diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4807 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9816849816849818,
+  "eval_steps": 500,
+  "global_step": 682,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0029304029304029304,
+      "grad_norm": 5.236616966747348,
+      "learning_rate": 3.3333333333333335e-07,
+      "loss": 1.6037,
+      "step": 1
+    },
+    {
+      "epoch": 0.005860805860805861,
+      "grad_norm": 4.277013116202448,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 1.5162,
+      "step": 2
+    },
+    {
+      "epoch": 0.008791208791208791,
+      "grad_norm": 5.329467381575862,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.5559,
+      "step": 3
+    },
+    {
+      "epoch": 0.011721611721611722,
+      "grad_norm": 4.6957386419713805,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 1.5187,
+      "step": 4
+    },
+    {
+      "epoch": 0.014652014652014652,
+      "grad_norm": 4.177469884648981,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.4766,
+      "step": 5
+    },
+    {
+      "epoch": 0.017582417582417582,
+      "grad_norm": 5.102737364259812,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.5765,
+      "step": 6
+    },
+    {
+      "epoch": 0.020512820512820513,
+      "grad_norm": 3.257410120888959,
+      "learning_rate": 2.3333333333333336e-06,
+      "loss": 1.4471,
+      "step": 7
+    },
+    {
+      "epoch": 0.023443223443223443,
+      "grad_norm": 3.115059473576919,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 1.3834,
+      "step": 8
+    },
+    {
+      "epoch": 0.026373626373626374,
+      "grad_norm": 2.1258574044969984,
+      "learning_rate": 3e-06,
+      "loss": 1.3919,
+      "step": 9
+    },
+    {
+      "epoch": 0.029304029304029304,
+      "grad_norm": 2.0531292894889805,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.3591,
+      "step": 10
+    },
+    {
+      "epoch": 0.03223443223443224,
+      "grad_norm": 2.06956179185827,
+      "learning_rate": 3.6666666666666666e-06,
+      "loss": 1.4242,
+      "step": 11
+    },
+    {
+      "epoch": 0.035164835164835165,
+      "grad_norm": 2.303837943491906,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.3092,
+      "step": 12
+    },
+    {
+      "epoch": 0.0380952380952381,
+      "grad_norm": 2.2605964849950344,
+      "learning_rate": 4.333333333333334e-06,
+      "loss": 1.2986,
+      "step": 13
+    },
+    {
+      "epoch": 0.041025641025641026,
+      "grad_norm": 1.9951383138249523,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 1.2942,
+      "step": 14
+    },
+    {
+      "epoch": 0.04395604395604396,
+      "grad_norm": 1.4898524610077184,
+      "learning_rate": 5e-06,
+      "loss": 1.2985,
+      "step": 15
+    },
+    {
+      "epoch": 0.046886446886446886,
+      "grad_norm": 1.3109439673909382,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 1.244,
+      "step": 16
+    },
+    {
+      "epoch": 0.04981684981684982,
+      "grad_norm": 1.1655623220122422,
+      "learning_rate": 5.666666666666667e-06,
+      "loss": 1.1808,
+      "step": 17
+    },
+    {
+      "epoch": 0.05274725274725275,
+      "grad_norm": 1.099398694116044,
+      "learning_rate": 6e-06,
+      "loss": 1.2938,
+      "step": 18
+    },
+    {
+      "epoch": 0.05567765567765568,
+      "grad_norm": 1.1377973094630334,
+      "learning_rate": 6.333333333333333e-06,
+      "loss": 1.2727,
+      "step": 19
+    },
+    {
+      "epoch": 0.05860805860805861,
+      "grad_norm": 0.9305342408980034,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.2282,
+      "step": 20
+    },
+    {
+      "epoch": 0.06153846153846154,
+      "grad_norm": 0.8573945958738098,
+      "learning_rate": 7e-06,
+      "loss": 1.2088,
+      "step": 21
+    },
+    {
+      "epoch": 0.06446886446886448,
+      "grad_norm": 0.9176308767455088,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 1.213,
+      "step": 22
+    },
+    {
+      "epoch": 0.0673992673992674,
+      "grad_norm": 1.0602477117511504,
+      "learning_rate": 7.666666666666667e-06,
+      "loss": 1.256,
+      "step": 23
+    },
+    {
+      "epoch": 0.07032967032967033,
+      "grad_norm": 0.9793994066565616,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.197,
+      "step": 24
+    },
+    {
+      "epoch": 0.07326007326007326,
+      "grad_norm": 1.0029818180142296,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.137,
+      "step": 25
+    },
+    {
+      "epoch": 0.0761904761904762,
+      "grad_norm": 1.0195465294285324,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 1.2386,
+      "step": 26
+    },
+    {
+      "epoch": 0.07912087912087912,
+      "grad_norm": 1.0759683772319115,
+      "learning_rate": 9e-06,
+      "loss": 1.2376,
+      "step": 27
+    },
+    {
+      "epoch": 0.08205128205128205,
+      "grad_norm": 0.8911061397225324,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 1.1831,
+      "step": 28
+    },
+    {
+      "epoch": 0.08498168498168499,
+      "grad_norm": 0.861354167362915,
+      "learning_rate": 9.666666666666667e-06,
+      "loss": 1.1511,
+      "step": 29
+    },
+    {
+      "epoch": 0.08791208791208792,
+      "grad_norm": 0.8881533584446896,
+      "learning_rate": 1e-05,
+      "loss": 1.228,
+      "step": 30
+    },
+    {
+      "epoch": 0.09084249084249084,
+      "grad_norm": 0.8512130063857013,
+      "learning_rate": 9.999941957819112e-06,
+      "loss": 1.2142,
+      "step": 31
+    },
+    {
+      "epoch": 0.09377289377289377,
+      "grad_norm": 0.880628181688528,
+      "learning_rate": 9.999767832624e-06,
+      "loss": 1.2078,
+      "step": 32
+    },
+    {
+      "epoch": 0.0967032967032967,
+      "grad_norm": 0.8855214361195558,
+      "learning_rate": 9.99947762845731e-06,
+      "loss": 1.2384,
+      "step": 33
+    },
+    {
+      "epoch": 0.09963369963369964,
+      "grad_norm": 0.9160758544701845,
+      "learning_rate": 9.999071352056676e-06,
+      "loss": 1.1859,
+      "step": 34
+    },
+    {
+      "epoch": 0.10256410256410256,
+      "grad_norm": 0.8031365392857486,
+      "learning_rate": 9.998549012854562e-06,
+      "loss": 1.1149,
+      "step": 35
+    },
+    {
+      "epoch": 0.1054945054945055,
+      "grad_norm": 0.7882701972390256,
+      "learning_rate": 9.99791062297805e-06,
+      "loss": 1.1174,
+      "step": 36
+    },
+    {
+      "epoch": 0.10842490842490843,
+      "grad_norm": 0.8172216850403872,
+      "learning_rate": 9.997156197248564e-06,
+      "loss": 1.2042,
+      "step": 37
+    },
+    {
+      "epoch": 0.11135531135531136,
+      "grad_norm": 0.9084324585899033,
+      "learning_rate": 9.996285753181499e-06,
+      "loss": 1.1251,
+      "step": 38
+    },
+    {
+      "epoch": 0.11428571428571428,
+      "grad_norm": 0.8107056469020013,
+      "learning_rate": 9.99529931098585e-06,
+      "loss": 1.1847,
+      "step": 39
+    },
+    {
+      "epoch": 0.11721611721611722,
+      "grad_norm": 0.8449584359735586,
+      "learning_rate": 9.994196893563722e-06,
+      "loss": 1.1505,
+      "step": 40
+    },
+    {
+      "epoch": 0.12014652014652015,
+      "grad_norm": 0.8831422669635972,
+      "learning_rate": 9.992978526509794e-06,
+      "loss": 1.1699,
+      "step": 41
+    },
+    {
+      "epoch": 0.12307692307692308,
+      "grad_norm": 0.8644767120795944,
+      "learning_rate": 9.991644238110741e-06,
+      "loss": 1.2478,
+      "step": 42
+    },
+    {
+      "epoch": 0.126007326007326,
+      "grad_norm": 0.9482277505129767,
+      "learning_rate": 9.990194059344565e-06,
+      "loss": 1.2268,
+      "step": 43
+    },
+    {
+      "epoch": 0.12893772893772895,
+      "grad_norm": 0.8617016918844874,
+      "learning_rate": 9.988628023879883e-06,
+      "loss": 1.1817,
+      "step": 44
+    },
+    {
+      "epoch": 0.13186813186813187,
+      "grad_norm": 0.8061790018852762,
+      "learning_rate": 9.98694616807514e-06,
+      "loss": 1.1482,
+      "step": 45
+    },
+    {
+      "epoch": 0.1347985347985348,
+      "grad_norm": 0.7832884251827952,
+      "learning_rate": 9.985148530977767e-06,
+      "loss": 1.1317,
+      "step": 46
+    },
+    {
+      "epoch": 0.13772893772893774,
+      "grad_norm": 0.9771610956162755,
+      "learning_rate": 9.983235154323276e-06,
+      "loss": 1.2133,
+      "step": 47
+    },
+    {
+      "epoch": 0.14065934065934066,
+      "grad_norm": 0.6863448092272799,
+      "learning_rate": 9.981206082534287e-06,
+      "loss": 1.0658,
+      "step": 48
+    },
+    {
+      "epoch": 0.14358974358974358,
+      "grad_norm": 0.7171230395493451,
+      "learning_rate": 9.979061362719502e-06,
+      "loss": 1.0827,
+      "step": 49
+    },
+    {
+      "epoch": 0.14652014652014653,
+      "grad_norm": 0.8336856657882589,
+      "learning_rate": 9.976801044672608e-06,
+      "loss": 1.1807,
+      "step": 50
+    },
+    {
+      "epoch": 0.14945054945054945,
+      "grad_norm": 0.7702137284348907,
+      "learning_rate": 9.974425180871118e-06,
+      "loss": 1.0692,
+      "step": 51
+    },
+    {
+      "epoch": 0.1523809523809524,
+      "grad_norm": 0.7532169814888361,
+      "learning_rate": 9.971933826475162e-06,
+      "loss": 1.1026,
+      "step": 52
+    },
+    {
+      "epoch": 0.15531135531135531,
+      "grad_norm": 0.6885436983052671,
+      "learning_rate": 9.969327039326193e-06,
+      "loss": 1.1867,
+      "step": 53
+    },
+    {
+      "epoch": 0.15824175824175823,
+      "grad_norm": 0.8133906289524008,
+      "learning_rate": 9.966604879945659e-06,
+      "loss": 1.1775,
+      "step": 54
+    },
+    {
+      "epoch": 0.16117216117216118,
+      "grad_norm": 0.7968024513483275,
+      "learning_rate": 9.963767411533584e-06,
+      "loss": 1.113,
+      "step": 55
+    },
+    {
+      "epoch": 0.1641025641025641,
+      "grad_norm": 0.8313193163815186,
+      "learning_rate": 9.960814699967112e-06,
+      "loss": 1.1637,
+      "step": 56
+    },
+    {
+      "epoch": 0.16703296703296702,
+      "grad_norm": 0.9517211139026077,
+      "learning_rate": 9.95774681379897e-06,
+      "loss": 1.134,
+      "step": 57
+    },
+    {
+      "epoch": 0.16996336996336997,
+      "grad_norm": 0.8301120009736043,
+      "learning_rate": 9.954563824255879e-06,
+      "loss": 1.1291,
+      "step": 58
+    },
+    {
+      "epoch": 0.1728937728937729,
+      "grad_norm": 0.8270284845492123,
+      "learning_rate": 9.951265805236903e-06,
+      "loss": 1.1349,
+      "step": 59
+    },
+    {
+      "epoch": 0.17582417582417584,
+      "grad_norm": 0.7017311057617309,
+      "learning_rate": 9.947852833311725e-06,
+      "loss": 1.1389,
+      "step": 60
+    },
+    {
+      "epoch": 0.17875457875457876,
+      "grad_norm": 0.7621669276766957,
+      "learning_rate": 9.94432498771888e-06,
+      "loss": 1.1036,
+      "step": 61
+    },
+    {
+      "epoch": 0.18168498168498168,
+      "grad_norm": 0.8320291744185084,
+      "learning_rate": 9.940682350363913e-06,
+      "loss": 1.0958,
+      "step": 62
+    },
+    {
+      "epoch": 0.18461538461538463,
+      "grad_norm": 0.8835787608753064,
+      "learning_rate": 9.936925005817464e-06,
+      "loss": 1.1471,
+      "step": 63
+    },
+    {
+      "epoch": 0.18754578754578755,
+      "grad_norm": 0.749274429167846,
+      "learning_rate": 9.933053041313325e-06,
+      "loss": 1.1325,
+      "step": 64
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 0.7262497057804643,
+      "learning_rate": 9.929066546746403e-06,
+      "loss": 1.1456,
+      "step": 65
+    },
+    {
+      "epoch": 0.1934065934065934,
+      "grad_norm": 0.8670981015624153,
+      "learning_rate": 9.924965614670629e-06,
+      "loss": 1.138,
+      "step": 66
+    },
+    {
+      "epoch": 0.19633699633699633,
+      "grad_norm": 0.8432940149428723,
+      "learning_rate": 9.920750340296823e-06,
+      "loss": 1.1176,
+      "step": 67
+    },
+    {
+      "epoch": 0.19926739926739928,
+      "grad_norm": 0.7312143321816789,
+      "learning_rate": 9.916420821490474e-06,
+      "loss": 1.1434,
+      "step": 68
+    },
+    {
+      "epoch": 0.2021978021978022,
+      "grad_norm": 0.7605667410112772,
+      "learning_rate": 9.911977158769461e-06,
+      "loss": 1.1023,
+      "step": 69
+    },
+    {
+      "epoch": 0.20512820512820512,
+      "grad_norm": 0.803430465139554,
+      "learning_rate": 9.90741945530174e-06,
+      "loss": 1.1953,
+      "step": 70
+    },
+    {
+      "epoch": 0.20805860805860807,
+      "grad_norm": 0.9220290395479969,
+      "learning_rate": 9.90274781690293e-06,
+      "loss": 1.1115,
+      "step": 71
+    },
+    {
+      "epoch": 0.210989010989011,
+      "grad_norm": 0.8183474218674964,
+      "learning_rate": 9.89796235203386e-06,
+      "loss": 1.1527,
+      "step": 72
+    },
+    {
+      "epoch": 0.2139194139194139,
+      "grad_norm": 0.7126405845950673,
+      "learning_rate": 9.89306317179806e-06,
+      "loss": 1.0852,
+      "step": 73
+    },
+    {
+      "epoch": 0.21684981684981686,
+      "grad_norm": 0.7601543510464269,
+      "learning_rate": 9.888050389939172e-06,
+      "loss": 1.1428,
+      "step": 74
+    },
+    {
+      "epoch": 0.21978021978021978,
+      "grad_norm": 0.8916999196971415,
+      "learning_rate": 9.882924122838311e-06,
+      "loss": 1.074,
+      "step": 75
+    },
+    {
+      "epoch": 0.22271062271062272,
+      "grad_norm": 0.8688542443917322,
+      "learning_rate": 9.877684489511367e-06,
+      "loss": 1.0831,
+      "step": 76
+    },
+    {
+      "epoch": 0.22564102564102564,
+      "grad_norm": 0.8488868510757361,
+      "learning_rate": 9.872331611606238e-06,
+      "loss": 1.1393,
+      "step": 77
+    },
+    {
+      "epoch": 0.22857142857142856,
+      "grad_norm": 0.7370262143749565,
+      "learning_rate": 9.866865613400008e-06,
+      "loss": 1.0948,
+      "step": 78
+    },
+    {
+      "epoch": 0.2315018315018315,
+      "grad_norm": 0.7505253317469636,
+      "learning_rate": 9.861286621796056e-06,
+      "loss": 1.1462,
+      "step": 79
+    },
+    {
+      "epoch": 0.23443223443223443,
+      "grad_norm": 0.9094343725835037,
+      "learning_rate": 9.855594766321122e-06,
+      "loss": 1.1325,
+      "step": 80
+    },
+    {
+      "epoch": 0.23736263736263735,
+      "grad_norm": 0.7886762129451516,
+      "learning_rate": 9.849790179122287e-06,
+      "loss": 1.1147,
+      "step": 81
+    },
+    {
+      "epoch": 0.2402930402930403,
+      "grad_norm": 0.7613394543040073,
+      "learning_rate": 9.843872994963912e-06,
+      "loss": 1.0538,
+      "step": 82
+    },
+    {
+      "epoch": 0.24322344322344322,
+      "grad_norm": 0.754376414929907,
+      "learning_rate": 9.837843351224504e-06,
+      "loss": 1.0924,
+      "step": 83
+    },
+    {
+      "epoch": 0.24615384615384617,
+      "grad_norm": 0.8764280950179428,
+      "learning_rate": 9.831701387893533e-06,
+      "loss": 1.1504,
+      "step": 84
+    },
+    {
+      "epoch": 0.2490842490842491,
+      "grad_norm": 0.695623410235394,
+      "learning_rate": 9.825447247568177e-06,
+      "loss": 1.1245,
+      "step": 85
+    },
+    {
+      "epoch": 0.252014652014652,
+      "grad_norm": 0.6969634561214898,
+      "learning_rate": 9.819081075450014e-06,
