diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6594 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9936034115138592,
+  "eval_steps": 500,
+  "global_step": 936,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0031982942430703624,
+      "grad_norm": 7.622738078733055,
+      "learning_rate": 1.0638297872340426e-07,
+      "loss": 1.1909,
+      "step": 1
+    },
+    {
+      "epoch": 0.006396588486140725,
+      "grad_norm": 7.574176961529125,
+      "learning_rate": 2.1276595744680852e-07,
+      "loss": 1.1838,
+      "step": 2
+    },
+    {
+      "epoch": 0.009594882729211088,
+      "grad_norm": 7.4636690419508005,
+      "learning_rate": 3.1914893617021275e-07,
+      "loss": 1.1971,
+      "step": 3
+    },
+    {
+      "epoch": 0.01279317697228145,
+      "grad_norm": 7.513735532368614,
+      "learning_rate": 4.2553191489361704e-07,
+      "loss": 1.1795,
+      "step": 4
+    },
+    {
+      "epoch": 0.015991471215351813,
+      "grad_norm": 7.333440889375827,
+      "learning_rate": 5.319148936170213e-07,
+      "loss": 1.1448,
+      "step": 5
+    },
+    {
+      "epoch": 0.019189765458422176,
+      "grad_norm": 7.3639120181383575,
+      "learning_rate": 6.382978723404255e-07,
+      "loss": 1.1893,
+      "step": 6
+    },
+    {
+      "epoch": 0.022388059701492536,
+      "grad_norm": 7.438753810381023,
+      "learning_rate": 7.446808510638298e-07,
+      "loss": 1.2042,
+      "step": 7
+    },
+    {
+      "epoch": 0.0255863539445629,
+      "grad_norm": 6.96598869339596,
+      "learning_rate": 8.510638297872341e-07,
+      "loss": 1.1527,
+      "step": 8
+    },
+    {
+      "epoch": 0.028784648187633263,
+      "grad_norm": 6.606115410466868,
+      "learning_rate": 9.574468085106384e-07,
+      "loss": 1.1691,
+      "step": 9
+    },
+    {
+      "epoch": 0.031982942430703626,
+      "grad_norm": 6.428618826670432,
+      "learning_rate": 1.0638297872340427e-06,
+      "loss": 1.1713,
+      "step": 10
+    },
+    {
+      "epoch": 0.035181236673773986,
+      "grad_norm": 5.30277143517507,
+      "learning_rate": 1.170212765957447e-06,
+      "loss": 1.1162,
+      "step": 11
+    },
+    {
+      "epoch": 0.03837953091684435,
+      "grad_norm": 5.216323252768823,
+      "learning_rate": 1.276595744680851e-06,
+      "loss": 1.1101,
+      "step": 12
+    },
+    {
+      "epoch": 0.04157782515991471,
+      "grad_norm": 5.091458168528318,
+      "learning_rate": 1.3829787234042555e-06,
+      "loss": 1.1071,
+      "step": 13
+    },
+    {
+      "epoch": 0.04477611940298507,
+      "grad_norm": 3.3709915569255746,
+      "learning_rate": 1.4893617021276596e-06,
+      "loss": 1.0666,
+      "step": 14
+    },
+    {
+      "epoch": 0.04797441364605544,
+      "grad_norm": 3.3278073990556303,
+      "learning_rate": 1.595744680851064e-06,
+      "loss": 1.0571,
+      "step": 15
+    },
+    {
+      "epoch": 0.0511727078891258,
+      "grad_norm": 3.1457570812535685,
+      "learning_rate": 1.7021276595744682e-06,
+      "loss": 1.0419,
+      "step": 16
+    },
+    {
+      "epoch": 0.054371002132196165,
+      "grad_norm": 3.0524886673021165,
+      "learning_rate": 1.8085106382978727e-06,
+      "loss": 1.0643,
+      "step": 17
+    },
+    {
+      "epoch": 0.057569296375266525,
+      "grad_norm": 2.9940425825944055,
+      "learning_rate": 1.9148936170212767e-06,
+      "loss": 1.0184,
+      "step": 18
+    },
+    {
+      "epoch": 0.060767590618336885,
+      "grad_norm": 3.353851239083325,
+      "learning_rate": 2.021276595744681e-06,
+      "loss": 1.0121,
+      "step": 19
+    },
+    {
+      "epoch": 0.06396588486140725,
+      "grad_norm": 3.7889395105244073,
+      "learning_rate": 2.1276595744680853e-06,
+      "loss": 0.9889,
+      "step": 20
+    },
+    {
+      "epoch": 0.06716417910447761,
+      "grad_norm": 3.8783603137474665,
+      "learning_rate": 2.2340425531914894e-06,
+      "loss": 0.996,
+      "step": 21
+    },
+    {
+      "epoch": 0.07036247334754797,
+      "grad_norm": 3.5352669616023475,
+      "learning_rate": 2.340425531914894e-06,
+      "loss": 0.9742,
+      "step": 22
+    },
+    {
+      "epoch": 0.07356076759061833,
+      "grad_norm": 3.2857152101090605,
+      "learning_rate": 2.446808510638298e-06,
+      "loss": 0.953,
+      "step": 23
+    },
+    {
+      "epoch": 0.0767590618336887,
+      "grad_norm": 2.7570947843045857,
+      "learning_rate": 2.553191489361702e-06,
+      "loss": 0.9587,
+      "step": 24
+    },
+    {
+      "epoch": 0.07995735607675906,
+      "grad_norm": 1.9278968128889937,
+      "learning_rate": 2.6595744680851065e-06,
+      "loss": 0.9329,
+      "step": 25
+    },
+    {
+      "epoch": 0.08315565031982942,
+      "grad_norm": 1.78597612548795,
+      "learning_rate": 2.765957446808511e-06,
+      "loss": 0.9337,
+      "step": 26
+    },
+    {
+      "epoch": 0.08635394456289978,
+      "grad_norm": 1.9761748056190767,
+      "learning_rate": 2.8723404255319155e-06,
+      "loss": 0.9235,
+      "step": 27
+    },
+    {
+      "epoch": 0.08955223880597014,
+      "grad_norm": 1.9959330898097134,
+      "learning_rate": 2.978723404255319e-06,
+      "loss": 0.9095,
+      "step": 28
+    },
+    {
+      "epoch": 0.09275053304904052,
+      "grad_norm": 1.6188409417305705,
+      "learning_rate": 3.0851063829787237e-06,
+      "loss": 0.8924,
+      "step": 29
+    },
+    {
+      "epoch": 0.09594882729211088,
+      "grad_norm": 1.5343331514431935,
+      "learning_rate": 3.191489361702128e-06,
+      "loss": 0.8759,
+      "step": 30
+    },
+    {
+      "epoch": 0.09914712153518124,
+      "grad_norm": 1.431667627920631,
+      "learning_rate": 3.297872340425532e-06,
+      "loss": 0.8694,
+      "step": 31
+    },
+    {
+      "epoch": 0.1023454157782516,
+      "grad_norm": 1.2543200154352037,
+      "learning_rate": 3.4042553191489363e-06,
+      "loss": 0.8512,
+      "step": 32
+    },
+    {
+      "epoch": 0.10554371002132196,
+      "grad_norm": 1.0539218969611874,
+      "learning_rate": 3.510638297872341e-06,
+      "loss": 0.8458,
+      "step": 33
+    },
+    {
+      "epoch": 0.10874200426439233,
+      "grad_norm": 1.1170568972274193,
+      "learning_rate": 3.6170212765957453e-06,
+      "loss": 0.8899,
+      "step": 34
+    },
+    {
+      "epoch": 0.11194029850746269,
+      "grad_norm": 1.2202744700302341,
+      "learning_rate": 3.723404255319149e-06,
+      "loss": 0.8393,
+      "step": 35
+    },
+    {
+      "epoch": 0.11513859275053305,
+      "grad_norm": 1.0830349710322051,
+      "learning_rate": 3.8297872340425535e-06,
+      "loss": 0.871,
+      "step": 36
+    },
+    {
+      "epoch": 0.11833688699360341,
+      "grad_norm": 0.9052734617810034,
+      "learning_rate": 3.936170212765958e-06,
+      "loss": 0.8202,
+      "step": 37
+    },
+    {
+      "epoch": 0.12153518123667377,
+      "grad_norm": 0.8993984549295001,
+      "learning_rate": 4.042553191489362e-06,
+      "loss": 0.801,
+      "step": 38
+    },
+    {
+      "epoch": 0.12473347547974413,
+      "grad_norm": 0.9725408899239824,
+      "learning_rate": 4.148936170212766e-06,
+      "loss": 0.8701,
+      "step": 39
+    },
+    {
+      "epoch": 0.1279317697228145,
+      "grad_norm": 0.943110853985239,
+      "learning_rate": 4.255319148936171e-06,
+      "loss": 0.8114,
+      "step": 40
+    },
+    {
+      "epoch": 0.13113006396588486,
+      "grad_norm": 0.9352658574729733,
+      "learning_rate": 4.361702127659575e-06,
+      "loss": 0.8401,
+      "step": 41
+    },
+    {
+      "epoch": 0.13432835820895522,
+      "grad_norm": 1.0170142598852345,
+      "learning_rate": 4.468085106382979e-06,
+      "loss": 0.8198,
+      "step": 42
+    },
+    {
+      "epoch": 0.13752665245202558,
+      "grad_norm": 0.9009283876561643,
+      "learning_rate": 4.574468085106383e-06,
+      "loss": 0.8055,
+      "step": 43
+    },
+    {
+      "epoch": 0.14072494669509594,
+      "grad_norm": 0.8173324407768994,
+      "learning_rate": 4.680851063829788e-06,
+      "loss": 0.8042,
+      "step": 44
+    },
+    {
+      "epoch": 0.1439232409381663,
+      "grad_norm": 0.8483547771193565,
+      "learning_rate": 4.787234042553192e-06,
+      "loss": 0.7773,
+      "step": 45
+    },
+    {
+      "epoch": 0.14712153518123666,
+      "grad_norm": 0.7692184274159833,
+      "learning_rate": 4.893617021276596e-06,
+      "loss": 0.7997,
+      "step": 46
+    },
+    {
+      "epoch": 0.15031982942430705,
+      "grad_norm": 0.81536863431857,
+      "learning_rate": 5e-06,
+      "loss": 0.7892,
+      "step": 47
+    },
+    {
+      "epoch": 0.1535181236673774,
+      "grad_norm": 0.8158497316472542,
+      "learning_rate": 5.106382978723404e-06,
+      "loss": 0.7978,
+      "step": 48
+    },
+    {
+      "epoch": 0.15671641791044777,
+      "grad_norm": 0.9305475572050518,
+      "learning_rate": 5.212765957446809e-06,
+      "loss": 0.8087,
+      "step": 49
+    },
+    {
+      "epoch": 0.15991471215351813,
+      "grad_norm": 0.8236050831503633,
+      "learning_rate": 5.319148936170213e-06,
+      "loss": 0.7607,
+      "step": 50
+    },
+    {
+      "epoch": 0.1631130063965885,
+      "grad_norm": 0.7846154878148738,
+      "learning_rate": 5.425531914893617e-06,
+      "loss": 0.8175,
+      "step": 51
+    },
+    {
+      "epoch": 0.16631130063965885,
+      "grad_norm": 0.9528828885498349,
+      "learning_rate": 5.531914893617022e-06,
+      "loss": 0.791,
+      "step": 52
+    },
+    {
+      "epoch": 0.1695095948827292,
+      "grad_norm": 0.7710572192032797,
+      "learning_rate": 5.638297872340426e-06,
+      "loss": 0.7986,
+      "step": 53
+    },
+    {
+      "epoch": 0.17270788912579957,
+      "grad_norm": 0.8318885111022261,
+      "learning_rate": 5.744680851063831e-06,
+      "loss": 0.7732,
+      "step": 54
+    },
+    {
+      "epoch": 0.17590618336886993,
+      "grad_norm": 0.8923328928381091,
+      "learning_rate": 5.851063829787235e-06,
+      "loss": 0.7506,
+      "step": 55
+    },
+    {
+      "epoch": 0.1791044776119403,
+      "grad_norm": 0.8476738749734117,
+      "learning_rate": 5.957446808510638e-06,
+      "loss": 0.8041,
+      "step": 56
+    },
+    {
+      "epoch": 0.18230277185501065,
+      "grad_norm": 0.6991963183642709,
+      "learning_rate": 6.063829787234044e-06,
+      "loss": 0.743,
+      "step": 57
+    },
+    {
+      "epoch": 0.18550106609808104,
+      "grad_norm": 0.9527151297958727,
+      "learning_rate": 6.170212765957447e-06,
+      "loss": 0.7783,
+      "step": 58
+    },
+    {
+      "epoch": 0.1886993603411514,
+      "grad_norm": 0.8889593346162057,
+      "learning_rate": 6.276595744680851e-06,
+      "loss": 0.7437,
+      "step": 59
+    },
+    {
+      "epoch": 0.19189765458422176,
+      "grad_norm": 0.6699184888475535,
+      "learning_rate": 6.382978723404256e-06,
+      "loss": 0.7714,
+      "step": 60
+    },
+    {
+      "epoch": 0.19509594882729211,
+      "grad_norm": 0.7705563645271767,
+      "learning_rate": 6.48936170212766e-06,
+      "loss": 0.7744,
+      "step": 61
+    },
+    {
+      "epoch": 0.19829424307036247,
+      "grad_norm": 0.8167415683795367,
+      "learning_rate": 6.595744680851064e-06,
+      "loss": 0.7561,
+      "step": 62
+    },
+    {
+      "epoch": 0.20149253731343283,
+      "grad_norm": 0.81962712226257,
+      "learning_rate": 6.702127659574469e-06,
+      "loss": 0.7891,
+      "step": 63
+    },
+    {
+      "epoch": 0.2046908315565032,
+      "grad_norm": 0.6986640424503381,
+      "learning_rate": 6.808510638297873e-06,
+      "loss": 0.7572,
+      "step": 64
+    },
+    {
+      "epoch": 0.20788912579957355,
+      "grad_norm": 0.756369209825303,
+      "learning_rate": 6.914893617021278e-06,
+      "loss": 0.7458,
+      "step": 65
+    },
+    {
+      "epoch": 0.21108742004264391,
+      "grad_norm": 0.7270975619851376,
+      "learning_rate": 7.021276595744682e-06,
+      "loss": 0.7116,
+      "step": 66
+    },
+    {
+      "epoch": 0.21428571428571427,
+      "grad_norm": 0.7077904574727415,
+      "learning_rate": 7.127659574468085e-06,
+      "loss": 0.741,
+      "step": 67
+    },
+    {
+      "epoch": 0.21748400852878466,
+      "grad_norm": 0.7367201939644344,
+      "learning_rate": 7.234042553191491e-06,
+      "loss": 0.738,
+      "step": 68
+    },
+    {
+      "epoch": 0.22068230277185502,
+      "grad_norm": 0.7461379677761271,
+      "learning_rate": 7.340425531914894e-06,
+      "loss": 0.7402,
+      "step": 69
+    },
+    {
+      "epoch": 0.22388059701492538,
+      "grad_norm": 0.7635469236526181,
+      "learning_rate": 7.446808510638298e-06,
+      "loss": 0.7539,
+      "step": 70
+    },
+    {
+      "epoch": 0.22707889125799574,
+      "grad_norm": 0.8662247901922688,
+      "learning_rate": 7.553191489361703e-06,
+      "loss": 0.7515,
+      "step": 71
+    },
+    {
+      "epoch": 0.2302771855010661,
+      "grad_norm": 0.7957789030344851,
+      "learning_rate": 7.659574468085107e-06,
+      "loss": 0.7416,
+      "step": 72
+    },
+    {
+      "epoch": 0.23347547974413646,
+      "grad_norm": 0.7423199681203461,
+      "learning_rate": 7.765957446808511e-06,
+      "loss": 0.7629,
+      "step": 73
+    },
+    {
+      "epoch": 0.23667377398720682,
+      "grad_norm": 0.9173283664095695,
+      "learning_rate": 7.872340425531916e-06,
+      "loss": 0.7452,
+      "step": 74
+    },
+    {
+      "epoch": 0.23987206823027718,
+      "grad_norm": 0.78368199600868,
+      "learning_rate": 7.97872340425532e-06,
+      "loss": 0.7268,
+      "step": 75
+    },
+    {
+      "epoch": 0.24307036247334754,
+      "grad_norm": 0.8361765532940366,
+      "learning_rate": 8.085106382978723e-06,
+      "loss": 0.7584,
+      "step": 76
+    },
+    {
+      "epoch": 0.2462686567164179,
+      "grad_norm": 0.809289414930847,
+      "learning_rate": 8.191489361702128e-06,
+      "loss": 0.7323,
+      "step": 77
+    },
+    {
+      "epoch": 0.24946695095948826,
+      "grad_norm": 0.7351470133306395,
+      "learning_rate": 8.297872340425532e-06,
+      "loss": 0.7363,
+      "step": 78
+    },
+    {
+      "epoch": 0.2526652452025586,
+      "grad_norm": 0.8056016102088669,
+      "learning_rate": 8.404255319148937e-06,
+      "loss": 0.723,
+      "step": 79
+    },
+    {
+      "epoch": 0.255863539445629,
+      "grad_norm": 0.7345894297808727,
+      "learning_rate": 8.510638297872341e-06,
+      "loss": 0.769,
+      "step": 80
+    },
+    {
+      "epoch": 0.25906183368869934,
+      "grad_norm": 0.7863593332470072,
+      "learning_rate": 8.617021276595746e-06,
+      "loss": 0.728,
+      "step": 81
+    },
+    {
+      "epoch": 0.2622601279317697,
+      "grad_norm": 0.7627382259597176,
+      "learning_rate": 8.72340425531915e-06,
+      "loss": 0.7283,
+      "step": 82
+    },
+    {
+      "epoch": 0.26545842217484006,
+      "grad_norm": 0.7824566533257207,
+      "learning_rate": 8.829787234042555e-06,
+      "loss": 0.7529,
+      "step": 83
+    },
+    {
+      "epoch": 0.26865671641791045,
+      "grad_norm": 0.7231171750656687,
+      "learning_rate": 8.936170212765958e-06,
+      "loss": 0.7335,
+      "step": 84
+    },
+    {
+      "epoch": 0.27185501066098083,
+      "grad_norm": 0.9121689099240826,
+      "learning_rate": 9.042553191489362e-06,
+      "loss": 0.756,
+      "step": 85
+    },
+    {
+      "epoch": 0.27505330490405117,
+      "grad_norm": 0.7192041394994152,
+      "learning_rate": 9.148936170212767e-06,
+      "loss": 0.7008,
+      "step": 86
+    },
+    {
+      "epoch": 0.27825159914712155,
+      "grad_norm": 0.926261732519985,
+      "learning_rate": 9.255319148936171e-06,
+      "loss": 0.7169,
+      "step": 87
+    },
+    {
+      "epoch": 0.2814498933901919,
+      "grad_norm": 0.7106373516758131,
+      "learning_rate": 9.361702127659576e-06,
+      "loss": 0.7019,
+      "step": 88
+    },
+    {
+      "epoch": 0.2846481876332623,
+      "grad_norm": 0.8650610615775703,
+      "learning_rate": 9.46808510638298e-06,
+      "loss": 0.7579,
+      "step": 89
+    },
+    {
+      "epoch": 0.2878464818763326,
+      "grad_norm": 0.8537551564881841,
+      "learning_rate": 9.574468085106385e-06,
+      "loss": 0.7151,
+      "step": 90
+    },
+    {
+      "epoch": 0.291044776119403,
+      "grad_norm": 0.9816090964885784,
+      "learning_rate": 9.680851063829787e-06,
+      "loss": 0.7361,
+      "step": 91
+    },
+    {
+      "epoch": 0.2942430703624733,
+      "grad_norm": 0.7896344914068675,
+      "learning_rate": 9.787234042553192e-06,
+      "loss": 0.7241,
+      "step": 92
+    },
+    {
+      "epoch": 0.2974413646055437,
+      "grad_norm": 0.8361537117215864,
+      "learning_rate": 9.893617021276596e-06,
+      "loss": 0.7063,
+      "step": 93
+    },
+    {
+      "epoch": 0.3006396588486141,
+      "grad_norm": 0.8817821003662552,
+      "learning_rate": 1e-05,
+      "loss": 0.711,
+      "step": 94
+    },
+    {
+      "epoch": 0.30383795309168443,
+      "grad_norm": 1.1654827546629623,
+      "learning_rate": 9.999965197129365e-06,
+      "loss": 0.7493,
+      "step": 95
+    },
+    {
+      "epoch": 0.3070362473347548,
+      "grad_norm": 0.6905754899045732,
+      "learning_rate": 9.999860789001947e-06,
+      "loss": 0.7189,
+      "step": 96
+    },
+    {
+      "epoch": 0.31023454157782515,
+      "grad_norm": 1.0193923266199096,
+      "learning_rate": 9.999686777071233e-06,
+      "loss": 0.748,
+      "step": 97
+    },
+    {
+      "epoch": 0.31343283582089554,
+      "grad_norm": 0.8631527647207492,
+      "learning_rate": 9.999443163759669e-06,
+      "loss": 0.7297,
+      "step": 98
+    },
+    {
+      "epoch": 0.31663113006396587,
+      "grad_norm": 0.7959272086157606,
+      "learning_rate": 9.999129952458628e-06,
+      "loss": 0.6892,
+      "step": 99
+    },
+    {
+      "epoch": 0.31982942430703626,
+      "grad_norm": 1.019986780308216,
+      "learning_rate": 9.998747147528375e-06,
+      "loss": 0.6945,
+      "step": 100
+    },
+    {
+      "epoch": 0.3230277185501066,
+      "grad_norm": 0.9796997640078874,
+      "learning_rate": 9.998294754297992e-06,
+      "loss": 0.7683,
+      "step": 101
+    },
+    {
+      "epoch": 0.326226012793177,
+      "grad_norm": 0.9025864669755898,
+      "learning_rate": 9.997772779065312e-06,
+      "loss": 0.707,
+      "step": 102
+    },
+    {
+      "epoch": 0.3294243070362473,
+      "grad_norm": 0.9406972010353718,
+      "learning_rate": 9.997181229096831e-06,
+      "loss": 0.7148,
+      "step": 103
+    },
+    {
+      "epoch": 0.3326226012793177,
+      "grad_norm": 0.8270166326880105,
+      "learning_rate": 9.996520112627602e-06,
+      "loss": 0.7217,
+      "step": 104
+    },
+    {
+      "epoch": 0.3358208955223881,
+      "grad_norm": 0.8843538366343328,
+      "learning_rate": 9.995789438861128e-06,
+      "loss": 0.7195,
+      "step": 105
+    },
+    {
+      "epoch": 0.3390191897654584,
+      "grad_norm": 0.8701458072422218,
+      "learning_rate": 9.994989217969224e-06,
+      "loss": 0.7175,
+      "step": 106
+    },
+    {
+      "epoch": 0.3422174840085288,
+      "grad_norm": 0.7418076755999989,
+      "learning_rate": 9.994119461091885e-06,
+      "loss": 0.7158,
+      "step": 107
+    },
+    {
+      "epoch": 0.34541577825159914,
+      "grad_norm": 0.8703976870763576,
+      "learning_rate": 9.993180180337126e-06,
+      "loss": 0.6813,
+      "step": 108
+    },
+    {
+      "epoch": 0.3486140724946695,
+      "grad_norm": 0.9178054118198938,
+      "learning_rate": 9.992171388780814e-06,
+      "loss": 0.712,
+      "step": 109
+    },
+    {
+      "epoch": 0.35181236673773986,
+      "grad_norm": 0.8589547671449834,
+      "learning_rate": 9.991093100466482e-06,
+      "loss": 0.7499,
+      "step": 110
+    },
+    {
+      "epoch": 0.35501066098081024,
+      "grad_norm": 0.8726951716749137,
+      "learning_rate": 9.989945330405146e-06,
+      "loss": 0.6837,
+      "step": 111
+    },
+    {
+      "epoch": 0.3582089552238806,
+      "grad_norm": 0.927713037073185,
+      "learning_rate": 9.988728094575082e-06,
+      "loss": 0.7342,
+      "step": 112
+    },
+    {
+      "epoch": 0.36140724946695096,
+      "grad_norm": 0.7505991388636275,
+      "learning_rate": 9.98744140992161e-06,
+      "loss": 0.7094,
+      "step": 113
+    },
+    {
+      "epoch": 0.3646055437100213,
+      "grad_norm": 0.894268442242807,
+      "learning_rate": 9.986085294356858e-06,
+      "loss": 0.6777,
+      "step": 114
+    },
+    {
+      "epoch": 0.3678038379530917,
+      "grad_norm": 0.8024161706862449,
+      "learning_rate": 9.98465976675951e-06,
+      "loss": 0.7261,
+      "step": 115
+    },
+    {
+      "epoch": 0.37100213219616207,
+      "grad_norm": 0.9360632386011705,
+      "learning_rate": 9.983164846974549e-06,
+      "loss": 0.7111,
+      "step": 116
+    },
+    {
+      "epoch": 0.3742004264392324,
+      "grad_norm": 0.850822368769409,
