diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6292 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 894,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0011185682326621924,
+      "grad_norm": 4.768556118011475,
+      "learning_rate": 4.99998456401763e-05,
+      "loss": 4.817,
+      "step": 1
+    },
+    {
+      "epoch": 0.0022371364653243847,
+      "grad_norm": 11.73538875579834,
+      "learning_rate": 4.9999382562611344e-05,
+      "loss": 4.6414,
+      "step": 2
+    },
+    {
+      "epoch": 0.003355704697986577,
+      "grad_norm": 4.775832176208496,
+      "learning_rate": 4.999861077302358e-05,
+      "loss": 4.6803,
+      "step": 3
+    },
+    {
+      "epoch": 0.0044742729306487695,
+      "grad_norm": 7.621553421020508,
+      "learning_rate": 4.9997530280943684e-05,
+      "loss": 4.8222,
+      "step": 4
+    },
+    {
+      "epoch": 0.005592841163310962,
+      "grad_norm": 4.997500896453857,
+      "learning_rate": 4.9996141099714405e-05,
+      "loss": 4.581,
+      "step": 5
+    },
+    {
+      "epoch": 0.006711409395973154,
+      "grad_norm": 4.785826683044434,
+      "learning_rate": 4.999444324649045e-05,
+      "loss": 4.4922,
+      "step": 6
+    },
+    {
+      "epoch": 0.007829977628635347,
+      "grad_norm": 6.647776126861572,
+      "learning_rate": 4.999243674223826e-05,
+      "loss": 4.8455,
+      "step": 7
+    },
+    {
+      "epoch": 0.008948545861297539,
+      "grad_norm": 6.596885681152344,
+      "learning_rate": 4.9990121611735704e-05,
+      "loss": 4.5977,
+      "step": 8
+    },
+    {
+      "epoch": 0.010067114093959731,
+      "grad_norm": 3.9681015014648438,
+      "learning_rate": 4.998749788357184e-05,
+      "loss": 4.5543,
+      "step": 9
+    },
+    {
+      "epoch": 0.011185682326621925,
+      "grad_norm": 3.7588727474212646,
+      "learning_rate": 4.998456559014653e-05,
+      "loss": 4.5815,
+      "step": 10
+    },
+    {
+      "epoch": 0.012304250559284116,
+      "grad_norm": 4.819014072418213,
+      "learning_rate": 4.9981324767670034e-05,
+      "loss": 4.5079,
+      "step": 11
+    },
+    {
+      "epoch": 0.013422818791946308,
+      "grad_norm": 3.4477319717407227,
+      "learning_rate": 4.997777545616258e-05,
+      "loss": 4.299,
+      "step": 12
+    },
+    {
+      "epoch": 0.0145413870246085,
+      "grad_norm": 4.113775730133057,
+      "learning_rate": 4.997391769945385e-05,
+      "loss": 4.3105,
+      "step": 13
+    },
+    {
+      "epoch": 0.015659955257270694,
+      "grad_norm": 3.848585844039917,
+      "learning_rate": 4.996975154518245e-05,
+      "loss": 4.4019,
+      "step": 14
+    },
+    {
+      "epoch": 0.016778523489932886,
+      "grad_norm": 3.787041425704956,
+      "learning_rate": 4.996527704479535e-05,
+      "loss": 4.4983,
+      "step": 15
+    },
+    {
+      "epoch": 0.017897091722595078,
+      "grad_norm": 4.073084354400635,
+      "learning_rate": 4.996049425354718e-05,
+      "loss": 4.4303,
+      "step": 16
+    },
+    {
+      "epoch": 0.01901565995525727,
+      "grad_norm": 3.8887457847595215,
+      "learning_rate": 4.99554032304996e-05,
+      "loss": 4.3752,
+      "step": 17
+    },
+    {
+      "epoch": 0.020134228187919462,
+      "grad_norm": 4.539124011993408,
+      "learning_rate": 4.995000403852057e-05,
+      "loss": 4.4398,
+      "step": 18
+    },
+    {
+      "epoch": 0.021252796420581657,
+      "grad_norm": 4.554653167724609,
+      "learning_rate": 4.994429674428356e-05,
+      "loss": 4.5068,
+      "step": 19
+    },
+    {
+      "epoch": 0.02237136465324385,
+      "grad_norm": 4.2434892654418945,
+      "learning_rate": 4.9938281418266717e-05,
+      "loss": 4.2128,
+      "step": 20
+    },
+    {
+      "epoch": 0.02348993288590604,
+      "grad_norm": 3.5199990272521973,
+      "learning_rate": 4.993195813475202e-05,
+      "loss": 4.3802,
+      "step": 21
+    },
+    {
+      "epoch": 0.024608501118568233,
+      "grad_norm": 4.401162624359131,
+      "learning_rate": 4.9925326971824345e-05,
+      "loss": 4.3791,
+      "step": 22
+    },
+    {
+      "epoch": 0.025727069351230425,
+      "grad_norm": 3.330929756164551,
+      "learning_rate": 4.9918388011370496e-05,
+      "loss": 4.3586,
+      "step": 23
+    },
+    {
+      "epoch": 0.026845637583892617,
+      "grad_norm": 3.9935824871063232,
+      "learning_rate": 4.9911141339078215e-05,
+      "loss": 4.4506,
+      "step": 24
+    },
+    {
+      "epoch": 0.02796420581655481,
+      "grad_norm": 5.406892776489258,
+      "learning_rate": 4.990358704443511e-05,
+      "loss": 4.1189,
+      "step": 25
+    },
+    {
+      "epoch": 0.029082774049217,
+      "grad_norm": 3.9377341270446777,
+      "learning_rate": 4.989572522072753e-05,
+      "loss": 4.4841,
+      "step": 26
+    },
+    {
+      "epoch": 0.030201342281879196,
+      "grad_norm": 3.408912181854248,
+      "learning_rate": 4.988755596503948e-05,
+      "loss": 4.4523,
+      "step": 27
+    },
+    {
+      "epoch": 0.03131991051454139,
+      "grad_norm": 3.9062108993530273,
+      "learning_rate": 4.987907937825133e-05,
+      "loss": 4.2521,
+      "step": 28
+    },
+    {
+      "epoch": 0.03243847874720358,
+      "grad_norm": 3.1871280670166016,
+      "learning_rate": 4.987029556503864e-05,
+      "loss": 3.9971,
+      "step": 29
+    },
+    {
+      "epoch": 0.03355704697986577,
+      "grad_norm": 3.458665609359741,
+      "learning_rate": 4.986120463387084e-05,
+      "loss": 4.3549,
+      "step": 30
+    },
+    {
+      "epoch": 0.03467561521252797,
+      "grad_norm": 3.2417380809783936,
+      "learning_rate": 4.985180669700989e-05,
+      "loss": 4.452,
+      "step": 31
+    },
+    {
+      "epoch": 0.035794183445190156,
+      "grad_norm": 2.737922430038452,
+      "learning_rate": 4.9842101870508904e-05,
+      "loss": 4.5525,
+      "step": 32
+    },
+    {
+      "epoch": 0.03691275167785235,
+      "grad_norm": 3.5653915405273438,
+      "learning_rate": 4.9832090274210714e-05,
+      "loss": 4.3004,
+      "step": 33
+    },
+    {
+      "epoch": 0.03803131991051454,
+      "grad_norm": 3.541865348815918,
+      "learning_rate": 4.982177203174636e-05,
+      "loss": 4.1722,
+      "step": 34
+    },
+    {
+      "epoch": 0.039149888143176735,
+      "grad_norm": 3.923772096633911,
+      "learning_rate": 4.981114727053362e-05,
+      "loss": 4.4216,
+      "step": 35
+    },
+    {
+      "epoch": 0.040268456375838924,
+      "grad_norm": 4.278195381164551,
+      "learning_rate": 4.98002161217754e-05,
+      "loss": 4.2207,
+      "step": 36
+    },
+    {
+      "epoch": 0.04138702460850112,
+      "grad_norm": 2.4082083702087402,
+      "learning_rate": 4.9788978720458104e-05,
+      "loss": 4.2682,
+      "step": 37
+    },
+    {
+      "epoch": 0.042505592841163314,
+      "grad_norm": 3.5256547927856445,
+      "learning_rate": 4.977743520535001e-05,
+      "loss": 4.2313,
+      "step": 38
+    },
+    {
+      "epoch": 0.0436241610738255,
+      "grad_norm": 3.8092055320739746,
+      "learning_rate": 4.9765585718999495e-05,
+      "loss": 4.3788,
+      "step": 39
+    },
+    {
+      "epoch": 0.0447427293064877,
+      "grad_norm": 3.763134241104126,
+      "learning_rate": 4.975343040773335e-05,
+      "loss": 4.1468,
+      "step": 40
+    },
+    {
+      "epoch": 0.04586129753914989,
+      "grad_norm": 4.660019874572754,
+      "learning_rate": 4.974096942165489e-05,
+      "loss": 4.1907,
+      "step": 41
+    },
+    {
+      "epoch": 0.04697986577181208,
+      "grad_norm": 4.472564220428467,
+      "learning_rate": 4.9728202914642183e-05,
+      "loss": 4.5382,
+      "step": 42
+    },
+    {
+      "epoch": 0.04809843400447427,
+      "grad_norm": 4.786443710327148,
+      "learning_rate": 4.9715131044346084e-05,
+      "loss": 4.2274,
+      "step": 43
+    },
+    {
+      "epoch": 0.049217002237136466,
+      "grad_norm": 4.188578128814697,
+      "learning_rate": 4.970175397218832e-05,
+      "loss": 4.0529,
+      "step": 44
+    },
+    {
+      "epoch": 0.050335570469798654,
+      "grad_norm": 3.1671926975250244,
+      "learning_rate": 4.9688071863359484e-05,
+      "loss": 4.1644,
+      "step": 45
+    },
+    {
+      "epoch": 0.05145413870246085,
+      "grad_norm": 3.790748357772827,
+      "learning_rate": 4.9674084886817016e-05,
+      "loss": 3.9051,
+      "step": 46
+    },
+    {
+      "epoch": 0.052572706935123045,
+      "grad_norm": 3.404773473739624,
+      "learning_rate": 4.965979321528309e-05,
+      "loss": 4.5011,
+      "step": 47
+    },
+    {
+      "epoch": 0.053691275167785234,
+      "grad_norm": 3.7244088649749756,
+      "learning_rate": 4.9645197025242506e-05,
+      "loss": 4.3524,
+      "step": 48
+    },
+    {
+      "epoch": 0.05480984340044743,
+      "grad_norm": 2.824608564376831,
+      "learning_rate": 4.963029649694049e-05,
+      "loss": 4.3211,
+      "step": 49
+    },
+    {
+      "epoch": 0.05592841163310962,
+      "grad_norm": 3.457947254180908,
+      "learning_rate": 4.9615091814380465e-05,
+      "loss": 4.3325,
+      "step": 50
+    },
+    {
+      "epoch": 0.05704697986577181,
+      "grad_norm": 2.658190965652466,
+      "learning_rate": 4.959958316532181e-05,
+      "loss": 4.2936,
+      "step": 51
+    },
+    {
+      "epoch": 0.058165548098434,
+      "grad_norm": 3.7741482257843018,
+      "learning_rate": 4.9583770741277505e-05,
+      "loss": 4.184,
+      "step": 52
+    },
+    {
+      "epoch": 0.0592841163310962,
+      "grad_norm": 2.5784506797790527,
+      "learning_rate": 4.9567654737511794e-05,
+      "loss": 4.2936,
+      "step": 53
+    },
+    {
+      "epoch": 0.06040268456375839,
+      "grad_norm": 3.018603563308716,
+      "learning_rate": 4.955123535303776e-05,
+      "loss": 4.1786,
+      "step": 54
+    },
+    {
+      "epoch": 0.06152125279642058,
+      "grad_norm": 3.7826101779937744,
+      "learning_rate": 4.953451279061485e-05,
+      "loss": 4.207,
+      "step": 55
+    },
+    {
+      "epoch": 0.06263982102908278,
+      "grad_norm": 4.488926887512207,
+      "learning_rate": 4.951748725674643e-05,
+      "loss": 4.6036,
+      "step": 56
+    },
+    {
+      "epoch": 0.06375838926174497,
+      "grad_norm": 4.047707557678223,
+      "learning_rate": 4.950015896167716e-05,
+      "loss": 3.9239,
+      "step": 57
+    },
+    {
+      "epoch": 0.06487695749440715,
+      "grad_norm": 2.5932610034942627,
+      "learning_rate": 4.9482528119390435e-05,
+      "loss": 4.1098,
+      "step": 58
+    },
+    {
+      "epoch": 0.06599552572706935,
+      "grad_norm": 2.828861951828003,
+      "learning_rate": 4.946459494760578e-05,
+      "loss": 4.2837,
+      "step": 59
+    },
+    {
+      "epoch": 0.06711409395973154,
+      "grad_norm": 3.5818116664886475,
+      "learning_rate": 4.9446359667776065e-05,
+      "loss": 4.3673,
+      "step": 60
+    },
+    {
+      "epoch": 0.06823266219239374,
+      "grad_norm": 2.7391154766082764,
+      "learning_rate": 4.9427822505084874e-05,
+      "loss": 4.5594,
+      "step": 61
+    },
+    {
+      "epoch": 0.06935123042505593,
+      "grad_norm": 2.6267666816711426,
+      "learning_rate": 4.9408983688443654e-05,
+      "loss": 4.3283,
+      "step": 62
+    },
+    {
+      "epoch": 0.07046979865771812,
+      "grad_norm": 2.574420690536499,
+      "learning_rate": 4.938984345048892e-05,
+      "loss": 4.2056,
+      "step": 63
+    },
+    {
+      "epoch": 0.07158836689038031,
+      "grad_norm": 3.568028450012207,
+      "learning_rate": 4.937040202757937e-05,
+      "loss": 4.2298,
+      "step": 64
+    },
+    {
+      "epoch": 0.07270693512304251,
+      "grad_norm": 3.778529405593872,
+      "learning_rate": 4.9350659659792976e-05,
+      "loss": 4.1133,
+      "step": 65
+    },
+    {
+      "epoch": 0.0738255033557047,
+      "grad_norm": 2.817169427871704,
+      "learning_rate": 4.933061659092401e-05,
+      "loss": 4.183,
+      "step": 66
+    },
+    {
+      "epoch": 0.07494407158836688,
+      "grad_norm": 3.1611924171447754,
+      "learning_rate": 4.931027306848004e-05,
+      "loss": 4.1715,
+      "step": 67
+    },
+    {
+      "epoch": 0.07606263982102908,
+      "grad_norm": 3.813957929611206,
+      "learning_rate": 4.9289629343678864e-05,
+      "loss": 4.2382,
+      "step": 68
+    },
+    {
+      "epoch": 0.07718120805369127,
+      "grad_norm": 2.7094550132751465,
+      "learning_rate": 4.926868567144543e-05,
+      "loss": 4.107,
+      "step": 69
+    },
+    {
+      "epoch": 0.07829977628635347,
+      "grad_norm": 2.958317995071411,
+      "learning_rate": 4.924744231040864e-05,
+      "loss": 4.2052,
+      "step": 70
+    },
+    {
+      "epoch": 0.07941834451901567,
+      "grad_norm": 2.8046460151672363,
+      "learning_rate": 4.9225899522898236e-05,
+      "loss": 4.1191,
+      "step": 71
+    },
+    {
+      "epoch": 0.08053691275167785,
+      "grad_norm": 2.8792548179626465,
+      "learning_rate": 4.920405757494147e-05,
+      "loss": 3.9608,
+      "step": 72
+    },
+    {
+      "epoch": 0.08165548098434004,
+      "grad_norm": 2.953927516937256,
+      "learning_rate": 4.91819167362599e-05,
+      "loss": 4.6505,
+      "step": 73
+    },
+    {
+      "epoch": 0.08277404921700224,
+      "grad_norm": 2.8335180282592773,
+      "learning_rate": 4.915947728026598e-05,
+      "loss": 4.5291,
+      "step": 74
+    },
+    {
+      "epoch": 0.08389261744966443,
+      "grad_norm": 2.0205912590026855,
+      "learning_rate": 4.9136739484059766e-05,
+      "loss": 4.1919,
+      "step": 75
+    },
+    {
+      "epoch": 0.08501118568232663,
+      "grad_norm": 4.005690097808838,
+      "learning_rate": 4.911370362842543e-05,
+      "loss": 4.2847,
+      "step": 76
+    },
+    {
+      "epoch": 0.08612975391498881,
+      "grad_norm": 2.3989546298980713,
+      "learning_rate": 4.9090369997827826e-05,
+      "loss": 4.2773,
+      "step": 77
+    },
+    {
+      "epoch": 0.087248322147651,
+      "grad_norm": 2.6297130584716797,
+      "learning_rate": 4.9066738880408945e-05,
+      "loss": 4.1242,
+      "step": 78
+    },
+    {
+      "epoch": 0.0883668903803132,
+      "grad_norm": 3.310232639312744,
+      "learning_rate": 4.904281056798441e-05,
+      "loss": 4.4331,
+      "step": 79
+    },
+    {
+      "epoch": 0.0894854586129754,
+      "grad_norm": 1.7795904874801636,
+      "learning_rate": 4.901858535603983e-05,
+      "loss": 4.1656,
+      "step": 80
+    },
+    {
+      "epoch": 0.09060402684563758,
+      "grad_norm": 2.2798268795013428,
+      "learning_rate": 4.899406354372715e-05,
+      "loss": 4.1374,
+      "step": 81
+    },
+    {
+      "epoch": 0.09172259507829977,
+      "grad_norm": 2.3790340423583984,
+      "learning_rate": 4.896924543386099e-05,
+      "loss": 4.1786,
+      "step": 82
+    },
+    {
+      "epoch": 0.09284116331096197,
+      "grad_norm": 2.345724105834961,
+      "learning_rate": 4.894413133291488e-05,
+      "loss": 3.9511,
+      "step": 83
+    },
+    {
+      "epoch": 0.09395973154362416,
+      "grad_norm": 3.110872507095337,