+      "loss": 1.0707,
+      "step": 86
+    },
+    {
+      "epoch": 0.2549450549450549,
+      "grad_norm": 0.7156303735303894,
+      "learning_rate": 9.812603019341652e-06,
+      "loss": 1.1286,
+      "step": 87
+    },
+    {
+      "epoch": 0.2578754578754579,
+      "grad_norm": 0.8220793729909976,
+      "learning_rate": 9.80601322964329e-06,
+      "loss": 1.1302,
+      "step": 88
+    },
+    {
+      "epoch": 0.2608058608058608,
+      "grad_norm": 0.9670245180950947,
+      "learning_rate": 9.799311859349235e-06,
+      "loss": 1.122,
+      "step": 89
+    },
+    {
+      "epoch": 0.26373626373626374,
+      "grad_norm": 0.7021589814665796,
+      "learning_rate": 9.792499064044343e-06,
+      "loss": 1.0646,
+      "step": 90
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 0.7446117525493868,
+      "learning_rate": 9.785575001900418e-06,
+      "loss": 1.1038,
+      "step": 91
+    },
+    {
+      "epoch": 0.2695970695970696,
+      "grad_norm": 0.7963884145520905,
+      "learning_rate": 9.778539833672525e-06,
+      "loss": 1.1319,
+      "step": 92
+    },
+    {
+      "epoch": 0.2725274725274725,
+      "grad_norm": 0.8173788201228416,
+      "learning_rate": 9.771393722695267e-06,
+      "loss": 1.1246,
+      "step": 93
+    },
+    {
+      "epoch": 0.2754578754578755,
+      "grad_norm": 0.740082695855197,
+      "learning_rate": 9.764136834878987e-06,
+      "loss": 1.1088,
+      "step": 94
+    },
+    {
+      "epoch": 0.2783882783882784,
+      "grad_norm": 0.7467817240548424,
+      "learning_rate": 9.756769338705927e-06,
+      "loss": 1.0909,
+      "step": 95
+    },
+    {
+      "epoch": 0.2813186813186813,
+      "grad_norm": 0.7619316456122908,
+      "learning_rate": 9.749291405226304e-06,
+      "loss": 1.1233,
+      "step": 96
+    },
+    {
+      "epoch": 0.28424908424908424,
+      "grad_norm": 0.6936244573726563,
+      "learning_rate": 9.741703208054347e-06,
+      "loss": 1.0595,
+      "step": 97
+    },
+    {
+      "epoch": 0.28717948717948716,
+      "grad_norm": 0.8077496531848941,
+      "learning_rate": 9.734004923364258e-06,
+      "loss": 1.1384,
+      "step": 98
+    },
+    {
+      "epoch": 0.29010989010989013,
+      "grad_norm": 0.705333491953162,
+      "learning_rate": 9.72619672988613e-06,
+      "loss": 1.1248,
+      "step": 99
+    },
+    {
+      "epoch": 0.29304029304029305,
+      "grad_norm": 0.7150948818560614,
+      "learning_rate": 9.718278808901797e-06,
+      "loss": 1.0189,
+      "step": 100
+    },
+    {
+      "epoch": 0.295970695970696,
+      "grad_norm": 0.8190821657116661,
+      "learning_rate": 9.710251344240619e-06,
+      "loss": 1.1314,
+      "step": 101
+    },
+    {
+      "epoch": 0.2989010989010989,
+      "grad_norm": 0.7993491005692807,
+      "learning_rate": 9.702114522275216e-06,
+      "loss": 1.1379,
+      "step": 102
+    },
+    {
+      "epoch": 0.3018315018315018,
+      "grad_norm": 0.77941047322087,
+      "learning_rate": 9.69386853191715e-06,
+      "loss": 1.0411,
+      "step": 103
+    },
+    {
+      "epoch": 0.3047619047619048,
+      "grad_norm": 0.7858691393792162,
+      "learning_rate": 9.685513564612521e-06,
+      "loss": 1.1251,
+      "step": 104
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.7337514917747404,
+      "learning_rate": 9.67704981433754e-06,
+      "loss": 1.1191,
+      "step": 105
+    },
+    {
+      "epoch": 0.31062271062271063,
+      "grad_norm": 0.767317429310955,
+      "learning_rate": 9.668477477594021e-06,
+      "loss": 1.093,
+      "step": 106
+    },
+    {
+      "epoch": 0.31355311355311355,
+      "grad_norm": 0.7668402066152609,
+      "learning_rate": 9.659796753404806e-06,
+      "loss": 1.0853,
+      "step": 107
+    },
+    {
+      "epoch": 0.31648351648351647,
+      "grad_norm": 0.7172206747621077,
+      "learning_rate": 9.651007843309164e-06,
+      "loss": 1.0838,
+      "step": 108
+    },
+    {
+      "epoch": 0.3194139194139194,
+      "grad_norm": 0.7579043380498024,
+      "learning_rate": 9.642110951358097e-06,
+      "loss": 1.1491,
+      "step": 109
+    },
+    {
+      "epoch": 0.32234432234432236,
+      "grad_norm": 0.7733555528433625,
+      "learning_rate": 9.633106284109612e-06,
+      "loss": 1.1435,
+      "step": 110
+    },
+    {
+      "epoch": 0.3252747252747253,
+      "grad_norm": 3.3284789531656016,
+      "learning_rate": 9.623994050623917e-06,
+      "loss": 1.0951,
+      "step": 111
+    },
+    {
+      "epoch": 0.3282051282051282,
+      "grad_norm": 0.7820590656550864,
+      "learning_rate": 9.614774462458573e-06,
+      "loss": 1.1173,
+      "step": 112
+    },
+    {
+      "epoch": 0.3311355311355311,
+      "grad_norm": 0.7056629851514702,
+      "learning_rate": 9.605447733663586e-06,
+      "loss": 1.1012,
+      "step": 113
+    },
+    {
+      "epoch": 0.33406593406593404,
+      "grad_norm": 0.7628154605662072,
+      "learning_rate": 9.596014080776424e-06,
+      "loss": 1.1192,
+      "step": 114
+    },
+    {
+      "epoch": 0.336996336996337,
+      "grad_norm": 0.6924942007287507,
+      "learning_rate": 9.586473722817002e-06,
+      "loss": 1.0647,
+      "step": 115
+    },
+    {
+      "epoch": 0.33992673992673994,
+      "grad_norm": 0.7030223107571508,
+      "learning_rate": 9.576826881282595e-06,
+      "loss": 1.1056,
+      "step": 116
+    },
+    {
+      "epoch": 0.34285714285714286,
+      "grad_norm": 0.7341686688600246,
+      "learning_rate": 9.56707378014269e-06,
+      "loss": 1.0371,
+      "step": 117
+    },
+    {
+      "epoch": 0.3457875457875458,
+      "grad_norm": 0.7727927029586001,
+      "learning_rate": 9.557214645833792e-06,
+      "loss": 1.1259,
+      "step": 118
+    },
+    {
+      "epoch": 0.3487179487179487,
+      "grad_norm": 0.7310833198855431,
+      "learning_rate": 9.547249707254166e-06,
+      "loss": 1.0888,
+      "step": 119
+    },
+    {
+      "epoch": 0.3516483516483517,
+      "grad_norm": 0.6667796341322038,
+      "learning_rate": 9.537179195758513e-06,
+      "loss": 1.0823,
+      "step": 120
+    },
+    {
+      "epoch": 0.3545787545787546,
+      "grad_norm": 0.665405837005823,
+      "learning_rate": 9.527003345152616e-06,
+      "loss": 1.0777,
+      "step": 121
+    },
+    {
+      "epoch": 0.3575091575091575,
+      "grad_norm": 0.6809160644579101,
+      "learning_rate": 9.516722391687903e-06,
+      "loss": 1.053,
+      "step": 122
+    },
+    {
+      "epoch": 0.36043956043956044,
+      "grad_norm": 0.728118239149536,
+      "learning_rate": 9.506336574055954e-06,
+      "loss": 1.1185,
+      "step": 123
+    },
+    {
+      "epoch": 0.36336996336996336,
+      "grad_norm": 0.7065965453050594,
+      "learning_rate": 9.495846133382973e-06,
+      "loss": 1.1062,
+      "step": 124
+    },
+    {
+      "epoch": 0.3663003663003663,
+      "grad_norm": 0.7014383586245783,
+      "learning_rate": 9.485251313224181e-06,
+      "loss": 1.0878,
+      "step": 125
+    },
+    {
+      "epoch": 0.36923076923076925,
+      "grad_norm": 0.681702034528789,
+      "learning_rate": 9.474552359558167e-06,
+      "loss": 1.0704,
+      "step": 126
+    },
+    {
+      "epoch": 0.37216117216117217,
+      "grad_norm": 0.7187395064100738,
+      "learning_rate": 9.463749520781172e-06,
+      "loss": 1.0812,
+      "step": 127
+    },
+    {
+      "epoch": 0.3750915750915751,
+      "grad_norm": 0.7184519689161065,
+      "learning_rate": 9.452843047701324e-06,
+      "loss": 1.1057,
+      "step": 128
+    },
+    {
+      "epoch": 0.378021978021978,
+      "grad_norm": 0.7690003114746435,
+      "learning_rate": 9.441833193532817e-06,
+      "loss": 1.0689,
+      "step": 129
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 0.7164964859738834,
+      "learning_rate": 9.43072021389003e-06,
+      "loss": 1.117,
+      "step": 130
+    },
+    {
+      "epoch": 0.3838827838827839,
+      "grad_norm": 0.6974378217023173,
+      "learning_rate": 9.419504366781591e-06,
+      "loss": 1.0734,
+      "step": 131
+    },
+    {
+      "epoch": 0.3868131868131868,
+      "grad_norm": 0.7254779614689418,
+      "learning_rate": 9.408185912604395e-06,
+      "loss": 1.0559,
+      "step": 132
+    },
+    {
+      "epoch": 0.38974358974358975,
+      "grad_norm": 0.7303990545301333,
+      "learning_rate": 9.396765114137544e-06,
+      "loss": 1.0918,
+      "step": 133
+    },
+    {
+      "epoch": 0.39267399267399267,
+      "grad_norm": 0.6651968867581742,
+      "learning_rate": 9.385242236536259e-06,
+      "loss": 1.078,
+      "step": 134
+    },
+    {
+      "epoch": 0.3956043956043956,
+      "grad_norm": 0.6364456851581658,
+      "learning_rate": 9.37361754732572e-06,
+      "loss": 1.0218,
+      "step": 135
+    },
+    {
+      "epoch": 0.39853479853479856,
+      "grad_norm": 0.8125339033682204,
+      "learning_rate": 9.36189131639485e-06,
+      "loss": 1.1092,
+      "step": 136
+    },
+    {
+      "epoch": 0.4014652014652015,
+      "grad_norm": 0.7445751488678471,
+      "learning_rate": 9.35006381599006e-06,
+      "loss": 1.1637,
+      "step": 137
+    },
+    {
+      "epoch": 0.4043956043956044,
+      "grad_norm": 0.7911370318492177,
+      "learning_rate": 9.338135320708912e-06,
+      "loss": 1.1167,
+      "step": 138
+    },
+    {
+      "epoch": 0.4073260073260073,
+      "grad_norm": 0.7108436127581649,
+      "learning_rate": 9.326106107493762e-06,
+      "loss": 1.0756,
+      "step": 139
+    },
+    {
+      "epoch": 0.41025641025641024,
+      "grad_norm": 0.6595709769127521,
+      "learning_rate": 9.313976455625316e-06,
+      "loss": 1.0266,
+      "step": 140
+    },
+    {
+      "epoch": 0.41318681318681316,
+      "grad_norm": 0.7843564857029468,
+      "learning_rate": 9.301746646716152e-06,
+      "loss": 1.1095,
+      "step": 141
+    },
+    {
+      "epoch": 0.41611721611721614,
+      "grad_norm": 0.7585560414429611,
+      "learning_rate": 9.289416964704186e-06,
+      "loss": 1.1141,
+      "step": 142
+    },
+    {
+      "epoch": 0.41904761904761906,
+      "grad_norm": 0.7573484270766634,
+      "learning_rate": 9.27698769584607e-06,
+      "loss": 1.0725,
+      "step": 143
+    },
+    {
+      "epoch": 0.421978021978022,
+      "grad_norm": 0.8449861185856159,
+      "learning_rate": 9.26445912871055e-06,
+      "loss": 1.1368,
+      "step": 144
+    },
+    {
+      "epoch": 0.4249084249084249,
+      "grad_norm": 0.8534050421722909,
+      "learning_rate": 9.251831554171773e-06,
+      "loss": 1.0788,
+      "step": 145
+    },
+    {
+      "epoch": 0.4278388278388278,
+      "grad_norm": 0.7638318082003831,
+      "learning_rate": 9.239105265402525e-06,
+      "loss": 1.1393,
+      "step": 146
+    },
+    {
+      "epoch": 0.4307692307692308,
+      "grad_norm": 0.830680728213633,
+      "learning_rate": 9.226280557867428e-06,
+      "loss": 1.1028,
+      "step": 147
+    },
+    {
+      "epoch": 0.4336996336996337,
+      "grad_norm": 0.6744236125985169,
+      "learning_rate": 9.213357729316077e-06,
+      "loss": 1.0371,
+      "step": 148
+    },
+    {
+      "epoch": 0.43663003663003663,
+      "grad_norm": 0.7708023870173715,
+      "learning_rate": 9.200337079776136e-06,
+      "loss": 1.0292,
+      "step": 149
+    },
+    {
+      "epoch": 0.43956043956043955,
+      "grad_norm": 0.7392502681190622,
+      "learning_rate": 9.187218911546363e-06,
+      "loss": 1.1144,
+      "step": 150
+    },
+    {
+      "epoch": 0.4424908424908425,
+      "grad_norm": 0.8223130697058154,
+      "learning_rate": 9.174003529189592e-06,
+      "loss": 1.1267,
+      "step": 151
+    },
+    {
+      "epoch": 0.44542124542124545,
+      "grad_norm": 0.6534093776081946,
+      "learning_rate": 9.160691239525675e-06,
+      "loss": 1.0506,
+      "step": 152
+    },
+    {
+      "epoch": 0.44835164835164837,
+      "grad_norm": 0.7725854247329561,
+      "learning_rate": 9.147282351624334e-06,
+      "loss": 1.0608,
+      "step": 153
+    },
+    {
+      "epoch": 0.4512820512820513,
+      "grad_norm": 0.6816968972584797,
+      "learning_rate": 9.133777176798013e-06,
+      "loss": 1.0555,
+      "step": 154
+    },
+    {
+      "epoch": 0.4542124542124542,
+      "grad_norm": 0.9669303786054149,
+      "learning_rate": 9.120176028594632e-06,
+      "loss": 1.1101,
+      "step": 155
+    },
+    {
+      "epoch": 0.45714285714285713,
+      "grad_norm": 0.7037824443977663,
+      "learning_rate": 9.106479222790312e-06,
+      "loss": 1.0933,
+      "step": 156
+    },
+    {
+      "epoch": 0.46007326007326005,
+      "grad_norm": 0.8568980072792112,
+      "learning_rate": 9.092687077382043e-06,
+      "loss": 1.0872,
+      "step": 157
+    },
+    {
+      "epoch": 0.463003663003663,
+      "grad_norm": 1.0093527220046434,
+      "learning_rate": 9.078799912580305e-06,
+      "loss": 1.0833,
+      "step": 158
+    },
+    {
+      "epoch": 0.46593406593406594,
+      "grad_norm": 0.8145192278685842,
+      "learning_rate": 9.064818050801634e-06,
+      "loss": 1.0981,
+      "step": 159
+    },
+    {
+      "epoch": 0.46886446886446886,
+      "grad_norm": 0.7087077852924482,
+      "learning_rate": 9.050741816661128e-06,
+      "loss": 1.0728,
+      "step": 160
+    },
+    {
+      "epoch": 0.4717948717948718,
+      "grad_norm": 0.8436314316973053,
+      "learning_rate": 9.036571536964915e-06,
+      "loss": 1.1254,
+      "step": 161
+    },
+    {
+      "epoch": 0.4747252747252747,
+      "grad_norm": 0.8154298111795745,
+      "learning_rate": 9.022307540702576e-06,
+      "loss": 1.1122,
+      "step": 162
+    },
+    {
+      "epoch": 0.4776556776556777,
+      "grad_norm": 0.6530298783851548,
+      "learning_rate": 9.007950159039488e-06,
+      "loss": 1.0955,
+      "step": 163
+    },
+    {
+      "epoch": 0.4805860805860806,
+      "grad_norm": 0.7845249406652348,
+      "learning_rate": 8.993499725309148e-06,
+      "loss": 1.051,
+      "step": 164
+    },
+    {
+      "epoch": 0.4835164835164835,
+      "grad_norm": 0.8429786879957137,
+      "learning_rate": 8.978956575005432e-06,
+      "loss": 1.1001,
+      "step": 165
+    },
+    {
+      "epoch": 0.48644688644688644,
+      "grad_norm": 0.7070651182594344,
+      "learning_rate": 8.964321045774808e-06,
+      "loss": 1.078,
+      "step": 166
+    },
+    {
+      "epoch": 0.48937728937728936,
+      "grad_norm": 0.7491716944247487,
+      "learning_rate": 8.949593477408485e-06,
+      "loss": 1.0743,
+      "step": 167
+    },
+    {
+      "epoch": 0.49230769230769234,
+      "grad_norm": 0.8088630208896083,
+      "learning_rate": 8.934774211834538e-06,
+      "loss": 1.0151,
+      "step": 168
+    },
+    {
+      "epoch": 0.49523809523809526,
+      "grad_norm": 0.7439375920034751,
+      "learning_rate": 8.919863593109967e-06,
+      "loss": 1.0513,
+      "step": 169
+    },
+    {
+      "epoch": 0.4981684981684982,
+      "grad_norm": 0.815503037318965,
+      "learning_rate": 8.904861967412702e-06,
+      "loss": 1.0963,
+      "step": 170
+    },
+    {
+      "epoch": 0.5010989010989011,
+      "grad_norm": 0.8588369274428411,
+      "learning_rate": 8.889769683033573e-06,
+      "loss": 1.0378,
+      "step": 171
+    },
+    {
