 2.953485031715342e-07,
+      "loss": 0.4801,
+      "step": 8310
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.0214669098322595,
+      "learning_rate": 2.947592604489807e-07,
+      "loss": 0.5068,
+      "step": 8311
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.807832636607853,
+      "learning_rate": 2.9417058825308874e-07,
+      "loss": 0.4748,
+      "step": 8312
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.3015663772295145,
+      "learning_rate": 2.935824866552367e-07,
+      "loss": 0.4667,
+      "step": 8313
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.39558092283643675,
+      "learning_rate": 2.929949557267331e-07,
+      "loss": 0.2766,
+      "step": 8314
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6029448838988847,
+      "learning_rate": 2.9240799553881813e-07,
+      "loss": 0.4865,
+      "step": 8315
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.764350147388825,
+      "learning_rate": 2.918216061626644e-07,
+      "loss": 0.5308,
+      "step": 8316
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.600527221640901,
+      "learning_rate": 2.912357876693711e-07,
+      "loss": 0.5235,
+      "step": 8317
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.63261502514191,
+      "learning_rate": 2.9065054012997305e-07,
+      "loss": 0.4685,
+      "step": 8318
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3681283826279133,
+      "learning_rate": 2.900658636154324e-07,
+      "loss": 0.2637,
+      "step": 8319
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.136590893785943,
+      "learning_rate": 2.894817581966436e-07,
+      "loss": 0.4737,
+      "step": 8320
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.965719746610781,
+      "learning_rate": 2.8889822394443047e-07,
+      "loss": 0.4761,
+      "step": 8321
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3755854176921987,
+      "learning_rate": 2.883152609295503e-07,
+      "loss": 0.4812,
+      "step": 8322
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.727397288368218,
+      "learning_rate": 2.8773286922268885e-07,
+      "loss": 0.5015,
+      "step": 8323
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.023904351049628,
+      "learning_rate": 2.871510488944612e-07,
+      "loss": 0.502,
+      "step": 8324
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.76674995179598,
+      "learning_rate": 2.8656980001541766e-07,
+      "loss": 0.4892,
+      "step": 8325
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.38106546823637166,
+      "learning_rate": 2.859891226560346e-07,
+      "loss": 0.2647,
+      "step": 8326
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.121651715554508,
+      "learning_rate": 2.854090168867224e-07,
+      "loss": 0.4437,
+      "step": 8327
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7934743997328564,
+      "learning_rate": 2.848294827778214e-07,
+      "loss": 0.5062,
+      "step": 8328
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.561131611918471,
+      "learning_rate": 2.8425052039959986e-07,
+      "loss": 0.4718,
+      "step": 8329
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.287367018914808,
+      "learning_rate": 2.836721298222611e-07,
+      "loss": 0.4657,
+      "step": 8330
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.23777225374048,
+      "learning_rate": 2.830943111159368e-07,
+      "loss": 0.4679,
+      "step": 8331
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.3166103349759535,
+      "learning_rate": 2.8251706435068806e-07,
+      "loss": 0.4835,
+      "step": 8332
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.917597962553783,
+      "learning_rate": 2.8194038959650895e-07,
+      "loss": 0.5035,
+      "step": 8333
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4258057443482222,
+      "learning_rate": 2.813642869233235e-07,
+      "loss": 0.2736,
+      "step": 8334
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0739862415406485,
+      "learning_rate": 2.8078875640098646e-07,
+      "loss": 0.4895,
+      "step": 8335
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.435292113369288,
+      "learning_rate": 2.8021379809928137e-07,
+      "loss": 0.4848,
+      "step": 8336
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.320877486532518,
+      "learning_rate": 2.796394120879259e-07,
+      "loss": 0.494,
+      "step": 8337
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.337775507487395,
+      "learning_rate": 2.790655984365642e-07,
+      "loss": 0.4567,
+      "step": 8338
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.396518401160575,
+      "learning_rate": 2.7849235721477407e-07,
+      "loss": 0.5167,
+      "step": 8339
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.218245254275644,
+      "learning_rate": 2.779196884920643e-07,
+      "loss": 0.4287,
+      "step": 8340
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3231156890424844,
+      "learning_rate": 2.7734759233787046e-07,
+      "loss": 0.4356,
+      "step": 8341
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9953003168111,
+      "learning_rate": 2.7677606882156314e-07,
+      "loss": 0.5091,
+      "step": 8342
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.916428241102304,
+      "learning_rate": 2.762051180124414e-07,
+      "loss": 0.5032,
+      "step": 8343
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.505023532848287,
+      "learning_rate": 2.7563473997973434e-07,
+      "loss": 0.4773,
+      "step": 8344
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.233391658297153,
+      "learning_rate": 2.750649347926021e-07,
+      "loss": 0.4434,
+      "step": 8345
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.38653255132440906,
+      "learning_rate": 2.7449570252013557e-07,
+      "loss": 0.2657,
+      "step": 8346
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.963904391144444,
+      "learning_rate": 2.7392704323135676e-07,
+      "loss": 0.5169,
+      "step": 8347
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.770398600815006,
+      "learning_rate": 2.733589569952172e-07,
+      "loss": 0.4477,
+      "step": 8348
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.352662773688024,
+      "learning_rate": 2.7279144388059895e-07,
+      "loss": 0.4369,
+      "step": 8349
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.983376399929851,
+      "learning_rate": 2.7222450395631593e-07,
+      "loss": 0.4516,
+      "step": 8350
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.615932996931086,
+      "learning_rate": 2.7165813729111035e-07,
+      "loss": 0.484,
+      "step": 8351
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.222441838570564,
+      "learning_rate": 2.7109234395365666e-07,
+      "loss": 0.4649,
+      "step": 8352
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2115421616037225,
+      "learning_rate": 2.705271240125601e-07,
+      "loss": 0.4786,
+      "step": 8353
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6083094736802357,
+      "learning_rate": 2.699624775363541e-07,
+      "loss": 0.4653,
+      "step": 8354
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.40236108172726126,
+      "learning_rate": 2.6939840459350497e-07,
+      "loss": 0.2741,
+      "step": 8355
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.045162066974496,
+      "learning_rate": 2.688349052524081e-07,
+      "loss": 0.4658,
+      "step": 8356
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.351190789397109,
+      "learning_rate": 2.6827197958138927e-07,
+      "loss": 0.4408,
+      "step": 8357
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.7438769868946755,
+      "learning_rate": 2.6770962764870625e-07,
+      "loss": 0.4869,
+      "step": 8358
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.619873779978012,
+      "learning_rate": 2.6714784952254546e-07,
+      "loss": 0.4457,
+      "step": 8359
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.202561796442311,
+      "learning_rate": 2.665866452710242e-07,
+      "loss": 0.4786,
+      "step": 8360
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.068368899165916,
+      "learning_rate": 2.6602601496219073e-07,
+      "loss": 0.4682,
+      "step": 8361
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4132198726345276,
+      "learning_rate": 2.65465958664024e-07,
+      "loss": 0.2773,
+      "step": 8362
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.39120384349095466,
+      "learning_rate": 2.6490647644443145e-07,
+      "loss": 0.2661,
+      "step": 8363
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.135416609931957,
+      "learning_rate": 2.643475683712532e-07,
+      "loss": 0.5348,
+      "step": 8364
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 24.2279025105478,
+      "learning_rate": 2.6378923451225893e-07,
+      "loss": 0.4684,
+      "step": 8365
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.073901721404017,
+      "learning_rate": 2.632314749351483e-07,
+      "loss": 0.5214,
+      "step": 8366
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.308960059177089,
+      "learning_rate": 2.6267428970755124e-07,
+      "loss": 0.5239,
+      "step": 8367
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.696484252871334,
+      "learning_rate": 2.6211767889702813e-07,
+      "loss": 0.4505,
+      "step": 8368
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.140959437462924,
+      "learning_rate": 2.61561642571071e-07,
+      "loss": 0.5055,
+      "step": 8369
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.8186571615152225,
+      "learning_rate": 2.610061807971004e-07,
+      "loss": 0.4713,
+      "step": 8370
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.4088052039468204,
+      "learning_rate": 2.6045129364246857e-07,
+      "loss": 0.4321,
+      "step": 8371
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.366643378953596,
+      "learning_rate": 2.598969811744562e-07,
+      "loss": 0.2802,
+      "step": 8372
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.374064094749522,
+      "learning_rate": 2.593432434602766e-07,
+      "loss": 0.5078,
+      "step": 8373
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.706016860415992,
+      "learning_rate": 2.5879008056707286e-07,
+      "loss": 0.5104,
+      "step": 8374
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.582646952190097,
+      "learning_rate": 2.5823749256191686e-07,
+      "loss": 0.4611,
+      "step": 8375
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.613471705179284,
+      "learning_rate": 2.5768547951181276e-07,
+      "loss": 0.4841,
+      "step": 8376
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.536552603108902,
+      "learning_rate": 2.5713404148369316e-07,
+      "loss": 0.5111,
+      "step": 8377
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.389807407976676,
+      "learning_rate": 2.5658317854442283e-07,
+      "loss": 0.5304,
+      "step": 8378
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.629363772188186,
+      "learning_rate": 2.56032890760794e-07,
+      "loss": 0.5079,
+      "step": 8379
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.446516044641298,
+      "learning_rate": 2.5548317819953205e-07,
+      "loss": 0.4902,
+      "step": 8380
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.605274604052965,
+      "learning_rate": 2.5493404092729267e-07,
+      "loss": 0.5112,
+      "step": 8381
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.4167673931876723,
+      "learning_rate": 2.543854790106587e-07,
+      "loss": 0.4635,
+      "step": 8382
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.228359811433122,
+      "learning_rate": 2.538374925161463e-07,
+      "loss": 0.4679,
+      "step": 8383
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.39535394775238974,
+      "learning_rate": 2.5329008151020076e-07,
+      "loss": 0.2822,
+      "step": 8384
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6269818702578482,
+      "learning_rate": 2.5274324605919664e-07,
+      "loss": 0.549,
+      "step": 8385
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.810147443777816,
+      "learning_rate": 2.5219698622944044e-07,
+      "loss": 0.474,
+      "step": 8386
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.631829244921488,
+      "learning_rate": 2.5165130208716913e-07,
+      "loss": 0.5594,
+      "step": 8387
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.5942738803820977,
+      "learning_rate": 2.5110619369854594e-07,
+      "loss": 0.4305,
+      "step": 8388
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.228801811552604,
+      "learning_rate": 2.505616611296685e-07,
+      "loss": 0.4686,
+      "step": 8389
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1887946195361954,
+      "learning_rate": 2.500177044465646e-07,
+      "loss": 0.443,
+      "step": 8390
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.898618822832122,
+      "learning_rate": 2.494743237151892e-07,
+      "loss": 0.4445,
+      "step": 8391
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.518307333212721,
+      "learning_rate": 2.489315190014291e-07,
+      "loss": 0.4582,
+      "step": 8392
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.232436151688414,
+      "learning_rate": 2.483892903711027e-07,
+      "loss": 0.4597,
+      "step": 8393
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.309607534548218,
+      "learning_rate": 2.4784763788995527e-07,
+      "loss": 0.5008,
+      "step": 8394
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2588793099446822,
+      "learning_rate": 2.473065616236647e-07,
+      "loss": 0.4391,
+      "step": 8395
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.5739808870816323,
+      "learning_rate": 2.467660616378398e-07,
+      "loss": 0.5088,
+      "step": 8396