+      "learning_rate": 9.981600555812975e-06,
+      "loss": 0.6618,
+      "step": 117
+    },
+    {
+      "epoch": 0.3773987206823028,
+      "grad_norm": 0.8087687979800354,
+      "learning_rate": 9.979966915051517e-06,
+      "loss": 0.7151,
+      "step": 118
+    },
+    {
+      "epoch": 0.3805970149253731,
+      "grad_norm": 1.019238061277482,
+      "learning_rate": 9.978263947432331e-06,
+      "loss": 0.7419,
+      "step": 119
+    },
+    {
+      "epoch": 0.3837953091684435,
+      "grad_norm": 0.6679781593773177,
+      "learning_rate": 9.976491676662679e-06,
+      "loss": 0.6984,
+      "step": 120
+    },
+    {
+      "epoch": 0.38699360341151384,
+      "grad_norm": 0.8025666983281575,
+      "learning_rate": 9.974650127414609e-06,
+      "loss": 0.7041,
+      "step": 121
+    },
+    {
+      "epoch": 0.39019189765458423,
+      "grad_norm": 0.6935416167824833,
+      "learning_rate": 9.972739325324596e-06,
+      "loss": 0.6972,
+      "step": 122
+    },
+    {
+      "epoch": 0.39339019189765456,
+      "grad_norm": 0.8012590306354953,
+      "learning_rate": 9.970759296993205e-06,
+      "loss": 0.6719,
+      "step": 123
+    },
+    {
+      "epoch": 0.39658848614072495,
+      "grad_norm": 0.7351333900689639,
+      "learning_rate": 9.968710069984699e-06,
+      "loss": 0.6897,
+      "step": 124
+    },
+    {
+      "epoch": 0.3997867803837953,
+      "grad_norm": 0.8515682416300615,
+      "learning_rate": 9.966591672826674e-06,
+      "loss": 0.6911,
+      "step": 125
+    },
+    {
+      "epoch": 0.40298507462686567,
+      "grad_norm": 0.7745318085363246,
+      "learning_rate": 9.964404135009649e-06,
+      "loss": 0.6768,
+      "step": 126
+    },
+    {
+      "epoch": 0.40618336886993606,
+      "grad_norm": 0.7565019041685611,
+      "learning_rate": 9.962147486986664e-06,
+      "loss": 0.7018,
+      "step": 127
+    },
+    {
+      "epoch": 0.4093816631130064,
+      "grad_norm": 0.675609674290225,
+      "learning_rate": 9.959821760172849e-06,
+      "loss": 0.6978,
+      "step": 128
+    },
+    {
+      "epoch": 0.4125799573560768,
+      "grad_norm": 0.8186758848215644,
+      "learning_rate": 9.957426986944994e-06,
+      "loss": 0.6931,
+      "step": 129
+    },
+    {
+      "epoch": 0.4157782515991471,
+      "grad_norm": 0.7207234783292092,
+      "learning_rate": 9.95496320064109e-06,
+      "loss": 0.7082,
+      "step": 130
+    },
+    {
+      "epoch": 0.4189765458422175,
+      "grad_norm": 0.7354378059056297,
+      "learning_rate": 9.952430435559873e-06,
+      "loss": 0.7363,
+      "step": 131
+    },
+    {
+      "epoch": 0.42217484008528783,
+      "grad_norm": 0.8258670771003666,
+      "learning_rate": 9.94982872696034e-06,
+      "loss": 0.71,
+      "step": 132
+    },
+    {
+      "epoch": 0.4253731343283582,
+      "grad_norm": 0.7436932431094343,
+      "learning_rate": 9.947158111061263e-06,
+      "loss": 0.7276,
+      "step": 133
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 0.6981052362583808,
+      "learning_rate": 9.94441862504068e-06,
+      "loss": 0.6774,
+      "step": 134
+    },
+    {
+      "epoch": 0.43176972281449894,
+      "grad_norm": 0.8840543705470293,
+      "learning_rate": 9.941610307035385e-06,
+      "loss": 0.7257,
+      "step": 135
+    },
+    {
+      "epoch": 0.4349680170575693,
+      "grad_norm": 0.6694638600717385,
+      "learning_rate": 9.938733196140386e-06,
+      "loss": 0.7029,
+      "step": 136
+    },
+    {
+      "epoch": 0.43816631130063965,
+      "grad_norm": 0.8441420481093084,
+      "learning_rate": 9.935787332408375e-06,
+      "loss": 0.7274,
+      "step": 137
+    },
+    {
+      "epoch": 0.44136460554371004,
+      "grad_norm": 0.8505776286992377,
+      "learning_rate": 9.932772756849152e-06,
+      "loss": 0.6935,
+      "step": 138
+    },
+    {
+      "epoch": 0.4445628997867804,
+      "grad_norm": 0.9226301253646606,
+      "learning_rate": 9.929689511429075e-06,
+      "loss": 0.7264,
+      "step": 139
+    },
+    {
+      "epoch": 0.44776119402985076,
+      "grad_norm": 0.7579619597097683,
+      "learning_rate": 9.926537639070457e-06,
+      "loss": 0.7076,
+      "step": 140
+    },
+    {
+      "epoch": 0.4509594882729211,
+      "grad_norm": 0.8076488456889905,
+      "learning_rate": 9.923317183650985e-06,
+      "loss": 0.7003,
+      "step": 141
+    },
+    {
+      "epoch": 0.4541577825159915,
+      "grad_norm": 0.8485727454100752,
+      "learning_rate": 9.92002819000309e-06,
+      "loss": 0.7211,
+      "step": 142
+    },
+    {
+      "epoch": 0.4573560767590618,
+      "grad_norm": 0.751313097685783,
+      "learning_rate": 9.916670703913345e-06,
+      "loss": 0.6859,
+      "step": 143
+    },
+    {
+      "epoch": 0.4605543710021322,
+      "grad_norm": 0.7996244442440978,
+      "learning_rate": 9.913244772121811e-06,
+      "loss": 0.7048,
+      "step": 144
+    },
+    {
+      "epoch": 0.46375266524520253,
+      "grad_norm": 0.7586742602393676,
+      "learning_rate": 9.90975044232139e-06,
+      "loss": 0.7232,
+      "step": 145
+    },
+    {
+      "epoch": 0.4669509594882729,
+      "grad_norm": 0.7333771413773464,
+      "learning_rate": 9.90618776315717e-06,
+      "loss": 0.6973,
+      "step": 146
+    },
+    {
+      "epoch": 0.4701492537313433,
+      "grad_norm": 0.7462538908947492,
+      "learning_rate": 9.902556784225729e-06,
+      "loss": 0.6956,
+      "step": 147
+    },
+    {
+      "epoch": 0.47334754797441364,
+      "grad_norm": 0.7953075557133956,
+      "learning_rate": 9.898857556074469e-06,
+      "loss": 0.7225,
+      "step": 148
+    },
+    {
+      "epoch": 0.47654584221748403,
+      "grad_norm": 0.7060064737270609,
+      "learning_rate": 9.895090130200889e-06,
+      "loss": 0.6814,
+      "step": 149
+    },
+    {
+      "epoch": 0.47974413646055436,
+      "grad_norm": 0.8120142107026623,
+      "learning_rate": 9.891254559051886e-06,
+      "loss": 0.6731,
+      "step": 150
+    },
+    {
+      "epoch": 0.48294243070362475,
+      "grad_norm": 0.9453211383140191,
+      "learning_rate": 9.887350896023015e-06,
+      "loss": 0.6725,
+      "step": 151
+    },
+    {
+      "epoch": 0.4861407249466951,
+      "grad_norm": 0.753978978180176,
+      "learning_rate": 9.883379195457747e-06,
+      "loss": 0.6931,
+      "step": 152
+    },
+    {
+      "epoch": 0.48933901918976547,
+      "grad_norm": 0.9761958786101086,
+      "learning_rate": 9.879339512646714e-06,
+      "loss": 0.6964,
+      "step": 153
+    },
+    {
+      "epoch": 0.4925373134328358,
+      "grad_norm": 0.8092682567275233,
+      "learning_rate": 9.875231903826936e-06,
+      "loss": 0.7033,
+      "step": 154
+    },
+    {
+      "epoch": 0.4957356076759062,
+      "grad_norm": 0.8755474532977587,
+      "learning_rate": 9.871056426181052e-06,
+      "loss": 0.697,
+      "step": 155
+    },
+    {
+      "epoch": 0.4989339019189765,
+      "grad_norm": 0.7311549338301705,
+      "learning_rate": 9.8668131378365e-06,
+      "loss": 0.6999,
+      "step": 156
+    },
+    {
+      "epoch": 0.502132196162047,
+      "grad_norm": 0.7037792105922891,
+      "learning_rate": 9.862502097864726e-06,
+      "loss": 0.6792,
+      "step": 157
+    },
+    {
+      "epoch": 0.5053304904051172,
+      "grad_norm": 0.7713310350403229,
+      "learning_rate": 9.858123366280358e-06,
+      "loss": 0.7028,
+      "step": 158
+    },
+    {
+      "epoch": 0.5085287846481876,
+      "grad_norm": 0.7608743370139368,
+      "learning_rate": 9.853677004040368e-06,
+      "loss": 0.7262,
+      "step": 159
+    },
+    {
+      "epoch": 0.511727078891258,
+      "grad_norm": 0.7523293064381485,
+      "learning_rate": 9.849163073043223e-06,
+      "loss": 0.6968,
+      "step": 160
+    },
+    {
+      "epoch": 0.5149253731343284,
+      "grad_norm": 0.7701188802305937,
+      "learning_rate": 9.844581636128025e-06,
+      "loss": 0.6878,
+      "step": 161
+    },
+    {
+      "epoch": 0.5181236673773987,
+      "grad_norm": 0.6850118378660099,
+      "learning_rate": 9.83993275707364e-06,
+      "loss": 0.6843,
+      "step": 162
+    },
+    {
+      "epoch": 0.5213219616204691,
+      "grad_norm": 0.7633903463897005,
+      "learning_rate": 9.835216500597797e-06,
+      "loss": 0.6625,
+      "step": 163
+    },
+    {
+      "epoch": 0.5245202558635395,
+      "grad_norm": 0.7032490531444039,
+      "learning_rate": 9.830432932356207e-06,
+      "loss": 0.6787,
+      "step": 164
+    },
+    {
+      "epoch": 0.5277185501066098,
+      "grad_norm": 0.8377452657391034,
+      "learning_rate": 9.82558211894163e-06,
+      "loss": 0.6485,
+      "step": 165
+    },
+    {
+      "epoch": 0.5309168443496801,
+      "grad_norm": 0.6618917554232225,
+      "learning_rate": 9.820664127882958e-06,
+      "loss": 0.6789,
+      "step": 166
+    },
+    {
+      "epoch": 0.5341151385927505,
+      "grad_norm": 0.8178971073023995,
+      "learning_rate": 9.815679027644273e-06,
+      "loss": 0.7136,
+      "step": 167
+    },
+    {
+      "epoch": 0.5373134328358209,
+      "grad_norm": 0.7226926940415138,
+      "learning_rate": 9.8106268876239e-06,
+      "loss": 0.6973,
+      "step": 168
+    },
+    {
+      "epoch": 0.5405117270788913,
+      "grad_norm": 0.8219860052512269,
+      "learning_rate": 9.805507778153423e-06,
+      "loss": 0.6833,
+      "step": 169
+    },
+    {
+      "epoch": 0.5437100213219617,
+      "grad_norm": 0.8090438198755149,
+      "learning_rate": 9.800321770496726e-06,
+      "loss": 0.6683,
+      "step": 170
+    },
+    {
+      "epoch": 0.5469083155650319,
+      "grad_norm": 0.8076219868171068,
+      "learning_rate": 9.79506893684899e-06,
+      "loss": 0.6742,
+      "step": 171
+    },
+    {
+      "epoch": 0.5501066098081023,
+      "grad_norm": 0.7650394887530159,
+      "learning_rate": 9.789749350335693e-06,
+      "loss": 0.6809,
+      "step": 172
+    },
+    {
+      "epoch": 0.5533049040511727,
+      "grad_norm": 0.6909569362578575,
+      "learning_rate": 9.784363085011587e-06,
+      "loss": 0.6987,
+      "step": 173
+    },
+    {
+      "epoch": 0.5565031982942431,
+      "grad_norm": 0.7544714952212014,
+      "learning_rate": 9.778910215859666e-06,
+      "loss": 0.7286,
+      "step": 174
+    },
+    {
+      "epoch": 0.5597014925373134,
+      "grad_norm": 0.6910806053162777,
+      "learning_rate": 9.773390818790136e-06,
+      "loss": 0.7004,
+      "step": 175
+    },
+    {
+      "epoch": 0.5628997867803838,
+      "grad_norm": 0.728635086997373,
+      "learning_rate": 9.767804970639338e-06,
+      "loss": 0.7107,
+      "step": 176
+    },
+    {
+      "epoch": 0.5660980810234542,
+      "grad_norm": 0.7164830915823774,
+      "learning_rate": 9.762152749168693e-06,
+      "loss": 0.6659,
+      "step": 177
+    },
+    {
+      "epoch": 0.5692963752665245,
+      "grad_norm": 0.7185222557170264,
+      "learning_rate": 9.756434233063616e-06,
+      "loss": 0.7204,
+      "step": 178
+    },
+    {
+      "epoch": 0.5724946695095949,
+      "grad_norm": 0.6376198071084583,
+      "learning_rate": 9.750649501932414e-06,
+      "loss": 0.6851,
+      "step": 179
+    },
+    {
+      "epoch": 0.5756929637526652,
+      "grad_norm": 0.7372459805811605,
+      "learning_rate": 9.744798636305189e-06,
+      "loss": 0.6655,
+      "step": 180
+    },
+    {
+      "epoch": 0.5788912579957356,
+      "grad_norm": 0.7192794699200221,
+      "learning_rate": 9.738881717632709e-06,
+      "loss": 0.6981,
+      "step": 181
+    },
+    {
+      "epoch": 0.582089552238806,
+      "grad_norm": 0.8372267929782506,
+      "learning_rate": 9.732898828285273e-06,
+      "loss": 0.6609,
+      "step": 182
+    },
+    {
+      "epoch": 0.5852878464818764,
+      "grad_norm": 0.6890860653831234,
+      "learning_rate": 9.726850051551575e-06,
+      "loss": 0.6849,
+      "step": 183
+    },
+    {
+      "epoch": 0.5884861407249466,
+      "grad_norm": 1.0979021153485466,
+      "learning_rate": 9.72073547163753e-06,
+      "loss": 0.7055,
+      "step": 184
+    },
+    {
+      "epoch": 0.591684434968017,
+      "grad_norm": 0.718530118293507,
+      "learning_rate": 9.714555173665112e-06,
+      "loss": 0.6793,
+      "step": 185
+    },
+    {
+      "epoch": 0.5948827292110874,
+      "grad_norm": 0.9568746124941592,
+      "learning_rate": 9.708309243671167e-06,
+      "loss": 0.6805,
+      "step": 186
+    },
+    {
+      "epoch": 0.5980810234541578,
+      "grad_norm": 0.9190645493855029,
+      "learning_rate": 9.701997768606209e-06,
+      "loss": 0.6578,
+      "step": 187
+    },
+    {
+      "epoch": 0.6012793176972282,
+      "grad_norm": 0.8452449705256939,
+      "learning_rate": 9.695620836333219e-06,
+      "loss": 0.6752,
+      "step": 188
+    },
+    {
+      "epoch": 0.6044776119402985,
+      "grad_norm": 0.912469389514112,
+      "learning_rate": 9.68917853562642e-06,
+      "loss": 0.677,
+      "step": 189
+    },
+    {
+      "epoch": 0.6076759061833689,
+      "grad_norm": 0.8380041206744561,
+      "learning_rate": 9.68267095617003e-06,
+      "loss": 0.6985,
+      "step": 190
+    },
+    {
+      "epoch": 0.6108742004264393,
+      "grad_norm": 0.863868290234638,
+      "learning_rate": 9.676098188557032e-06,
+      "loss": 0.7313,
+      "step": 191
+    },
+    {
+      "epoch": 0.6140724946695096,
+      "grad_norm": 0.8174466263379166,
+      "learning_rate": 9.669460324287899e-06,
+      "loss": 0.7053,
+      "step": 192
+    },
+    {
+      "epoch": 0.6172707889125799,
+      "grad_norm": 0.8909312106674119,
+      "learning_rate": 9.662757455769317e-06,
+      "loss": 0.682,
+      "step": 193
+    },
+    {
+      "epoch": 0.6204690831556503,
+      "grad_norm": 0.7302375395625422,
+      "learning_rate": 9.655989676312918e-06,
+      "loss": 0.6594,
+      "step": 194
+    },
+    {
+      "epoch": 0.6236673773987207,
+      "grad_norm": 0.8735527453330153,
+      "learning_rate": 9.649157080133962e-06,
+      "loss": 0.6674,
+      "step": 195
+    },
+    {
+      "epoch": 0.6268656716417911,
+      "grad_norm": 0.7064746459902383,
+      "learning_rate": 9.642259762350034e-06,
+      "loss": 0.7017,
+      "step": 196
+    },
+    {
+      "epoch": 0.6300639658848614,
+      "grad_norm": 0.8655024689230618,
+      "learning_rate": 9.635297818979715e-06,
+      "loss": 0.6661,
+      "step": 197
+    },
+    {
+      "epoch": 0.6332622601279317,
+      "grad_norm": 0.728341775629865,
+      "learning_rate": 9.628271346941252e-06,
+      "loss": 0.6925,
+      "step": 198
+    },
+    {
+      "epoch": 0.6364605543710021,
+      "grad_norm": 0.6968228679265215,
+      "learning_rate": 9.621180444051206e-06,
+      "loss": 0.685,
+      "step": 199
+    },
+    {
+      "epoch": 0.6396588486140725,
+      "grad_norm": 0.8921095235064438,
+      "learning_rate": 9.614025209023084e-06,
+      "loss": 0.7016,
+      "step": 200
+    },
+    {
+      "epoch": 0.6428571428571429,
+      "grad_norm": 0.7770028881285537,
+      "learning_rate": 9.606805741465977e-06,
+      "loss": 0.7077,
+      "step": 201
+    },
+    {
+      "epoch": 0.6460554371002132,
+      "grad_norm": 0.7270295337970796,
+      "learning_rate": 9.59952214188316e-06,
+      "loss": 0.6779,
+      "step": 202
+    },
+    {
+      "epoch": 0.6492537313432836,
+      "grad_norm": 0.8302887272339062,
+      "learning_rate": 9.592174511670704e-06,
+      "loss": 0.6722,
+      "step": 203
+    },
+    {
+      "epoch": 0.652452025586354,
+      "grad_norm": 0.8688719467779429,
+      "learning_rate": 9.58476295311606e-06,
+      "loss": 0.7104,
+      "step": 204
+    },
+    {
+      "epoch": 0.6556503198294243,
+      "grad_norm": 0.6323690135784534,
+      "learning_rate": 9.577287569396632e-06,
+      "loss": 0.6927,
+      "step": 205
+    },
+    {
+      "epoch": 0.6588486140724946,
+      "grad_norm": 0.8941706514509578,
+      "learning_rate": 9.569748464578343e-06,
+      "loss": 0.6926,
+      "step": 206
+    },
+    {
+      "epoch": 0.662046908315565,
+      "grad_norm": 0.7211897394095068,
+      "learning_rate": 9.562145743614193e-06,
+      "loss": 0.7006,
+      "step": 207
+    },
+    {
+      "epoch": 0.6652452025586354,
+      "grad_norm": 0.655783741036758,
+      "learning_rate": 9.554479512342785e-06,
+      "loss": 0.7108,
+      "step": 208
+    },
+    {
+      "epoch": 0.6684434968017058,
+      "grad_norm": 0.7467548640108442,
+      "learning_rate": 9.54674987748686e-06,
+      "loss": 0.6649,
+      "step": 209
+    },
+    {
+      "epoch": 0.6716417910447762,
+      "grad_norm": 0.7171171346221873,
+      "learning_rate": 9.538956946651816e-06,
+      "loss": 0.6531,
+      "step": 210
+    },
+    {
+      "epoch": 0.6748400852878464,
+      "grad_norm": 0.7120324790582493,
+      "learning_rate": 9.531100828324191e-06,
+      "loss": 0.6795,
+      "step": 211
+    },
+    {
+      "epoch": 0.6780383795309168,
+      "grad_norm": 0.675292328763824,
+      "learning_rate": 9.52318163187018e-06,
+      "loss": 0.6883,
+      "step": 212
+    },
+    {
+      "epoch": 0.6812366737739872,
+      "grad_norm": 0.9706242893642346,
+      "learning_rate": 9.515199467534086e-06,
+      "loss": 0.7105,
+      "step": 213
+    },
+    {
+      "epoch": 0.6844349680170576,
+      "grad_norm": 0.6339606603859417,
+      "learning_rate": 9.507154446436806e-06,
+      "loss": 0.7006,
+      "step": 214
+    },
+    {
+      "epoch": 0.6876332622601279,
+      "grad_norm": 0.8270660302761437,
+      "learning_rate": 9.499046680574267e-06,
+      "loss": 0.6937,
+      "step": 215
+    },
+    {
+      "epoch": 0.6908315565031983,
+      "grad_norm": 0.8339411710564606,
+      "learning_rate": 9.490876282815884e-06,
+      "loss": 0.7089,
+      "step": 216
+    },
+    {
+      "epoch": 0.6940298507462687,
+      "grad_norm": 0.7095006207653165,
+      "learning_rate": 9.482643366902972e-06,
+      "loss": 0.6774,
+      "step": 217
+    },
+    {
+      "epoch": 0.697228144989339,
+      "grad_norm": 0.8701992961050132,
+      "learning_rate": 9.474348047447177e-06,
+      "loss": 0.6987,
+      "step": 218
+    },
+    {
+      "epoch": 0.7004264392324094,
+      "grad_norm": 0.8555157892062962,
+      "learning_rate": 9.465990439928868e-06,
+      "loss": 0.6888,
+      "step": 219
+    },
+    {
+      "epoch": 0.7036247334754797,
+      "grad_norm": 0.7482188301136834,
+      "learning_rate": 9.457570660695542e-06,
+      "loss": 0.6924,
+      "step": 220
+    },
+    {
+      "epoch": 0.7068230277185501,
+      "grad_norm": 0.6679769126751529,
+      "learning_rate": 9.449088826960187e-06,
+      "loss": 0.6814,
+      "step": 221
+    },
+    {
+      "epoch": 0.7100213219616205,
+      "grad_norm": 0.6481228296820869,
+      "learning_rate": 9.440545056799677e-06,
+      "loss": 0.6927,
+      "step": 222
+    },
+    {
+      "epoch": 0.7132196162046909,
+      "grad_norm": 0.6138493570700504,
+      "learning_rate": 9.431939469153096e-06,
+      "loss": 0.6848,
+      "step": 223
+    },
+    {
+      "epoch": 0.7164179104477612,
+      "grad_norm": 0.6857392440178816,
+      "learning_rate": 9.423272183820109e-06,
+      "loss": 0.7016,
+      "step": 224
+    },
+    {
+      "epoch": 0.7196162046908315,
+      "grad_norm": 0.6565822633668216,
+      "learning_rate": 9.41454332145928e-06,
+      "loss": 0.684,
+      "step": 225
+    },
+    {
+      "epoch": 0.7228144989339019,
+      "grad_norm": 0.7686352189035123,
+      "learning_rate": 9.405753003586396e-06,
+      "loss": 0.662,
+      "step": 226
+    },
+    {
+      "epoch": 0.7260127931769723,
+      "grad_norm": 0.6889440928778451,
+      "learning_rate": 9.396901352572771e-06,
+      "loss": 0.6757,
+      "step": 227
+    },
+    {
+      "epoch": 0.7292110874200426,
+      "grad_norm": 0.6644883797889718,
+      "learning_rate": 9.387988491643558e-06,
+      "loss": 0.69,
+      "step": 228
+    },
+    {
+      "epoch": 0.732409381663113,
+      "grad_norm": 0.7626373390965248,
+      "learning_rate": 9.379014544876011e-06,
+      "loss": 0.7187,
+      "step": 229
+    },
+    {
+      "epoch": 0.7356076759061834,
+      "grad_norm": 0.704507663933334,
+      "learning_rate": 9.369979637197774e-06,
+      "loss": 0.6585,
+      "step": 230
+    },
+    {
+      "epoch": 0.7388059701492538,
+      "grad_norm": 0.7336157000029004,
+      "learning_rate": 9.360883894385137e-06,
+      "loss": 0.7127,
+      "step": 231
+    },
+    {
+      "epoch": 0.7420042643923241,
+      "grad_norm": 0.6546636167664962,
+      "learning_rate": 9.351727443061284e-06,
+      "loss": 0.704,
+      "step": 232
+    },
+    {
+      "epoch": 0.7452025586353944,
+      "grad_norm": 0.7043313515356551,
+      "learning_rate": 9.342510410694529e-06,
+      "loss": 0.6852,
+      "step": 233
+    },
+    {
+      "epoch": 0.7484008528784648,
+      "grad_norm": 0.9011726584289667,
+      "learning_rate": 9.33323292559655e-06,
+      "loss": 0.7081,
+      "step": 234