+      "epoch": 0.504029304029304,
+      "grad_norm": 0.7890693152650229,
+      "learning_rate": 8.874587090368221e-06,
+      "loss": 1.0849,
+      "step": 172
+    },
+    {
+      "epoch": 0.5069597069597069,
+      "grad_norm": 0.679463169659677,
+      "learning_rate": 8.859314541908957e-06,
+      "loss": 1.0481,
+      "step": 173
+    },
+    {
+      "epoch": 0.5098901098901099,
+      "grad_norm": 0.7507974578994663,
+      "learning_rate": 8.843952392236595e-06,
+      "loss": 1.0806,
+      "step": 174
+    },
+    {
+      "epoch": 0.5128205128205128,
+      "grad_norm": 0.7459226590677729,
+      "learning_rate": 8.8285009980122e-06,
+      "loss": 1.0841,
+      "step": 175
+    },
+    {
+      "epoch": 0.5157509157509158,
+      "grad_norm": 0.7750302070712438,
+      "learning_rate": 8.81296071796882e-06,
+      "loss": 1.0865,
+      "step": 176
+    },
+    {
+      "epoch": 0.5186813186813187,
+      "grad_norm": 0.7147033919188667,
+      "learning_rate": 8.797331912903152e-06,
+      "loss": 1.0832,
+      "step": 177
+    },
+    {
+      "epoch": 0.5216117216117216,
+      "grad_norm": 0.6594455064471973,
+      "learning_rate": 8.78161494566717e-06,
+      "loss": 1.0843,
+      "step": 178
+    },
+    {
+      "epoch": 0.5245421245421246,
+      "grad_norm": 0.719823747496003,
+      "learning_rate": 8.765810181159696e-06,
+      "loss": 1.0775,
+      "step": 179
+    },
+    {
+      "epoch": 0.5274725274725275,
+      "grad_norm": 0.6848428843395724,
+      "learning_rate": 8.74991798631793e-06,
+      "loss": 1.0865,
+      "step": 180
+    },
+    {
+      "epoch": 0.5304029304029304,
+      "grad_norm": 0.6724971412658745,
+      "learning_rate": 8.733938730108928e-06,
+      "loss": 1.061,
+      "step": 181
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 0.6900267089498971,
+      "learning_rate": 8.717872783521048e-06,
+      "loss": 1.0811,
+      "step": 182
+    },
+    {
+      "epoch": 0.5362637362637362,
+      "grad_norm": 0.7534508660091974,
+      "learning_rate": 8.701720519555315e-06,
+      "loss": 1.0978,
+      "step": 183
+    },
+    {
+      "epoch": 0.5391941391941392,
+      "grad_norm": 0.7769245974744959,
+      "learning_rate": 8.685482313216784e-06,
+      "loss": 1.1021,
+      "step": 184
+    },
+    {
+      "epoch": 0.5421245421245421,
+      "grad_norm": 0.7339756524664363,
+      "learning_rate": 8.669158541505817e-06,
+      "loss": 1.0857,
+      "step": 185
+    },
+    {
+      "epoch": 0.545054945054945,
+      "grad_norm": 0.6938946896380407,
+      "learning_rate": 8.65274958340934e-06,
+      "loss": 1.1024,
+      "step": 186
+    },
+    {
+      "epoch": 0.547985347985348,
+      "grad_norm": 0.7373481468386694,
+      "learning_rate": 8.636255819892037e-06,
+      "loss": 1.0731,
+      "step": 187
+    },
+    {
+      "epoch": 0.550915750915751,
+      "grad_norm": 0.7356675883617693,
+      "learning_rate": 8.61967763388751e-06,
+      "loss": 1.0833,
+      "step": 188
+    },
+    {
+      "epoch": 0.5538461538461539,
+      "grad_norm": 0.6721670812209259,
+      "learning_rate": 8.603015410289387e-06,
+      "loss": 1.0692,
+      "step": 189
+    },
+    {
+      "epoch": 0.5567765567765568,
+      "grad_norm": 0.7164842748809434,
+      "learning_rate": 8.586269535942386e-06,
+      "loss": 1.0276,
+      "step": 190
+    },
+    {
+      "epoch": 0.5597069597069597,
+      "grad_norm": 0.7138024169545026,
+      "learning_rate": 8.569440399633337e-06,
+      "loss": 1.0394,
+      "step": 191
+    },
+    {
+      "epoch": 0.5626373626373626,
+      "grad_norm": 0.7406122273657452,
+      "learning_rate": 8.552528392082147e-06,
+      "loss": 1.1192,
+      "step": 192
+    },
+    {
+      "epoch": 0.5655677655677656,
+      "grad_norm": 0.6235431206762926,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 1.0834,
+      "step": 193
+    },
+    {
+      "epoch": 0.5684981684981685,
+      "grad_norm": 0.7124256759895987,
+      "learning_rate": 8.518457335743927e-06,
+      "loss": 1.0753,
+      "step": 194
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.6677579355944584,
+      "learning_rate": 8.50129907798026e-06,
+      "loss": 1.064,
+      "step": 195
+    },
+    {
+      "epoch": 0.5743589743589743,
+      "grad_norm": 0.7499647765747881,
+      "learning_rate": 8.484059531002822e-06,
+      "loss": 1.1573,
+      "step": 196
+    },
+    {
+      "epoch": 0.5772893772893772,
+      "grad_norm": 0.7294026788868488,
+      "learning_rate": 8.46673909505997e-06,
+      "loss": 1.1011,
+      "step": 197
+    },
+    {
+      "epoch": 0.5802197802197803,
+      "grad_norm": 0.7114933216227827,
+      "learning_rate": 8.44933817227806e-06,
+      "loss": 1.0818,
+      "step": 198
+    },
+    {
+      "epoch": 0.5831501831501832,
+      "grad_norm": 0.7768894842987932,
+      "learning_rate": 8.43185716665209e-06,
+      "loss": 1.089,
+      "step": 199
+    },
+    {
+      "epoch": 0.5860805860805861,
+      "grad_norm": 0.7115463290600534,
+      "learning_rate": 8.41429648403634e-06,
+      "loss": 1.1403,
+      "step": 200
+    },
+    {
+      "epoch": 0.589010989010989,
+      "grad_norm": 0.744930428011927,
+      "learning_rate": 8.396656532134934e-06,
+      "loss": 1.0643,
+      "step": 201
+    },
+    {
+      "epoch": 0.591941391941392,
+      "grad_norm": 0.7155769890648402,
+      "learning_rate": 8.378937720492384e-06,
+      "loss": 1.0786,
+      "step": 202
+    },
+    {
+      "epoch": 0.5948717948717949,
+      "grad_norm": 0.7127918438707657,
+      "learning_rate": 8.361140460484081e-06,
+      "loss": 1.0324,
+      "step": 203
+    },
+    {
+      "epoch": 0.5978021978021978,
+      "grad_norm": 0.7402459212974355,
+      "learning_rate": 8.343265165306736e-06,
+      "loss": 1.0155,
+      "step": 204
+    },
+    {
+      "epoch": 0.6007326007326007,
+      "grad_norm": 0.7886204426553328,
+      "learning_rate": 8.325312249968795e-06,
+      "loss": 1.0667,
+      "step": 205
+    },
+    {
+      "epoch": 0.6036630036630036,
+      "grad_norm": 0.9134639382550404,
+      "learning_rate": 8.307282131280805e-06,
+      "loss": 1.0971,
+      "step": 206
+    },
+    {
+      "epoch": 0.6065934065934065,
+      "grad_norm": 0.7673543125270567,
+      "learning_rate": 8.289175227845727e-06,
+      "loss": 1.0703,
+      "step": 207
+    },
+    {
+      "epoch": 0.6095238095238096,
+      "grad_norm": 0.7709392914165542,
+      "learning_rate": 8.270991960049231e-06,
+      "loss": 1.0553,
+      "step": 208
+    },
+    {
+      "epoch": 0.6124542124542125,
+      "grad_norm": 0.7555209074220831,
+      "learning_rate": 8.252732750049918e-06,
+      "loss": 1.0162,
+      "step": 209
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.7247403658877494,
+      "learning_rate": 8.234398021769541e-06,
+      "loss": 1.0665,
+      "step": 210
+    },
+    {
+      "epoch": 0.6183150183150183,
+      "grad_norm": 0.7491705240437763,
+      "learning_rate": 8.215988200883145e-06,
+      "loss": 1.0822,
+      "step": 211
+    },
+    {
+      "epoch": 0.6212454212454213,
+      "grad_norm": 0.7786072749364591,
+      "learning_rate": 8.19750371480919e-06,
+      "loss": 1.0759,
+      "step": 212
+    },
+    {
+      "epoch": 0.6241758241758242,
+      "grad_norm": 0.7502719607254296,
+      "learning_rate": 8.178944992699632e-06,
+      "loss": 1.0494,
+      "step": 213
+    },
+    {
+      "epoch": 0.6271062271062271,
+      "grad_norm": 0.980326365407937,
+      "learning_rate": 8.160312465429952e-06,
+      "loss": 1.0222,
+      "step": 214
+    },
+    {
+      "epoch": 0.63003663003663,
+      "grad_norm": 0.8323917117064283,
+      "learning_rate": 8.141606565589157e-06,
+      "loss": 1.0473,
+      "step": 215
+    },
+    {
+      "epoch": 0.6329670329670329,
+      "grad_norm": 0.711816328710808,
+      "learning_rate": 8.122827727469737e-06,
+      "loss": 1.0875,
+      "step": 216
+    },
+    {
+      "epoch": 0.6358974358974359,
+      "grad_norm": 0.7441109264008124,
+      "learning_rate": 8.103976387057579e-06,
+      "loss": 1.0823,
+      "step": 217
+    },
+    {
+      "epoch": 0.6388278388278388,
+      "grad_norm": 0.7262059806258652,
+      "learning_rate": 8.085052982021849e-06,
+      "loss": 1.0286,
+      "step": 218
+    },
+    {
+      "epoch": 0.6417582417582418,
+      "grad_norm": 0.7928850120911566,
+      "learning_rate": 8.066057951704821e-06,
+      "loss": 1.123,
+      "step": 219
+    },
+    {
+      "epoch": 0.6446886446886447,
+      "grad_norm": 0.7685198156864494,
+      "learning_rate": 8.046991737111696e-06,
+      "loss": 1.0818,
+      "step": 220
+    },
+    {
+      "epoch": 0.6476190476190476,
+      "grad_norm": 0.7198626738261616,
+      "learning_rate": 8.027854780900339e-06,
+      "loss": 1.0191,
+      "step": 221
+    },
+    {
+      "epoch": 0.6505494505494506,
+      "grad_norm": 0.7326930302530626,
+      "learning_rate": 8.008647527371022e-06,
+      "loss": 1.0473,
+      "step": 222
+    },
+    {
+      "epoch": 0.6534798534798535,
+      "grad_norm": 0.7625974111690114,
+      "learning_rate": 7.989370422456098e-06,
+      "loss": 1.0573,
+      "step": 223
+    },
+    {
+      "epoch": 0.6564102564102564,
+      "grad_norm": 0.7927397844502478,
+      "learning_rate": 7.970023913709652e-06,
+      "loss": 1.0741,
+      "step": 224
+    },
+    {
+      "epoch": 0.6593406593406593,
+      "grad_norm": 0.8105334583256553,
+      "learning_rate": 7.95060845029711e-06,
+      "loss": 1.0406,
+      "step": 225
+    },
+    {
+      "epoch": 0.6622710622710622,
+      "grad_norm": 0.819096758560569,
+      "learning_rate": 7.931124482984802e-06,
+      "loss": 1.0952,
+      "step": 226
+    },
+    {
+      "epoch": 0.6652014652014652,
+      "grad_norm": 0.7225216871665473,
+      "learning_rate": 7.911572464129515e-06,
+      "loss": 1.0514,
+      "step": 227
+    },
+    {
+      "epoch": 0.6681318681318681,
+      "grad_norm": 0.8288654068393345,
+      "learning_rate": 7.891952847667973e-06,
+      "loss": 1.1164,
+      "step": 228
+    },
+    {
+      "epoch": 0.671062271062271,
+      "grad_norm": 0.6759229558936937,
+      "learning_rate": 7.872266089106309e-06,
+      "loss": 1.0495,
+      "step": 229
+    },
+    {
+      "epoch": 0.673992673992674,
+      "grad_norm": 0.7292237419206736,
+      "learning_rate": 7.85251264550948e-06,
+      "loss": 1.0418,
+      "step": 230
+    },
+    {
+      "epoch": 0.676923076923077,
+      "grad_norm": 0.8129512318365157,
+      "learning_rate": 7.832692975490668e-06,
+      "loss": 1.0824,
+      "step": 231
+    },
+    {
+      "epoch": 0.6798534798534799,
+      "grad_norm": 0.7437869057766316,
+      "learning_rate": 7.812807539200622e-06,
+      "loss": 1.096,
+      "step": 232
+    },
+    {
+      "epoch": 0.6827838827838828,
+      "grad_norm": 0.7490823810391557,
+      "learning_rate": 7.792856798316975e-06,
+      "loss": 1.0763,
+      "step": 233
+    },
+    {
+      "epoch": 0.6857142857142857,
+      "grad_norm": 0.7898358276478571,
+      "learning_rate": 7.772841216033534e-06,
+      "loss": 1.0771,
+      "step": 234
+    },
+    {
+      "epoch": 0.6886446886446886,
+      "grad_norm": 0.6343598620923301,
+      "learning_rate": 7.752761257049517e-06,
+      "loss": 1.0322,
+      "step": 235
+    },
+    {
+      "epoch": 0.6915750915750916,
+      "grad_norm": 0.7433500237046824,
+      "learning_rate": 7.732617387558769e-06,
+      "loss": 1.0797,
+      "step": 236
+    },
+    {
+      "epoch": 0.6945054945054945,
+      "grad_norm": 0.748028902224525,
+      "learning_rate": 7.712410075238936e-06,
+      "loss": 1.0574,
+      "step": 237
+    },
+    {
+      "epoch": 0.6974358974358974,
+      "grad_norm": 0.7738403008694008,
+      "learning_rate": 7.692139789240611e-06,
+      "loss": 1.1001,
+      "step": 238
+    },
+    {
+      "epoch": 0.7003663003663003,
+      "grad_norm": 0.7228116763923162,
+      "learning_rate": 7.671807000176434e-06,
+      "loss": 1.0799,
+      "step": 239
+    },
+    {
+      "epoch": 0.7032967032967034,
+      "grad_norm": 0.7253527243733535,
+      "learning_rate": 7.651412180110176e-06,
+      "loss": 1.0707,
+      "step": 240
+    },
+    {
+      "epoch": 0.7062271062271063,
+      "grad_norm": 0.7348620682799656,
+      "learning_rate": 7.630955802545768e-06,
+      "loss": 1.0474,
+      "step": 241
+    },
+    {
+      "epoch": 0.7091575091575092,
+      "grad_norm": 0.6836835745144434,
+      "learning_rate": 7.61043834241632e-06,
+      "loss": 1.0812,
+      "step": 242
+    },
+    {
+      "epoch": 0.7120879120879121,
+      "grad_norm": 0.6837971336343311,
+      "learning_rate": 7.589860276073082e-06,
+      "loss": 1.0473,
+      "step": 243
+    },
+    {
+      "epoch": 0.715018315018315,
+      "grad_norm": 0.7230704080796915,
+      "learning_rate": 7.569222081274396e-06,
+      "loss": 1.1134,
+      "step": 244
+    },
+    {
+      "epoch": 0.717948717948718,
+      "grad_norm": 0.7127149830099806,
+      "learning_rate": 7.548524237174595e-06,
+      "loss": 1.064,
+      "step": 245
+    },
+    {
+      "epoch": 0.7208791208791209,
+      "grad_norm": 0.7393242144731675,
+      "learning_rate": 7.527767224312883e-06,
+      "loss": 1.055,
+      "step": 246
+    },
+    {
+      "epoch": 0.7238095238095238,
+      "grad_norm": 0.7070740978028522,
+      "learning_rate": 7.50695152460218e-06,
+      "loss": 1.0514,
+      "step": 247
+    },
+    {
+      "epoch": 0.7267399267399267,
+      "grad_norm": 0.6285827841364233,
+      "learning_rate": 7.4860776213179264e-06,
+      "loss": 1.0663,
+      "step": 248
+    },
+    {
+      "epoch": 0.7296703296703296,
+      "grad_norm": 0.722505271587716,
+      "learning_rate": 7.465145999086874e-06,
+      "loss": 1.0247,
+      "step": 249
+    },
+    {
+      "epoch": 0.7326007326007326,
+      "grad_norm": 0.7574691065276626,
+      "learning_rate": 7.44415714387582e-06,
+      "loss": 1.0806,
+      "step": 250
+    },
+    {
+      "epoch": 0.7355311355311356,
+      "grad_norm": 0.6671911844141701,
+      "learning_rate": 7.4231115429803405e-06,
+      "loss": 1.0718,
+      "step": 251
+    },
+    {
+      "epoch": 0.7384615384615385,
+      "grad_norm": 0.6739232630180421,
+      "learning_rate": 7.4020096850134635e-06,
+      "loss": 1.054,
+      "step": 252
+    },
+    {
+      "epoch": 0.7413919413919414,
+      "grad_norm": 0.6816160363185811,
+      "learning_rate": 7.380852059894331e-06,
+      "loss": 1.0135,
+      "step": 253
+    },
+    {
+      "epoch": 0.7443223443223443,
+      "grad_norm": 0.7255625077642197,
+      "learning_rate": 7.359639158836828e-06,
+      "loss": 1.0765,
+      "step": 254
+    },
+    {
+      "epoch": 0.7472527472527473,
+      "grad_norm": 0.7389412886246451,
+      "learning_rate": 7.3383714743381664e-06,
+      "loss": 1.0734,
+      "step": 255
+    },
+    {
+      "epoch": 0.7501831501831502,
+      "grad_norm": 0.6974028362998995,
+      "learning_rate": 7.317049500167466e-06,
+      "loss": 1.0225,
+      "step": 256
+    },
+    {
+      "epoch": 0.7531135531135531,
+      "grad_norm": 0.706462620691865,
+      "learning_rate": 7.295673731354278e-06,
