diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,2029 +1,3821 @@
 {
-  "best_metric": 1.5016273260116577,
-  "best_model_checkpoint": "lora_lr/mistralai/Mistral-7B-Instruct-v0.2/unaligned/checkpoint-240",
-  "epoch": 0.327575175943698,
-  "eval_steps": 10,
-  "global_step": 256,
+  "best_metric": 1.633476734161377,
+  "best_model_checkpoint": "lora_lr_pad/meta-llama/Meta-Llama-3-8B-Instruct/unaligned/checkpoint-500",
+  "epoch": 0.655150351887396,
+  "eval_steps": 20,
+  "global_step": 512,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.0,
-      "grad_norm": 1.484375,
+      "epoch": 0.0012795905310300703,
+      "grad_norm": 0.1796875,
       "learning_rate": 2.0000000000000003e-06,
-      "loss": 3.7421,
+      "loss": 2.2847,
       "step": 1
     },
     {
-      "epoch": 0.0,
-      "grad_norm": 1.4921875,
+      "epoch": 0.0025591810620601407,
+      "grad_norm": 0.1884765625,
       "learning_rate": 4.000000000000001e-06,
-      "loss": 3.6136,
+      "loss": 2.1736,
       "step": 2
     },
     {
-      "epoch": 0.0,
-      "grad_norm": 1.4140625,
+      "epoch": 0.003838771593090211,
+      "grad_norm": 0.1572265625,
       "learning_rate": 6e-06,
-      "loss": 3.5641,
+      "loss": 2.205,
       "step": 3
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 1.59375,
+      "epoch": 0.005118362124120281,
+      "grad_norm": 0.1943359375,
       "learning_rate": 8.000000000000001e-06,
-      "loss": 3.8168,
+      "loss": 2.2475,
       "step": 4
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 1.5703125,
+      "epoch": 0.006397952655150352,
+      "grad_norm": 0.189453125,
       "learning_rate": 1e-05,
-      "loss": 3.7471,
+      "loss": 2.2441,
       "step": 5
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 1.4609375,
+      "epoch": 0.007677543186180422,
+      "grad_norm": 0.18359375,
       "learning_rate": 1.2e-05,
-      "loss": 3.6939,
+      "loss": 2.2687,
       "step": 6
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 1.5546875,
+      "epoch": 0.008957133717210493,
+      "grad_norm": 0.201171875,
       "learning_rate": 1.4000000000000001e-05,
-      "loss": 3.5964,
+      "loss": 2.1679,
       "step": 7
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 1.5625,
+      "epoch": 0.010236724248240563,
+      "grad_norm": 0.1884765625,
       "learning_rate": 1.6000000000000003e-05,
-      "loss": 3.7708,
+      "loss": 2.2092,
       "step": 8
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 1.5703125,
+      "epoch": 0.011516314779270634,
+      "grad_norm": 0.1904296875,
       "learning_rate": 1.8e-05,
-      "loss": 3.7418,
+      "loss": 2.2664,
       "step": 9
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 1.609375,
+      "epoch": 0.012795905310300703,
+      "grad_norm": 0.189453125,
       "learning_rate": 2e-05,
-      "loss": 3.8394,
+      "loss": 2.2509,
       "step": 10
     },
     {
-      "epoch": 0.01,
-      "eval_loss": 3.672560214996338,
-      "eval_runtime": 108.3497,
-      "eval_samples_per_second": 46.147,
-      "eval_steps_per_second": 1.449,
-      "step": 10
-    },
-    {
-      "epoch": 0.01,
-      "grad_norm": 1.5078125,
+      "epoch": 0.014075495841330775,
+      "grad_norm": 0.1748046875,
       "learning_rate": 2.2000000000000003e-05,
-      "loss": 3.6075,
+      "loss": 2.1772,
       "step": 11
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 1.4296875,
+      "epoch": 0.015355086372360844,
+      "grad_norm": 0.1796875,
       "learning_rate": 2.4e-05,
-      "loss": 3.4171,
+      "loss": 2.2352,
       "step": 12
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 1.6484375,
+      "epoch": 0.016634676903390915,
+      "grad_norm": 0.208984375,
       "learning_rate": 2.6000000000000002e-05,
-      "loss": 3.7848,
+      "loss": 2.2991,
       "step": 13
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 1.6328125,
+      "epoch": 0.017914267434420986,
+      "grad_norm": 0.201171875,
       "learning_rate": 2.8000000000000003e-05,
-      "loss": 3.7797,
+      "loss": 2.3595,
       "step": 14
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 1.609375,
+      "epoch": 0.019193857965451054,
+      "grad_norm": 0.1943359375,
       "learning_rate": 3e-05,
-      "loss": 3.7132,
+      "loss": 2.2419,
       "step": 15
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 1.5390625,
+      "epoch": 0.020473448496481125,
+      "grad_norm": 0.201171875,
       "learning_rate": 3.2000000000000005e-05,
-      "loss": 3.5353,
+      "loss": 2.2519,
       "step": 16
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 1.828125,
+      "epoch": 0.021753039027511197,
+      "grad_norm": 0.2275390625,
       "learning_rate": 3.4000000000000007e-05,
-      "loss": 3.687,
+      "loss": 2.1917,
       "step": 17
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 1.65625,
+      "epoch": 0.023032629558541268,
+      "grad_norm": 0.2373046875,
       "learning_rate": 3.6e-05,
-      "loss": 3.4312,
+      "loss": 2.26,
       "step": 18
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 1.8203125,
+      "epoch": 0.02431222008957134,
+      "grad_norm": 0.236328125,
       "learning_rate": 3.8e-05,
-      "loss": 3.4971,
+      "loss": 2.2416,
       "step": 19
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 1.875,
+      "epoch": 0.025591810620601407,
+      "grad_norm": 0.25,
       "learning_rate": 4e-05,
-      "loss": 3.3386,
+      "loss": 2.2174,
       "step": 20
     },
     {
-      "epoch": 0.03,
-      "eval_loss": 3.376652240753174,
-      "eval_runtime": 108.6461,
-      "eval_samples_per_second": 46.021,
-      "eval_steps_per_second": 1.445,
+      "epoch": 0.025591810620601407,
+      "eval_loss": 2.2057077884674072,
+      "eval_runtime": 103.2988,
+      "eval_samples_per_second": 48.403,
+      "eval_steps_per_second": 1.52,
       "step": 20
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 1.9921875,
+      "epoch": 0.026871401151631478,
+      "grad_norm": 0.283203125,
       "learning_rate": 4.2e-05,
-      "loss": 3.3568,
+      "loss": 2.1936,
       "step": 21
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 2.109375,
+      "epoch": 0.02815099168266155,
+      "grad_norm": 0.294921875,
       "learning_rate": 4.4000000000000006e-05,
-      "loss": 3.3976,
+      "loss": 2.2211,
       "step": 22
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 2.125,
+      "epoch": 0.02943058221369162,
+      "grad_norm": 0.345703125,
       "learning_rate": 4.600000000000001e-05,
-      "loss": 3.3644,
+      "loss": 2.2378,
       "step": 23
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 2.28125,
+      "epoch": 0.030710172744721688,
+      "grad_norm": 0.353515625,
       "learning_rate": 4.8e-05,
-      "loss": 3.1825,
+      "loss": 2.2521,
       "step": 24
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 2.421875,
+      "epoch": 0.03198976327575176,
+      "grad_norm": 0.37109375,
       "learning_rate": 5e-05,
-      "loss": 3.2047,
+      "loss": 2.1821,
       "step": 25
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 2.359375,
+      "epoch": 0.03326935380678183,
+      "grad_norm": 0.34765625,
       "learning_rate": 5.2000000000000004e-05,
-      "loss": 3.098,
+      "loss": 2.2399,
       "step": 26
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 2.390625,
+      "epoch": 0.0345489443378119,
+      "grad_norm": 0.400390625,
       "learning_rate": 5.4000000000000005e-05,
-      "loss": 2.9831,
+      "loss": 2.1444,
       "step": 27
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 2.359375,
+      "epoch": 0.03582853486884197,
+      "grad_norm": 0.451171875,
       "learning_rate": 5.6000000000000006e-05,
-      "loss": 2.9573,
+      "loss": 2.1987,
       "step": 28
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 2.609375,
+      "epoch": 0.037108125399872044,
+      "grad_norm": 0.427734375,
       "learning_rate": 5.8e-05,
-      "loss": 2.9497,
+      "loss": 2.159,
       "step": 29
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 2.125,
+      "epoch": 0.03838771593090211,
+      "grad_norm": 0.404296875,
       "learning_rate": 6e-05,
-      "loss": 2.5156,
-      "step": 30
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 2.578493356704712,
-      "eval_runtime": 109.4296,
-      "eval_samples_per_second": 45.691,
-      "eval_steps_per_second": 1.435,
+      "loss": 2.0387,
       "step": 30
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 2.25,
+      "epoch": 0.03966730646193218,
+      "grad_norm": 0.470703125,
       "learning_rate": 6.2e-05,
-      "loss": 2.6802,
+      "loss": 2.144,
       "step": 31
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 1.984375,
+      "epoch": 0.04094689699296225,
+      "grad_norm": 0.431640625,
       "learning_rate": 6.400000000000001e-05,
-      "loss": 2.4563,
+      "loss": 2.0673,
       "step": 32
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 2.046875,
+      "epoch": 0.04222648752399232,
+      "grad_norm": 0.44921875,
       "learning_rate": 6.6e-05,
-      "loss": 2.4709,
+      "loss": 2.0571,
       "step": 33
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 2.046875,
+      "epoch": 0.04350607805502239,
+      "grad_norm": 0.41796875,
       "learning_rate": 6.800000000000001e-05,
-      "loss": 2.4096,
+      "loss": 2.0175,
       "step": 34
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 1.75,
+      "epoch": 0.044785668586052464,
+      "grad_norm": 0.42578125,
       "learning_rate": 7e-05,
-      "loss": 2.2541,
+      "loss": 2.039,
       "step": 35
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 1.78125,
+      "epoch": 0.046065259117082535,
+      "grad_norm": 0.390625,
       "learning_rate": 7.2e-05,
-      "loss": 2.1879,
+      "loss": 1.9393,
       "step": 36
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 1.6484375,
+      "epoch": 0.04734484964811261,
+      "grad_norm": 0.423828125,
       "learning_rate": 7.4e-05,
-      "loss": 2.2291,
+      "loss": 2.0395,
       "step": 37
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 1.4140625,
+      "epoch": 0.04862444017914268,
+      "grad_norm": 0.357421875,
       "learning_rate": 7.6e-05,
-      "loss": 2.028,
+      "loss": 1.9196,
       "step": 38
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 1.1640625,
+      "epoch": 0.04990403071017274,
+      "grad_norm": 0.310546875,
       "learning_rate": 7.800000000000001e-05,
-      "loss": 1.9569,
+      "loss": 1.8573,
       "step": 39
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 0.9296875,
+      "epoch": 0.05118362124120281,
+      "grad_norm": 0.283203125,
       "learning_rate": 8e-05,
-      "loss": 1.8834,
+      "loss": 1.8748,
       "step": 40
     },
     {
-      "epoch": 0.05,
-      "eval_loss": 1.834492564201355,
-      "eval_runtime": 109.4942,
-      "eval_samples_per_second": 45.665,
-      "eval_steps_per_second": 1.434,
+      "epoch": 0.05118362124120281,
+      "eval_loss": 1.8459974527359009,
+      "eval_runtime": 103.1757,
+      "eval_samples_per_second": 48.461,
+      "eval_steps_per_second": 1.522,
       "step": 40
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 0.765625,
+      "epoch": 0.052463211772232884,
+      "grad_norm": 0.287109375,
       "learning_rate": 8.2e-05,
-      "loss": 1.9118,
+      "loss": 1.9095,
       "step": 41
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 0.640625,
+      "epoch": 0.053742802303262956,
+      "grad_norm": 0.3203125,
       "learning_rate": 8.4e-05,
-      "loss": 1.8096,
+      "loss": 1.854,
       "step": 42
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.515625,
+      "epoch": 0.05502239283429303,
+      "grad_norm": 0.341796875,
       "learning_rate": 8.6e-05,
-      "loss": 1.8097,
+      "loss": 1.8241,
       "step": 43
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.49609375,
+      "epoch": 0.0563019833653231,
+      "grad_norm": 0.3359375,
       "learning_rate": 8.800000000000001e-05,
-      "loss": 1.7625,
+      "loss": 1.8081,
       "step": 44
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.578125,
+      "epoch": 0.05758157389635317,
+      "grad_norm": 0.2890625,
       "learning_rate": 9e-05,
-      "loss": 1.7719,
+      "loss": 1.7557,
       "step": 45
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.490234375,
+      "epoch": 0.05886116442738324,
+      "grad_norm": 0.25,
       "learning_rate": 9.200000000000001e-05,
-      "loss": 1.7783,
+      "loss": 1.7961,
       "step": 46
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.5625,
+      "epoch": 0.060140754958413305,
+      "grad_norm": 0.23828125,
       "learning_rate": 9.4e-05,
-      "loss": 1.7888,
+      "loss": 1.812,
       "step": 47
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.55078125,
+      "epoch": 0.061420345489443376,
+      "grad_norm": 0.26953125,
       "learning_rate": 9.6e-05,
-      "loss": 1.7738,
+      "loss": 1.8099,
       "step": 48
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.5859375,
+      "epoch": 0.06269993602047345,
+      "grad_norm": 0.2451171875,
       "learning_rate": 9.8e-05,
-      "loss": 1.7428,
+      "loss": 1.7337,
       "step": 49
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 0.55078125,
+      "epoch": 0.06397952655150352,
+      "grad_norm": 0.279296875,
       "learning_rate": 0.0001,
-      "loss": 1.7651,
-      "step": 50
-    },
-    {
-      "epoch": 0.06,
-      "eval_loss": 1.744554042816162,
-      "eval_runtime": 109.7416,
-      "eval_samples_per_second": 45.562,
-      "eval_steps_per_second": 1.431,
+      "loss": 1.7997,
       "step": 50
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.58984375,
+      "epoch": 0.06525911708253358,
+      "grad_norm": 0.2099609375,
       "learning_rate": 0.00010200000000000001,
-      "loss": 1.8558,
+      "loss": 1.8664,
       "step": 51
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.6171875,
+      "epoch": 0.06653870761356366,
+      "grad_norm": 0.2109375,
       "learning_rate": 0.00010400000000000001,
-      "loss": 1.7136,
+      "loss": 1.7195,
       "step": 52
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.6328125,
+      "epoch": 0.06781829814459372,
+      "grad_norm": 0.185546875,
       "learning_rate": 0.00010600000000000002,
-      "loss": 1.6953,
+      "loss": 1.7255,
       "step": 53
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.79296875,
+      "epoch": 0.0690978886756238,
+      "grad_norm": 0.1904296875,
       "learning_rate": 0.00010800000000000001,
-      "loss": 1.7225,
+      "loss": 1.7667,
       "step": 54
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.83203125,
+      "epoch": 0.07037747920665387,
+      "grad_norm": 0.1728515625,
       "learning_rate": 0.00011000000000000002,
-      "loss": 1.6275,
+      "loss": 1.6752,
       "step": 55
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.84375,
+      "epoch": 0.07165706973768395,
+      "grad_norm": 0.173828125,
       "learning_rate": 0.00011200000000000001,
-      "loss": 1.6724,
+      "loss": 1.7413,
       "step": 56
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.5546875,
+      "epoch": 0.07293666026871401,