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.836530747359646,
+      "learning_rate": 2.462261379980163e-07,
+      "loss": 0.5116,
+      "step": 8397
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1816250021882353,
+      "learning_rate": 2.4568679076966197e-07,
+      "loss": 0.4746,
+      "step": 8398
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.474278505273215,
+      "learning_rate": 2.451480200181744e-07,
+      "loss": 0.4843,
+      "step": 8399
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7828825729978037,
+      "learning_rate": 2.44609825808883e-07,
+      "loss": 0.4995,
+      "step": 8400
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.598460926061491,
+      "learning_rate": 2.44072208207044e-07,
+      "loss": 0.5384,
+      "step": 8401
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6997032486769714,
+      "learning_rate": 2.435351672778463e-07,
+      "loss": 0.4876,
+      "step": 8402
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.700918069350987,
+      "learning_rate": 2.4299870308640726e-07,
+      "loss": 0.4725,
+      "step": 8403
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9661941260514517,
+      "learning_rate": 2.4246281569777485e-07,
+      "loss": 0.4766,
+      "step": 8404
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.697670304250903,
+      "learning_rate": 2.4192750517692877e-07,
+      "loss": 0.4553,
+      "step": 8405
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.272630712069542,
+      "learning_rate": 2.4139277158877537e-07,
+      "loss": 0.4474,
+      "step": 8406
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.047957173234134,
+      "learning_rate": 2.40858614998154e-07,
+      "loss": 0.4762,
+      "step": 8407
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.131853051420306,
+      "learning_rate": 2.403250354698333e-07,
+      "loss": 0.4301,
+      "step": 8408
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.184968921442552,
+      "learning_rate": 2.3979203306851165e-07,
+      "loss": 0.4423,
+      "step": 8409
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.483886341322782,
+      "learning_rate": 2.3925960785881554e-07,
+      "loss": 0.5208,
+      "step": 8410
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.042742395198748,
+      "learning_rate": 2.3872775990530505e-07,
+      "loss": 0.4595,
+      "step": 8411
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.308612656695638,
+      "learning_rate": 2.381964892724692e-07,
+      "loss": 0.4721,
+      "step": 8412
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8434004643351454,
+      "learning_rate": 2.3766579602472474e-07,
+      "loss": 0.5042,
+      "step": 8413
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0735644662416934,
+      "learning_rate": 2.3713568022642018e-07,
+      "loss": 0.4572,
+      "step": 8414
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.8228169694490255,
+      "learning_rate": 2.3660614194183584e-07,
+      "loss": 0.5302,
+      "step": 8415
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2990128954383553,
+      "learning_rate": 2.3607718123517753e-07,
+      "loss": 0.5031,
+      "step": 8416
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.375471351136477,
+      "learning_rate": 2.3554879817058507e-07,
+      "loss": 0.4793,
+      "step": 8417
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0287615756005426,
+      "learning_rate": 2.3502099281212775e-07,
+      "loss": 0.4374,
+      "step": 8418
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 16.193061395229776,
+      "learning_rate": 2.3449376522380106e-07,
+      "loss": 0.4721,
+      "step": 8419
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.166132701042524,
+      "learning_rate": 2.3396711546953444e-07,
+      "loss": 0.4779,
+      "step": 8420
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.173722685470219,
+      "learning_rate": 2.3344104361318677e-07,
+      "loss": 0.4418,
+      "step": 8421
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.42652142168197577,
+      "learning_rate": 2.329155497185448e-07,
+      "loss": 0.2856,
+      "step": 8422
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.36833057701833083,
+      "learning_rate": 2.3239063384932703e-07,
+      "loss": 0.2811,
+      "step": 8423
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.347163932365568,
+      "learning_rate": 2.3186629606918198e-07,
+      "loss": 0.4682,
+      "step": 8424
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.002119284070395,
+      "learning_rate": 2.3134253644168604e-07,
+      "loss": 0.4918,
+      "step": 8425
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.414264132481698,
+      "learning_rate": 2.308193550303478e-07,
+      "loss": 0.4844,
+      "step": 8426
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0236623367676496,
+      "learning_rate": 2.3029675189860546e-07,
+      "loss": 0.4743,
+      "step": 8427
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.582284846233034,
+      "learning_rate": 2.297747271098244e-07,
+      "loss": 0.4986,
+      "step": 8428
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5266806803874617,
+      "learning_rate": 2.2925328072730402e-07,
+      "loss": 0.4745,
+      "step": 8429
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.650983531983849,
+      "learning_rate": 2.287324128142704e-07,
+      "loss": 0.4863,
+      "step": 8430
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3523731411486,
+      "learning_rate": 2.2821212343388077e-07,
+      "loss": 0.5185,
+      "step": 8431
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.306120187563471,
+      "learning_rate": 2.2769241264922194e-07,
+      "loss": 0.4595,
+      "step": 8432
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.7477635922730315,
+      "learning_rate": 2.2717328052331122e-07,
+      "loss": 0.471,
+      "step": 8433
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.428780125588758,
+      "learning_rate": 2.2665472711909387e-07,
+      "loss": 0.468,
+      "step": 8434
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.216177634699311,
+      "learning_rate": 2.2613675249944677e-07,
+      "loss": 0.5008,
+      "step": 8435
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.505815871917193,
+      "learning_rate": 2.256193567271775e-07,
+      "loss": 0.512,
+      "step": 8436
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8912249474317737,
+      "learning_rate": 2.251025398650203e-07,
+      "loss": 0.5298,
+      "step": 8437
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.392805506335058,
+      "learning_rate": 2.2458630197564225e-07,
+      "loss": 0.512,
+      "step": 8438
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3802271563810535,
+      "learning_rate": 2.240706431216383e-07,
+      "loss": 0.4662,
+      "step": 8439
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3221194396829827,
+      "learning_rate": 2.2355556336553452e-07,
+      "loss": 0.4965,
+      "step": 8440
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.3438981841537485,
+      "learning_rate": 2.2304106276978432e-07,
+      "loss": 0.4668,
+      "step": 8441
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.284285226539707,
+      "learning_rate": 2.2252714139677445e-07,
+      "loss": 0.4973,
+      "step": 8442
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.873381882939235,
+      "learning_rate": 2.2201379930882006e-07,
+      "loss": 0.4537,
+      "step": 8443
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.676152334972608,
+      "learning_rate": 2.2150103656816357e-07,
+      "loss": 0.5101,
+      "step": 8444
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.978934924830429,
+      "learning_rate": 2.2098885323698027e-07,
+      "loss": 0.4705,
+      "step": 8445
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.856098808841864,
+      "learning_rate": 2.2047724937737547e-07,
+      "loss": 0.4692,
+      "step": 8446
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.630454740974109,
+      "learning_rate": 2.1996622505138064e-07,
+      "loss": 0.4834,
+      "step": 8447
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.49782795131801,
+      "learning_rate": 2.1945578032096016e-07,
+      "loss": 0.4936,
+      "step": 8448
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.527504291954565,
+      "learning_rate": 2.1894591524800835e-07,
+      "loss": 0.5101,
+      "step": 8449
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.5524146470747855,
+      "learning_rate": 2.1843662989434687e-07,
+      "loss": 0.4642,
+      "step": 8450
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.318007930231444,
+      "learning_rate": 2.179279243217286e-07,
+      "loss": 0.5012,
+      "step": 8451
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.71695391202553,
+      "learning_rate": 2.1741979859183582e-07,
+      "loss": 0.49,
+      "step": 8452
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.474192414455861,
+      "learning_rate": 2.169122527662798e-07,
+      "loss": 0.5129,
+      "step": 8453
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.536041627330342,
+      "learning_rate": 2.1640528690660302e-07,
+      "loss": 0.5066,
+      "step": 8454
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.270386686601826,
+      "learning_rate": 2.158989010742779e-07,
+      "loss": 0.4946,
+      "step": 8455
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7937491049522647,
+      "learning_rate": 2.1539309533070318e-07,
+      "loss": 0.5044,
+      "step": 8456
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.378915984642704,
+      "learning_rate": 2.1488786973721088e-07,
+      "loss": 0.4591,
+      "step": 8457
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.7873909052492944,
+      "learning_rate": 2.1438322435506198e-07,
+      "loss": 0.5178,
+      "step": 8458
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0836603688222364,
+      "learning_rate": 2.138791592454448e-07,
+      "loss": 0.5124,
+      "step": 8459
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.42150942816716,
+      "learning_rate": 2.1337567446947982e-07,
+      "loss": 0.4577,
+      "step": 8460
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.228347739984215,
+      "learning_rate": 2.1287277008821662e-07,
+      "loss": 0.5028,
+      "step": 8461
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.860392248299571,
+      "learning_rate": 2.1237044616263413e-07,
+      "loss": 0.4532,
+      "step": 8462
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7838551702115373,
+      "learning_rate": 2.118687027536398e-07,
+      "loss": 0.4639,
+      "step": 8463
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.294320282550556,
+      "learning_rate": 2.113675399220727e-07,
+      "loss": 0.5083,
+      "step": 8464
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1725078953220582,
+      "learning_rate": 2.1086695772870092e-07,
+      "loss": 0.4875,
+      "step": 8465
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.36956325664551287,
+      "learning_rate": 2.1036695623422033e-07,
+      "loss": 0.2701,
+      "step": 8466
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6287331688476843,
+      "learning_rate": 2.0986753549925963e-07,
+      "loss": 0.4561,
+      "step": 8467
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.054497555320505,
+      "learning_rate": 2.0936869558437322e-07,
+      "loss": 0.4547,
+      "step": 8468
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.796390626501031,
+      "learning_rate": 2.0887043655004825e-07,
+      "loss": 0.5115,
+      "step": 8469
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.8372478007766295,
+      "learning_rate": 2.0837275845670135e-07,
+      "loss": 0.4582,
+      "step": 8470
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.197855493341651,
+      "learning_rate": 2.0787566136467708e-07,
+      "loss": 0.4257,
+      "step": 8471
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.40131385160904254,
+      "learning_rate": 2.0737914533424885e-07,
+      "loss": 0.2735,
+      "step": 8472
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.8782187381080018,
+      "learning_rate": 2.0688321042562186e-07,
+      "loss": 0.4448,
+      "step": 8473
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.52675906860291,
+      "learning_rate": 2.0638785669893026e-07,
+      "loss": 0.4539,
+      "step": 8474
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.075272166773583,
+      "learning_rate": 2.0589308421423704e-07,
+      "loss": 0.4633,
+      "step": 8475
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5829988181134866,
+      "learning_rate": 2.053988930315348e-07,
+      "loss": 0.4807,
+      "step": 8476
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.40356954225767916,
+      "learning_rate": 2.0490528321074732e-07,
+      "loss": 0.2858,
+      "step": 8477
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.239540104690527,
+      "learning_rate": 2.0441225481172444e-07,
+      "loss": 0.4987,
+      "step": 8478
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4055408075764671,
+      "learning_rate": 2.0391980789424892e-07,
+      "loss": 0.2818,
+      "step": 8479
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.258222744057416,
+      "learning_rate": 2.034279425180319e-07,
+      "loss": 0.5033,
+      "step": 8480
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1618122220972587,
+      "learning_rate": 2.0293665874271228e-07,
+      "loss": 0.4493,
+      "step": 8481
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.149526167087179,
+      "learning_rate": 2.0244595662786138e-07,
+      "loss": 0.4592,
+      "step": 8482
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.026448185697821,