+      "loss": 1.0433,
+      "step": 257
+    },
+    {
+      "epoch": 0.756043956043956,
+      "grad_norm": 0.6939088050227432,
+      "learning_rate": 7.2742446641770985e-06,
+      "loss": 1.0283,
+      "step": 258
+    },
+    {
+      "epoch": 0.7589743589743589,
+      "grad_norm": 0.6738265352397054,
+      "learning_rate": 7.252762796151843e-06,
+      "loss": 1.0338,
+      "step": 259
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 0.7509847899569063,
+      "learning_rate": 7.231228626020303e-06,
+      "loss": 1.0851,
+      "step": 260
+    },
+    {
+      "epoch": 0.7648351648351648,
+      "grad_norm": 0.7111929689141302,
+      "learning_rate": 7.209642653738555e-06,
+      "loss": 1.0467,
+      "step": 261
+    },
+    {
+      "epoch": 0.7677655677655678,
+      "grad_norm": 0.6700991516973885,
+      "learning_rate": 7.188005380465365e-06,
+      "loss": 1.0886,
+      "step": 262
+    },
+    {
+      "epoch": 0.7706959706959707,
+      "grad_norm": 0.7620776406324662,
+      "learning_rate": 7.166317308550541e-06,
+      "loss": 1.0914,
+      "step": 263
+    },
+    {
+      "epoch": 0.7736263736263737,
+      "grad_norm": 0.6438553939939631,
+      "learning_rate": 7.144578941523283e-06,
+      "loss": 1.0139,
+      "step": 264
+    },
+    {
+      "epoch": 0.7765567765567766,
+      "grad_norm": 0.6627461325502493,
+      "learning_rate": 7.122790784080482e-06,
+      "loss": 0.978,
+      "step": 265
+    },
+    {
+      "epoch": 0.7794871794871795,
+      "grad_norm": 0.7162460216190278,
+      "learning_rate": 7.10095334207501e-06,
+      "loss": 1.0949,
+      "step": 266
+    },
+    {
+      "epoch": 0.7824175824175824,
+      "grad_norm": 0.6889558669766283,
+      "learning_rate": 7.079067122503969e-06,
+      "loss": 1.0872,
+      "step": 267
+    },
+    {
+      "epoch": 0.7853479853479853,
+      "grad_norm": 0.6988411158347763,
+      "learning_rate": 7.057132633496924e-06,
+      "loss": 1.0609,
+      "step": 268
+    },
+    {
+      "epoch": 0.7882783882783883,
+      "grad_norm": 0.8465993660592789,
+      "learning_rate": 7.03515038430411e-06,
+      "loss": 1.0754,
+      "step": 269
+    },
+    {
+      "epoch": 0.7912087912087912,
+      "grad_norm": 0.7433786645831059,
+      "learning_rate": 7.013120885284599e-06,
+      "loss": 1.0521,
+      "step": 270
+    },
+    {
+      "epoch": 0.7941391941391941,
+      "grad_norm": 0.6471784629568724,
+      "learning_rate": 6.991044647894456e-06,
+      "loss": 1.0222,
+      "step": 271
+    },
+    {
+      "epoch": 0.7970695970695971,
+      "grad_norm": 0.7717525416061395,
+      "learning_rate": 6.968922184674868e-06,
+      "loss": 1.0598,
+      "step": 272
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7375028451227608,
+      "learning_rate": 6.946754009240241e-06,
+      "loss": 1.0879,
+      "step": 273
+    },
+    {
+      "epoch": 0.802930402930403,
+      "grad_norm": 0.7395816953663372,
+      "learning_rate": 6.924540636266272e-06,
+      "loss": 1.0495,
+      "step": 274
+    },
+    {
+      "epoch": 0.8058608058608059,
+      "grad_norm": 0.6534788778148224,
+      "learning_rate": 6.902282581478009e-06,
+      "loss": 1.0484,
+      "step": 275
+    },
+    {
+      "epoch": 0.8087912087912088,
+      "grad_norm": 0.6831894193923664,
+      "learning_rate": 6.879980361637865e-06,
+      "loss": 1.0709,
+      "step": 276
+    },
+    {
+      "epoch": 0.8117216117216117,
+      "grad_norm": 0.7986776852595957,
+      "learning_rate": 6.857634494533636e-06,
+      "loss": 1.0989,
+      "step": 277
+    },
+    {
+      "epoch": 0.8146520146520146,
+      "grad_norm": 0.7138986323229565,
+      "learning_rate": 6.835245498966461e-06,
+      "loss": 1.1026,
+      "step": 278
+    },
+    {
+      "epoch": 0.8175824175824176,
+      "grad_norm": 0.7650967662850501,
+      "learning_rate": 6.8128138947387966e-06,
+      "loss": 1.0906,
+      "step": 279
+    },
+    {
+      "epoch": 0.8205128205128205,
+      "grad_norm": 0.6705870456445562,
+      "learning_rate": 6.790340202642333e-06,
+      "loss": 1.0306,
+      "step": 280
+    },
+    {
+      "epoch": 0.8234432234432234,
+      "grad_norm": 0.6924029470535028,
+      "learning_rate": 6.76782494444591e-06,
+      "loss": 1.0682,
+      "step": 281
+    },
+    {
+      "epoch": 0.8263736263736263,
+      "grad_norm": 0.7640179483164876,
+      "learning_rate": 6.7452686428834045e-06,
+      "loss": 1.0251,
+      "step": 282
+    },
+    {
+      "epoch": 0.8293040293040294,
+      "grad_norm": 0.6821634834350029,
+      "learning_rate": 6.722671821641591e-06,
+      "loss": 1.0446,
+      "step": 283
+    },
+    {
+      "epoch": 0.8322344322344323,
+      "grad_norm": 0.7451959173472754,
+      "learning_rate": 6.700035005347983e-06,
+      "loss": 1.1235,
+      "step": 284
+    },
+    {
+      "epoch": 0.8351648351648352,
+      "grad_norm": 0.6649020603168196,
+      "learning_rate": 6.677358719558655e-06,
+      "loss": 0.9895,
+      "step": 285
+    },
+    {
+      "epoch": 0.8380952380952381,
+      "grad_norm": 0.710844339419014,
+      "learning_rate": 6.654643490746042e-06,
+      "loss": 1.0464,
+      "step": 286
+    },
+    {
+      "epoch": 0.841025641025641,
+      "grad_norm": 0.6558219608160121,
+      "learning_rate": 6.6318898462867086e-06,
+      "loss": 1.0789,
+      "step": 287
+    },
+    {
+      "epoch": 0.843956043956044,
+      "grad_norm": 0.7103315959032128,
+      "learning_rate": 6.609098314449116e-06,
+      "loss": 1.0693,
+      "step": 288
+    },
+    {
+      "epoch": 0.8468864468864469,
+      "grad_norm": 0.6642938199213403,
+      "learning_rate": 6.586269424381349e-06,
+      "loss": 1.0111,
+      "step": 289
+    },
+    {
+      "epoch": 0.8498168498168498,
+      "grad_norm": 0.6686652730007528,
+      "learning_rate": 6.563403706098833e-06,
+      "loss": 1.0493,
+      "step": 290
+    },
+    {
+      "epoch": 0.8527472527472527,
+      "grad_norm": 0.707080494234748,
+      "learning_rate": 6.540501690472032e-06,
+      "loss": 1.0933,
+      "step": 291
+    },
+    {
+      "epoch": 0.8556776556776556,
+      "grad_norm": 0.8070591206820166,
+      "learning_rate": 6.517563909214119e-06,
+      "loss": 1.0598,
+      "step": 292
+    },
+    {
+      "epoch": 0.8586080586080586,
+      "grad_norm": 0.6443876519255367,
+      "learning_rate": 6.4945908948686355e-06,
+      "loss": 1.0436,
+      "step": 293
+    },
+    {
+      "epoch": 0.8615384615384616,
+      "grad_norm": 0.6358308118659253,
+      "learning_rate": 6.471583180797121e-06,
+      "loss": 1.0205,
+      "step": 294
+    },
+    {
+      "epoch": 0.8644688644688645,
+      "grad_norm": 0.6980748843159623,
+      "learning_rate": 6.448541301166737e-06,
+      "loss": 1.0481,
+      "step": 295
+    },
+    {
+      "epoch": 0.8673992673992674,
+      "grad_norm": 0.721201397418447,
+      "learning_rate": 6.4254657909378615e-06,
+      "loss": 1.0533,
+      "step": 296
+    },
+    {
+      "epoch": 0.8703296703296703,
+      "grad_norm": 0.6902627729970224,
+      "learning_rate": 6.402357185851669e-06,
+      "loss": 1.0786,
+      "step": 297
+    },
+    {
+      "epoch": 0.8732600732600733,
+      "grad_norm": 0.7134804889991092,
+      "learning_rate": 6.379216022417695e-06,
+      "loss": 1.0654,
+      "step": 298
+    },
+    {
+      "epoch": 0.8761904761904762,
+      "grad_norm": 0.6409021568091943,
+      "learning_rate": 6.3560428379013795e-06,
+      "loss": 1.0456,
+      "step": 299
+    },
+    {
+      "epoch": 0.8791208791208791,
+      "grad_norm": 0.7159300653418184,
+      "learning_rate": 6.332838170311586e-06,
+      "loss": 1.0795,
+      "step": 300
+    },
+    {
+      "epoch": 0.882051282051282,
+      "grad_norm": 0.6634422162154384,
+      "learning_rate": 6.309602558388122e-06,
+      "loss": 1.0906,
+      "step": 301
+    },
+    {
+      "epoch": 0.884981684981685,
+      "grad_norm": 0.6748316902652748,
+      "learning_rate": 6.286336541589224e-06,
+      "loss": 1.0344,
+      "step": 302
+    },
+    {
+      "epoch": 0.8879120879120879,
+      "grad_norm": 0.762311805582648,
+      "learning_rate": 6.263040660079031e-06,
+      "loss": 1.0328,
+      "step": 303
+    },
+    {
+      "epoch": 0.8908424908424909,
+      "grad_norm": 0.6567073940057072,
+      "learning_rate": 6.239715454715054e-06,
+      "loss": 1.0326,
+      "step": 304
+    },
+    {
+      "epoch": 0.8937728937728938,
+      "grad_norm": 0.7360191370508475,
+      "learning_rate": 6.216361467035608e-06,
+      "loss": 1.0671,
+      "step": 305
+    },
+    {
+      "epoch": 0.8967032967032967,
+      "grad_norm": 0.6536751414819448,
+      "learning_rate": 6.192979239247243e-06,
+      "loss": 1.04,
+      "step": 306
+    },
+    {
+      "epoch": 0.8996336996336997,
+      "grad_norm": 0.688711476034779,
+      "learning_rate": 6.169569314212157e-06,
+      "loss": 1.0425,
+      "step": 307
+    },
+    {
+      "epoch": 0.9025641025641026,
+      "grad_norm": 0.7128048107980386,
+      "learning_rate": 6.146132235435591e-06,
+      "loss": 1.0479,
+      "step": 308
+    },
+    {
+      "epoch": 0.9054945054945055,
+      "grad_norm": 0.7027292353974949,
+      "learning_rate": 6.1226685470532125e-06,
+      "loss": 0.9922,
+      "step": 309
+    },
+    {
+      "epoch": 0.9084249084249084,
+      "grad_norm": 0.662332343110812,
+      "learning_rate": 6.099178793818479e-06,
+      "loss": 1.0564,
+      "step": 310
+    },
+    {
+      "epoch": 0.9113553113553113,
+      "grad_norm": 0.6578408725334101,
+      "learning_rate": 6.075663521089994e-06,
+      "loss": 1.0408,
+      "step": 311
+    },
+    {
+      "epoch": 0.9142857142857143,
+      "grad_norm": 0.7225530601139032,
+      "learning_rate": 6.0521232748188416e-06,
+      "loss": 1.0386,
+      "step": 312
+    },
+    {
+      "epoch": 0.9172161172161172,
+      "grad_norm": 0.6884615641249574,
+      "learning_rate": 6.028558601535915e-06,
+      "loss": 1.0422,
+      "step": 313
+    },
+    {
+      "epoch": 0.9201465201465201,
+      "grad_norm": 0.6237151400719672,
+      "learning_rate": 6.0049700483392256e-06,
+      "loss": 0.9923,
+      "step": 314
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.6729617533435394,
+      "learning_rate": 5.981358162881202e-06,
+      "loss": 1.0802,
+      "step": 315
+    },
+    {
+      "epoch": 0.926007326007326,
+      "grad_norm": 0.6283161670002942,
+      "learning_rate": 5.957723493355977e-06,
+      "loss": 1.0057,
+      "step": 316
+    },
+    {
+      "epoch": 0.928937728937729,
+      "grad_norm": 0.7106310162079014,
+      "learning_rate": 5.9340665884866535e-06,
+      "loss": 1.0452,
+      "step": 317
+    },
+    {
+      "epoch": 0.9318681318681319,
+      "grad_norm": 0.6307561472458387,
+      "learning_rate": 5.910387997512573e-06,
+      "loss": 1.0678,
+      "step": 318
+    },
+    {
+      "epoch": 0.9347985347985348,
+      "grad_norm": 0.6660079952439634,
+      "learning_rate": 5.8866882701765605e-06,
+      "loss": 1.011,
+      "step": 319
+    },
+    {
+      "epoch": 0.9377289377289377,
+      "grad_norm": 0.6621973288526832,
+      "learning_rate": 5.86296795671216e-06,
+      "loss": 1.0777,
+      "step": 320
+    },
+    {
+      "epoch": 0.9406593406593406,
+      "grad_norm": 0.6542364386949882,
+      "learning_rate": 5.839227607830862e-06,
+      "loss": 1.0337,
+      "step": 321
+    },
+    {
+      "epoch": 0.9435897435897436,
+      "grad_norm": 0.6839429524505124,
+      "learning_rate": 5.815467774709314e-06,
+      "loss": 1.0398,
+      "step": 322
+    },
+    {
+      "epoch": 0.9465201465201465,
+      "grad_norm": 0.7895999257535952,
+      "learning_rate": 5.791689008976531e-06,
+      "loss": 1.0965,
+      "step": 323
+    },
+    {
+      "epoch": 0.9494505494505494,
+      "grad_norm": 0.7076974699228543,
+      "learning_rate": 5.767891862701081e-06,
+      "loss": 1.0508,
+      "step": 324
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 0.6560722941227365,
+      "learning_rate": 5.744076888378272e-06,
+      "loss": 1.0347,
+      "step": 325
+    },
+    {
+      "epoch": 0.9553113553113554,
+      "grad_norm": 0.7153841740442476,
+      "learning_rate": 5.7202446389173225e-06,
+      "loss": 1.048,
+      "step": 326
+    },
+    {
+      "epoch": 0.9582417582417583,
+      "grad_norm": 0.6840752335867581,
+      "learning_rate": 5.696395667628526e-06,
+      "loss": 1.0706,
+      "step": 327
+    },
+    {
+      "epoch": 0.9611721611721612,
+      "grad_norm": 0.7367756826844876,
+      "learning_rate": 5.672530528210405e-06,
+      "loss": 1.0631,
+      "step": 328
+    },
+    {
+      "epoch": 0.9641025641025641,
+      "grad_norm": 0.739456837074914,
+      "learning_rate": 5.648649774736855e-06,
+      "loss": 1.0655,
+      "step": 329
+    },
+    {
+      "epoch": 0.967032967032967,
+      "grad_norm": 0.7489432381420208,
+      "learning_rate": 5.624753961644281e-06,
+      "loss": 1.0103,
+      "step": 330
+    },
+    {
+      "epoch": 0.96996336996337,
+      "grad_norm": 0.9214056170953451,
+      "learning_rate": 5.600843643718728e-06,
+      "loss": 1.0586,
+      "step": 331
+    },
+    {
+      "epoch": 0.9728937728937729,
+      "grad_norm": 0.6946485880358703,
+      "learning_rate": 5.57691937608299e-06,
+      "loss": 1.0673,
+      "step": 332
+    },
+    {
+      "epoch": 0.9758241758241758,
+      "grad_norm": 0.6814888711345364,
+      "learning_rate": 5.552981714183738e-06,
+      "loss": 1.0312,
+      "step": 333
+    },
+    {
+      "epoch": 0.9787545787545787,
+      "grad_norm": 0.7851305269555706,
+      "learning_rate": 5.529031213778615e-06,
+      "loss": 1.0929,
+      "step": 334
+    },
+    {
+      "epoch": 0.9816849816849816,
+      "grad_norm": 0.698622025095992,
+      "learning_rate": 5.505068430923327e-06,
+      "loss": 1.0721,
+      "step": 335
+    },
+    {
+      "epoch": 0.9846153846153847,
+      "grad_norm": 0.6534426116496003,
+      "learning_rate": 5.481093921958749e-06,
+      "loss": 1.0953,
+      "step": 336
+    },
+    {
+      "epoch": 0.9875457875457876,
+      "grad_norm": 0.6625287425989687,
+      "learning_rate": 5.457108243497993e-06,
+      "loss": 0.9991,
+      "step": 337
+    },
+    {
+      "epoch": 0.9904761904761905,
+      "grad_norm": 0.701327887950468,
+      "learning_rate": 5.433111952413496e-06,
+      "loss": 1.0658,
+      "step": 338
+    },
+    {
+      "epoch": 0.9934065934065934,
+      "grad_norm": 0.6716839301392191,
+      "learning_rate": 5.409105605824082e-06,
+      "loss": 1.0413,
+      "step": 339
+    },
+    {
+      "epoch": 0.9963369963369964,
+      "grad_norm": 0.7609181677574756,
+      "learning_rate": 5.385089761082039e-06,
+      "loss": 0.9849,
+      "step": 340
+    },
+    {
+      "epoch": 0.9992673992673993,
+      "grad_norm": 0.6928967498595602,
+      "learning_rate": 5.361064975760166e-06,
+      "loss": 1.025,
+      "step": 341
+    },
+    {
+      "epoch": 1.0021978021978022,
+      "grad_norm": 0.6340076640585811,
+      "learning_rate": 5.3370318076388405e-06,
+      "loss": 1.06,
+      "step": 342
+    },
+    {
+      "epoch": 1.005128205128205,