+      "grad_norm": 0.177734375,
       "learning_rate": 0.00011399999999999999,
-      "loss": 1.627,
+      "loss": 1.7498,
       "step": 57
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.388671875,
+      "epoch": 0.07421625079974409,
+      "grad_norm": 0.1494140625,
       "learning_rate": 0.000116,
-      "loss": 1.6624,
+      "loss": 1.7908,
       "step": 58
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 0.34375,
+      "epoch": 0.07549584133077415,
+      "grad_norm": 0.1533203125,
       "learning_rate": 0.000118,
-      "loss": 1.592,
+      "loss": 1.7423,
       "step": 59
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 0.306640625,
+      "epoch": 0.07677543186180422,
+      "grad_norm": 0.154296875,
       "learning_rate": 0.00012,
-      "loss": 1.638,
+      "loss": 1.7723,
       "step": 60
     },
     {
-      "epoch": 0.08,
-      "eval_loss": 1.5958868265151978,
-      "eval_runtime": 109.6677,
-      "eval_samples_per_second": 45.592,
-      "eval_steps_per_second": 1.432,
+      "epoch": 0.07677543186180422,
+      "eval_loss": 1.7295056581497192,
+      "eval_runtime": 103.1327,
+      "eval_samples_per_second": 48.481,
+      "eval_steps_per_second": 1.522,
       "step": 60
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 0.302734375,
+      "epoch": 0.0780550223928343,
+      "grad_norm": 0.1552734375,
       "learning_rate": 0.000122,
-      "loss": 1.5942,
+      "loss": 1.747,
       "step": 61
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 0.30078125,
+      "epoch": 0.07933461292386436,
+      "grad_norm": 0.1396484375,
       "learning_rate": 0.000124,
-      "loss": 1.6169,
+      "loss": 1.7531,
       "step": 62
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 0.296875,
+      "epoch": 0.08061420345489444,
+      "grad_norm": 0.1650390625,
       "learning_rate": 0.000126,
-      "loss": 1.6183,
+      "loss": 1.7453,
       "step": 63
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 0.3046875,
+      "epoch": 0.0818937939859245,
+      "grad_norm": 0.1484375,
       "learning_rate": 0.00012800000000000002,
-      "loss": 1.5629,
+      "loss": 1.6982,
       "step": 64
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 0.3203125,
+      "epoch": 0.08317338451695458,
+      "grad_norm": 0.140625,
       "learning_rate": 0.00013000000000000002,
-      "loss": 1.6236,
+      "loss": 1.7718,
       "step": 65
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 0.267578125,
+      "epoch": 0.08445297504798464,
+      "grad_norm": 0.1630859375,
       "learning_rate": 0.000132,
-      "loss": 1.6701,
+      "loss": 1.815,
       "step": 66
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.294921875,
+      "epoch": 0.08573256557901472,
+      "grad_norm": 0.1552734375,
       "learning_rate": 0.000134,
-      "loss": 1.6067,
+      "loss": 1.7436,
       "step": 67
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.30859375,
+      "epoch": 0.08701215611004479,
+      "grad_norm": 0.1494140625,
       "learning_rate": 0.00013600000000000003,
-      "loss": 1.5935,
+      "loss": 1.7252,
       "step": 68
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.27734375,
+      "epoch": 0.08829174664107485,
+      "grad_norm": 0.146484375,
       "learning_rate": 0.000138,
-      "loss": 1.578,
+      "loss": 1.6938,
       "step": 69
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.26953125,
+      "epoch": 0.08957133717210493,
+      "grad_norm": 0.138671875,
       "learning_rate": 0.00014,
-      "loss": 1.592,
-      "step": 70
-    },
-    {
-      "epoch": 0.09,
-      "eval_loss": 1.571932077407837,
-      "eval_runtime": 109.6303,
-      "eval_samples_per_second": 45.608,
-      "eval_steps_per_second": 1.432,
+      "loss": 1.688,
       "step": 70
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.28125,
+      "epoch": 0.09085092770313499,
+      "grad_norm": 0.1484375,
       "learning_rate": 0.000142,
-      "loss": 1.5923,
+      "loss": 1.7185,
       "step": 71
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.283203125,
+      "epoch": 0.09213051823416507,
+      "grad_norm": 0.1357421875,
       "learning_rate": 0.000144,
-      "loss": 1.5838,
+      "loss": 1.7416,
       "step": 72
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.26171875,
+      "epoch": 0.09341010876519514,
+      "grad_norm": 0.12890625,
       "learning_rate": 0.000146,
-      "loss": 1.5646,
+      "loss": 1.6831,
       "step": 73
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 0.255859375,
+      "epoch": 0.09468969929622521,
+      "grad_norm": 0.13671875,
       "learning_rate": 0.000148,
-      "loss": 1.6265,
+      "loss": 1.7872,
       "step": 74
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.271484375,
+      "epoch": 0.09596928982725528,
+      "grad_norm": 0.142578125,
       "learning_rate": 0.00015000000000000001,
-      "loss": 1.622,
+      "loss": 1.7634,
       "step": 75
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.2373046875,
+      "epoch": 0.09724888035828536,
+      "grad_norm": 0.1298828125,
       "learning_rate": 0.000152,
-      "loss": 1.5504,
+      "loss": 1.7029,
       "step": 76
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.267578125,
+      "epoch": 0.09852847088931542,
+      "grad_norm": 0.1376953125,
       "learning_rate": 0.000154,
-      "loss": 1.6184,
+      "loss": 1.7621,
       "step": 77
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.228515625,
+      "epoch": 0.09980806142034548,
+      "grad_norm": 0.1708984375,
       "learning_rate": 0.00015600000000000002,
-      "loss": 1.5829,
+      "loss": 1.7248,
       "step": 78
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.2451171875,
+      "epoch": 0.10108765195137556,
+      "grad_norm": 0.1611328125,
       "learning_rate": 0.00015800000000000002,
-      "loss": 1.5305,
+      "loss": 1.664,
       "step": 79
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.25,
+      "epoch": 0.10236724248240563,
+      "grad_norm": 0.1513671875,
       "learning_rate": 0.00016,
-      "loss": 1.554,
+      "loss": 1.694,
       "step": 80
     },
     {
-      "epoch": 0.1,
-      "eval_loss": 1.5540077686309814,
-      "eval_runtime": 109.5859,
-      "eval_samples_per_second": 45.626,
-      "eval_steps_per_second": 1.433,
+      "epoch": 0.10236724248240563,
+      "eval_loss": 1.6964631080627441,
+      "eval_runtime": 103.1498,
+      "eval_samples_per_second": 48.473,
+      "eval_steps_per_second": 1.522,
       "step": 80
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.24609375,
+      "epoch": 0.1036468330134357,
+      "grad_norm": 0.14453125,
       "learning_rate": 0.000162,
-      "loss": 1.5691,
+      "loss": 1.6872,
       "step": 81
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 0.2470703125,
+      "epoch": 0.10492642354446577,
+      "grad_norm": 0.1474609375,
       "learning_rate": 0.000164,
-      "loss": 1.542,
+      "loss": 1.684,
       "step": 82
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.2431640625,
+      "epoch": 0.10620601407549585,
+      "grad_norm": 0.1826171875,
       "learning_rate": 0.000166,
-      "loss": 1.5844,
+      "loss": 1.7423,
       "step": 83
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.2333984375,
+      "epoch": 0.10748560460652591,
+      "grad_norm": 0.169921875,
       "learning_rate": 0.000168,
-      "loss": 1.5429,
+      "loss": 1.7016,
       "step": 84
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.2255859375,
+      "epoch": 0.10876519513755598,
+      "grad_norm": 0.150390625,
       "learning_rate": 0.00017,
-      "loss": 1.5836,
+      "loss": 1.7208,
       "step": 85
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.25390625,
+      "epoch": 0.11004478566858605,
+      "grad_norm": 0.16015625,
       "learning_rate": 0.000172,
-      "loss": 1.5227,
+      "loss": 1.6444,
       "step": 86
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.2412109375,
+      "epoch": 0.11132437619961612,
+      "grad_norm": 0.1533203125,
       "learning_rate": 0.000174,
-      "loss": 1.5495,
+      "loss": 1.7113,
       "step": 87
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.21484375,
+      "epoch": 0.1126039667306462,
+      "grad_norm": 0.1513671875,
       "learning_rate": 0.00017600000000000002,
-      "loss": 1.6124,
+      "loss": 1.7502,
       "step": 88
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 0.2470703125,
+      "epoch": 0.11388355726167626,
+      "grad_norm": 0.2080078125,
       "learning_rate": 0.00017800000000000002,
-      "loss": 1.5469,
+      "loss": 1.7002,
       "step": 89
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.2470703125,
+      "epoch": 0.11516314779270634,
+      "grad_norm": 0.1796875,
       "learning_rate": 0.00018,
-      "loss": 1.5215,
-      "step": 90
-    },
-    {
-      "epoch": 0.12,
-      "eval_loss": 1.5420269966125488,
-      "eval_runtime": 109.6203,
-      "eval_samples_per_second": 45.612,
-      "eval_steps_per_second": 1.432,
+      "loss": 1.6666,
       "step": 90
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.22265625,
+      "epoch": 0.1164427383237364,
+      "grad_norm": 0.1484375,
       "learning_rate": 0.000182,
-      "loss": 1.5654,
+      "loss": 1.7207,
       "step": 91
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.2392578125,
+      "epoch": 0.11772232885476648,
+      "grad_norm": 0.162109375,
       "learning_rate": 0.00018400000000000003,
-      "loss": 1.494,
+      "loss": 1.6261,
       "step": 92
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.216796875,
+      "epoch": 0.11900191938579655,
+      "grad_norm": 0.1787109375,
       "learning_rate": 0.00018600000000000002,
-      "loss": 1.6285,
+      "loss": 1.7819,
       "step": 93
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.2451171875,
+      "epoch": 0.12028150991682661,
+      "grad_norm": 0.1572265625,
       "learning_rate": 0.000188,
-      "loss": 1.5177,
+      "loss": 1.6319,
       "step": 94
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.2177734375,
+      "epoch": 0.12156110044785669,
+      "grad_norm": 0.171875,
       "learning_rate": 0.00019,
-      "loss": 1.5659,
+      "loss": 1.6837,
       "step": 95
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.220703125,
+      "epoch": 0.12284069097888675,
+      "grad_norm": 0.1884765625,
       "learning_rate": 0.000192,
-      "loss": 1.6005,
+      "loss": 1.7428,
       "step": 96
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 0.251953125,
+      "epoch": 0.12412028150991683,
+      "grad_norm": 0.1748046875,
       "learning_rate": 0.000194,
-      "loss": 1.5713,
+      "loss": 1.686,
       "step": 97
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.2294921875,
+      "epoch": 0.1253998720409469,
+      "grad_norm": 0.1513671875,
       "learning_rate": 0.000196,
-      "loss": 1.4815,
+      "loss": 1.6477,
       "step": 98
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.216796875,
+      "epoch": 0.12667946257197696,
+      "grad_norm": 0.1904296875,
       "learning_rate": 0.00019800000000000002,
-      "loss": 1.489,
+      "loss": 1.6311,
       "step": 99
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.216796875,
+      "epoch": 0.12795905310300704,
+      "grad_norm": 0.1923828125,
       "learning_rate": 0.0002,
-      "loss": 1.497,
+      "loss": 1.6155,
       "step": 100
     },
     {
-      "epoch": 0.13,
-      "eval_loss": 1.5331306457519531,
-      "eval_runtime": 108.6942,
-      "eval_samples_per_second": 46.001,
-      "eval_steps_per_second": 1.444,
+      "epoch": 0.12795905310300704,
+      "eval_loss": 1.6780136823654175,
+      "eval_runtime": 103.1139,
+      "eval_samples_per_second": 48.49,
+      "eval_steps_per_second": 1.523,
       "step": 100
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.23046875,
-      "learning_rate": 0.00019871794871794874,
-      "loss": 1.4933,
+      "epoch": 0.12923864363403711,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019951456310679614,
+      "loss": 1.6283,
       "step": 101
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.23828125,
-      "learning_rate": 0.00019743589743589744,
-      "loss": 1.5385,
+      "epoch": 0.13051823416506717,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019902912621359224,
+      "loss": 1.6891,
       "step": 102
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.2431640625,
-      "learning_rate": 0.00019615384615384615,
-      "loss": 1.5745,
+      "epoch": 0.13179782469609724,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00019854368932038837,
+      "loss": 1.738,
       "step": 103
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.00019487179487179487,
-      "loss": 1.5485,
+      "epoch": 0.13307741522712732,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019805825242718447,
+      "loss": 1.7,
       "step": 104
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 0.248046875,
-      "learning_rate": 0.0001935897435897436,
-      "loss": 1.5391,
+      "epoch": 0.1343570057581574,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001975728155339806,
+      "loss": 1.6759,
       "step": 105
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.23046875,
-      "learning_rate": 0.00019230769230769233,
-      "loss": 1.5649,
+      "epoch": 0.13563659628918745,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0001970873786407767,
+      "loss": 1.7018,
       "step": 106
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.212890625,
-      "learning_rate": 0.00019102564102564104,
-      "loss": 1.5543,
+      "epoch": 0.13691618682021753,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00019660194174757283,
+      "loss": 1.6875,
       "step": 107
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.23046875,
-      "learning_rate": 0.00018974358974358974,
-      "loss": 1.5377,
+      "epoch": 0.1381957773512476,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019611650485436895,
+      "loss": 1.6614,
       "step": 108
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.00018846153846153847,
-      "loss": 1.534,
+      "epoch": 0.13947536788227768,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019563106796116505,
+      "loss": 1.6804,
       "step": 109
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.0001871794871794872,
-      "loss": 1.5583,
-      "step": 110
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 1.526591181755066,
-      "eval_runtime": 108.4319,
-      "eval_samples_per_second": 46.112,
-      "eval_steps_per_second": 1.448,
+      "epoch": 0.14075495841330773,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019514563106796118,
+      "loss": 1.7319,
       "step": 110
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.234375,
-      "learning_rate": 0.0001858974358974359,
-      "loss": 1.4938,
+      "epoch": 0.1420345489443378,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00019466019417475728,
+      "loss": 1.6407,
       "step": 111
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.248046875,
-      "learning_rate": 0.00018461538461538463,
-      "loss": 1.5327,