+      "learning_rate": 2.019558362329782e-07,
+      "loss": 0.4751,
+      "step": 8483
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.452597203478537,
+      "learning_rate": 2.014662976174908e-07,
+      "loss": 0.4431,
+      "step": 8484
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.001590812716567,
+      "learning_rate": 2.0097734084075726e-07,
+      "loss": 0.5033,
+      "step": 8485
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.545714848687938,
+      "learning_rate": 2.004889659620668e-07,
+      "loss": 0.4503,
+      "step": 8486
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.29857541966741,
+      "learning_rate": 2.0000117304063483e-07,
+      "loss": 0.4728,
+      "step": 8487
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.7537868883769,
+      "learning_rate": 1.9951396213560904e-07,
+      "loss": 0.5062,
+      "step": 8488
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.705300193197994,
+      "learning_rate": 1.9902733330606605e-07,
+      "loss": 0.4699,
+      "step": 8489
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5036210127549086,
+      "learning_rate": 1.9854128661100924e-07,
+      "loss": 0.4759,
+      "step": 8490
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.048691788078977,
+      "learning_rate": 1.980558221093748e-07,
+      "loss": 0.4991,
+      "step": 8491
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 15.943128688759383,
+      "learning_rate": 1.9757093986002729e-07,
+      "loss": 0.4548,
+      "step": 8492
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0677098906144953,
+      "learning_rate": 1.9708663992175969e-07,
+      "loss": 0.4964,
+      "step": 8493
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.267766319134077,
+      "learning_rate": 1.966029223532945e-07,
+      "loss": 0.4702,
+      "step": 8494
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 23.52885519353142,
+      "learning_rate": 1.9611978721328473e-07,
+      "loss": 0.5107,
+      "step": 8495
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.190520949592765,
+      "learning_rate": 1.9563723456031303e-07,
+      "loss": 0.4284,
+      "step": 8496
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.662122979098868,
+      "learning_rate": 1.9515526445288923e-07,
+      "loss": 0.4993,
+      "step": 8497
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.274045017898971,
+      "learning_rate": 1.946738769494555e-07,
+      "loss": 0.479,
+      "step": 8498
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.103409287956581,
+      "learning_rate": 1.9419307210837955e-07,
+      "loss": 0.4672,
+      "step": 8499
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.271574959896244,
+      "learning_rate": 1.9371284998796146e-07,
+      "loss": 0.497,
+      "step": 8500
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.319121354128411,
+      "learning_rate": 1.9323321064643131e-07,
+      "loss": 0.4396,
+      "step": 8501
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.39532129256154425,
+      "learning_rate": 1.9275415414194476e-07,
+      "loss": 0.275,
+      "step": 8502
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4209001651275715,
+      "learning_rate": 1.9227568053259094e-07,
+      "loss": 0.2665,
+      "step": 8503
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.177363725930472,
+      "learning_rate": 1.9179778987638508e-07,
+      "loss": 0.4478,
+      "step": 8504
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.169643252209534,
+      "learning_rate": 1.913204822312742e-07,
+      "loss": 0.458,
+      "step": 8505
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.307833408994178,
+      "learning_rate": 1.9084375765513197e-07,
+      "loss": 0.5154,
+      "step": 8506
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5251514743711625,
+      "learning_rate": 1.9036761620576438e-07,
+      "loss": 0.4772,
+      "step": 8507
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.932834494402399,
+      "learning_rate": 1.898920579409047e-07,
+      "loss": 0.4774,
+      "step": 8508
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.520062139824719,
+      "learning_rate": 1.894170829182157e-07,
+      "loss": 0.4602,
+      "step": 8509
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.596601188406675,
+      "learning_rate": 1.8894269119528963e-07,
+      "loss": 0.5306,
+      "step": 8510
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.183920885549393,
+      "learning_rate": 1.8846888282964936e-07,
+      "loss": 0.5065,
+      "step": 8511
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.936903963182984,
+      "learning_rate": 1.8799565787874397e-07,
+      "loss": 0.4391,
+      "step": 8512
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.40337149148500034,
+      "learning_rate": 1.8752301639995528e-07,
+      "loss": 0.2774,
+      "step": 8513
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.4654515202091463,
+      "learning_rate": 1.8705095845059252e-07,
+      "loss": 0.4889,
+      "step": 8514
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.698483309795595,
+      "learning_rate": 1.8657948408789261e-07,
+      "loss": 0.5233,
+      "step": 8515
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.3298713838160285,
+      "learning_rate": 1.8610859336902488e-07,
+      "loss": 0.4863,
+      "step": 8516
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.300308795218516,
+      "learning_rate": 1.8563828635108694e-07,
+      "loss": 0.4639,
+      "step": 8517
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.370478169013373,
+      "learning_rate": 1.8516856309110375e-07,
+      "loss": 0.2619,
+      "step": 8518
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.141980083538687,
+      "learning_rate": 1.8469942364603145e-07,
+      "loss": 0.5016,
+      "step": 8519
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.912659905660343,
+      "learning_rate": 1.8423086807275558e-07,
+      "loss": 0.5015,
+      "step": 8520
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.980845487048708,
+      "learning_rate": 1.8376289642808854e-07,
+      "loss": 0.4492,
+      "step": 8521
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.00479487302828,
+      "learning_rate": 1.8329550876877488e-07,
+      "loss": 0.476,
+      "step": 8522
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.28082654285319,
+      "learning_rate": 1.8282870515148654e-07,
+      "loss": 0.4681,
+      "step": 8523
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3744572092432679,
+      "learning_rate": 1.8236248563282545e-07,
+      "loss": 0.2758,
+      "step": 8524
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.497542722896531,
+      "learning_rate": 1.8189685026932137e-07,
+      "loss": 0.4708,
+      "step": 8525
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.344887136230116,
+      "learning_rate": 1.8143179911743415e-07,
+      "loss": 0.4916,
+      "step": 8526
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4372059323999279,
+      "learning_rate": 1.8096733223355477e-07,
+      "loss": 0.2922,
+      "step": 8527
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9678473568212227,
+      "learning_rate": 1.805034496739988e-07,
+      "loss": 0.5215,
+      "step": 8528
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5488500830253704,
+      "learning_rate": 1.8004015149501564e-07,
+      "loss": 0.4596,
+      "step": 8529
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.429556963062183,
+      "learning_rate": 1.795774377527809e-07,
+      "loss": 0.2853,
+      "step": 8530
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.36749275105749546,
+      "learning_rate": 1.7911530850339976e-07,
+      "loss": 0.2773,
+      "step": 8531
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.176622631465613,
+      "learning_rate": 1.7865376380290845e-07,
+      "loss": 0.4747,
+      "step": 8532
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3816889887904104,
+      "learning_rate": 1.7819280370726943e-07,
+      "loss": 0.2686,
+      "step": 8533
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3845783915801841,
+      "learning_rate": 1.7773242827237635e-07,
+      "loss": 0.2673,
+      "step": 8534
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.752493631616692,
+      "learning_rate": 1.7727263755405178e-07,
+      "loss": 0.4917,
+      "step": 8535
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.633625673029752,
+      "learning_rate": 1.7681343160804608e-07,
+      "loss": 0.474,
+      "step": 8536
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.36942425763581016,
+      "learning_rate": 1.7635481049003979e-07,
+      "loss": 0.2517,
+      "step": 8537
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.922860087412812,
+      "learning_rate": 1.7589677425564222e-07,
+      "loss": 0.498,
+      "step": 8538
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.476653998790395,
+      "learning_rate": 1.7543932296039234e-07,
+      "loss": 0.4672,
+      "step": 8539
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.709347630795656,
+      "learning_rate": 1.7498245665975689e-07,
+      "loss": 0.4691,
+      "step": 8540
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.051590430544429,
+      "learning_rate": 1.7452617540913265e-07,
+      "loss": 0.485,
+      "step": 8541
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.31775909061491,
+      "learning_rate": 1.7407047926384646e-07,
+      "loss": 0.5019,
+      "step": 8542
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.030349072291054,
+      "learning_rate": 1.7361536827915138e-07,
+      "loss": 0.496,
+      "step": 8543
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.195234058245526,
+      "learning_rate": 1.7316084251023214e-07,
+      "loss": 0.5196,
+      "step": 8544
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 30.636174943087187,
+      "learning_rate": 1.727069020122024e-07,
+      "loss": 0.4941,
+      "step": 8545
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.750301441191228,
+      "learning_rate": 1.722535468401021e-07,
+      "loss": 0.4811,
+      "step": 8546
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.709973710692066,
+      "learning_rate": 1.7180077704890274e-07,
+      "loss": 0.5047,
+      "step": 8547
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.31321973778015,
+      "learning_rate": 1.7134859269350546e-07,
+      "loss": 0.5304,
+      "step": 8548
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5289925787133405,
+      "learning_rate": 1.7089699382873748e-07,
+      "loss": 0.4855,
+      "step": 8549
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.733006991849277,
+      "learning_rate": 1.7044598050935722e-07,
+      "loss": 0.4832,
+      "step": 8550
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.377520532904826,
+      "learning_rate": 1.6999555279005264e-07,
+      "loss": 0.5268,
+      "step": 8551
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.584739116603017,
+      "learning_rate": 1.695457107254378e-07,
+      "loss": 0.4644,
+      "step": 8552
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.527353728233938,
+      "learning_rate": 1.6909645437005906e-07,
+      "loss": 0.4661,
+      "step": 8553
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2399847851449155,
+      "learning_rate": 1.6864778377839064e-07,
+      "loss": 0.4616,
+      "step": 8554
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.669439990274473,
+      "learning_rate": 1.6819969900483347e-07,
+      "loss": 0.5194,
+      "step": 8555
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.8247011503774155,
+      "learning_rate": 1.6775220010372185e-07,
+      "loss": 0.4473,
+      "step": 8556
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.6012408753429344,
+      "learning_rate": 1.6730528712931404e-07,
+      "loss": 0.5064,
+      "step": 8557
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.6746259176108875,
+      "learning_rate": 1.6685896013580226e-07,
+      "loss": 0.4574,
+      "step": 8558
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.416119125126853,
+      "learning_rate": 1.6641321917730268e-07,
+      "loss": 0.4814,
+      "step": 8559
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.565739079526291,
+      "learning_rate": 1.6596806430786428e-07,
+      "loss": 0.4716,
+      "step": 8560
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5509242697356185,
+      "learning_rate": 1.6552349558146442e-07,
+      "loss": 0.4467,
+      "step": 8561
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.742811387809005,
+      "learning_rate": 1.650795130520072e-07,
+      "loss": 0.4582,
+      "step": 8562
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.797983019477226,
+      "learning_rate": 1.6463611677332792e-07,
+      "loss": 0.4973,
+      "step": 8563
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9768968109365814,
+      "learning_rate": 1.6419330679918855e-07,
+      "loss": 0.5154,
+      "step": 8564
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.665346603681885,
+      "learning_rate": 1.6375108318328282e-07,
+      "loss": 0.5051,
+      "step": 8565
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.783912842571154,
+      "learning_rate": 1.633094459792317e-07,
+      "loss": 0.5195,
+      "step": 8566
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.808841631739329,
+      "learning_rate": 1.6286839524058462e-07,
+      "loss": 0.5362,
+      "step": 8567
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.053989138516079,
+      "learning_rate": 1.6242793102082043e-07,
+      "loss": 0.4545,
+      "step": 8568
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2253591431440256,