+      "learning_rate": 4.891872155101746e-05,
+      "loss": 4.2832,
+      "step": 84
+    },
+    {
+      "epoch": 0.09507829977628636,
+      "grad_norm": 2.6279842853546143,
+      "learning_rate": 4.889301640194869e-05,
+      "loss": 4.4677,
+      "step": 85
+    },
+    {
+      "epoch": 0.09619686800894854,
+      "grad_norm": 2.8604133129119873,
+      "learning_rate": 4.886701620313595e-05,
+      "loss": 4.3158,
+      "step": 86
+    },
+    {
+      "epoch": 0.09731543624161074,
+      "grad_norm": 4.6412858963012695,
+      "learning_rate": 4.884072127565014e-05,
+      "loss": 4.4612,
+      "step": 87
+    },
+    {
+      "epoch": 0.09843400447427293,
+      "grad_norm": 2.649590492248535,
+      "learning_rate": 4.881413194420169e-05,
+      "loss": 3.9387,
+      "step": 88
+    },
+    {
+      "epoch": 0.09955257270693513,
+      "grad_norm": 3.56644344329834,
+      "learning_rate": 4.878724853713655e-05,
+      "loss": 4.2099,
+      "step": 89
+    },
+    {
+      "epoch": 0.10067114093959731,
+      "grad_norm": 2.599231243133545,
+      "learning_rate": 4.876007138643216e-05,
+      "loss": 4.3014,
+      "step": 90
+    },
+    {
+      "epoch": 0.1017897091722595,
+      "grad_norm": 2.8511576652526855,
+      "learning_rate": 4.8732600827693344e-05,
+      "loss": 3.8494,
+      "step": 91
+    },
+    {
+      "epoch": 0.1029082774049217,
+      "grad_norm": 2.3662049770355225,
+      "learning_rate": 4.870483720014814e-05,
+      "loss": 3.9008,
+      "step": 92
+    },
+    {
+      "epoch": 0.1040268456375839,
+      "grad_norm": 3.010821580886841,
+      "learning_rate": 4.8676780846643644e-05,
+      "loss": 4.1022,
+      "step": 93
+    },
+    {
+      "epoch": 0.10514541387024609,
+      "grad_norm": 3.158817768096924,
+      "learning_rate": 4.864843211364176e-05,
+      "loss": 4.2633,
+      "step": 94
+    },
+    {
+      "epoch": 0.10626398210290827,
+      "grad_norm": 2.9291787147521973,
+      "learning_rate": 4.861979135121493e-05,
+      "loss": 4.0874,
+      "step": 95
+    },
+    {
+      "epoch": 0.10738255033557047,
+      "grad_norm": 3.55708909034729,
+      "learning_rate": 4.859085891304178e-05,
+      "loss": 4.1598,
+      "step": 96
+    },
+    {
+      "epoch": 0.10850111856823266,
+      "grad_norm": 2.482081651687622,
+      "learning_rate": 4.85616351564028e-05,
+      "loss": 4.4488,
+      "step": 97
+    },
+    {
+      "epoch": 0.10961968680089486,
+      "grad_norm": 2.2329485416412354,
+      "learning_rate": 4.853212044217591e-05,
+      "loss": 4.1849,
+      "step": 98
+    },
+    {
+      "epoch": 0.11073825503355705,
+      "grad_norm": 2.4485504627227783,
+      "learning_rate": 4.8502315134832e-05,
+      "loss": 4.2842,
+      "step": 99
+    },
+    {
+      "epoch": 0.11185682326621924,
+      "grad_norm": 2.239509344100952,
+      "learning_rate": 4.847221960243041e-05,
+      "loss": 4.4354,
+      "step": 100
+    },
+    {
+      "epoch": 0.11297539149888143,
+      "grad_norm": 2.5421688556671143,
+      "learning_rate": 4.8441834216614454e-05,
+      "loss": 4.4723,
+      "step": 101
+    },
+    {
+      "epoch": 0.11409395973154363,
+      "grad_norm": 3.123340129852295,
+      "learning_rate": 4.8411159352606734e-05,
+      "loss": 3.812,
+      "step": 102
+    },
+    {
+      "epoch": 0.11521252796420582,
+      "grad_norm": 1.9788379669189453,
+      "learning_rate": 4.838019538920458e-05,
+      "loss": 4.2419,
+      "step": 103
+    },
+    {
+      "epoch": 0.116331096196868,
+      "grad_norm": 2.201848268508911,
+      "learning_rate": 4.834894270877536e-05,
+      "loss": 4.1965,
+      "step": 104
+    },
+    {
+      "epoch": 0.1174496644295302,
+      "grad_norm": 2.6100194454193115,
+      "learning_rate": 4.831740169725172e-05,
+      "loss": 4.3533,
+      "step": 105
+    },
+    {
+      "epoch": 0.1185682326621924,
+      "grad_norm": 2.866046905517578,
+      "learning_rate": 4.8285572744126854e-05,
+      "loss": 3.9282,
+      "step": 106
+    },
+    {
+      "epoch": 0.11968680089485459,
+      "grad_norm": 2.9005253314971924,
+      "learning_rate": 4.8253456242449704e-05,
+      "loss": 4.2894,
+      "step": 107
+    },
+    {
+      "epoch": 0.12080536912751678,
+      "grad_norm": 2.173900842666626,
+      "learning_rate": 4.822105258882007e-05,
+      "loss": 4.3334,
+      "step": 108
+    },
+    {
+      "epoch": 0.12192393736017897,
+      "grad_norm": 2.949611186981201,
+      "learning_rate": 4.818836218338373e-05,
+      "loss": 4.2813,
+      "step": 109
+    },
+    {
+      "epoch": 0.12304250559284116,
+      "grad_norm": 3.2048230171203613,
+      "learning_rate": 4.81553854298275e-05,
+      "loss": 4.3333,
+      "step": 110
+    },
+    {
+      "epoch": 0.12416107382550336,
+      "grad_norm": 2.2151906490325928,
+      "learning_rate": 4.812212273537425e-05,
+      "loss": 4.0185,
+      "step": 111
+    },
+    {
+      "epoch": 0.12527964205816555,
+      "grad_norm": 3.4764933586120605,
+      "learning_rate": 4.808857451077788e-05,
+      "loss": 4.0535,
+      "step": 112
+    },
+    {
+      "epoch": 0.12639821029082773,
+      "grad_norm": 2.3762760162353516,
+      "learning_rate": 4.805474117031822e-05,
+      "loss": 4.3605,
+      "step": 113
+    },
+    {
+      "epoch": 0.12751677852348994,
+      "grad_norm": 2.3675270080566406,
+      "learning_rate": 4.802062313179595e-05,
+      "loss": 4.1901,
+      "step": 114
+    },
+    {
+      "epoch": 0.12863534675615212,
+      "grad_norm": 2.8210885524749756,
+      "learning_rate": 4.798622081652743e-05,
+      "loss": 4.5484,
+      "step": 115
+    },
+    {
+      "epoch": 0.1297539149888143,
+      "grad_norm": 2.597954511642456,
+      "learning_rate": 4.795153464933948e-05,
+      "loss": 4.3055,
+      "step": 116
+    },
+    {
+      "epoch": 0.13087248322147652,
+      "grad_norm": 2.793868064880371,
+      "learning_rate": 4.7916565058564155e-05,
+      "loss": 4.4277,
+      "step": 117
+    },
+    {
+      "epoch": 0.1319910514541387,
+      "grad_norm": 3.061800718307495,
+      "learning_rate": 4.788131247603345e-05,
+      "loss": 4.2027,
+      "step": 118
+    },
+    {
+      "epoch": 0.1331096196868009,
+      "grad_norm": 3.1223654747009277,
+      "learning_rate": 4.784577733707394e-05,
+      "loss": 4.3513,
+      "step": 119
+    },
+    {
+      "epoch": 0.1342281879194631,
+      "grad_norm": 2.714157819747925,
+      "learning_rate": 4.7809960080501464e-05,
+      "loss": 4.2208,
+      "step": 120
+    },
+    {
+      "epoch": 0.13534675615212527,
+      "grad_norm": 2.218320369720459,
+      "learning_rate": 4.777386114861565e-05,
+      "loss": 4.2255,
+      "step": 121
+    },
+    {
+      "epoch": 0.13646532438478748,
+      "grad_norm": 2.74094557762146,
+      "learning_rate": 4.7737480987194484e-05,
+      "loss": 4.4531,
+      "step": 122
+    },
+    {
+      "epoch": 0.13758389261744966,
+      "grad_norm": 2.0145950317382812,
+      "learning_rate": 4.7700820045488783e-05,
+      "loss": 4.2616,
+      "step": 123
+    },
+    {
+      "epoch": 0.13870246085011187,
+      "grad_norm": 3.6711690425872803,
+      "learning_rate": 4.766387877621667e-05,
+      "loss": 4.3626,
+      "step": 124
+    },
+    {
+      "epoch": 0.13982102908277405,
+      "grad_norm": 2.7168869972229004,
+      "learning_rate": 4.762665763555797e-05,
+      "loss": 4.1299,
+      "step": 125
+    },
+    {
+      "epoch": 0.14093959731543623,
+      "grad_norm": 2.299030065536499,
+      "learning_rate": 4.758915708314858e-05,
+      "loss": 4.378,
+      "step": 126
+    },
+    {
+      "epoch": 0.14205816554809844,
+      "grad_norm": 2.2006888389587402,
+      "learning_rate": 4.755137758207479e-05,
+      "loss": 4.5967,
+      "step": 127
+    },
+    {
+      "epoch": 0.14317673378076062,
+      "grad_norm": 2.05462908744812,
+      "learning_rate": 4.751331959886758e-05,
+      "loss": 4.235,
+      "step": 128
+    },
+    {
+      "epoch": 0.14429530201342283,
+      "grad_norm": 1.8818479776382446,
+      "learning_rate": 4.7474983603496815e-05,
+      "loss": 4.1797,
+      "step": 129
+    },
+    {
+      "epoch": 0.14541387024608501,
+      "grad_norm": 2.008810043334961,
+      "learning_rate": 4.7436370069365524e-05,
+      "loss": 4.26,
+      "step": 130
+    },
+    {
+      "epoch": 0.1465324384787472,
+      "grad_norm": 2.405534029006958,
+      "learning_rate": 4.739747947330394e-05,
+      "loss": 3.8538,
+      "step": 131
+    },
+    {
+      "epoch": 0.1476510067114094,
+      "grad_norm": 2.6731414794921875,
+      "learning_rate": 4.7358312295563734e-05,
+      "loss": 4.1691,
+      "step": 132
+    },
+    {
+      "epoch": 0.1487695749440716,
+      "grad_norm": 3.544353723526001,
+      "learning_rate": 4.7318869019811986e-05,
+      "loss": 4.0219,
+      "step": 133
+    },
+    {
+      "epoch": 0.14988814317673377,
+      "grad_norm": 2.278864622116089,
+      "learning_rate": 4.727915013312526e-05,
+      "loss": 4.3926,
+      "step": 134
+    },
+    {
+      "epoch": 0.15100671140939598,
+      "grad_norm": 2.240647554397583,
+      "learning_rate": 4.7239156125983594e-05,
+      "loss": 4.3835,
+      "step": 135
+    },
+    {
+      "epoch": 0.15212527964205816,
+      "grad_norm": 3.1606392860412598,
+      "learning_rate": 4.7198887492264416e-05,
+      "loss": 3.9971,
+      "step": 136
+    },
+    {
+      "epoch": 0.15324384787472037,
+      "grad_norm": 2.4671874046325684,
+      "learning_rate": 4.7158344729236454e-05,
+      "loss": 4.4052,
+      "step": 137
+    },
+    {
+      "epoch": 0.15436241610738255,
+      "grad_norm": 1.9113795757293701,
+      "learning_rate": 4.711752833755362e-05,
+      "loss": 4.1998,
+      "step": 138
+    },
+    {
+      "epoch": 0.15548098434004473,
+      "grad_norm": 2.9461162090301514,
+      "learning_rate": 4.707643882124878e-05,
+      "loss": 4.212,
+      "step": 139
+    },
+    {
+      "epoch": 0.15659955257270694,
+      "grad_norm": 3.278677463531494,
+      "learning_rate": 4.70350766877276e-05,
+      "loss": 4.4147,
+      "step": 140
+    },
+    {
+      "epoch": 0.15771812080536912,
+      "grad_norm": 3.2217676639556885,
+      "learning_rate": 4.699344244776218e-05,
+      "loss": 4.2919,
+      "step": 141
+    },
+    {
+      "epoch": 0.15883668903803133,
+      "grad_norm": 1.8135348558425903,
+      "learning_rate": 4.6951536615484854e-05,
+      "loss": 4.0982,
+      "step": 142
+    },
+    {
+      "epoch": 0.1599552572706935,
+      "grad_norm": 3.3648054599761963,
+      "learning_rate": 4.6909359708381775e-05,
+      "loss": 4.3256,
+      "step": 143
+    },
+    {
+      "epoch": 0.1610738255033557,
+      "grad_norm": 2.8032402992248535,
+      "learning_rate": 4.686691224728652e-05,
+      "loss": 4.3029,
+      "step": 144
+    },
+    {
+      "epoch": 0.1621923937360179,
+      "grad_norm": 2.585247039794922,
+      "learning_rate": 4.682419475637372e-05,
+      "loss": 4.1191,
+      "step": 145
+    },
+    {
+      "epoch": 0.16331096196868009,
+      "grad_norm": 2.0398333072662354,
+      "learning_rate": 4.678120776315251e-05,
+      "loss": 4.2553,
+      "step": 146
+    },
+    {
+      "epoch": 0.1644295302013423,
+      "grad_norm": 2.485276460647583,
+      "learning_rate": 4.673795179846007e-05,
+      "loss": 4.2602,
+      "step": 147
+    },
+    {
+      "epoch": 0.16554809843400448,
+      "grad_norm": 2.1474947929382324,
+      "learning_rate": 4.669442739645506e-05,
+      "loss": 4.2519,
+      "step": 148
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 2.4320313930511475,
+      "learning_rate": 4.665063509461097e-05,
+      "loss": 4.2698,
+      "step": 149
+    },
+    {
+      "epoch": 0.16778523489932887,
+      "grad_norm": 3.272096872329712,
+      "learning_rate": 4.660657543370958e-05,
+      "loss": 4.2166,
+      "step": 150
+    },
+    {
+      "epoch": 0.16890380313199105,
+      "grad_norm": 2.540706157684326,
+      "learning_rate": 4.656224895783421e-05,
+      "loss": 4.133,
+      "step": 151
+    },
+    {
+      "epoch": 0.17002237136465326,
+      "grad_norm": 2.285813808441162,
+      "learning_rate": 4.651765621436303e-05,
+      "loss": 4.4777,
+      "step": 152
+    },
+    {
+      "epoch": 0.17114093959731544,
+      "grad_norm": 3.893460273742676,
+      "learning_rate": 4.6472797753962246e-05,
+      "loss": 4.2023,
+      "step": 153
+    },
+    {
+      "epoch": 0.17225950782997762,
+      "grad_norm": 2.54040789604187,
+      "learning_rate": 4.6427674130579424e-05,
+      "loss": 4.2924,
+      "step": 154
+    },
+    {
+      "epoch": 0.17337807606263983,
+      "grad_norm": 2.005213975906372,
+      "learning_rate": 4.6382285901436495e-05,
+      "loss": 3.8809,
+      "step": 155
+    },
+    {
+      "epoch": 0.174496644295302,
+      "grad_norm": 3.2139828205108643,
+      "learning_rate": 4.633663362702299e-05,
+      "loss": 4.5209,
+      "step": 156
+    },
+    {
+      "epoch": 0.1756152125279642,
+      "grad_norm": 2.509523868560791,
+      "learning_rate": 4.62907178710891e-05,
+      "loss": 4.3067,
+      "step": 157
+    },
+    {
+      "epoch": 0.1767337807606264,
+      "grad_norm": 2.318631649017334,
+      "learning_rate": 4.6244539200638626e-05,
+      "loss": 4.0036,
+      "step": 158
+    },
+    {
+      "epoch": 0.17785234899328858,
+      "grad_norm": 2.146442413330078,
+      "learning_rate": 4.61980981859221e-05,
+      "loss": 4.3332,
+      "step": 159
+    },
+    {
+      "epoch": 0.1789709172259508,
+      "grad_norm": 2.574519634246826,
+      "learning_rate": 4.615139540042966e-05,
+      "loss": 4.0637,
+      "step": 160
+    },
+    {
+      "epoch": 0.18008948545861297,
+      "grad_norm": 2.722486972808838,
+      "learning_rate": 4.610443142088402e-05,
+      "loss": 4.1284,
+      "step": 161
+    },
+    {
+      "epoch": 0.18120805369127516,
+      "grad_norm": 2.3555033206939697,
+      "learning_rate": 4.60572068272333e-05,
+      "loss": 4.2058,
+      "step": 162
+    },
+    {
+      "epoch": 0.18232662192393737,
+      "grad_norm": 2.389615297317505,
+      "learning_rate": 4.60097222026439e-05,
+      "loss": 4.0544,
+      "step": 163
+    },
+    {
+      "epoch": 0.18344519015659955,
+      "grad_norm": 2.4774980545043945,
+      "learning_rate": 4.596197813349328e-05,
+      "loss": 4.0337,
+      "step": 164
+    },
+    {
+      "epoch": 0.18456375838926176,
+      "grad_norm": 2.1499996185302734,
+      "learning_rate": 4.591397520936271e-05,
+      "loss": 4.1015,
+      "step": 165
+    },
+    {
+      "epoch": 0.18568232662192394,
+      "grad_norm": 2.3946421146392822,
+      "learning_rate": 4.586571402303006e-05,
+      "loss": 3.8225,
+      "step": 166
+    },
+    {
+      "epoch": 0.18680089485458612,
+      "grad_norm": 2.3064377307891846,
+      "learning_rate": 4.581719517046236e-05,
+      "loss": 4.0789,
+      "step": 167
+    },
+    {
+      "epoch": 0.18791946308724833,
+      "grad_norm": 2.358966827392578,