+      "epoch": 0.1433141394753679,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001941747572815534,
+      "loss": 1.6765,
       "step": 112
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 0.2294921875,
-      "learning_rate": 0.00018333333333333334,
-      "loss": 1.5475,
+      "epoch": 0.14459373000639794,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019368932038834954,
+      "loss": 1.6793,
       "step": 113
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.2431640625,
-      "learning_rate": 0.00018205128205128207,
-      "loss": 1.5718,
+      "epoch": 0.14587332053742802,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019320388349514564,
+      "loss": 1.685,
       "step": 114
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.23046875,
-      "learning_rate": 0.00018076923076923077,
-      "loss": 1.53,
+      "epoch": 0.1471529110684581,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00019271844660194177,
+      "loss": 1.6685,
       "step": 115
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.2177734375,
-      "learning_rate": 0.0001794871794871795,
-      "loss": 1.5675,
+      "epoch": 0.14843250159948818,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019223300970873787,
+      "loss": 1.689,
       "step": 116
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.2431640625,
-      "learning_rate": 0.00017820512820512823,
-      "loss": 1.6159,
+      "epoch": 0.14971209213051823,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.000191747572815534,
+      "loss": 1.7595,
       "step": 117
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.00017692307692307693,
-      "loss": 1.5419,
+      "epoch": 0.1509916826615483,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001912621359223301,
+      "loss": 1.6845,
       "step": 118
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.2294921875,
-      "learning_rate": 0.00017564102564102566,
-      "loss": 1.4955,
+      "epoch": 0.15227127319257838,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00019077669902912623,
+      "loss": 1.6387,
       "step": 119
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 0.00017435897435897436,
-      "loss": 1.5639,
+      "epoch": 0.15355086372360843,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019029126213592236,
+      "loss": 1.7231,
       "step": 120
     },
     {
-      "epoch": 0.15,
-      "eval_loss": 1.521650791168213,
-      "eval_runtime": 108.8276,
-      "eval_samples_per_second": 45.944,
-      "eval_steps_per_second": 1.443,
+      "epoch": 0.15355086372360843,
+      "eval_loss": 1.667124629020691,
+      "eval_runtime": 103.1502,
+      "eval_samples_per_second": 48.473,
+      "eval_steps_per_second": 1.522,
       "step": 120
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 0.2119140625,
-      "learning_rate": 0.0001730769230769231,
-      "loss": 1.5569,
+      "epoch": 0.1548304542546385,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00018980582524271846,
+      "loss": 1.7072,
       "step": 121
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.24609375,
-      "learning_rate": 0.0001717948717948718,
-      "loss": 1.527,
+      "epoch": 0.1561100447856686,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00018932038834951458,
+      "loss": 1.6321,
       "step": 122
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 0.00017051282051282053,
-      "loss": 1.538,
+      "epoch": 0.15738963531669867,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00018883495145631069,
+      "loss": 1.6628,
       "step": 123
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.00016923076923076923,
-      "loss": 1.5353,
+      "epoch": 0.15866922584772872,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00018834951456310681,
+      "loss": 1.6786,
       "step": 124
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.00016794871794871796,
-      "loss": 1.5495,
+      "epoch": 0.1599488163787588,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00018786407766990291,
+      "loss": 1.6748,
       "step": 125
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.2421875,
-      "learning_rate": 0.0001666666666666667,
-      "loss": 1.5366,
+      "epoch": 0.16122840690978887,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00018737864077669904,
+      "loss": 1.6798,
       "step": 126
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.228515625,
-      "learning_rate": 0.0001653846153846154,
-      "loss": 1.5695,
+      "epoch": 0.16250799744081892,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00018689320388349517,
+      "loss": 1.7003,
       "step": 127
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.0001641025641025641,
-      "loss": 1.5112,
+      "epoch": 0.163787587971849,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00018640776699029127,
+      "loss": 1.6482,
       "step": 128
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.234375,
-      "learning_rate": 0.00016282051282051282,
-      "loss": 1.5919,
+      "epoch": 0.16506717850287908,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001859223300970874,
+      "loss": 1.7206,
       "step": 129
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.00016153846153846155,
-      "loss": 1.5785,
-      "step": 130
-    },
-    {
-      "epoch": 0.17,
-      "eval_loss": 1.517764925956726,
-      "eval_runtime": 109.0406,
-      "eval_samples_per_second": 45.854,
-      "eval_steps_per_second": 1.44,
+      "epoch": 0.16634676903390916,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0001854368932038835,
+      "loss": 1.7277,
       "step": 130
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.248046875,
-      "learning_rate": 0.00016025641025641028,
-      "loss": 1.5056,
+      "epoch": 0.1676263595649392,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00018495145631067963,
+      "loss": 1.633,
       "step": 131
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.00015897435897435896,
-      "loss": 1.5427,
+      "epoch": 0.1689059500959693,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00018446601941747576,
+      "loss": 1.7117,
       "step": 132
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.23046875,
-      "learning_rate": 0.0001576923076923077,
-      "loss": 1.5357,
+      "epoch": 0.17018554062699937,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00018398058252427186,
+      "loss": 1.7057,
       "step": 133
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.21875,
-      "learning_rate": 0.00015641025641025642,
-      "loss": 1.5671,
+      "epoch": 0.17146513115802944,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00018349514563106799,
+      "loss": 1.7182,
       "step": 134
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.2421875,
-      "learning_rate": 0.00015512820512820515,
-      "loss": 1.5247,
+      "epoch": 0.1727447216890595,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001830097087378641,
+      "loss": 1.68,
       "step": 135
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 0.00015384615384615385,
-      "loss": 1.557,
+      "epoch": 0.17402431222008957,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00018252427184466022,
+      "loss": 1.6948,
       "step": 136
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.00015256410256410255,
-      "loss": 1.5042,
+      "epoch": 0.17530390275111965,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00018203883495145632,
+      "loss": 1.6332,
       "step": 137
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 0.2373046875,
-      "learning_rate": 0.00015128205128205128,
-      "loss": 1.5061,
+      "epoch": 0.1765834932821497,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00018155339805825244,
+      "loss": 1.6327,
       "step": 138
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 0.26953125,
-      "learning_rate": 0.00015000000000000001,
-      "loss": 1.5608,
+      "epoch": 0.17786308381317978,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00018106796116504857,
+      "loss": 1.7119,
       "step": 139
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 0.25,
-      "learning_rate": 0.00014871794871794872,
-      "loss": 1.5235,
+      "epoch": 0.17914267434420986,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00018058252427184467,
+      "loss": 1.6484,
       "step": 140
     },
     {
-      "epoch": 0.18,
-      "eval_loss": 1.5145833492279053,
-      "eval_runtime": 109.0187,
-      "eval_samples_per_second": 45.864,
-      "eval_steps_per_second": 1.44,
+      "epoch": 0.17914267434420986,
+      "eval_loss": 1.6604257822036743,
+      "eval_runtime": 103.1306,
+      "eval_samples_per_second": 48.482,
+      "eval_steps_per_second": 1.522,
       "step": 140
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.00014743589743589745,
-      "loss": 1.5099,
+      "epoch": 0.18042226487523993,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0001800970873786408,
+      "loss": 1.6833,
       "step": 141
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 0.228515625,
-      "learning_rate": 0.00014615384615384615,
-      "loss": 1.5413,
+      "epoch": 0.18170185540626999,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001796116504854369,
+      "loss": 1.6802,
       "step": 142
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 0.2373046875,
-      "learning_rate": 0.00014487179487179488,
-      "loss": 1.5383,
+      "epoch": 0.18298144593730006,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00017912621359223303,
+      "loss": 1.6759,
       "step": 143
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 0.234375,
-      "learning_rate": 0.0001435897435897436,
-      "loss": 1.5671,
+      "epoch": 0.18426103646833014,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00017864077669902913,
+      "loss": 1.7033,
       "step": 144
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.244140625,
-      "learning_rate": 0.0001423076923076923,
-      "loss": 1.5102,
+      "epoch": 0.1855406269993602,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00017815533980582526,
+      "loss": 1.6981,
       "step": 145
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.00014102564102564104,
-      "loss": 1.545,
+      "epoch": 0.18682021753039027,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001776699029126214,
+      "loss": 1.7084,
       "step": 146
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.228515625,
-      "learning_rate": 0.00013974358974358974,
-      "loss": 1.5223,
+      "epoch": 0.18809980806142035,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001771844660194175,
+      "loss": 1.6641,
       "step": 147
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.2275390625,
-      "learning_rate": 0.00013846153846153847,
-      "loss": 1.433,
+      "epoch": 0.18937939859245043,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00017669902912621362,
+      "loss": 1.5855,
       "step": 148
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.00013717948717948718,
-      "loss": 1.554,
+      "epoch": 0.19065898912348048,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00017621359223300972,
+      "loss": 1.7034,
       "step": 149
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.0001358974358974359,
-      "loss": 1.5447,
-      "step": 150
-    },
-    {
-      "epoch": 0.19,
-      "eval_loss": 1.5118788480758667,
-      "eval_runtime": 109.1366,
-      "eval_samples_per_second": 45.814,
-      "eval_steps_per_second": 1.439,
+      "epoch": 0.19193857965451055,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017572815533980585,
+      "loss": 1.7136,
       "step": 150
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.259765625,
-      "learning_rate": 0.00013461538461538464,
-      "loss": 1.4771,
+      "epoch": 0.19321817018554063,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00017524271844660195,
+      "loss": 1.6045,
       "step": 151
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.00013333333333333334,
-      "loss": 1.5853,
+      "epoch": 0.1944977607165707,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00017475728155339805,
+      "loss": 1.7295,
       "step": 152
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.2197265625,
-      "learning_rate": 0.00013205128205128204,
-      "loss": 1.5237,
+      "epoch": 0.19577735124760076,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00017427184466019418,
+      "loss": 1.6617,
       "step": 153
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.00013076923076923077,
-      "loss": 1.4963,
+      "epoch": 0.19705694177863084,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00017378640776699028,
+      "loss": 1.6196,
       "step": 154
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.2373046875,
-      "learning_rate": 0.0001294871794871795,
-      "loss": 1.5393,
+      "epoch": 0.19833653230966092,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001733009708737864,
+      "loss": 1.6836,
       "step": 155
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.00012820512820512823,
-      "loss": 1.4725,
+      "epoch": 0.19961612284069097,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017281553398058253,
+      "loss": 1.6273,
       "step": 156
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.220703125,
-      "learning_rate": 0.00012692307692307693,
-      "loss": 1.5187,
+      "epoch": 0.20089571337172105,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00017233009708737864,
+      "loss": 1.6758,
       "step": 157
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.00012564102564102564,
-      "loss": 1.5263,
+      "epoch": 0.20217530390275112,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00017184466019417476,
+      "loss": 1.6883,
       "step": 158
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 0.00012435897435897437,
-      "loss": 1.5388,
+      "epoch": 0.2034548944337812,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017135922330097086,
+      "loss": 1.676,
       "step": 159
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.224609375,
-      "learning_rate": 0.0001230769230769231,
-      "loss": 1.4606,
+      "epoch": 0.20473448496481125,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.000170873786407767,
+      "loss": 1.6166,
       "step": 160
     },
     {
-      "epoch": 0.2,
-      "eval_loss": 1.5093975067138672,
-      "eval_runtime": 109.3522,
-      "eval_samples_per_second": 45.724,
-      "eval_steps_per_second": 1.436,
+      "epoch": 0.20473448496481125,
+      "eval_loss": 1.6558058261871338,
+      "eval_runtime": 103.0857,
+      "eval_samples_per_second": 48.503,
+      "eval_steps_per_second": 1.523,
       "step": 160
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 0.00012179487179487179,
-      "loss": 1.5584,
+      "epoch": 0.20601407549584133,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001703883495145631,
+      "loss": 1.6962,
       "step": 161
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.224609375,
-      "learning_rate": 0.00012051282051282052,
-      "loss": 1.4941,