+      "grad_norm": 0.7147887444305453,
+      "learning_rate": 5.3129908146930565e-06,
+      "loss": 1.0402,
+      "step": 343
+    },
+    {
+      "epoch": 1.008058608058608,
+      "grad_norm": 0.7081726636264288,
+      "learning_rate": 5.288942555079479e-06,
+      "loss": 1.0214,
+      "step": 344
+    },
+    {
+      "epoch": 1.010989010989011,
+      "grad_norm": 0.6931248026006267,
+      "learning_rate": 5.264887587123483e-06,
+      "loss": 1.0355,
+      "step": 345
+    },
+    {
+      "epoch": 1.0139194139194139,
+      "grad_norm": 0.6614729802440827,
+      "learning_rate": 5.240826469306187e-06,
+      "loss": 1.0336,
+      "step": 346
+    },
+    {
+      "epoch": 1.0168498168498168,
+      "grad_norm": 0.72843052045215,
+      "learning_rate": 5.216759760251493e-06,
+      "loss": 1.0959,
+      "step": 347
+    },
+    {
+      "epoch": 1.002930402930403,
+      "grad_norm": 0.9856684936564257,
+      "learning_rate": 5.1926880187131134e-06,
+      "loss": 0.8665,
+      "step": 348
+    },
+    {
+      "epoch": 1.0058608058608058,
+      "grad_norm": 0.8924138774764413,
+      "learning_rate": 5.168611803561599e-06,
+      "loss": 0.8482,
+      "step": 349
+    },
+    {
+      "epoch": 1.0087912087912088,
+      "grad_norm": 0.80619364515724,
+      "learning_rate": 5.144531673771364e-06,
+      "loss": 0.904,
+      "step": 350
+    },
+    {
+      "epoch": 1.0117216117216117,
+      "grad_norm": 0.8307330430576558,
+      "learning_rate": 5.1204481884077075e-06,
+      "loss": 0.824,
+      "step": 351
+    },
+    {
+      "epoch": 1.0146520146520146,
+      "grad_norm": 0.8223666575265466,
+      "learning_rate": 5.096361906613836e-06,
+      "loss": 0.8194,
+      "step": 352
+    },
+    {
+      "epoch": 1.0175824175824175,
+      "grad_norm": 0.882567742233647,
+      "learning_rate": 5.072273387597877e-06,
+      "loss": 0.8007,
+      "step": 353
+    },
+    {
+      "epoch": 1.0205128205128204,
+      "grad_norm": 0.852155527533341,
+      "learning_rate": 5.048183190619904e-06,
+      "loss": 0.8299,
+      "step": 354
+    },
+    {
+      "epoch": 1.0234432234432234,
+      "grad_norm": 0.6698748000250594,
+      "learning_rate": 5.0240918749789455e-06,
+      "loss": 0.8609,
+      "step": 355
+    },
+    {
+      "epoch": 1.0263736263736263,
+      "grad_norm": 0.740093168594684,
+      "learning_rate": 5e-06,
+      "loss": 0.8179,
+      "step": 356
+    },
+    {
+      "epoch": 1.0293040293040292,
+      "grad_norm": 0.9241467975042712,
+      "learning_rate": 4.975908125021055e-06,
+      "loss": 0.8817,
+      "step": 357
+    },
+    {
+      "epoch": 1.0322344322344323,
+      "grad_norm": 0.8242101021976431,
+      "learning_rate": 4.951816809380098e-06,
+      "loss": 0.8461,
+      "step": 358
+    },
+    {
+      "epoch": 1.0351648351648353,
+      "grad_norm": 0.6778411446141664,
+      "learning_rate": 4.9277266124021245e-06,
+      "loss": 0.8251,
+      "step": 359
+    },
+    {
+      "epoch": 1.0380952380952382,
+      "grad_norm": 0.8162644593862449,
+      "learning_rate": 4.903638093386167e-06,
+      "loss": 0.8391,
+      "step": 360
+    },
+    {
+      "epoch": 1.041025641025641,
+      "grad_norm": 0.6831525068683001,
+      "learning_rate": 4.879551811592295e-06,
+      "loss": 0.8454,
+      "step": 361
+    },
+    {
+      "epoch": 1.043956043956044,
+      "grad_norm": 0.7366137812958817,
+      "learning_rate": 4.855468326228638e-06,
+      "loss": 0.8383,
+      "step": 362
+    },
+    {
+      "epoch": 1.046886446886447,
+      "grad_norm": 0.7091031619746667,
+      "learning_rate": 4.831388196438402e-06,
+      "loss": 0.826,
+      "step": 363
+    },
+    {
+      "epoch": 1.0498168498168499,
+      "grad_norm": 0.6708047011362692,
+      "learning_rate": 4.807311981286888e-06,
+      "loss": 0.7795,
+      "step": 364
+    },
+    {
+      "epoch": 1.0527472527472528,
+      "grad_norm": 0.7189951484271022,
+      "learning_rate": 4.783240239748509e-06,
+      "loss": 0.8212,
+      "step": 365
+    },
+    {
+      "epoch": 1.0556776556776557,
+      "grad_norm": 0.6929263617105769,
+      "learning_rate": 4.7591735306938144e-06,
+      "loss": 0.8404,
+      "step": 366
+    },
+    {
+      "epoch": 1.0586080586080586,
+      "grad_norm": 0.7090536517310597,
+      "learning_rate": 4.7351124128765185e-06,
+      "loss": 0.7743,
+      "step": 367
+    },
+    {
+      "epoch": 1.0615384615384615,
+      "grad_norm": 0.7200636917743535,
+      "learning_rate": 4.711057444920522e-06,
+      "loss": 0.8388,
+      "step": 368
+    },
+    {
+      "epoch": 1.0644688644688645,
+      "grad_norm": 0.710980769195996,
+      "learning_rate": 4.687009185306945e-06,
+      "loss": 0.8322,
+      "step": 369
+    },
+    {
+      "epoch": 1.0673992673992674,
+      "grad_norm": 0.6240248501038852,
+      "learning_rate": 4.662968192361161e-06,
+      "loss": 0.7604,
+      "step": 370
+    },
+    {
+      "epoch": 1.0703296703296703,
+      "grad_norm": 0.6841882570704744,
+      "learning_rate": 4.6389350242398354e-06,
+      "loss": 0.8292,
+      "step": 371
+    },
+    {
+      "epoch": 1.0732600732600732,
+      "grad_norm": 0.6356314258852013,
+      "learning_rate": 4.614910238917963e-06,
+      "loss": 0.8112,
+      "step": 372
+    },
+    {
+      "epoch": 1.0761904761904761,
+      "grad_norm": 0.6777660171190972,
+      "learning_rate": 4.59089439417592e-06,
+      "loss": 0.8422,
+      "step": 373
+    },
+    {
+      "epoch": 1.079120879120879,
+      "grad_norm": 0.7302941927547204,
+      "learning_rate": 4.5668880475865074e-06,
+      "loss": 0.844,
+      "step": 374
+    },
+    {
+      "epoch": 1.082051282051282,
+      "grad_norm": 0.7403000182698058,
+      "learning_rate": 4.542891756502008e-06,
+      "loss": 0.8252,
+      "step": 375
+    },
+    {
+      "epoch": 1.084981684981685,
+      "grad_norm": 0.6270811120486014,
+      "learning_rate": 4.518906078041252e-06,
+      "loss": 0.8351,
+      "step": 376
+    },
+    {
+      "epoch": 1.0879120879120878,
+      "grad_norm": 0.6128334495594291,
+      "learning_rate": 4.494931569076674e-06,
+      "loss": 0.7953,
+      "step": 377
+    },
+    {
+      "epoch": 1.0908424908424907,
+      "grad_norm": 0.6647669320211729,
+      "learning_rate": 4.4709687862213866e-06,
+      "loss": 0.8271,
+      "step": 378
+    },
+    {
+      "epoch": 1.0937728937728939,
+      "grad_norm": 0.6015242764545017,
+      "learning_rate": 4.447018285816263e-06,
+      "loss": 0.7774,
+      "step": 379
+    },
+    {
+      "epoch": 1.0967032967032968,
+      "grad_norm": 0.6850007287026474,
+      "learning_rate": 4.423080623917012e-06,
+      "loss": 0.8323,
+      "step": 380
+    },
+    {
+      "epoch": 1.0996336996336997,
+      "grad_norm": 0.7137348622498122,
+      "learning_rate": 4.399156356281274e-06,
+      "loss": 0.8385,
+      "step": 381
+    },
+    {
+      "epoch": 1.1025641025641026,
+      "grad_norm": 0.6268533396942314,
+      "learning_rate": 4.3752460383557195e-06,
+      "loss": 0.7795,
+      "step": 382
+    },
+    {
+      "epoch": 1.1054945054945056,
+      "grad_norm": 0.6553350909418485,
+      "learning_rate": 4.351350225263147e-06,
+      "loss": 0.7732,
+      "step": 383
+    },
+    {
+      "epoch": 1.1084249084249085,
+      "grad_norm": 0.7239076024187311,
+      "learning_rate": 4.327469471789597e-06,
+      "loss": 0.7988,
+      "step": 384
+    },
+    {
+      "epoch": 1.1113553113553114,
+      "grad_norm": 0.6603332573386838,
+      "learning_rate": 4.303604332371476e-06,
+      "loss": 0.8234,
+      "step": 385
+    },
+    {
+      "epoch": 1.1142857142857143,
+      "grad_norm": 0.6836658490120479,
+      "learning_rate": 4.27975536108268e-06,
+      "loss": 0.8262,
+      "step": 386
+    },
+    {
+      "epoch": 1.1172161172161172,
+      "grad_norm": 0.6782684413264028,
+      "learning_rate": 4.2559231116217305e-06,
+      "loss": 0.8224,
+      "step": 387
+    },
+    {
+      "epoch": 1.1201465201465202,
+      "grad_norm": 0.6700675127409191,
+      "learning_rate": 4.2321081372989195e-06,
+      "loss": 0.7931,
+      "step": 388
+    },
+    {
+      "epoch": 1.123076923076923,
+      "grad_norm": 0.6991274806941559,
+      "learning_rate": 4.208310991023469e-06,
+      "loss": 0.8165,
+      "step": 389
+    },
+    {
+      "epoch": 1.126007326007326,
+      "grad_norm": 0.6764322388131341,
+      "learning_rate": 4.184532225290687e-06,
+      "loss": 0.7884,
+      "step": 390
+    },
+    {
+      "epoch": 1.128937728937729,
+      "grad_norm": 0.6640121837001143,
+      "learning_rate": 4.16077239216914e-06,
+      "loss": 0.8233,
+      "step": 391
+    },
+    {
+      "epoch": 1.1318681318681318,
+      "grad_norm": 0.6910556702389824,
+      "learning_rate": 4.137032043287841e-06,
+      "loss": 0.8162,
+      "step": 392
+    },
+    {
+      "epoch": 1.1347985347985348,
+      "grad_norm": 0.6709419218301458,
+      "learning_rate": 4.11331172982344e-06,
+      "loss": 0.7851,
+      "step": 393
+    },
+    {
+      "epoch": 1.1377289377289377,
+      "grad_norm": 0.6928725094292173,
+      "learning_rate": 4.089612002487428e-06,
+      "loss": 0.7892,
+      "step": 394
+    },
+    {
+      "epoch": 1.1406593406593406,
+      "grad_norm": 0.6562007454380986,
+      "learning_rate": 4.065933411513349e-06,
+      "loss": 0.8029,
+      "step": 395
+    },
+    {
+      "epoch": 1.1435897435897435,
+      "grad_norm": 0.6548393048781599,
+      "learning_rate": 4.042276506644024e-06,
+      "loss": 0.8241,
+      "step": 396
+    },
+    {
+      "epoch": 1.1465201465201464,
+      "grad_norm": 0.6896835546766348,
+      "learning_rate": 4.018641837118799e-06,
+      "loss": 0.8635,
+      "step": 397
+    },
+    {
+      "epoch": 1.1494505494505494,
+      "grad_norm": 0.7120720872162148,
+      "learning_rate": 3.995029951660777e-06,
+      "loss": 0.8453,
+      "step": 398
+    },
+    {
+      "epoch": 1.1523809523809523,
+      "grad_norm": 0.675423559668541,
+      "learning_rate": 3.971441398464088e-06,
+      "loss": 0.8502,
+      "step": 399
+    },
+    {
+      "epoch": 1.1553113553113552,
+      "grad_norm": 0.6666783325846857,
+      "learning_rate": 3.94787672518116e-06,
+      "loss": 0.764,
+      "step": 400
+    },
+    {
+      "epoch": 1.1582417582417581,
+      "grad_norm": 0.6792452894265647,
+      "learning_rate": 3.924336478910007e-06,
+      "loss": 0.837,
+      "step": 401
+    },
+    {
+      "epoch": 1.1611721611721613,
+      "grad_norm": 0.6695779092340519,
+      "learning_rate": 3.900821206181521e-06,
+      "loss": 0.8553,
+      "step": 402
+    },
+    {
+      "epoch": 1.1641025641025642,
+      "grad_norm": 0.6245571473231076,
+      "learning_rate": 3.8773314529467875e-06,
+      "loss": 0.8628,
+      "step": 403
+    },
+    {
+      "epoch": 1.167032967032967,
+      "grad_norm": 0.6270158605449955,
+      "learning_rate": 3.853867764564409e-06,
+      "loss": 0.7977,
+      "step": 404
+    },
+    {
+      "epoch": 1.16996336996337,
+      "grad_norm": 0.6407892922208382,
+      "learning_rate": 3.830430685787844e-06,
+      "loss": 0.7771,
+      "step": 405
+    },
+    {
+      "epoch": 1.172893772893773,
+      "grad_norm": 0.6223015359803294,
+      "learning_rate": 3.8070207607527587e-06,
+      "loss": 0.8103,
+      "step": 406
+    },
+    {
+      "epoch": 1.1758241758241759,
+      "grad_norm": 0.7174693913851876,
+      "learning_rate": 3.7836385329643933e-06,
+      "loss": 0.8235,
+      "step": 407
+    },
+    {
+      "epoch": 1.1787545787545788,
+      "grad_norm": 0.6872440597677548,
+      "learning_rate": 3.760284545284947e-06,
+      "loss": 0.8005,
+      "step": 408
+    },
+    {
+      "epoch": 1.1816849816849817,
+      "grad_norm": 0.653437969039063,
+      "learning_rate": 3.7369593399209704e-06,
+      "loss": 0.8539,
+      "step": 409
+    },
+    {
+      "epoch": 1.1846153846153846,
+      "grad_norm": 0.6473384600035247,
+      "learning_rate": 3.7136634584107787e-06,
+      "loss": 0.862,
+      "step": 410
+    },
+    {
+      "epoch": 1.1875457875457875,
+      "grad_norm": 0.6751113614880868,
+      "learning_rate": 3.6903974416118786e-06,
+      "loss": 0.8373,
+      "step": 411
+    },
+    {
+      "epoch": 1.1904761904761905,
+      "grad_norm": 0.717978123237241,
+      "learning_rate": 3.6671618296884147e-06,
+      "loss": 0.8654,
+      "step": 412
+    },
+    {
+      "epoch": 1.1934065934065934,
+      "grad_norm": 0.6512581688639196,
+      "learning_rate": 3.6439571620986213e-06,
+      "loss": 0.7892,
+      "step": 413
+    },
+    {
+      "epoch": 1.1963369963369963,
+      "grad_norm": 0.6777361292192612,
+      "learning_rate": 3.620783977582305e-06,
+      "loss": 0.8695,
+      "step": 414
+    },
+    {
+      "epoch": 1.1992673992673992,
+      "grad_norm": 0.7516797430775108,
+      "learning_rate": 3.5976428141483326e-06,
+      "loss": 0.7964,
+      "step": 415
+    },
+    {
+      "epoch": 1.2021978021978021,
+      "grad_norm": 0.6227714905254692,
+      "learning_rate": 3.5745342090621406e-06,
+      "loss": 0.8005,
+      "step": 416
+    },
+    {
+      "epoch": 1.205128205128205,
+      "grad_norm": 0.7119874654244905,
+      "learning_rate": 3.551458698833265e-06,
+      "loss": 0.8489,
+      "step": 417
+    },
+    {
+      "epoch": 1.208058608058608,
+      "grad_norm": 0.6742791369516543,
+      "learning_rate": 3.528416819202881e-06,
+      "loss": 0.8181,
+      "step": 418
+    },
+    {
+      "epoch": 1.210989010989011,
+      "grad_norm": 0.6685215351425255,
+      "learning_rate": 3.5054091051313666e-06,
+      "loss": 0.8546,
+      "step": 419
+    },
+    {
+      "epoch": 1.2139194139194138,
+      "grad_norm": 0.6446298190079693,
+      "learning_rate": 3.4824360907858824e-06,
+      "loss": 0.8238,
+      "step": 420
+    },
+    {
+      "epoch": 1.216849816849817,
+      "grad_norm": 0.6752602907676645,
+      "learning_rate": 3.45949830952797e-06,
+      "loss": 0.8587,
+      "step": 421
+    },
+    {
+      "epoch": 1.2197802197802199,
+      "grad_norm": 0.6536486616817935,
+      "learning_rate": 3.43659629390117e-06,
+      "loss": 0.8482,
+      "step": 422
+    },
+    {
+      "epoch": 1.2227106227106228,
+      "grad_norm": 0.6662928213764988,
+      "learning_rate": 3.4137305756186543e-06,
+      "loss": 0.8215,
+      "step": 423
+    },
+    {
+      "epoch": 1.2256410256410257,
+      "grad_norm": 0.7247368921308519,
+      "learning_rate": 3.390901685550887e-06,
+      "loss": 0.8503,
+      "step": 424
+    },
+    {
+      "epoch": 1.2285714285714286,
+      "grad_norm": 0.7081609071329151,
+      "learning_rate": 3.3681101537132914e-06,
+      "loss": 0.7612,
+      "step": 425
+    },
+    {
+      "epoch": 1.2315018315018316,
+      "grad_norm": 0.6894290798631326,
+      "learning_rate": 3.3453565092539586e-06,
+      "loss": 0.8197,
+      "step": 426
+    },
+    {
+      "epoch": 1.2344322344322345,
+      "grad_norm": 0.6314938401797467,
+      "learning_rate": 3.3226412804413444e-06,
+      "loss": 0.8906,
+      "step": 427
+    },
+    {
+      "epoch": 1.2373626373626374,