+    },
+    {
+      "epoch": 0.7515991471215352,
+      "grad_norm": 0.7292470054254178,
+      "learning_rate": 9.323895116920591e-06,
+      "loss": 0.669,
+      "step": 235
+    },
+    {
+      "epoch": 0.7547974413646056,
+      "grad_norm": 0.8382635859763398,
+      "learning_rate": 9.31449711465967e-06,
+      "loss": 0.6763,
+      "step": 236
+    },
+    {
+      "epoch": 0.7579957356076759,
+      "grad_norm": 0.7483080444786336,
+      "learning_rate": 9.305039049644772e-06,
+      "loss": 0.6567,
+      "step": 237
+    },
+    {
+      "epoch": 0.7611940298507462,
+      "grad_norm": 0.7261935584991153,
+      "learning_rate": 9.29552105354302e-06,
+      "loss": 0.6714,
+      "step": 238
+    },
+    {
+      "epoch": 0.7643923240938166,
+      "grad_norm": 0.8645856822885789,
+      "learning_rate": 9.28594325885585e-06,
+      "loss": 0.6625,
+      "step": 239
+    },
+    {
+      "epoch": 0.767590618336887,
+      "grad_norm": 0.7568040003927768,
+      "learning_rate": 9.27630579891716e-06,
+      "loss": 0.6191,
+      "step": 240
+    },
+    {
+      "epoch": 0.7707889125799574,
+      "grad_norm": 0.7959983248684005,
+      "learning_rate": 9.266608807891459e-06,
+      "loss": 0.72,
+      "step": 241
+    },
+    {
+      "epoch": 0.7739872068230277,
+      "grad_norm": 0.6158464477516697,
+      "learning_rate": 9.256852420771999e-06,
+      "loss": 0.6984,
+      "step": 242
+    },
+    {
+      "epoch": 0.7771855010660981,
+      "grad_norm": 0.6475971949577826,
+      "learning_rate": 9.24703677337889e-06,
+      "loss": 0.6707,
+      "step": 243
+    },
+    {
+      "epoch": 0.7803837953091685,
+      "grad_norm": 0.7514772530157263,
+      "learning_rate": 9.237162002357214e-06,
+      "loss": 0.7025,
+      "step": 244
+    },
+    {
+      "epoch": 0.7835820895522388,
+      "grad_norm": 0.7153934193506174,
+      "learning_rate": 9.227228245175127e-06,
+      "loss": 0.7084,
+      "step": 245
+    },
+    {
+      "epoch": 0.7867803837953091,
+      "grad_norm": 0.6510532625278841,
+      "learning_rate": 9.217235640121927e-06,
+      "loss": 0.7032,
+      "step": 246
+    },
+    {
+      "epoch": 0.7899786780383795,
+      "grad_norm": 0.648491391649832,
+      "learning_rate": 9.207184326306155e-06,
+      "loss": 0.703,
+      "step": 247
+    },
+    {
+      "epoch": 0.7931769722814499,
+      "grad_norm": 0.8034299471010478,
+      "learning_rate": 9.197074443653643e-06,
+      "loss": 0.7118,
+      "step": 248
+    },
+    {
+      "epoch": 0.7963752665245203,
+      "grad_norm": 0.7868890537454554,
+      "learning_rate": 9.186906132905563e-06,
+      "loss": 0.6907,
+      "step": 249
+    },
+    {
+      "epoch": 0.7995735607675906,
+      "grad_norm": 0.7573924459698659,
+      "learning_rate": 9.176679535616477e-06,
+      "loss": 0.7133,
+      "step": 250
+    },
+    {
+      "epoch": 0.802771855010661,
+      "grad_norm": 0.7039169561600006,
+      "learning_rate": 9.166394794152363e-06,
+      "loss": 0.6749,
+      "step": 251
+    },
+    {
+      "epoch": 0.8059701492537313,
+      "grad_norm": 0.8423685923463412,
+      "learning_rate": 9.156052051688633e-06,
+      "loss": 0.7378,
+      "step": 252
+    },
+    {
+      "epoch": 0.8091684434968017,
+      "grad_norm": 0.6131291481922342,
+      "learning_rate": 9.145651452208133e-06,
+      "loss": 0.6415,
+      "step": 253
+    },
+    {
+      "epoch": 0.8123667377398721,
+      "grad_norm": 0.7195983825244481,
+      "learning_rate": 9.135193140499155e-06,
+      "loss": 0.6613,
+      "step": 254
+    },
+    {
+      "epoch": 0.8155650319829424,
+      "grad_norm": 0.8577511387252152,
+      "learning_rate": 9.124677262153405e-06,
+      "loss": 0.721,
+      "step": 255
+    },
+    {
+      "epoch": 0.8187633262260128,
+      "grad_norm": 0.7278841089447043,
+      "learning_rate": 9.114103963563986e-06,
+      "loss": 0.6738,
+      "step": 256
+    },
+    {
+      "epoch": 0.8219616204690832,
+      "grad_norm": 0.6856569850641855,
+      "learning_rate": 9.103473391923354e-06,
+      "loss": 0.6588,
+      "step": 257
+    },
+    {
+      "epoch": 0.8251599147121536,
+      "grad_norm": 0.9021119379041687,
+      "learning_rate": 9.092785695221271e-06,
+      "loss": 0.7182,
+      "step": 258
+    },
+    {
+      "epoch": 0.8283582089552238,
+      "grad_norm": 0.8796683162834094,
+      "learning_rate": 9.08204102224275e-06,
+      "loss": 0.6816,
+      "step": 259
+    },
+    {
+      "epoch": 0.8315565031982942,
+      "grad_norm": 0.847520407573671,
+      "learning_rate": 9.071239522565978e-06,
+      "loss": 0.6508,
+      "step": 260
+    },
+    {
+      "epoch": 0.8347547974413646,
+      "grad_norm": 0.9049944252517715,
+      "learning_rate": 9.06038134656023e-06,
+      "loss": 0.6925,
+      "step": 261
+    },
+    {
+      "epoch": 0.837953091684435,
+      "grad_norm": 0.8571499479329611,
+      "learning_rate": 9.049466645383785e-06,
+      "loss": 0.681,
+      "step": 262
+    },
+    {
+      "epoch": 0.8411513859275054,
+      "grad_norm": 0.745522815593341,
+      "learning_rate": 9.038495570981814e-06,
+      "loss": 0.6525,
+      "step": 263
+    },
+    {
+      "epoch": 0.8443496801705757,
+      "grad_norm": 0.947292350163867,
+      "learning_rate": 9.027468276084274e-06,
+      "loss": 0.6696,
+      "step": 264
+    },
+    {
+      "epoch": 0.847547974413646,
+      "grad_norm": 0.8373012552761833,
+      "learning_rate": 9.016384914203771e-06,
+      "loss": 0.6782,
+      "step": 265
+    },
+    {
+      "epoch": 0.8507462686567164,
+      "grad_norm": 0.7211699745049627,
+      "learning_rate": 9.00524563963343e-06,
+      "loss": 0.7015,
+      "step": 266
+    },
+    {
+      "epoch": 0.8539445628997868,
+      "grad_norm": 0.973744577195585,
+      "learning_rate": 8.99405060744474e-06,
+      "loss": 0.6598,
+      "step": 267
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.7600474078060623,
+      "learning_rate": 8.982799973485407e-06,
+      "loss": 0.6853,
+      "step": 268
+    },
+    {
+      "epoch": 0.8603411513859275,
+      "grad_norm": 0.7568774174674711,
+      "learning_rate": 8.971493894377174e-06,
+      "loss": 0.6963,
+      "step": 269
+    },
+    {
+      "epoch": 0.8635394456289979,
+      "grad_norm": 1.2096518035065966,
+      "learning_rate": 8.960132527513642e-06,
+      "loss": 0.6901,
+      "step": 270
+    },
+    {
+      "epoch": 0.8667377398720683,
+      "grad_norm": 0.8943559219904526,
+      "learning_rate": 8.94871603105809e-06,
+      "loss": 0.6782,
+      "step": 271
+    },
+    {
+      "epoch": 0.8699360341151386,
+      "grad_norm": 0.7461346678625066,
+      "learning_rate": 8.937244563941248e-06,
+      "loss": 0.7018,
+      "step": 272
+    },
+    {
+      "epoch": 0.8731343283582089,
+      "grad_norm": 1.0000505503295154,
+      "learning_rate": 8.925718285859118e-06,
+      "loss": 0.6875,
+      "step": 273
+    },
+    {
+      "epoch": 0.8763326226012793,
+      "grad_norm": 0.7858176236082158,
+      "learning_rate": 8.914137357270723e-06,
+      "loss": 0.7121,
+      "step": 274
+    },
+    {
+      "epoch": 0.8795309168443497,
+      "grad_norm": 0.7063872884648656,
+      "learning_rate": 8.902501939395887e-06,
+      "loss": 0.6695,
+      "step": 275
+    },
+    {
+      "epoch": 0.8827292110874201,
+      "grad_norm": 0.8044416558628715,
+      "learning_rate": 8.890812194212987e-06,
+      "loss": 0.7076,
+      "step": 276
+    },
+    {
+      "epoch": 0.8859275053304904,
+      "grad_norm": 0.6393740328761567,
+      "learning_rate": 8.879068284456702e-06,
+      "loss": 0.6556,
+      "step": 277
+    },
+    {
+      "epoch": 0.8891257995735607,
+      "grad_norm": 0.857231208160294,
+      "learning_rate": 8.867270373615735e-06,
+      "loss": 0.7148,
+      "step": 278
+    },
+    {
+      "epoch": 0.8923240938166311,
+      "grad_norm": 0.6337872608316819,
+      "learning_rate": 8.855418625930556e-06,
+      "loss": 0.7365,
+      "step": 279
+    },
+    {
+      "epoch": 0.8955223880597015,
+      "grad_norm": 0.6706185344174747,
+      "learning_rate": 8.8435132063911e-06,
+      "loss": 0.6699,
+      "step": 280
+    },
+    {
+      "epoch": 0.8987206823027718,
+      "grad_norm": 0.7688016478702769,
+      "learning_rate": 8.83155428073448e-06,
+      "loss": 0.6695,
+      "step": 281
+    },
+    {
+      "epoch": 0.9019189765458422,
+      "grad_norm": 0.6859750413913875,
+      "learning_rate": 8.81954201544267e-06,
+      "loss": 0.7105,
+      "step": 282
+    },
+    {
+      "epoch": 0.9051172707889126,
+      "grad_norm": 0.736049871487562,
+      "learning_rate": 8.8074765777402e-06,
+      "loss": 0.6566,
+      "step": 283
+    },
+    {
+      "epoch": 0.908315565031983,
+      "grad_norm": 0.6566455673708244,
+      "learning_rate": 8.79535813559181e-06,
+      "loss": 0.7087,
+      "step": 284
+    },
+    {
+      "epoch": 0.9115138592750534,
+      "grad_norm": 0.6678755774132027,
+      "learning_rate": 8.783186857700137e-06,
+      "loss": 0.6803,
+      "step": 285
+    },
+    {
+      "epoch": 0.9147121535181236,
+      "grad_norm": 0.6358862703989032,
+      "learning_rate": 8.77096291350334e-06,
+      "loss": 0.6477,
+      "step": 286
+    },
+    {
+      "epoch": 0.917910447761194,
+      "grad_norm": 0.621135300076204,
+      "learning_rate": 8.75868647317276e-06,
+      "loss": 0.6853,
+      "step": 287
+    },
+    {
+      "epoch": 0.9211087420042644,
+      "grad_norm": 0.8269166477046754,
+      "learning_rate": 8.746357707610544e-06,
+      "loss": 0.7068,
+      "step": 288
+    },
+    {
+      "epoch": 0.9243070362473348,
+      "grad_norm": 0.754354903003167,
+      "learning_rate": 8.733976788447265e-06,
+      "loss": 0.6775,
+      "step": 289
+    },
+    {
+      "epoch": 0.9275053304904051,
+      "grad_norm": 0.6328527198398567,
+      "learning_rate": 8.721543888039534e-06,
+      "loss": 0.6738,
+      "step": 290
+    },
+    {
+      "epoch": 0.9307036247334755,
+      "grad_norm": 0.7230763976135067,
+      "learning_rate": 8.709059179467598e-06,
+      "loss": 0.6448,
+      "step": 291
+    },
+    {
+      "epoch": 0.9339019189765458,
+      "grad_norm": 0.6877836548216805,
+      "learning_rate": 8.69652283653294e-06,
+      "loss": 0.6798,
+      "step": 292
+    },
+    {
+      "epoch": 0.9371002132196162,
+      "grad_norm": 0.7564591211606494,
+      "learning_rate": 8.683935033755848e-06,
+      "loss": 0.6953,
+      "step": 293
+    },
+    {
+      "epoch": 0.9402985074626866,
+      "grad_norm": 0.668332745907526,
+      "learning_rate": 8.671295946372989e-06,
+      "loss": 0.681,
+      "step": 294
+    },
+    {
+      "epoch": 0.9434968017057569,
+      "grad_norm": 0.7159527856011911,
+      "learning_rate": 8.658605750334972e-06,
+      "loss": 0.6988,
+      "step": 295
+    },
+    {
+      "epoch": 0.9466950959488273,
+      "grad_norm": 0.6848922808640394,
+      "learning_rate": 8.6458646223039e-06,
+      "loss": 0.641,
+      "step": 296
+    },
+    {
+      "epoch": 0.9498933901918977,
+      "grad_norm": 0.7548334675392807,
+      "learning_rate": 8.6330727396509e-06,
+      "loss": 0.6713,
+      "step": 297
+    },
+    {
+      "epoch": 0.9530916844349681,
+      "grad_norm": 0.5861356337358591,
+      "learning_rate": 8.620230280453672e-06,
+      "loss": 0.691,
+      "step": 298
+    },
+    {
+      "epoch": 0.9562899786780383,
+      "grad_norm": 0.7949755015478441,
+      "learning_rate": 8.607337423493996e-06,
+      "loss": 0.6808,
+      "step": 299
+    },
+    {
+      "epoch": 0.9594882729211087,
+      "grad_norm": 0.6765008542247006,
+      "learning_rate": 8.594394348255239e-06,
+      "loss": 0.6898,
+      "step": 300
+    },
+    {
+      "epoch": 0.9626865671641791,
+      "grad_norm": 0.5806572889436659,
+      "learning_rate": 8.581401234919873e-06,
+      "loss": 0.6492,
+      "step": 301
+    },
+    {
+      "epoch": 0.9658848614072495,
+      "grad_norm": 0.7469524089876209,
+      "learning_rate": 8.568358264366958e-06,
+      "loss": 0.6821,
+      "step": 302
+    },
+    {
+      "epoch": 0.9690831556503199,
+      "grad_norm": 0.7249833696216659,
+      "learning_rate": 8.555265618169615e-06,
+      "loss": 0.6522,
+      "step": 303
+    },
+    {
+      "epoch": 0.9722814498933902,
+      "grad_norm": 0.6218765386847785,
+      "learning_rate": 8.542123478592518e-06,
+      "loss": 0.6843,
+      "step": 304
+    },
+    {
+      "epoch": 0.9754797441364605,
+      "grad_norm": 0.7551061996851178,
+      "learning_rate": 8.528932028589337e-06,
+      "loss": 0.6728,
+      "step": 305
+    },
+    {
+      "epoch": 0.9786780383795309,
+      "grad_norm": 0.674338314023206,
+      "learning_rate": 8.515691451800206e-06,
+      "loss": 0.6736,
+      "step": 306
+    },
+    {
+      "epoch": 0.9818763326226013,
+      "grad_norm": 0.7849066941451676,
+      "learning_rate": 8.502401932549154e-06,
+      "loss": 0.6878,
+      "step": 307
+    },
+    {
+      "epoch": 0.9850746268656716,
+      "grad_norm": 0.7498368721683543,
+      "learning_rate": 8.489063655841552e-06,
+      "loss": 0.6903,
+      "step": 308
+    },
+    {
+      "epoch": 0.988272921108742,
+      "grad_norm": 0.665126766182557,
+      "learning_rate": 8.475676807361526e-06,
+      "loss": 0.6552,
+      "step": 309
+    },
+    {
+      "epoch": 0.9914712153518124,
+      "grad_norm": 0.637731666313635,
+      "learning_rate": 8.462241573469378e-06,
+      "loss": 0.6769,
+      "step": 310
+    },
+    {
+      "epoch": 0.9946695095948828,
+      "grad_norm": 0.6112924731775418,
+      "learning_rate": 8.448758141198991e-06,
+      "loss": 0.677,
+      "step": 311
+    },
+    {
+      "epoch": 0.997867803837953,
+      "grad_norm": 0.7383946127305492,
+      "learning_rate": 8.435226698255228e-06,
+      "loss": 0.6746,
+      "step": 312
+    },
+    {
+      "epoch": 1.0010660980810235,
+      "grad_norm": 0.626462069677627,
+      "learning_rate": 8.421647433011306e-06,
+      "loss": 0.6566,
+      "step": 313
+    },
+    {
+      "epoch": 1.004264392324094,
+      "grad_norm": 0.7510688924872979,
+      "learning_rate": 8.408020534506195e-06,
+      "loss": 0.6281,
+      "step": 314
+    },
+    {
+      "epoch": 1.007462686567164,
+      "grad_norm": 0.7595637013734888,
+      "learning_rate": 8.394346192441967e-06,
+      "loss": 0.6484,
+      "step": 315
+    },
+    {
+      "epoch": 1.0106609808102345,
+      "grad_norm": 0.8191267012599663,
+      "learning_rate": 8.380624597181165e-06,
+      "loss": 0.6269,
+      "step": 316
+    },
+    {
+      "epoch": 1.0138592750533049,
+      "grad_norm": 0.6920756752948478,
+      "learning_rate": 8.366855939744152e-06,
+      "loss": 0.6407,
+      "step": 317
+    },
+    {
+      "epoch": 1.0170575692963753,
+      "grad_norm": 0.8113121478207607,
+      "learning_rate": 8.353040411806449e-06,
+      "loss": 0.6379,
+      "step": 318
+    },
+    {
+      "epoch": 1.0202558635394456,
+      "grad_norm": 0.6038342001520437,
+      "learning_rate": 8.339178205696067e-06,
+      "loss": 0.6219,
+      "step": 319
+    },
+    {
+      "epoch": 1.023454157782516,
+      "grad_norm": 0.6525041753408835,
+      "learning_rate": 8.325269514390835e-06,
+      "loss": 0.6451,
+      "step": 320
+    },
+    {
+      "epoch": 1.0266524520255864,
+      "grad_norm": 0.7876363756773992,
+      "learning_rate": 8.311314531515707e-06,
+      "loss": 0.6556,
+      "step": 321
+    },
+    {
+      "epoch": 1.0298507462686568,
+      "grad_norm": 0.6729418158706656,
+      "learning_rate": 8.297313451340064e-06,
+      "loss": 0.6278,
+      "step": 322
+    },
+    {
+      "epoch": 1.033049040511727,
+      "grad_norm": 0.7323640919082973,
+      "learning_rate": 8.283266468775024e-06,
+      "loss": 0.635,
+      "step": 323
+    },
+    {
+      "epoch": 1.0362473347547974,
+      "grad_norm": 0.7456586009899282,
+      "learning_rate": 8.269173779370712e-06,
+      "loss": 0.6301,
+      "step": 324
+    },
+    {
+      "epoch": 1.0394456289978677,
+      "grad_norm": 0.6472864535960438,
+      "learning_rate": 8.255035579313545e-06,
+      "loss": 0.6514,
+      "step": 325
+    },
+    {
+      "epoch": 1.0426439232409381,
+      "grad_norm": 0.7261880506431115,
+      "learning_rate": 8.240852065423507e-06,
+      "loss": 0.6255,
+      "step": 326
+    },
+    {
+      "epoch": 1.0458422174840085,
+      "grad_norm": 0.7277147979365199,
+      "learning_rate": 8.226623435151389e-06,
+      "loss": 0.5958,
+      "step": 327
+    },
+    {
+      "epoch": 1.049040511727079,
+      "grad_norm": 0.6624168018368254,
+      "learning_rate": 8.21234988657607e-06,
+      "loss": 0.6179,
+      "step": 328
+    },
+    {
+      "epoch": 1.0522388059701493,
+      "grad_norm": 0.7453935828883017,
+      "learning_rate": 8.198031618401733e-06,
+      "loss": 0.6108,
+      "step": 329
+    },
+    {
+      "epoch": 1.0554371002132197,
+      "grad_norm": 0.7264796554563445,
+      "learning_rate": 8.183668829955111e-06,
+      "loss": 0.6238,
+      "step": 330
+    },
+    {
+      "epoch": 1.05863539445629,
+      "grad_norm": 0.7682551654401545,
+      "learning_rate": 8.169261721182715e-06,
+      "loss": 0.6191,
+      "step": 331
+    },
+    {
+      "epoch": 1.0618336886993602,
+      "grad_norm": 0.6228417217082994,
+      "learning_rate": 8.154810492648038e-06,
+      "loss": 0.6316,
+      "step": 332
+    },
+    {
+      "epoch": 1.0650319829424306,
+      "grad_norm": 0.7327288572841154,
+      "learning_rate": 8.140315345528778e-06,
+      "loss": 0.6586,
+      "step": 333
+    },
+    {
+      "epoch": 1.068230277185501,
+      "grad_norm": 0.6964883436081012,
+      "learning_rate": 8.125776481614025e-06,
+      "loss": 0.6414,
+      "step": 334
+    },
+    {
+      "epoch": 1.0714285714285714,
+      "grad_norm": 0.6054633221728152,
+      "learning_rate": 8.111194103301461e-06,
+      "loss": 0.5893,
+      "step": 335
+    },
+    {
+      "epoch": 1.0746268656716418,
+      "grad_norm": 0.6959078284441366,
+      "learning_rate": 8.096568413594533e-06,
+      "loss": 0.6209,
+      "step": 336
+    },
+    {
+      "epoch": 1.0778251599147122,
+      "grad_norm": 0.6687442847144212,
+      "learning_rate": 8.081899616099638e-06,
+      "loss": 0.6395,
+      "step": 337
+    },
+    {
+      "epoch": 1.0810234541577826,
+      "grad_norm": 0.6672132910162282,
+      "learning_rate": 8.067187915023283e-06,
+      "loss": 0.6453,
+      "step": 338
+    },
+    {
+      "epoch": 1.084221748400853,
+      "grad_norm": 0.7011871375952131,
+      "learning_rate": 8.052433515169235e-06,
+      "loss": 0.6578,
+      "step": 339
+    },
+    {
+      "epoch": 1.0874200426439233,
+      "grad_norm": 0.6341715440450284,
+      "learning_rate": 8.037636621935686e-06,
+      "loss": 0.5993,
+      "step": 340
+    },
+    {
+      "epoch": 1.0906183368869935,
+      "grad_norm": 0.7942056782846189,
+      "learning_rate": 8.022797441312376e-06,
+      "loss": 0.6377,
+      "step": 341
+    },
+    {
+      "epoch": 1.0938166311300639,
+      "grad_norm": 0.6703609675609049,
+      "learning_rate": 8.007916179877742e-06,
+      "loss": 0.6262,
+      "step": 342
+    },
+    {
+      "epoch": 1.0970149253731343,
+      "grad_norm": 0.6839776213487608,
+      "learning_rate": 7.99299304479603e-06,
+      "loss": 0.6244,
+      "step": 343
+    },
+    {
+      "epoch": 1.1002132196162047,
+      "grad_norm": 0.7336708760979732,
+      "learning_rate": 7.978028243814416e-06,
+      "loss": 0.6607,
+      "step": 344
+    },
+    {
+      "epoch": 1.103411513859275,
+      "grad_norm": 0.6632430703502449,
+      "learning_rate": 7.96302198526011e-06,
+      "loss": 0.6556,
+      "step": 345
+    },
+    {
+      "epoch": 1.1066098081023454,
+      "grad_norm": 0.695919842154835,
+      "learning_rate": 7.947974478037468e-06,
+      "loss": 0.6194,
+      "step": 346
+    },
+    {
+      "epoch": 1.1098081023454158,
+      "grad_norm": 0.654985383307004,
+      "learning_rate": 7.932885931625063e-06,
+      "loss": 0.6015,
+      "step": 347
+    },
+    {
+      "epoch": 1.1130063965884862,
+      "grad_norm": 0.7284161828955869,
+      "learning_rate": 7.917756556072792e-06,
+      "loss": 0.6238,
+      "step": 348
+    },
+    {
+      "epoch": 1.1162046908315566,
+      "grad_norm": 0.7633948849076417,
+      "learning_rate": 7.902586561998928e-06,
+      "loss": 0.6829,
+      "step": 349
+    },
+    {
+      "epoch": 1.1194029850746268,
+      "grad_norm": 0.7808526782166596,
+      "learning_rate": 7.887376160587214e-06,
+      "loss": 0.6286,
+      "step": 350
+    },
+    {
+      "epoch": 1.1226012793176972,
+      "grad_norm": 0.7168210695937199,
+      "learning_rate": 7.8721255635839e-06,
+      "loss": 0.655,
+      "step": 351
+    },
+    {
+      "epoch": 1.1257995735607675,
+      "grad_norm": 0.7093615169947094,
+      "learning_rate": 7.85683498329481e-06,