+      "epoch": 0.2072936660268714,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00016990291262135922,
+      "loss": 1.643,
       "step": 162
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.00011923076923076923,
-      "loss": 1.5283,
+      "epoch": 0.20857325655790146,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00016941747572815535,
+      "loss": 1.6716,
       "step": 163
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.2373046875,
-      "learning_rate": 0.00011794871794871796,
-      "loss": 1.4864,
+      "epoch": 0.20985284708893154,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00016893203883495145,
+      "loss": 1.6173,
       "step": 164
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.24609375,
-      "learning_rate": 0.00011666666666666668,
-      "loss": 1.4398,
+      "epoch": 0.21113243761996162,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00016844660194174758,
+      "loss": 1.5833,
       "step": 165
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.251953125,
-      "learning_rate": 0.00011538461538461538,
-      "loss": 1.514,
+      "epoch": 0.2124120281509917,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00016796116504854368,
+      "loss": 1.6686,
       "step": 166
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.25390625,
-      "learning_rate": 0.0001141025641025641,
-      "loss": 1.5578,
+      "epoch": 0.21369161868202174,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0001674757281553398,
+      "loss": 1.6964,
       "step": 167
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 0.00011282051282051283,
-      "loss": 1.5682,
+      "epoch": 0.21497120921305182,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00016699029126213594,
+      "loss": 1.6934,
       "step": 168
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 0.23046875,
-      "learning_rate": 0.00011153846153846154,
-      "loss": 1.4911,
+      "epoch": 0.2162507997440819,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00016650485436893204,
+      "loss": 1.6628,
       "step": 169
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 0.236328125,
-      "learning_rate": 0.00011025641025641027,
-      "loss": 1.5906,
-      "step": 170
-    },
-    {
-      "epoch": 0.22,
-      "eval_loss": 1.5074635744094849,
-      "eval_runtime": 108.9797,
-      "eval_samples_per_second": 45.88,
-      "eval_steps_per_second": 1.441,
+      "epoch": 0.21753039027511195,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00016601941747572817,
+      "loss": 1.7518,
       "step": 170
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 0.271484375,
-      "learning_rate": 0.00010897435897435896,
-      "loss": 1.5266,
+      "epoch": 0.21880998080614203,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00016553398058252427,
+      "loss": 1.6653,
       "step": 171
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 0.23828125,
-      "learning_rate": 0.0001076923076923077,
-      "loss": 1.4515,
+      "epoch": 0.2200895713371721,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001650485436893204,
+      "loss": 1.6052,
       "step": 172
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 0.2451171875,
-      "learning_rate": 0.00010641025641025641,
-      "loss": 1.5492,
+      "epoch": 0.22136916186820219,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001645631067961165,
+      "loss": 1.6775,
       "step": 173
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 0.00010512820512820514,
-      "loss": 1.4961,
+      "epoch": 0.22264875239923224,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00016407766990291262,
+      "loss": 1.6516,
       "step": 174
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 0.2421875,
-      "learning_rate": 0.00010384615384615386,
-      "loss": 1.5594,
+      "epoch": 0.22392834293026231,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00016359223300970875,
+      "loss": 1.7263,
       "step": 175
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.2275390625,
-      "learning_rate": 0.00010256410256410256,
-      "loss": 1.5469,
+      "epoch": 0.2252079334612924,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00016310679611650485,
+      "loss": 1.6874,
       "step": 176
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.240234375,
-      "learning_rate": 0.00010128205128205129,
-      "loss": 1.4925,
+      "epoch": 0.22648752399232247,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00016262135922330098,
+      "loss": 1.6519,
       "step": 177
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.2578125,
-      "learning_rate": 0.0001,
-      "loss": 1.4808,
+      "epoch": 0.22776711452335252,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00016213592233009708,
+      "loss": 1.6241,
       "step": 178
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.251953125,
-      "learning_rate": 9.871794871794872e-05,
-      "loss": 1.4963,
+      "epoch": 0.2290467050543826,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001616504854368932,
+      "loss": 1.6657,
       "step": 179
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.2578125,
-      "learning_rate": 9.743589743589744e-05,
-      "loss": 1.5397,
+      "epoch": 0.23032629558541268,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0001611650485436893,
+      "loss": 1.6658,
       "step": 180
     },
     {
-      "epoch": 0.23,
-      "eval_loss": 1.5060302019119263,
-      "eval_runtime": 109.1672,
-      "eval_samples_per_second": 45.801,
-      "eval_steps_per_second": 1.438,
+      "epoch": 0.23032629558541268,
+      "eval_loss": 1.6510590314865112,
+      "eval_runtime": 103.101,
+      "eval_samples_per_second": 48.496,
+      "eval_steps_per_second": 1.523,
       "step": 180
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.2451171875,
-      "learning_rate": 9.615384615384617e-05,
-      "loss": 1.5202,
+      "epoch": 0.23160588611644273,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00016067961165048544,
+      "loss": 1.6732,
       "step": 181
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.2158203125,
-      "learning_rate": 9.487179487179487e-05,
-      "loss": 1.488,
+      "epoch": 0.2328854766474728,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00016019417475728157,
+      "loss": 1.6288,
       "step": 182
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 0.25390625,
-      "learning_rate": 9.35897435897436e-05,
-      "loss": 1.5124,
+      "epoch": 0.23416506717850288,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00015970873786407767,
+      "loss": 1.65,
       "step": 183
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.2578125,
-      "learning_rate": 9.230769230769232e-05,
-      "loss": 1.4791,
+      "epoch": 0.23544465770953296,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0001592233009708738,
+      "loss": 1.643,
       "step": 184
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.2421875,
-      "learning_rate": 9.102564102564103e-05,
-      "loss": 1.6039,
+      "epoch": 0.236724248240563,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001587378640776699,
+      "loss": 1.7358,
       "step": 185
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 8.974358974358975e-05,
-      "loss": 1.5389,
+      "epoch": 0.2380038387715931,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015825242718446603,
+      "loss": 1.6775,
       "step": 186
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 8.846153846153847e-05,
-      "loss": 1.5542,
+      "epoch": 0.23928342930262317,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00015776699029126213,
+      "loss": 1.7007,
       "step": 187
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 8.717948717948718e-05,
-      "loss": 1.5131,
+      "epoch": 0.24056301983365322,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00015728155339805825,
+      "loss": 1.6372,
       "step": 188
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.248046875,
-      "learning_rate": 8.58974358974359e-05,
-      "loss": 1.5371,
+      "epoch": 0.2418426103646833,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00015679611650485438,
+      "loss": 1.7057,
       "step": 189
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.2392578125,
-      "learning_rate": 8.461538461538461e-05,
-      "loss": 1.5264,
-      "step": 190
-    },
-    {
-      "epoch": 0.24,
-      "eval_loss": 1.5048484802246094,
-      "eval_runtime": 109.1034,
-      "eval_samples_per_second": 45.828,
-      "eval_steps_per_second": 1.439,
+      "epoch": 0.24312220089571338,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00015631067961165048,
+      "loss": 1.6965,
       "step": 190
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 0.236328125,
-      "learning_rate": 8.333333333333334e-05,
-      "loss": 1.5295,
+      "epoch": 0.24440179142674345,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0001558252427184466,
+      "loss": 1.6888,
       "step": 191
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.2333984375,
-      "learning_rate": 8.205128205128205e-05,
-      "loss": 1.5677,
+      "epoch": 0.2456813819577735,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0001553398058252427,
+      "loss": 1.7146,
       "step": 192
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.255859375,
-      "learning_rate": 8.076923076923078e-05,
-      "loss": 1.4473,
+      "epoch": 0.24696097248880358,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00015485436893203884,
+      "loss": 1.5767,
       "step": 193
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 7.948717948717948e-05,
-      "loss": 1.5756,
+      "epoch": 0.24824056301983366,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00015436893203883497,
+      "loss": 1.7411,
       "step": 194
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.236328125,
-      "learning_rate": 7.820512820512821e-05,
-      "loss": 1.4832,
+      "epoch": 0.2495201535508637,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00015388349514563107,
+      "loss": 1.6321,
       "step": 195
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.2294921875,
-      "learning_rate": 7.692307692307693e-05,
-      "loss": 1.5144,
+      "epoch": 0.2507997440818938,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001533980582524272,
+      "loss": 1.6406,
       "step": 196
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.248046875,
-      "learning_rate": 7.564102564102564e-05,
-      "loss": 1.5077,
+      "epoch": 0.25207933461292387,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0001529126213592233,
+      "loss": 1.6426,
       "step": 197
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.21875,
-      "learning_rate": 7.435897435897436e-05,
-      "loss": 1.5042,
+      "epoch": 0.2533589251439539,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015242718446601943,
+      "loss": 1.643,
       "step": 198
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 0.22265625,
-      "learning_rate": 7.307692307692307e-05,
-      "loss": 1.5022,
+      "epoch": 0.254638515674984,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00015194174757281553,
+      "loss": 1.6484,
       "step": 199
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.265625,
-      "learning_rate": 7.17948717948718e-05,
-      "loss": 1.484,
+      "epoch": 0.2559181062060141,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00015145631067961166,
+      "loss": 1.6357,
       "step": 200
     },
     {
-      "epoch": 0.26,
-      "eval_loss": 1.503655195236206,
-      "eval_runtime": 117.4418,
-      "eval_samples_per_second": 42.574,
-      "eval_steps_per_second": 1.337,
+      "epoch": 0.2559181062060141,
+      "eval_loss": 1.6480427980422974,
+      "eval_runtime": 103.1119,
+      "eval_samples_per_second": 48.491,
+      "eval_steps_per_second": 1.523,
       "step": 200
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.2431640625,
-      "learning_rate": 7.051282051282052e-05,
-      "loss": 1.4973,
+      "epoch": 0.2571976967370441,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00015097087378640778,
+      "loss": 1.6421,
       "step": 201
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.2314453125,
-      "learning_rate": 6.923076923076924e-05,
-      "loss": 1.5267,
+      "epoch": 0.25847728726807423,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00015048543689320389,
+      "loss": 1.6995,
       "step": 202
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.2236328125,
-      "learning_rate": 6.794871794871795e-05,
-      "loss": 1.5657,
+      "epoch": 0.2597568777991043,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 1.7139,
       "step": 203
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.23046875,
-      "learning_rate": 6.666666666666667e-05,
-      "loss": 1.5184,
+      "epoch": 0.26103646833013433,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00014951456310679611,
+      "loss": 1.6852,
       "step": 204
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.23828125,
-      "learning_rate": 6.538461538461539e-05,
-      "loss": 1.4522,
+      "epoch": 0.26231605886116444,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00014902912621359224,
+      "loss": 1.6057,
       "step": 205
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.2451171875,
-      "learning_rate": 6.410256410256412e-05,
-      "loss": 1.5309,
+      "epoch": 0.2635956493921945,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00014854368932038834,
+      "loss": 1.6795,
       "step": 206
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 0.2373046875,
-      "learning_rate": 6.282051282051282e-05,
-      "loss": 1.4503,
+      "epoch": 0.2648752399232246,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00014805825242718447,
+      "loss": 1.5875,
       "step": 207
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.2265625,
-      "learning_rate": 6.153846153846155e-05,
-      "loss": 1.5173,
+      "epoch": 0.26615483045425464,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001475728155339806,
+      "loss": 1.6799,
       "step": 208
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.275390625,
-      "learning_rate": 6.025641025641026e-05,
-      "loss": 1.5112,
+      "epoch": 0.2674344209852847,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0001470873786407767,
+      "loss": 1.6385,
       "step": 209
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.234375,
-      "learning_rate": 5.897435897435898e-05,
-      "loss": 1.5342,
-      "step": 210
-    },
-    {
-      "epoch": 0.27,
-      "eval_loss": 1.5029850006103516,
-      "eval_runtime": 118.7123,
-      "eval_samples_per_second": 42.119,
-      "eval_steps_per_second": 1.323,
+      "epoch": 0.2687140115163148,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014660194174757283,
+      "loss": 1.7025,
       "step": 210
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.2333984375,
-      "learning_rate": 5.769230769230769e-05,
-      "loss": 1.5216,
+      "epoch": 0.26999360204734485,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00014611650485436893,
+      "loss": 1.6543,
       "step": 211