+      "learning_rate": 4.576841925080853e-05,
+      "loss": 3.9453,
+      "step": 168
+    },
+    {
+      "epoch": 0.1890380313199105,
+      "grad_norm": 1.9163504838943481,
+      "learning_rate": 4.5719386866391976e-05,
+      "loss": 3.9758,
+      "step": 169
+    },
+    {
+      "epoch": 0.19015659955257272,
+      "grad_norm": 2.147257089614868,
+      "learning_rate": 4.56700986227031e-05,
+      "loss": 4.0456,
+      "step": 170
+    },
+    {
+      "epoch": 0.1912751677852349,
+      "grad_norm": 2.340684413909912,
+      "learning_rate": 4.5620555128391884e-05,
+      "loss": 3.9842,
+      "step": 171
+    },
+    {
+      "epoch": 0.19239373601789708,
+      "grad_norm": 1.9256740808486938,
+      "learning_rate": 4.557075699526032e-05,
+      "loss": 3.9314,
+      "step": 172
+    },
+    {
+      "epoch": 0.1935123042505593,
+      "grad_norm": 2.1142566204071045,
+      "learning_rate": 4.552070483825489e-05,
+      "loss": 4.216,
+      "step": 173
+    },
+    {
+      "epoch": 0.19463087248322147,
+      "grad_norm": 2.934136152267456,
+      "learning_rate": 4.5470399275458985e-05,
+      "loss": 4.3844,
+      "step": 174
+    },
+    {
+      "epoch": 0.19574944071588368,
+      "grad_norm": 2.802530527114868,
+      "learning_rate": 4.541984092808521e-05,
+      "loss": 4.1614,
+      "step": 175
+    },
+    {
+      "epoch": 0.19686800894854586,
+      "grad_norm": 3.0535061359405518,
+      "learning_rate": 4.536903042046777e-05,
+      "loss": 4.0901,
+      "step": 176
+    },
+    {
+      "epoch": 0.19798657718120805,
+      "grad_norm": 3.6601295471191406,
+      "learning_rate": 4.531796838005477e-05,
+      "loss": 4.1447,
+      "step": 177
+    },
+    {
+      "epoch": 0.19910514541387025,
+      "grad_norm": 2.675509214401245,
+      "learning_rate": 4.526665543740038e-05,
+      "loss": 4.0328,
+      "step": 178
+    },
+    {
+      "epoch": 0.20022371364653244,
+      "grad_norm": 3.0935540199279785,
+      "learning_rate": 4.5215092226157165e-05,
+      "loss": 4.0755,
+      "step": 179
+    },
+    {
+      "epoch": 0.20134228187919462,
+      "grad_norm": 2.0524206161499023,
+      "learning_rate": 4.516327938306818e-05,
+      "loss": 4.2309,
+      "step": 180
+    },
+    {
+      "epoch": 0.20246085011185683,
+      "grad_norm": 2.2571775913238525,
+      "learning_rate": 4.5111217547959114e-05,
+      "loss": 4.1893,
+      "step": 181
+    },
+    {
+      "epoch": 0.203579418344519,
+      "grad_norm": 3.665895462036133,
+      "learning_rate": 4.505890736373045e-05,
+      "loss": 4.2883,
+      "step": 182
+    },
+    {
+      "epoch": 0.20469798657718122,
+      "grad_norm": 2.241145372390747,
+      "learning_rate": 4.500634947634943e-05,
+      "loss": 4.1803,
+      "step": 183
+    },
+    {
+      "epoch": 0.2058165548098434,
+      "grad_norm": 2.3424227237701416,
+      "learning_rate": 4.495354453484216e-05,
+      "loss": 3.9584,
+      "step": 184
+    },
+    {
+      "epoch": 0.20693512304250558,
+      "grad_norm": 2.7996134757995605,
+      "learning_rate": 4.4900493191285554e-05,
+      "loss": 3.977,
+      "step": 185
+    },
+    {
+      "epoch": 0.2080536912751678,
+      "grad_norm": 2.122816324234009,
+      "learning_rate": 4.48471961007993e-05,
+      "loss": 4.1925,
+      "step": 186
+    },
+    {
+      "epoch": 0.20917225950782997,
+      "grad_norm": 2.3053228855133057,
+      "learning_rate": 4.479365392153776e-05,
+      "loss": 4.3152,
+      "step": 187
+    },
+    {
+      "epoch": 0.21029082774049218,
+      "grad_norm": 2.4579198360443115,
+      "learning_rate": 4.473986731468183e-05,
+      "loss": 3.8768,
+      "step": 188
+    },
+    {
+      "epoch": 0.21140939597315436,
+      "grad_norm": 4.130041599273682,
+      "learning_rate": 4.4685836944430816e-05,
+      "loss": 3.9151,
+      "step": 189
+    },
+    {
+      "epoch": 0.21252796420581654,
+      "grad_norm": 3.137632369995117,
+      "learning_rate": 4.4631563477994184e-05,
+      "loss": 4.2563,
+      "step": 190
+    },
+    {
+      "epoch": 0.21364653243847875,
+      "grad_norm": 2.419900894165039,
+      "learning_rate": 4.457704758558335e-05,
+      "loss": 4.0084,
+      "step": 191
+    },
+    {
+      "epoch": 0.21476510067114093,
+      "grad_norm": 2.5032031536102295,
+      "learning_rate": 4.4522289940403404e-05,
+      "loss": 4.2155,
+      "step": 192
+    },
+    {
+      "epoch": 0.21588366890380314,
+      "grad_norm": 2.497797727584839,
+      "learning_rate": 4.446729121864478e-05,
+      "loss": 3.8909,
+      "step": 193
+    },
+    {
+      "epoch": 0.21700223713646533,
+      "grad_norm": 3.136139154434204,
+      "learning_rate": 4.441205209947491e-05,
+      "loss": 4.45,
+      "step": 194
+    },
+    {
+      "epoch": 0.2181208053691275,
+      "grad_norm": 2.3615260124206543,
+      "learning_rate": 4.435657326502986e-05,
+      "loss": 4.3952,
+      "step": 195
+    },
+    {
+      "epoch": 0.21923937360178972,
+      "grad_norm": 2.7231903076171875,
+      "learning_rate": 4.430085540040587e-05,
+      "loss": 4.0338,
+      "step": 196
+    },
+    {
+      "epoch": 0.2203579418344519,
+      "grad_norm": 2.807499408721924,
+      "learning_rate": 4.4244899193650933e-05,
+      "loss": 4.4239,
+      "step": 197
+    },
+    {
+      "epoch": 0.2214765100671141,
+      "grad_norm": 2.6443018913269043,
+      "learning_rate": 4.418870533575625e-05,
+      "loss": 4.1129,
+      "step": 198
+    },
+    {
+      "epoch": 0.2225950782997763,
+      "grad_norm": 2.2121267318725586,
+      "learning_rate": 4.4132274520647754e-05,
+      "loss": 4.2689,
+      "step": 199
+    },
+    {
+      "epoch": 0.22371364653243847,
+      "grad_norm": 2.53283953666687,
+      "learning_rate": 4.407560744517749e-05,
+      "loss": 4.1621,
+      "step": 200
+    },
+    {
+      "epoch": 0.22483221476510068,
+      "grad_norm": 2.109092950820923,
+      "learning_rate": 4.401870480911504e-05,
+      "loss": 4.2749,
+      "step": 201
+    },
+    {
+      "epoch": 0.22595078299776286,
+      "grad_norm": 2.3910768032073975,
+      "learning_rate": 4.3961567315138885e-05,
+      "loss": 4.2594,
+      "step": 202
+    },
+    {
+      "epoch": 0.22706935123042504,
+      "grad_norm": 3.1483852863311768,
+      "learning_rate": 4.3904195668827697e-05,
+      "loss": 3.8371,
+      "step": 203
+    },
+    {
+      "epoch": 0.22818791946308725,
+      "grad_norm": 3.3514859676361084,
+      "learning_rate": 4.384659057865165e-05,
+      "loss": 4.1402,
+      "step": 204
+    },
+    {
+      "epoch": 0.22930648769574943,
+      "grad_norm": 4.020583629608154,
+      "learning_rate": 4.378875275596367e-05,
+      "loss": 3.8622,
+      "step": 205
+    },
+    {
+      "epoch": 0.23042505592841164,
+      "grad_norm": 2.541602611541748,
+      "learning_rate": 4.3730682914990653e-05,
+      "loss": 4.3944,
+      "step": 206
+    },
+    {
+      "epoch": 0.23154362416107382,
+      "grad_norm": 2.3997037410736084,
+      "learning_rate": 4.3672381772824615e-05,
+      "loss": 4.1781,
+      "step": 207
+    },
+    {
+      "epoch": 0.232662192393736,
+      "grad_norm": 3.2126035690307617,
+      "learning_rate": 4.36138500494139e-05,
+      "loss": 4.115,
+      "step": 208
+    },
+    {
+      "epoch": 0.23378076062639822,
+      "grad_norm": 2.839704990386963,
+      "learning_rate": 4.355508846755422e-05,
+      "loss": 3.9238,
+      "step": 209
+    },
+    {
+      "epoch": 0.2348993288590604,
+      "grad_norm": 3.05448317527771,
+      "learning_rate": 4.3496097752879764e-05,
+      "loss": 4.11,
+      "step": 210
+    },
+    {
+      "epoch": 0.2360178970917226,
+      "grad_norm": 2.8709630966186523,
+      "learning_rate": 4.343687863385425e-05,
+      "loss": 3.9743,
+      "step": 211
+    },
+    {
+      "epoch": 0.2371364653243848,
+      "grad_norm": 2.701570749282837,
+      "learning_rate": 4.3377431841761875e-05,
+      "loss": 3.8432,
+      "step": 212
+    },
+    {
+      "epoch": 0.23825503355704697,
+      "grad_norm": 2.5893900394439697,
+      "learning_rate": 4.331775811069837e-05,
+      "loss": 4.0372,
+      "step": 213
+    },
+    {
+      "epoch": 0.23937360178970918,
+      "grad_norm": 2.2405996322631836,
+      "learning_rate": 4.325785817756186e-05,
+      "loss": 3.8983,
+      "step": 214
+    },
+    {
+      "epoch": 0.24049217002237136,
+      "grad_norm": 2.6390066146850586,
+      "learning_rate": 4.3197732782043784e-05,
+      "loss": 4.2161,
+      "step": 215
+    },
+    {
+      "epoch": 0.24161073825503357,
+      "grad_norm": 2.5067667961120605,
+      "learning_rate": 4.3137382666619783e-05,
+      "loss": 4.3013,
+      "step": 216
+    },
+    {
+      "epoch": 0.24272930648769575,
+      "grad_norm": 2.50183367729187,
+      "learning_rate": 4.307680857654052e-05,
+      "loss": 4.1306,
+      "step": 217
+    },
+    {
+      "epoch": 0.24384787472035793,
+      "grad_norm": 2.526111364364624,
+      "learning_rate": 4.301601125982245e-05,
+      "loss": 3.9925,
+      "step": 218
+    },
+    {
+      "epoch": 0.24496644295302014,
+      "grad_norm": 2.189241886138916,
+      "learning_rate": 4.2954991467238634e-05,
+      "loss": 4.1998,
+      "step": 219
+    },
+    {
+      "epoch": 0.24608501118568232,
+      "grad_norm": 2.3332066535949707,
+      "learning_rate": 4.289374995230942e-05,
+      "loss": 4.1606,
+      "step": 220
+    },
+    {
+      "epoch": 0.24720357941834453,
+      "grad_norm": 3.0104753971099854,
+      "learning_rate": 4.2832287471293155e-05,
+      "loss": 4.3294,
+      "step": 221
+    },
+    {
+      "epoch": 0.2483221476510067,
+      "grad_norm": 3.3456063270568848,
+      "learning_rate": 4.277060478317687e-05,
+      "loss": 4.3516,
+      "step": 222
+    },
+    {
+      "epoch": 0.2494407158836689,
+      "grad_norm": 1.9993607997894287,
+      "learning_rate": 4.270870264966687e-05,
+      "loss": 4.2042,
+      "step": 223
+    },
+    {
+      "epoch": 0.2505592841163311,
+      "grad_norm": 3.882202625274658,
+      "learning_rate": 4.264658183517935e-05,
+      "loss": 4.1279,
+      "step": 224
+    },
+    {
+      "epoch": 0.2516778523489933,
+      "grad_norm": 2.162135124206543,
+      "learning_rate": 4.258424310683094e-05,
+      "loss": 4.0197,
+      "step": 225
+    },
+    {
+      "epoch": 0.25279642058165547,
+      "grad_norm": 1.9756934642791748,
+      "learning_rate": 4.2521687234429264e-05,
+      "loss": 4.1365,
+      "step": 226
+    },
+    {
+      "epoch": 0.2539149888143177,
+      "grad_norm": 2.928903579711914,
+      "learning_rate": 4.245891499046338e-05,
+      "loss": 4.1895,
+      "step": 227
+    },
+    {
+      "epoch": 0.2550335570469799,
+      "grad_norm": 2.3427515029907227,
+      "learning_rate": 4.239592715009429e-05,
+      "loss": 4.3197,
+      "step": 228
+    },
+    {
+      "epoch": 0.25615212527964204,
+      "grad_norm": 2.0358877182006836,
+      "learning_rate": 4.2332724491145374e-05,
+      "loss": 4.0241,
+      "step": 229
+    },
+    {
+      "epoch": 0.25727069351230425,
+      "grad_norm": 3.108318567276001,
+      "learning_rate": 4.226930779409271e-05,
+      "loss": 3.8592,
+      "step": 230
+    },
+    {
+      "epoch": 0.25838926174496646,
+      "grad_norm": 2.0987706184387207,
+      "learning_rate": 4.2205677842055516e-05,
+      "loss": 4.1741,
+      "step": 231
+    },
+    {
+      "epoch": 0.2595078299776286,
+      "grad_norm": 2.86515474319458,
+      "learning_rate": 4.214183542078646e-05,
+      "loss": 4.0576,
+      "step": 232
+    },
+    {
+      "epoch": 0.2606263982102908,
+      "grad_norm": 3.104260206222534,
+      "learning_rate": 4.207778131866191e-05,
+      "loss": 3.9961,
+      "step": 233
+    },
+    {
+      "epoch": 0.26174496644295303,
+      "grad_norm": 2.444136142730713,
+      "learning_rate": 4.2013516326672273e-05,
+      "loss": 4.0128,
+      "step": 234
+    },
+    {
+      "epoch": 0.26286353467561524,
+      "grad_norm": 2.4717466831207275,
+      "learning_rate": 4.194904123841218e-05,
+      "loss": 4.4341,
+      "step": 235
+    },
+    {
+      "epoch": 0.2639821029082774,
+      "grad_norm": 2.2069778442382812,
+      "learning_rate": 4.188435685007069e-05,
+      "loss": 3.9367,
+      "step": 236
+    },
+    {
+      "epoch": 0.2651006711409396,
+      "grad_norm": 2.673962354660034,
+      "learning_rate": 4.1819463960421454e-05,
+      "loss": 3.9492,
+      "step": 237
+    },
+    {
+      "epoch": 0.2662192393736018,
+      "grad_norm": 2.869675397872925,
+      "learning_rate": 4.175436337081289e-05,
+      "loss": 4.2262,
+      "step": 238
+    },
+    {
+      "epoch": 0.26733780760626397,
+      "grad_norm": 2.8593602180480957,
+      "learning_rate": 4.168905588515822e-05,
+      "loss": 3.9327,
+      "step": 239
+    },
+    {
+      "epoch": 0.2684563758389262,
+      "grad_norm": 3.113313674926758,
+      "learning_rate": 4.162354230992562e-05,
+      "loss": 4.3994,
+      "step": 240
+    },
+    {
+      "epoch": 0.2695749440715884,
+      "grad_norm": 3.0361926555633545,
+      "learning_rate": 4.155782345412819e-05,
+      "loss": 4.2951,
+      "step": 241
+    },
+    {
+      "epoch": 0.27069351230425054,
+      "grad_norm": 2.609422206878662,
+      "learning_rate": 4.149190012931402e-05,
+      "loss": 4.1109,
+      "step": 242
+    },
+    {
+      "epoch": 0.27181208053691275,
+      "grad_norm": 2.932523250579834,
+      "learning_rate": 4.1425773149556134e-05,
+      "loss": 4.1605,
+      "step": 243
+    },
+    {
+      "epoch": 0.27293064876957496,
+      "grad_norm": 2.0494723320007324,
+      "learning_rate": 4.135944333144244e-05,
+      "loss": 4.2708,
+      "step": 244
+    },
+    {
+      "epoch": 0.2740492170022371,
+      "grad_norm": 2.5390541553497314,
+      "learning_rate": 4.129291149406567e-05,
+      "loss": 3.9803,
+      "step": 245
+    },
+    {
+      "epoch": 0.2751677852348993,
+      "grad_norm": 2.236856460571289,
+      "learning_rate": 4.122617845901322e-05,
+      "loss": 4.1316,
+      "step": 246
+    },
+    {
+      "epoch": 0.27628635346756153,
+      "grad_norm": 2.3391566276550293,
+      "learning_rate": 4.1159245050357065e-05,
+      "loss": 4.2234,
+      "step": 247
+    },
+    {
+      "epoch": 0.27740492170022374,
+      "grad_norm": 2.809293031692505,
+      "learning_rate": 4.1092112094643543e-05,
+      "loss": 4.0531,
+      "step": 248
+    },
+    {
+      "epoch": 0.2785234899328859,
+      "grad_norm": 2.943033456802368,
+      "learning_rate": 4.102478042088315e-05,
+      "loss": 3.9868,
+      "step": 249
+    },
+    {
+      "epoch": 0.2796420581655481,
+      "grad_norm": 1.9977362155914307,
+      "learning_rate": 4.095725086054029e-05,
+      "loss": 3.7822,
+      "step": 250
+    },
+    {
+      "epoch": 0.2807606263982103,
+      "grad_norm": 2.405503988265991,
+      "learning_rate": 4.088952424752307e-05,
+      "loss": 4.1802,
+      "step": 251
+    },
+    {
+      "epoch": 0.28187919463087246,
+      "grad_norm": 2.7163617610931396,