+      "loss": 0.6082,
+      "step": 352
+    },
+    {
+      "epoch": 1.128997867803838,
+      "grad_norm": 0.7319031414944822,
+      "learning_rate": 7.841504632582378e-06,
+      "loss": 0.6046,
+      "step": 353
+    },
+    {
+      "epoch": 1.1321961620469083,
+      "grad_norm": 0.7551618483349399,
+      "learning_rate": 7.826134724862687e-06,
+      "loss": 0.6229,
+      "step": 354
+    },
+    {
+      "epoch": 1.1353944562899787,
+      "grad_norm": 0.5826096940443887,
+      "learning_rate": 7.810725474102504e-06,
+      "loss": 0.6491,
+      "step": 355
+    },
+    {
+      "epoch": 1.138592750533049,
+      "grad_norm": 0.6417225168262856,
+      "learning_rate": 7.795277094816292e-06,
+      "loss": 0.6261,
+      "step": 356
+    },
+    {
+      "epoch": 1.1417910447761195,
+      "grad_norm": 0.6906660028796798,
+      "learning_rate": 7.779789802063229e-06,
+      "loss": 0.6716,
+      "step": 357
+    },
+    {
+      "epoch": 1.1449893390191899,
+      "grad_norm": 0.6599388631575054,
+      "learning_rate": 7.764263811444214e-06,
+      "loss": 0.6223,
+      "step": 358
+    },
+    {
+      "epoch": 1.14818763326226,
+      "grad_norm": 0.6878046102310988,
+      "learning_rate": 7.748699339098864e-06,
+      "loss": 0.6535,
+      "step": 359
+    },
+    {
+      "epoch": 1.1513859275053304,
+      "grad_norm": 0.6983975410121682,
+      "learning_rate": 7.733096601702508e-06,
+      "loss": 0.6175,
+      "step": 360
+    },
+    {
+      "epoch": 1.1545842217484008,
+      "grad_norm": 0.7101392030170859,
+      "learning_rate": 7.717455816463161e-06,
+      "loss": 0.6064,
+      "step": 361
+    },
+    {
+      "epoch": 1.1577825159914712,
+      "grad_norm": 0.7138468401070005,
+      "learning_rate": 7.70177720111852e-06,
+      "loss": 0.6318,
+      "step": 362
+    },
+    {
+      "epoch": 1.1609808102345416,
+      "grad_norm": 0.8020962425971306,
+      "learning_rate": 7.68606097393291e-06,
+      "loss": 0.5551,
+      "step": 363
+    },
+    {
+      "epoch": 1.164179104477612,
+      "grad_norm": 0.5838755573325491,
+      "learning_rate": 7.67030735369426e-06,
+      "loss": 0.6681,
+      "step": 364
+    },
+    {
+      "epoch": 1.1673773987206824,
+      "grad_norm": 0.7126521508489613,
+      "learning_rate": 7.654516559711053e-06,
+      "loss": 0.6146,
+      "step": 365
+    },
+    {
+      "epoch": 1.1705756929637527,
+      "grad_norm": 0.786856054891589,
+      "learning_rate": 7.638688811809274e-06,
+      "loss": 0.592,
+      "step": 366
+    },
+    {
+      "epoch": 1.1737739872068231,
+      "grad_norm": 0.6150312500511127,
+      "learning_rate": 7.622824330329345e-06,
+      "loss": 0.6454,
+      "step": 367
+    },
+    {
+      "epoch": 1.1769722814498933,
+      "grad_norm": 0.6533779011208625,
+      "learning_rate": 7.6069233361230696e-06,
+      "loss": 0.5972,
+      "step": 368
+    },
+    {
+      "epoch": 1.1801705756929637,
+      "grad_norm": 0.7040587378616768,
+      "learning_rate": 7.590986050550542e-06,
+      "loss": 0.6506,
+      "step": 369
+    },
+    {
+      "epoch": 1.183368869936034,
+      "grad_norm": 0.6498966757491511,
+      "learning_rate": 7.575012695477076e-06,
+      "loss": 0.6686,
+      "step": 370
+    },
+    {
+      "epoch": 1.1865671641791045,
+      "grad_norm": 0.686323024280706,
+      "learning_rate": 7.55900349327012e-06,
+      "loss": 0.6156,
+      "step": 371
+    },
+    {
+      "epoch": 1.1897654584221748,
+      "grad_norm": 0.6423904326022212,
+      "learning_rate": 7.542958666796149e-06,
+      "loss": 0.6074,
+      "step": 372
+    },
+    {
+      "epoch": 1.1929637526652452,
+      "grad_norm": 0.6178701938146715,
+      "learning_rate": 7.526878439417572e-06,
+      "loss": 0.6224,
+      "step": 373
+    },
+    {
+      "epoch": 1.1961620469083156,
+      "grad_norm": 0.6496299199143347,
+      "learning_rate": 7.510763034989616e-06,
+      "loss": 0.6122,
+      "step": 374
+    },
+    {
+      "epoch": 1.199360341151386,
+      "grad_norm": 0.6906674571020524,
+      "learning_rate": 7.494612677857218e-06,
+      "loss": 0.5982,
+      "step": 375
+    },
+    {
+      "epoch": 1.2025586353944564,
+      "grad_norm": 0.7648845383563382,
+      "learning_rate": 7.478427592851894e-06,
+      "loss": 0.6388,
+      "step": 376
+    },
+    {
+      "epoch": 1.2057569296375266,
+      "grad_norm": 0.696212830007392,
+      "learning_rate": 7.462208005288609e-06,
+      "loss": 0.6195,
+      "step": 377
+    },
+    {
+      "epoch": 1.208955223880597,
+      "grad_norm": 0.8576480726323676,
+      "learning_rate": 7.44595414096265e-06,
+      "loss": 0.62,
+      "step": 378
+    },
+    {
+      "epoch": 1.2121535181236673,
+      "grad_norm": 0.6976210286882664,
+      "learning_rate": 7.429666226146468e-06,
+      "loss": 0.6416,
+      "step": 379
+    },
+    {
+      "epoch": 1.2153518123667377,
+      "grad_norm": 0.720435239108808,
+      "learning_rate": 7.413344487586542e-06,
+      "loss": 0.663,
+      "step": 380
+    },
+    {
+      "epoch": 1.2185501066098081,
+      "grad_norm": 0.6855025363671802,
+      "learning_rate": 7.396989152500215e-06,
+      "loss": 0.6244,
+      "step": 381
+    },
+    {
+      "epoch": 1.2217484008528785,
+      "grad_norm": 0.7874191554629412,
+      "learning_rate": 7.380600448572532e-06,
+      "loss": 0.669,
+      "step": 382
+    },
+    {
+      "epoch": 1.224946695095949,
+      "grad_norm": 0.7434853121491252,
+      "learning_rate": 7.364178603953066e-06,
+      "loss": 0.6448,
+      "step": 383
+    },
+    {
+      "epoch": 1.2281449893390193,
+      "grad_norm": 0.61487183758893,
+      "learning_rate": 7.347723847252756e-06,
+      "loss": 0.6406,
+      "step": 384
+    },
+    {
+      "epoch": 1.2313432835820897,
+      "grad_norm": 0.6917610834136999,
+      "learning_rate": 7.331236407540704e-06,
+      "loss": 0.6311,
+      "step": 385
+    },
+    {
+      "epoch": 1.2345415778251598,
+      "grad_norm": 0.7293794438462118,
+      "learning_rate": 7.314716514341007e-06,
+      "loss": 0.6496,
+      "step": 386
+    },
+    {
+      "epoch": 1.2377398720682302,
+      "grad_norm": 0.6943306944663377,
+      "learning_rate": 7.298164397629545e-06,
+      "loss": 0.695,
+      "step": 387
+    },
+    {
+      "epoch": 1.2409381663113006,
+      "grad_norm": 0.727606577778182,
+      "learning_rate": 7.28158028783079e-06,
+      "loss": 0.6472,
+      "step": 388
+    },
+    {
+      "epoch": 1.244136460554371,
+      "grad_norm": 0.6612780887089154,
+      "learning_rate": 7.2649644158145925e-06,
+      "loss": 0.6581,
+      "step": 389
+    },
+    {
+      "epoch": 1.2473347547974414,
+      "grad_norm": 0.6796926109344666,
+      "learning_rate": 7.248317012892969e-06,
+      "loss": 0.6455,
+      "step": 390
+    },
+    {
+      "epoch": 1.2505330490405118,
+      "grad_norm": 0.7347393619920833,
+      "learning_rate": 7.231638310816888e-06,
+      "loss": 0.6135,
+      "step": 391
+    },
+    {
+      "epoch": 1.2537313432835822,
+      "grad_norm": 0.6507765447015836,
+      "learning_rate": 7.214928541773027e-06,
+      "loss": 0.6528,
+      "step": 392
+    },
+    {
+      "epoch": 1.2569296375266523,
+      "grad_norm": 0.6378971746133411,
+      "learning_rate": 7.198187938380565e-06,
+      "loss": 0.64,
+      "step": 393
+    },
+    {
+      "epoch": 1.260127931769723,
+      "grad_norm": 0.6973995690775606,
+      "learning_rate": 7.1814167336879195e-06,
+      "loss": 0.6372,
+      "step": 394
+    },
+    {
+      "epoch": 1.263326226012793,
+      "grad_norm": 0.6611835574299032,
+      "learning_rate": 7.164615161169518e-06,
+      "loss": 0.641,
+      "step": 395
+    },
+    {
+      "epoch": 1.2665245202558635,
+      "grad_norm": 0.5638998333261266,
+      "learning_rate": 7.147783454722545e-06,
+      "loss": 0.6191,
+      "step": 396
+    },
+    {
+      "epoch": 1.2697228144989339,
+      "grad_norm": 0.5840735647196549,
+      "learning_rate": 7.130921848663678e-06,
+      "loss": 0.6461,
+      "step": 397
+    },
+    {
+      "epoch": 1.2729211087420043,
+      "grad_norm": 0.6067150183156588,
+      "learning_rate": 7.1140305777258355e-06,
+      "loss": 0.6534,
+      "step": 398
+    },
+    {
+      "epoch": 1.2761194029850746,
+      "grad_norm": 0.5533289520902701,
+      "learning_rate": 7.097109877054906e-06,
+      "loss": 0.6251,
+      "step": 399
+    },
+    {
+      "epoch": 1.279317697228145,
+      "grad_norm": 0.6286453359485137,
+      "learning_rate": 7.080159982206471e-06,
+      "loss": 0.6192,
+      "step": 400
+    },
+    {
+      "epoch": 1.2825159914712154,
+      "grad_norm": 0.5518718217422934,
+      "learning_rate": 7.06318112914253e-06,
+      "loss": 0.6014,
+      "step": 401
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 0.5849867238933993,
+      "learning_rate": 7.046173554228213e-06,
+      "loss": 0.6131,
+      "step": 402
+    },
+    {
+      "epoch": 1.2889125799573562,
+      "grad_norm": 0.6401106522610778,
+      "learning_rate": 7.029137494228491e-06,
+      "loss": 0.6638,
+      "step": 403
+    },
+    {
+      "epoch": 1.2921108742004264,
+      "grad_norm": 0.571567700082923,
+      "learning_rate": 7.012073186304885e-06,
+      "loss": 0.6347,
+      "step": 404
+    },
+    {
+      "epoch": 1.2953091684434968,
+      "grad_norm": 0.5618388361367395,
+      "learning_rate": 6.994980868012151e-06,
+      "loss": 0.6374,
+      "step": 405
+    },
+    {
+      "epoch": 1.2985074626865671,
+      "grad_norm": 0.5861018343702368,
+      "learning_rate": 6.9778607772949894e-06,
+      "loss": 0.6427,
+      "step": 406
+    },
+    {
+      "epoch": 1.3017057569296375,
+      "grad_norm": 0.5732316822425729,
+      "learning_rate": 6.9607131524847175e-06,
+      "loss": 0.6153,
+      "step": 407
+    },
+    {
+      "epoch": 1.304904051172708,
+      "grad_norm": 0.6579936163217828,
+      "learning_rate": 6.943538232295965e-06,
+      "loss": 0.6822,
+      "step": 408
+    },
+    {
+      "epoch": 1.3081023454157783,
+      "grad_norm": 0.6429931495365935,
+      "learning_rate": 6.926336255823341e-06,
+      "loss": 0.6703,
+      "step": 409
+    },
+    {
+      "epoch": 1.3113006396588487,
+      "grad_norm": 0.6970899348524371,
+      "learning_rate": 6.909107462538113e-06,
+      "loss": 0.6118,
+      "step": 410
+    },
+    {
+      "epoch": 1.3144989339019189,
+      "grad_norm": 0.6680718435346977,
+      "learning_rate": 6.891852092284863e-06,
+      "loss": 0.6485,
+      "step": 411
+    },
+    {
+      "epoch": 1.3176972281449895,
+      "grad_norm": 0.6600380168580939,
+      "learning_rate": 6.874570385278161e-06,
+      "loss": 0.6717,
+      "step": 412
+    },
+    {
+      "epoch": 1.3208955223880596,
+      "grad_norm": 0.802744507326931,
+      "learning_rate": 6.857262582099209e-06,
+      "loss": 0.6225,
+      "step": 413
+    },
+    {
+      "epoch": 1.32409381663113,
+      "grad_norm": 0.5662862687056852,
+      "learning_rate": 6.839928923692505e-06,
+      "loss": 0.6327,
+      "step": 414
+    },
+    {
+      "epoch": 1.3272921108742004,
+      "grad_norm": 0.7774960950807415,
+      "learning_rate": 6.822569651362475e-06,
+      "loss": 0.629,
+      "step": 415
+    },
+    {
+      "epoch": 1.3304904051172708,
+      "grad_norm": 0.7049572238409981,
+      "learning_rate": 6.805185006770125e-06,
+      "loss": 0.6693,
+      "step": 416
+    },
+    {
+      "epoch": 1.3336886993603412,
+      "grad_norm": 0.583155396015153,
+      "learning_rate": 6.787775231929666e-06,
+      "loss": 0.6237,
+      "step": 417
+    },
+    {
+      "epoch": 1.3368869936034116,
+      "grad_norm": 0.7111519870233683,
+      "learning_rate": 6.7703405692051585e-06,
+      "loss": 0.6394,
+      "step": 418
+    },
+    {
+      "epoch": 1.340085287846482,
+      "grad_norm": 0.5653594640288497,
+      "learning_rate": 6.752881261307125e-06,
+      "loss": 0.6155,
+      "step": 419
+    },
+    {
+      "epoch": 1.3432835820895521,
+      "grad_norm": 0.5953314550960257,
+      "learning_rate": 6.735397551289179e-06,
+      "loss": 0.6768,
+      "step": 420
+    },
+    {
+      "epoch": 1.3464818763326227,
+      "grad_norm": 0.6173871359335188,
+      "learning_rate": 6.717889682544641e-06,
+      "loss": 0.6524,
+      "step": 421
+    },
+    {
+      "epoch": 1.349680170575693,
+      "grad_norm": 0.5904374139677531,
+      "learning_rate": 6.700357898803146e-06,
+      "loss": 0.6171,
+      "step": 422
+    },
+    {
+      "epoch": 1.3528784648187633,
+      "grad_norm": 0.6013270780619727,
+      "learning_rate": 6.6828024441272554e-06,
+      "loss": 0.6331,
+      "step": 423
+    },
+    {
+      "epoch": 1.3560767590618337,
+      "grad_norm": 0.7234400477846492,
+      "learning_rate": 6.665223562909058e-06,
+      "loss": 0.6048,
+      "step": 424
+    },
+    {
+      "epoch": 1.359275053304904,
+      "grad_norm": 0.5513590207931245,
+      "learning_rate": 6.647621499866762e-06,
+      "loss": 0.6315,
+      "step": 425
+    },
+    {
+      "epoch": 1.3624733475479744,
+      "grad_norm": 0.6666248996294633,
+      "learning_rate": 6.629996500041299e-06,
+      "loss": 0.6611,
+      "step": 426
+    },
+    {
+      "epoch": 1.3656716417910448,
+      "grad_norm": 0.7473392889269375,
+      "learning_rate": 6.612348808792904e-06,
+      "loss": 0.6509,
+      "step": 427
+    },
+    {
+      "epoch": 1.3688699360341152,
+      "grad_norm": 0.6178536139476686,
+      "learning_rate": 6.5946786717977026e-06,
+      "loss": 0.5846,
+      "step": 428
+    },
+    {
+      "epoch": 1.3720682302771854,
+      "grad_norm": 0.7286227764621949,
+      "learning_rate": 6.576986335044292e-06,
+      "loss": 0.6225,
+      "step": 429
+    },
+    {
+      "epoch": 1.375266524520256,
+      "grad_norm": 0.6891105977017067,
+      "learning_rate": 6.5592720448303174e-06,
+      "loss": 0.6247,
+      "step": 430
+    },
+    {
+      "epoch": 1.3784648187633262,
+      "grad_norm": 0.548934245482671,
+      "learning_rate": 6.541536047759034e-06,
+      "loss": 0.6446,
+      "step": 431
+    },
+    {
+      "epoch": 1.3816631130063965,
+      "grad_norm": 0.6602568691180308,
+      "learning_rate": 6.523778590735892e-06,
+      "loss": 0.6262,
+      "step": 432
+    },
+    {
+      "epoch": 1.384861407249467,
+      "grad_norm": 0.6490608935621192,
+      "learning_rate": 6.5059999209650795e-06,
+      "loss": 0.6272,
+      "step": 433
+    },
+    {
+      "epoch": 1.3880597014925373,
+      "grad_norm": 0.6052733094339158,
+      "learning_rate": 6.488200285946094e-06,
+      "loss": 0.6625,
+      "step": 434
+    },
+    {
+      "epoch": 1.3912579957356077,
+      "grad_norm": 0.7850931175623906,
+      "learning_rate": 6.470379933470296e-06,
+      "loss": 0.6289,
+      "step": 435
+    },
+    {
+      "epoch": 1.394456289978678,
+      "grad_norm": 0.6381779692800353,
+      "learning_rate": 6.452539111617454e-06,
+      "loss": 0.6403,
+      "step": 436
+    },
+    {
+      "epoch": 1.3976545842217485,
+      "grad_norm": 0.6118616893209208,
+      "learning_rate": 6.434678068752293e-06,
+      "loss": 0.6407,
+      "step": 437
+    },
+    {
+      "epoch": 1.4008528784648187,
+      "grad_norm": 0.6834821963612697,
+      "learning_rate": 6.416797053521039e-06,
+      "loss": 0.6274,
+      "step": 438
+    },
+    {
+      "epoch": 1.4040511727078893,
+      "grad_norm": 0.6399343872362031,
+      "learning_rate": 6.398896314847954e-06,
+      "loss": 0.5884,
+      "step": 439
+    },
+    {
+      "epoch": 1.4072494669509594,
+      "grad_norm": 0.650012682249965,
+      "learning_rate": 6.380976101931879e-06,
+      "loss": 0.5957,
+      "step": 440
+    },
+    {
+      "epoch": 1.4104477611940298,
+      "grad_norm": 0.7005173466154051,
+      "learning_rate": 6.363036664242751e-06,
+      "loss": 0.6632,
+      "step": 441
+    },
+    {
+      "epoch": 1.4136460554371002,
+      "grad_norm": 0.6793406210914823,
+      "learning_rate": 6.345078251518144e-06,
+      "loss": 0.6597,
+      "step": 442
+    },
+    {
+      "epoch": 1.4168443496801706,
+      "grad_norm": 0.6296952630265025,
+      "learning_rate": 6.327101113759783e-06,
+      "loss": 0.6719,
+      "step": 443
+    },
+    {
+      "epoch": 1.420042643923241,
+      "grad_norm": 0.6793002098632206,
+      "learning_rate": 6.3091055012300675e-06,
+      "loss": 0.6683,
+      "step": 444
+    },
+    {
+      "epoch": 1.4232409381663114,
+      "grad_norm": 0.632818749841171,
+      "learning_rate": 6.291091664448589e-06,
+      "loss": 0.636,
+      "step": 445
+    },
+    {
+      "epoch": 1.4264392324093818,
+      "grad_norm": 0.6812898757622275,
+      "learning_rate": 6.273059854188636e-06,
+      "loss": 0.6437,
+      "step": 446
+    },
+    {
+      "epoch": 1.429637526652452,
+      "grad_norm": 0.6587981529111426,
+      "learning_rate": 6.25501032147372e-06,
+      "loss": 0.683,
+      "step": 447
+    },
+    {
+      "epoch": 1.4328358208955223,
+      "grad_norm": 0.6112506718049283,
+      "learning_rate": 6.236943317574054e-06,
+      "loss": 0.6291,
+      "step": 448
+    },
+    {
+      "epoch": 1.4360341151385927,
+      "grad_norm": 0.6816015509553757,
+      "learning_rate": 6.218859094003082e-06,
+      "loss": 0.6389,
+      "step": 449
+    },
+    {
+      "epoch": 1.439232409381663,
+      "grad_norm": 0.5678562801569867,
+      "learning_rate": 6.200757902513962e-06,
+      "loss": 0.655,
+      "step": 450
+    },
+    {
+      "epoch": 1.4424307036247335,
+      "grad_norm": 0.5811458853598366,
+      "learning_rate": 6.182639995096061e-06,
+      "loss": 0.6762,
+      "step": 451
+    },
+    {
+      "epoch": 1.4456289978678039,
+      "grad_norm": 0.7324330200735513,
+      "learning_rate": 6.164505623971458e-06,
+      "loss": 0.6298,
+      "step": 452
+    },
+    {
+      "epoch": 1.4488272921108742,
+      "grad_norm": 0.5725759197205573,
+      "learning_rate": 6.146355041591419e-06,
+      "loss": 0.6573,
+      "step": 453
+    },
+    {
+      "epoch": 1.4520255863539446,
+      "grad_norm": 0.6093969811734463,
+      "learning_rate": 6.128188500632892e-06,
+      "loss": 0.6226,
+      "step": 454
+    },
+    {
+      "epoch": 1.455223880597015,
+      "grad_norm": 0.5642498697251256,
+      "learning_rate": 6.11000625399499e-06,
+      "loss": 0.6047,
+      "step": 455
+    },
+    {
+      "epoch": 1.4584221748400852,
+      "grad_norm": 0.5740303890176889,
+      "learning_rate": 6.091808554795462e-06,
+      "loss": 0.6318,
+      "step": 456
+    },
+    {
+      "epoch": 1.4616204690831556,
+      "grad_norm": 0.6288307609988142,
+      "learning_rate": 6.073595656367175e-06,
+      "loss": 0.647,
+      "step": 457
+    },
+    {
+      "epoch": 1.464818763326226,
+      "grad_norm": 0.5866496634459744,
+      "learning_rate": 6.055367812254592e-06,
+      "loss": 0.6098,
+      "step": 458
+    },
+    {
+      "epoch": 1.4680170575692963,
+      "grad_norm": 0.5931796642802533,
+      "learning_rate": 6.037125276210229e-06,
+      "loss": 0.6682,
+      "step": 459
+    },
+    {
+      "epoch": 1.4712153518123667,
+      "grad_norm": 0.5952005451682014,
+      "learning_rate": 6.0188683021911394e-06,
+      "loss": 0.6415,
+      "step": 460
+    },
+    {
+      "epoch": 1.4744136460554371,
+      "grad_norm": 0.5818037194226144,
+      "learning_rate": 6.000597144355361e-06,
+      "loss": 0.6301,
+      "step": 461
+    },
+    {
+      "epoch": 1.4776119402985075,
+      "grad_norm": 0.5536985639025898,
+      "learning_rate": 5.982312057058392e-06,
+      "loss": 0.6189,
+      "step": 462
+    },
+    {
+      "epoch": 1.480810234541578,
+      "grad_norm": 0.5574476870925947,
+      "learning_rate": 5.964013294849646e-06,
+      "loss": 0.6513,
+      "step": 463
+    },
+    {
+      "epoch": 1.4840085287846483,
+      "grad_norm": 0.6007921638296236,
+      "learning_rate": 5.9457011124689025e-06,
+      "loss": 0.6562,
+      "step": 464
+    },
+    {
+      "epoch": 1.4872068230277184,
+      "grad_norm": 0.6484774122884718,
+      "learning_rate": 5.927375764842766e-06,
+      "loss": 0.6415,
+      "step": 465
+    },
+    {
+      "epoch": 1.4904051172707888,
+      "grad_norm": 0.5367744850334513,
+      "learning_rate": 5.9090375070811215e-06,
+      "loss": 0.6324,
+      "step": 466
+    },
+    {
+      "epoch": 1.4936034115138592,
+      "grad_norm": 0.6330458661332963,
+      "learning_rate": 5.890686594473571e-06,
+      "loss": 0.6773,
+      "step": 467
+    },
+    {
+      "epoch": 1.4968017057569296,
+      "grad_norm": 0.6239236549424172,
+      "learning_rate": 5.872323282485889e-06,
+      "loss": 0.62,
+      "step": 468
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7714169311784487,
+      "learning_rate": 5.853947826756465e-06,
+      "loss": 0.635,
+      "step": 469
+    },
+    {
+      "epoch": 1.5031982942430704,