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.2373046875,
-      "learning_rate": 5.6410256410256414e-05,
-      "loss": 1.5005,
+      "epoch": 0.2712731925783749,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00014563106796116506,
+      "loss": 1.6421,
       "step": 212
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.23046875,
-      "learning_rate": 5.512820512820514e-05,
-      "loss": 1.5372,
+      "epoch": 0.272552783109405,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0001451456310679612,
+      "loss": 1.6629,
       "step": 213
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 5.384615384615385e-05,
-      "loss": 1.4916,
+      "epoch": 0.27383237364043506,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0001446601941747573,
+      "loss": 1.632,
       "step": 214
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.232421875,
-      "learning_rate": 5.256410256410257e-05,
-      "loss": 1.4511,
+      "epoch": 0.2751119641714651,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00014417475728155342,
+      "loss": 1.59,
       "step": 215
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.251953125,
-      "learning_rate": 5.128205128205128e-05,
-      "loss": 1.4984,
+      "epoch": 0.2763915547024952,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00014368932038834952,
+      "loss": 1.6291,
       "step": 216
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.275390625,
-      "learning_rate": 5e-05,
-      "loss": 1.5332,
+      "epoch": 0.27767114523352526,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00014320388349514565,
+      "loss": 1.6759,
       "step": 217
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 4.871794871794872e-05,
-      "loss": 1.558,
+      "epoch": 0.27895073576455537,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00014271844660194175,
+      "loss": 1.6989,
       "step": 218
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.255859375,
-      "learning_rate": 4.7435897435897435e-05,
-      "loss": 1.5057,
+      "epoch": 0.2802303262955854,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00014223300970873787,
+      "loss": 1.6291,
       "step": 219
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.2431640625,
-      "learning_rate": 4.615384615384616e-05,
-      "loss": 1.5374,
+      "epoch": 0.28150991682661547,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.000141747572815534,
+      "loss": 1.6757,
       "step": 220
     },
     {
-      "epoch": 0.28,
-      "eval_loss": 1.5021294355392456,
-      "eval_runtime": 120.5989,
-      "eval_samples_per_second": 41.46,
-      "eval_steps_per_second": 1.302,
+      "epoch": 0.28150991682661547,
+      "eval_loss": 1.645837664604187,
+      "eval_runtime": 103.1004,
+      "eval_samples_per_second": 48.496,
+      "eval_steps_per_second": 1.523,
       "step": 220
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.26953125,
-      "learning_rate": 4.4871794871794874e-05,
-      "loss": 1.5332,
+      "epoch": 0.2827895073576456,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0001412621359223301,
+      "loss": 1.6866,
       "step": 221
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 0.236328125,
-      "learning_rate": 4.358974358974359e-05,
-      "loss": 1.5205,
+      "epoch": 0.2840690978886756,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00014077669902912623,
+      "loss": 1.6816,
       "step": 222
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.224609375,
-      "learning_rate": 4.230769230769231e-05,
-      "loss": 1.497,
+      "epoch": 0.2853486884197057,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00014029126213592233,
+      "loss": 1.6683,
       "step": 223
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.22265625,
-      "learning_rate": 4.1025641025641023e-05,
-      "loss": 1.5158,
+      "epoch": 0.2866282789507358,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00013980582524271846,
+      "loss": 1.6537,
       "step": 224
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.2421875,
-      "learning_rate": 3.974358974358974e-05,
-      "loss": 1.4847,
+      "epoch": 0.28790786948176583,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.00013932038834951456,
+      "loss": 1.6292,
       "step": 225
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.244140625,
-      "learning_rate": 3.846153846153846e-05,
-      "loss": 1.4456,
+      "epoch": 0.2891874600127959,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0001388349514563107,
+      "loss": 1.5684,
       "step": 226
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.228515625,
-      "learning_rate": 3.717948717948718e-05,
-      "loss": 1.5334,
+      "epoch": 0.290467050543826,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00013834951456310682,
+      "loss": 1.6822,
       "step": 227
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.25390625,
-      "learning_rate": 3.58974358974359e-05,
-      "loss": 1.5238,
+      "epoch": 0.29174664107485604,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00013786407766990292,
+      "loss": 1.668,
       "step": 228
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.2490234375,
-      "learning_rate": 3.461538461538462e-05,
-      "loss": 1.4388,
+      "epoch": 0.2930262316058861,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00013737864077669905,
+      "loss": 1.564,
       "step": 229
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 3.3333333333333335e-05,
-      "loss": 1.5555,
-      "step": 230
-    },
-    {
-      "epoch": 0.29,
-      "eval_loss": 1.5019102096557617,
-      "eval_runtime": 109.1608,
-      "eval_samples_per_second": 45.804,
-      "eval_steps_per_second": 1.438,
+      "epoch": 0.2943058221369162,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00013689320388349515,
+      "loss": 1.6721,
       "step": 230
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.267578125,
-      "learning_rate": 3.205128205128206e-05,
-      "loss": 1.5535,
+      "epoch": 0.29558541266794625,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.00013640776699029128,
+      "loss": 1.6735,
       "step": 231
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.240234375,
-      "learning_rate": 3.0769230769230774e-05,
-      "loss": 1.4322,
+      "epoch": 0.29686500319897635,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0001359223300970874,
+      "loss": 1.5751,
       "step": 232
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 2.948717948717949e-05,
-      "loss": 1.471,
+      "epoch": 0.2981445937300064,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0001354368932038835,
+      "loss": 1.6004,
       "step": 233
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.25,
-      "learning_rate": 2.8205128205128207e-05,
-      "loss": 1.5444,
+      "epoch": 0.29942418426103645,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.00013495145631067963,
+      "loss": 1.6797,
       "step": 234
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.25,
-      "learning_rate": 2.6923076923076923e-05,
-      "loss": 1.5331,
+      "epoch": 0.30070377479206656,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00013446601941747573,
+      "loss": 1.6688,
       "step": 235
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.236328125,
-      "learning_rate": 2.564102564102564e-05,
-      "loss": 1.5355,
+      "epoch": 0.3019833653230966,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.00013398058252427186,
+      "loss": 1.6719,
       "step": 236
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.244140625,
-      "learning_rate": 2.435897435897436e-05,
-      "loss": 1.4883,
+      "epoch": 0.30326295585412666,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00013349514563106796,
+      "loss": 1.6186,
       "step": 237
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 0.263671875,
-      "learning_rate": 2.307692307692308e-05,
-      "loss": 1.5266,
+      "epoch": 0.30454254638515676,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0001330097087378641,
+      "loss": 1.6694,
       "step": 238
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.2451171875,
-      "learning_rate": 2.1794871794871795e-05,
-      "loss": 1.5576,
+      "epoch": 0.3058221369161868,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00013252427184466022,
+      "loss": 1.7104,
       "step": 239
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.259765625,
-      "learning_rate": 2.0512820512820512e-05,
-      "loss": 1.5615,
+      "epoch": 0.30710172744721687,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00013203883495145632,
+      "loss": 1.6939,
       "step": 240
     },
     {
-      "epoch": 0.31,
-      "eval_loss": 1.5016273260116577,
-      "eval_runtime": 109.191,
-      "eval_samples_per_second": 45.791,
-      "eval_steps_per_second": 1.438,
+      "epoch": 0.30710172744721687,
+      "eval_loss": 1.6442595720291138,
+      "eval_runtime": 103.1037,
+      "eval_samples_per_second": 48.495,
+      "eval_steps_per_second": 1.523,
       "step": 240
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.23828125,
-      "learning_rate": 1.923076923076923e-05,
-      "loss": 1.531,
+      "epoch": 0.30838131797824697,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00013155339805825245,
+      "loss": 1.6986,
       "step": 241
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.2470703125,
-      "learning_rate": 1.794871794871795e-05,
-      "loss": 1.4563,
+      "epoch": 0.309660908509277,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00013106796116504855,
+      "loss": 1.5877,
       "step": 242
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.24609375,
-      "learning_rate": 1.6666666666666667e-05,
-      "loss": 1.4773,
+      "epoch": 0.31094049904030713,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00013058252427184468,
+      "loss": 1.6292,
       "step": 243
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.2431640625,
-      "learning_rate": 1.5384615384615387e-05,
-      "loss": 1.4496,
+      "epoch": 0.3122200895713372,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00013009708737864078,
+      "loss": 1.6097,
       "step": 244
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.228515625,
-      "learning_rate": 1.4102564102564104e-05,
-      "loss": 1.5677,
+      "epoch": 0.31349968010236723,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001296116504854369,
+      "loss": 1.7044,
       "step": 245
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 0.2578125,
-      "learning_rate": 1.282051282051282e-05,
-      "loss": 1.4935,
+      "epoch": 0.31477927063339733,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00012912621359223304,
+      "loss": 1.6325,
       "step": 246
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 1.153846153846154e-05,
-      "loss": 1.5357,
+      "epoch": 0.3160588611644274,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00012864077669902914,
+      "loss": 1.6831,
       "step": 247
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 0.259765625,
-      "learning_rate": 1.0256410256410256e-05,
-      "loss": 1.5693,
+      "epoch": 0.31733845169545744,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.00012815533980582526,
+      "loss": 1.7119,
       "step": 248
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 0.271484375,
-      "learning_rate": 8.974358974358976e-06,
-      "loss": 1.5037,
+      "epoch": 0.31861804222648754,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00012766990291262137,
+      "loss": 1.6437,
       "step": 249
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 0.236328125,
-      "learning_rate": 7.692307692307694e-06,
-      "loss": 1.4786,
-      "step": 250
-    },
-    {
-      "epoch": 0.32,
-      "eval_loss": 1.501659631729126,
-      "eval_runtime": 109.1682,
-      "eval_samples_per_second": 45.801,
-      "eval_steps_per_second": 1.438,
+      "epoch": 0.3198976327575176,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0001271844660194175,
+      "loss": 1.6205,
       "step": 250
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 0.2412109375,
-      "learning_rate": 6.41025641025641e-06,
-      "loss": 1.5624,
+      "epoch": 0.32117722328854764,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00012669902912621362,
+      "loss": 1.7224,
       "step": 251
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 0.232421875,
-      "learning_rate": 5.128205128205128e-06,
-      "loss": 1.4851,
+      "epoch": 0.32245681381957775,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00012621359223300972,
+      "loss": 1.6414,
       "step": 252
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 0.2451171875,
-      "learning_rate": 3.846153846153847e-06,
-      "loss": 1.4184,
+      "epoch": 0.3237364043506078,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.00012572815533980585,
+      "loss": 1.5297,
       "step": 253
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 0.2353515625,
-      "learning_rate": 2.564102564102564e-06,
-      "loss": 1.4701,
+      "epoch": 0.32501599488163785,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00012524271844660195,
+      "loss": 1.5789,
       "step": 254
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 0.244140625,
-      "learning_rate": 1.282051282051282e-06,
-      "loss": 1.5079,
+      "epoch": 0.32629558541266795,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00012475728155339805,
+      "loss": 1.6328,
       "step": 255
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 0.2216796875,
-      "learning_rate": 0.0,
-      "loss": 1.5386,
+      "epoch": 0.327575175943698,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00012427184466019418,
+      "loss": 1.6872,
       "step": 256
     },
     {
-      "epoch": 0.33,
-      "step": 256,
-      "total_flos": 1.8280284932328653e+17,
-      "train_loss": 1.802713230252266,
-      "train_runtime": 3920.872,
-      "train_samples_per_second": 4.179,
-      "train_steps_per_second": 0.065
+      "epoch": 0.3288547664747281,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00012378640776699028,
+      "loss": 1.6293,
+      "step": 257
     },
     {
-      "epoch": 0.33,
-      "eval_loss": 1.5016273260116577,
-      "eval_runtime": 109.2744,
-      "eval_samples_per_second": 45.756,
-      "eval_steps_per_second": 1.437,
-      "step": 256
+      "epoch": 0.33013435700575816,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0001233009708737864,
+      "loss": 1.6105,
+      "step": 258
+    },
+    {
+      "epoch": 0.3314139475367882,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0001228155339805825,
+      "loss": 1.6522,
+      "step": 259
+    },
+    {
+      "epoch": 0.3326935380678183,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00012233009708737864,
+      "loss": 1.6715,
+      "step": 260
+    },
+    {
+      "epoch": 0.3326935380678183,
+      "eval_loss": 1.6419228315353394,
+      "eval_runtime": 103.1115,
+      "eval_samples_per_second": 48.491,
+      "eval_steps_per_second": 1.523,
+      "step": 260
+    },
+    {