+      "learning_rate": 4.082160141817293e-05,
+      "loss": 4.1736,
+      "step": 252
+    },
+    {
+      "epoch": 0.2829977628635347,
+      "grad_norm": 2.2036311626434326,
+      "learning_rate": 4.075348321125433e-05,
+      "loss": 3.9607,
+      "step": 253
+    },
+    {
+      "epoch": 0.2841163310961969,
+      "grad_norm": 3.2602269649505615,
+      "learning_rate": 4.068517046794443e-05,
+      "loss": 3.984,
+      "step": 254
+    },
+    {
+      "epoch": 0.28523489932885904,
+      "grad_norm": 3.4510889053344727,
+      "learning_rate": 4.0616664031822684e-05,
+      "loss": 4.1701,
+      "step": 255
+    },
+    {
+      "epoch": 0.28635346756152125,
+      "grad_norm": 2.2935943603515625,
+      "learning_rate": 4.0547964748860386e-05,
+      "loss": 4.1199,
+      "step": 256
+    },
+    {
+      "epoch": 0.28747203579418346,
+      "grad_norm": 3.6087756156921387,
+      "learning_rate": 4.0479073467410286e-05,
+      "loss": 4.0622,
+      "step": 257
+    },
+    {
+      "epoch": 0.28859060402684567,
+      "grad_norm": 3.2882027626037598,
+      "learning_rate": 4.040999103819606e-05,
+      "loss": 4.2735,
+      "step": 258
+    },
+    {
+      "epoch": 0.2897091722595078,
+      "grad_norm": 2.9699313640594482,
+      "learning_rate": 4.034071831430184e-05,
+      "loss": 3.9871,
+      "step": 259
+    },
+    {
+      "epoch": 0.29082774049217003,
+      "grad_norm": 2.889160394668579,
+      "learning_rate": 4.0271256151161664e-05,
+      "loss": 4.0421,
+      "step": 260
+    },
+    {
+      "epoch": 0.29194630872483224,
+      "grad_norm": 2.320803165435791,
+      "learning_rate": 4.020160540654891e-05,
+      "loss": 4.4205,
+      "step": 261
+    },
+    {
+      "epoch": 0.2930648769574944,
+      "grad_norm": 3.7101683616638184,
+      "learning_rate": 4.0131766940565715e-05,
+      "loss": 3.9198,
+      "step": 262
+    },
+    {
+      "epoch": 0.2941834451901566,
+      "grad_norm": 2.4428420066833496,
+      "learning_rate": 4.006174161563233e-05,
+      "loss": 4.1809,
+      "step": 263
+    },
+    {
+      "epoch": 0.2953020134228188,
+      "grad_norm": 3.8638551235198975,
+      "learning_rate": 3.999153029647651e-05,
+      "loss": 3.9782,
+      "step": 264
+    },
+    {
+      "epoch": 0.29642058165548096,
+      "grad_norm": 1.7899916172027588,
+      "learning_rate": 3.9921133850122805e-05,
+      "loss": 4.3221,
+      "step": 265
+    },
+    {
+      "epoch": 0.2975391498881432,
+      "grad_norm": 3.8810582160949707,
+      "learning_rate": 3.9850553145881854e-05,
+      "loss": 3.8899,
+      "step": 266
+    },
+    {
+      "epoch": 0.2986577181208054,
+      "grad_norm": 2.2642791271209717,
+      "learning_rate": 3.9779789055339656e-05,
+      "loss": 4.0838,
+      "step": 267
+    },
+    {
+      "epoch": 0.29977628635346754,
+      "grad_norm": 2.4321177005767822,
+      "learning_rate": 3.9708842452346836e-05,
+      "loss": 4.2623,
+      "step": 268
+    },
+    {
+      "epoch": 0.30089485458612975,
+      "grad_norm": 2.04886794090271,
+      "learning_rate": 3.963771421300777e-05,
+      "loss": 4.1379,
+      "step": 269
+    },
+    {
+      "epoch": 0.30201342281879195,
+      "grad_norm": 1.9282593727111816,
+      "learning_rate": 3.956640521566989e-05,
+      "loss": 4.2569,
+      "step": 270
+    },
+    {
+      "epoch": 0.30313199105145416,
+      "grad_norm": 2.2058980464935303,
+      "learning_rate": 3.949491634091272e-05,
+      "loss": 4.7175,
+      "step": 271
+    },
+    {
+      "epoch": 0.3042505592841163,
+      "grad_norm": 2.039492130279541,
+      "learning_rate": 3.9423248471537065e-05,
+      "loss": 4.234,
+      "step": 272
+    },
+    {
+      "epoch": 0.3053691275167785,
+      "grad_norm": 1.9331101179122925,
+      "learning_rate": 3.935140249255412e-05,
+      "loss": 3.9734,
+      "step": 273
+    },
+    {
+      "epoch": 0.30648769574944074,
+      "grad_norm": 2.2190299034118652,
+      "learning_rate": 3.9279379291174465e-05,
+      "loss": 4.1752,
+      "step": 274
+    },
+    {
+      "epoch": 0.3076062639821029,
+      "grad_norm": 2.8630807399749756,
+      "learning_rate": 3.920717975679723e-05,
+      "loss": 4.0549,
+      "step": 275
+    },
+    {
+      "epoch": 0.3087248322147651,
+      "grad_norm": 2.7145142555236816,
+      "learning_rate": 3.913480478099897e-05,
+      "loss": 4.1861,
+      "step": 276
+    },
+    {
+      "epoch": 0.3098434004474273,
+      "grad_norm": 3.505417823791504,
+      "learning_rate": 3.9062255257522794e-05,
+      "loss": 4.0303,
+      "step": 277
+    },
+    {
+      "epoch": 0.31096196868008946,
+      "grad_norm": 3.250004291534424,
+      "learning_rate": 3.8989532082267225e-05,
+      "loss": 4.2068,
+      "step": 278
+    },
+    {
+      "epoch": 0.31208053691275167,
+      "grad_norm": 2.1549458503723145,
+      "learning_rate": 3.891663615327518e-05,
+      "loss": 3.921,
+      "step": 279
+    },
+    {
+      "epoch": 0.3131991051454139,
+      "grad_norm": 2.008209705352783,
+      "learning_rate": 3.884356837072288e-05,
+      "loss": 3.8706,
+      "step": 280
+    },
+    {
+      "epoch": 0.3143176733780761,
+      "grad_norm": 3.1446456909179688,
+      "learning_rate": 3.877032963690873e-05,
+      "loss": 4.0936,
+      "step": 281
+    },
+    {
+      "epoch": 0.31543624161073824,
+      "grad_norm": 2.9265387058258057,
+      "learning_rate": 3.8696920856242174e-05,
+      "loss": 4.2987,
+      "step": 282
+    },
+    {
+      "epoch": 0.31655480984340045,
+      "grad_norm": 2.246251344680786,
+      "learning_rate": 3.8623342935232525e-05,
+      "loss": 4.1966,
+      "step": 283
+    },
+    {
+      "epoch": 0.31767337807606266,
+      "grad_norm": 2.046858549118042,
+      "learning_rate": 3.854959678247778e-05,
+      "loss": 4.3037,
+      "step": 284
+    },
+    {
+      "epoch": 0.3187919463087248,
+      "grad_norm": 2.070281744003296,
+      "learning_rate": 3.847568330865338e-05,
+      "loss": 3.8673,
+      "step": 285
+    },
+    {
+      "epoch": 0.319910514541387,
+      "grad_norm": 2.2606728076934814,
+      "learning_rate": 3.8401603426501e-05,
+      "loss": 4.3719,
+      "step": 286
+    },
+    {
+      "epoch": 0.32102908277404923,
+      "grad_norm": 2.7562570571899414,
+      "learning_rate": 3.8327358050817234e-05,
+      "loss": 3.8372,
+      "step": 287
+    },
+    {
+      "epoch": 0.3221476510067114,
+      "grad_norm": 2.6188101768493652,
+      "learning_rate": 3.8252948098442344e-05,
+      "loss": 3.9952,
+      "step": 288
+    },
+    {
+      "epoch": 0.3232662192393736,
+      "grad_norm": 3.580763101577759,
+      "learning_rate": 3.817837448824888e-05,
+      "loss": 4.0735,
+      "step": 289
+    },
+    {
+      "epoch": 0.3243847874720358,
+      "grad_norm": 2.4125702381134033,
+      "learning_rate": 3.81036381411304e-05,
+      "loss": 4.0671,
+      "step": 290
+    },
+    {
+      "epoch": 0.32550335570469796,
+      "grad_norm": 2.0095221996307373,
+      "learning_rate": 3.8028739979990066e-05,
+      "loss": 4.1353,
+      "step": 291
+    },
+    {
+      "epoch": 0.32662192393736017,
+      "grad_norm": 2.5612494945526123,
+      "learning_rate": 3.7953680929729215e-05,
+      "loss": 3.9072,
+      "step": 292
+    },
+    {
+      "epoch": 0.3277404921700224,
+      "grad_norm": 2.8321776390075684,
+      "learning_rate": 3.787846191723599e-05,
+      "loss": 4.1067,
+      "step": 293
+    },
+    {
+      "epoch": 0.3288590604026846,
+      "grad_norm": 2.400554895401001,
+      "learning_rate": 3.780308387137387e-05,
+      "loss": 4.1338,
+      "step": 294
+    },
+    {
+      "epoch": 0.32997762863534674,
+      "grad_norm": 2.4202167987823486,
+      "learning_rate": 3.772754772297022e-05,
+      "loss": 4.139,
+      "step": 295
+    },
+    {
+      "epoch": 0.33109619686800895,
+      "grad_norm": 2.7618842124938965,
+      "learning_rate": 3.7651854404804755e-05,
+      "loss": 3.7908,
+      "step": 296
+    },
+    {
+      "epoch": 0.33221476510067116,
+      "grad_norm": 3.0693230628967285,
+      "learning_rate": 3.757600485159805e-05,
+      "loss": 3.9235,
+      "step": 297
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 1.86801016330719,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 4.1265,
+      "step": 298
+    },
+    {
+      "epoch": 0.3344519015659955,
+      "grad_norm": 2.0775158405303955,
+      "learning_rate": 3.742384078857824e-05,
+      "loss": 4.0009,
+      "step": 299
+    },
+    {
+      "epoch": 0.33557046979865773,
+      "grad_norm": 2.6333956718444824,
+      "learning_rate": 3.7347528157806586e-05,
+      "loss": 4.161,
+      "step": 300
+    },
+    {
+      "epoch": 0.3366890380313199,
+      "grad_norm": 2.197185754776001,
+      "learning_rate": 3.727106305005336e-05,
+      "loss": 3.9653,
+      "step": 301
+    },
+    {
+      "epoch": 0.3378076062639821,
+      "grad_norm": 2.323561191558838,
+      "learning_rate": 3.719444640956981e-05,
+      "loss": 4.0114,
+      "step": 302
+    },
+    {
+      "epoch": 0.3389261744966443,
+      "grad_norm": 2.5178816318511963,
+      "learning_rate": 3.7117679182478417e-05,
+      "loss": 3.8501,
+      "step": 303
+    },
+    {
+      "epoch": 0.3400447427293065,
+      "grad_norm": 2.5967633724212646,
+      "learning_rate": 3.704076231676125e-05,
+      "loss": 3.7106,
+      "step": 304
+    },
+    {
+      "epoch": 0.34116331096196867,
+      "grad_norm": 2.9387967586517334,
+      "learning_rate": 3.696369676224819e-05,
+      "loss": 3.9001,
+      "step": 305
+    },
+    {
+      "epoch": 0.3422818791946309,
+      "grad_norm": 2.7642838954925537,
+      "learning_rate": 3.688648347060529e-05,
+      "loss": 4.0384,
+      "step": 306
+    },
+    {
+      "epoch": 0.3434004474272931,
+      "grad_norm": 2.5087080001831055,
+      "learning_rate": 3.680912339532296e-05,
+      "loss": 4.1227,
+      "step": 307
+    },
+    {
+      "epoch": 0.34451901565995524,
+      "grad_norm": 2.084219217300415,
+      "learning_rate": 3.67316174917042e-05,
+      "loss": 4.0968,
+      "step": 308
+    },
+    {
+      "epoch": 0.34563758389261745,
+      "grad_norm": 3.1434361934661865,
+      "learning_rate": 3.66539667168528e-05,
+      "loss": 4.3648,
+      "step": 309
+    },
+    {
+      "epoch": 0.34675615212527966,
+      "grad_norm": 2.9753198623657227,
+      "learning_rate": 3.657617202966158e-05,
+      "loss": 3.8644,
+      "step": 310
+    },
+    {
+      "epoch": 0.3478747203579418,
+      "grad_norm": 1.990665316581726,
+      "learning_rate": 3.649823439080047e-05,
+      "loss": 4.0785,
+      "step": 311
+    },
+    {
+      "epoch": 0.348993288590604,
+      "grad_norm": 3.1740851402282715,
+      "learning_rate": 3.6420154762704686e-05,
+      "loss": 4.1794,
+      "step": 312
+    },
+    {
+      "epoch": 0.35011185682326623,
+      "grad_norm": 2.5907697677612305,
+      "learning_rate": 3.634193410956282e-05,
+      "loss": 3.887,
+      "step": 313
+    },
+    {
+      "epoch": 0.3512304250559284,
+      "grad_norm": 2.6295742988586426,
+      "learning_rate": 3.6263573397305e-05,
+      "loss": 4.0013,
+      "step": 314
+    },
+    {
+      "epoch": 0.3523489932885906,
+      "grad_norm": 3.465045213699341,
+      "learning_rate": 3.618507359359087e-05,
+      "loss": 3.9178,
+      "step": 315
+    },
+    {
+      "epoch": 0.3534675615212528,
+      "grad_norm": 2.5845601558685303,
+      "learning_rate": 3.6106435667797685e-05,
+      "loss": 3.9463,
+      "step": 316
+    },
+    {
+      "epoch": 0.354586129753915,
+      "grad_norm": 2.198699712753296,
+      "learning_rate": 3.602766059100838e-05,
+      "loss": 3.6981,
+      "step": 317
+    },
+    {
+      "epoch": 0.35570469798657717,
+      "grad_norm": 2.5952913761138916,
+      "learning_rate": 3.5948749335999496e-05,
+      "loss": 4.3058,
+      "step": 318
+    },
+    {
+      "epoch": 0.3568232662192394,
+      "grad_norm": 2.6661579608917236,
+      "learning_rate": 3.586970287722923e-05,
+      "loss": 3.9699,
+      "step": 319
+    },
+    {
+      "epoch": 0.3579418344519016,
+      "grad_norm": 2.2353670597076416,
+      "learning_rate": 3.5790522190825365e-05,
+      "loss": 3.9394,
+      "step": 320
+    },
+    {
+      "epoch": 0.35906040268456374,
+      "grad_norm": 1.800933599472046,
+      "learning_rate": 3.571120825457327e-05,
+      "loss": 3.9037,
+      "step": 321
+    },
+    {
+      "epoch": 0.36017897091722595,
+      "grad_norm": 2.2671263217926025,
+      "learning_rate": 3.563176204790374e-05,
+      "loss": 4.2906,
+      "step": 322
+    },
+    {
+      "epoch": 0.36129753914988816,
+      "grad_norm": 2.249171733856201,
+      "learning_rate": 3.555218455188099e-05,
+      "loss": 4.0766,
+      "step": 323
+    },
+    {
+      "epoch": 0.3624161073825503,
+      "grad_norm": 2.222735643386841,
+      "learning_rate": 3.547247674919046e-05,
+      "loss": 4.0171,
+      "step": 324
+    },
+    {
+      "epoch": 0.3635346756152125,
+      "grad_norm": 2.537754774093628,
+      "learning_rate": 3.539263962412676e-05,
+      "loss": 3.9333,
+      "step": 325
+    },
+    {
+      "epoch": 0.36465324384787473,
+      "grad_norm": 3.2057714462280273,
+      "learning_rate": 3.5312674162581436e-05,
+      "loss": 4.1573,
+      "step": 326
+    },
+    {
+      "epoch": 0.36577181208053694,
+      "grad_norm": 3.4859671592712402,
+      "learning_rate": 3.523258135203087e-05,
+      "loss": 3.8762,
+      "step": 327
+    },
+    {
+      "epoch": 0.3668903803131991,
+      "grad_norm": 3.155507802963257,
+      "learning_rate": 3.5152362181524014e-05,
+      "loss": 3.9051,
+      "step": 328
+    },
+    {
+      "epoch": 0.3680089485458613,
+      "grad_norm": 2.1552867889404297,
+      "learning_rate": 3.507201764167024e-05,
+      "loss": 3.8245,
+      "step": 329
+    },
+    {
+      "epoch": 0.3691275167785235,
+      "grad_norm": 2.492128372192383,
+      "learning_rate": 3.4991548724627054e-05,
+      "loss": 3.9806,
+      "step": 330
+    },
+    {
+      "epoch": 0.37024608501118567,
+      "grad_norm": 2.913975715637207,
+      "learning_rate": 3.491095642408791e-05,
+      "loss": 4.0708,
+      "step": 331
+    },
+    {
+      "epoch": 0.3713646532438479,
+      "grad_norm": 1.982664942741394,
+      "learning_rate": 3.483024173526985e-05,
+      "loss": 3.7897,
+      "step": 332
+    },
+    {
+      "epoch": 0.3724832214765101,
+      "grad_norm": 2.7133326530456543,
+      "learning_rate": 3.4749405654901294e-05,
+      "loss": 4.0167,
+      "step": 333
+    },
+    {
+      "epoch": 0.37360178970917224,
+      "grad_norm": 2.769604444503784,
+      "learning_rate": 3.46684491812097e-05,
+      "loss": 4.2449,
+      "step": 334
+    },
+    {
+      "epoch": 0.37472035794183445,
+      "grad_norm": 2.3599345684051514,
+      "learning_rate": 3.45873733139092e-05,
+      "loss": 4.0715,
+      "step": 335
+    },
+    {
+      "epoch": 0.37583892617449666,
+      "grad_norm": 3.5112671852111816,