+      "grad_norm": 0.6052673944971001,
+      "learning_rate": 5.835560483092743e-06,
+      "loss": 0.6212,
+      "step": 470
+    },
+    {
+      "epoch": 1.5063965884861408,
+      "grad_norm": 0.6500878797817035,
+      "learning_rate": 5.8171615074676615e-06,
+      "loss": 0.6235,
+      "step": 471
+    },
+    {
+      "epoch": 1.509594882729211,
+      "grad_norm": 0.5806431215575556,
+      "learning_rate": 5.798751156016085e-06,
+      "loss": 0.618,
+      "step": 472
+    },
+    {
+      "epoch": 1.5127931769722816,
+      "grad_norm": 0.6174497537734422,
+      "learning_rate": 5.780329685031247e-06,
+      "loss": 0.6218,
+      "step": 473
+    },
+    {
+      "epoch": 1.5159914712153517,
+      "grad_norm": 0.6376322903314644,
+      "learning_rate": 5.7618973509611755e-06,
+      "loss": 0.6133,
+      "step": 474
+    },
+    {
+      "epoch": 1.5191897654584223,
+      "grad_norm": 0.5842117132744907,
+      "learning_rate": 5.743454410405126e-06,
+      "loss": 0.6135,
+      "step": 475
+    },
+    {
+      "epoch": 1.5223880597014925,
+      "grad_norm": 0.7067488456038941,
+      "learning_rate": 5.72500112011001e-06,
+      "loss": 0.6172,
+      "step": 476
+    },
+    {
+      "epoch": 1.5255863539445629,
+      "grad_norm": 0.563785993728458,
+      "learning_rate": 5.706537736966814e-06,
+      "loss": 0.598,
+      "step": 477
+    },
+    {
+      "epoch": 1.5287846481876333,
+      "grad_norm": 0.6261356329525944,
+      "learning_rate": 5.688064518007036e-06,
+      "loss": 0.6362,
+      "step": 478
+    },
+    {
+      "epoch": 1.5319829424307037,
+      "grad_norm": 0.7762542900583126,
+      "learning_rate": 5.669581720399094e-06,
+      "loss": 0.6108,
+      "step": 479
+    },
+    {
+      "epoch": 1.535181236673774,
+      "grad_norm": 0.5837610551252048,
+      "learning_rate": 5.651089601444752e-06,
+      "loss": 0.6516,
+      "step": 480
+    },
+    {
+      "epoch": 1.5383795309168442,
+      "grad_norm": 0.5716180642723976,
+      "learning_rate": 5.632588418575542e-06,
+      "loss": 0.6345,
+      "step": 481
+    },
+    {
+      "epoch": 1.5415778251599148,
+      "grad_norm": 0.5946826841288883,
+      "learning_rate": 5.614078429349172e-06,
+      "loss": 0.5934,
+      "step": 482
+    },
+    {
+      "epoch": 1.544776119402985,
+      "grad_norm": 0.6512015058357723,
+      "learning_rate": 5.5955598914459465e-06,
+      "loss": 0.6457,
+      "step": 483
+    },
+    {
+      "epoch": 1.5479744136460556,
+      "grad_norm": 0.5539302739317081,
+      "learning_rate": 5.577033062665179e-06,
+      "loss": 0.6234,
+      "step": 484
+    },
+    {
+      "epoch": 1.5511727078891258,
+      "grad_norm": 0.5788744632809979,
+      "learning_rate": 5.558498200921597e-06,
+      "loss": 0.5965,
+      "step": 485
+    },
+    {
+      "epoch": 1.5543710021321961,
+      "grad_norm": 0.5723934422902696,
+      "learning_rate": 5.53995556424176e-06,
+      "loss": 0.6148,
+      "step": 486
+    },
+    {
+      "epoch": 1.5575692963752665,
+      "grad_norm": 0.6217220728990693,
+      "learning_rate": 5.521405410760462e-06,
+      "loss": 0.6306,
+      "step": 487
+    },
+    {
+      "epoch": 1.560767590618337,
+      "grad_norm": 0.5798449514904306,
+      "learning_rate": 5.50284799871714e-06,
+      "loss": 0.6523,
+      "step": 488
+    },
+    {
+      "epoch": 1.5639658848614073,
+      "grad_norm": 0.5884980910713539,
+      "learning_rate": 5.484283586452279e-06,
+      "loss": 0.6409,
+      "step": 489
+    },
+    {
+      "epoch": 1.5671641791044775,
+      "grad_norm": 0.5451220864593269,
+      "learning_rate": 5.465712432403812e-06,
+      "loss": 0.64,
+      "step": 490
+    },
+    {
+      "epoch": 1.570362473347548,
+      "grad_norm": 0.6378194747314097,
+      "learning_rate": 5.447134795103531e-06,
+      "loss": 0.6369,
+      "step": 491
+    },
+    {
+      "epoch": 1.5735607675906182,
+      "grad_norm": 0.6422794734982141,
+      "learning_rate": 5.428550933173476e-06,
+      "loss": 0.6315,
+      "step": 492
+    },
+    {
+      "epoch": 1.5767590618336889,
+      "grad_norm": 0.6165632358558669,
+      "learning_rate": 5.409961105322347e-06,
+      "loss": 0.606,
+      "step": 493
+    },
+    {
+      "epoch": 1.579957356076759,
+      "grad_norm": 0.5655676781911688,
+      "learning_rate": 5.391365570341893e-06,
+      "loss": 0.6454,
+      "step": 494
+    },
+    {
+      "epoch": 1.5831556503198294,
+      "grad_norm": 0.5938708749925583,
+      "learning_rate": 5.372764587103309e-06,
+      "loss": 0.5788,
+      "step": 495
+    },
+    {
+      "epoch": 1.5863539445628998,
+      "grad_norm": 0.6088992599619968,
+      "learning_rate": 5.3541584145536475e-06,
+      "loss": 0.6072,
+      "step": 496
+    },
+    {
+      "epoch": 1.5895522388059702,
+      "grad_norm": 0.5784074732476567,
+      "learning_rate": 5.335547311712188e-06,
+      "loss": 0.6217,
+      "step": 497
+    },
+    {
+      "epoch": 1.5927505330490406,
+      "grad_norm": 0.6355703007033081,
+      "learning_rate": 5.3169315376668566e-06,
+      "loss": 0.6003,
+      "step": 498
+    },
+    {
+      "epoch": 1.5959488272921107,
+      "grad_norm": 0.5969121792808912,
+      "learning_rate": 5.2983113515706045e-06,
+      "loss": 0.6141,
+      "step": 499
+    },
+    {
+      "epoch": 1.5991471215351813,
+      "grad_norm": 0.5840423887981219,
+      "learning_rate": 5.279687012637798e-06,
+      "loss": 0.6454,
+      "step": 500
+    },
+    {
+      "epoch": 1.6023454157782515,
+      "grad_norm": 0.632640436994868,
+      "learning_rate": 5.2610587801406256e-06,
+      "loss": 0.619,
+      "step": 501
+    },
+    {
+      "epoch": 1.6055437100213221,
+      "grad_norm": 0.6215694598661943,
+      "learning_rate": 5.242426913405471e-06,
+      "loss": 0.612,
+      "step": 502
+    },
+    {
+      "epoch": 1.6087420042643923,
+      "grad_norm": 0.6102315942715076,
+      "learning_rate": 5.223791671809314e-06,
+      "loss": 0.6246,
+      "step": 503
+    },
+    {
+      "epoch": 1.6119402985074627,
+      "grad_norm": 0.6343359621017866,
+      "learning_rate": 5.2051533147761155e-06,
+      "loss": 0.6361,
+      "step": 504
+    },
+    {
+      "epoch": 1.615138592750533,
+      "grad_norm": 0.6647629539626027,
+      "learning_rate": 5.186512101773206e-06,
+      "loss": 0.6227,
+      "step": 505
+    },
+    {
+      "epoch": 1.6183368869936035,
+      "grad_norm": 0.6372800082536703,
+      "learning_rate": 5.167868292307679e-06,
+      "loss": 0.6478,
+      "step": 506
+    },
+    {
+      "epoch": 1.6215351812366738,
+      "grad_norm": 0.5768049939693273,
+      "learning_rate": 5.149222145922765e-06,
+      "loss": 0.6459,
+      "step": 507
+    },
+    {
+      "epoch": 1.624733475479744,
+      "grad_norm": 0.625505130556027,
+      "learning_rate": 5.130573922194236e-06,
+      "loss": 0.6587,
+      "step": 508
+    },
+    {
+      "epoch": 1.6279317697228146,
+      "grad_norm": 0.6164713574466578,
+      "learning_rate": 5.111923880726779e-06,
+      "loss": 0.631,
+      "step": 509
+    },
+    {
+      "epoch": 1.6311300639658848,
+      "grad_norm": 0.6064673645339017,
+      "learning_rate": 5.093272281150383e-06,
+      "loss": 0.6425,
+      "step": 510
+    },
+    {
+      "epoch": 1.6343283582089554,
+      "grad_norm": 0.6081052704243078,
+      "learning_rate": 5.074619383116733e-06,
+      "loss": 0.5795,
+      "step": 511
+    },
+    {
+      "epoch": 1.6375266524520256,
+      "grad_norm": 0.6412746235926382,
+      "learning_rate": 5.05596544629559e-06,
+      "loss": 0.6341,
+      "step": 512
+    },
+    {
+      "epoch": 1.640724946695096,
+      "grad_norm": 0.6297169827461854,
+      "learning_rate": 5.03731073037117e-06,
+      "loss": 0.6413,
+      "step": 513
+    },
+    {
+      "epoch": 1.6439232409381663,
+      "grad_norm": 0.5638619759271226,
+      "learning_rate": 5.018655495038542e-06,
+      "loss": 0.642,
+      "step": 514
+    },
+    {
+      "epoch": 1.6471215351812367,
+      "grad_norm": 0.556408547120617,
+      "learning_rate": 5e-06,
+      "loss": 0.6713,
+      "step": 515
+    },
+    {
+      "epoch": 1.650319829424307,
+      "grad_norm": 0.6515376681950392,
+      "learning_rate": 4.981344504961459e-06,
+      "loss": 0.6701,
+      "step": 516
+    },
+    {
+      "epoch": 1.6535181236673773,
+      "grad_norm": 0.5492041340566489,
+      "learning_rate": 4.962689269628832e-06,
+      "loss": 0.6173,
+      "step": 517
+    },
+    {
+      "epoch": 1.6567164179104479,
+      "grad_norm": 0.5686773096619224,
+      "learning_rate": 4.944034553704412e-06,
+      "loss": 0.6461,
+      "step": 518
+    },
+    {
+      "epoch": 1.659914712153518,
+      "grad_norm": 0.6435263865701101,
+      "learning_rate": 4.9253806168832685e-06,
+      "loss": 0.6123,
+      "step": 519
+    },
+    {
+      "epoch": 1.6631130063965884,
+      "grad_norm": 0.5691664452687136,
+      "learning_rate": 4.906727718849619e-06,
+      "loss": 0.6238,
+      "step": 520
+    },
+    {
+      "epoch": 1.6663113006396588,
+      "grad_norm": 0.567539010985681,
+      "learning_rate": 4.888076119273223e-06,
+      "loss": 0.6184,
+      "step": 521
+    },
+    {
+      "epoch": 1.6695095948827292,
+      "grad_norm": 0.6144632127897085,
+      "learning_rate": 4.8694260778057655e-06,
+      "loss": 0.5979,
+      "step": 522
+    },
+    {
+      "epoch": 1.6727078891257996,
+      "grad_norm": 0.5911520451028203,
+      "learning_rate": 4.850777854077235e-06,
+      "loss": 0.6469,
+      "step": 523
+    },
+    {
+      "epoch": 1.67590618336887,
+      "grad_norm": 0.5454031941868988,
+      "learning_rate": 4.832131707692322e-06,
+      "loss": 0.6122,
+      "step": 524
+    },
+    {
+      "epoch": 1.6791044776119404,
+      "grad_norm": 0.5981940676142077,
+      "learning_rate": 4.813487898226794e-06,
+      "loss": 0.6427,
+      "step": 525
+    },
+    {
+      "epoch": 1.6823027718550105,
+      "grad_norm": 0.5547199384817155,
+      "learning_rate": 4.7948466852238844e-06,
+      "loss": 0.6278,
+      "step": 526
+    },
+    {
+      "epoch": 1.6855010660980811,
+      "grad_norm": 0.5468385314356093,
+      "learning_rate": 4.7762083281906864e-06,
+      "loss": 0.594,
+      "step": 527
+    },
+    {
+      "epoch": 1.6886993603411513,
+      "grad_norm": 0.5456445809907,
+      "learning_rate": 4.757573086594529e-06,
+      "loss": 0.5884,
+      "step": 528
+    },
+    {
+      "epoch": 1.6918976545842217,
+      "grad_norm": 0.6646249117314513,
+      "learning_rate": 4.738941219859375e-06,
+      "loss": 0.6306,
+      "step": 529
+    },
+    {
+      "epoch": 1.695095948827292,
+      "grad_norm": 0.6292949980039241,
+      "learning_rate": 4.720312987362204e-06,
+      "loss": 0.6619,
+      "step": 530
+    },
+    {
+      "epoch": 1.6982942430703625,
+      "grad_norm": 0.6322989923958425,
+      "learning_rate": 4.701688648429399e-06,
+      "loss": 0.6266,
+      "step": 531
+    },
+    {
+      "epoch": 1.7014925373134329,
+      "grad_norm": 0.5305082924051039,
+      "learning_rate": 4.683068462333144e-06,
+      "loss": 0.6342,
+      "step": 532
+    },
+    {
+      "epoch": 1.7046908315565032,
+      "grad_norm": 0.550929491236182,
+      "learning_rate": 4.6644526882878145e-06,
+      "loss": 0.6392,
+      "step": 533
+    },
+    {
+      "epoch": 1.7078891257995736,
+      "grad_norm": 0.6232015245825243,
+      "learning_rate": 4.645841585446356e-06,
+      "loss": 0.6064,
+      "step": 534
+    },
+    {
+      "epoch": 1.7110874200426438,
+      "grad_norm": 0.5434413619942932,
+      "learning_rate": 4.6272354128966924e-06,
+      "loss": 0.6501,
+      "step": 535
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 0.5191107189759604,
+      "learning_rate": 4.6086344296581095e-06,
+      "loss": 0.6448,
+      "step": 536
+    },
+    {
+      "epoch": 1.7174840085287846,
+      "grad_norm": 0.5434292262769143,
+      "learning_rate": 4.590038894677653e-06,
+      "loss": 0.6074,
+      "step": 537
+    },
+    {
+      "epoch": 1.720682302771855,
+      "grad_norm": 0.4655865353721599,
+      "learning_rate": 4.5714490668265245e-06,
+      "loss": 0.6236,
+      "step": 538
+    },
+    {
+      "epoch": 1.7238805970149254,
+      "grad_norm": 0.550629570360553,
+      "learning_rate": 4.55286520489647e-06,
+      "loss": 0.6397,
+      "step": 539
+    },
+    {
+      "epoch": 1.7270788912579957,
+      "grad_norm": 0.6004187629555796,
+      "learning_rate": 4.534287567596189e-06,
+      "loss": 0.6385,
+      "step": 540
+    },
+    {
+      "epoch": 1.7302771855010661,
+      "grad_norm": 0.5176720613714986,
+      "learning_rate": 4.515716413547722e-06,
+      "loss": 0.6429,
+      "step": 541
+    },
+    {
+      "epoch": 1.7334754797441365,
+      "grad_norm": 0.5115763781808625,
+      "learning_rate": 4.497152001282861e-06,
+      "loss": 0.6382,
+      "step": 542
+    },
+    {
+      "epoch": 1.736673773987207,
+      "grad_norm": 0.5562269449292555,
+      "learning_rate": 4.478594589239539e-06,
+      "loss": 0.6236,
+      "step": 543
+    },
+    {
+      "epoch": 1.739872068230277,
+      "grad_norm": 0.5670914639063015,
+      "learning_rate": 4.460044435758241e-06,
+      "loss": 0.6441,
+      "step": 544
+    },
+    {
+      "epoch": 1.7430703624733477,
+      "grad_norm": 0.5489747231617312,
+      "learning_rate": 4.441501799078405e-06,
+      "loss": 0.613,
+      "step": 545
+    },
+    {
+      "epoch": 1.7462686567164178,
+      "grad_norm": 0.5424103892867449,
+      "learning_rate": 4.4229669373348225e-06,
+      "loss": 0.6253,
+      "step": 546
+    },
+    {
+      "epoch": 1.7494669509594882,
+      "grad_norm": 0.5562973456307682,
+      "learning_rate": 4.404440108554055e-06,
+      "loss": 0.6168,
+      "step": 547
+    },
+    {
+      "epoch": 1.7526652452025586,
+      "grad_norm": 0.533299067462362,
+      "learning_rate": 4.3859215706508295e-06,
+      "loss": 0.6485,
+      "step": 548
+    },
+    {
+      "epoch": 1.755863539445629,
+      "grad_norm": 0.5172786400673339,
+      "learning_rate": 4.3674115814244595e-06,
+      "loss": 0.6323,
+      "step": 549
+    },
+    {
+      "epoch": 1.7590618336886994,
+      "grad_norm": 0.5085177296011396,
+      "learning_rate": 4.348910398555249e-06,
+      "loss": 0.5899,
+      "step": 550
+    },
+    {
+      "epoch": 1.7622601279317696,
+      "grad_norm": 0.47810924329720716,
+      "learning_rate": 4.330418279600907e-06,
+      "loss": 0.6379,
+      "step": 551
+    },
+    {
+      "epoch": 1.7654584221748402,
+      "grad_norm": 0.5321362592336698,
+      "learning_rate": 4.311935481992965e-06,
+      "loss": 0.6151,
+      "step": 552
+    },
+    {
+      "epoch": 1.7686567164179103,
+      "grad_norm": 0.5446514716404816,
+      "learning_rate": 4.2934622630331855e-06,
+      "loss": 0.6079,
+      "step": 553
+    },
+    {
+      "epoch": 1.771855010660981,
+      "grad_norm": 0.5237047324414964,
+      "learning_rate": 4.274998879889991e-06,
+      "loss": 0.6427,
+      "step": 554
+    },
+    {
+      "epoch": 1.775053304904051,
+      "grad_norm": 0.5365142222456774,
+      "learning_rate": 4.2565455895948745e-06,
+      "loss": 0.582,
+      "step": 555
+    },
+    {
+      "epoch": 1.7782515991471215,
+      "grad_norm": 0.5308651285729982,
+      "learning_rate": 4.238102649038825e-06,
+      "loss": 0.6575,
+      "step": 556
+    },
+    {
+      "epoch": 1.7814498933901919,
+      "grad_norm": 0.5717647379657843,
+      "learning_rate": 4.219670314968754e-06,
+      "loss": 0.62,
+      "step": 557
+    },
+    {
+      "epoch": 1.7846481876332623,
+      "grad_norm": 0.5050163481740146,
+      "learning_rate": 4.2012488439839185e-06,
+      "loss": 0.6636,
+      "step": 558
+    },
+    {
+      "epoch": 1.7878464818763327,
+      "grad_norm": 0.49227167259915716,
+      "learning_rate": 4.182838492532342e-06,
+      "loss": 0.5945,
+      "step": 559
+    },
+    {
+      "epoch": 1.7910447761194028,
+      "grad_norm": 0.5302015513388415,
+      "learning_rate": 4.164439516907258e-06,
+      "loss": 0.6106,
+      "step": 560
+    },
+    {
+      "epoch": 1.7942430703624734,
+      "grad_norm": 0.64011360289916,
+      "learning_rate": 4.146052173243538e-06,
+      "loss": 0.6593,
+      "step": 561
+    },
+    {
+      "epoch": 1.7974413646055436,
+      "grad_norm": 0.601687625944857,
+      "learning_rate": 4.127676717514114e-06,
+      "loss": 0.6092,
+      "step": 562
+    },
+    {
+      "epoch": 1.8006396588486142,
+      "grad_norm": 0.524585479629224,
+      "learning_rate": 4.109313405526433e-06,
+      "loss": 0.6446,
+      "step": 563
+    },
+    {
+      "epoch": 1.8038379530916844,
+      "grad_norm": 0.6161148649503772,
+      "learning_rate": 4.090962492918881e-06,
+      "loss": 0.6233,
+      "step": 564
+    },
+    {
+      "epoch": 1.8070362473347548,
+      "grad_norm": 0.6062559672756939,
+      "learning_rate": 4.072624235157234e-06,
+      "loss": 0.6623,
+      "step": 565
+    },
+    {
+      "epoch": 1.8102345415778252,
+      "grad_norm": 0.5985336173857585,
+      "learning_rate": 4.054298887531099e-06,
+      "loss": 0.582,
+      "step": 566
+    },
+    {
+      "epoch": 1.8134328358208955,
+      "grad_norm": 0.5632646612772408,
+      "learning_rate": 4.035986705150355e-06,
+      "loss": 0.6159,
+      "step": 567
+    },
+    {
+      "epoch": 1.816631130063966,
+      "grad_norm": 0.7068557752731836,
+      "learning_rate": 4.017687942941609e-06,
+      "loss": 0.6536,
+      "step": 568
+    },
+    {
+      "epoch": 1.819829424307036,
+      "grad_norm": 0.5730952893121638,
+      "learning_rate": 3.9994028556446404e-06,
+      "loss": 0.6241,
+      "step": 569
+    },
+    {
+      "epoch": 1.8230277185501067,
+      "grad_norm": 0.5175887707114529,
+      "learning_rate": 3.981131697808862e-06,
+      "loss": 0.6627,
+      "step": 570
+    },
+    {
+      "epoch": 1.8262260127931769,
+      "grad_norm": 0.5796987142687964,
+      "learning_rate": 3.9628747237897715e-06,
+      "loss": 0.6536,
+      "step": 571
+    },
+    {
+      "epoch": 1.8294243070362475,
+      "grad_norm": 0.5271014946968099,
+      "learning_rate": 3.94463218774541e-06,
+      "loss": 0.6467,
+      "step": 572
+    },
+    {
+      "epoch": 1.8326226012793176,
+      "grad_norm": 0.5704417597925869,
+      "learning_rate": 3.926404343632826e-06,
+      "loss": 0.6831,
+      "step": 573
+    },
+    {
+      "epoch": 1.835820895522388,
+      "grad_norm": 0.5292627931377012,
+      "learning_rate": 3.90819144520454e-06,
+      "loss": 0.6408,
+      "step": 574
+    },
+    {
+      "epoch": 1.8390191897654584,
+      "grad_norm": 0.5646881580541779,
+      "learning_rate": 3.889993746005011e-06,
+      "loss": 0.6348,
+      "step": 575
+    },
+    {
+      "epoch": 1.8422174840085288,
+      "grad_norm": 0.5401410573662442,
+      "learning_rate": 3.8718114993671086e-06,
+      "loss": 0.5917,
+      "step": 576
+    },
+    {
+      "epoch": 1.8454157782515992,
+      "grad_norm": 0.5033737411697443,
+      "learning_rate": 3.853644958408582e-06,
+      "loss": 0.6454,
+      "step": 577
+    },
+    {
+      "epoch": 1.8486140724946694,
+      "grad_norm": 0.5317406038132683,
+      "learning_rate": 3.835494376028544e-06,
+      "loss": 0.579,
+      "step": 578
+    },
+    {
+      "epoch": 1.85181236673774,
+      "grad_norm": 0.5247944872571422,
+      "learning_rate": 3.817360004903939e-06,
+      "loss": 0.6443,
+      "step": 579
+    },
+    {
+      "epoch": 1.8550106609808101,
+      "grad_norm": 0.5255266393479685,
+      "learning_rate": 3.799242097486038e-06,
+      "loss": 0.6548,
+      "step": 580
+    },
+    {
+      "epoch": 1.8582089552238807,
+      "grad_norm": 0.5562031814775014,
+      "learning_rate": 3.7811409059969177e-06,
+      "loss": 0.6368,
+      "step": 581
+    },
+    {
+      "epoch": 1.861407249466951,
+      "grad_norm": 0.5490844678533507,
+      "learning_rate": 3.7630566824259456e-06,
+      "loss": 0.6001,
+      "step": 582
+    },
+    {
+      "epoch": 1.8646055437100213,
+      "grad_norm": 0.5257106510728633,
+      "learning_rate": 3.7449896785262817e-06,
+      "loss": 0.6092,
+      "step": 583
+    },
+    {
+      "epoch": 1.8678038379530917,
+      "grad_norm": 0.5410953577384838,
+      "learning_rate": 3.726940145811363e-06,
+      "loss": 0.643,
+      "step": 584
+    },
+    {
+      "epoch": 1.871002132196162,
+      "grad_norm": 0.5205935779573823,
+      "learning_rate": 3.708908335551412e-06,
+      "loss": 0.6609,
+      "step": 585
+    },
+    {
+      "epoch": 1.8742004264392325,
+      "grad_norm": 0.5715057818226926,
+      "learning_rate": 3.6908944987699346e-06,
+      "loss": 0.6191,
+      "step": 586
+    },
+    {
+      "epoch": 1.8773987206823026,
+      "grad_norm": 0.5410333169459866,
+      "learning_rate": 3.67289888624022e-06,
+      "loss": 0.6169,