+      "learning_rate": 1.6198805337334755e-07,
+      "loss": 0.5321,
+      "step": 8569
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2576859160134153,
+      "learning_rate": 1.6154876235150273e-07,
+      "loss": 0.4926,
+      "step": 8570
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.37888197052444805,
+      "learning_rate": 1.6111005800855063e-07,
+      "loss": 0.2506,
+      "step": 8571
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.7509920723893306,
+      "learning_rate": 1.6067194039768585e-07,
+      "loss": 0.4484,
+      "step": 8572
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8165826543954338,
+      "learning_rate": 1.6023440957203262e-07,
+      "loss": 0.4669,
+      "step": 8573
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2403976999585957,
+      "learning_rate": 1.5979746558464237e-07,
+      "loss": 0.4294,
+      "step": 8574
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.7810410397844345,
+      "learning_rate": 1.593611084884955e-07,
+      "loss": 0.4991,
+      "step": 8575
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.1016066206934365,
+      "learning_rate": 1.589253383365036e-07,
+      "loss": 0.4445,
+      "step": 8576
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.068145338639536,
+      "learning_rate": 1.584901551815038e-07,
+      "loss": 0.4653,
+      "step": 8577
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.140838814542984,
+      "learning_rate": 1.5805555907626336e-07,
+      "loss": 0.4546,
+      "step": 8578
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.882328263025543,
+      "learning_rate": 1.5762155007347957e-07,
+      "loss": 0.4547,
+      "step": 8579
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.1860960756413075,
+      "learning_rate": 1.57188128225777e-07,
+      "loss": 0.461,
+      "step": 8580
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.037615811857658,
+      "learning_rate": 1.5675529358570918e-07,
+      "loss": 0.4725,
+      "step": 8581
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.000784865346864,
+      "learning_rate": 1.5632304620575965e-07,
+      "loss": 0.4921,
+      "step": 8582
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8424364947509924,
+      "learning_rate": 1.5589138613833933e-07,
+      "loss": 0.4979,
+      "step": 8583
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.466192459807887,
+      "learning_rate": 1.5546031343578793e-07,
+      "loss": 0.4339,
+      "step": 8584
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.3419897830366425,
+      "learning_rate": 1.5502982815037648e-07,
+      "loss": 0.5293,
+      "step": 8585
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.028554527183518,
+      "learning_rate": 1.5459993033430043e-07,
+      "loss": 0.4631,
+      "step": 8586
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0281345306335816,
+      "learning_rate": 1.5417062003968808e-07,
+      "loss": 0.5258,
+      "step": 8587
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.463431962303145,
+      "learning_rate": 1.537418973185939e-07,
+      "loss": 0.4275,
+      "step": 8588
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.20599529434043,
+      "learning_rate": 1.5331376222300297e-07,
+      "loss": 0.5252,
+      "step": 8589
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.138068151095498,
+      "learning_rate": 1.5288621480482658e-07,
+      "loss": 0.5351,
+      "step": 8590
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.197713839417613,
+      "learning_rate": 1.524592551159071e-07,
+      "loss": 0.4978,
+      "step": 8591
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1314182008325497,
+      "learning_rate": 1.520328832080159e-07,
+      "loss": 0.4477,
+      "step": 8592
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.81783894679705,
+      "learning_rate": 1.5160709913285e-07,
+      "loss": 0.454,
+      "step": 8593
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.3189244894693815,
+      "learning_rate": 1.5118190294203915e-07,
+      "loss": 0.4683,
+      "step": 8594
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.754763183682165,
+      "learning_rate": 1.5075729468713829e-07,
+      "loss": 0.511,
+      "step": 8595
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.529540051375486,
+      "learning_rate": 1.5033327441963398e-07,
+      "loss": 0.5358,
+      "step": 8596
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.891588251199776,
+      "learning_rate": 1.4990984219093897e-07,
+      "loss": 0.4654,
+      "step": 8597
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.202881865459107,
+      "learning_rate": 1.4948699805239718e-07,
+      "loss": 0.4914,
+      "step": 8598
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.723596146945328,
+      "learning_rate": 1.490647420552782e-07,
+      "loss": 0.495,
+      "step": 8599
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.37985834425030734,
+      "learning_rate": 1.486430742507833e-07,
+      "loss": 0.2827,
+      "step": 8600
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8540475968298673,
+      "learning_rate": 1.4822199469004094e-07,
+      "loss": 0.4579,
+      "step": 8601
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.783641487218594,
+      "learning_rate": 1.4780150342410814e-07,
+      "loss": 0.4683,
+      "step": 8602
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.09075059125622,
+      "learning_rate": 1.473816005039713e-07,
+      "loss": 0.4737,
+      "step": 8603
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.379546589878345,
+      "learning_rate": 1.4696228598054473e-07,
+      "loss": 0.5102,
+      "step": 8604
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.41765714331588377,
+      "learning_rate": 1.4654355990467217e-07,
+      "loss": 0.2904,
+      "step": 8605
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.487846467209492,
+      "learning_rate": 1.4612542232712522e-07,
+      "loss": 0.5094,
+      "step": 8606
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.150321318629627,
+      "learning_rate": 1.4570787329860504e-07,
+      "loss": 0.4684,
+      "step": 8607
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.654838103928217,
+      "learning_rate": 1.4529091286973994e-07,
+      "loss": 0.4518,
+      "step": 8608
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.824559970638505,
+      "learning_rate": 1.44874541091089e-07,
+      "loss": 0.5098,
+      "step": 8609
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.267706052239964,
+      "learning_rate": 1.444587580131379e-07,
+      "loss": 0.4857,
+      "step": 8610
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9902473790553334,
+      "learning_rate": 1.4404356368630134e-07,
+      "loss": 0.454,
+      "step": 8611
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.041208723958112,
+      "learning_rate": 1.4362895816092403e-07,
+      "loss": 0.4656,
+      "step": 8612
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.38695308834149583,
+      "learning_rate": 1.4321494148727855e-07,
+      "loss": 0.2689,
+      "step": 8613
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.879692514865846,
+      "learning_rate": 1.4280151371556472e-07,
+      "loss": 0.5409,
+      "step": 8614
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.605676085243845,
+      "learning_rate": 1.4238867489591302e-07,
+      "loss": 0.4711,
+      "step": 8615
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.096761261438603,
+      "learning_rate": 1.4197642507838115e-07,
+      "loss": 0.4978,
+      "step": 8616
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.307564188778468,
+      "learning_rate": 1.4156476431295586e-07,
+      "loss": 0.4721,
+      "step": 8617
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.1930805748207085,
+      "learning_rate": 1.4115369264955215e-07,
+      "loss": 0.4265,
+      "step": 8618
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.655063555129131,
+      "learning_rate": 1.4074321013801462e-07,
+      "loss": 0.5085,
+      "step": 8619
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.083493616661105,
+      "learning_rate": 1.4033331682811568e-07,
+      "loss": 0.4372,
+      "step": 8620
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.654254640463673,
+      "learning_rate": 1.3992401276955502e-07,
+      "loss": 0.4754,
+      "step": 8621
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9428513218218053,
+      "learning_rate": 1.3951529801196295e-07,
+      "loss": 0.5491,
+      "step": 8622
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3603930879083423,
+      "learning_rate": 1.391071726048987e-07,
+      "loss": 0.2831,
+      "step": 8623
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.26762399631875,
+      "learning_rate": 1.3869963659784657e-07,
+      "loss": 0.4645,
+      "step": 8624
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.832744462054685,
+      "learning_rate": 1.3829269004022426e-07,
+      "loss": 0.4839,
+      "step": 8625
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.109887441871438,
+      "learning_rate": 1.3788633298137288e-07,
+      "loss": 0.4955,
+      "step": 8626
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.799538323770272,
+      "learning_rate": 1.3748056547056632e-07,
+      "loss": 0.5113,
+      "step": 8627
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.558024848587687,
+      "learning_rate": 1.3707538755700522e-07,
+      "loss": 0.4875,
+      "step": 8628
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1290101231474003,
+      "learning_rate": 1.3667079928981807e-07,
+      "loss": 0.5031,
+      "step": 8629
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.815160455525418,
+      "learning_rate": 1.3626680071806342e-07,
+      "loss": 0.4825,
+      "step": 8630
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0296938457265274,
+      "learning_rate": 1.358633918907265e-07,
+      "loss": 0.4621,
+      "step": 8631
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.684790611607264,
+      "learning_rate": 1.354605728567232e-07,
+      "loss": 0.4223,
+      "step": 8632
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4597434803548324,
+      "learning_rate": 1.3505834366489502e-07,
+      "loss": 0.4951,
+      "step": 8633
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9481726499003136,
+      "learning_rate": 1.3465670436401513e-07,
+      "loss": 0.4782,
+      "step": 8634
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.36453233388199946,
+      "learning_rate": 1.3425565500278348e-07,
+      "loss": 0.2623,
+      "step": 8635
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.0017490344149005,
+      "learning_rate": 1.3385519562982842e-07,
+      "loss": 0.475,
+      "step": 8636
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.159307230599198,
+      "learning_rate": 1.334553262937066e-07,
+      "loss": 0.4957,
+      "step": 8637
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.5234688100393505,
+      "learning_rate": 1.3305604704290432e-07,
+      "loss": 0.4791,
+      "step": 8638
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 15.135858691391636,
+      "learning_rate": 1.32657357925835e-07,
+      "loss": 0.5502,
+      "step": 8639
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.042738045122541,
+      "learning_rate": 1.322592589908417e-07,
+      "loss": 0.4845,
+      "step": 8640
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.578021016004308,
+      "learning_rate": 1.318617502861952e-07,
+      "loss": 0.4721,
+      "step": 8641
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.0983418950880255,
+      "learning_rate": 1.3146483186009418e-07,
+      "loss": 0.462,
+      "step": 8642
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.38245255692943614,
+      "learning_rate": 1.3106850376066626e-07,
+      "loss": 0.2697,
+      "step": 8643
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.39955171606719286,
+      "learning_rate": 1.3067276603596856e-07,
+      "loss": 0.2659,
+      "step": 8644
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0527696839230685,
+      "learning_rate": 1.3027761873398437e-07,
+      "loss": 0.5375,
+      "step": 8645
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3414591356437526,
+      "learning_rate": 1.2988306190262757e-07,
+      "loss": 0.5029,
+      "step": 8646
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2174603744790726,
+      "learning_rate": 1.2948909558974045e-07,
+      "loss": 0.4968,
+      "step": 8647
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.8599923842406385,
+      "learning_rate": 1.2909571984309032e-07,
+      "loss": 0.4615,
+      "step": 8648
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.523392945539515,
+      "learning_rate": 1.2870293471037743e-07,
+      "loss": 0.4402,
+      "step": 8649
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.747916441612022,
+      "learning_rate": 1.2831074023922752e-07,
+      "loss": 0.4401,
+      "step": 8650
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3817325999802095,
+      "learning_rate": 1.279191364771959e-07,
+      "loss": 0.4835,
+      "step": 8651
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.023270670577837,
+      "learning_rate": 1.2752812347176514e-07,
+      "loss": 0.4682,
+      "step": 8652
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.228015247632366,
+      "learning_rate": 1.271377012703473e-07,
+      "loss": 0.5154,
+      "step": 8653
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.002824931173263,
+      "learning_rate": 1.2674786992028287e-07,
+      "loss": 0.4586,
+      "step": 8654
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3789304628498259,
+      "learning_rate": 1.263586294688396e-07,