+      "learning_rate": 3.4506179054188344e-05,
+      "loss": 3.896,
+      "step": 336
+    },
+    {
+      "epoch": 0.3769574944071588,
+      "grad_norm": 2.5493431091308594,
+      "learning_rate": 3.442486740469766e-05,
+      "loss": 3.8645,
+      "step": 337
+    },
+    {
+      "epoch": 0.378076062639821,
+      "grad_norm": 2.665292501449585,
+      "learning_rate": 3.434343936953729e-05,
+      "loss": 3.9132,
+      "step": 338
+    },
+    {
+      "epoch": 0.37919463087248323,
+      "grad_norm": 2.7781031131744385,
+      "learning_rate": 3.426189595424461e-05,
+      "loss": 4.0039,
+      "step": 339
+    },
+    {
+      "epoch": 0.38031319910514544,
+      "grad_norm": 3.5199902057647705,
+      "learning_rate": 3.41802381657818e-05,
+      "loss": 4.0221,
+      "step": 340
+    },
+    {
+      "epoch": 0.3814317673378076,
+      "grad_norm": 2.529172897338867,
+      "learning_rate": 3.4098467012523404e-05,
+      "loss": 3.9565,
+      "step": 341
+    },
+    {
+      "epoch": 0.3825503355704698,
+      "grad_norm": 3.5351314544677734,
+      "learning_rate": 3.401658350424389e-05,
+      "loss": 4.2232,
+      "step": 342
+    },
+    {
+      "epoch": 0.383668903803132,
+      "grad_norm": 2.705423593521118,
+      "learning_rate": 3.393458865210516e-05,
+      "loss": 3.8003,
+      "step": 343
+    },
+    {
+      "epoch": 0.38478747203579416,
+      "grad_norm": 2.4735591411590576,
+      "learning_rate": 3.38524834686441e-05,
+      "loss": 3.9762,
+      "step": 344
+    },
+    {
+      "epoch": 0.3859060402684564,
+      "grad_norm": 3.4464943408966064,
+      "learning_rate": 3.377026896776003e-05,
+      "loss": 4.2064,
+      "step": 345
+    },
+    {
+      "epoch": 0.3870246085011186,
+      "grad_norm": 2.1392390727996826,
+      "learning_rate": 3.368794616470222e-05,
+      "loss": 3.9257,
+      "step": 346
+    },
+    {
+      "epoch": 0.38814317673378074,
+      "grad_norm": 3.5217792987823486,
+      "learning_rate": 3.360551607605735e-05,
+      "loss": 3.7378,
+      "step": 347
+    },
+    {
+      "epoch": 0.38926174496644295,
+      "grad_norm": 2.3355872631073,
+      "learning_rate": 3.3522979719736926e-05,
+      "loss": 4.3103,
+      "step": 348
+    },
+    {
+      "epoch": 0.39038031319910516,
+      "grad_norm": 1.909463882446289,
+      "learning_rate": 3.344033811496475e-05,
+      "loss": 3.9676,
+      "step": 349
+    },
+    {
+      "epoch": 0.39149888143176736,
+      "grad_norm": 2.3093831539154053,
+      "learning_rate": 3.3357592282264296e-05,
+      "loss": 3.915,
+      "step": 350
+    },
+    {
+      "epoch": 0.3926174496644295,
+      "grad_norm": 2.086834192276001,
+      "learning_rate": 3.327474324344614e-05,
+      "loss": 3.9272,
+      "step": 351
+    },
+    {
+      "epoch": 0.39373601789709173,
+      "grad_norm": 2.3872201442718506,
+      "learning_rate": 3.3191792021595316e-05,
+      "loss": 3.8731,
+      "step": 352
+    },
+    {
+      "epoch": 0.39485458612975394,
+      "grad_norm": 2.4958016872406006,
+      "learning_rate": 3.310873964105872e-05,
+      "loss": 3.9696,
+      "step": 353
+    },
+    {
+      "epoch": 0.3959731543624161,
+      "grad_norm": 2.3622846603393555,
+      "learning_rate": 3.302558712743241e-05,
+      "loss": 3.6666,
+      "step": 354
+    },
+    {
+      "epoch": 0.3970917225950783,
+      "grad_norm": 3.684812068939209,
+      "learning_rate": 3.2942335507548966e-05,
+      "loss": 3.9342,
+      "step": 355
+    },
+    {
+      "epoch": 0.3982102908277405,
+      "grad_norm": 2.6267335414886475,
+      "learning_rate": 3.285898580946482e-05,
+      "loss": 4.0503,
+      "step": 356
+    },
+    {
+      "epoch": 0.39932885906040266,
+      "grad_norm": 2.092944860458374,
+      "learning_rate": 3.277553906244756e-05,
+      "loss": 4.1497,
+      "step": 357
+    },
+    {
+      "epoch": 0.4004474272930649,
+      "grad_norm": 2.3737025260925293,
+      "learning_rate": 3.2691996296963186e-05,
+      "loss": 3.8479,
+      "step": 358
+    },
+    {
+      "epoch": 0.4015659955257271,
+      "grad_norm": 2.4298672676086426,
+      "learning_rate": 3.260835854466342e-05,
+      "loss": 4.0085,
+      "step": 359
+    },
+    {
+      "epoch": 0.40268456375838924,
+      "grad_norm": 2.19952130317688,
+      "learning_rate": 3.252462683837297e-05,
+      "loss": 4.4371,
+      "step": 360
+    },
+    {
+      "epoch": 0.40380313199105144,
+      "grad_norm": 2.4244048595428467,
+      "learning_rate": 3.244080221207674e-05,
+      "loss": 4.2144,
+      "step": 361
+    },
+    {
+      "epoch": 0.40492170022371365,
+      "grad_norm": 2.4055869579315186,
+      "learning_rate": 3.23568857009071e-05,
+      "loss": 3.9106,
+      "step": 362
+    },
+    {
+      "epoch": 0.40604026845637586,
+      "grad_norm": 2.3508224487304688,
+      "learning_rate": 3.2272878341131075e-05,
+      "loss": 4.237,
+      "step": 363
+    },
+    {
+      "epoch": 0.407158836689038,
+      "grad_norm": 2.4762392044067383,
+      "learning_rate": 3.218878117013756e-05,
+      "loss": 4.0447,
+      "step": 364
+    },
+    {
+      "epoch": 0.4082774049217002,
+      "grad_norm": 2.0015032291412354,
+      "learning_rate": 3.210459522642452e-05,
+      "loss": 4.0781,
+      "step": 365
+    },
+    {
+      "epoch": 0.40939597315436244,
+      "grad_norm": 2.745539665222168,
+      "learning_rate": 3.2020321549586154e-05,
+      "loss": 3.8968,
+      "step": 366
+    },
+    {
+      "epoch": 0.4105145413870246,
+      "grad_norm": 2.3843181133270264,
+      "learning_rate": 3.193596118030005e-05,
+      "loss": 4.0991,
+      "step": 367
+    },
+    {
+      "epoch": 0.4116331096196868,
+      "grad_norm": 2.248856782913208,
+      "learning_rate": 3.185151516031434e-05,
+      "loss": 4.034,
+      "step": 368
+    },
+    {
+      "epoch": 0.412751677852349,
+      "grad_norm": 2.3670918941497803,
+      "learning_rate": 3.1766984532434853e-05,
+      "loss": 4.0161,
+      "step": 369
+    },
+    {
+      "epoch": 0.41387024608501116,
+      "grad_norm": 2.298100709915161,
+      "learning_rate": 3.1682370340512217e-05,
+      "loss": 4.1546,
+      "step": 370
+    },
+    {
+      "epoch": 0.41498881431767337,
+      "grad_norm": 1.788293719291687,
+      "learning_rate": 3.159767362942896e-05,
+      "loss": 4.1429,
+      "step": 371
+    },
+    {
+      "epoch": 0.4161073825503356,
+      "grad_norm": 2.500016212463379,
+      "learning_rate": 3.1512895445086636e-05,
+      "loss": 3.7988,
+      "step": 372
+    },
+    {
+      "epoch": 0.4172259507829978,
+      "grad_norm": 3.096670627593994,
+      "learning_rate": 3.14280368343929e-05,
+      "loss": 4.406,
+      "step": 373
+    },
+    {
+      "epoch": 0.41834451901565994,
+      "grad_norm": 2.44936466217041,
+      "learning_rate": 3.134309884524856e-05,
+      "loss": 4.1445,
+      "step": 374
+    },
+    {
+      "epoch": 0.41946308724832215,
+      "grad_norm": 2.5851917266845703,
+      "learning_rate": 3.125808252653466e-05,
+      "loss": 4.6243,
+      "step": 375
+    },
+    {
+      "epoch": 0.42058165548098436,
+      "grad_norm": 2.472642183303833,
+      "learning_rate": 3.1172988928099525e-05,
+      "loss": 4.2828,
+      "step": 376
+    },
+    {
+      "epoch": 0.4217002237136465,
+      "grad_norm": 2.0510940551757812,
+      "learning_rate": 3.108781910074578e-05,
+      "loss": 3.7741,
+      "step": 377
+    },
+    {
+      "epoch": 0.4228187919463087,
+      "grad_norm": 2.4596452713012695,
+      "learning_rate": 3.100257409621738e-05,
+      "loss": 3.974,
+      "step": 378
+    },
+    {
+      "epoch": 0.42393736017897093,
+      "grad_norm": 2.0561556816101074,
+      "learning_rate": 3.0917254967186635e-05,
+      "loss": 3.8241,
+      "step": 379
+    },
+    {
+      "epoch": 0.4250559284116331,
+      "grad_norm": 1.8558775186538696,
+      "learning_rate": 3.0831862767241205e-05,
+      "loss": 4.1329,
+      "step": 380
+    },
+    {
+      "epoch": 0.4261744966442953,
+      "grad_norm": 2.460034132003784,
+      "learning_rate": 3.074639855087109e-05,
+      "loss": 4.0921,
+      "step": 381
+    },
+    {
+      "epoch": 0.4272930648769575,
+      "grad_norm": 2.0531368255615234,
+      "learning_rate": 3.0660863373455595e-05,
+      "loss": 3.8236,
+      "step": 382
+    },
+    {
+      "epoch": 0.42841163310961966,
+      "grad_norm": 2.9021363258361816,
+      "learning_rate": 3.057525829125032e-05,
+      "loss": 3.8496,
+      "step": 383
+    },
+    {
+      "epoch": 0.42953020134228187,
+      "grad_norm": 2.5614798069000244,
+      "learning_rate": 3.0489584361374074e-05,
+      "loss": 4.1064,
+      "step": 384
+    },
+    {
+      "epoch": 0.4306487695749441,
+      "grad_norm": 2.2442526817321777,
+      "learning_rate": 3.04038426417959e-05,
+      "loss": 4.1327,
+      "step": 385
+    },
+    {
+      "epoch": 0.4317673378076063,
+      "grad_norm": 2.7794744968414307,
+      "learning_rate": 3.031803419132192e-05,
+      "loss": 4.1708,
+      "step": 386
+    },
+    {
+      "epoch": 0.43288590604026844,
+      "grad_norm": 2.464383363723755,
+      "learning_rate": 3.0232160069582332e-05,
+      "loss": 4.1961,
+      "step": 387
+    },
+    {
+      "epoch": 0.43400447427293065,
+      "grad_norm": 1.990610122680664,
+      "learning_rate": 3.014622133701826e-05,
+      "loss": 3.8704,
+      "step": 388
+    },
+    {
+      "epoch": 0.43512304250559286,
+      "grad_norm": 2.630527973175049,
+      "learning_rate": 3.0060219054868727e-05,
+      "loss": 4.0868,
+      "step": 389
+    },
+    {
+      "epoch": 0.436241610738255,
+      "grad_norm": 1.9832707643508911,
+      "learning_rate": 2.9974154285157497e-05,
+      "loss": 4.0184,
+      "step": 390
+    },
+    {
+      "epoch": 0.4373601789709172,
+      "grad_norm": 2.482058048248291,
+      "learning_rate": 2.9888028090679982e-05,
+      "loss": 4.0066,
+      "step": 391
+    },
+    {
+      "epoch": 0.43847874720357943,
+      "grad_norm": 2.484102487564087,
+      "learning_rate": 2.9801841534990115e-05,
+      "loss": 4.1228,
+      "step": 392
+    },
+    {
+      "epoch": 0.4395973154362416,
+      "grad_norm": 2.7631897926330566,
+      "learning_rate": 2.9715595682387242e-05,
+      "loss": 3.863,
+      "step": 393
+    },
+    {
+      "epoch": 0.4407158836689038,
+      "grad_norm": 2.0470075607299805,
+      "learning_rate": 2.9629291597902898e-05,
+      "loss": 3.8224,
+      "step": 394
+    },
+    {
+      "epoch": 0.441834451901566,
+      "grad_norm": 1.8407243490219116,
+      "learning_rate": 2.954293034728776e-05,
+      "loss": 3.8581,
+      "step": 395
+    },
+    {
+      "epoch": 0.4429530201342282,
+      "grad_norm": 1.9803217649459839,
+      "learning_rate": 2.9456512996998424e-05,
+      "loss": 3.914,
+      "step": 396
+    },
+    {
+      "epoch": 0.44407158836689037,
+      "grad_norm": 2.7368860244750977,
+      "learning_rate": 2.9370040614184245e-05,
+      "loss": 4.0718,
+      "step": 397
+    },
+    {
+      "epoch": 0.4451901565995526,
+      "grad_norm": 5.358017444610596,
+      "learning_rate": 2.9283514266674168e-05,
+      "loss": 3.7782,
+      "step": 398
+    },
+    {
+      "epoch": 0.4463087248322148,
+      "grad_norm": 2.5471818447113037,
+      "learning_rate": 2.9196935022963525e-05,
+      "loss": 3.7342,
+      "step": 399
+    },
+    {
+      "epoch": 0.44742729306487694,
+      "grad_norm": 2.3943824768066406,
+      "learning_rate": 2.9110303952200863e-05,
+      "loss": 4.0393,
+      "step": 400
+    },
+    {
+      "epoch": 0.44854586129753915,
+      "grad_norm": 2.130612373352051,
+      "learning_rate": 2.902362212417472e-05,
+      "loss": 3.7082,
+      "step": 401
+    },
+    {
+      "epoch": 0.44966442953020136,
+      "grad_norm": 2.9800944328308105,
+      "learning_rate": 2.893689060930045e-05,
+      "loss": 3.8505,
+      "step": 402
+    },
+    {
+      "epoch": 0.4507829977628635,
+      "grad_norm": 2.35976243019104,
+      "learning_rate": 2.8850110478606938e-05,
+      "loss": 4.1292,
+      "step": 403
+    },
+    {
+      "epoch": 0.4519015659955257,
+      "grad_norm": 2.627814292907715,
+      "learning_rate": 2.876328280372346e-05,
+      "loss": 3.6394,
+      "step": 404
+    },
+    {
+      "epoch": 0.45302013422818793,
+      "grad_norm": 2.3630151748657227,
+      "learning_rate": 2.8676408656866353e-05,
+      "loss": 3.7281,
+      "step": 405
+    },
+    {
+      "epoch": 0.4541387024608501,
+      "grad_norm": 2.5293986797332764,
+      "learning_rate": 2.8589489110825897e-05,
+      "loss": 3.7605,
+      "step": 406
+    },
+    {
+      "epoch": 0.4552572706935123,
+      "grad_norm": 2.503148078918457,
+      "learning_rate": 2.8502525238952916e-05,
+      "loss": 4.1672,
+      "step": 407
+    },
+    {
+      "epoch": 0.4563758389261745,
+      "grad_norm": 2.9510624408721924,
+      "learning_rate": 2.8415518115145674e-05,
+      "loss": 3.9833,
+      "step": 408
+    },
+    {
+      "epoch": 0.4574944071588367,
+      "grad_norm": 2.3186118602752686,
+      "learning_rate": 2.8328468813836493e-05,
+      "loss": 4.1216,
+      "step": 409
+    },
+    {
+      "epoch": 0.45861297539149887,
+      "grad_norm": 3.173356294631958,
+      "learning_rate": 2.824137840997858e-05,
+      "loss": 3.9926,
+      "step": 410
+    },
+    {
+      "epoch": 0.4597315436241611,
+      "grad_norm": 1.9115232229232788,
+      "learning_rate": 2.8154247979032665e-05,
+      "loss": 4.0426,
+      "step": 411
+    },
+    {
+      "epoch": 0.4608501118568233,
+      "grad_norm": 2.5544466972351074,
+      "learning_rate": 2.8067078596953796e-05,
+      "loss": 3.9857,
+      "step": 412
+    },
+    {
+      "epoch": 0.46196868008948544,
+      "grad_norm": 1.8539265394210815,
+      "learning_rate": 2.7979871340178003e-05,
+      "loss": 3.894,
+      "step": 413
+    },
+    {
+      "epoch": 0.46308724832214765,
+      "grad_norm": 2.6608541011810303,
+      "learning_rate": 2.7892627285609035e-05,
+      "loss": 4.2197,
+      "step": 414
+    },
+    {
+      "epoch": 0.46420581655480986,
+      "grad_norm": 2.6225533485412598,
+      "learning_rate": 2.780534751060504e-05,
+      "loss": 3.8299,
+      "step": 415
+    },
+    {
+      "epoch": 0.465324384787472,
+      "grad_norm": 3.5249505043029785,
+      "learning_rate": 2.771803309296527e-05,
+      "loss": 3.4724,
+      "step": 416
+    },
+    {
+      "epoch": 0.4664429530201342,
+      "grad_norm": 2.1255197525024414,
+      "learning_rate": 2.7630685110916778e-05,
+      "loss": 4.12,
+      "step": 417
+    },
+    {
+      "epoch": 0.46756152125279643,
+      "grad_norm": 2.8229820728302,
+      "learning_rate": 2.754330464310108e-05,
+      "loss": 4.3171,
+      "step": 418
+    },
+    {
+      "epoch": 0.46868008948545864,
+      "grad_norm": 2.2736423015594482,
+      "learning_rate": 2.7455892768560888e-05,
+      "loss": 3.9784,
+      "step": 419
+    },
+    {
+      "epoch": 0.4697986577181208,
+      "grad_norm": 1.8942090272903442,