+      "step": 587
+    },
+    {
+      "epoch": 1.8805970149253732,
+      "grad_norm": 0.5321189795175738,
+      "learning_rate": 3.6549217484818576e-06,
+      "loss": 0.5957,
+      "step": 588
+    },
+    {
+      "epoch": 1.8837953091684434,
+      "grad_norm": 0.6038607670908029,
+      "learning_rate": 3.6369633357572514e-06,
+      "loss": 0.6112,
+      "step": 589
+    },
+    {
+      "epoch": 1.886993603411514,
+      "grad_norm": 0.5382547649391735,
+      "learning_rate": 3.6190238980681235e-06,
+      "loss": 0.6446,
+      "step": 590
+    },
+    {
+      "epoch": 1.8901918976545842,
+      "grad_norm": 0.509428705908829,
+      "learning_rate": 3.6011036851520465e-06,
+      "loss": 0.6245,
+      "step": 591
+    },
+    {
+      "epoch": 1.8933901918976546,
+      "grad_norm": 0.5331030357626905,
+      "learning_rate": 3.583202946478963e-06,
+      "loss": 0.6667,
+      "step": 592
+    },
+    {
+      "epoch": 1.896588486140725,
+      "grad_norm": 0.5957179693199653,
+      "learning_rate": 3.5653219312477085e-06,
+      "loss": 0.6164,
+      "step": 593
+    },
+    {
+      "epoch": 1.8997867803837953,
+      "grad_norm": 0.540978235449404,
+      "learning_rate": 3.5474608883825475e-06,
+      "loss": 0.6814,
+      "step": 594
+    },
+    {
+      "epoch": 1.9029850746268657,
+      "grad_norm": 0.5609593877190009,
+      "learning_rate": 3.529620066529704e-06,
+      "loss": 0.6337,
+      "step": 595
+    },
+    {
+      "epoch": 1.906183368869936,
+      "grad_norm": 0.5759309632996505,
+      "learning_rate": 3.5117997140539073e-06,
+      "loss": 0.637,
+      "step": 596
+    },
+    {
+      "epoch": 1.9093816631130065,
+      "grad_norm": 0.5913831985711084,
+      "learning_rate": 3.4940000790349226e-06,
+      "loss": 0.6288,
+      "step": 597
+    },
+    {
+      "epoch": 1.9125799573560767,
+      "grad_norm": 0.5075699761691839,
+      "learning_rate": 3.47622140926411e-06,
+      "loss": 0.6173,
+      "step": 598
+    },
+    {
+      "epoch": 1.9157782515991473,
+      "grad_norm": 0.6653450535384141,
+      "learning_rate": 3.458463952240967e-06,
+      "loss": 0.6523,
+      "step": 599
+    },
+    {
+      "epoch": 1.9189765458422174,
+      "grad_norm": 0.628552652910662,
+      "learning_rate": 3.4407279551696846e-06,
+      "loss": 0.6358,
+      "step": 600
+    },
+    {
+      "epoch": 1.9221748400852878,
+      "grad_norm": 0.5333242599508652,
+      "learning_rate": 3.4230136649557087e-06,
+      "loss": 0.6288,
+      "step": 601
+    },
+    {
+      "epoch": 1.9253731343283582,
+      "grad_norm": 0.47146898112937263,
+      "learning_rate": 3.4053213282022983e-06,
+      "loss": 0.6596,
+      "step": 602
+    },
+    {
+      "epoch": 1.9285714285714286,
+      "grad_norm": 0.6621044867075212,
+      "learning_rate": 3.387651191207097e-06,
+      "loss": 0.5934,
+      "step": 603
+    },
+    {
+      "epoch": 1.931769722814499,
+      "grad_norm": 0.5923983943553847,
+      "learning_rate": 3.370003499958703e-06,
+      "loss": 0.6332,
+      "step": 604
+    },
+    {
+      "epoch": 1.9349680170575692,
+      "grad_norm": 0.5262532242135438,
+      "learning_rate": 3.352378500133239e-06,
+      "loss": 0.6164,
+      "step": 605
+    },
+    {
+      "epoch": 1.9381663113006398,
+      "grad_norm": 0.6174684357894091,
+      "learning_rate": 3.334776437090944e-06,
+      "loss": 0.6276,
+      "step": 606
+    },
+    {
+      "epoch": 1.94136460554371,
+      "grad_norm": 0.6066479008889557,
+      "learning_rate": 3.317197555872745e-06,
+      "loss": 0.6061,
+      "step": 607
+    },
+    {
+      "epoch": 1.9445628997867805,
+      "grad_norm": 0.5282593114168354,
+      "learning_rate": 3.2996421011968546e-06,
+      "loss": 0.5957,
+      "step": 608
+    },
+    {
+      "epoch": 1.9477611940298507,
+      "grad_norm": 0.5428512832531854,
+      "learning_rate": 3.28211031745536e-06,
+      "loss": 0.6274,
+      "step": 609
+    },
+    {
+      "epoch": 1.950959488272921,
+      "grad_norm": 0.563963200457514,
+      "learning_rate": 3.264602448710822e-06,
+      "loss": 0.5977,
+      "step": 610
+    },
+    {
+      "epoch": 1.9541577825159915,
+      "grad_norm": 0.506718113639374,
+      "learning_rate": 3.2471187386928766e-06,
+      "loss": 0.5975,
+      "step": 611
+    },
+    {
+      "epoch": 1.9573560767590619,
+      "grad_norm": 0.5077307814187035,
+      "learning_rate": 3.2296594307948428e-06,
+      "loss": 0.6401,
+      "step": 612
+    },
+    {
+      "epoch": 1.9605543710021323,
+      "grad_norm": 0.5128850695519592,
+      "learning_rate": 3.212224768070334e-06,
+      "loss": 0.6381,
+      "step": 613
+    },
+    {
+      "epoch": 1.9637526652452024,
+      "grad_norm": 0.5083765825134322,
+      "learning_rate": 3.194814993229878e-06,
+      "loss": 0.62,
+      "step": 614
+    },
+    {
+      "epoch": 1.966950959488273,
+      "grad_norm": 0.47099407882432265,
+      "learning_rate": 3.177430348637527e-06,
+      "loss": 0.6125,
+      "step": 615
+    },
+    {
+      "epoch": 1.9701492537313432,
+      "grad_norm": 0.5530258695290763,
+      "learning_rate": 3.1600710763074972e-06,
+      "loss": 0.6356,
+      "step": 616
+    },
+    {
+      "epoch": 1.9733475479744138,
+      "grad_norm": 0.5347670899530949,
+      "learning_rate": 3.142737417900793e-06,
+      "loss": 0.6611,
+      "step": 617
+    },
+    {
+      "epoch": 1.976545842217484,
+      "grad_norm": 0.5188746097889174,
+      "learning_rate": 3.125429614721842e-06,
+      "loss": 0.6185,
+      "step": 618
+    },
+    {
+      "epoch": 1.9797441364605544,
+      "grad_norm": 0.47423520536928904,
+      "learning_rate": 3.1081479077151387e-06,
+      "loss": 0.6207,
+      "step": 619
+    },
+    {
+      "epoch": 1.9829424307036247,
+      "grad_norm": 0.5191715465473807,
+      "learning_rate": 3.090892537461889e-06,
+      "loss": 0.604,
+      "step": 620
+    },
+    {
+      "epoch": 1.9861407249466951,
+      "grad_norm": 0.6101222943327476,
+      "learning_rate": 3.0736637441766594e-06,
+      "loss": 0.6234,
+      "step": 621
+    },
+    {
+      "epoch": 1.9893390191897655,
+      "grad_norm": 0.5351791226481845,
+      "learning_rate": 3.056461767704037e-06,
+      "loss": 0.6411,
+      "step": 622
+    },
+    {
+      "epoch": 1.9925373134328357,
+      "grad_norm": 0.5193133031919753,
+      "learning_rate": 3.039286847515284e-06,
+      "loss": 0.6175,
+      "step": 623
+    },
+    {
+      "epoch": 1.9957356076759063,
+      "grad_norm": 0.5316524549516811,
+      "learning_rate": 3.0221392227050126e-06,
+      "loss": 0.6408,
+      "step": 624
+    },
+    {
+      "epoch": 1.9989339019189765,
+      "grad_norm": 0.5093874593326244,
+      "learning_rate": 3.00501913198785e-06,
+      "loss": 0.6261,
+      "step": 625
+    },
+    {
+      "epoch": 2.002132196162047,
+      "grad_norm": 0.6163744912473614,
+      "learning_rate": 2.9879268136951163e-06,
+      "loss": 0.5951,
+      "step": 626
+    },
+    {
+      "epoch": 2.0053304904051172,
+      "grad_norm": 0.5783808952696761,
+      "learning_rate": 2.970862505771509e-06,
+      "loss": 0.5622,
+      "step": 627
+    },
+    {
+      "epoch": 2.008528784648188,
+      "grad_norm": 0.49273447076344123,
+      "learning_rate": 2.953826445771788e-06,
+      "loss": 0.5673,
+      "step": 628
+    },
+    {
+      "epoch": 2.011727078891258,
+      "grad_norm": 0.5859845266864249,
+      "learning_rate": 2.9368188708574706e-06,
+      "loss": 0.6246,
+      "step": 629
+    },
+    {
+      "epoch": 2.014925373134328,
+      "grad_norm": 0.5611571675265489,
+      "learning_rate": 2.9198400177935303e-06,
+      "loss": 0.6277,
+      "step": 630
+    },
+    {
+      "epoch": 2.018123667377399,
+      "grad_norm": 0.5276964961204649,
+      "learning_rate": 2.902890122945096e-06,
+      "loss": 0.5577,
+      "step": 631
+    },
+    {
+      "epoch": 2.021321961620469,
+      "grad_norm": 0.525544540323068,
+      "learning_rate": 2.8859694222741653e-06,
+      "loss": 0.6051,
+      "step": 632
+    },
+    {
+      "epoch": 2.0245202558635396,
+      "grad_norm": 0.5673683781918863,
+      "learning_rate": 2.869078151336323e-06,
+      "loss": 0.5808,
+      "step": 633
+    },
+    {
+      "epoch": 2.0277185501066097,
+      "grad_norm": 0.5589784105831517,
+      "learning_rate": 2.852216545277456e-06,
+      "loss": 0.6106,
+      "step": 634
+    },
+    {
+      "epoch": 2.0309168443496803,
+      "grad_norm": 0.5030496312589189,
+      "learning_rate": 2.835384838830481e-06,
+      "loss": 0.5541,
+      "step": 635
+    },
+    {
+      "epoch": 2.0341151385927505,
+      "grad_norm": 0.6708437250927518,
+      "learning_rate": 2.8185832663120817e-06,
+      "loss": 0.6271,
+      "step": 636
+    },
+    {
+      "epoch": 2.0373134328358207,
+      "grad_norm": 0.5763870321226472,
+      "learning_rate": 2.8018120616194356e-06,
+      "loss": 0.5507,
+      "step": 637
+    },
+    {
+      "epoch": 2.0405117270788913,
+      "grad_norm": 0.5752129249821096,
+      "learning_rate": 2.785071458226972e-06,
+      "loss": 0.5978,
+      "step": 638
+    },
+    {
+      "epoch": 2.0437100213219614,
+      "grad_norm": 0.5054941801603793,
+      "learning_rate": 2.768361689183113e-06,
+      "loss": 0.6203,
+      "step": 639
+    },
+    {
+      "epoch": 2.046908315565032,
+      "grad_norm": 0.5481782008730065,
+      "learning_rate": 2.7516829871070295e-06,
+      "loss": 0.6093,
+      "step": 640
+    },
+    {
+      "epoch": 2.050106609808102,
+      "grad_norm": 0.5323921761861825,
+      "learning_rate": 2.735035584185409e-06,
+      "loss": 0.6007,
+      "step": 641
+    },
+    {
+      "epoch": 2.053304904051173,
+      "grad_norm": 0.5293554299631367,
+      "learning_rate": 2.718419712169213e-06,
+      "loss": 0.5884,
+      "step": 642
+    },
+    {
+      "epoch": 2.056503198294243,
+      "grad_norm": 0.553435984362884,
+      "learning_rate": 2.7018356023704574e-06,
+      "loss": 0.5326,
+      "step": 643
+    },
+    {
+      "epoch": 2.0597014925373136,
+      "grad_norm": 0.5139320287446054,
+      "learning_rate": 2.685283485658995e-06,
+      "loss": 0.5639,
+      "step": 644
+    },
+    {
+      "epoch": 2.0628997867803838,
+      "grad_norm": 0.49577644020424133,
+      "learning_rate": 2.668763592459297e-06,
+      "loss": 0.5571,
+      "step": 645
+    },
+    {
+      "epoch": 2.066098081023454,
+      "grad_norm": 0.5089676714103052,
+      "learning_rate": 2.6522761527472464e-06,
+      "loss": 0.6163,
+      "step": 646
+    },
+    {
+      "epoch": 2.0692963752665245,
+      "grad_norm": 0.5593194742250533,
+      "learning_rate": 2.6358213960469357e-06,
+      "loss": 0.5808,
+      "step": 647
+    },
+    {
+      "epoch": 2.0724946695095947,
+      "grad_norm": 0.5130651147305723,
+      "learning_rate": 2.6193995514274705e-06,
+      "loss": 0.5953,
+      "step": 648
+    },
+    {
+      "epoch": 2.0756929637526653,
+      "grad_norm": 0.5609831284876498,
+      "learning_rate": 2.6030108474997854e-06,
+      "loss": 0.5841,
+      "step": 649
+    },
+    {
+      "epoch": 2.0788912579957355,
+      "grad_norm": 0.47312462257776017,
+      "learning_rate": 2.586655512413458e-06,
+      "loss": 0.5967,
+      "step": 650
+    },
+    {
+      "epoch": 2.082089552238806,
+      "grad_norm": 0.5001915039170618,
+      "learning_rate": 2.5703337738535324e-06,
+      "loss": 0.5485,
+      "step": 651
+    },
+    {
+      "epoch": 2.0852878464818763,
+      "grad_norm": 0.5785964658330373,
+      "learning_rate": 2.554045859037353e-06,
+      "loss": 0.566,
+      "step": 652
+    },
+    {
+      "epoch": 2.088486140724947,
+      "grad_norm": 0.5912116290504553,
+      "learning_rate": 2.5377919947113917e-06,
+      "loss": 0.5856,
+      "step": 653
+    },
+    {
+      "epoch": 2.091684434968017,
+      "grad_norm": 0.5209304706365177,
+      "learning_rate": 2.521572407148107e-06,
+      "loss": 0.5556,
+      "step": 654
+    },
+    {
+      "epoch": 2.094882729211087,
+      "grad_norm": 0.5101895622941909,
+      "learning_rate": 2.505387322142782e-06,
+      "loss": 0.5776,
+      "step": 655
+    },
+    {
+      "epoch": 2.098081023454158,
+      "grad_norm": 0.4880680240587064,
+      "learning_rate": 2.4892369650103837e-06,
+      "loss": 0.6056,
+      "step": 656
+    },
+    {
+      "epoch": 2.101279317697228,
+      "grad_norm": 0.5514381593453722,
+      "learning_rate": 2.4731215605824304e-06,
+      "loss": 0.6006,
+      "step": 657
+    },
+    {
+      "epoch": 2.1044776119402986,
+      "grad_norm": 0.5751639987148548,
+      "learning_rate": 2.4570413332038523e-06,
+      "loss": 0.5665,
+      "step": 658
+    },
+    {
+      "epoch": 2.1076759061833688,
+      "grad_norm": 0.5804924762978739,
+      "learning_rate": 2.440996506729881e-06,
+      "loss": 0.5973,
+      "step": 659
+    },
+    {
+      "epoch": 2.1108742004264394,
+      "grad_norm": 0.4918030532858259,
+      "learning_rate": 2.4249873045229244e-06,
+      "loss": 0.57,
+      "step": 660
+    },
+    {
+      "epoch": 2.1140724946695095,
+      "grad_norm": 0.5243890793969488,
+      "learning_rate": 2.4090139494494596e-06,
+      "loss": 0.5871,
+      "step": 661
+    },
+    {
+      "epoch": 2.11727078891258,
+      "grad_norm": 0.4806747676363039,
+      "learning_rate": 2.3930766638769325e-06,
+      "loss": 0.5598,
+      "step": 662
+    },
+    {
+      "epoch": 2.1204690831556503,
+      "grad_norm": 0.5175543377119891,
+      "learning_rate": 2.3771756696706553e-06,
+      "loss": 0.6051,
+      "step": 663
+    },
+    {
+      "epoch": 2.1236673773987205,
+      "grad_norm": 0.5246309451300202,
+      "learning_rate": 2.3613111881907273e-06,
+      "loss": 0.598,
+      "step": 664
+    },
+    {
+      "epoch": 2.126865671641791,
+      "grad_norm": 0.526456441833946,
+      "learning_rate": 2.345483440288947e-06,
+      "loss": 0.547,
+      "step": 665
+    },
+    {
+      "epoch": 2.1300639658848612,
+      "grad_norm": 0.47184799877761335,
+      "learning_rate": 2.3296926463057396e-06,
+      "loss": 0.6138,
+      "step": 666
+    },
+    {
+      "epoch": 2.133262260127932,
+      "grad_norm": 0.48133254553344923,
+      "learning_rate": 2.313939026067091e-06,
+      "loss": 0.583,
+      "step": 667
+    },
+    {
+      "epoch": 2.136460554371002,
+      "grad_norm": 0.5064732476569918,
+      "learning_rate": 2.29822279888148e-06,
+      "loss": 0.5454,
+      "step": 668
+    },
+    {
+      "epoch": 2.1396588486140726,
+      "grad_norm": 0.5781888574823206,
+      "learning_rate": 2.2825441835368377e-06,
+      "loss": 0.5712,
+      "step": 669
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 0.5978766513888759,
+      "learning_rate": 2.2669033982974946e-06,
+      "loss": 0.5695,
+      "step": 670
+    },
+    {
+      "epoch": 2.1460554371002134,
+      "grad_norm": 0.4555096407611019,
+      "learning_rate": 2.2513006609011365e-06,
+      "loss": 0.5555,
+      "step": 671
+    },
+    {
+      "epoch": 2.1492537313432836,
+      "grad_norm": 0.4676695942649127,
+      "learning_rate": 2.235736188555787e-06,
+      "loss": 0.5735,
+      "step": 672
+    },
+    {
+      "epoch": 2.1524520255863537,
+      "grad_norm": 0.4763343046497065,
+      "learning_rate": 2.2202101979367735e-06,
+      "loss": 0.5955,
+      "step": 673
+    },
+    {
+      "epoch": 2.1556503198294243,
+      "grad_norm": 0.4657805538089298,
+      "learning_rate": 2.2047229051837107e-06,
+      "loss": 0.6076,
+      "step": 674
+    },
+    {
+      "epoch": 2.1588486140724945,
+      "grad_norm": 0.4838763721788115,
+      "learning_rate": 2.189274525897498e-06,
+      "loss": 0.616,
+      "step": 675
+    },
+    {
+      "epoch": 2.162046908315565,
+      "grad_norm": 0.5013682671987199,
+      "learning_rate": 2.173865275137314e-06,
+      "loss": 0.5809,
+      "step": 676
+    },
+    {
+      "epoch": 2.1652452025586353,
+      "grad_norm": 0.4730326559560223,
+      "learning_rate": 2.158495367417625e-06,
+      "loss": 0.5758,
+      "step": 677
+    },
+    {
+      "epoch": 2.168443496801706,
+      "grad_norm": 0.49708415063251904,
+      "learning_rate": 2.143165016705192e-06,
+      "loss": 0.5859,
+      "step": 678
+    },
+    {
+      "epoch": 2.171641791044776,
+      "grad_norm": 0.4952662090916961,
+      "learning_rate": 2.1278744364161007e-06,
+      "loss": 0.5575,
+      "step": 679
+    },
+    {
+      "epoch": 2.1748400852878467,
+      "grad_norm": 0.46833442631122313,
+      "learning_rate": 2.1126238394127868e-06,
+      "loss": 0.5974,
+      "step": 680
+    },
+    {
+      "epoch": 2.178038379530917,
+      "grad_norm": 0.4714670060045907,
+      "learning_rate": 2.0974134380010726e-06,
+      "loss": 0.5751,
+      "step": 681
+    },
+    {
+      "epoch": 2.181236673773987,
+      "grad_norm": 0.4772805496941911,
+      "learning_rate": 2.082243443927212e-06,
+      "loss": 0.5996,
+      "step": 682
+    },
+    {
+      "epoch": 2.1844349680170576,
+      "grad_norm": 0.4606017759107387,
+      "learning_rate": 2.0671140683749386e-06,
+      "loss": 0.5817,
+      "step": 683
+    },
+    {
+      "epoch": 2.1876332622601278,
+      "grad_norm": 0.4774965126191024,
+      "learning_rate": 2.052025521962534e-06,
+      "loss": 0.5961,
+      "step": 684
+    },
+    {
+      "epoch": 2.1908315565031984,
+      "grad_norm": 0.49760796326925766,
+      "learning_rate": 2.03697801473989e-06,
+      "loss": 0.5788,
+      "step": 685
+    },
+    {
+      "epoch": 2.1940298507462686,
+      "grad_norm": 0.4871186232939493,
+      "learning_rate": 2.0219717561855857e-06,
+      "loss": 0.5879,
+      "step": 686
+    },
+    {
+      "epoch": 2.197228144989339,
+      "grad_norm": 0.464109823915951,
+      "learning_rate": 2.0070069552039722e-06,
+      "loss": 0.6154,
+      "step": 687
+    },
+    {
+      "epoch": 2.2004264392324093,
+      "grad_norm": 0.4856649377645587,
+      "learning_rate": 1.992083820122259e-06,
+      "loss": 0.5788,
+      "step": 688
+    },
+    {
+      "epoch": 2.20362473347548,
+      "grad_norm": 0.46996237636688565,
+      "learning_rate": 1.9772025586876252e-06,
+      "loss": 0.5953,
+      "step": 689
+    },
+    {
+      "epoch": 2.20682302771855,
+      "grad_norm": 0.46859069921217344,
+      "learning_rate": 1.962363378064316e-06,
+      "loss": 0.6083,
+      "step": 690
+    },
+    {
+      "epoch": 2.2100213219616203,
+      "grad_norm": 0.5168988401256138,
+      "learning_rate": 1.947566484830765e-06,
+      "loss": 0.5814,
+      "step": 691
+    },
+    {
+      "epoch": 2.213219616204691,
+      "grad_norm": 0.5406960885215742,
+      "learning_rate": 1.9328120849767198e-06,
+      "loss": 0.5561,
+      "step": 692
+    },
+    {
+      "epoch": 2.216417910447761,
+      "grad_norm": 0.44334448566836304,
+      "learning_rate": 1.9181003839003627e-06,
+      "loss": 0.5927,
+      "step": 693
+    },
+    {
+      "epoch": 2.2196162046908317,
+      "grad_norm": 0.4557583490316895,
+      "learning_rate": 1.9034315864054682e-06,
+      "loss": 0.5862,
+      "step": 694
+    },
+    {
+      "epoch": 2.222814498933902,
+      "grad_norm": 0.5092922861926251,
+      "learning_rate": 1.8888058966985407e-06,
+      "loss": 0.5668,
+      "step": 695
+    },
+    {
+      "epoch": 2.2260127931769724,
+      "grad_norm": 0.5004818455277625,
+      "learning_rate": 1.8742235183859747e-06,
+      "loss": 0.5968,
+      "step": 696
+    },
+    {
+      "epoch": 2.2292110874200426,
+      "grad_norm": 0.4949014258712852,
+      "learning_rate": 1.8596846544712233e-06,
+      "loss": 0.622,
+      "step": 697
+    },
+    {
+      "epoch": 2.232409381663113,
+      "grad_norm": 0.47178385456983946,
+      "learning_rate": 1.8451895073519643e-06,
+      "loss": 0.5725,
+      "step": 698
+    },
+    {
+      "epoch": 2.2356076759061834,
+      "grad_norm": 0.4709195571648972,
+      "learning_rate": 1.8307382788172877e-06,
+      "loss": 0.5991,
+      "step": 699
+    },
+    {
+      "epoch": 2.2388059701492535,
+      "grad_norm": 0.4806780947771992,
+      "learning_rate": 1.8163311700448899e-06,
+      "loss": 0.5976,
+      "step": 700
+    },
+    {
+      "epoch": 2.242004264392324,
+      "grad_norm": 0.46974071200274314,
+      "learning_rate": 1.8019683815982691e-06,
+      "loss": 0.6091,
+      "step": 701
+    },
+    {
+      "epoch": 2.2452025586353943,
+      "grad_norm": 0.45361098939294514,
+      "learning_rate": 1.7876501134239316e-06,
+      "loss": 0.5933,
+      "step": 702
+    },
+    {
+      "epoch": 2.248400852878465,
+      "grad_norm": 0.4850717880695125,
+      "learning_rate": 1.7733765648486134e-06,
+      "loss": 0.5597,
+      "step": 703
+    },
+    {
+      "epoch": 2.251599147121535,
+      "grad_norm": 0.4692790998035614,
+      "learning_rate": 1.7591479345764972e-06,
+      "loss": 0.6007,
+      "step": 704
+    },