+      "grad_norm": 0.700020304527246,
+      "learning_rate": 3.299964994652017e-06,
+      "loss": 0.8305,
+      "step": 428
+    },
+    {
+      "epoch": 1.2402930402930403,
+      "grad_norm": 0.6672421569303719,
+      "learning_rate": 3.2773281783584104e-06,
+      "loss": 0.7996,
+      "step": 429
+    },
+    {
+      "epoch": 1.2432234432234432,
+      "grad_norm": 0.6523896330990022,
+      "learning_rate": 3.254731357116597e-06,
+      "loss": 0.8177,
+      "step": 430
+    },
+    {
+      "epoch": 1.2461538461538462,
+      "grad_norm": 0.7438977310052812,
+      "learning_rate": 3.232175055554091e-06,
+      "loss": 0.7981,
+      "step": 431
+    },
+    {
+      "epoch": 1.249084249084249,
+      "grad_norm": 0.6616594726095909,
+      "learning_rate": 3.209659797357669e-06,
+      "loss": 0.858,
+      "step": 432
+    },
+    {
+      "epoch": 1.252014652014652,
+      "grad_norm": 0.6357225071174895,
+      "learning_rate": 3.187186105261205e-06,
+      "loss": 0.7919,
+      "step": 433
+    },
+    {
+      "epoch": 1.254945054945055,
+      "grad_norm": 0.6645514669695376,
+      "learning_rate": 3.16475450103354e-06,
+      "loss": 0.8445,
+      "step": 434
+    },
+    {
+      "epoch": 1.2578754578754578,
+      "grad_norm": 0.580411587537406,
+      "learning_rate": 3.1423655054663665e-06,
+      "loss": 0.8118,
+      "step": 435
+    },
+    {
+      "epoch": 1.2608058608058608,
+      "grad_norm": 0.6502200181839426,
+      "learning_rate": 3.1200196383621363e-06,
+      "loss": 0.8128,
+      "step": 436
+    },
+    {
+      "epoch": 1.2637362637362637,
+      "grad_norm": 0.7118079489594058,
+      "learning_rate": 3.0977174185219937e-06,
+      "loss": 0.8457,
+      "step": 437
+    },
+    {
+      "epoch": 1.2666666666666666,
+      "grad_norm": 0.6399422284684246,
+      "learning_rate": 3.0754593637337276e-06,
+      "loss": 0.8285,
+      "step": 438
+    },
+    {
+      "epoch": 1.2695970695970695,
+      "grad_norm": 0.6476474872969088,
+      "learning_rate": 3.05324599075976e-06,
+      "loss": 0.8328,
+      "step": 439
+    },
+    {
+      "epoch": 1.2725274725274724,
+      "grad_norm": 0.6527054020963707,
+      "learning_rate": 3.0310778153251325e-06,
+      "loss": 0.8183,
+      "step": 440
+    },
+    {
+      "epoch": 1.2754578754578754,
+      "grad_norm": 0.6892021510002114,
+      "learning_rate": 3.0089553521055455e-06,
+      "loss": 0.8117,
+      "step": 441
+    },
+    {
+      "epoch": 1.2783882783882783,
+      "grad_norm": 0.6938622596098429,
+      "learning_rate": 2.986879114715403e-06,
+      "loss": 0.7997,
+      "step": 442
+    },
+    {
+      "epoch": 1.2813186813186812,
+      "grad_norm": 0.6760167477421419,
+      "learning_rate": 2.964849615695892e-06,
+      "loss": 0.78,
+      "step": 443
+    },
+    {
+      "epoch": 1.2842490842490841,
+      "grad_norm": 0.6630159721868387,
+      "learning_rate": 2.9428673665030772e-06,
+      "loss": 0.7888,
+      "step": 444
+    },
+    {
+      "epoch": 1.287179487179487,
+      "grad_norm": 0.6422717314512306,
+      "learning_rate": 2.9209328774960327e-06,
+      "loss": 0.7959,
+      "step": 445
+    },
+    {
+      "epoch": 1.2901098901098902,
+      "grad_norm": 0.6049124816690329,
+      "learning_rate": 2.899046657924992e-06,
+      "loss": 0.8261,
+      "step": 446
+    },
+    {
+      "epoch": 1.293040293040293,
+      "grad_norm": 0.6538410790869666,
+      "learning_rate": 2.8772092159195193e-06,
+      "loss": 0.7815,
+      "step": 447
+    },
+    {
+      "epoch": 1.295970695970696,
+      "grad_norm": 0.6674785244689053,
+      "learning_rate": 2.8554210584767188e-06,
+      "loss": 0.844,
+      "step": 448
+    },
+    {
+      "epoch": 1.298901098901099,
+      "grad_norm": 0.634998782888407,
+      "learning_rate": 2.8336826914494607e-06,
+      "loss": 0.8069,
+      "step": 449
+    },
+    {
+      "epoch": 1.3018315018315019,
+      "grad_norm": 0.6837741896832135,
+      "learning_rate": 2.8119946195346375e-06,
+      "loss": 0.8087,
+      "step": 450
+    },
+    {
+      "epoch": 1.3047619047619048,
+      "grad_norm": 0.622038243341648,
+      "learning_rate": 2.790357346261445e-06,
+      "loss": 0.8064,
+      "step": 451
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.632121822028228,
+      "learning_rate": 2.7687713739796972e-06,
+      "loss": 0.8385,
+      "step": 452
+    },
+    {
+      "epoch": 1.3106227106227106,
+      "grad_norm": 0.660621163031169,
+      "learning_rate": 2.7472372038481575e-06,
+      "loss": 0.8076,
+      "step": 453
+    },
+    {
+      "epoch": 1.3135531135531135,
+      "grad_norm": 0.6456025313112773,
+      "learning_rate": 2.725755335822903e-06,
+      "loss": 0.8417,
+      "step": 454
+    },
+    {
+      "epoch": 1.3164835164835165,
+      "grad_norm": 0.6339639780226668,
+      "learning_rate": 2.704326268645724e-06,
+      "loss": 0.7822,
+      "step": 455
+    },
+    {
+      "epoch": 1.3194139194139194,
+      "grad_norm": 0.679968992896686,
+      "learning_rate": 2.6829504998325352e-06,
+      "loss": 0.8068,
+      "step": 456
+    },
+    {
+      "epoch": 1.3223443223443223,
+      "grad_norm": 0.6865250171224553,
+      "learning_rate": 2.661628525661835e-06,
+      "loss": 0.8148,
+      "step": 457
+    },
+    {
+      "epoch": 1.3252747252747252,
+      "grad_norm": 0.6684828897934624,
+      "learning_rate": 2.6403608411631744e-06,
+      "loss": 0.8542,
+      "step": 458
+    },
+    {
+      "epoch": 1.3282051282051281,
+      "grad_norm": 0.6758098656078194,
+      "learning_rate": 2.61914794010567e-06,
+      "loss": 0.7891,
+      "step": 459
+    },
+    {
+      "epoch": 1.331135531135531,
+      "grad_norm": 0.6302841257314736,
+      "learning_rate": 2.5979903149865386e-06,
+      "loss": 0.8028,
+      "step": 460
+    },
+    {
+      "epoch": 1.334065934065934,
+      "grad_norm": 0.6195066900214882,
+      "learning_rate": 2.5768884570196615e-06,
+      "loss": 0.8105,
+      "step": 461
+    },
+    {
+      "epoch": 1.3369963369963371,
+      "grad_norm": 0.6527285262735081,
+      "learning_rate": 2.555842856124182e-06,
+      "loss": 0.8416,
+      "step": 462
+    },
+    {
+      "epoch": 1.33992673992674,
+      "grad_norm": 0.6844711291773308,
+      "learning_rate": 2.5348540009131283e-06,
+      "loss": 0.8731,
+      "step": 463
+    },
+    {
+      "epoch": 1.342857142857143,
+      "grad_norm": 0.6917219880994996,
+      "learning_rate": 2.513922378682075e-06,
+      "loss": 0.8617,
+      "step": 464
+    },
+    {
+      "epoch": 1.345787545787546,
+      "grad_norm": 0.7505357241697591,
+      "learning_rate": 2.4930484753978224e-06,
+      "loss": 0.8734,
+      "step": 465
+    },
+    {
+      "epoch": 1.3487179487179488,
+      "grad_norm": 0.6280358853585012,
+      "learning_rate": 2.472232775687119e-06,
+      "loss": 0.8073,
+      "step": 466
+    },
+    {
+      "epoch": 1.3516483516483517,
+      "grad_norm": 0.6357985057852017,
+      "learning_rate": 2.4514757628254075e-06,
+      "loss": 0.8643,
+      "step": 467
+    },
+    {
+      "epoch": 1.3545787545787547,
+      "grad_norm": 0.618985879452095,
+      "learning_rate": 2.4307779187256064e-06,
+      "loss": 0.8588,
+      "step": 468
+    },
+    {
+      "epoch": 1.3575091575091576,
+      "grad_norm": 0.6816901320558723,
+      "learning_rate": 2.4101397239269202e-06,
+      "loss": 0.8974,
+      "step": 469
+    },
+    {
+      "epoch": 1.3604395604395605,
+      "grad_norm": 0.6569693608830611,
+      "learning_rate": 2.389561657583681e-06,
+      "loss": 0.847,
+      "step": 470
+    },
+    {
+      "epoch": 1.3633699633699634,
+      "grad_norm": 0.6722649284724962,
+      "learning_rate": 2.3690441974542322e-06,
+      "loss": 0.8382,
+      "step": 471
+    },
+    {
+      "epoch": 1.3663003663003663,
+      "grad_norm": 0.6738703659901211,
+      "learning_rate": 2.3485878198898253e-06,
+      "loss": 0.8096,
+      "step": 472
+    },
+    {
+      "epoch": 1.3692307692307693,
+      "grad_norm": 0.6199220534522856,
+      "learning_rate": 2.3281929998235664e-06,
+      "loss": 0.8546,
+      "step": 473
+    },
+    {
+      "epoch": 1.3721611721611722,
+      "grad_norm": 0.637774103707959,
+      "learning_rate": 2.30786021075939e-06,
+      "loss": 0.8496,
+      "step": 474
+    },
+    {
+      "epoch": 1.375091575091575,
+      "grad_norm": 0.6745414429909864,
+      "learning_rate": 2.287589924761065e-06,
+      "loss": 0.8613,
+      "step": 475
+    },
+    {
+      "epoch": 1.378021978021978,
+      "grad_norm": 0.6720281480914894,
+      "learning_rate": 2.2673826124412314e-06,
+      "loss": 0.7859,
+      "step": 476
+    },
+    {
+      "epoch": 1.380952380952381,
+      "grad_norm": 0.7515360163059726,
+      "learning_rate": 2.247238742950483e-06,
+      "loss": 0.8675,
+      "step": 477
+    },
+    {
+      "epoch": 1.3838827838827839,
+      "grad_norm": 0.6457379923431958,
+      "learning_rate": 2.2271587839664673e-06,
+      "loss": 0.7631,
+      "step": 478
+    },
+    {
+      "epoch": 1.3868131868131868,
+      "grad_norm": 0.6854082468341819,
+      "learning_rate": 2.2071432016830257e-06,
+      "loss": 0.8032,
+      "step": 479
+    },
+    {
+      "epoch": 1.3897435897435897,
+      "grad_norm": 0.6599311122434474,
+      "learning_rate": 2.18719246079938e-06,
+      "loss": 0.8608,
+      "step": 480
+    },
+    {
+      "epoch": 1.3926739926739926,
+      "grad_norm": 0.6303175902931171,
+      "learning_rate": 2.1673070245093324e-06,
+      "loss": 0.7762,
+      "step": 481
+    },
+    {
+      "epoch": 1.3956043956043955,
+      "grad_norm": 0.6570693148213644,
+      "learning_rate": 2.1474873544905204e-06,
+      "loss": 0.8644,
+      "step": 482
+    },
+    {
+      "epoch": 1.3985347985347985,
+      "grad_norm": 0.6856939265944957,
+      "learning_rate": 2.127733910893693e-06,
+      "loss": 0.8534,
+      "step": 483
+    },
+    {
+      "epoch": 1.4014652014652014,
+      "grad_norm": 0.6239443590408076,
+      "learning_rate": 2.108047152332028e-06,
+      "loss": 0.8024,
+      "step": 484
+    },
+    {
+      "epoch": 1.4043956043956043,
+      "grad_norm": 0.7125717109839642,
+      "learning_rate": 2.0884275358704863e-06,
+      "loss": 0.8313,
+      "step": 485
+    },
+    {
+      "epoch": 1.4073260073260072,
+      "grad_norm": 0.6724094211474902,
+      "learning_rate": 2.0688755170152e-06,
+      "loss": 0.8008,
+      "step": 486
+    },
+    {
+      "epoch": 1.4102564102564101,
+      "grad_norm": 0.6349895866455858,
+      "learning_rate": 2.0493915497028933e-06,
+      "loss": 0.7776,
+      "step": 487
+    },
+    {
+      "epoch": 1.413186813186813,
+      "grad_norm": 0.638362138133294,
+      "learning_rate": 2.029976086290347e-06,
+      "loss": 0.833,
+      "step": 488
+    },
+    {
+      "epoch": 1.4161172161172162,
+      "grad_norm": 0.654564338267121,
+      "learning_rate": 2.0106295775439018e-06,
+      "loss": 0.8209,
+      "step": 489
+    },
+    {
+      "epoch": 1.4190476190476191,
+      "grad_norm": 0.6357815141984277,
+      "learning_rate": 1.9913524726289784e-06,
+      "loss": 0.8488,
+      "step": 490
+    },
+    {
+      "epoch": 1.421978021978022,
+      "grad_norm": 0.684936203470749,
+      "learning_rate": 1.972145219099662e-06,
+      "loss": 0.8553,
+      "step": 491
+    },
+    {
+      "epoch": 1.424908424908425,
+      "grad_norm": 0.7598638244867787,
+      "learning_rate": 1.9530082628883058e-06,
+      "loss": 0.792,
+      "step": 492
+    },
+    {
+      "epoch": 1.4278388278388279,
+      "grad_norm": 0.6233735788941451,
+      "learning_rate": 1.9339420482951794e-06,
+      "loss": 0.8084,
+      "step": 493
+    },
+    {
+      "epoch": 1.4307692307692308,
+      "grad_norm": 0.7179491144851546,
+      "learning_rate": 1.914947017978153e-06,
+      "loss": 0.8926,
+      "step": 494
+    },
+    {
+      "epoch": 1.4336996336996337,
+      "grad_norm": 0.6555466049636046,
+      "learning_rate": 1.896023612942422e-06,
+      "loss": 0.8343,
+      "step": 495
+    },
+    {
+      "epoch": 1.4366300366300366,
+      "grad_norm": 0.645826989084998,
+      "learning_rate": 1.8771722725302644e-06,
+      "loss": 0.7947,
+      "step": 496
+    },
+    {
+      "epoch": 1.4395604395604396,
+      "grad_norm": 0.6482726366595238,
+      "learning_rate": 1.8583934344108446e-06,
+      "loss": 0.7992,
+      "step": 497
+    },
+    {
+      "epoch": 1.4424908424908425,
+      "grad_norm": 0.6376596791410226,
+      "learning_rate": 1.8396875345700498e-06,
+      "loss": 0.8553,
+      "step": 498
+    },
+    {
+      "epoch": 1.4454212454212454,
+      "grad_norm": 0.6903116633272584,
+      "learning_rate": 1.8210550073003701e-06,
+      "loss": 0.8495,
+      "step": 499
+    },
+    {
+      "epoch": 1.4483516483516483,
+      "grad_norm": 0.64069962019644,
+      "learning_rate": 1.8024962851908106e-06,
+      "loss": 0.8217,
+      "step": 500
+    },
+    {
+      "epoch": 1.4512820512820512,
+      "grad_norm": 0.6558252227810359,
+      "learning_rate": 1.7840117991168566e-06,
+      "loss": 0.8505,
+      "step": 501
+    },
+    {
+      "epoch": 1.4542124542124542,
+      "grad_norm": 0.7079418419069864,
+      "learning_rate": 1.7656019782304602e-06,
+      "loss": 0.8472,
+      "step": 502
+    },
+    {
+      "epoch": 1.457142857142857,
+      "grad_norm": 0.705103453747584,
+      "learning_rate": 1.7472672499500837e-06,
+      "loss": 0.8183,
+      "step": 503
+    },
+    {
+      "epoch": 1.46007326007326,
+      "grad_norm": 0.6971164190084713,
+      "learning_rate": 1.729008039950772e-06,
+      "loss": 0.8294,
+      "step": 504
+    },
+    {
+      "epoch": 1.4630036630036631,
+      "grad_norm": 0.6219980479611457,
+      "learning_rate": 1.7108247721542725e-06,
+      "loss": 0.8172,
+      "step": 505
+    },
+    {
+      "epoch": 1.465934065934066,
+      "grad_norm": 0.6616829143182322,
+      "learning_rate": 1.6927178687191953e-06,
+      "loss": 0.8351,
+      "step": 506
+    },
+    {
+      "epoch": 1.468864468864469,
+      "grad_norm": 0.6102810407037182,
+      "learning_rate": 1.6746877500312054e-06,
+      "loss": 0.8267,
+      "step": 507
+    },
+    {
+      "epoch": 1.471794871794872,
+      "grad_norm": 0.6186344377928403,
+      "learning_rate": 1.656734834693266e-06,
+      "loss": 0.8209,
+      "step": 508
+    },
+    {
+      "epoch": 1.4747252747252748,
+      "grad_norm": 0.6028243081632925,
+      "learning_rate": 1.6388595395159207e-06,
+      "loss": 0.8012,
+      "step": 509
+    },
+    {
+      "epoch": 1.4776556776556777,
+      "grad_norm": 0.6358843664209652,
+      "learning_rate": 1.621062279507617e-06,
+      "loss": 0.8406,
+      "step": 510
+    },
+    {
+      "epoch": 1.4805860805860807,
+      "grad_norm": 0.6192894548917643,
+      "learning_rate": 1.6033434678650684e-06,
+      "loss": 0.8013,
+      "step": 511
+    },
+    {
+      "epoch": 1.4835164835164836,
+      "grad_norm": 0.6578313717018371,
+      "learning_rate": 1.5857035159636625e-06,
+      "loss": 0.7908,
+      "step": 512
+    },
+    {
+      "epoch": 1.4864468864468865,
+      "grad_norm": 0.6865547442963875,