+      "learning_rate": 2.736845056672671e-05,
+      "loss": 3.7484,
+      "step": 420
+    },
+    {
+      "epoch": 0.470917225950783,
+      "grad_norm": 3.5483438968658447,
+      "learning_rate": 2.7280979117403575e-05,
+      "loss": 4.1573,
+      "step": 421
+    },
+    {
+      "epoch": 0.4720357941834452,
+      "grad_norm": 2.8570756912231445,
+      "learning_rate": 2.7193479500757685e-05,
+      "loss": 3.8323,
+      "step": 422
+    },
+    {
+      "epoch": 0.47315436241610737,
+      "grad_norm": 2.425438404083252,
+      "learning_rate": 2.710595279730308e-05,
+      "loss": 4.0871,
+      "step": 423
+    },
+    {
+      "epoch": 0.4742729306487696,
+      "grad_norm": 2.321178674697876,
+      "learning_rate": 2.7018400087888263e-05,
+      "loss": 4.1138,
+      "step": 424
+    },
+    {
+      "epoch": 0.4753914988814318,
+      "grad_norm": 2.9220635890960693,
+      "learning_rate": 2.6930822453682915e-05,
+      "loss": 3.8734,
+      "step": 425
+    },
+    {
+      "epoch": 0.47651006711409394,
+      "grad_norm": 3.5046780109405518,
+      "learning_rate": 2.684322097616448e-05,
+      "loss": 4.3458,
+      "step": 426
+    },
+    {
+      "epoch": 0.47762863534675615,
+      "grad_norm": 3.008169651031494,
+      "learning_rate": 2.675559673710485e-05,
+      "loss": 3.7159,
+      "step": 427
+    },
+    {
+      "epoch": 0.47874720357941836,
+      "grad_norm": 2.289055109024048,
+      "learning_rate": 2.6667950818556993e-05,
+      "loss": 4.2936,
+      "step": 428
+    },
+    {
+      "epoch": 0.4798657718120805,
+      "grad_norm": 1.8744932413101196,
+      "learning_rate": 2.658028430284159e-05,
+      "loss": 3.9657,
+      "step": 429
+    },
+    {
+      "epoch": 0.4809843400447427,
+      "grad_norm": 2.0840201377868652,
+      "learning_rate": 2.649259827253368e-05,
+      "loss": 4.265,
+      "step": 430
+    },
+    {
+      "epoch": 0.48210290827740493,
+      "grad_norm": 2.352440118789673,
+      "learning_rate": 2.6404893810449272e-05,
+      "loss": 3.7011,
+      "step": 431
+    },
+    {
+      "epoch": 0.48322147651006714,
+      "grad_norm": 2.6247146129608154,
+      "learning_rate": 2.631717199963199e-05,
+      "loss": 3.5364,
+      "step": 432
+    },
+    {
+      "epoch": 0.4843400447427293,
+      "grad_norm": 3.915102243423462,
+      "learning_rate": 2.6229433923339696e-05,
+      "loss": 4.2285,
+      "step": 433
+    },
+    {
+      "epoch": 0.4854586129753915,
+      "grad_norm": 2.190736770629883,
+      "learning_rate": 2.6141680665031116e-05,
+      "loss": 3.9863,
+      "step": 434
+    },
+    {
+      "epoch": 0.4865771812080537,
+      "grad_norm": 2.417031764984131,
+      "learning_rate": 2.6053913308352428e-05,
+      "loss": 3.9451,
+      "step": 435
+    },
+    {
+      "epoch": 0.48769574944071586,
+      "grad_norm": 2.4824278354644775,
+      "learning_rate": 2.596613293712396e-05,
+      "loss": 3.9539,
+      "step": 436
+    },
+    {
+      "epoch": 0.4888143176733781,
+      "grad_norm": 2.9509193897247314,
+      "learning_rate": 2.5878340635326686e-05,
+      "loss": 3.8236,
+      "step": 437
+    },
+    {
+      "epoch": 0.4899328859060403,
+      "grad_norm": 2.627767562866211,
+      "learning_rate": 2.5790537487088974e-05,
+      "loss": 4.0342,
+      "step": 438
+    },
+    {
+      "epoch": 0.49105145413870244,
+      "grad_norm": 2.270059108734131,
+      "learning_rate": 2.5702724576673088e-05,
+      "loss": 4.3256,
+      "step": 439
+    },
+    {
+      "epoch": 0.49217002237136465,
+      "grad_norm": 2.08921217918396,
+      "learning_rate": 2.561490298846186e-05,
+      "loss": 3.8144,
+      "step": 440
+    },
+    {
+      "epoch": 0.49328859060402686,
+      "grad_norm": 2.151421308517456,
+      "learning_rate": 2.5527073806945278e-05,
+      "loss": 3.881,
+      "step": 441
+    },
+    {
+      "epoch": 0.49440715883668906,
+      "grad_norm": 3.5347208976745605,
+      "learning_rate": 2.5439238116707102e-05,
+      "loss": 4.1678,
+      "step": 442
+    },
+    {
+      "epoch": 0.4955257270693512,
+      "grad_norm": 1.9540199041366577,
+      "learning_rate": 2.5351397002411477e-05,
+      "loss": 4.028,
+      "step": 443
+    },
+    {
+      "epoch": 0.4966442953020134,
+      "grad_norm": 2.151306629180908,
+      "learning_rate": 2.5263551548789495e-05,
+      "loss": 3.6328,
+      "step": 444
+    },
+    {
+      "epoch": 0.49776286353467564,
+      "grad_norm": 3.1425187587738037,
+      "learning_rate": 2.517570284062586e-05,
+      "loss": 3.9881,
+      "step": 445
+    },
+    {
+      "epoch": 0.4988814317673378,
+      "grad_norm": 2.3265583515167236,
+      "learning_rate": 2.5087851962745468e-05,
+      "loss": 3.9283,
+      "step": 446
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.091590166091919,
+      "learning_rate": 2.5e-05,
+      "loss": 3.6892,
+      "step": 447
+    },
+    {
+      "epoch": 0.5011185682326622,
+      "grad_norm": 2.2790603637695312,
+      "learning_rate": 2.4912148037254535e-05,
+      "loss": 3.7534,
+      "step": 448
+    },
+    {
+      "epoch": 0.5022371364653244,
+      "grad_norm": 2.3313305377960205,
+      "learning_rate": 2.4824297159374142e-05,
+      "loss": 3.8916,
+      "step": 449
+    },
+    {
+      "epoch": 0.5033557046979866,
+      "grad_norm": 2.9091544151306152,
+      "learning_rate": 2.473644845121051e-05,
+      "loss": 4.1545,
+      "step": 450
+    },
+    {
+      "epoch": 0.5044742729306487,
+      "grad_norm": 1.9648659229278564,
+      "learning_rate": 2.464860299758854e-05,
+      "loss": 3.9658,
+      "step": 451
+    },
+    {
+      "epoch": 0.5055928411633109,
+      "grad_norm": 2.6131553649902344,
+      "learning_rate": 2.45607618832929e-05,
+      "loss": 3.9665,
+      "step": 452
+    },
+    {
+      "epoch": 0.5067114093959731,
+      "grad_norm": 2.341771364212036,
+      "learning_rate": 2.4472926193054728e-05,
+      "loss": 3.9947,
+      "step": 453
+    },
+    {
+      "epoch": 0.5078299776286354,
+      "grad_norm": 2.526048183441162,
+      "learning_rate": 2.4385097011538144e-05,
+      "loss": 4.0324,
+      "step": 454
+    },
+    {
+      "epoch": 0.5089485458612976,
+      "grad_norm": 3.4752910137176514,
+      "learning_rate": 2.4297275423326918e-05,
+      "loss": 4.0246,
+      "step": 455
+    },
+    {
+      "epoch": 0.5100671140939598,
+      "grad_norm": 2.2056233882904053,
+      "learning_rate": 2.420946251291103e-05,
+      "loss": 3.8516,
+      "step": 456
+    },
+    {
+      "epoch": 0.5111856823266219,
+      "grad_norm": 2.7161202430725098,
+      "learning_rate": 2.412165936467332e-05,
+      "loss": 4.2025,
+      "step": 457
+    },
+    {
+      "epoch": 0.5123042505592841,
+      "grad_norm": 2.7404301166534424,
+      "learning_rate": 2.4033867062876052e-05,
+      "loss": 3.7109,
+      "step": 458
+    },
+    {
+      "epoch": 0.5134228187919463,
+      "grad_norm": 3.4687747955322266,
+      "learning_rate": 2.3946086691647575e-05,
+      "loss": 3.8982,
+      "step": 459
+    },
+    {
+      "epoch": 0.5145413870246085,
+      "grad_norm": 2.116055488586426,
+      "learning_rate": 2.3858319334968893e-05,
+      "loss": 3.9515,
+      "step": 460
+    },
+    {
+      "epoch": 0.5156599552572707,
+      "grad_norm": 2.8268518447875977,
+      "learning_rate": 2.377056607666031e-05,
+      "loss": 4.4184,
+      "step": 461
+    },
+    {
+      "epoch": 0.5167785234899329,
+      "grad_norm": 2.4674103260040283,
+      "learning_rate": 2.368282800036801e-05,
+      "loss": 4.0095,
+      "step": 462
+    },
+    {
+      "epoch": 0.5178970917225951,
+      "grad_norm": 2.0366992950439453,
+      "learning_rate": 2.359510618955073e-05,
+      "loss": 4.4928,
+      "step": 463
+    },
+    {
+      "epoch": 0.5190156599552572,
+      "grad_norm": 2.1218111515045166,
+      "learning_rate": 2.350740172746633e-05,
+      "loss": 4.149,
+      "step": 464
+    },
+    {
+      "epoch": 0.5201342281879194,
+      "grad_norm": 3.7365167140960693,
+      "learning_rate": 2.3419715697158416e-05,
+      "loss": 3.6304,
+      "step": 465
+    },
+    {
+      "epoch": 0.5212527964205816,
+      "grad_norm": 2.4339170455932617,
+      "learning_rate": 2.3332049181443016e-05,
+      "loss": 3.9618,
+      "step": 466
+    },
+    {
+      "epoch": 0.5223713646532439,
+      "grad_norm": 2.091418981552124,
+      "learning_rate": 2.3244403262895153e-05,
+      "loss": 4.3328,
+      "step": 467
+    },
+    {
+      "epoch": 0.5234899328859061,
+      "grad_norm": 2.5647706985473633,
+      "learning_rate": 2.3156779023835525e-05,
+      "loss": 3.9168,
+      "step": 468
+    },
+    {
+      "epoch": 0.5246085011185683,
+      "grad_norm": 3.0460050106048584,
+      "learning_rate": 2.3069177546317087e-05,
+      "loss": 3.6819,
+      "step": 469
+    },
+    {
+      "epoch": 0.5257270693512305,
+      "grad_norm": 1.7986963987350464,
+      "learning_rate": 2.2981599912111736e-05,
+      "loss": 4.0007,
+      "step": 470
+    },
+    {
+      "epoch": 0.5268456375838926,
+      "grad_norm": 2.6966090202331543,
+      "learning_rate": 2.289404720269693e-05,
+      "loss": 4.2807,
+      "step": 471
+    },
+    {
+      "epoch": 0.5279642058165548,
+      "grad_norm": 2.5532920360565186,
+      "learning_rate": 2.280652049924232e-05,
+      "loss": 4.1665,
+      "step": 472
+    },
+    {
+      "epoch": 0.529082774049217,
+      "grad_norm": 2.1815285682678223,
+      "learning_rate": 2.2719020882596427e-05,
+      "loss": 3.6647,
+      "step": 473
+    },
+    {
+      "epoch": 0.5302013422818792,
+      "grad_norm": 2.97127103805542,
+      "learning_rate": 2.2631549433273293e-05,
+      "loss": 3.8236,
+      "step": 474
+    },
+    {
+      "epoch": 0.5313199105145414,
+      "grad_norm": 2.294546365737915,
+      "learning_rate": 2.2544107231439114e-05,
+      "loss": 3.6576,
+      "step": 475
+    },
+    {
+      "epoch": 0.5324384787472036,
+      "grad_norm": 2.5872814655303955,
+      "learning_rate": 2.2456695356898916e-05,
+      "loss": 4.077,
+      "step": 476
+    },
+    {
+      "epoch": 0.5335570469798657,
+      "grad_norm": 2.548797369003296,
+      "learning_rate": 2.2369314889083235e-05,
+      "loss": 3.7512,
+      "step": 477
+    },
+    {
+      "epoch": 0.5346756152125279,
+      "grad_norm": 2.8900959491729736,
+      "learning_rate": 2.2281966907034733e-05,
+      "loss": 4.3499,
+      "step": 478
+    },
+    {
+      "epoch": 0.5357941834451901,
+      "grad_norm": 3.7104897499084473,
+      "learning_rate": 2.2194652489394967e-05,
+      "loss": 3.5961,
+      "step": 479
+    },
+    {
+      "epoch": 0.5369127516778524,
+      "grad_norm": 2.2205984592437744,
+      "learning_rate": 2.2107372714390974e-05,
+      "loss": 3.6201,
+      "step": 480
+    },
+    {
+      "epoch": 0.5380313199105146,
+      "grad_norm": 1.851414442062378,
+      "learning_rate": 2.2020128659822e-05,
+      "loss": 4.0772,
+      "step": 481
+    },
+    {
+      "epoch": 0.5391498881431768,
+      "grad_norm": 2.894435167312622,
+      "learning_rate": 2.1932921403046207e-05,
+      "loss": 3.6691,
+      "step": 482
+    },
+    {
+      "epoch": 0.540268456375839,
+      "grad_norm": 2.332059621810913,
+      "learning_rate": 2.1845752020967337e-05,
+      "loss": 3.7673,
+      "step": 483
+    },
+    {
+      "epoch": 0.5413870246085011,
+      "grad_norm": 2.175400972366333,
+      "learning_rate": 2.1758621590021426e-05,
+      "loss": 3.8803,
+      "step": 484
+    },
+    {
+      "epoch": 0.5425055928411633,
+      "grad_norm": 2.4179916381835938,
+      "learning_rate": 2.1671531186163512e-05,
+      "loss": 3.6459,
+      "step": 485
+    },
+    {
+      "epoch": 0.5436241610738255,
+      "grad_norm": 3.060016393661499,
+      "learning_rate": 2.158448188485433e-05,
+      "loss": 3.8407,
+      "step": 486
+    },
+    {
+      "epoch": 0.5447427293064877,
+      "grad_norm": 2.318657875061035,
+      "learning_rate": 2.1497474761047086e-05,
+      "loss": 3.7452,
+      "step": 487
+    },
+    {
+      "epoch": 0.5458612975391499,
+      "grad_norm": 2.2398786544799805,
+      "learning_rate": 2.141051088917411e-05,
+      "loss": 4.078,
+      "step": 488
+    },
+    {
+      "epoch": 0.5469798657718121,
+      "grad_norm": 2.4985172748565674,
+      "learning_rate": 2.1323591343133646e-05,
+      "loss": 3.8253,
+      "step": 489
+    },
+    {
+      "epoch": 0.5480984340044742,
+      "grad_norm": 2.5000696182250977,
+      "learning_rate": 2.1236717196276558e-05,
+      "loss": 3.753,
+      "step": 490
+    },
+    {
+      "epoch": 0.5492170022371364,
+      "grad_norm": 2.586850643157959,
+      "learning_rate": 2.114988952139307e-05,
+      "loss": 4.1803,
+      "step": 491
+    },
+    {
+      "epoch": 0.5503355704697986,
+      "grad_norm": 2.7555367946624756,
+      "learning_rate": 2.106310939069956e-05,
+      "loss": 3.8567,
+      "step": 492
+    },
+    {
+      "epoch": 0.5514541387024608,
+      "grad_norm": 2.633790969848633,
+      "learning_rate": 2.0976377875825283e-05,
+      "loss": 4.1461,
+      "step": 493
+    },
+    {
+      "epoch": 0.5525727069351231,
+      "grad_norm": 3.8133838176727295,
+      "learning_rate": 2.0889696047799143e-05,
+      "loss": 4.0689,
+      "step": 494
+    },
+    {
+      "epoch": 0.5536912751677853,
+      "grad_norm": 1.6977475881576538,
+      "learning_rate": 2.0803064977036478e-05,
+      "loss": 3.9074,
+      "step": 495
+    },
+    {
+      "epoch": 0.5548098434004475,
+      "grad_norm": 3.245305299758911,
+      "learning_rate": 2.071648573332583e-05,
+      "loss": 3.7151,
+      "step": 496
+    },
+    {
+      "epoch": 0.5559284116331096,
+      "grad_norm": 2.754373550415039,
+      "learning_rate": 2.0629959385815757e-05,
+      "loss": 4.0098,
+      "step": 497
+    },
+    {
+      "epoch": 0.5570469798657718,
+      "grad_norm": 2.561891794204712,
+      "learning_rate": 2.054348700300158e-05,
+      "loss": 3.8599,
+      "step": 498
+    },
+    {
+      "epoch": 0.558165548098434,
+      "grad_norm": 2.573894500732422,
+      "learning_rate": 2.0457069652712242e-05,
+      "loss": 4.1772,
+      "step": 499
+    },
+    {
+      "epoch": 0.5592841163310962,
+      "grad_norm": 2.2721850872039795,
+      "learning_rate": 2.037070840209711e-05,
+      "loss": 4.2124,
+      "step": 500
+    },
+    {
+      "epoch": 0.5604026845637584,
+      "grad_norm": 2.0952377319335938,
+      "learning_rate": 2.0284404317612764e-05,
+      "loss": 3.9423,
+      "step": 501
+    },
+    {
+      "epoch": 0.5615212527964206,
+      "grad_norm": 3.6689555644989014,
+      "learning_rate": 2.019815846500988e-05,
+      "loss": 3.8274,
+      "step": 502
+    },
+    {
+      "epoch": 0.5626398210290827,
+      "grad_norm": 2.6612648963928223,
+      "learning_rate": 2.0111971909320027e-05,
+      "loss": 3.9185,
+      "step": 503
+    },
+    {
+      "epoch": 0.5637583892617449,
+      "grad_norm": 2.889272928237915,
+      "learning_rate": 2.0025845714842516e-05,