+    {
+      "epoch": 2.2547974413646057,
+      "grad_norm": 0.4906873053697701,
+      "learning_rate": 1.7449644206864564e-06,
+      "loss": 0.552,
+      "step": 705
+    },
+    {
+      "epoch": 2.257995735607676,
+      "grad_norm": 0.481979324803139,
+      "learning_rate": 1.7308262206292898e-06,
+      "loss": 0.6125,
+      "step": 706
+    },
+    {
+      "epoch": 2.2611940298507465,
+      "grad_norm": 0.5016963903077071,
+      "learning_rate": 1.7167335312249766e-06,
+      "loss": 0.6017,
+      "step": 707
+    },
+    {
+      "epoch": 2.2643923240938166,
+      "grad_norm": 0.47149916005013454,
+      "learning_rate": 1.7026865486599375e-06,
+      "loss": 0.5869,
+      "step": 708
+    },
+    {
+      "epoch": 2.267590618336887,
+      "grad_norm": 0.507514553246801,
+      "learning_rate": 1.6886854684842962e-06,
+      "loss": 0.5926,
+      "step": 709
+    },
+    {
+      "epoch": 2.2707889125799574,
+      "grad_norm": 0.49251780946608525,
+      "learning_rate": 1.6747304856091662e-06,
+      "loss": 0.5874,
+      "step": 710
+    },
+    {
+      "epoch": 2.2739872068230276,
+      "grad_norm": 0.47492882189499497,
+      "learning_rate": 1.660821794303934e-06,
+      "loss": 0.6,
+      "step": 711
+    },
+    {
+      "epoch": 2.277185501066098,
+      "grad_norm": 0.4403913104531341,
+      "learning_rate": 1.6469595881935523e-06,
+      "loss": 0.5673,
+      "step": 712
+    },
+    {
+      "epoch": 2.2803837953091683,
+      "grad_norm": 0.5039434479895408,
+      "learning_rate": 1.6331440602558501e-06,
+      "loss": 0.5908,
+      "step": 713
+    },
+    {
+      "epoch": 2.283582089552239,
+      "grad_norm": 0.4893793347672981,
+      "learning_rate": 1.6193754028188363e-06,
+      "loss": 0.5757,
+      "step": 714
+    },
+    {
+      "epoch": 2.286780383795309,
+      "grad_norm": 0.4833544188283998,
+      "learning_rate": 1.6056538075580342e-06,
+      "loss": 0.6034,
+      "step": 715
+    },
+    {
+      "epoch": 2.2899786780383797,
+      "grad_norm": 0.46660784623966395,
+      "learning_rate": 1.591979465493806e-06,
+      "loss": 0.554,
+      "step": 716
+    },
+    {
+      "epoch": 2.29317697228145,
+      "grad_norm": 0.4941030437078769,
+      "learning_rate": 1.5783525669886934e-06,
+      "loss": 0.5615,
+      "step": 717
+    },
+    {
+      "epoch": 2.29637526652452,
+      "grad_norm": 0.467606067398601,
+      "learning_rate": 1.5647733017447741e-06,
+      "loss": 0.608,
+      "step": 718
+    },
+    {
+      "epoch": 2.2995735607675907,
+      "grad_norm": 0.47853133402269155,
+      "learning_rate": 1.5512418588010086e-06,
+      "loss": 0.5894,
+      "step": 719
+    },
+    {
+      "epoch": 2.302771855010661,
+      "grad_norm": 0.4972054054110685,
+      "learning_rate": 1.5377584265306222e-06,
+      "loss": 0.5282,
+      "step": 720
+    },
+    {
+      "epoch": 2.3059701492537314,
+      "grad_norm": 0.4727141579634826,
+      "learning_rate": 1.5243231926384744e-06,
+      "loss": 0.6045,
+      "step": 721
+    },
+    {
+      "epoch": 2.3091684434968016,
+      "grad_norm": 0.46328977747110367,
+      "learning_rate": 1.510936344158448e-06,
+      "loss": 0.6181,
+      "step": 722
+    },
+    {
+      "epoch": 2.3123667377398722,
+      "grad_norm": 0.46878962646890604,
+      "learning_rate": 1.4975980674508472e-06,
+      "loss": 0.5932,
+      "step": 723
+    },
+    {
+      "epoch": 2.3155650319829424,
+      "grad_norm": 0.46930491695796345,
+      "learning_rate": 1.484308548199796e-06,
+      "loss": 0.5943,
+      "step": 724
+    },
+    {
+      "epoch": 2.318763326226013,
+      "grad_norm": 0.4819028591025632,
+      "learning_rate": 1.4710679714106635e-06,
+      "loss": 0.6015,
+      "step": 725
+    },
+    {
+      "epoch": 2.321961620469083,
+      "grad_norm": 0.4490778160016328,
+      "learning_rate": 1.4578765214074842e-06,
+      "loss": 0.5885,
+      "step": 726
+    },
+    {
+      "epoch": 2.3251599147121533,
+      "grad_norm": 0.439589868568974,
+      "learning_rate": 1.444734381830386e-06,
+      "loss": 0.5752,
+      "step": 727
+    },
+    {
+      "epoch": 2.328358208955224,
+      "grad_norm": 0.4595533358265758,
+      "learning_rate": 1.4316417356330441e-06,
+      "loss": 0.5933,
+      "step": 728
+    },
+    {
+      "epoch": 2.331556503198294,
+      "grad_norm": 0.46808692540880764,
+      "learning_rate": 1.4185987650801286e-06,
+      "loss": 0.5593,
+      "step": 729
+    },
+    {
+      "epoch": 2.3347547974413647,
+      "grad_norm": 0.4422620648261011,
+      "learning_rate": 1.4056056517447637e-06,
+      "loss": 0.5725,
+      "step": 730
+    },
+    {
+      "epoch": 2.337953091684435,
+      "grad_norm": 0.47954403115117483,
+      "learning_rate": 1.392662576506007e-06,
+      "loss": 0.5761,
+      "step": 731
+    },
+    {
+      "epoch": 2.3411513859275055,
+      "grad_norm": 0.4546489872200875,
+      "learning_rate": 1.3797697195463278e-06,
+      "loss": 0.5871,
+      "step": 732
+    },
+    {
+      "epoch": 2.3443496801705757,
+      "grad_norm": 0.4500895428397602,
+      "learning_rate": 1.3669272603491002e-06,
+      "loss": 0.5644,
+      "step": 733
+    },
+    {
+      "epoch": 2.3475479744136463,
+      "grad_norm": 0.4805002284443759,
+      "learning_rate": 1.3541353776961035e-06,
+      "loss": 0.5714,
+      "step": 734
+    },
+    {
+      "epoch": 2.3507462686567164,
+      "grad_norm": 0.450604856228356,
+      "learning_rate": 1.3413942496650301e-06,
+      "loss": 0.5569,
+      "step": 735
+    },
+    {
+      "epoch": 2.3539445628997866,
+      "grad_norm": 0.44449130255853053,
+      "learning_rate": 1.3287040536270135e-06,
+      "loss": 0.5673,
+      "step": 736
+    },
+    {
+      "epoch": 2.357142857142857,
+      "grad_norm": 0.4311651330481667,
+      "learning_rate": 1.3160649662441532e-06,
+      "loss": 0.5649,
+      "step": 737
+    },
+    {
+      "epoch": 2.3603411513859274,
+      "grad_norm": 0.5481566773492595,
+      "learning_rate": 1.30347716346706e-06,
+      "loss": 0.5805,
+      "step": 738
+    },
+    {
+      "epoch": 2.363539445628998,
+      "grad_norm": 0.4694122862360678,
+      "learning_rate": 1.290940820532403e-06,
+      "loss": 0.5862,
+      "step": 739
+    },
+    {
+      "epoch": 2.366737739872068,
+      "grad_norm": 0.4808027009258574,
+      "learning_rate": 1.2784561119604683e-06,
+      "loss": 0.5844,
+      "step": 740
+    },
+    {
+      "epoch": 2.3699360341151388,
+      "grad_norm": 0.4741021908385686,
+      "learning_rate": 1.266023211552736e-06,
+      "loss": 0.5927,
+      "step": 741
+    },
+    {
+      "epoch": 2.373134328358209,
+      "grad_norm": 0.5070129440283249,
+      "learning_rate": 1.2536422923894565e-06,
+      "loss": 0.5989,
+      "step": 742
+    },
+    {
+      "epoch": 2.3763326226012795,
+      "grad_norm": 0.49982577127543343,
+      "learning_rate": 1.2413135268272403e-06,
+      "loss": 0.6086,
+      "step": 743
+    },
+    {
+      "epoch": 2.3795309168443497,
+      "grad_norm": 0.5171582506065449,
+      "learning_rate": 1.2290370864966623e-06,
+      "loss": 0.5935,
+      "step": 744
+    },
+    {
+      "epoch": 2.38272921108742,
+      "grad_norm": 0.5058849568215932,
+      "learning_rate": 1.2168131422998653e-06,
+      "loss": 0.5788,
+      "step": 745
+    },
+    {
+      "epoch": 2.3859275053304905,
+      "grad_norm": 0.49576030363447515,
+      "learning_rate": 1.2046418644081904e-06,
+      "loss": 0.6084,
+      "step": 746
+    },
+    {
+      "epoch": 2.3891257995735606,
+      "grad_norm": 0.4737787057177936,
+      "learning_rate": 1.192523422259802e-06,
+      "loss": 0.5667,
+      "step": 747
+    },
+    {
+      "epoch": 2.3923240938166312,
+      "grad_norm": 0.4628087834904147,
+      "learning_rate": 1.1804579845573288e-06,
+      "loss": 0.6369,
+      "step": 748
+    },
+    {
+      "epoch": 2.3955223880597014,
+      "grad_norm": 0.44818958645228585,
+      "learning_rate": 1.1684457192655207e-06,
+      "loss": 0.5785,
+      "step": 749
+    },
+    {
+      "epoch": 2.398720682302772,
+      "grad_norm": 0.44830220311328195,
+      "learning_rate": 1.156486793608899e-06,
+      "loss": 0.5932,
+      "step": 750
+    },
+    {
+      "epoch": 2.401918976545842,
+      "grad_norm": 0.5032388277593316,
+      "learning_rate": 1.144581374069444e-06,
+      "loss": 0.6266,
+      "step": 751
+    },
+    {
+      "epoch": 2.405117270788913,
+      "grad_norm": 0.5016006802472114,
+      "learning_rate": 1.1327296263842653e-06,
+      "loss": 0.5956,
+      "step": 752
+    },
+    {
+      "epoch": 2.408315565031983,
+      "grad_norm": 0.4643916700878555,
+      "learning_rate": 1.120931715543299e-06,
+      "loss": 0.5694,
+      "step": 753
+    },
+    {
+      "epoch": 2.411513859275053,
+      "grad_norm": 0.44920509828052607,
+      "learning_rate": 1.1091878057870137e-06,
+      "loss": 0.6153,
+      "step": 754
+    },
+    {
+      "epoch": 2.4147121535181237,
+      "grad_norm": 0.45686584478530823,
+      "learning_rate": 1.0974980606041152e-06,
+      "loss": 0.5805,
+      "step": 755
+    },
+    {
+      "epoch": 2.417910447761194,
+      "grad_norm": 0.46553788908049154,
+      "learning_rate": 1.0858626427292796e-06,
+      "loss": 0.599,
+      "step": 756
+    },
+    {
+      "epoch": 2.4211087420042645,
+      "grad_norm": 0.5153605958479428,
+      "learning_rate": 1.074281714140884e-06,
+      "loss": 0.6158,
+      "step": 757
+    },
+    {
+      "epoch": 2.4243070362473347,
+      "grad_norm": 0.5316883972134206,
+      "learning_rate": 1.0627554360587533e-06,
+      "loss": 0.6284,
+      "step": 758
+    },
+    {
+      "epoch": 2.4275053304904053,
+      "grad_norm": 0.4612691902886173,
+      "learning_rate": 1.0512839689419124e-06,
+      "loss": 0.629,
+      "step": 759
+    },
+    {
+      "epoch": 2.4307036247334755,
+      "grad_norm": 0.45681038140195945,
+      "learning_rate": 1.0398674724863584e-06,
+      "loss": 0.5693,
+      "step": 760
+    },
+    {
+      "epoch": 2.433901918976546,
+      "grad_norm": 0.48533257603226454,
+      "learning_rate": 1.0285061056228273e-06,
+      "loss": 0.5905,
+      "step": 761
+    },
+    {
+      "epoch": 2.4371002132196162,
+      "grad_norm": 0.4746224830922052,
+      "learning_rate": 1.0172000265145938e-06,
+      "loss": 0.6055,
+      "step": 762
+    },
+    {
+      "epoch": 2.4402985074626864,
+      "grad_norm": 0.5086026458920127,
+      "learning_rate": 1.0059493925552604e-06,
+      "loss": 0.5744,
+      "step": 763
+    },
+    {
+      "epoch": 2.443496801705757,
+      "grad_norm": 0.4514225973264305,
+      "learning_rate": 9.947543603665711e-07,
+      "loss": 0.5964,
+      "step": 764
+    },
+    {
+      "epoch": 2.446695095948827,
+      "grad_norm": 0.475954682216827,
+      "learning_rate": 9.836150857962296e-07,
+      "loss": 0.6195,
+      "step": 765
+    },
+    {
+      "epoch": 2.449893390191898,
+      "grad_norm": 0.516019394227879,
+      "learning_rate": 9.72531723915726e-07,
+      "loss": 0.5606,
+      "step": 766
+    },
+    {
+      "epoch": 2.453091684434968,
+      "grad_norm": 0.4863472899308536,
+      "learning_rate": 9.615044290181863e-07,
+      "loss": 0.5711,
+      "step": 767
+    },
+    {
+      "epoch": 2.4562899786780386,
+      "grad_norm": 0.433161350310435,
+      "learning_rate": 9.505333546162171e-07,
+      "loss": 0.61,
+      "step": 768
+    },
+    {
+      "epoch": 2.4594882729211087,
+      "grad_norm": 0.4744290062548913,
+      "learning_rate": 9.396186534397711e-07,
+      "loss": 0.5657,
+      "step": 769
+    },
+    {
+      "epoch": 2.4626865671641793,
+      "grad_norm": 0.4438296651802267,
+      "learning_rate": 9.287604774340236e-07,
+      "loss": 0.6051,
+      "step": 770
+    },
+    {
+      "epoch": 2.4658848614072495,
+      "grad_norm": 0.4527860416616376,
+      "learning_rate": 9.179589777572496e-07,
+      "loss": 0.5531,
+      "step": 771
+    },
+    {
+      "epoch": 2.4690831556503197,
+      "grad_norm": 0.4863675854140087,
+      "learning_rate": 9.07214304778729e-07,
+      "loss": 0.5948,
+      "step": 772
+    },
+    {
+      "epoch": 2.4722814498933903,
+      "grad_norm": 0.45267687004556684,
+      "learning_rate": 8.965266080766471e-07,
+      "loss": 0.5954,
+      "step": 773
+    },
+    {
+      "epoch": 2.4754797441364604,
+      "grad_norm": 0.4577039509757653,
+      "learning_rate": 8.858960364360142e-07,
+      "loss": 0.5749,
+      "step": 774
+    },
+    {
+      "epoch": 2.478678038379531,
+      "grad_norm": 0.42098442226808497,
+      "learning_rate": 8.753227378465956e-07,
+      "loss": 0.5503,
+      "step": 775
+    },
+    {
+      "epoch": 2.481876332622601,
+      "grad_norm": 0.4593978666976355,
+      "learning_rate": 8.648068595008458e-07,
+      "loss": 0.6078,
+      "step": 776
+    },
+    {
+      "epoch": 2.485074626865672,
+      "grad_norm": 0.4859190992732641,
+      "learning_rate": 8.543485477918672e-07,
+      "loss": 0.5579,
+      "step": 777
+    },
+    {
+      "epoch": 2.488272921108742,
+      "grad_norm": 0.43344193599128084,
+      "learning_rate": 8.439479483113683e-07,
+      "loss": 0.578,
+      "step": 778
+    },
+    {
+      "epoch": 2.4914712153518126,
+      "grad_norm": 0.46299605001742955,
+      "learning_rate": 8.336052058476374e-07,
+      "loss": 0.5936,
+      "step": 779
+    },
+    {
+      "epoch": 2.4946695095948828,
+      "grad_norm": 0.48867807045497547,
+      "learning_rate": 8.233204643835235e-07,
+      "loss": 0.5904,
+      "step": 780
+    },
+    {
+      "epoch": 2.497867803837953,
+      "grad_norm": 0.49771350857070334,
+      "learning_rate": 8.130938670944377e-07,
+      "loss": 0.6007,
+      "step": 781
+    },
+    {
+      "epoch": 2.5010660980810235,
+      "grad_norm": 0.4577972103614504,
+      "learning_rate": 8.029255563463589e-07,
+      "loss": 0.5895,
+      "step": 782
+    },
+    {
+      "epoch": 2.5042643923240937,
+      "grad_norm": 0.47989050108275627,
+      "learning_rate": 7.928156736938458e-07,
+      "loss": 0.5949,
+      "step": 783
+    },
+    {
+      "epoch": 2.5074626865671643,
+      "grad_norm": 0.4371318457371128,
+      "learning_rate": 7.827643598780748e-07,
+      "loss": 0.5528,
+      "step": 784
+    },
+    {
+      "epoch": 2.5106609808102345,
+      "grad_norm": 0.46786863175003857,
+      "learning_rate": 7.72771754824877e-07,
+      "loss": 0.6073,
+      "step": 785
+    },
+    {
+      "epoch": 2.5138592750533046,
+      "grad_norm": 0.44519658517690325,
+      "learning_rate": 7.628379976427868e-07,
+      "loss": 0.6135,
+      "step": 786
+    },
+    {
+      "epoch": 2.5170575692963753,
+      "grad_norm": 0.44205578681704094,
+      "learning_rate": 7.529632266211112e-07,
+      "loss": 0.5942,
+      "step": 787
+    },
+    {
+      "epoch": 2.520255863539446,
+      "grad_norm": 0.4875271600407386,
+      "learning_rate": 7.431475792280018e-07,
+      "loss": 0.5807,
+      "step": 788
+    },
+    {
+      "epoch": 2.523454157782516,
+      "grad_norm": 0.4710353166878217,
+      "learning_rate": 7.333911921085418e-07,
+      "loss": 0.6056,
+      "step": 789
+    },
+    {
+      "epoch": 2.526652452025586,
+      "grad_norm": 0.48054394292285285,
+      "learning_rate": 7.23694201082843e-07,
+      "loss": 0.5591,
+      "step": 790
+    },
+    {
+      "epoch": 2.529850746268657,
+      "grad_norm": 0.4868072363712072,
+      "learning_rate": 7.140567411441529e-07,
+      "loss": 0.5832,
+      "step": 791
+    },
+    {
+      "epoch": 2.533049040511727,
+      "grad_norm": 0.4412560252403492,
+      "learning_rate": 7.044789464569817e-07,
+      "loss": 0.5937,
+      "step": 792
+    },
+    {
+      "epoch": 2.5362473347547976,
+      "grad_norm": 0.4511415013667377,
+      "learning_rate": 6.94960950355229e-07,
+      "loss": 0.6048,
+      "step": 793
+    },
+    {
+      "epoch": 2.5394456289978677,
+      "grad_norm": 0.4454888744137283,
+      "learning_rate": 6.855028853403295e-07,
+      "loss": 0.6137,
+      "step": 794
+    },
+    {
+      "epoch": 2.542643923240938,
+      "grad_norm": 0.5107375323074698,
+      "learning_rate": 6.761048830794098e-07,
+      "loss": 0.6191,
+      "step": 795
+    },
+    {
+      "epoch": 2.5458422174840085,
+      "grad_norm": 0.46547987418755626,
+      "learning_rate": 6.667670744034498e-07,
+      "loss": 0.5814,
+      "step": 796
+    },
+    {
+      "epoch": 2.549040511727079,
+      "grad_norm": 0.4703517207855011,
+      "learning_rate": 6.574895893054711e-07,
+      "loss": 0.5778,
+      "step": 797
+    },
+    {
+      "epoch": 2.5522388059701493,
+      "grad_norm": 0.4746163195915809,
+      "learning_rate": 6.482725569387171e-07,
+      "loss": 0.5706,
+      "step": 798
+    },
+    {
+      "epoch": 2.5554371002132195,
+      "grad_norm": 0.470907057244819,
+      "learning_rate": 6.391161056148637e-07,
+      "loss": 0.5868,
+      "step": 799
+    },
+    {
+      "epoch": 2.55863539445629,
+      "grad_norm": 0.4623234041016843,
+      "learning_rate": 6.300203628022272e-07,
+      "loss": 0.5983,
+      "step": 800
+    },
+    {
+      "epoch": 2.5618336886993602,
+      "grad_norm": 0.42562483606080814,
+      "learning_rate": 6.209854551239902e-07,
+      "loss": 0.5927,
+      "step": 801
+    },
+    {
+      "epoch": 2.565031982942431,
+      "grad_norm": 0.5128197818499242,
+      "learning_rate": 6.120115083564432e-07,
+      "loss": 0.5741,
+      "step": 802
+    },
+    {
+      "epoch": 2.568230277185501,
+      "grad_norm": 0.47049712813520317,
+      "learning_rate": 6.030986474272288e-07,
+      "loss": 0.584,
+      "step": 803
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.44693735376914034,
+      "learning_rate": 5.942469964136055e-07,
+      "loss": 0.5903,
+      "step": 804
+    },
+    {
+      "epoch": 2.574626865671642,
+      "grad_norm": 0.42861119739729087,
+      "learning_rate": 5.854566785407212e-07,
+      "loss": 0.597,
+      "step": 805
+    },
+    {
+      "epoch": 2.5778251599147124,
+      "grad_norm": 0.44886124383973774,
+      "learning_rate": 5.767278161798912e-07,
+      "loss": 0.5721,
+      "step": 806
+    },
+    {
+      "epoch": 2.5810234541577826,
+      "grad_norm": 0.4202452468317384,
+      "learning_rate": 5.680605308469045e-07,
+      "loss": 0.5828,
+      "step": 807
+    },
+    {
+      "epoch": 2.5842217484008527,
+      "grad_norm": 0.4235020132956652,
+      "learning_rate": 5.594549432003244e-07,
+      "loss": 0.5846,
+      "step": 808
+    },
+    {
+      "epoch": 2.5874200426439233,
+      "grad_norm": 0.4603963689104671,
+      "learning_rate": 5.509111730398125e-07,
+      "loss": 0.5753,
+      "step": 809
+    },
+    {
+      "epoch": 2.5906183368869935,
+      "grad_norm": 0.47640520792175567,
+      "learning_rate": 5.42429339304461e-07,
+      "loss": 0.5813,
+      "step": 810
+    },
+    {
+      "epoch": 2.593816631130064,
+      "grad_norm": 0.4964363064529672,
+      "learning_rate": 5.340095600711343e-07,
+      "loss": 0.5483,
+      "step": 811
+    },
+    {
+      "epoch": 2.5970149253731343,
+      "grad_norm": 0.4188032394181886,
+      "learning_rate": 5.256519525528254e-07,
+      "loss": 0.6258,
+      "step": 812
+    },
+    {
+      "epoch": 2.6002132196162044,
+      "grad_norm": 0.4418308069828843,
+      "learning_rate": 5.173566330970286e-07,
+      "loss": 0.5858,
+      "step": 813
+    },
+    {
+      "epoch": 2.603411513859275,
+      "grad_norm": 0.4729321718897582,
+      "learning_rate": 5.091237171841173e-07,
+      "loss": 0.56,
+      "step": 814
+    },
+    {
+      "epoch": 2.6066098081023457,
+      "grad_norm": 0.4393226839327768,
+      "learning_rate": 5.009533194257332e-07,
+      "loss": 0.5494,
+      "step": 815
+    },
+    {
+      "epoch": 2.609808102345416,
+      "grad_norm": 0.44734542269470695,
+      "learning_rate": 4.92845553563196e-07,
+      "loss": 0.626,
+      "step": 816
+    },
+    {
+      "epoch": 2.613006396588486,
+      "grad_norm": 0.4501393477711489,
+      "learning_rate": 4.848005324659144e-07,
+      "loss": 0.5966,
+      "step": 817
+    },
+    {
+      "epoch": 2.6162046908315566,
+      "grad_norm": 0.4524819111331228,
+      "learning_rate": 4.768183681298211e-07,
+      "loss": 0.5816,
+      "step": 818
+    },
+    {
+      "epoch": 2.6194029850746268,
+      "grad_norm": 0.44348410421643397,
+      "learning_rate": 4.6889917167580903e-07,
+      "loss": 0.5752,
+      "step": 819
+    },
+    {
+      "epoch": 2.6226012793176974,
+      "grad_norm": 0.46496313577314546,
+      "learning_rate": 4.6104305334818577e-07,
+      "loss": 0.5894,
+      "step": 820
+    },
+    {
+      "epoch": 2.6257995735607675,
+      "grad_norm": 0.45530128557907246,
+      "learning_rate": 4.532501225131408e-07,
+      "loss": 0.628,
+      "step": 821
+    },
+    {
+      "epoch": 2.6289978678038377,