+      "loss": 0.2786,
+      "step": 8655
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.385866951774461,
+      "learning_rate": 1.259699799632147e-07,
+      "loss": 0.4924,
+      "step": 8656
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5654764859100707,
+      "learning_rate": 1.2558192145053327e-07,
+      "loss": 0.462,
+      "step": 8657
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1157064912575048,
+      "learning_rate": 1.251944539778477e-07,
+      "loss": 0.4704,
+      "step": 8658
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2453597954032998,
+      "learning_rate": 1.2480757759214145e-07,
+      "loss": 0.4739,
+      "step": 8659
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9877141550786903,
+      "learning_rate": 1.2442129234032373e-07,
+      "loss": 0.467,
+      "step": 8660
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.714326004871507,
+      "learning_rate": 1.2403559826923205e-07,
+      "loss": 0.394,
+      "step": 8661
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.447843161296101,
+      "learning_rate": 1.236504954256351e-07,
+      "loss": 0.4702,
+      "step": 8662
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.8227851656395275,
+      "learning_rate": 1.2326598385622723e-07,
+      "loss": 0.4595,
+      "step": 8663
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.273664824720929,
+      "learning_rate": 1.2288206360763056e-07,
+      "loss": 0.4416,
+      "step": 8664
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.491781957723075,
+      "learning_rate": 1.2249873472639783e-07,
+      "loss": 0.4758,
+      "step": 8665
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.678980902082391,
+      "learning_rate": 1.2211599725900914e-07,
+      "loss": 0.5025,
+      "step": 8666
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.345642909506876,
+      "learning_rate": 1.2173385125187177e-07,
+      "loss": 0.4974,
+      "step": 8667
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.807305818164364,
+      "learning_rate": 1.213522967513231e-07,
+      "loss": 0.4211,
+      "step": 8668
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.220445737163746,
+      "learning_rate": 1.2097133380362835e-07,
+      "loss": 0.484,
+      "step": 8669
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.004284060112766,
+      "learning_rate": 1.2059096245497948e-07,
+      "loss": 0.4281,
+      "step": 8670
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.902736052515889,
+      "learning_rate": 1.202111827514979e-07,
+      "loss": 0.4772,
+      "step": 8671
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.38275022737966025,
+      "learning_rate": 1.1983199473923457e-07,
+      "loss": 0.2749,
+      "step": 8672
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.540456443541057,
+      "learning_rate": 1.1945339846416603e-07,
+      "loss": 0.5294,
+      "step": 8673
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6260349015835365,
+      "learning_rate": 1.1907539397219837e-07,
+      "loss": 0.4447,
+      "step": 8674
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.180986795610461,
+      "learning_rate": 1.1869798130916654e-07,
+      "loss": 0.4905,
+      "step": 8675
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.322169851594932,
+      "learning_rate": 1.1832116052083232e-07,
+      "loss": 0.4998,
+      "step": 8676
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.841025913517013,
+      "learning_rate": 1.1794493165288745e-07,
+      "loss": 0.4588,
+      "step": 8677
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.8963261663139135,
+      "learning_rate": 1.1756929475095103e-07,
+      "loss": 0.4536,
+      "step": 8678
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.052800282138882,
+      "learning_rate": 1.1719424986056938e-07,
+      "loss": 0.4827,
+      "step": 8679
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9787414064942945,
+      "learning_rate": 1.168197970272189e-07,
+      "loss": 0.4628,
+      "step": 8680
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.6636088569090335,
+      "learning_rate": 1.1644593629630274e-07,
+      "loss": 0.4999,
+      "step": 8681
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.5683237187542676,
+      "learning_rate": 1.1607266771315295e-07,
+      "loss": 0.4797,
+      "step": 8682
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.170336359901062,
+      "learning_rate": 1.1569999132303001e-07,
+      "loss": 0.4077,
+      "step": 8683
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.435579104072131,
+      "learning_rate": 1.1532790717112164e-07,
+      "loss": 0.511,
+      "step": 8684
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.553660366985488,
+      "learning_rate": 1.1495641530254454e-07,
+      "loss": 0.4762,
+      "step": 8685
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2030446999425197,
+      "learning_rate": 1.1458551576234322e-07,
+      "loss": 0.4872,
+      "step": 8686
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.1381904958773035,
+      "learning_rate": 1.1421520859549063e-07,
+      "loss": 0.4777,
+      "step": 8687
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.883739911253023,
+      "learning_rate": 1.1384549384688804e-07,
+      "loss": 0.4853,
+      "step": 8688
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0989909618401197,
+      "learning_rate": 1.134763715613646e-07,
+      "loss": 0.4707,
+      "step": 8689
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.226927878247467,
+      "learning_rate": 1.1310784178367728e-07,
+      "loss": 0.4625,
+      "step": 8690
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.40364270152786125,
+      "learning_rate": 1.1273990455851203e-07,
+      "loss": 0.258,
+      "step": 8691
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.374733735879726,
+      "learning_rate": 1.1237255993048202e-07,
+      "loss": 0.5117,
+      "step": 8692
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6139958654196414,
+      "learning_rate": 1.1200580794412997e-07,
+      "loss": 0.4267,
+      "step": 8693
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 14.338928289191278,
+      "learning_rate": 1.1163964864392474e-07,
+      "loss": 0.4283,
+      "step": 8694
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.1453455591404635,
+      "learning_rate": 1.1127408207426471e-07,
+      "loss": 0.4576,
+      "step": 8695
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.38312893066728126,
+      "learning_rate": 1.109091082794761e-07,
+      "loss": 0.2999,
+      "step": 8696
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.553207131259288,
+      "learning_rate": 1.1054472730381405e-07,
+      "loss": 0.4543,
+      "step": 8697
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7141148192689215,
+      "learning_rate": 1.1018093919145989e-07,
+      "loss": 0.4913,
+      "step": 8698
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.524259876870368,
+      "learning_rate": 1.0981774398652446e-07,
+      "loss": 0.4775,
+      "step": 8699
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.696579336453592,
+      "learning_rate": 1.0945514173304695e-07,
+      "loss": 0.449,
+      "step": 8700
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.930926081879493,
+      "learning_rate": 1.0909313247499331e-07,
+      "loss": 0.4878,
+      "step": 8701
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.741962762402157,
+      "learning_rate": 1.08731716256259e-07,
+      "loss": 0.4731,
+      "step": 8702
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.153654348034106,
+      "learning_rate": 1.0837089312066729e-07,
+      "loss": 0.4786,
+      "step": 8703
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.741206287957141,
+      "learning_rate": 1.0801066311196872e-07,
+      "loss": 0.4415,
+      "step": 8704
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.6697353946934825,
+      "learning_rate": 1.0765102627384227e-07,
+      "loss": 0.4608,
+      "step": 8705
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.241282088180614,
+      "learning_rate": 1.0729198264989582e-07,
+      "loss": 0.467,
+      "step": 8706
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9550551262950737,
+      "learning_rate": 1.0693353228366344e-07,
+      "loss": 0.4036,
+      "step": 8707
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.3518244346975345,
+      "learning_rate": 1.0657567521860978e-07,
+      "loss": 0.4981,
+      "step": 8708
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.473548086595823,
+      "learning_rate": 1.0621841149812573e-07,
+      "loss": 0.463,
+      "step": 8709
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.1755932252848496,
+      "learning_rate": 1.0586174116552994e-07,
+      "loss": 0.4862,
+      "step": 8710
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.454593446989599,
+      "learning_rate": 1.0550566426407117e-07,
+      "loss": 0.4644,
+      "step": 8711
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.063784622618975,
+      "learning_rate": 1.0515018083692486e-07,
+      "loss": 0.5276,
+      "step": 8712
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5012945646670905,
+      "learning_rate": 1.0479529092719376e-07,
+      "loss": 0.4977,
+      "step": 8713
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.564364439149957,
+      "learning_rate": 1.0444099457791013e-07,
+      "loss": 0.4674,
+      "step": 8714
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3810686945116815,
+      "learning_rate": 1.0408729183203403e-07,
+      "loss": 0.258,
+      "step": 8715
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.6803719195073565,
+      "learning_rate": 1.0373418273245228e-07,
+      "loss": 0.5055,
+      "step": 8716
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.010022432613141,
+      "learning_rate": 1.0338166732198062e-07,
+      "loss": 0.478,
+      "step": 8717
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.101322987881904,
+      "learning_rate": 1.0302974564336265e-07,
+      "loss": 0.5043,
+      "step": 8718
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.396607766897713,
+      "learning_rate": 1.0267841773927089e-07,
+      "loss": 0.2636,
+      "step": 8719
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.4615011073054753,
+      "learning_rate": 1.0232768365230462e-07,
+      "loss": 0.4185,
+      "step": 8720
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.6385922027120317,
+      "learning_rate": 1.01977543424992e-07,
+      "loss": 0.4993,
+      "step": 8721
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2288974703402564,
+      "learning_rate": 1.0162799709978743e-07,
+      "loss": 0.486,
+      "step": 8722
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.083901929452404,
+      "learning_rate": 1.012790447190759e-07,
+      "loss": 0.4633,
+      "step": 8723
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.2582805467031015,
+      "learning_rate": 1.0093068632516855e-07,
+      "loss": 0.4821,
+      "step": 8724
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.937537889078509,
+      "learning_rate": 1.005829219603055e-07,
+      "loss": 0.4825,
+      "step": 8725
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.641573450331607,
+      "learning_rate": 1.0023575166665356e-07,
+      "loss": 0.5308,
+      "step": 8726
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4926822499087424,
+      "learning_rate": 9.98891754863085e-08,
+      "loss": 0.4676,
+      "step": 8727
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7129140975781905,
+      "learning_rate": 9.954319346129504e-08,
+      "loss": 0.4372,
+      "step": 8728
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.565732553424755,
+      "learning_rate": 9.919780563356296e-08,
+      "loss": 0.4459,
+      "step": 8729
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.221895039221924,
+      "learning_rate": 9.885301204499321e-08,
+      "loss": 0.4664,
+      "step": 8730
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.8052736239229406,
+      "learning_rate": 9.850881273739232e-08,
+      "loss": 0.4984,
+      "step": 8731
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.648107932852618,
+      "learning_rate": 9.816520775249583e-08,
+      "loss": 0.4984,
+      "step": 8732
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.8273702374855456,
+      "learning_rate": 9.782219713196706e-08,
+      "loss": 0.4485,
+      "step": 8733
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3961347726389275,
+      "learning_rate": 9.747978091739774e-08,
+      "loss": 0.4795,
+      "step": 8734
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.859328961863652,
+      "learning_rate": 9.713795915030577e-08,
+      "loss": 0.46,
+      "step": 8735
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.271018401123852,
+      "learning_rate": 9.679673187214023e-08,
+      "loss": 0.4719,
+      "step": 8736
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.806888080211213,
+      "learning_rate": 9.645609912427412e-08,
+      "loss": 0.4679,
+      "step": 8737
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8334068896304734,
+      "learning_rate": 9.611606094801052e-08,
+      "loss": 0.4796,
+      "step": 8738
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.005956156277415,
+      "learning_rate": 9.577661738458144e-08,
+      "loss": 0.4897,
+      "step": 8739
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6106929515630117,
+      "learning_rate": 9.543776847514508e-08,
+      "loss": 0.4659,
+      "step": 8740
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.215691285302694,
+      "learning_rate": 9.509951426078745e-08,
+      "loss": 0.436,
+      "step": 8741