+      "learning_rate": 1.5681428333479104e-06,
+      "loss": 0.8147,
+      "step": 513
+    },
+    {
+      "epoch": 1.4893772893772894,
+      "grad_norm": 0.6391710767212113,
+      "learning_rate": 1.550661827721941e-06,
+      "loss": 0.8002,
+      "step": 514
+    },
+    {
+      "epoch": 1.4923076923076923,
+      "grad_norm": 0.6697114193966748,
+      "learning_rate": 1.5332609049400298e-06,
+      "loss": 0.8738,
+      "step": 515
+    },
+    {
+      "epoch": 1.4952380952380953,
+      "grad_norm": 0.6484578925753415,
+      "learning_rate": 1.5159404689971797e-06,
+      "loss": 0.8248,
+      "step": 516
+    },
+    {
+      "epoch": 1.4981684981684982,
+      "grad_norm": 0.6727152397518047,
+      "learning_rate": 1.4987009220197406e-06,
+      "loss": 0.8182,
+      "step": 517
+    },
+    {
+      "epoch": 1.501098901098901,
+      "grad_norm": 0.6222439914425902,
+      "learning_rate": 1.4815426642560753e-06,
+      "loss": 0.7844,
+      "step": 518
+    },
+    {
+      "epoch": 1.504029304029304,
+      "grad_norm": 0.5928698513150097,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.827,
+      "step": 519
+    },
+    {
+      "epoch": 1.506959706959707,
+      "grad_norm": 0.628709550563957,
+      "learning_rate": 1.4474716079178541e-06,
+      "loss": 0.8124,
+      "step": 520
+    },
+    {
+      "epoch": 1.5098901098901099,
+      "grad_norm": 0.6926451262517697,
+      "learning_rate": 1.430559600366665e-06,
+      "loss": 0.801,
+      "step": 521
+    },
+    {
+      "epoch": 1.5128205128205128,
+      "grad_norm": 0.6320651537323465,
+      "learning_rate": 1.4137304640576161e-06,
+      "loss": 0.7778,
+      "step": 522
+    },
+    {
+      "epoch": 1.5157509157509157,
+      "grad_norm": 0.6302945508979699,
+      "learning_rate": 1.3969845897106166e-06,
+      "loss": 0.8078,
+      "step": 523
+    },
+    {
+      "epoch": 1.5186813186813186,
+      "grad_norm": 0.6487308175519874,
+      "learning_rate": 1.3803223661124938e-06,
+      "loss": 0.8335,
+      "step": 524
+    },
+    {
+      "epoch": 1.5216117216117215,
+      "grad_norm": 0.5828653402715643,
+      "learning_rate": 1.3637441801079666e-06,
+      "loss": 0.7563,
+      "step": 525
+    },
+    {
+      "epoch": 1.5245421245421245,
+      "grad_norm": 0.6433886325941865,
+      "learning_rate": 1.3472504165906614e-06,
+      "loss": 0.8079,
+      "step": 526
+    },
+    {
+      "epoch": 1.5274725274725274,
+      "grad_norm": 0.7453574785666953,
+      "learning_rate": 1.330841458494183e-06,
+      "loss": 0.8451,
+      "step": 527
+    },
+    {
+      "epoch": 1.5304029304029303,
+      "grad_norm": 0.6816287097521287,
+      "learning_rate": 1.3145176867832165e-06,
+      "loss": 0.8464,
+      "step": 528
+    },
+    {
+      "epoch": 1.5333333333333332,
+      "grad_norm": 0.6036938583818248,
+      "learning_rate": 1.2982794804446858e-06,
+      "loss": 0.7685,
+      "step": 529
+    },
+    {
+      "epoch": 1.5362637362637361,
+      "grad_norm": 0.6463872641656683,
+      "learning_rate": 1.2821272164789544e-06,
+      "loss": 0.8205,
+      "step": 530
+    },
+    {
+      "epoch": 1.539194139194139,
+      "grad_norm": 0.6230426878687141,
+      "learning_rate": 1.266061269891073e-06,
+      "loss": 0.8272,
+      "step": 531
+    },
+    {
+      "epoch": 1.542124542124542,
+      "grad_norm": 0.6178970771612478,
+      "learning_rate": 1.2500820136820735e-06,
+      "loss": 0.7795,
+      "step": 532
+    },
+    {
+      "epoch": 1.545054945054945,
+      "grad_norm": 0.6002590996506842,
+      "learning_rate": 1.2341898188403068e-06,
+      "loss": 0.7897,
+      "step": 533
+    },
+    {
+      "epoch": 1.547985347985348,
+      "grad_norm": 0.6434409409742198,
+      "learning_rate": 1.2183850543328313e-06,
+      "loss": 0.7969,
+      "step": 534
+    },
+    {
+      "epoch": 1.550915750915751,
+      "grad_norm": 0.6417718520851279,
+      "learning_rate": 1.2026680870968494e-06,
+      "loss": 0.812,
+      "step": 535
+    },
+    {
+      "epoch": 1.5538461538461539,
+      "grad_norm": 0.6425763487232372,
+      "learning_rate": 1.187039282031182e-06,
+      "loss": 0.8454,
+      "step": 536
+    },
+    {
+      "epoch": 1.5567765567765568,
+      "grad_norm": 0.6049459246679286,
+      "learning_rate": 1.171499001987802e-06,
+      "loss": 0.805,
+      "step": 537
+    },
+    {
+      "epoch": 1.5597069597069597,
+      "grad_norm": 0.6857584110796993,
+      "learning_rate": 1.156047607763407e-06,
+      "loss": 0.8451,
+      "step": 538
+    },
+    {
+      "epoch": 1.5626373626373626,
+      "grad_norm": 0.6802912313005144,
+      "learning_rate": 1.1406854580910426e-06,
+      "loss": 0.8573,
+      "step": 539
+    },
+    {
+      "epoch": 1.5655677655677656,
+      "grad_norm": 0.6283637449982267,
+      "learning_rate": 1.1254129096317807e-06,
+      "loss": 0.8468,
+      "step": 540
+    },
+    {
+      "epoch": 1.5684981684981685,
+      "grad_norm": 0.6999745340643583,
+      "learning_rate": 1.110230316966427e-06,
+      "loss": 0.8222,
+      "step": 541
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.6683547293175401,
+      "learning_rate": 1.095138032587298e-06,
+      "loss": 0.8226,
+      "step": 542
+    },
+    {
+      "epoch": 1.5743589743589743,
+      "grad_norm": 0.6395602646701126,
+      "learning_rate": 1.0801364068900334e-06,
+      "loss": 0.7993,
+      "step": 543
+    },
+    {
+      "epoch": 1.5772893772893772,
+      "grad_norm": 0.666258880747308,
+      "learning_rate": 1.0652257881654625e-06,
+      "loss": 0.8272,
+      "step": 544
+    },
+    {
+      "epoch": 1.5802197802197804,
+      "grad_norm": 0.6081578536350295,
+      "learning_rate": 1.0504065225915166e-06,
+      "loss": 0.7984,
+      "step": 545
+    },
+    {
+      "epoch": 1.5831501831501833,
+      "grad_norm": 0.6964583691213134,
+      "learning_rate": 1.0356789542251939e-06,
+      "loss": 0.834,
+      "step": 546
+    },
+    {
+      "epoch": 1.5860805860805862,
+      "grad_norm": 0.6364703392936907,
+      "learning_rate": 1.0210434249945677e-06,
+      "loss": 0.8448,
+      "step": 547
+    },
+    {
+      "epoch": 1.5890109890109891,
+      "grad_norm": 0.6616419125303836,
+      "learning_rate": 1.0065002746908532e-06,
+      "loss": 0.7907,
+      "step": 548
+    },
+    {
+      "epoch": 1.591941391941392,
+      "grad_norm": 0.6089769062029402,
+      "learning_rate": 9.92049840960514e-07,
+      "loss": 0.8168,
+      "step": 549
+    },
+    {
+      "epoch": 1.594871794871795,
+      "grad_norm": 0.6496974721223905,
+      "learning_rate": 9.776924592974257e-07,
+      "loss": 0.7637,
+      "step": 550
+    },
+    {
+      "epoch": 1.597802197802198,
+      "grad_norm": 0.59655757686596,
+      "learning_rate": 9.63428463035085e-07,
+      "loss": 0.7466,
+      "step": 551
+    },
+    {
+      "epoch": 1.6007326007326008,
+      "grad_norm": 0.6797373377804693,
+      "learning_rate": 9.492581833388736e-07,
+      "loss": 0.8493,
+      "step": 552
+    },
+    {
+      "epoch": 1.6036630036630037,
+      "grad_norm": 0.6540093484659664,
+      "learning_rate": 9.351819491983671e-07,
+      "loss": 0.8414,
+      "step": 553
+    },
+    {
+      "epoch": 1.6065934065934067,
+      "grad_norm": 0.6192544935560744,
+      "learning_rate": 9.212000874196953e-07,
+      "loss": 0.7949,
+      "step": 554
+    },
+    {
+      "epoch": 1.6095238095238096,
+      "grad_norm": 0.6413798335452731,
+      "learning_rate": 9.073129226179589e-07,
+      "loss": 0.8367,
+      "step": 555
+    },
+    {
+      "epoch": 1.6124542124542125,
+      "grad_norm": 0.6793141292931023,
+      "learning_rate": 8.935207772096904e-07,
+      "loss": 0.8514,
+      "step": 556
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.6134133075817998,
+      "learning_rate": 8.798239714053692e-07,
+      "loss": 0.7971,
+      "step": 557
+    },
+    {
+      "epoch": 1.6183150183150183,
+      "grad_norm": 0.6483216170375673,
+      "learning_rate": 8.662228232019876e-07,
+      "loss": 0.805,
+      "step": 558
+    },
+    {
+      "epoch": 1.6212454212454213,
+      "grad_norm": 0.6121384170993905,
+      "learning_rate": 8.527176483756671e-07,
+      "loss": 0.8169,
+      "step": 559
+    },
+    {
+      "epoch": 1.6241758241758242,
+      "grad_norm": 0.6068175290937734,
+      "learning_rate": 8.393087604743283e-07,
+      "loss": 0.789,
+      "step": 560
+    },
+    {
+      "epoch": 1.627106227106227,
+      "grad_norm": 0.669787004717848,
+      "learning_rate": 8.259964708104091e-07,
+      "loss": 0.8504,
+      "step": 561
+    },
+    {
+      "epoch": 1.63003663003663,
+      "grad_norm": 0.6283516029134004,
+      "learning_rate": 8.127810884536402e-07,
+      "loss": 0.7839,
+      "step": 562
+    },
+    {
+      "epoch": 1.632967032967033,
+      "grad_norm": 0.6356496138544873,
+      "learning_rate": 7.996629202238654e-07,
+      "loss": 0.8213,
+      "step": 563
+    },
+    {
+      "epoch": 1.6358974358974359,
+      "grad_norm": 0.6409340166886885,
+      "learning_rate": 7.866422706839239e-07,
+      "loss": 0.8164,
+      "step": 564
+    },
+    {
+      "epoch": 1.6388278388278388,
+      "grad_norm": 0.6102325495945577,
+      "learning_rate": 7.737194421325739e-07,
+      "loss": 0.8133,
+      "step": 565
+    },
+    {
+      "epoch": 1.6417582417582417,
+      "grad_norm": 0.6263579578543954,
+      "learning_rate": 7.60894734597476e-07,
+      "loss": 0.8469,
+      "step": 566
+    },
+    {
+      "epoch": 1.6446886446886446,
+      "grad_norm": 0.6110697107228826,
+      "learning_rate": 7.481684458282273e-07,
+      "loss": 0.78,
+      "step": 567
+    },
+    {
+      "epoch": 1.6476190476190475,
+      "grad_norm": 0.623376405697899,
+      "learning_rate": 7.355408712894508e-07,
+      "loss": 0.7951,
+      "step": 568
+    },
+    {
+      "epoch": 1.6505494505494505,
+      "grad_norm": 0.6305238108003742,
+      "learning_rate": 7.23012304153931e-07,
+      "loss": 0.8202,
+      "step": 569
+    },
+    {
+      "epoch": 1.6534798534798534,
+      "grad_norm": 0.6780370401869151,
+      "learning_rate": 7.105830352958143e-07,
+      "loss": 0.8482,
+      "step": 570
+    },
+    {
+      "epoch": 1.6564102564102563,
+      "grad_norm": 0.618188299040753,
+      "learning_rate": 6.982533532838481e-07,
+      "loss": 0.8142,
+      "step": 571
+    },
+    {
+      "epoch": 1.6593406593406592,
+      "grad_norm": 0.6162603773878373,
+      "learning_rate": 6.86023544374686e-07,
+      "loss": 0.8058,
+      "step": 572
+    },
+    {
+      "epoch": 1.6622710622710621,
+      "grad_norm": 0.5988741632305926,
+      "learning_rate": 6.738938925062399e-07,
+      "loss": 0.7909,
+      "step": 573
+    },
+    {
+      "epoch": 1.665201465201465,
+      "grad_norm": 0.6352751552587808,
+      "learning_rate": 6.618646792910893e-07,
+      "loss": 0.8238,
+      "step": 574
+    },
+    {
+      "epoch": 1.668131868131868,
+      "grad_norm": 0.653372698236446,
+      "learning_rate": 6.499361840099421e-07,
+      "loss": 0.8317,
+      "step": 575
+    },
+    {
+      "epoch": 1.671062271062271,
+      "grad_norm": 0.6882980816759717,
+      "learning_rate": 6.381086836051498e-07,
+      "loss": 0.8632,
+      "step": 576
+    },
+    {
+      "epoch": 1.673992673992674,
+      "grad_norm": 0.7343311172190679,
+      "learning_rate": 6.263824526742807e-07,
+      "loss": 0.7804,
+      "step": 577
+    },
+    {
+      "epoch": 1.676923076923077,
+      "grad_norm": 0.6758598163365385,
+      "learning_rate": 6.147577634637413e-07,
+      "loss": 0.8298,
+      "step": 578
+    },
+    {
+      "epoch": 1.6798534798534799,
+      "grad_norm": 0.6132872813883782,
+      "learning_rate": 6.03234885862457e-07,
+      "loss": 0.8346,
+      "step": 579
+    },
+    {
+      "epoch": 1.6827838827838828,
+      "grad_norm": 0.6370171751728234,
+      "learning_rate": 5.918140873956063e-07,
+      "loss": 0.8204,
+      "step": 580
+    },
+    {
+      "epoch": 1.6857142857142857,
+      "grad_norm": 0.6033052646359772,
+      "learning_rate": 5.804956332184092e-07,
+      "loss": 0.8057,
+      "step": 581
+    },
+    {
+      "epoch": 1.6886446886446886,
+      "grad_norm": 0.608486749709341,
+      "learning_rate": 5.692797861099719e-07,
+      "loss": 0.802,
+      "step": 582
+    },
+    {
+      "epoch": 1.6915750915750916,
+      "grad_norm": 0.6205330824144688,
+      "learning_rate": 5.581668064671847e-07,
+      "loss": 0.788,
+      "step": 583
+    },
+    {
+      "epoch": 1.6945054945054945,
+      "grad_norm": 0.6139277202253846,
+      "learning_rate": 5.471569522986775e-07,
+      "loss": 0.808,
+      "step": 584
+    },
+    {
+      "epoch": 1.6974358974358974,
+      "grad_norm": 0.6107771472698981,
+      "learning_rate": 5.3625047921883e-07,
+      "loss": 0.7571,
+      "step": 585
+    },
+    {
+      "epoch": 1.7003663003663003,
+      "grad_norm": 0.63711314046744,
+      "learning_rate": 5.254476404418341e-07,
+      "loss": 0.8345,
+      "step": 586
+    },
+    {
+      "epoch": 1.7032967032967035,
+      "grad_norm": 0.6420503886385798,
+      "learning_rate": 5.147486867758201e-07,
+      "loss": 0.7937,
+      "step": 587
+    },
+    {
+      "epoch": 1.7062271062271064,
+      "grad_norm": 0.6623827998358632,
+      "learning_rate": 5.041538666170282e-07,
+      "loss": 0.8063,
+      "step": 588
+    },
+    {
+      "epoch": 1.7091575091575093,
+      "grad_norm": 0.6311887431693253,
+      "learning_rate": 4.93663425944047e-07,
+      "loss": 0.813,
+      "step": 589
+    },
+    {
+      "epoch": 1.7120879120879122,
+      "grad_norm": 0.6894109166940522,
+      "learning_rate": 4.832776083120983e-07,
+      "loss": 0.8284,
+      "step": 590
+    },
+    {
+      "epoch": 1.7150183150183151,
+      "grad_norm": 0.6344417770301126,
+      "learning_rate": 4.729966548473841e-07,
+      "loss": 0.8173,
+      "step": 591
+    },
+    {
+      "epoch": 1.717948717948718,
+      "grad_norm": 0.6423843551998889,
+      "learning_rate": 4.628208042414889e-07,
+      "loss": 0.7984,
+      "step": 592
+    },
+    {
+      "epoch": 1.720879120879121,
+      "grad_norm": 0.6306008472792638,
+      "learning_rate": 4.5275029274583715e-07,
+      "loss": 0.7507,
+      "step": 593
+    },
+    {
+      "epoch": 1.723809523809524,
+      "grad_norm": 0.6593797867797567,
+      "learning_rate": 4.4278535416620914e-07,
+      "loss": 0.828,
+      "step": 594
+    },
+    {
+      "epoch": 1.7267399267399268,
+      "grad_norm": 0.6859435430823481,
+      "learning_rate": 4.329262198573109e-07,
+      "loss": 0.845,
+      "step": 595
+    },
+    {
+      "epoch": 1.7296703296703297,
+      "grad_norm": 0.668055176901726,
+      "learning_rate": 4.231731187174065e-07,
+      "loss": 0.8841,
+      "step": 596
+    },
+    {
+      "epoch": 1.7326007326007327,
+      "grad_norm": 0.65219754190525,
+      "learning_rate": 4.1352627718299953e-07,
+      "loss": 0.7924,
+      "step": 597
+    },
+    {
+      "epoch": 1.7355311355311356,
+      "grad_norm": 0.6187950860853839,
+      "learning_rate": 4.0398591922357787e-07,
+      "loss": 0.7908,