+      "epoch": 0.33397312859884837,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.00012184466019417475,
+      "loss": 1.6435,
+      "step": 261
+    },
+    {
+      "epoch": 0.3352527191298784,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.00012135922330097087,
+      "loss": 1.6339,
+      "step": 262
+    },
+    {
+      "epoch": 0.3365323096609085,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.00012087378640776698,
+      "loss": 1.6233,
+      "step": 263
+    },
+    {
+      "epoch": 0.3378119001919386,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0001203883495145631,
+      "loss": 1.61,
+      "step": 264
+    },
+    {
+      "epoch": 0.3390914907229686,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.00011990291262135923,
+      "loss": 1.6532,
+      "step": 265
+    },
+    {
+      "epoch": 0.34037108125399873,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00011941747572815534,
+      "loss": 1.6994,
+      "step": 266
+    },
+    {
+      "epoch": 0.3416506717850288,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.00011893203883495146,
+      "loss": 1.5847,
+      "step": 267
+    },
+    {
+      "epoch": 0.3429302623160589,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00011844660194174757,
+      "loss": 1.6299,
+      "step": 268
+    },
+    {
+      "epoch": 0.34420985284708894,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00011796116504854368,
+      "loss": 1.63,
+      "step": 269
+    },
+    {
+      "epoch": 0.345489443378119,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0001174757281553398,
+      "loss": 1.6556,
+      "step": 270
+    },
+    {
+      "epoch": 0.3467690339091491,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00011699029126213593,
+      "loss": 1.6967,
+      "step": 271
+    },
+    {
+      "epoch": 0.34804862444017914,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.00011650485436893204,
+      "loss": 1.7115,
+      "step": 272
+    },
+    {
+      "epoch": 0.3493282149712092,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00011601941747572816,
+      "loss": 1.7578,
+      "step": 273
+    },
+    {
+      "epoch": 0.3506078055022393,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.00011553398058252427,
+      "loss": 1.6873,
+      "step": 274
+    },
+    {
+      "epoch": 0.35188739603326935,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00011504854368932039,
+      "loss": 1.6566,
+      "step": 275
+    },
+    {
+      "epoch": 0.3531669865642994,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0001145631067961165,
+      "loss": 1.7911,
+      "step": 276
+    },
+    {
+      "epoch": 0.3544465770953295,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.00011407766990291261,
+      "loss": 1.6738,
+      "step": 277
+    },
+    {
+      "epoch": 0.35572616762635956,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.00011359223300970874,
+      "loss": 1.6299,
+      "step": 278
+    },
+    {
+      "epoch": 0.3570057581573896,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00011310679611650486,
+      "loss": 1.7136,
+      "step": 279
+    },
+    {
+      "epoch": 0.3582853486884197,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00011262135922330097,
+      "loss": 1.6764,
+      "step": 280
+    },
+    {
+      "epoch": 0.3582853486884197,
+      "eval_loss": 1.6408168077468872,
+      "eval_runtime": 103.085,
+      "eval_samples_per_second": 48.504,
+      "eval_steps_per_second": 1.523,
+      "step": 280
+    },
+    {
+      "epoch": 0.35956493921944976,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00011213592233009709,
+      "loss": 1.6869,
+      "step": 281
+    },
+    {
+      "epoch": 0.36084452975047987,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0001116504854368932,
+      "loss": 1.6637,
+      "step": 282
+    },
+    {
+      "epoch": 0.3621241202815099,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00011116504854368932,
+      "loss": 1.6449,
+      "step": 283
+    },
+    {
+      "epoch": 0.36340371081253997,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.00011067961165048544,
+      "loss": 1.6964,
+      "step": 284
+    },
+    {
+      "epoch": 0.3646833013435701,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00011019417475728156,
+      "loss": 1.7385,
+      "step": 285
+    },
+    {
+      "epoch": 0.3659628918746001,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00010970873786407767,
+      "loss": 1.6712,
+      "step": 286
+    },
+    {
+      "epoch": 0.3672424824056302,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00010922330097087379,
+      "loss": 1.675,
+      "step": 287
+    },
+    {
+      "epoch": 0.3685220729366603,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0001087378640776699,
+      "loss": 1.6172,
+      "step": 288
+    },
+    {
+      "epoch": 0.36980166346769033,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00010825242718446602,
+      "loss": 1.6774,
+      "step": 289
+    },
+    {
+      "epoch": 0.3710812539987204,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00010776699029126213,
+      "loss": 1.6645,
+      "step": 290
+    },
+    {
+      "epoch": 0.3723608445297505,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00010728155339805826,
+      "loss": 1.6674,
+      "step": 291
+    },
+    {
+      "epoch": 0.37364043506078054,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.00010679611650485437,
+      "loss": 1.6028,
+      "step": 292
+    },
+    {
+      "epoch": 0.37492002559181065,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00010631067961165049,
+      "loss": 1.7152,
+      "step": 293
+    },
+    {
+      "epoch": 0.3761996161228407,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001058252427184466,
+      "loss": 1.6715,
+      "step": 294
+    },
+    {
+      "epoch": 0.37747920665387075,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00010533980582524272,
+      "loss": 1.6071,
+      "step": 295
+    },
+    {
+      "epoch": 0.37875879718490085,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.00010485436893203883,
+      "loss": 1.6542,
+      "step": 296
+    },
+    {
+      "epoch": 0.3800383877159309,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.00010436893203883496,
+      "loss": 1.734,
+      "step": 297
+    },
+    {
+      "epoch": 0.38131797824696095,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.00010388349514563107,
+      "loss": 1.577,
+      "step": 298
+    },
+    {
+      "epoch": 0.38259756877799106,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.00010339805825242719,
+      "loss": 1.6174,
+      "step": 299
+    },
+    {
+      "epoch": 0.3838771593090211,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0001029126213592233,
+      "loss": 1.6505,
+      "step": 300
+    },
+    {
+      "epoch": 0.3838771593090211,
+      "eval_loss": 1.6392818689346313,
+      "eval_runtime": 103.1129,
+      "eval_samples_per_second": 48.491,
+      "eval_steps_per_second": 1.523,
+      "step": 300
+    },
+    {
+      "epoch": 0.38515674984005116,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.00010242718446601942,
+      "loss": 1.6486,
+      "step": 301
+    },
+    {
+      "epoch": 0.38643634037108127,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00010194174757281553,
+      "loss": 1.6358,
+      "step": 302
+    },
+    {
+      "epoch": 0.3877159309021113,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.00010145631067961166,
+      "loss": 1.6483,
+      "step": 303
+    },
+    {
+      "epoch": 0.3889955214331414,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.00010097087378640778,
+      "loss": 1.7252,
+      "step": 304
+    },
+    {
+      "epoch": 0.3902751119641715,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00010048543689320389,
+      "loss": 1.7207,
+      "step": 305
+    },
+    {
+      "epoch": 0.3915547024952015,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0001,
+      "loss": 1.6304,
+      "step": 306
+    },
+    {
+      "epoch": 0.39283429302623163,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.951456310679612e-05,
+      "loss": 1.6087,
+      "step": 307
+    },
+    {
+      "epoch": 0.3941138835572617,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.902912621359223e-05,
+      "loss": 1.6207,
+      "step": 308
+    },
+    {
+      "epoch": 0.39539347408829173,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 9.854368932038835e-05,
+      "loss": 1.618,
+      "step": 309
+    },
+    {
+      "epoch": 0.39667306461932184,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.805825242718448e-05,
+      "loss": 1.5887,
+      "step": 310
+    },
+    {
+      "epoch": 0.3979526551503519,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 9.757281553398059e-05,
+      "loss": 1.6387,
+      "step": 311
+    },
+    {
+      "epoch": 0.39923224568138194,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.70873786407767e-05,
+      "loss": 1.6049,
+      "step": 312
+    },
+    {
+      "epoch": 0.40051183621241204,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 9.660194174757282e-05,
+      "loss": 1.6714,
+      "step": 313
+    },
+    {
+      "epoch": 0.4017914267434421,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.611650485436893e-05,
+      "loss": 1.6432,
+      "step": 314
+    },
+    {
+      "epoch": 0.40307101727447214,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.563106796116505e-05,
+      "loss": 1.6556,
+      "step": 315
+    },
+    {
+      "epoch": 0.40435060780550225,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.514563106796118e-05,
+      "loss": 1.7007,
+      "step": 316
+    },
+    {
+      "epoch": 0.4056301983365323,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.466019417475729e-05,
+      "loss": 1.6269,
+      "step": 317
+    },
+    {
+      "epoch": 0.4069097888675624,
+      "grad_norm": 0.19140625,
+      "learning_rate": 9.417475728155341e-05,
+      "loss": 1.602,
+      "step": 318
+    },
+    {
+      "epoch": 0.40818937939859246,
+      "grad_norm": 0.19140625,
+      "learning_rate": 9.368932038834952e-05,
+      "loss": 1.6387,
+      "step": 319
+    },
+    {
+      "epoch": 0.4094689699296225,
+      "grad_norm": 0.201171875,
+      "learning_rate": 9.320388349514564e-05,
+      "loss": 1.6849,
+      "step": 320
+    },
+    {
+      "epoch": 0.4094689699296225,
+      "eval_loss": 1.6380316019058228,
+      "eval_runtime": 103.1068,
+      "eval_samples_per_second": 48.493,
+      "eval_steps_per_second": 1.523,
+      "step": 320
+    },
+    {
+      "epoch": 0.4107485604606526,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.271844660194175e-05,
+      "loss": 1.6638,
+      "step": 321
+    },
+    {
+      "epoch": 0.41202815099168266,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.223300970873788e-05,
+      "loss": 1.6501,
+      "step": 322
+    },
+    {
+      "epoch": 0.4133077415227127,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 9.174757281553399e-05,
+      "loss": 1.624,
+      "step": 323
+    },
+    {
+      "epoch": 0.4145873320537428,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.126213592233011e-05,
+      "loss": 1.5555,
+      "step": 324
+    },
+    {
+      "epoch": 0.41586692258477287,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.077669902912622e-05,
+      "loss": 1.6896,
+      "step": 325
+    },
+    {
+      "epoch": 0.4171465131158029,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.029126213592234e-05,
+      "loss": 1.696,
+      "step": 326
+    },
+    {
+      "epoch": 0.418426103646833,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.980582524271845e-05,
+      "loss": 1.7137,
+      "step": 327
+    },
+    {
+      "epoch": 0.4197056941778631,
+      "grad_norm": 0.189453125,
+      "learning_rate": 8.932038834951457e-05,
+      "loss": 1.705,
+      "step": 328
+    },
+    {
+      "epoch": 0.4209852847088932,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 8.88349514563107e-05,
+      "loss": 1.63,
+      "step": 329
+    },
+    {
+      "epoch": 0.42226487523992323,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.834951456310681e-05,
+      "loss": 1.7132,
+      "step": 330
+    },
+    {
+      "epoch": 0.4235444657709533,
+      "grad_norm": 0.1875,
+      "learning_rate": 8.786407766990292e-05,
+      "loss": 1.6763,
+      "step": 331
+    },
+    {
+      "epoch": 0.4248240563019834,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 8.737864077669902e-05,
+      "loss": 1.6609,
+      "step": 332
+    },
+    {
+      "epoch": 0.42610364683301344,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 8.689320388349514e-05,
+      "loss": 1.738,
+      "step": 333
+    },
+    {
+      "epoch": 0.4273832373640435,
+      "grad_norm": 0.19140625,
+      "learning_rate": 8.640776699029127e-05,
+      "loss": 1.6413,
+      "step": 334
+    },
+    {
+      "epoch": 0.4286628278950736,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 8.592233009708738e-05,
+      "loss": 1.6171,
+      "step": 335
+    },
+    {
+      "epoch": 0.42994241842610365,
+      "grad_norm": 0.203125,
+      "learning_rate": 8.54368932038835e-05,
+      "loss": 1.6703,
+      "step": 336
+    },
+    {
+      "epoch": 0.4312220089571337,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.495145631067961e-05,
+      "loss": 1.7195,
+      "step": 337
+    },
+    {
+      "epoch": 0.4325015994881638,
+      "grad_norm": 0.20703125,
+      "learning_rate": 8.446601941747573e-05,
+      "loss": 1.5734,
+      "step": 338
+    },
+    {
+      "epoch": 0.43378119001919385,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.398058252427184e-05,
+      "loss": 1.697,
+      "step": 339
+    },
+    {
+      "epoch": 0.4350607805502239,
+      "grad_norm": 0.216796875,
+      "learning_rate": 8.349514563106797e-05,
+      "loss": 1.7021,
+      "step": 340
+    },
+    {
+      "epoch": 0.4350607805502239,
+      "eval_loss": 1.637184739112854,
+      "eval_runtime": 103.0987,
+      "eval_samples_per_second": 48.497,
+      "eval_steps_per_second": 1.523,
+      "step": 340
+    },
+    {
+      "epoch": 0.436340371081254,
+      "grad_norm": 0.197265625,
+      "learning_rate": 8.300970873786408e-05,
+      "loss": 1.6074,
+      "step": 341
+    },
+    {
+      "epoch": 0.43761996161228406,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.25242718446602e-05,
+      "loss": 1.6612,
+      "step": 342
+    },
+    {
+      "epoch": 0.43889955214331416,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 8.203883495145631e-05,
+      "loss": 1.6748,
+      "step": 343
+    },
+    {
+      "epoch": 0.4401791426743442,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.155339805825243e-05,
+      "loss": 1.6782,
+      "step": 344
+    },
+    {
+      "epoch": 0.44145873320537427,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.106796116504854e-05,
+      "loss": 1.6676,
+      "step": 345
+    },
+    {
+      "epoch": 0.44273832373640437,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.058252427184466e-05,
+      "loss": 1.6316,
+      "step": 346
+    },
+    {