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.732944715316738,
+      "learning_rate": 9.476185478252354e-08,
+      "loss": 0.4876,
+      "step": 8742
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.38130904334274157,
+      "learning_rate": 9.442479008129557e-08,
+      "loss": 0.2827,
+      "step": 8743
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3876274714112574,
+      "learning_rate": 9.40883201979742e-08,
+      "loss": 0.2719,
+      "step": 8744
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.9052112423105,
+      "learning_rate": 9.37524451733568e-08,
+      "loss": 0.5474,
+      "step": 8745
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.3219184768860925,
+      "learning_rate": 9.341716504817022e-08,
+      "loss": 0.5024,
+      "step": 8746
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.5490671087431345,
+      "learning_rate": 9.308247986306862e-08,
+      "loss": 0.5117,
+      "step": 8747
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3738831444503264,
+      "learning_rate": 9.274838965863175e-08,
+      "loss": 0.2675,
+      "step": 8748
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.769633315480667,
+      "learning_rate": 9.241489447537111e-08,
+      "loss": 0.4779,
+      "step": 8749
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.6158683192226646,
+      "learning_rate": 9.208199435372379e-08,
+      "loss": 0.4495,
+      "step": 8750
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3634744858966187,
+      "learning_rate": 9.174968933405415e-08,
+      "loss": 0.2681,
+      "step": 8751
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.06615638600499,
+      "learning_rate": 9.141797945665609e-08,
+      "loss": 0.5276,
+      "step": 8752
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.832479218797265,
+      "learning_rate": 9.108686476175132e-08,
+      "loss": 0.5222,
+      "step": 8753
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.108965180475343,
+      "learning_rate": 9.075634528948718e-08,
+      "loss": 0.4823,
+      "step": 8754
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1722747353321834,
+      "learning_rate": 9.042642107994103e-08,
+      "loss": 0.4571,
+      "step": 8755
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.030714185860235,
+      "learning_rate": 9.009709217311702e-08,
+      "loss": 0.4609,
+      "step": 8756
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.278340595258182,
+      "learning_rate": 8.976835860894762e-08,
+      "loss": 0.4761,
+      "step": 8757
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.507950855251037,
+      "learning_rate": 8.944022042729317e-08,
+      "loss": 0.5104,
+      "step": 8758
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.819343086622712,
+      "learning_rate": 8.91126776679413e-08,
+      "loss": 0.4677,
+      "step": 8759
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5404995645413915,
+      "learning_rate": 8.878573037060689e-08,
+      "loss": 0.4538,
+      "step": 8760
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4037883641350614,
+      "learning_rate": 8.845937857493492e-08,
+      "loss": 0.2477,
+      "step": 8761
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.559107493153639,
+      "learning_rate": 8.813362232049594e-08,
+      "loss": 0.5195,
+      "step": 8762
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.921561984155913,
+      "learning_rate": 8.780846164678836e-08,
+      "loss": 0.4524,
+      "step": 8763
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.804047725578889,
+      "learning_rate": 8.748389659324008e-08,
+      "loss": 0.4712,
+      "step": 8764
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.675705947718729,
+      "learning_rate": 8.715992719920575e-08,
+      "loss": 0.4659,
+      "step": 8765
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2790028911926603,
+      "learning_rate": 8.683655350396724e-08,
+      "loss": 0.4429,
+      "step": 8766
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1816161948111206,
+      "learning_rate": 8.651377554673435e-08,
+      "loss": 0.5071,
+      "step": 8767
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.874060366099913,
+      "learning_rate": 8.619159336664684e-08,
+      "loss": 0.4603,
+      "step": 8768
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 14.940311449580989,
+      "learning_rate": 8.587000700276793e-08,
+      "loss": 0.446,
+      "step": 8769
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.926387653206238,
+      "learning_rate": 8.554901649409253e-08,
+      "loss": 0.4629,
+      "step": 8770
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.283841439385385,
+      "learning_rate": 8.522862187954173e-08,
+      "loss": 0.4826,
+      "step": 8771
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.015269593674128,
+      "learning_rate": 8.49088231979639e-08,
+      "loss": 0.4756,
+      "step": 8772
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.064233904371267,
+      "learning_rate": 8.458962048813634e-08,
+      "loss": 0.5279,
+      "step": 8773
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.970401937879649,
+      "learning_rate": 8.427101378876368e-08,
+      "loss": 0.488,
+      "step": 8774
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2262883058216913,
+      "learning_rate": 8.39530031384772e-08,
+      "loss": 0.4858,
+      "step": 8775
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.189202547257496,
+      "learning_rate": 8.36355885758372e-08,
+      "loss": 0.5082,
+      "step": 8776
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.242348029833922,
+      "learning_rate": 8.331877013933176e-08,
+      "loss": 0.4508,
+      "step": 8777
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.145728830037351,
+      "learning_rate": 8.300254786737627e-08,
+      "loss": 0.4704,
+      "step": 8778
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.448417594285461,
+      "learning_rate": 8.268692179831228e-08,
+      "loss": 0.5183,
+      "step": 8779
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.780379689740045,
+      "learning_rate": 8.237189197041196e-08,
+      "loss": 0.5186,
+      "step": 8780
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.626228125458384,
+      "learning_rate": 8.205745842187363e-08,
+      "loss": 0.4081,
+      "step": 8781
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.203092505429106,
+      "learning_rate": 8.174362119082291e-08,
+      "loss": 0.4591,
+      "step": 8782
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 15.64936080385862,
+      "learning_rate": 8.143038031531381e-08,
+      "loss": 0.4833,
+      "step": 8783
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.384242567896703,
+      "learning_rate": 8.111773583332871e-08,
+      "loss": 0.4723,
+      "step": 8784
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.697131346934249,
+      "learning_rate": 8.080568778277509e-08,
+      "loss": 0.506,
+      "step": 8785
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.939244620258785,
+      "learning_rate": 8.049423620149155e-08,
+      "loss": 0.4629,
+      "step": 8786
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.617577946704121,
+      "learning_rate": 8.018338112724178e-08,
+      "loss": 0.4498,
+      "step": 8787
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.011468397341661,
+      "learning_rate": 7.98731225977184e-08,
+      "loss": 0.4372,
+      "step": 8788
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.610948780601649,
+      "learning_rate": 7.956346065054132e-08,
+      "loss": 0.4503,
+      "step": 8789
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.553507683267007,
+      "learning_rate": 7.925439532325774e-08,
+      "loss": 0.4956,
+      "step": 8790
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4492662381763015,
+      "learning_rate": 7.894592665334266e-08,
+      "loss": 0.4947,
+      "step": 8791
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.032212549177937,
+      "learning_rate": 7.863805467820007e-08,
+      "loss": 0.5267,
+      "step": 8792
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.947294590930455,
+      "learning_rate": 7.833077943515954e-08,
+      "loss": 0.5206,
+      "step": 8793
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.036808274855425,
+      "learning_rate": 7.802410096147961e-08,
+      "loss": 0.5011,
+      "step": 8794
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.38568123590532016,
+      "learning_rate": 7.771801929434608e-08,
+      "loss": 0.2775,
+      "step": 8795
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4016714255495586,
+      "learning_rate": 7.74125344708726e-08,
+      "loss": 0.4773,
+      "step": 8796
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2845356413524986,
+      "learning_rate": 7.71076465281001e-08,
+      "loss": 0.4833,
+      "step": 8797
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.173126918433267,
+      "learning_rate": 7.680335550299678e-08,
+      "loss": 0.4898,
+      "step": 8798
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.819960195831742,
+      "learning_rate": 7.64996614324598e-08,
+      "loss": 0.475,
+      "step": 8799
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.41080409730483,
+      "learning_rate": 7.619656435331302e-08,
+      "loss": 0.4518,
+      "step": 8800
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.360632638113598,
+      "learning_rate": 7.589406430230706e-08,
+      "loss": 0.513,
+      "step": 8801
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.799659483479415,
+      "learning_rate": 7.559216131612256e-08,
+      "loss": 0.4806,
+      "step": 8802
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.154014223160058,
+      "learning_rate": 7.529085543136472e-08,
+      "loss": 0.4529,
+      "step": 8803
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.275179072277843,
+      "learning_rate": 7.499014668456872e-08,
+      "loss": 0.5291,
+      "step": 8804
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.605257764110895,
+      "learning_rate": 7.469003511219708e-08,
+      "loss": 0.5246,
+      "step": 8805
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.017956804345102,
+      "learning_rate": 7.439052075063847e-08,
+      "loss": 0.4946,
+      "step": 8806
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 38.768432979786375,
+      "learning_rate": 7.409160363621049e-08,
+      "loss": 0.4965,
+      "step": 8807
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2077755935276904,
+      "learning_rate": 7.379328380515805e-08,
+      "loss": 0.4775,
+      "step": 8808
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.93763815264077,
+      "learning_rate": 7.349556129365276e-08,
+      "loss": 0.4476,
+      "step": 8809
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.895008956942785,
+      "learning_rate": 7.31984361377952e-08,
+      "loss": 0.4818,
+      "step": 8810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.736914028482886,
+      "learning_rate": 7.290190837361211e-08,
+      "loss": 0.453,
+      "step": 8811
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.070089938286251,
+      "learning_rate": 7.260597803705971e-08,
+      "loss": 0.4443,
+      "step": 8812
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.883374045181473,
+      "learning_rate": 7.231064516401932e-08,
+      "loss": 0.4492,
+      "step": 8813
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.250356105467037,
+      "learning_rate": 7.201590979030116e-08,
+      "loss": 0.5124,
+      "step": 8814
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.37526350209854664,
+      "learning_rate": 7.172177195164387e-08,
+      "loss": 0.2632,
+      "step": 8815
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.633608836886533,
+      "learning_rate": 7.142823168371171e-08,
+      "loss": 0.4428,
+      "step": 8816
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5910367731324087,
+      "learning_rate": 7.113528902209787e-08,
+      "loss": 0.4871,
+      "step": 8817
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.3352487196786935,
+      "learning_rate": 7.084294400232283e-08,
+      "loss": 0.4735,
+      "step": 8818
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.39056826940605416,
+      "learning_rate": 7.055119665983379e-08,
+      "loss": 0.2729,
+      "step": 8819
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9019406435242825,
+      "learning_rate": 7.026004703000689e-08,
+      "loss": 0.4389,
+      "step": 8820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.545127147463427,
+      "learning_rate": 6.9969495148145e-08,
+      "loss": 0.5405,
+      "step": 8821
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.840989643846297,
+      "learning_rate": 6.967954104947771e-08,
+      "loss": 0.4693,
+      "step": 8822
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3911035144512888,
+      "learning_rate": 6.939018476916359e-08,
+      "loss": 0.276,
+      "step": 8823
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0928811022853955,
+      "learning_rate": 6.91014263422879e-08,
+      "loss": 0.5024,
+      "step": 8824
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0965945857019266,
+      "learning_rate": 6.881326580386372e-08,
+      "loss": 0.4533,
+      "step": 8825
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.452831845044329,
+      "learning_rate": 6.852570318883145e-08,
+      "loss": 0.4666,
+      "step": 8826
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.28942370894143,
+      "learning_rate": 6.823873853205931e-08,
+      "loss": 0.5259,
+      "step": 8827
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6947646283867703,
+      "learning_rate": 6.79523718683417e-08,