+      "grad_norm": 0.4557231787890459,
+      "learning_rate": 4.455204876572172e-07,
+      "loss": 0.5963,
+      "step": 822
+    },
+    {
+      "epoch": 2.6321961620469083,
+      "grad_norm": 0.413730075405882,
+      "learning_rate": 4.3785425638580847e-07,
+      "loss": 0.5895,
+      "step": 823
+    },
+    {
+      "epoch": 2.635394456289979,
+      "grad_norm": 0.46478458824267105,
+      "learning_rate": 4.3025153542165744e-07,
+      "loss": 0.5628,
+      "step": 824
+    },
+    {
+      "epoch": 2.638592750533049,
+      "grad_norm": 0.4660432242750011,
+      "learning_rate": 4.2271243060336976e-07,
+      "loss": 0.5843,
+      "step": 825
+    },
+    {
+      "epoch": 2.6417910447761193,
+      "grad_norm": 0.45857321057691036,
+      "learning_rate": 4.1523704688394176e-07,
+      "loss": 0.569,
+      "step": 826
+    },
+    {
+      "epoch": 2.64498933901919,
+      "grad_norm": 0.454757225375927,
+      "learning_rate": 4.0782548832929646e-07,
+      "loss": 0.5631,
+      "step": 827
+    },
+    {
+      "epoch": 2.64818763326226,
+      "grad_norm": 0.4219569542186656,
+      "learning_rate": 4.0047785811684116e-07,
+      "loss": 0.5923,
+      "step": 828
+    },
+    {
+      "epoch": 2.6513859275053306,
+      "grad_norm": 0.407295453366224,
+      "learning_rate": 3.931942585340243e-07,
+      "loss": 0.6345,
+      "step": 829
+    },
+    {
+      "epoch": 2.654584221748401,
+      "grad_norm": 0.4422205956943115,
+      "learning_rate": 3.8597479097691626e-07,
+      "loss": 0.6016,
+      "step": 830
+    },
+    {
+      "epoch": 2.657782515991471,
+      "grad_norm": 0.4322372658598502,
+      "learning_rate": 3.788195559487956e-07,
+      "loss": 0.5717,
+      "step": 831
+    },
+    {
+      "epoch": 2.6609808102345416,
+      "grad_norm": 0.446074402964079,
+      "learning_rate": 3.717286530587483e-07,
+      "loss": 0.5933,
+      "step": 832
+    },
+    {
+      "epoch": 2.664179104477612,
+      "grad_norm": 0.44149384007245823,
+      "learning_rate": 3.6470218102028607e-07,
+      "loss": 0.5935,
+      "step": 833
+    },
+    {
+      "epoch": 2.6673773987206824,
+      "grad_norm": 0.48606195110523104,
+      "learning_rate": 3.577402376499672e-07,
+      "loss": 0.5632,
+      "step": 834
+    },
+    {
+      "epoch": 2.6705756929637525,
+      "grad_norm": 0.45172402428831737,
+      "learning_rate": 3.508429198660379e-07,
+      "loss": 0.6271,
+      "step": 835
+    },
+    {
+      "epoch": 2.673773987206823,
+      "grad_norm": 0.4783802058190356,
+      "learning_rate": 3.440103236870823e-07,
+      "loss": 0.5614,
+      "step": 836
+    },
+    {
+      "epoch": 2.6769722814498933,
+      "grad_norm": 0.4125951168982884,
+      "learning_rate": 3.372425442306837e-07,
+      "loss": 0.6403,
+      "step": 837
+    },
+    {
+      "epoch": 2.680170575692964,
+      "grad_norm": 0.45804236331506576,
+      "learning_rate": 3.3053967571210375e-07,
+      "loss": 0.6235,
+      "step": 838
+    },
+    {
+      "epoch": 2.683368869936034,
+      "grad_norm": 0.4481909177996369,
+      "learning_rate": 3.2390181144296815e-07,
+      "loss": 0.6079,
+      "step": 839
+    },
+    {
+      "epoch": 2.6865671641791042,
+      "grad_norm": 0.44229637332098864,
+      "learning_rate": 3.1732904382996975e-07,
+      "loss": 0.6046,
+      "step": 840
+    },
+    {
+      "epoch": 2.689765458422175,
+      "grad_norm": 0.456497771982713,
+      "learning_rate": 3.108214643735813e-07,
+      "loss": 0.6003,
+      "step": 841
+    },
+    {
+      "epoch": 2.6929637526652455,
+      "grad_norm": 0.43916739323489756,
+      "learning_rate": 3.04379163666782e-07,
+      "loss": 0.596,
+      "step": 842
+    },
+    {
+      "epoch": 2.6961620469083156,
+      "grad_norm": 0.4833880509304977,
+      "learning_rate": 2.98002231393793e-07,
+      "loss": 0.602,
+      "step": 843
+    },
+    {
+      "epoch": 2.699360341151386,
+      "grad_norm": 0.44190751175823845,
+      "learning_rate": 2.916907563288357e-07,
+      "loss": 0.6089,
+      "step": 844
+    },
+    {
+      "epoch": 2.7025586353944564,
+      "grad_norm": 0.4409073668802216,
+      "learning_rate": 2.854448263348891e-07,
+      "loss": 0.6085,
+      "step": 845
+    },
+    {
+      "epoch": 2.7057569296375266,
+      "grad_norm": 0.44376266623437516,
+      "learning_rate": 2.792645283624712e-07,
+      "loss": 0.5517,
+      "step": 846
+    },
+    {
+      "epoch": 2.708955223880597,
+      "grad_norm": 0.44030030505529505,
+      "learning_rate": 2.7314994844842623e-07,
+      "loss": 0.5704,
+      "step": 847
+    },
+    {
+      "epoch": 2.7121535181236673,
+      "grad_norm": 0.4811909570823317,
+      "learning_rate": 2.671011717147276e-07,
+      "loss": 0.5887,
+      "step": 848
+    },
+    {
+      "epoch": 2.7153518123667375,
+      "grad_norm": 0.46435424416208204,
+      "learning_rate": 2.611182823672931e-07,
+      "loss": 0.5753,
+      "step": 849
+    },
+    {
+      "epoch": 2.718550106609808,
+      "grad_norm": 0.45062900256380783,
+      "learning_rate": 2.5520136369481194e-07,
+      "loss": 0.5956,
+      "step": 850
+    },
+    {
+      "epoch": 2.7217484008528787,
+      "grad_norm": 0.4373848417355698,
+      "learning_rate": 2.493504980675865e-07,
+      "loss": 0.5725,
+      "step": 851
+    },
+    {
+      "epoch": 2.724946695095949,
+      "grad_norm": 0.41456459520431993,
+      "learning_rate": 2.4356576693638555e-07,
+      "loss": 0.5989,
+      "step": 852
+    },
+    {
+      "epoch": 2.728144989339019,
+      "grad_norm": 0.45641172841307615,
+      "learning_rate": 2.3784725083130678e-07,
+      "loss": 0.5394,
+      "step": 853
+    },
+    {
+      "epoch": 2.7313432835820897,
+      "grad_norm": 0.41453418124512204,
+      "learning_rate": 2.3219502936066228e-07,
+      "loss": 0.5822,
+      "step": 854
+    },
+    {
+      "epoch": 2.73454157782516,
+      "grad_norm": 0.47178126843622886,
+      "learning_rate": 2.266091812098642e-07,
+      "loss": 0.6153,
+      "step": 855
+    },
+    {
+      "epoch": 2.7377398720682304,
+      "grad_norm": 0.415357561340448,
+      "learning_rate": 2.210897841403331e-07,
+      "loss": 0.6319,
+      "step": 856
+    },
+    {
+      "epoch": 2.7409381663113006,
+      "grad_norm": 0.429713391850993,
+      "learning_rate": 2.1563691498841465e-07,
+      "loss": 0.6076,
+      "step": 857
+    },
+    {
+      "epoch": 2.7441364605543708,
+      "grad_norm": 0.4364411910317568,
+      "learning_rate": 2.1025064966430697e-07,
+      "loss": 0.5809,
+      "step": 858
+    },
+    {
+      "epoch": 2.7473347547974414,
+      "grad_norm": 0.4438187825256016,
+      "learning_rate": 2.0493106315100987e-07,
+      "loss": 0.6077,
+      "step": 859
+    },
+    {
+      "epoch": 2.750533049040512,
+      "grad_norm": 0.4396721972544317,
+      "learning_rate": 1.9967822950327453e-07,
+      "loss": 0.5649,
+      "step": 860
+    },
+    {
+      "epoch": 2.753731343283582,
+      "grad_norm": 0.4087723994643433,
+      "learning_rate": 1.944922218465778e-07,
+      "loss": 0.5983,
+      "step": 861
+    },
+    {
+      "epoch": 2.7569296375266523,
+      "grad_norm": 0.45839789417365656,
+      "learning_rate": 1.8937311237610168e-07,
+      "loss": 0.5666,
+      "step": 862
+    },
+    {
+      "epoch": 2.760127931769723,
+      "grad_norm": 0.4623951503889097,
+      "learning_rate": 1.8432097235572655e-07,
+      "loss": 0.5814,
+      "step": 863
+    },
+    {
+      "epoch": 2.763326226012793,
+      "grad_norm": 0.42397071107319473,
+      "learning_rate": 1.793358721170435e-07,
+      "loss": 0.6227,
+      "step": 864
+    },
+    {
+      "epoch": 2.7665245202558637,
+      "grad_norm": 0.46632194080900585,
+      "learning_rate": 1.7441788105837133e-07,
+      "loss": 0.5687,
+      "step": 865
+    },
+    {
+      "epoch": 2.769722814498934,
+      "grad_norm": 0.42383172618691023,
+      "learning_rate": 1.6956706764379438e-07,
+      "loss": 0.5571,
+      "step": 866
+    },
+    {
+      "epoch": 2.772921108742004,
+      "grad_norm": 0.42444537413289984,
+      "learning_rate": 1.6478349940220294e-07,
+      "loss": 0.5672,
+      "step": 867
+    },
+    {
+      "epoch": 2.7761194029850746,
+      "grad_norm": 0.442713023638936,
+      "learning_rate": 1.6006724292636166e-07,
+      "loss": 0.6251,
+      "step": 868
+    },
+    {
+      "epoch": 2.7793176972281453,
+      "grad_norm": 0.44882872466351875,
+      "learning_rate": 1.5541836387197528e-07,
+      "loss": 0.6309,
+      "step": 869
+    },
+    {
+      "epoch": 2.7825159914712154,
+      "grad_norm": 0.4426652587385554,
+      "learning_rate": 1.508369269567783e-07,
+      "loss": 0.6038,
+      "step": 870
+    },
+    {
+      "epoch": 2.7857142857142856,
+      "grad_norm": 0.43159736313225455,
+      "learning_rate": 1.4632299595963294e-07,
+      "loss": 0.6136,
+      "step": 871
+    },
+    {
+      "epoch": 2.788912579957356,
+      "grad_norm": 0.4328150604393648,
+      "learning_rate": 1.418766337196431e-07,
+      "loss": 0.5821,
+      "step": 872
+    },
+    {
+      "epoch": 2.7921108742004264,
+      "grad_norm": 0.4510225598844259,
+      "learning_rate": 1.374979021352757e-07,
+      "loss": 0.544,
+      "step": 873
+    },
+    {
+      "epoch": 2.795309168443497,
+      "grad_norm": 0.42414299491852264,
+      "learning_rate": 1.3318686216350241e-07,
+      "loss": 0.5917,
+      "step": 874
+    },
+    {
+      "epoch": 2.798507462686567,
+      "grad_norm": 0.4491433631888564,
+      "learning_rate": 1.2894357381894984e-07,
+      "loss": 0.5709,
+      "step": 875
+    },
+    {
+      "epoch": 2.8017057569296373,
+      "grad_norm": 0.435024650739532,
+      "learning_rate": 1.2476809617306408e-07,
+      "loss": 0.5768,
+      "step": 876
+    },
+    {
+      "epoch": 2.804904051172708,
+      "grad_norm": 0.4376084724071413,
+      "learning_rate": 1.206604873532885e-07,
+      "loss": 0.5747,
+      "step": 877
+    },
+    {
+      "epoch": 2.8081023454157785,
+      "grad_norm": 0.4430099826070007,
+      "learning_rate": 1.166208045422551e-07,
+      "loss": 0.5671,
+      "step": 878
+    },
+    {
+      "epoch": 2.8113006396588487,
+      "grad_norm": 0.45821945748201703,
+      "learning_rate": 1.1264910397698614e-07,
+      "loss": 0.5801,
+      "step": 879
+    },
+    {
+      "epoch": 2.814498933901919,
+      "grad_norm": 0.42968896616005114,
+      "learning_rate": 1.0874544094811424e-07,
+      "loss": 0.609,
+      "step": 880
+    },
+    {
+      "epoch": 2.8176972281449895,
+      "grad_norm": 0.4392112009989005,
+      "learning_rate": 1.0490986979911189e-07,
+      "loss": 0.6004,
+      "step": 881
+    },
+    {
+      "epoch": 2.8208955223880596,
+      "grad_norm": 0.4678179782402967,
+      "learning_rate": 1.0114244392553318e-07,
+      "loss": 0.5799,
+      "step": 882
+    },
+    {
+      "epoch": 2.8240938166311302,
+      "grad_norm": 0.4649676388307849,
+      "learning_rate": 9.744321577427218e-08,
+      "loss": 0.5694,
+      "step": 883
+    },
+    {
+      "epoch": 2.8272921108742004,
+      "grad_norm": 0.4410645863970879,
+      "learning_rate": 9.381223684283291e-08,
+      "loss": 0.5792,
+      "step": 884
+    },
+    {
+      "epoch": 2.8304904051172706,
+      "grad_norm": 0.41344134456514686,
+      "learning_rate": 9.024955767861054e-08,
+      "loss": 0.6179,
+      "step": 885
+    },
+    {
+      "epoch": 2.833688699360341,
+      "grad_norm": 0.45207918900612454,
+      "learning_rate": 8.675522787819023e-08,
+      "loss": 0.5765,
+      "step": 886
+    },
+    {
+      "epoch": 2.836886993603412,
+      "grad_norm": 0.4815854246181784,
+      "learning_rate": 8.332929608665553e-08,
+      "loss": 0.5622,
+      "step": 887
+    },
+    {
+      "epoch": 2.840085287846482,
+      "grad_norm": 0.4104529223314507,
+      "learning_rate": 7.997180999691101e-08,
+      "loss": 0.6061,
+      "step": 888
+    },
+    {
+      "epoch": 2.843283582089552,
+      "grad_norm": 0.43863504264752207,
+      "learning_rate": 7.668281634901686e-08,
+      "loss": 0.6205,
+      "step": 889
+    },
+    {
+      "epoch": 2.8464818763326227,
+      "grad_norm": 0.42107787916551426,
+      "learning_rate": 7.346236092954318e-08,
+      "loss": 0.6112,
+      "step": 890
+    },
+    {
+      "epoch": 2.849680170575693,
+      "grad_norm": 0.45281531695898647,
+      "learning_rate": 7.031048857092604e-08,
+      "loss": 0.5897,
+      "step": 891
+    },
+    {
+      "epoch": 2.8528784648187635,
+      "grad_norm": 0.4461369389804025,
+      "learning_rate": 6.722724315084805e-08,
+      "loss": 0.5908,
+      "step": 892
+    },
+    {
+      "epoch": 2.8560767590618337,
+      "grad_norm": 0.4523838631460964,
+      "learning_rate": 6.421266759162659e-08,
+      "loss": 0.6059,
+      "step": 893
+    },
+    {
+      "epoch": 2.859275053304904,
+      "grad_norm": 0.4201076758194753,
+      "learning_rate": 6.12668038596137e-08,
+      "loss": 0.6017,
+      "step": 894
+    },
+    {
+      "epoch": 2.8624733475479744,
+      "grad_norm": 0.4576782992482849,
+      "learning_rate": 5.838969296461605e-08,
+      "loss": 0.5595,
+      "step": 895
+    },
+    {
+      "epoch": 2.8656716417910446,
+      "grad_norm": 0.452789058187118,
+      "learning_rate": 5.5581374959320366e-08,
+      "loss": 0.6069,
+      "step": 896
+    },
+    {
+      "epoch": 2.868869936034115,
+      "grad_norm": 0.4502806418223326,
+      "learning_rate": 5.2841888938738314e-08,
+      "loss": 0.595,
+      "step": 897
+    },
+    {
+      "epoch": 2.8720682302771854,
+      "grad_norm": 0.4291345074640387,
+      "learning_rate": 5.017127303966085e-08,
+      "loss": 0.5737,
+      "step": 898
+    },
+    {
+      "epoch": 2.875266524520256,
+      "grad_norm": 0.4275943368139046,
+      "learning_rate": 4.7569564440128055e-08,
+      "loss": 0.5802,
+      "step": 899
+    },
+    {
+      "epoch": 2.878464818763326,
+      "grad_norm": 0.4783054811845979,
+      "learning_rate": 4.50367993589107e-08,
+      "loss": 0.5523,
+      "step": 900
+    },
+    {
+      "epoch": 2.8816631130063968,
+      "grad_norm": 0.4207771911644098,
+      "learning_rate": 4.257301305500672e-08,
+      "loss": 0.6295,
+      "step": 901
+    },
+    {
+      "epoch": 2.884861407249467,
+      "grad_norm": 0.4709474401474473,
+      "learning_rate": 4.0178239827151077e-08,
+      "loss": 0.5824,
+      "step": 902
+    },
+    {
+      "epoch": 2.888059701492537,
+      "grad_norm": 0.47155045759162584,
+      "learning_rate": 3.785251301333726e-08,
+      "loss": 0.594,
+      "step": 903
+    },
+    {
+      "epoch": 2.8912579957356077,
+      "grad_norm": 0.4456059748610983,
+      "learning_rate": 3.559586499035206e-08,
+      "loss": 0.5662,
+      "step": 904
+    },
+    {
+      "epoch": 2.894456289978678,
+      "grad_norm": 0.4575365472232499,
+      "learning_rate": 3.340832717332765e-08,
+      "loss": 0.596,
+      "step": 905
+    },
+    {
+      "epoch": 2.8976545842217485,
+      "grad_norm": 0.45897611846221986,
+      "learning_rate": 3.128993001530245e-08,
+      "loss": 0.5467,
+      "step": 906
+    },
+    {
+      "epoch": 2.9008528784648187,
+      "grad_norm": 0.46086480691517584,
+      "learning_rate": 2.9240703006797044e-08,
+      "loss": 0.591,
+      "step": 907
+    },
+    {
+      "epoch": 2.9040511727078893,
+      "grad_norm": 0.4578137617838968,
+      "learning_rate": 2.7260674675404498e-08,
+      "loss": 0.5807,
+      "step": 908
+    },
+    {
+      "epoch": 2.9072494669509594,
+      "grad_norm": 0.4744681575485193,
+      "learning_rate": 2.5349872585392898e-08,
+      "loss": 0.5882,
+      "step": 909
+    },
+    {
+      "epoch": 2.91044776119403,
+      "grad_norm": 0.43778348787470217,
+      "learning_rate": 2.3508323337321225e-08,
+      "loss": 0.555,
+      "step": 910
+    },
+    {
+      "epoch": 2.9136460554371,
+      "grad_norm": 0.43656355895512294,
+      "learning_rate": 2.1736052567670195e-08,
+      "loss": 0.5777,
+      "step": 911
+    },
+    {
+      "epoch": 2.9168443496801704,
+      "grad_norm": 0.4664832575605358,
+      "learning_rate": 2.0033084948483104e-08,
+      "loss": 0.5442,
+      "step": 912
+    },
+    {
+      "epoch": 2.920042643923241,
+      "grad_norm": 0.438143790231801,
+      "learning_rate": 1.8399444187024995e-08,
+      "loss": 0.5745,
+      "step": 913
+    },
+    {
+      "epoch": 2.923240938166311,
+      "grad_norm": 0.42188051822462064,
+      "learning_rate": 1.6835153025451246e-08,
+      "loss": 0.6018,
+      "step": 914
+    },
+    {
+      "epoch": 2.9264392324093818,
+      "grad_norm": 0.41865517546309167,
+      "learning_rate": 1.534023324049061e-08,
+      "loss": 0.5891,
+      "step": 915
+    },
+    {
+      "epoch": 2.929637526652452,
+      "grad_norm": 0.41837714838468193,
+      "learning_rate": 1.3914705643143788e-08,
+      "loss": 0.6085,
+      "step": 916
+    },
+    {
+      "epoch": 2.9328358208955225,
+      "grad_norm": 0.465979632617469,
+      "learning_rate": 1.2558590078390886e-08,
+      "loss": 0.5841,
+      "step": 917
+    },
+    {
+      "epoch": 2.9360341151385927,
+      "grad_norm": 0.45654976609627784,
+      "learning_rate": 1.1271905424918294e-08,
+      "loss": 0.6035,
+      "step": 918
+    },
+    {
+      "epoch": 2.9392324093816633,
+      "grad_norm": 0.45629479117279054,
+      "learning_rate": 1.0054669594853905e-08,
+      "loss": 0.5988,
+      "step": 919
+    },
+    {
+      "epoch": 2.9424307036247335,
+      "grad_norm": 0.44536389990804115,
+      "learning_rate": 8.906899533517866e-09,
+      "loss": 0.6208,
+      "step": 920
+    },
+    {
+      "epoch": 2.9456289978678036,
+      "grad_norm": 0.4513455406012691,
+      "learning_rate": 7.828611219187765e-09,
+      "loss": 0.6199,
+      "step": 921
+    },
+    {
+      "epoch": 2.9488272921108742,
+      "grad_norm": 0.43738097436145257,
+      "learning_rate": 6.819819662874372e-09,
+      "loss": 0.5942,
+      "step": 922
+    },
+    {
+      "epoch": 2.9520255863539444,
+      "grad_norm": 0.45522291904606826,
+      "learning_rate": 5.88053890811513e-09,
+      "loss": 0.6035,
+      "step": 923
+    },
+    {
+      "epoch": 2.955223880597015,
+      "grad_norm": 0.4521268106702668,
+      "learning_rate": 5.0107820307770945e-09,
+      "loss": 0.6053,
+      "step": 924
+    },
+    {
+      "epoch": 2.958422174840085,
+      "grad_norm": 0.4313674140155561,
+      "learning_rate": 4.210561138873193e-09,
+      "loss": 0.58,
+      "step": 925
+    },
+    {
+      "epoch": 2.961620469083156,
+      "grad_norm": 0.4389755845527305,
+      "learning_rate": 3.4798873723984604e-09,
+      "loss": 0.6198,
+      "step": 926
+    },
+    {
+      "epoch": 2.964818763326226,
+      "grad_norm": 0.4358918838142038,
+      "learning_rate": 2.818770903170176e-09,
+      "loss": 0.6201,
+      "step": 927
+    },
+    {
+      "epoch": 2.9680170575692966,
+      "grad_norm": 0.446788281214219,
+      "learning_rate": 2.2272209346885233e-09,
+      "loss": 0.5517,
+      "step": 928
+    },
+    {
+      "epoch": 2.9712153518123667,
+      "grad_norm": 0.4627514418027653,
+      "learning_rate": 1.7052457020089175e-09,
+      "loss": 0.602,
+      "step": 929
+    },
+    {
+      "epoch": 2.974413646055437,
+      "grad_norm": 0.46487447660834663,
+      "learning_rate": 1.2528524716259872e-09,
+      "loss": 0.5523,
+      "step": 930
+    },
+    {
+      "epoch": 2.9776119402985075,
+      "grad_norm": 0.4291177591260577,
+      "learning_rate": 8.700475413719877e-10,
+      "loss": 0.6038,
+      "step": 931
+    },
+    {
+      "epoch": 2.9808102345415777,
+      "grad_norm": 0.4440151991197288,
+      "learning_rate": 5.568362403318706e-10,
+      "loss": 0.5722,
+      "step": 932
+    },
+    {
+      "epoch": 2.9840085287846483,
+      "grad_norm": 0.41546760769750524,
+      "learning_rate": 3.132229287666766e-10,
+      "loss": 0.6158,
+      "step": 933
+    },
+    {
+      "epoch": 2.9872068230277184,
+      "grad_norm": 0.47012397918132254,
+      "learning_rate": 1.3921099805302985e-10,
+      "loss": 0.612,
+      "step": 934
+    },
+    {
+      "epoch": 2.990405117270789,
+      "grad_norm": 0.4452241770705769,
+      "learning_rate": 3.480287063706289e-11,
+      "loss": 0.5822,
+      "step": 935
+    },
+    {
+      "epoch": 2.9936034115138592,
+      "grad_norm": 0.4503894897532014,
+      "learning_rate": 0.0,
+      "loss": 0.5513,
+      "step": 936
+    },
+    {
+      "epoch": 2.9936034115138592,
+      "step": 936,
+      "total_flos": 9.051650706620744e+17,
+      "train_loss": 0.0,
+      "train_runtime": 4.9645,
+      "train_samples_per_second": 18128.643,
+      "train_steps_per_second": 188.538
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 936,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 9.051650706620744e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}