+      "epoch": 0.4440179142674344,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 8.009708737864078e-05,
+      "loss": 1.6504,
+      "step": 347
+    },
+    {
+      "epoch": 0.44529750479846447,
+      "grad_norm": 0.22265625,
+      "learning_rate": 7.96116504854369e-05,
+      "loss": 1.5952,
+      "step": 348
+    },
+    {
+      "epoch": 0.4465770953294946,
+      "grad_norm": 0.212890625,
+      "learning_rate": 7.912621359223301e-05,
+      "loss": 1.5702,
+      "step": 349
+    },
+    {
+      "epoch": 0.44785668586052463,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 7.864077669902913e-05,
+      "loss": 1.661,
+      "step": 350
+    },
+    {
+      "epoch": 0.4491362763915547,
+      "grad_norm": 0.21484375,
+      "learning_rate": 7.815533980582524e-05,
+      "loss": 1.6323,
+      "step": 351
+    },
+    {
+      "epoch": 0.4504158669225848,
+      "grad_norm": 0.220703125,
+      "learning_rate": 7.766990291262136e-05,
+      "loss": 1.6799,
+      "step": 352
+    },
+    {
+      "epoch": 0.45169545745361483,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.718446601941748e-05,
+      "loss": 1.6872,
+      "step": 353
+    },
+    {
+      "epoch": 0.45297504798464494,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 7.66990291262136e-05,
+      "loss": 1.6583,
+      "step": 354
+    },
+    {
+      "epoch": 0.454254638515675,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.621359223300971e-05,
+      "loss": 1.6758,
+      "step": 355
+    },
+    {
+      "epoch": 0.45553422904670504,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 7.572815533980583e-05,
+      "loss": 1.586,
+      "step": 356
+    },
+    {
+      "epoch": 0.45681381957773515,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 7.524271844660194e-05,
+      "loss": 1.5202,
+      "step": 357
+    },
+    {
+      "epoch": 0.4580934101087652,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 7.475728155339806e-05,
+      "loss": 1.6501,
+      "step": 358
+    },
+    {
+      "epoch": 0.45937300063979525,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 7.427184466019417e-05,
+      "loss": 1.6315,
+      "step": 359
+    },
+    {
+      "epoch": 0.46065259117082535,
+      "grad_norm": 0.19921875,
+      "learning_rate": 7.37864077669903e-05,
+      "loss": 1.661,
+      "step": 360
+    },
+    {
+      "epoch": 0.46065259117082535,
+      "eval_loss": 1.6363531351089478,
+      "eval_runtime": 103.1142,
+      "eval_samples_per_second": 48.49,
+      "eval_steps_per_second": 1.523,
+      "step": 360
+    },
+    {
+      "epoch": 0.4619321817018554,
+      "grad_norm": 0.2109375,
+      "learning_rate": 7.330097087378641e-05,
+      "loss": 1.6214,
+      "step": 361
+    },
+    {
+      "epoch": 0.46321177223288545,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 7.281553398058253e-05,
+      "loss": 1.645,
+      "step": 362
+    },
+    {
+      "epoch": 0.46449136276391556,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 7.233009708737864e-05,
+      "loss": 1.7353,
+      "step": 363
+    },
+    {
+      "epoch": 0.4657709532949456,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.184466019417476e-05,
+      "loss": 1.6297,
+      "step": 364
+    },
+    {
+      "epoch": 0.46705054382597566,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.135922330097087e-05,
+      "loss": 1.6419,
+      "step": 365
+    },
+    {
+      "epoch": 0.46833013435700577,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 7.0873786407767e-05,
+      "loss": 1.5767,
+      "step": 366
+    },
+    {
+      "epoch": 0.4696097248880358,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.038834951456312e-05,
+      "loss": 1.6454,
+      "step": 367
+    },
+    {
+      "epoch": 0.4708893154190659,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.990291262135923e-05,
+      "loss": 1.6399,
+      "step": 368
+    },
+    {
+      "epoch": 0.472168905950096,
+      "grad_norm": 0.19921875,
+      "learning_rate": 6.941747572815534e-05,
+      "loss": 1.7118,
+      "step": 369
+    },
+    {
+      "epoch": 0.473448496481126,
+      "grad_norm": 0.22265625,
+      "learning_rate": 6.893203883495146e-05,
+      "loss": 1.6869,
+      "step": 370
+    },
+    {
+      "epoch": 0.47472808701215613,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.844660194174757e-05,
+      "loss": 1.6411,
+      "step": 371
+    },
+    {
+      "epoch": 0.4760076775431862,
+      "grad_norm": 0.19921875,
+      "learning_rate": 6.79611650485437e-05,
+      "loss": 1.7214,
+      "step": 372
+    },
+    {
+      "epoch": 0.47728726807421623,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.747572815533982e-05,
+      "loss": 1.664,
+      "step": 373
+    },
+    {
+      "epoch": 0.47856685860524634,
+      "grad_norm": 0.20703125,
+      "learning_rate": 6.699029126213593e-05,
+      "loss": 1.6712,
+      "step": 374
+    },
+    {
+      "epoch": 0.4798464491362764,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 6.650485436893205e-05,
+      "loss": 1.662,
+      "step": 375
+    },
+    {
+      "epoch": 0.48112603966730644,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.601941747572816e-05,
+      "loss": 1.6388,
+      "step": 376
+    },
+    {
+      "epoch": 0.48240563019833654,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 6.553398058252428e-05,
+      "loss": 1.6927,
+      "step": 377
+    },
+    {
+      "epoch": 0.4836852207293666,
+      "grad_norm": 0.201171875,
+      "learning_rate": 6.504854368932039e-05,
+      "loss": 1.6294,
+      "step": 378
+    },
+    {
+      "epoch": 0.4849648112603967,
+      "grad_norm": 0.197265625,
+      "learning_rate": 6.456310679611652e-05,
+      "loss": 1.6538,
+      "step": 379
+    },
+    {
+      "epoch": 0.48624440179142675,
+      "grad_norm": 0.236328125,
+      "learning_rate": 6.407766990291263e-05,
+      "loss": 1.6242,
+      "step": 380
+    },
+    {
+      "epoch": 0.48624440179142675,
+      "eval_loss": 1.6356617212295532,
+      "eval_runtime": 103.1559,
+      "eval_samples_per_second": 48.47,
+      "eval_steps_per_second": 1.522,
+      "step": 380
+    },
+    {
+      "epoch": 0.4875239923224568,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.359223300970875e-05,
+      "loss": 1.6998,
+      "step": 381
+    },
+    {
+      "epoch": 0.4888035828534869,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.310679611650486e-05,
+      "loss": 1.6254,
+      "step": 382
+    },
+    {
+      "epoch": 0.49008317338451696,
+      "grad_norm": 0.203125,
+      "learning_rate": 6.262135922330098e-05,
+      "loss": 1.5919,
+      "step": 383
+    },
+    {
+      "epoch": 0.491362763915547,
+      "grad_norm": 0.20703125,
+      "learning_rate": 6.213592233009709e-05,
+      "loss": 1.6995,
+      "step": 384
+    },
+    {
+      "epoch": 0.4926423544465771,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.16504854368932e-05,
+      "loss": 1.6394,
+      "step": 385
+    },
+    {
+      "epoch": 0.49392194497760716,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.116504854368932e-05,
+      "loss": 1.6893,
+      "step": 386
+    },
+    {
+      "epoch": 0.4952015355086372,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 6.0679611650485434e-05,
+      "loss": 1.5551,
+      "step": 387
+    },
+    {
+      "epoch": 0.4964811260396673,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 6.019417475728155e-05,
+      "loss": 1.623,
+      "step": 388
+    },
+    {
+      "epoch": 0.49776071657069737,
+      "grad_norm": 0.205078125,
+      "learning_rate": 5.970873786407767e-05,
+      "loss": 1.5971,
+      "step": 389
+    },
+    {
+      "epoch": 0.4990403071017274,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 5.9223300970873785e-05,
+      "loss": 1.6355,
+      "step": 390
+    },
+    {
+      "epoch": 0.5003198976327575,
+      "grad_norm": 0.205078125,
+      "learning_rate": 5.87378640776699e-05,
+      "loss": 1.7074,
+      "step": 391
+    },
+    {
+      "epoch": 0.5015994881637876,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 5.825242718446602e-05,
+      "loss": 1.6935,
+      "step": 392
+    },
+    {
+      "epoch": 0.5028790786948176,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 5.7766990291262135e-05,
+      "loss": 1.7455,
+      "step": 393
+    },
+    {
+      "epoch": 0.5041586692258477,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 5.728155339805825e-05,
+      "loss": 1.6606,
+      "step": 394
+    },
+    {
+      "epoch": 0.5054382597568778,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 5.679611650485437e-05,
+      "loss": 1.6364,
+      "step": 395
+    },
+    {
+      "epoch": 0.5067178502879078,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.6310679611650486e-05,
+      "loss": 1.6301,
+      "step": 396
+    },
+    {
+      "epoch": 0.5079974408189379,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 5.58252427184466e-05,
+      "loss": 1.5651,
+      "step": 397
+    },
+    {
+      "epoch": 0.509277031349968,
+      "grad_norm": 0.1953125,
+      "learning_rate": 5.533980582524272e-05,
+      "loss": 1.6575,
+      "step": 398
+    },
+    {
+      "epoch": 0.510556621880998,
+      "grad_norm": 0.203125,
+      "learning_rate": 5.4854368932038836e-05,
+      "loss": 1.5667,
+      "step": 399
+    },
+    {
+      "epoch": 0.5118362124120281,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 5.436893203883495e-05,
+      "loss": 1.6508,
+      "step": 400
+    },
+    {
+      "epoch": 0.5118362124120281,
+      "eval_loss": 1.6350483894348145,
+      "eval_runtime": 103.234,
+      "eval_samples_per_second": 48.434,
+      "eval_steps_per_second": 1.521,
+      "step": 400
+    },
+    {
+      "epoch": 0.5131158029430583,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 5.3883495145631065e-05,
+      "loss": 1.6265,
+      "step": 401
+    },
+    {
+      "epoch": 0.5143953934740882,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.339805825242719e-05,
+      "loss": 1.6332,
+      "step": 402
+    },
+    {
+      "epoch": 0.5156749840051184,
+      "grad_norm": 0.205078125,
+      "learning_rate": 5.29126213592233e-05,
+      "loss": 1.6849,
+      "step": 403
+    },
+    {
+      "epoch": 0.5169545745361485,
+      "grad_norm": 0.228515625,
+      "learning_rate": 5.2427184466019416e-05,
+      "loss": 1.6917,
+      "step": 404
+    },
+    {
+      "epoch": 0.5182341650671785,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 5.194174757281554e-05,
+      "loss": 1.6444,
+      "step": 405
+    },
+    {
+      "epoch": 0.5195137555982086,
+      "grad_norm": 0.193359375,
+      "learning_rate": 5.145631067961165e-05,
+      "loss": 1.6165,
+      "step": 406
+    },
+    {
+      "epoch": 0.5207933461292387,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 5.0970873786407766e-05,
+      "loss": 1.6882,
+      "step": 407
+    },
+    {
+      "epoch": 0.5220729366602687,
+      "grad_norm": 0.19140625,
+      "learning_rate": 5.048543689320389e-05,
+      "loss": 1.65,
+      "step": 408
+    },
+    {
+      "epoch": 0.5233525271912988,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5e-05,
+      "loss": 1.6691,
+      "step": 409
+    },
+    {
+      "epoch": 0.5246321177223289,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 4.951456310679612e-05,
+      "loss": 1.6197,
+      "step": 410
+    },
+    {
+      "epoch": 0.525911708253359,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 4.902912621359224e-05,
+      "loss": 1.6359,
+      "step": 411
+    },
+    {
+      "epoch": 0.527191298784389,
+      "grad_norm": 0.177734375,
+      "learning_rate": 4.854368932038835e-05,
+      "loss": 1.6942,
+      "step": 412
+    },
+    {
+      "epoch": 0.5284708893154191,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 4.805825242718447e-05,
+      "loss": 1.6303,
+      "step": 413
+    },
+    {
+      "epoch": 0.5297504798464492,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 4.757281553398059e-05,
+      "loss": 1.6041,
+      "step": 414
+    },
+    {
+      "epoch": 0.5310300703774792,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 4.7087378640776703e-05,
+      "loss": 1.7167,
+      "step": 415
+    },
+    {
+      "epoch": 0.5323096609085093,
+      "grad_norm": 0.1953125,
+      "learning_rate": 4.660194174757282e-05,
+      "loss": 1.6828,
+      "step": 416
+    },
+    {
+      "epoch": 0.5335892514395394,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 4.611650485436894e-05,
+      "loss": 1.6386,
+      "step": 417
+    },
+    {
+      "epoch": 0.5348688419705694,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 4.5631067961165054e-05,
+      "loss": 1.6407,
+      "step": 418
+    },
+    {
+      "epoch": 0.5361484325015995,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 4.514563106796117e-05,
+      "loss": 1.6942,
+      "step": 419
+    },
+    {
+      "epoch": 0.5374280230326296,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 4.466019417475728e-05,
+      "loss": 1.6137,
+      "step": 420
+    },
+    {
+      "epoch": 0.5374280230326296,
+      "eval_loss": 1.634352684020996,
+      "eval_runtime": 103.4435,
+      "eval_samples_per_second": 48.336,
+      "eval_steps_per_second": 1.518,
+      "step": 420
+    },
+    {
+      "epoch": 0.5387076135636596,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 4.4174757281553404e-05,
+      "loss": 1.5895,
+      "step": 421
+    },
+    {
+      "epoch": 0.5399872040946897,
+      "grad_norm": 0.201171875,
+      "learning_rate": 4.368932038834951e-05,
+      "loss": 1.5607,
+      "step": 422
+    },
+    {
+      "epoch": 0.5412667946257198,
+      "grad_norm": 0.193359375,
+      "learning_rate": 4.3203883495145634e-05,
+      "loss": 1.6577,
+      "step": 423
+    },
+    {
+      "epoch": 0.5425463851567498,
+      "grad_norm": 0.20703125,
+      "learning_rate": 4.271844660194175e-05,
+      "loss": 1.6122,
+      "step": 424
+    },
+    {
+      "epoch": 0.5438259756877799,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 4.223300970873786e-05,
+      "loss": 1.6206,
+      "step": 425
+    },
+    {
+      "epoch": 0.54510556621881,
+      "grad_norm": 0.236328125,
+      "learning_rate": 4.1747572815533984e-05,
+      "loss": 1.6636,
+      "step": 426
+    },
+    {
+      "epoch": 0.54638515674984,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 4.12621359223301e-05,
+      "loss": 1.5838,
+      "step": 427
+    },
+    {
+      "epoch": 0.5476647472808701,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 4.077669902912621e-05,
+      "loss": 1.664,
+      "step": 428
+    },
+    {
+      "epoch": 0.5489443378119002,
+      "grad_norm": 0.1875,
+      "learning_rate": 4.029126213592233e-05,
+      "loss": 1.6365,
+      "step": 429
+    },
+    {
+      "epoch": 0.5502239283429302,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.980582524271845e-05,
+      "loss": 1.7319,
+      "step": 430
+    },
+    {
+      "epoch": 0.5515035188739603,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 3.9320388349514564e-05,
+      "loss": 1.5717,
+      "step": 431
+    },
+    {