+      "loss": 0.4834,
+      "step": 8828
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6788686043287275,
+      "learning_rate": 6.766660323240303e-08,
+      "loss": 0.5065,
+      "step": 8829
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.555351629393625,
+      "learning_rate": 6.738143265889285e-08,
+      "loss": 0.4714,
+      "step": 8830
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.533905725660719,
+      "learning_rate": 6.709686018238958e-08,
+      "loss": 0.4447,
+      "step": 8831
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.203109767769848,
+      "learning_rate": 6.681288583739787e-08,
+      "loss": 0.4776,
+      "step": 8832
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.441518384053841,
+      "learning_rate": 6.652950965835181e-08,
+      "loss": 0.443,
+      "step": 8833
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.70294849605885,
+      "learning_rate": 6.624673167961004e-08,
+      "loss": 0.5064,
+      "step": 8834
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4060354851030445,
+      "learning_rate": 6.59645519354618e-08,
+      "loss": 0.5117,
+      "step": 8835
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.724030384047118,
+      "learning_rate": 6.568297046012196e-08,
+      "loss": 0.4909,
+      "step": 8836
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0516441618047505,
+      "learning_rate": 6.540198728773262e-08,
+      "loss": 0.5379,
+      "step": 8837
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 75.6689339370689,
+      "learning_rate": 6.512160245236432e-08,
+      "loss": 0.5252,
+      "step": 8838
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.214800472301041,
+      "learning_rate": 6.484181598801542e-08,
+      "loss": 0.5085,
+      "step": 8839
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.172230055337022,
+      "learning_rate": 6.456262792861046e-08,
+      "loss": 0.4835,
+      "step": 8840
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.268277568958063,
+      "learning_rate": 6.428403830800123e-08,
+      "loss": 0.4578,
+      "step": 8841
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3878354777286888,
+      "learning_rate": 6.400604715996905e-08,
+      "loss": 0.2748,
+      "step": 8842
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.487291985582479,
+      "learning_rate": 6.372865451822086e-08,
+      "loss": 0.4429,
+      "step": 8843
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2471081298800573,
+      "learning_rate": 6.345186041639029e-08,
+      "loss": 0.4878,
+      "step": 8844
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.192988584191086,
+      "learning_rate": 6.317566488804105e-08,
+      "loss": 0.4401,
+      "step": 8845
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.018654638312231,
+      "learning_rate": 6.290006796666249e-08,
+      "loss": 0.494,
+      "step": 8846
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.419857241081244,
+      "learning_rate": 6.262506968567061e-08,
+      "loss": 0.544,
+      "step": 8847
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.737064289726117,
+      "learning_rate": 6.235067007841155e-08,
+      "loss": 0.459,
+      "step": 8848
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.819035988316389,
+      "learning_rate": 6.207686917815592e-08,
+      "loss": 0.4902,
+      "step": 8849
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.702259739126452,
+      "learning_rate": 6.18036670181038e-08,
+      "loss": 0.4814,
+      "step": 8850
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.8301111145514164,
+      "learning_rate": 6.153106363138151e-08,
+      "loss": 0.4835,
+      "step": 8851
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.117827422426964,
+      "learning_rate": 6.125905905104368e-08,
+      "loss": 0.4442,
+      "step": 8852
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.541044174599181,
+      "learning_rate": 6.098765331007061e-08,
+      "loss": 0.4336,
+      "step": 8853
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.046824320185265,
+      "learning_rate": 6.071684644137265e-08,
+      "loss": 0.4759,
+      "step": 8854
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.50099805531169,
+      "learning_rate": 6.044663847778521e-08,
+      "loss": 0.4638,
+      "step": 8855
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.011763780375813,
+      "learning_rate": 6.017702945207149e-08,
+      "loss": 0.4824,
+      "step": 8856
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9670423165529907,
+      "learning_rate": 5.990801939692315e-08,
+      "loss": 0.5156,
+      "step": 8857
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3937275695382332,
+      "learning_rate": 5.963960834495908e-08,
+      "loss": 0.2688,
+      "step": 8858
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.7560220711507064,
+      "learning_rate": 5.937179632872436e-08,
+      "loss": 0.4852,
+      "step": 8859
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.028162913409219,
+      "learning_rate": 5.910458338069192e-08,
+      "loss": 0.4838,
+      "step": 8860
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.8174912850867515,
+      "learning_rate": 5.8837969533263595e-08,
+      "loss": 0.5118,
+      "step": 8861
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.191064564380355,
+      "learning_rate": 5.857195481876521e-08,
+      "loss": 0.4532,
+      "step": 8862
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.5503552426698795,
+      "learning_rate": 5.830653926945373e-08,
+      "loss": 0.5255,
+      "step": 8863
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.854245103468305,
+      "learning_rate": 5.804172291751064e-08,
+      "loss": 0.49,
+      "step": 8864
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6996669517141427,
+      "learning_rate": 5.777750579504582e-08,
+      "loss": 0.4579,
+      "step": 8865
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.164799620146662,
+      "learning_rate": 5.751388793409696e-08,
+      "loss": 0.5278,
+      "step": 8866
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4006385885612774,
+      "learning_rate": 5.725086936662905e-08,
+      "loss": 0.2623,
+      "step": 8867
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.35175609959813986,
+      "learning_rate": 5.6988450124533244e-08,
+      "loss": 0.2594,
+      "step": 8868
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9072563169523296,
+      "learning_rate": 5.672663023962854e-08,
+      "loss": 0.5212,
+      "step": 8869
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.275825825562866,
+      "learning_rate": 5.646540974366288e-08,
+      "loss": 0.5162,
+      "step": 8870
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.193871933114768,
+      "learning_rate": 5.620478866830814e-08,
+      "loss": 0.5003,
+      "step": 8871
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1396133395422625,
+      "learning_rate": 5.594476704516738e-08,
+      "loss": 0.4505,
+      "step": 8872
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.357525109319964,
+      "learning_rate": 5.568534490576816e-08,
+      "loss": 0.462,
+      "step": 8873
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.614611602161748,
+      "learning_rate": 5.542652228156698e-08,
+      "loss": 0.4867,
+      "step": 8874
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.888087543208663,
+      "learning_rate": 5.5168299203945953e-08,
+      "loss": 0.4632,
+      "step": 8875
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.243720829269814,
+      "learning_rate": 5.491067570421615e-08,
+      "loss": 0.4584,
+      "step": 8876
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.783403564549749,
+      "learning_rate": 5.465365181361537e-08,
+      "loss": 0.4583,
+      "step": 8877
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.208920656457656,
+      "learning_rate": 5.439722756330812e-08,
+      "loss": 0.5096,
+      "step": 8878
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.296734142975376,
+      "learning_rate": 5.4141402984387304e-08,
+      "loss": 0.4555,
+      "step": 8879
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.209829868122966,
+      "learning_rate": 5.388617810787256e-08,
+      "loss": 0.4845,
+      "step": 8880
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.043954713253392,
+      "learning_rate": 5.363155296471079e-08,
+      "loss": 0.4502,
+      "step": 8881
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6245716866571063,
+      "learning_rate": 5.3377527585775635e-08,
+      "loss": 0.4776,
+      "step": 8882
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.520540558377754,
+      "learning_rate": 5.3124102001868574e-08,
+      "loss": 0.4637,
+      "step": 8883
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.608542315491864,
+      "learning_rate": 5.287127624371946e-08,
+      "loss": 0.4609,
+      "step": 8884
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.427097222656795,
+      "learning_rate": 5.261905034198267e-08,
+      "loss": 0.4568,
+      "step": 8885
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.37407837989036313,
+      "learning_rate": 5.236742432724262e-08,
+      "loss": 0.2801,
+      "step": 8886
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.8002465401924646,
+      "learning_rate": 5.211639823000936e-08,
+      "loss": 0.5277,
+      "step": 8887
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3959987574430573,
+      "learning_rate": 5.1865972080720195e-08,
+      "loss": 0.2809,
+      "step": 8888
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.541949386798863,
+      "learning_rate": 5.16161459097414e-08,
+      "loss": 0.5079,
+      "step": 8889
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.978471386928452,
+      "learning_rate": 5.13669197473643e-08,
+      "loss": 0.476,
+      "step": 8890
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.592427113826931,
+      "learning_rate": 5.11182936238086e-08,
+      "loss": 0.4439,
+      "step": 8891
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.1234498482199795,
+      "learning_rate": 5.0870267569221864e-08,
+      "loss": 0.489,
+      "step": 8892
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.412279755468691,
+      "learning_rate": 5.0622841613676695e-08,
+      "loss": 0.5672,
+      "step": 8893
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8931499196852006,
+      "learning_rate": 5.0376015787175215e-08,
+      "loss": 0.4531,
+      "step": 8894
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6789490417782043,
+      "learning_rate": 5.0129790119646804e-08,
+      "loss": 0.4494,
+      "step": 8895
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.144809530358154,
+      "learning_rate": 4.988416464094481e-08,
+      "loss": 0.4782,
+      "step": 8896
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.774960074504306,
+      "learning_rate": 4.963913938085374e-08,
+      "loss": 0.4471,
+      "step": 8897
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2416158737663845,
+      "learning_rate": 4.939471436908427e-08,
+      "loss": 0.4973,
+      "step": 8898
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.19265677093083,
+      "learning_rate": 4.9150889635272145e-08,
+      "loss": 0.4711,
+      "step": 8899
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.230303306098297,
+      "learning_rate": 4.890766520898316e-08,
+      "loss": 0.4914,
+      "step": 8900
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.51597334759627,
+      "learning_rate": 4.8665041119708714e-08,
+      "loss": 0.4744,
+      "step": 8901
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.869346043941346,
+      "learning_rate": 4.8423017396868054e-08,
+      "loss": 0.478,
+      "step": 8902
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.470849514520549,
+      "learning_rate": 4.818159406980716e-08,
+      "loss": 0.5075,
+      "step": 8903
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8941355891655345,
+      "learning_rate": 4.7940771167799274e-08,
+      "loss": 0.4473,
+      "step": 8904
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.084806902252136,
+      "learning_rate": 4.770054872004548e-08,
+      "loss": 0.4743,
+      "step": 8905
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5531704584296917,
+      "learning_rate": 4.7460926755673595e-08,
+      "loss": 0.4655,
+      "step": 8906
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8372141066779046,
+      "learning_rate": 4.72219053037376e-08,
+      "loss": 0.5222,
+      "step": 8907
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.915813779666017,
+      "learning_rate": 4.698348439322098e-08,
+      "loss": 0.4368,
+      "step": 8908
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.37358929904607663,
+      "learning_rate": 4.674566405303227e-08,
+      "loss": 0.265,
+      "step": 8909
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.295385236134482,
+      "learning_rate": 4.65084443120084e-08,
+      "loss": 0.4752,
+      "step": 8910
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3932609095403718,
+      "learning_rate": 4.627182519891305e-08,
+      "loss": 0.2766,
+      "step": 8911
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.885564389150201,
+      "learning_rate": 4.6035806742436575e-08,
+      "loss": 0.4345,
+      "step": 8912
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.957126998075086,
+      "learning_rate": 4.5800388971197764e-08,
+      "loss": 0.4957,
+      "step": 8913
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.088400368598818,
+      "learning_rate": 4.5565571913741005e-08,
+      "loss": 0.4897,
+      "step": 8914
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 14.45468469106885,