+      "loss": 3.7501,
+      "step": 504
+    },
+    {
+      "epoch": 0.5648769574944071,
+      "grad_norm": 3.470381736755371,
+      "learning_rate": 1.993978094513128e-05,
+      "loss": 3.6945,
+      "step": 505
+    },
+    {
+      "epoch": 0.5659955257270693,
+      "grad_norm": 2.4324817657470703,
+      "learning_rate": 1.9853778662981744e-05,
+      "loss": 4.0761,
+      "step": 506
+    },
+    {
+      "epoch": 0.5671140939597316,
+      "grad_norm": 2.1613693237304688,
+      "learning_rate": 1.9767839930417673e-05,
+      "loss": 4.3073,
+      "step": 507
+    },
+    {
+      "epoch": 0.5682326621923938,
+      "grad_norm": 2.2693519592285156,
+      "learning_rate": 1.968196580867808e-05,
+      "loss": 4.0687,
+      "step": 508
+    },
+    {
+      "epoch": 0.569351230425056,
+      "grad_norm": 1.9982450008392334,
+      "learning_rate": 1.9596157358204097e-05,
+      "loss": 3.6727,
+      "step": 509
+    },
+    {
+      "epoch": 0.5704697986577181,
+      "grad_norm": 3.018151044845581,
+      "learning_rate": 1.9510415638625932e-05,
+      "loss": 3.6253,
+      "step": 510
+    },
+    {
+      "epoch": 0.5715883668903803,
+      "grad_norm": 1.8852089643478394,
+      "learning_rate": 1.9424741708749695e-05,
+      "loss": 4.0768,
+      "step": 511
+    },
+    {
+      "epoch": 0.5727069351230425,
+      "grad_norm": 2.7679364681243896,
+      "learning_rate": 1.9339136626544407e-05,
+      "loss": 3.6646,
+      "step": 512
+    },
+    {
+      "epoch": 0.5738255033557047,
+      "grad_norm": 2.5092828273773193,
+      "learning_rate": 1.9253601449128914e-05,
+      "loss": 4.0424,
+      "step": 513
+    },
+    {
+      "epoch": 0.5749440715883669,
+      "grad_norm": 1.9775032997131348,
+      "learning_rate": 1.9168137232758797e-05,
+      "loss": 3.8255,
+      "step": 514
+    },
+    {
+      "epoch": 0.5760626398210291,
+      "grad_norm": 2.3674395084381104,
+      "learning_rate": 1.9082745032813368e-05,
+      "loss": 3.9264,
+      "step": 515
+    },
+    {
+      "epoch": 0.5771812080536913,
+      "grad_norm": 1.9700473546981812,
+      "learning_rate": 1.8997425903782627e-05,
+      "loss": 3.9236,
+      "step": 516
+    },
+    {
+      "epoch": 0.5782997762863534,
+      "grad_norm": 2.295315742492676,
+      "learning_rate": 1.891218089925423e-05,
+      "loss": 3.773,
+      "step": 517
+    },
+    {
+      "epoch": 0.5794183445190156,
+      "grad_norm": 1.7736364603042603,
+      "learning_rate": 1.8827011071900474e-05,
+      "loss": 3.8288,
+      "step": 518
+    },
+    {
+      "epoch": 0.5805369127516778,
+      "grad_norm": 2.1893913745880127,
+      "learning_rate": 1.874191747346534e-05,
+      "loss": 3.8251,
+      "step": 519
+    },
+    {
+      "epoch": 0.5816554809843401,
+      "grad_norm": 2.590500593185425,
+      "learning_rate": 1.865690115475144e-05,
+      "loss": 4.1849,
+      "step": 520
+    },
+    {
+      "epoch": 0.5827740492170023,
+      "grad_norm": 2.320377826690674,
+      "learning_rate": 1.8571963165607104e-05,
+      "loss": 3.8381,
+      "step": 521
+    },
+    {
+      "epoch": 0.5838926174496645,
+      "grad_norm": 2.334003210067749,
+      "learning_rate": 1.8487104554913363e-05,
+      "loss": 3.9728,
+      "step": 522
+    },
+    {
+      "epoch": 0.5850111856823266,
+      "grad_norm": 2.939434051513672,
+      "learning_rate": 1.8402326370571056e-05,
+      "loss": 3.9855,
+      "step": 523
+    },
+    {
+      "epoch": 0.5861297539149888,
+      "grad_norm": 2.41349720954895,
+      "learning_rate": 1.8317629659487796e-05,
+      "loss": 4.0445,
+      "step": 524
+    },
+    {
+      "epoch": 0.587248322147651,
+      "grad_norm": 2.5825881958007812,
+      "learning_rate": 1.8233015467565152e-05,
+      "loss": 3.852,
+      "step": 525
+    },
+    {
+      "epoch": 0.5883668903803132,
+      "grad_norm": 2.247124671936035,
+      "learning_rate": 1.8148484839685662e-05,
+      "loss": 4.341,
+      "step": 526
+    },
+    {
+      "epoch": 0.5894854586129754,
+      "grad_norm": 2.6871159076690674,
+      "learning_rate": 1.806403881969996e-05,
+      "loss": 3.9645,
+      "step": 527
+    },
+    {
+      "epoch": 0.5906040268456376,
+      "grad_norm": 3.4999427795410156,
+      "learning_rate": 1.7979678450413845e-05,
+      "loss": 4.1412,
+      "step": 528
+    },
+    {
+      "epoch": 0.5917225950782998,
+      "grad_norm": 3.4024899005889893,
+      "learning_rate": 1.789540477357548e-05,
+      "loss": 4.191,
+      "step": 529
+    },
+    {
+      "epoch": 0.5928411633109619,
+      "grad_norm": 2.1801340579986572,
+      "learning_rate": 1.781121882986245e-05,
+      "loss": 3.6334,
+      "step": 530
+    },
+    {
+      "epoch": 0.5939597315436241,
+      "grad_norm": 2.63584303855896,
+      "learning_rate": 1.7727121658868934e-05,
+      "loss": 3.7061,
+      "step": 531
+    },
+    {
+      "epoch": 0.5950782997762863,
+      "grad_norm": 2.664759635925293,
+      "learning_rate": 1.764311429909291e-05,
+      "loss": 4.0002,
+      "step": 532
+    },
+    {
+      "epoch": 0.5961968680089486,
+      "grad_norm": 3.4378206729888916,
+      "learning_rate": 1.7559197787923263e-05,
+      "loss": 3.8005,
+      "step": 533
+    },
+    {
+      "epoch": 0.5973154362416108,
+      "grad_norm": 2.902148485183716,
+      "learning_rate": 1.7475373161627034e-05,
+      "loss": 4.086,
+      "step": 534
+    },
+    {
+      "epoch": 0.598434004474273,
+      "grad_norm": 1.8565155267715454,
+      "learning_rate": 1.739164145533658e-05,
+      "loss": 3.6199,
+      "step": 535
+    },
+    {
+      "epoch": 0.5995525727069351,
+      "grad_norm": 2.2975833415985107,
+      "learning_rate": 1.730800370303683e-05,
+      "loss": 4.0409,
+      "step": 536
+    },
+    {
+      "epoch": 0.6006711409395973,
+      "grad_norm": 2.3823060989379883,
+      "learning_rate": 1.7224460937552446e-05,
+      "loss": 3.9938,
+      "step": 537
+    },
+    {
+      "epoch": 0.6017897091722595,
+      "grad_norm": 1.8713372945785522,
+      "learning_rate": 1.714101419053518e-05,
+      "loss": 3.9404,
+      "step": 538
+    },
+    {
+      "epoch": 0.6029082774049217,
+      "grad_norm": 2.7200701236724854,
+      "learning_rate": 1.7057664492451036e-05,
+      "loss": 3.6198,
+      "step": 539
+    },
+    {
+      "epoch": 0.6040268456375839,
+      "grad_norm": 2.899989604949951,
+      "learning_rate": 1.6974412872567597e-05,
+      "loss": 3.8505,
+      "step": 540
+    },
+    {
+      "epoch": 0.6051454138702461,
+      "grad_norm": 3.0191915035247803,
+      "learning_rate": 1.6891260358941276e-05,
+      "loss": 3.8781,
+      "step": 541
+    },
+    {
+      "epoch": 0.6062639821029083,
+      "grad_norm": 3.2697198390960693,
+      "learning_rate": 1.6808207978404683e-05,
+      "loss": 3.7293,
+      "step": 542
+    },
+    {
+      "epoch": 0.6073825503355704,
+      "grad_norm": 4.28400182723999,
+      "learning_rate": 1.672525675655387e-05,
+      "loss": 4.0706,
+      "step": 543
+    },
+    {
+      "epoch": 0.6085011185682326,
+      "grad_norm": 2.090829849243164,
+      "learning_rate": 1.664240771773571e-05,
+      "loss": 4.0724,
+      "step": 544
+    },
+    {
+      "epoch": 0.6096196868008948,
+      "grad_norm": 2.9155173301696777,
+      "learning_rate": 1.655966188503526e-05,
+      "loss": 3.7549,
+      "step": 545
+    },
+    {
+      "epoch": 0.610738255033557,
+      "grad_norm": 2.9214680194854736,
+      "learning_rate": 1.647702028026308e-05,
+      "loss": 3.9448,
+      "step": 546
+    },
+    {
+      "epoch": 0.6118568232662193,
+      "grad_norm": 1.5951290130615234,
+      "learning_rate": 1.6394483923942655e-05,
+      "loss": 4.1215,
+      "step": 547
+    },
+    {
+      "epoch": 0.6129753914988815,
+      "grad_norm": 2.0903103351593018,
+      "learning_rate": 1.6312053835297784e-05,
+      "loss": 3.8019,
+      "step": 548
+    },
+    {
+      "epoch": 0.6140939597315436,
+      "grad_norm": 2.552449941635132,
+      "learning_rate": 1.6229731032239988e-05,
+      "loss": 4.0168,
+      "step": 549
+    },
+    {
+      "epoch": 0.6152125279642058,
+      "grad_norm": 2.0024003982543945,
+      "learning_rate": 1.6147516531355912e-05,
+      "loss": 3.8326,
+      "step": 550
+    },
+    {
+      "epoch": 0.616331096196868,
+      "grad_norm": 2.165592908859253,
+      "learning_rate": 1.606541134789485e-05,
+      "loss": 3.798,
+      "step": 551
+    },
+    {
+      "epoch": 0.6174496644295302,
+      "grad_norm": 2.631091356277466,
+      "learning_rate": 1.5983416495756116e-05,
+      "loss": 3.8511,
+      "step": 552
+    },
+    {
+      "epoch": 0.6185682326621924,
+      "grad_norm": 2.507314682006836,
+      "learning_rate": 1.5901532987476598e-05,
+      "loss": 4.1623,
+      "step": 553
+    },
+    {
+      "epoch": 0.6196868008948546,
+      "grad_norm": 2.264253854751587,
+      "learning_rate": 1.58197618342182e-05,
+      "loss": 3.8904,
+      "step": 554
+    },
+    {
+      "epoch": 0.6208053691275168,
+      "grad_norm": 2.0525124073028564,
+      "learning_rate": 1.573810404575539e-05,
+      "loss": 3.8115,
+      "step": 555
+    },
+    {
+      "epoch": 0.6219239373601789,
+      "grad_norm": 2.31563401222229,
+      "learning_rate": 1.565656063046272e-05,
+      "loss": 3.9583,
+      "step": 556
+    },
+    {
+      "epoch": 0.6230425055928411,
+      "grad_norm": 2.9089858531951904,
+      "learning_rate": 1.5575132595302352e-05,
+      "loss": 3.9061,
+      "step": 557
+    },
+    {
+      "epoch": 0.6241610738255033,
+      "grad_norm": 2.3452372550964355,
+      "learning_rate": 1.549382094581166e-05,
+      "loss": 3.8342,
+      "step": 558
+    },
+    {
+      "epoch": 0.6252796420581656,
+      "grad_norm": 2.0573930740356445,
+      "learning_rate": 1.541262668609081e-05,
+      "loss": 3.7931,
+      "step": 559
+    },
+    {
+      "epoch": 0.6263982102908278,
+      "grad_norm": 2.165867328643799,
+      "learning_rate": 1.5331550818790315e-05,
+      "loss": 3.8872,
+      "step": 560
+    },
+    {
+      "epoch": 0.62751677852349,
+      "grad_norm": 2.885354518890381,
+      "learning_rate": 1.525059434509871e-05,
+      "loss": 3.9573,
+      "step": 561
+    },
+    {
+      "epoch": 0.6286353467561522,
+      "grad_norm": 2.424731731414795,
+      "learning_rate": 1.516975826473015e-05,
+      "loss": 3.7392,
+      "step": 562
+    },
+    {
+      "epoch": 0.6297539149888143,
+      "grad_norm": 1.858024001121521,
+      "learning_rate": 1.5089043575912099e-05,
+      "loss": 3.8223,
+      "step": 563
+    },
+    {
+      "epoch": 0.6308724832214765,
+      "grad_norm": 2.57297420501709,
+      "learning_rate": 1.5008451275372949e-05,
+      "loss": 3.4912,
+      "step": 564
+    },
+    {
+      "epoch": 0.6319910514541387,
+      "grad_norm": 2.9374403953552246,
+      "learning_rate": 1.4927982358329768e-05,
+      "loss": 4.142,
+      "step": 565
+    },
+    {
+      "epoch": 0.6331096196868009,
+      "grad_norm": 2.8601393699645996,
+      "learning_rate": 1.484763781847599e-05,
+      "loss": 4.0499,
+      "step": 566
+    },
+    {
+      "epoch": 0.6342281879194631,
+      "grad_norm": 2.611351251602173,
+      "learning_rate": 1.4767418647969133e-05,
+      "loss": 4.1721,
+      "step": 567
+    },
+    {
+      "epoch": 0.6353467561521253,
+      "grad_norm": 2.2778544425964355,
+      "learning_rate": 1.4687325837418563e-05,
+      "loss": 4.0221,
+      "step": 568
+    },
+    {
+      "epoch": 0.6364653243847874,
+      "grad_norm": 2.24289608001709,
+      "learning_rate": 1.4607360375873246e-05,
+      "loss": 4.0047,
+      "step": 569
+    },
+    {
+      "epoch": 0.6375838926174496,
+      "grad_norm": 2.1393930912017822,
+      "learning_rate": 1.4527523250809545e-05,
+      "loss": 3.7402,
+      "step": 570
+    },
+    {
+      "epoch": 0.6387024608501118,
+      "grad_norm": 2.4611830711364746,
+      "learning_rate": 1.4447815448119017e-05,
+      "loss": 4.1909,
+      "step": 571
+    },
+    {
+      "epoch": 0.639821029082774,
+      "grad_norm": 2.261955976486206,
+      "learning_rate": 1.4368237952096258e-05,
+      "loss": 3.6671,
+      "step": 572
+    },
+    {
+      "epoch": 0.6409395973154363,
+      "grad_norm": 2.2543327808380127,
+      "learning_rate": 1.4288791745426739e-05,
+      "loss": 3.719,
+      "step": 573
+    },
+    {
+      "epoch": 0.6420581655480985,
+      "grad_norm": 3.0173962116241455,
+      "learning_rate": 1.4209477809174634e-05,
+      "loss": 3.9079,
+      "step": 574
+    },
+    {
+      "epoch": 0.6431767337807607,
+      "grad_norm": 2.1777987480163574,
+      "learning_rate": 1.4130297122770774e-05,
+      "loss": 3.6232,
+      "step": 575
+    },
+    {
+      "epoch": 0.6442953020134228,
+      "grad_norm": 2.0531842708587646,
+      "learning_rate": 1.4051250664000515e-05,
+      "loss": 3.9533,
+      "step": 576
+    },
+    {
+      "epoch": 0.645413870246085,
+      "grad_norm": 2.7439255714416504,
+      "learning_rate": 1.3972339408991626e-05,
+      "loss": 3.8229,
+      "step": 577
+    },
+    {
+      "epoch": 0.6465324384787472,
+      "grad_norm": 2.413195848464966,
+      "learning_rate": 1.3893564332202319e-05,
+      "loss": 4.346,
+      "step": 578
+    },
+    {
+      "epoch": 0.6476510067114094,
+      "grad_norm": 2.382014274597168,
+      "learning_rate": 1.381492640640914e-05,
+      "loss": 3.771,
+      "step": 579
+    },
+    {
+      "epoch": 0.6487695749440716,
+      "grad_norm": 1.874355673789978,
+      "learning_rate": 1.3736426602694998e-05,
+      "loss": 3.7268,
+      "step": 580
+    },
+    {
+      "epoch": 0.6498881431767338,
+      "grad_norm": 1.958233118057251,
+      "learning_rate": 1.365806589043718e-05,
+      "loss": 3.9595,
+      "step": 581
+    },
+    {
+      "epoch": 0.6510067114093959,
+      "grad_norm": 2.6319327354431152,
+      "learning_rate": 1.357984523729533e-05,
+      "loss": 4.2425,
+      "step": 582
+    },
+    {
+      "epoch": 0.6521252796420581,
+      "grad_norm": 3.006190061569214,
+      "learning_rate": 1.3501765609199534e-05,
+      "loss": 3.8662,
+      "step": 583
+    },
+    {
+      "epoch": 0.6532438478747203,
+      "grad_norm": 2.89542818069458,
+      "learning_rate": 1.342382797033842e-05,
+      "loss": 3.8635,
+      "step": 584
+    },
+    {
+      "epoch": 0.6543624161073825,
+      "grad_norm": 2.149411916732788,
+      "learning_rate": 1.3346033283147196e-05,
+      "loss": 3.9787,
+      "step": 585
+    },
+    {
+      "epoch": 0.6554809843400448,
+      "grad_norm": 2.2859368324279785,
+      "learning_rate": 1.3268382508295812e-05,
+      "loss": 3.735,
+      "step": 586
+    },
+    {
+      "epoch": 0.656599552572707,
+      "grad_norm": 3.07099986076355,
+      "learning_rate": 1.3190876604677043e-05,
+      "loss": 3.9014,
+      "step": 587
+    },
+    {
+      "epoch": 0.6577181208053692,
+      "grad_norm": 2.119368553161621,
+      "learning_rate": 1.3113516529394704e-05,
+      "loss": 4.0083,