+      "epoch": 0.5527831094049904,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.883495145631068e-05,
+      "loss": 1.5898,
+      "step": 432
+    },
+    {
+      "epoch": 0.5540626999360204,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 3.83495145631068e-05,
+      "loss": 1.6142,
+      "step": 433
+    },
+    {
+      "epoch": 0.5553422904670505,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.7864077669902914e-05,
+      "loss": 1.6619,
+      "step": 434
+    },
+    {
+      "epoch": 0.5566218809980806,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.737864077669903e-05,
+      "loss": 1.624,
+      "step": 435
+    },
+    {
+      "epoch": 0.5579014715291107,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 3.689320388349515e-05,
+      "loss": 1.657,
+      "step": 436
+    },
+    {
+      "epoch": 0.5591810620601407,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.6407766990291265e-05,
+      "loss": 1.6744,
+      "step": 437
+    },
+    {
+      "epoch": 0.5604606525911708,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 3.592233009708738e-05,
+      "loss": 1.6852,
+      "step": 438
+    },
+    {
+      "epoch": 0.5617402431222009,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 3.54368932038835e-05,
+      "loss": 1.5807,
+      "step": 439
+    },
+    {
+      "epoch": 0.5630198336532309,
+      "grad_norm": 0.193359375,
+      "learning_rate": 3.4951456310679615e-05,
+      "loss": 1.7031,
+      "step": 440
+    },
+    {
+      "epoch": 0.5630198336532309,
+      "eval_loss": 1.6339551210403442,
+      "eval_runtime": 103.2337,
+      "eval_samples_per_second": 48.434,
+      "eval_steps_per_second": 1.521,
+      "step": 440
+    },
+    {
+      "epoch": 0.564299424184261,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.446601941747573e-05,
+      "loss": 1.624,
+      "step": 441
+    },
+    {
+      "epoch": 0.5655790147152912,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 3.398058252427185e-05,
+      "loss": 1.5942,
+      "step": 442
+    },
+    {
+      "epoch": 0.5668586052463211,
+      "grad_norm": 0.189453125,
+      "learning_rate": 3.3495145631067966e-05,
+      "loss": 1.6723,
+      "step": 443
+    },
+    {
+      "epoch": 0.5681381957773513,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 3.300970873786408e-05,
+      "loss": 1.6573,
+      "step": 444
+    },
+    {
+      "epoch": 0.5694177863083814,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 3.2524271844660195e-05,
+      "loss": 1.6232,
+      "step": 445
+    },
+    {
+      "epoch": 0.5706973768394114,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 3.2038834951456316e-05,
+      "loss": 1.6724,
+      "step": 446
+    },
+    {
+      "epoch": 0.5719769673704415,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 3.155339805825243e-05,
+      "loss": 1.7238,
+      "step": 447
+    },
+    {
+      "epoch": 0.5732565579014716,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.1067961165048545e-05,
+      "loss": 1.6028,
+      "step": 448
+    },
+    {
+      "epoch": 0.5745361484325016,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.058252427184466e-05,
+      "loss": 1.7009,
+      "step": 449
+    },
+    {
+      "epoch": 0.5758157389635317,
+      "grad_norm": 0.19921875,
+      "learning_rate": 3.0097087378640774e-05,
+      "loss": 1.5935,
+      "step": 450
+    },
+    {
+      "epoch": 0.5770953294945618,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.9611650485436892e-05,
+      "loss": 1.5787,
+      "step": 451
+    },
+    {
+      "epoch": 0.5783749200255918,
+      "grad_norm": 0.1953125,
+      "learning_rate": 2.912621359223301e-05,
+      "loss": 1.6732,
+      "step": 452
+    },
+    {
+      "epoch": 0.5796545105566219,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.8640776699029125e-05,
+      "loss": 1.665,
+      "step": 453
+    },
+    {
+      "epoch": 0.580934101087652,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.8155339805825243e-05,
+      "loss": 1.6404,
+      "step": 454
+    },
+    {
+      "epoch": 0.582213691618682,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.766990291262136e-05,
+      "loss": 1.6541,
+      "step": 455
+    },
+    {
+      "epoch": 0.5834932821497121,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.7184466019417475e-05,
+      "loss": 1.7034,
+      "step": 456
+    },
+    {
+      "epoch": 0.5847728726807422,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.6699029126213593e-05,
+      "loss": 1.6074,
+      "step": 457
+    },
+    {
+      "epoch": 0.5860524632117722,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.6213592233009708e-05,
+      "loss": 1.6595,
+      "step": 458
+    },
+    {
+      "epoch": 0.5873320537428023,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.5728155339805826e-05,
+      "loss": 1.6631,
+      "step": 459
+    },
+    {
+      "epoch": 0.5886116442738324,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.5242718446601944e-05,
+      "loss": 1.62,
+      "step": 460
+    },
+    {
+      "epoch": 0.5886116442738324,
+      "eval_loss": 1.6338707208633423,
+      "eval_runtime": 103.2695,
+      "eval_samples_per_second": 48.417,
+      "eval_steps_per_second": 1.52,
+      "step": 460
+    },
+    {
+      "epoch": 0.5898912348048625,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.475728155339806e-05,
+      "loss": 1.7321,
+      "step": 461
+    },
+    {
+      "epoch": 0.5911708253358925,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.4271844660194176e-05,
+      "loss": 1.6152,
+      "step": 462
+    },
+    {
+      "epoch": 0.5924504158669226,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.3786407766990294e-05,
+      "loss": 1.6654,
+      "step": 463
+    },
+    {
+      "epoch": 0.5937300063979527,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 2.330097087378641e-05,
+      "loss": 1.6801,
+      "step": 464
+    },
+    {
+      "epoch": 0.5950095969289827,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.2815533980582527e-05,
+      "loss": 1.5644,
+      "step": 465
+    },
+    {
+      "epoch": 0.5962891874600128,
+      "grad_norm": 0.1953125,
+      "learning_rate": 2.233009708737864e-05,
+      "loss": 1.6238,
+      "step": 466
+    },
+    {
+      "epoch": 0.5975687779910429,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.1844660194174756e-05,
+      "loss": 1.6235,
+      "step": 467
+    },
+    {
+      "epoch": 0.5988483685220729,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.1359223300970874e-05,
+      "loss": 1.611,
+      "step": 468
+    },
+    {
+      "epoch": 0.600127959053103,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.0873786407766992e-05,
+      "loss": 1.6158,
+      "step": 469
+    },
+    {
+      "epoch": 0.6014075495841331,
+      "grad_norm": 0.1953125,
+      "learning_rate": 2.0388349514563107e-05,
+      "loss": 1.6569,
+      "step": 470
+    },
+    {
+      "epoch": 0.6026871401151631,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.9902912621359225e-05,
+      "loss": 1.7015,
+      "step": 471
+    },
+    {
+      "epoch": 0.6039667306461932,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.941747572815534e-05,
+      "loss": 1.6282,
+      "step": 472
+    },
+    {
+      "epoch": 0.6052463211772233,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.8932038834951457e-05,
+      "loss": 1.6254,
+      "step": 473
+    },
+    {
+      "epoch": 0.6065259117082533,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.8446601941747575e-05,
+      "loss": 1.6283,
+      "step": 474
+    },
+    {
+      "epoch": 0.6078055022392834,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.796116504854369e-05,
+      "loss": 1.6671,
+      "step": 475
+    },
+    {
+      "epoch": 0.6090850927703135,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.7475728155339808e-05,
+      "loss": 1.6847,
+      "step": 476
+    },
+    {
+      "epoch": 0.6103646833013435,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.6990291262135926e-05,
+      "loss": 1.6453,
+      "step": 477
+    },
+    {
+      "epoch": 0.6116442738323736,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.650485436893204e-05,
+      "loss": 1.5761,
+      "step": 478
+    },
+    {
+      "epoch": 0.6129238643634037,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.6019417475728158e-05,
+      "loss": 1.6311,
+      "step": 479
+    },
+    {
+      "epoch": 0.6142034548944337,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.5533980582524273e-05,
+      "loss": 1.6448,
+      "step": 480
+    },
+    {
+      "epoch": 0.6142034548944337,
+      "eval_loss": 1.6335575580596924,
+      "eval_runtime": 103.3386,
+      "eval_samples_per_second": 48.385,
+      "eval_steps_per_second": 1.519,
+      "step": 480
+    },
+    {
+      "epoch": 0.6154830454254638,
+      "grad_norm": 0.19921875,
+      "learning_rate": 1.5048543689320387e-05,
+      "loss": 1.6385,
+      "step": 481
+    },
+    {
+      "epoch": 0.6167626359564939,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 1.4563106796116505e-05,
+      "loss": 1.5795,
+      "step": 482
+    },
+    {
+      "epoch": 0.6180422264875239,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.4077669902912621e-05,
+      "loss": 1.5808,
+      "step": 483
+    },
+    {
+      "epoch": 0.619321817018554,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.3592233009708738e-05,
+      "loss": 1.592,
+      "step": 484
+    },
+    {
+      "epoch": 0.6206014075495841,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.3106796116504854e-05,
+      "loss": 1.6375,
+      "step": 485
+    },
+    {
+      "epoch": 0.6218809980806143,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.2621359223300972e-05,
+      "loss": 1.671,
+      "step": 486
+    },
+    {
+      "epoch": 0.6231605886116443,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.2135922330097088e-05,
+      "loss": 1.6573,
+      "step": 487
+    },
+    {
+      "epoch": 0.6244401791426744,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.1650485436893204e-05,
+      "loss": 1.6253,
+      "step": 488
+    },
+    {
+      "epoch": 0.6257197696737045,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 1.116504854368932e-05,
+      "loss": 1.6306,
+      "step": 489
+    },
+    {
+      "epoch": 0.6269993602047345,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.0679611650485437e-05,
+      "loss": 1.6724,
+      "step": 490
+    },
+    {
+      "epoch": 0.6282789507357646,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.0194174757281553e-05,
+      "loss": 1.5965,
+      "step": 491
+    },
+    {
+      "epoch": 0.6295585412667947,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 9.70873786407767e-06,
+      "loss": 1.6145,
+      "step": 492
+    },
+    {
+      "epoch": 0.6308381317978247,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.223300970873788e-06,
+      "loss": 1.5935,
+      "step": 493
+    },
+    {
+      "epoch": 0.6321177223288548,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 8.737864077669904e-06,
+      "loss": 1.5623,
+      "step": 494
+    },
+    {
+      "epoch": 0.6333973128598849,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 8.25242718446602e-06,
+      "loss": 1.648,
+      "step": 495
+    },
+    {
+      "epoch": 0.6346769033909149,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 7.766990291262136e-06,
+      "loss": 1.6144,
+      "step": 496
+    },
+    {
+      "epoch": 0.635956493921945,
+      "grad_norm": 0.201171875,
+      "learning_rate": 7.281553398058253e-06,
+      "loss": 1.6338,
+      "step": 497
+    },
+    {
+      "epoch": 0.6372360844529751,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.796116504854369e-06,
+      "loss": 1.7038,
+      "step": 498
+    },
+    {
+      "epoch": 0.6385156749840051,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 6.310679611650486e-06,
+      "loss": 1.6286,
+      "step": 499
+    },
+    {
+      "epoch": 0.6397952655150352,
+      "grad_norm": 0.1796875,
+      "learning_rate": 5.825242718446602e-06,
+      "loss": 1.7244,
+      "step": 500
+    },
+    {
+      "epoch": 0.6397952655150352,
+      "eval_loss": 1.633476734161377,
+      "eval_runtime": 103.4328,
+      "eval_samples_per_second": 48.341,
+      "eval_steps_per_second": 1.518,
+      "step": 500
+    },
+    {
+      "epoch": 0.6410748560460653,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 5.3398058252427185e-06,
+      "loss": 1.6388,
+      "step": 501
+    },
+    {
+      "epoch": 0.6423544465770953,
+      "grad_norm": 0.205078125,
+      "learning_rate": 4.854368932038835e-06,
+      "loss": 1.6222,
+      "step": 502
+    },
+    {
+      "epoch": 0.6436340371081254,
+      "grad_norm": 0.197265625,
+      "learning_rate": 4.368932038834952e-06,
+      "loss": 1.6958,
+      "step": 503
+    },
+    {
+      "epoch": 0.6449136276391555,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.883495145631068e-06,
+      "loss": 1.6074,
+      "step": 504
+    },
+    {
+      "epoch": 0.6461932181701855,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 3.3980582524271844e-06,
+      "loss": 1.6576,
+      "step": 505
+    },
+    {
+      "epoch": 0.6474728087012156,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.912621359223301e-06,
+      "loss": 1.634,
+      "step": 506
+    },
+    {
+      "epoch": 0.6487523992322457,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.4271844660194174e-06,
+      "loss": 1.6171,
+      "step": 507
+    },
+    {
+      "epoch": 0.6500319897632757,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.941747572815534e-06,
+      "loss": 1.613,
+      "step": 508
+    },
+    {
+      "epoch": 0.6513115802943058,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.4563106796116506e-06,
+      "loss": 1.612,
+      "step": 509
+    },
+    {
+      "epoch": 0.6525911708253359,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.70873786407767e-07,
+      "loss": 1.6429,
+      "step": 510
+    },
+    {
+      "epoch": 0.653870761356366,
+      "grad_norm": 0.193359375,
+      "learning_rate": 4.854368932038835e-07,
+      "loss": 1.6898,
+      "step": 511
+    },
+    {
+      "epoch": 0.655150351887396,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0,
+      "loss": 1.6705,
+      "step": 512
+    },
+    {
+      "epoch": 0.655150351887396,
+      "step": 512,
+      "total_flos": 3.639412992104202e+17,
+      "train_loss": 1.7040968828368932,
+      "train_runtime": 4680.8022,
+      "train_samples_per_second": 7.001,
+      "train_steps_per_second": 0.109
+    },
+    {
+      "epoch": 0.655150351887396,
+      "eval_loss": 1.633476734161377,
+      "eval_runtime": 103.3998,
+      "eval_samples_per_second": 48.356,
+      "eval_steps_per_second": 1.518,
+      "step": 512
     }
   ],
   "logging_steps": 1,
-  "max_steps": 256,
+  "max_steps": 512,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 1,
-  "save_steps": 10,
-  "total_flos": 1.8280284932328653e+17,
+  "save_steps": 20,
+  "total_flos": 3.639412992104202e+17,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null