+      "learning_rate": 4.533135559853963e-08,
+      "loss": 0.469,
+      "step": 8915
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.684651622703728,
+      "learning_rate": 4.5097740053993144e-08,
+      "loss": 0.4709,
+      "step": 8916
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.146662881034305,
+      "learning_rate": 4.486472530842723e-08,
+      "loss": 0.4829,
+      "step": 8917
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.170496022190099,
+      "learning_rate": 4.463231139009649e-08,
+      "loss": 0.486,
+      "step": 8918
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8309438356867016,
+      "learning_rate": 4.440049832718174e-08,
+      "loss": 0.5075,
+      "step": 8919
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.029537655557752,
+      "learning_rate": 4.416928614779104e-08,
+      "loss": 0.4938,
+      "step": 8920
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.9207247904342775,
+      "learning_rate": 4.393867487995973e-08,
+      "loss": 0.4925,
+      "step": 8921
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.121103739351841,
+      "learning_rate": 4.370866455165046e-08,
+      "loss": 0.4959,
+      "step": 8922
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.093000016070896,
+      "learning_rate": 4.347925519075258e-08,
+      "loss": 0.4741,
+      "step": 8923
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.241095675634007,
+      "learning_rate": 4.325044682508217e-08,
+      "loss": 0.459,
+      "step": 8924
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.234389896569199,
+      "learning_rate": 4.3022239482384264e-08,
+      "loss": 0.4819,
+      "step": 8925
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.456332875427677,
+      "learning_rate": 4.279463319032895e-08,
+      "loss": 0.469,
+      "step": 8926
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5263423803394343,
+      "learning_rate": 4.256762797651415e-08,
+      "loss": 0.4398,
+      "step": 8927
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6884432635454947,
+      "learning_rate": 4.234122386846562e-08,
+      "loss": 0.487,
+      "step": 8928
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3931829525138657,
+      "learning_rate": 4.2115420893635296e-08,
+      "loss": 0.2693,
+      "step": 8929
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1131088929903057,
+      "learning_rate": 4.1890219079402384e-08,
+      "loss": 0.4813,
+      "step": 8930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4032782720442117,
+      "learning_rate": 4.166561845307393e-08,
+      "loss": 0.28,
+      "step": 8931
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.266334606517385,
+      "learning_rate": 4.1441619041883155e-08,
+      "loss": 0.5335,
+      "step": 8932
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.988195505742215,
+      "learning_rate": 4.1218220872990546e-08,
+      "loss": 0.4415,
+      "step": 8933
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.561820282088418,
+      "learning_rate": 4.099542397348444e-08,
+      "loss": 0.4625,
+      "step": 8934
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.993259404696838,
+      "learning_rate": 4.0773228370379334e-08,
+      "loss": 0.47,
+      "step": 8935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.493535772527986,
+      "learning_rate": 4.055163409061758e-08,
+      "loss": 0.4301,
+      "step": 8936
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.371206563797654,
+      "learning_rate": 4.033064116106766e-08,
+      "loss": 0.4712,
+      "step": 8937
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.7508627021113,
+      "learning_rate": 4.01102496085265e-08,
+      "loss": 0.5109,
+      "step": 8938
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.116027959370696,
+      "learning_rate": 3.989045945971659e-08,
+      "loss": 0.4768,
+      "step": 8939
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.031864517583749,
+      "learning_rate": 3.967127074128885e-08,
+      "loss": 0.4497,
+      "step": 8940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.870670798514981,
+      "learning_rate": 3.9452683479819783e-08,
+      "loss": 0.5235,
+      "step": 8941
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.669079717503596,
+      "learning_rate": 3.923469770181543e-08,
+      "loss": 0.4924,
+      "step": 8942
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.840824557902444,
+      "learning_rate": 3.901731343370574e-08,
+      "loss": 0.4752,
+      "step": 8943
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.198914048328498,
+      "learning_rate": 3.8800530701849634e-08,
+      "loss": 0.4975,
+      "step": 8944
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.283380044902781,
+      "learning_rate": 3.858434953253332e-08,
+      "loss": 0.5416,
+      "step": 8945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6163486627423773,
+      "learning_rate": 3.836876995196914e-08,
+      "loss": 0.4323,
+      "step": 8946
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.898187524410166,
+      "learning_rate": 3.8153791986297315e-08,
+      "loss": 0.5059,
+      "step": 8947
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.153904747573046,
+      "learning_rate": 3.793941566158421e-08,
+      "loss": 0.4378,
+      "step": 8948
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.263944000374155,
+      "learning_rate": 3.772564100382348e-08,
+      "loss": 0.4499,
+      "step": 8949
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3748477537945716,
+      "learning_rate": 3.75124680389366e-08,
+      "loss": 0.5049,
+      "step": 8950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.650893804489973,
+      "learning_rate": 3.729989679277124e-08,
+      "loss": 0.5203,
+      "step": 8951
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.9004771055640495,
+      "learning_rate": 3.7087927291102335e-08,
+      "loss": 0.4962,
+      "step": 8952
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.845192418991381,
+      "learning_rate": 3.687655955963154e-08,
+      "loss": 0.4874,
+      "step": 8953
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0320042828444413,
+      "learning_rate": 3.666579362398892e-08,
+      "loss": 0.4924,
+      "step": 8954
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3876257097993446,
+      "learning_rate": 3.645562950973014e-08,
+      "loss": 0.274,
+      "step": 8955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.830680907502666,
+      "learning_rate": 3.624606724233759e-08,
+      "loss": 0.474,
+      "step": 8956
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8750211445167926,
+      "learning_rate": 3.603710684722317e-08,
+      "loss": 0.477,
+      "step": 8957
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4384018594720556,
+      "learning_rate": 3.5828748349722164e-08,
+      "loss": 0.4399,
+      "step": 8958
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8866366297847907,
+      "learning_rate": 3.562099177509937e-08,
+      "loss": 0.4977,
+      "step": 8959
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.183173726931099,
+      "learning_rate": 3.541383714854685e-08,
+      "loss": 0.4774,
+      "step": 8960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3939996290134116,
+      "learning_rate": 3.520728449518118e-08,
+      "loss": 0.2812,
+      "step": 8961
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7362258373971686,
+      "learning_rate": 3.5001333840049e-08,
+      "loss": 0.4829,
+      "step": 8962
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.394742418198067,
+      "learning_rate": 3.4795985208122e-08,
+      "loss": 0.4946,
+      "step": 8963
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.813467947338091,
+      "learning_rate": 3.4591238624299696e-08,
+      "loss": 0.4673,
+      "step": 8964
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4128571875559771,
+      "learning_rate": 3.438709411340779e-08,
+      "loss": 0.2873,
+      "step": 8965
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6109996961032493,
+      "learning_rate": 3.418355170019982e-08,
+      "loss": 0.4539,
+      "step": 8966
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.006168680165029,
+      "learning_rate": 3.3980611409356045e-08,
+      "loss": 0.482,
+      "step": 8967
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.357337367966314,
+      "learning_rate": 3.377827326548344e-08,
+      "loss": 0.5324,
+      "step": 8968
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3487515332388893,
+      "learning_rate": 3.357653729311683e-08,
+      "loss": 0.2769,
+      "step": 8969
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8573964308287523,
+      "learning_rate": 3.337540351671664e-08,
+      "loss": 0.4819,
+      "step": 8970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.787347183297821,
+      "learning_rate": 3.317487196067115e-08,
+      "loss": 0.4689,
+      "step": 8971
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.38172452407628,
+      "learning_rate": 3.29749426492959e-08,
+      "loss": 0.4678,
+      "step": 8972
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.925520115677739,
+      "learning_rate": 3.277561560683262e-08,
+      "loss": 0.4874,
+      "step": 8973
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5838178393483013,
+      "learning_rate": 3.257689085745086e-08,
+      "loss": 0.5307,
+      "step": 8974
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.852671349485335,
+      "learning_rate": 3.237876842524634e-08,
+      "loss": 0.4452,
+      "step": 8975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5504011313964003,
+      "learning_rate": 3.2181248334242075e-08,
+      "loss": 0.472,
+      "step": 8976
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.964082136565591,
+      "learning_rate": 3.198433060838779e-08,
+      "loss": 0.4781,
+      "step": 8977
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.623117011709825,
+      "learning_rate": 3.1788015271561054e-08,
+      "loss": 0.5186,
+      "step": 8978
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.96716442555256,
+      "learning_rate": 3.159230234756561e-08,
+      "loss": 0.4732,
+      "step": 8979
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.763557871021792,
+      "learning_rate": 3.139719186013246e-08,
+      "loss": 0.4759,
+      "step": 8980
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.14078570217295,
+      "learning_rate": 3.1202683832918804e-08,
+      "loss": 0.4973,
+      "step": 8981
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.870485741435771,
+      "learning_rate": 3.100877828950965e-08,
+      "loss": 0.4569,
+      "step": 8982
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.711700161937866,
+      "learning_rate": 3.0815475253417326e-08,
+      "loss": 0.4662,
+      "step": 8983
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6717292326252986,
+      "learning_rate": 3.062277474807917e-08,
+      "loss": 0.4587,
+      "step": 8984
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.9420981414648204,
+      "learning_rate": 3.043067679686262e-08,
+      "loss": 0.5036,
+      "step": 8985
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.7571568574338245,
+      "learning_rate": 3.0239181423058485e-08,
+      "loss": 0.5299,
+      "step": 8986
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.332418857407382,
+      "learning_rate": 3.004828864988707e-08,
+      "loss": 0.4588,
+      "step": 8987
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.176844223005225,
+      "learning_rate": 2.985799850049487e-08,
+      "loss": 0.4826,
+      "step": 8988
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3991359266530295,
+      "learning_rate": 2.9668310997955085e-08,
+      "loss": 0.2818,
+      "step": 8989
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.227217195720618,
+      "learning_rate": 2.9479226165268216e-08,
+      "loss": 0.5245,
+      "step": 8990
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.822972738555277,
+      "learning_rate": 2.9290744025360916e-08,
+      "loss": 0.4915,
+      "step": 8991
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.909598704938074,
+      "learning_rate": 2.9102864601087133e-08,
+      "loss": 0.4759,
+      "step": 8992
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4090242194278059,
+      "learning_rate": 2.8915587915228638e-08,
+      "loss": 0.2779,
+      "step": 8993
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.589954724876772,
+      "learning_rate": 2.872891399049338e-08,
+      "loss": 0.474,
+      "step": 8994
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.649616163346752,
+      "learning_rate": 2.8542842849515474e-08,
+      "loss": 0.4817,
+      "step": 8995
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8008774927955677,
+      "learning_rate": 2.835737451485687e-08,
+      "loss": 0.4403,
+      "step": 8996
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9171944199973345,
+      "learning_rate": 2.8172509009006808e-08,
+      "loss": 0.4648,
+      "step": 8997
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0157481475947936,
+      "learning_rate": 2.798824635438069e-08,
+      "loss": 0.4915,
+      "step": 8998
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.427500294420459,
+      "learning_rate": 2.780458657332008e-08,
+      "loss": 0.5063,
+      "step": 8999
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.4155221830503635,
+      "learning_rate": 2.7621529688096062e-08,
+      "loss": 0.5307,
+      "step": 9000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 9302,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 9085203986644992.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}