+      "step": 598
+    },
+    {
+      "epoch": 1.7384615384615385,
+      "grad_norm": 0.6126484385304161,
+      "learning_rate": 3.945522663364154e-07,
+      "loss": 0.817,
+      "step": 599
+    },
+    {
+      "epoch": 1.7413919413919414,
+      "grad_norm": 0.6954213656596391,
+      "learning_rate": 3.852255375414271e-07,
+      "loss": 0.8631,
+      "step": 600
+    },
+    {
+      "epoch": 1.7443223443223443,
+      "grad_norm": 0.5983672679179807,
+      "learning_rate": 3.760059493760848e-07,
+      "loss": 0.8439,
+      "step": 601
+    },
+    {
+      "epoch": 1.7472527472527473,
+      "grad_norm": 0.6702815534109421,
+      "learning_rate": 3.6689371589039013e-07,
+      "loss": 0.81,
+      "step": 602
+    },
+    {
+      "epoch": 1.7501831501831502,
+      "grad_norm": 0.6558071370818314,
+      "learning_rate": 3.5788904864190346e-07,
+      "loss": 0.844,
+      "step": 603
+    },
+    {
+      "epoch": 1.753113553113553,
+      "grad_norm": 0.5830796594516852,
+      "learning_rate": 3.489921566908372e-07,
+      "loss": 0.7941,
+      "step": 604
+    },
+    {
+      "epoch": 1.756043956043956,
+      "grad_norm": 0.5804982345948221,
+      "learning_rate": 3.4020324659519487e-07,
+      "loss": 0.8018,
+      "step": 605
+    },
+    {
+      "epoch": 1.758974358974359,
+      "grad_norm": 0.6178976275532727,
+      "learning_rate": 3.315225224059809e-07,
+      "loss": 0.788,
+      "step": 606
+    },
+    {
+      "epoch": 1.7619047619047619,
+      "grad_norm": 0.6341256961341182,
+      "learning_rate": 3.2295018566246007e-07,
+      "loss": 0.7849,
+      "step": 607
+    },
+    {
+      "epoch": 1.7648351648351648,
+      "grad_norm": 0.6068231207313219,
+      "learning_rate": 3.1448643538748045e-07,
+      "loss": 0.802,
+      "step": 608
+    },
+    {
+      "epoch": 1.7677655677655677,
+      "grad_norm": 0.6249082935288197,
+      "learning_rate": 3.06131468082852e-07,
+      "loss": 0.8624,
+      "step": 609
+    },
+    {
+      "epoch": 1.7706959706959706,
+      "grad_norm": 0.6609124694758736,
+      "learning_rate": 2.9788547772478416e-07,
+      "loss": 0.8283,
+      "step": 610
+    },
+    {
+      "epoch": 1.7736263736263735,
+      "grad_norm": 0.6404899031591438,
+      "learning_rate": 2.897486557593826e-07,
+      "loss": 0.8682,
+      "step": 611
+    },
+    {
+      "epoch": 1.7765567765567765,
+      "grad_norm": 0.6406019381281695,
+      "learning_rate": 2.817211910982037e-07,
+      "loss": 0.7961,
+      "step": 612
+    },
+    {
+      "epoch": 1.7794871794871794,
+      "grad_norm": 0.6050989783512516,
+      "learning_rate": 2.7380327011387096e-07,
+      "loss": 0.7676,
+      "step": 613
+    },
+    {
+      "epoch": 1.7824175824175823,
+      "grad_norm": 0.6309104795298323,
+      "learning_rate": 2.6599507663574387e-07,
+      "loss": 0.8164,
+      "step": 614
+    },
+    {
+      "epoch": 1.7853479853479852,
+      "grad_norm": 0.6893839353883793,
+      "learning_rate": 2.582967919456547e-07,
+      "loss": 0.8334,
+      "step": 615
+    },
+    {
+      "epoch": 1.7882783882783881,
+      "grad_norm": 0.6679193801960764,
+      "learning_rate": 2.5070859477369645e-07,
+      "loss": 0.8355,
+      "step": 616
+    },
+    {
+      "epoch": 1.791208791208791,
+      "grad_norm": 0.597814457069523,
+      "learning_rate": 2.432306612940738e-07,
+      "loss": 0.7758,
+      "step": 617
+    },
+    {
+      "epoch": 1.794139194139194,
+      "grad_norm": 0.673036845863369,
+      "learning_rate": 2.3586316512101416e-07,
+      "loss": 0.815,
+      "step": 618
+    },
+    {
+      "epoch": 1.7970695970695971,
+      "grad_norm": 0.6829081733608965,
+      "learning_rate": 2.286062773047354e-07,
+      "loss": 0.7963,
+      "step": 619
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6663649626280688,
+      "learning_rate": 2.2146016632747624e-07,
+      "loss": 0.7936,
+      "step": 620
+    },
+    {
+      "epoch": 1.802930402930403,
+      "grad_norm": 0.6168148435228321,
+      "learning_rate": 2.1442499809958273e-07,
+      "loss": 0.8128,
+      "step": 621
+    },
+    {
+      "epoch": 1.8058608058608059,
+      "grad_norm": 0.6628611198238746,
+      "learning_rate": 2.0750093595565735e-07,
+      "loss": 0.8198,
+      "step": 622
+    },
+    {
+      "epoch": 1.8087912087912088,
+      "grad_norm": 0.6581682221674444,
+      "learning_rate": 2.006881406507677e-07,
+      "loss": 0.8204,
+      "step": 623
+    },
+    {
+      "epoch": 1.8117216117216117,
+      "grad_norm": 0.6698411962701205,
+      "learning_rate": 1.9398677035671222e-07,
+      "loss": 0.8394,
+      "step": 624
+    },
+    {
+      "epoch": 1.8146520146520146,
+      "grad_norm": 0.60519664432508,
+      "learning_rate": 1.8739698065834977e-07,
+      "loss": 0.819,
+      "step": 625
+    },
+    {
+      "epoch": 1.8175824175824176,
+      "grad_norm": 0.6606157797404275,
+      "learning_rate": 1.8091892454998595e-07,
+      "loss": 0.7976,
+      "step": 626
+    },
+    {
+      "epoch": 1.8205128205128205,
+      "grad_norm": 0.599099515272182,
+      "learning_rate": 1.7455275243182413e-07,
+      "loss": 0.7949,
+      "step": 627
+    },
+    {
+      "epoch": 1.8234432234432234,
+      "grad_norm": 0.6403542722056276,
+      "learning_rate": 1.682986121064689e-07,
+      "loss": 0.8309,
+      "step": 628
+    },
+    {
+      "epoch": 1.8263736263736263,
+      "grad_norm": 0.625773880009336,
+      "learning_rate": 1.6215664877549774e-07,
+      "loss": 0.8118,
+      "step": 629
+    },
+    {
+      "epoch": 1.8293040293040295,
+      "grad_norm": 0.6008354758535722,
+      "learning_rate": 1.561270050360897e-07,
+      "loss": 0.7888,
+      "step": 630
+    },
+    {
+      "epoch": 1.8322344322344324,
+      "grad_norm": 0.6697858393941617,
+      "learning_rate": 1.5020982087771362e-07,
+      "loss": 0.8638,
+      "step": 631
+    },
+    {
+      "epoch": 1.8351648351648353,
+      "grad_norm": 0.6349708657677041,
+      "learning_rate": 1.444052336788787e-07,
+      "loss": 0.8366,
+      "step": 632
+    },
+    {
+      "epoch": 1.8380952380952382,
+      "grad_norm": 0.654164591563027,
+      "learning_rate": 1.3871337820394404e-07,
+      "loss": 0.8657,
+      "step": 633
+    },
+    {
+      "epoch": 1.8410256410256411,
+      "grad_norm": 0.6797327121552983,
+      "learning_rate": 1.33134386599994e-07,
+      "loss": 0.7898,
+      "step": 634
+    },
+    {
+      "epoch": 1.843956043956044,
+      "grad_norm": 0.646590540827072,
+      "learning_rate": 1.2766838839376294e-07,
+      "loss": 0.7954,
+      "step": 635
+    },
+    {
+      "epoch": 1.846886446886447,
+      "grad_norm": 0.6614293793168924,
+      "learning_rate": 1.2231551048863421e-07,
+      "loss": 0.8019,
+      "step": 636
+    },
+    {
+      "epoch": 1.84981684981685,
+      "grad_norm": 0.5837576704664322,
+      "learning_rate": 1.1707587716169033e-07,
+      "loss": 0.7905,
+      "step": 637
+    },
+    {
+      "epoch": 1.8527472527472528,
+      "grad_norm": 0.613725137243765,
+      "learning_rate": 1.1194961006082972e-07,
+      "loss": 0.8264,
+      "step": 638
+    },
+    {
+      "epoch": 1.8556776556776557,
+      "grad_norm": 0.5981158519945132,
+      "learning_rate": 1.0693682820194062e-07,
+      "loss": 0.8222,
+      "step": 639
+    },
+    {
+      "epoch": 1.8586080586080587,
+      "grad_norm": 0.5939810574693122,
+      "learning_rate": 1.0203764796614057e-07,
+      "loss": 0.7919,
+      "step": 640
+    },
+    {
+      "epoch": 1.8615384615384616,
+      "grad_norm": 0.6380868259972026,
+      "learning_rate": 9.725218309707129e-08,
+      "loss": 0.7641,
+      "step": 641
+    },
+    {
+      "epoch": 1.8644688644688645,
+      "grad_norm": 0.6206777770699217,
+      "learning_rate": 9.258054469825972e-08,
+      "loss": 0.8119,
+      "step": 642
+    },
+    {
+      "epoch": 1.8673992673992674,
+      "grad_norm": 0.6299527540389706,
+      "learning_rate": 8.802284123053894e-08,
+      "loss": 0.8057,
+      "step": 643
+    },
+    {
+      "epoch": 1.8703296703296703,
+      "grad_norm": 0.608418876698341,
+      "learning_rate": 8.357917850952802e-08,
+      "loss": 0.8132,
+      "step": 644
+    },
+    {
+      "epoch": 1.8732600732600733,
+      "grad_norm": 0.621128299966682,
+      "learning_rate": 7.924965970317722e-08,
+      "loss": 0.8163,
+      "step": 645
+    },
+    {
+      "epoch": 1.8761904761904762,
+      "grad_norm": 0.655957461443404,
+      "learning_rate": 7.503438532937169e-08,
+      "loss": 0.8134,
+      "step": 646
+    },
+    {
+      "epoch": 1.879120879120879,
+      "grad_norm": 0.6306340219082865,
+      "learning_rate": 7.093345325359935e-08,
+      "loss": 0.8143,
+      "step": 647
+    },
+    {
+      "epoch": 1.882051282051282,
+      "grad_norm": 0.6451879681808641,
+      "learning_rate": 6.694695868667556e-08,
+      "loss": 0.8332,
+      "step": 648
+    },
+    {
+      "epoch": 1.884981684981685,
+      "grad_norm": 0.6742284480796318,
+      "learning_rate": 6.307499418253705e-08,
+      "loss": 0.8388,
+      "step": 649
+    },
+    {
+      "epoch": 1.8879120879120879,
+      "grad_norm": 0.6205039778686892,
+      "learning_rate": 5.9317649636088656e-08,
+      "loss": 0.8091,
+      "step": 650
+    },
+    {
+      "epoch": 1.8908424908424908,
+      "grad_norm": 0.6580483136532429,
+      "learning_rate": 5.5675012281119486e-08,
+      "loss": 0.8195,
+      "step": 651
+    },
+    {
+      "epoch": 1.8937728937728937,
+      "grad_norm": 0.6485138656930131,
+      "learning_rate": 5.214716668827558e-08,
+      "loss": 0.8119,
+      "step": 652
+    },
+    {
+      "epoch": 1.8967032967032966,
+      "grad_norm": 0.6013349089531645,
+      "learning_rate": 4.8734194763098706e-08,
+      "loss": 0.8124,
+      "step": 653
+    },
+    {
+      "epoch": 1.8996336996336995,
+      "grad_norm": 0.6442373339607835,
+      "learning_rate": 4.543617574412185e-08,
+      "loss": 0.7937,
+      "step": 654
+    },
+    {
+      "epoch": 1.9025641025641025,
+      "grad_norm": 0.6410125740001468,
+      "learning_rate": 4.2253186201031135e-08,
+      "loss": 0.8208,
+      "step": 655
+    },
+    {
+      "epoch": 1.9054945054945054,
+      "grad_norm": 0.6874802512487508,
+      "learning_rate": 3.9185300032889005e-08,
+      "loss": 0.7817,
+      "step": 656
+    },
+    {
+      "epoch": 1.9084249084249083,
+      "grad_norm": 0.6174107918080415,
+      "learning_rate": 3.6232588466417195e-08,
+      "loss": 0.8081,
+      "step": 657
+    },
+    {
+      "epoch": 1.9113553113553112,
+      "grad_norm": 0.6069693810447614,
+      "learning_rate": 3.339512005434309e-08,
+      "loss": 0.8471,
+      "step": 658
+    },
+    {
+      "epoch": 1.9142857142857141,
+      "grad_norm": 0.6571252565902649,
+      "learning_rate": 3.0672960673808205e-08,
+      "loss": 0.7734,
+      "step": 659
+    },
+    {
+      "epoch": 1.917216117216117,
+      "grad_norm": 0.6346303136090871,
+      "learning_rate": 2.8066173524839978e-08,
+      "loss": 0.8176,
+      "step": 660
+    },
+    {
+      "epoch": 1.92014652014652,
+      "grad_norm": 0.6209889706057522,
+      "learning_rate": 2.5574819128882933e-08,
+      "loss": 0.8267,
+      "step": 661
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.7020907999054834,
+      "learning_rate": 2.319895532739369e-08,
+      "loss": 0.8058,
+      "step": 662
+    },
+    {
+      "epoch": 1.926007326007326,
+      "grad_norm": 0.6758173840664298,
+      "learning_rate": 2.093863728049872e-08,
+      "loss": 0.8283,
+      "step": 663
+    },
+    {
+      "epoch": 1.928937728937729,
+      "grad_norm": 0.6385882790006528,
+      "learning_rate": 1.8793917465713686e-08,
+      "loss": 0.8188,
+      "step": 664
+    },
+    {
+      "epoch": 1.9318681318681319,
+      "grad_norm": 0.6438874910701458,
+      "learning_rate": 1.676484567672554e-08,
+      "loss": 0.8532,
+      "step": 665
+    },
+    {
+      "epoch": 1.9347985347985348,
+      "grad_norm": 0.7048203364520762,
+      "learning_rate": 1.4851469022234e-08,
+      "loss": 0.822,
+      "step": 666
+    },
+    {
+      "epoch": 1.9377289377289377,
+      "grad_norm": 0.6322493105177235,
+      "learning_rate": 1.3053831924861315e-08,
+      "loss": 0.8076,
+      "step": 667
+    },
+    {
+      "epoch": 1.9406593406593406,
+      "grad_norm": 0.6610364468359425,
+      "learning_rate": 1.137197612011809e-08,
+      "loss": 0.8645,
+      "step": 668
+    },
+    {
+      "epoch": 1.9435897435897436,
+      "grad_norm": 0.6200915881326821,
+      "learning_rate": 9.805940655436274e-09,
+      "loss": 0.7929,
+      "step": 669
+    },
+    {
+      "epoch": 1.9465201465201465,
+      "grad_norm": 0.6560677264326054,
+      "learning_rate": 8.35576188926046e-09,
+      "loss": 0.8775,
+      "step": 670
+    },
+    {
+      "epoch": 1.9494505494505494,
+      "grad_norm": 0.6631827694048288,
+      "learning_rate": 7.0214734902074314e-09,
+      "loss": 0.7636,
+      "step": 671
+    },
+    {
+      "epoch": 1.9523809523809523,
+      "grad_norm": 0.7712830976554526,
+      "learning_rate": 5.803106436279571e-09,
+      "loss": 0.8476,
+      "step": 672
+    },
+    {
+      "epoch": 1.9553113553113555,
+      "grad_norm": 0.6031190635724369,
+      "learning_rate": 4.700689014149329e-09,
+      "loss": 0.7796,
+      "step": 673
+    },
+    {
+      "epoch": 1.9582417582417584,
+      "grad_norm": 0.5840130736388364,
+      "learning_rate": 3.71424681850141e-09,
+      "loss": 0.7594,
+      "step": 674
+    },
+    {
+      "epoch": 1.9611721611721613,
+      "grad_norm": 0.652441280672971,
+      "learning_rate": 2.8438027514382517e-09,
+      "loss": 0.8257,
+      "step": 675
+    },
+    {
+      "epoch": 1.9641025641025642,
+      "grad_norm": 0.6572849884958872,
+      "learning_rate": 2.0893770219493347e-09,
+      "loss": 0.8316,
+      "step": 676
+    },
+    {
+      "epoch": 1.9670329670329672,
+      "grad_norm": 0.6233858205449643,
+      "learning_rate": 1.450987145439342e-09,
+      "loss": 0.8204,
+      "step": 677
+    },
+    {
+      "epoch": 1.96996336996337,
+      "grad_norm": 0.6331674610122936,
+      "learning_rate": 9.286479433257e-10,
+      "loss": 0.7992,
+      "step": 678
+    },
+    {
+      "epoch": 1.972893772893773,
+      "grad_norm": 0.7160803129297684,
+      "learning_rate": 5.2237154268997e-10,
+      "loss": 0.8252,
+      "step": 679
+    },
+    {
+      "epoch": 1.975824175824176,
+      "grad_norm": 0.7332845117682134,
+      "learning_rate": 2.321673760002918e-10,
+      "loss": 0.8221,
+      "step": 680
+    },
+    {
+      "epoch": 1.9787545787545788,
+      "grad_norm": 0.5902309209793818,
+      "learning_rate": 5.804218088933944e-11,
+      "loss": 0.8015,
+      "step": 681
+    },
+    {
+      "epoch": 1.9816849816849818,
+      "grad_norm": 0.6257835194161055,
+      "learning_rate": 0.0,
+      "loss": 0.8461,
+      "step": 682
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 682,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 171,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1064222230708224.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}