+      "step": 588
+    },
+    {
+      "epoch": 0.6588366890380313,
+      "grad_norm": 2.585334300994873,
+      "learning_rate": 1.3036303237751812e-05,
+      "loss": 3.7846,
+      "step": 589
+    },
+    {
+      "epoch": 0.6599552572706935,
+      "grad_norm": 2.7365801334381104,
+      "learning_rate": 1.2959237683238767e-05,
+      "loss": 3.6363,
+      "step": 590
+    },
+    {
+      "epoch": 0.6610738255033557,
+      "grad_norm": 2.1108875274658203,
+      "learning_rate": 1.2882320817521588e-05,
+      "loss": 3.8154,
+      "step": 591
+    },
+    {
+      "epoch": 0.6621923937360179,
+      "grad_norm": 2.750692129135132,
+      "learning_rate": 1.2805553590430197e-05,
+      "loss": 3.7864,
+      "step": 592
+    },
+    {
+      "epoch": 0.6633109619686801,
+      "grad_norm": 1.9360623359680176,
+      "learning_rate": 1.2728936949946638e-05,
+      "loss": 3.5591,
+      "step": 593
+    },
+    {
+      "epoch": 0.6644295302013423,
+      "grad_norm": 2.5163891315460205,
+      "learning_rate": 1.2652471842193415e-05,
+      "loss": 3.8985,
+      "step": 594
+    },
+    {
+      "epoch": 0.6655480984340044,
+      "grad_norm": 2.350154399871826,
+      "learning_rate": 1.2576159211421763e-05,
+      "loss": 3.8853,
+      "step": 595
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 2.915135622024536,
+      "learning_rate": 1.2500000000000006e-05,
+      "loss": 3.6496,
+      "step": 596
+    },
+    {
+      "epoch": 0.6677852348993288,
+      "grad_norm": 3.425151824951172,
+      "learning_rate": 1.2423995148401954e-05,
+      "loss": 4.1841,
+      "step": 597
+    },
+    {
+      "epoch": 0.668903803131991,
+      "grad_norm": 2.1344897747039795,
+      "learning_rate": 1.2348145595195245e-05,
+      "loss": 3.6813,
+      "step": 598
+    },
+    {
+      "epoch": 0.6700223713646533,
+      "grad_norm": 3.043466806411743,
+      "learning_rate": 1.2272452277029783e-05,
+      "loss": 3.717,
+      "step": 599
+    },
+    {
+      "epoch": 0.6711409395973155,
+      "grad_norm": 2.1800496578216553,
+      "learning_rate": 1.2196916128626126e-05,
+      "loss": 3.7066,
+      "step": 600
+    },
+    {
+      "epoch": 0.6722595078299777,
+      "grad_norm": 2.1941945552825928,
+      "learning_rate": 1.2121538082764009e-05,
+      "loss": 3.8536,
+      "step": 601
+    },
+    {
+      "epoch": 0.6733780760626398,
+      "grad_norm": 1.9257776737213135,
+      "learning_rate": 1.2046319070270792e-05,
+      "loss": 3.8676,
+      "step": 602
+    },
+    {
+      "epoch": 0.674496644295302,
+      "grad_norm": 1.9963043928146362,
+      "learning_rate": 1.1971260020009944e-05,
+      "loss": 4.0604,
+      "step": 603
+    },
+    {
+      "epoch": 0.6756152125279642,
+      "grad_norm": 4.963601112365723,
+      "learning_rate": 1.1896361858869598e-05,
+      "loss": 3.8178,
+      "step": 604
+    },
+    {
+      "epoch": 0.6767337807606264,
+      "grad_norm": 2.1134705543518066,
+      "learning_rate": 1.1821625511751122e-05,
+      "loss": 4.3728,
+      "step": 605
+    },
+    {
+      "epoch": 0.6778523489932886,
+      "grad_norm": 1.895908236503601,
+      "learning_rate": 1.174705190155766e-05,
+      "loss": 4.0292,
+      "step": 606
+    },
+    {
+      "epoch": 0.6789709172259508,
+      "grad_norm": 2.871042490005493,
+      "learning_rate": 1.1672641949182769e-05,
+      "loss": 3.6701,
+      "step": 607
+    },
+    {
+      "epoch": 0.680089485458613,
+      "grad_norm": 3.3493497371673584,
+      "learning_rate": 1.1598396573499005e-05,
+      "loss": 4.0769,
+      "step": 608
+    },
+    {
+      "epoch": 0.6812080536912751,
+      "grad_norm": 2.001706838607788,
+      "learning_rate": 1.152431669134663e-05,
+      "loss": 3.9975,
+      "step": 609
+    },
+    {
+      "epoch": 0.6823266219239373,
+      "grad_norm": 2.3988122940063477,
+      "learning_rate": 1.145040321752223e-05,
+      "loss": 4.1463,
+      "step": 610
+    },
+    {
+      "epoch": 0.6834451901565995,
+      "grad_norm": 2.896697759628296,
+      "learning_rate": 1.1376657064767485e-05,
+      "loss": 3.6005,
+      "step": 611
+    },
+    {
+      "epoch": 0.6845637583892618,
+      "grad_norm": 2.0190067291259766,
+      "learning_rate": 1.130307914375783e-05,
+      "loss": 3.8035,
+      "step": 612
+    },
+    {
+      "epoch": 0.685682326621924,
+      "grad_norm": 1.734345555305481,
+      "learning_rate": 1.122967036309127e-05,
+      "loss": 3.9921,
+      "step": 613
+    },
+    {
+      "epoch": 0.6868008948545862,
+      "grad_norm": 2.797583818435669,
+      "learning_rate": 1.1156431629277118e-05,
+      "loss": 3.7049,
+      "step": 614
+    },
+    {
+      "epoch": 0.6879194630872483,
+      "grad_norm": 2.020956039428711,
+      "learning_rate": 1.1083363846724822e-05,
+      "loss": 3.9454,
+      "step": 615
+    },
+    {
+      "epoch": 0.6890380313199105,
+      "grad_norm": 2.833671808242798,
+      "learning_rate": 1.1010467917732784e-05,
+      "loss": 3.9514,
+      "step": 616
+    },
+    {
+      "epoch": 0.6901565995525727,
+      "grad_norm": 1.816365122795105,
+      "learning_rate": 1.093774474247721e-05,
+      "loss": 4.1637,
+      "step": 617
+    },
+    {
+      "epoch": 0.6912751677852349,
+      "grad_norm": 2.2042744159698486,
+      "learning_rate": 1.086519521900103e-05,
+      "loss": 3.8951,
+      "step": 618
+    },
+    {
+      "epoch": 0.6923937360178971,
+      "grad_norm": 3.488060712814331,
+      "learning_rate": 1.0792820243202772e-05,
+      "loss": 3.942,
+      "step": 619
+    },
+    {
+      "epoch": 0.6935123042505593,
+      "grad_norm": 2.248445510864258,
+      "learning_rate": 1.0720620708825536e-05,
+      "loss": 3.8709,
+      "step": 620
+    },
+    {
+      "epoch": 0.6946308724832215,
+      "grad_norm": 2.1262619495391846,
+      "learning_rate": 1.0648597507445884e-05,
+      "loss": 3.9484,
+      "step": 621
+    },
+    {
+      "epoch": 0.6957494407158836,
+      "grad_norm": 2.552934169769287,
+      "learning_rate": 1.0576751528462935e-05,
+      "loss": 3.508,
+      "step": 622
+    },
+    {
+      "epoch": 0.6968680089485458,
+      "grad_norm": 2.420132875442505,
+      "learning_rate": 1.0505083659087284e-05,
+      "loss": 3.6366,
+      "step": 623
+    },
+    {
+      "epoch": 0.697986577181208,
+      "grad_norm": 2.1934938430786133,
+      "learning_rate": 1.043359478433012e-05,
+      "loss": 4.1458,
+      "step": 624
+    },
+    {
+      "epoch": 0.6991051454138703,
+      "grad_norm": 2.8820436000823975,
+      "learning_rate": 1.0362285786992231e-05,
+      "loss": 4.145,
+      "step": 625
+    },
+    {
+      "epoch": 0.7002237136465325,
+      "grad_norm": 2.036956548690796,
+      "learning_rate": 1.0291157547653172e-05,
+      "loss": 4.0621,
+      "step": 626
+    },
+    {
+      "epoch": 0.7013422818791947,
+      "grad_norm": 2.075648784637451,
+      "learning_rate": 1.0220210944660338e-05,
+      "loss": 4.1723,
+      "step": 627
+    },
+    {
+      "epoch": 0.7024608501118568,
+      "grad_norm": 2.007321357727051,
+      "learning_rate": 1.0149446854118153e-05,
+      "loss": 3.9694,
+      "step": 628
+    },
+    {
+      "epoch": 0.703579418344519,
+      "grad_norm": 2.436429977416992,
+      "learning_rate": 1.0078866149877206e-05,
+      "loss": 4.1056,
+      "step": 629
+    },
+    {
+      "epoch": 0.7046979865771812,
+      "grad_norm": 1.8887969255447388,
+      "learning_rate": 1.0008469703523493e-05,
+      "loss": 3.848,
+      "step": 630
+    },
+    {
+      "epoch": 0.7058165548098434,
+      "grad_norm": 2.362123489379883,
+      "learning_rate": 9.93825838436767e-06,
+      "loss": 4.1663,
+      "step": 631
+    },
+    {
+      "epoch": 0.7069351230425056,
+      "grad_norm": 2.105339765548706,
+      "learning_rate": 9.868233059434288e-06,
+      "loss": 3.6117,
+      "step": 632
+    },
+    {
+      "epoch": 0.7080536912751678,
+      "grad_norm": 2.77400541305542,
+      "learning_rate": 9.798394593451091e-06,
+      "loss": 4.035,
+      "step": 633
+    },
+    {
+      "epoch": 0.70917225950783,
+      "grad_norm": 3.4026005268096924,
+      "learning_rate": 9.728743848838333e-06,
+      "loss": 3.9344,
+      "step": 634
+    },
+    {
+      "epoch": 0.7102908277404921,
+      "grad_norm": 3.4259018898010254,
+      "learning_rate": 9.659281685698162e-06,
+      "loss": 4.2997,
+      "step": 635
+    },
+    {
+      "epoch": 0.7114093959731543,
+      "grad_norm": 2.261091470718384,
+      "learning_rate": 9.590008961803943e-06,
+      "loss": 3.7143,
+      "step": 636
+    },
+    {
+      "epoch": 0.7125279642058165,
+      "grad_norm": 3.5165352821350098,
+      "learning_rate": 9.520926532589725e-06,
+      "loss": 3.8508,
+      "step": 637
+    },
+    {
+      "epoch": 0.7136465324384788,
+      "grad_norm": 2.4055497646331787,
+      "learning_rate": 9.45203525113962e-06,
+      "loss": 3.6208,
+      "step": 638
+    },
+    {
+      "epoch": 0.714765100671141,
+      "grad_norm": 3.532019853591919,
+      "learning_rate": 9.383335968177323e-06,
+      "loss": 3.8882,
+      "step": 639
+    },
+    {
+      "epoch": 0.7158836689038032,
+      "grad_norm": 2.0139145851135254,
+      "learning_rate": 9.314829532055569e-06,
+      "loss": 3.9722,
+      "step": 640
+    },
+    {
+      "epoch": 0.7170022371364653,
+      "grad_norm": 3.198631525039673,
+      "learning_rate": 9.24651678874568e-06,
+      "loss": 3.7721,
+      "step": 641
+    },
+    {
+      "epoch": 0.7181208053691275,
+      "grad_norm": 2.2410922050476074,
+      "learning_rate": 9.178398581827085e-06,
+      "loss": 3.978,
+      "step": 642
+    },
+    {
+      "epoch": 0.7192393736017897,
+      "grad_norm": 2.0665810108184814,
+      "learning_rate": 9.110475752476935e-06,
+      "loss": 4.0286,
+      "step": 643
+    },
+    {
+      "epoch": 0.7203579418344519,
+      "grad_norm": 2.212315559387207,
+      "learning_rate": 9.04274913945971e-06,
+      "loss": 3.92,
+      "step": 644
+    },
+    {
+      "epoch": 0.7214765100671141,
+      "grad_norm": 2.330747127532959,
+      "learning_rate": 8.975219579116863e-06,
+      "loss": 4.0911,
+      "step": 645
+    },
+    {
+      "epoch": 0.7225950782997763,
+      "grad_norm": 2.7871272563934326,
+      "learning_rate": 8.90788790535646e-06,
+      "loss": 3.8816,
+      "step": 646
+    },
+    {
+      "epoch": 0.7237136465324385,
+      "grad_norm": 2.5167412757873535,
+      "learning_rate": 8.840754949642935e-06,
+      "loss": 3.8268,
+      "step": 647
+    },
+    {
+      "epoch": 0.7248322147651006,
+      "grad_norm": 1.9781793355941772,
+      "learning_rate": 8.77382154098679e-06,
+      "loss": 3.7334,
+      "step": 648
+    },
+    {
+      "epoch": 0.7259507829977628,
+      "grad_norm": 2.6232032775878906,
+      "learning_rate": 8.70708850593434e-06,
+      "loss": 3.7084,
+      "step": 649
+    },
+    {
+      "epoch": 0.727069351230425,
+      "grad_norm": 2.177091121673584,
+      "learning_rate": 8.64055666855757e-06,
+      "loss": 3.9966,
+      "step": 650
+    },
+    {
+      "epoch": 0.7281879194630873,
+      "grad_norm": 2.444685220718384,
+      "learning_rate": 8.574226850443872e-06,
+      "loss": 3.7267,
+      "step": 651
+    },
+    {
+      "epoch": 0.7293064876957495,
+      "grad_norm": 3.574202299118042,
+      "learning_rate": 8.50809987068598e-06,
+      "loss": 3.8466,
+      "step": 652
+    },
+    {
+      "epoch": 0.7304250559284117,
+      "grad_norm": 3.1503660678863525,
+      "learning_rate": 8.442176545871805e-06,
+      "loss": 3.7964,
+      "step": 653
+    },
+    {
+      "epoch": 0.7315436241610739,
+      "grad_norm": 1.677579641342163,
+      "learning_rate": 8.376457690074385e-06,
+      "loss": 3.8638,
+      "step": 654
+    },
+    {
+      "epoch": 0.732662192393736,
+      "grad_norm": 2.3635876178741455,
+      "learning_rate": 8.310944114841786e-06,
+      "loss": 3.8608,
+      "step": 655
+    },
+    {
+      "epoch": 0.7337807606263982,
+      "grad_norm": 3.185817241668701,
+      "learning_rate": 8.24563662918712e-06,
+      "loss": 4.1601,
+      "step": 656
+    },
+    {
+      "epoch": 0.7348993288590604,
+      "grad_norm": 3.7147934436798096,
+      "learning_rate": 8.180536039578545e-06,
+      "loss": 3.9529,
+      "step": 657
+    },
+    {
+      "epoch": 0.7360178970917226,
+      "grad_norm": 3.5860955715179443,
+      "learning_rate": 8.115643149929318e-06,
+      "loss": 4.2173,
+      "step": 658
+    },
+    {
+      "epoch": 0.7371364653243848,
+      "grad_norm": 2.398571252822876,
+      "learning_rate": 8.05095876158782e-06,
+      "loss": 3.7547,
+      "step": 659
+    },
+    {
+      "epoch": 0.738255033557047,
+      "grad_norm": 2.6575210094451904,
+      "learning_rate": 7.986483673327724e-06,
+      "loss": 3.9889,
+      "step": 660
+    },
+    {
+      "epoch": 0.7393736017897091,
+      "grad_norm": 2.9067556858062744,
+      "learning_rate": 7.922218681338095e-06,
+      "loss": 3.6771,
+      "step": 661
+    },
+    {
+      "epoch": 0.7404921700223713,
+      "grad_norm": 2.622535467147827,
+      "learning_rate": 7.858164579213547e-06,
+      "loss": 3.8512,
+      "step": 662
+    },
+    {
+      "epoch": 0.7416107382550335,
+      "grad_norm": 2.1957478523254395,
+      "learning_rate": 7.79432215794449e-06,
+      "loss": 3.9811,
+      "step": 663
+    },
+    {
+      "epoch": 0.7427293064876958,
+      "grad_norm": 2.333820343017578,
+      "learning_rate": 7.730692205907294e-06,
+      "loss": 3.5702,
+      "step": 664
+    },
+    {
+      "epoch": 0.743847874720358,
+      "grad_norm": 2.3756656646728516,
+      "learning_rate": 7.667275508854627e-06,
+      "loss": 4.0504,
+      "step": 665
+    },
+    {
+      "epoch": 0.7449664429530202,
+      "grad_norm": 2.2094883918762207,
+      "learning_rate": 7.604072849905708e-06,
+      "loss": 3.773,
+      "step": 666
+    },
+    {
+      "epoch": 0.7460850111856824,
+      "grad_norm": 2.3588247299194336,
+      "learning_rate": 7.541085009536625e-06,
+      "loss": 3.7015,
+      "step": 667
+    },
+    {
+      "epoch": 0.7472035794183445,
+      "grad_norm": 2.534830093383789,
+      "learning_rate": 7.4783127655707465e-06,
+      "loss": 3.8194,
+      "step": 668
+    },
+    {
+      "epoch": 0.7483221476510067,
+      "grad_norm": 2.694711208343506,
+      "learning_rate": 7.415756893169062e-06,
+      "loss": 3.9235,
+      "step": 669
+    },
+    {
+      "epoch": 0.7494407158836689,
+      "grad_norm": 3.596769332885742,
+      "learning_rate": 7.3534181648206555e-06,
+      "loss": 3.8338,
+      "step": 670
+    },
+    {
+      "epoch": 0.7505592841163311,
+      "grad_norm": 3.1873180866241455,
+      "learning_rate": 7.291297350333137e-06,
+      "loss": 3.6435,
+      "step": 671
+    },
+    {
+      "epoch": 0.7516778523489933,
+      "grad_norm": 1.9209433794021606,
+      "learning_rate": 7.2293952168231316e-06,
+      "loss": 3.7388,
+      "step": 672
+    },
+    {
+      "epoch": 0.7527964205816555,
+      "grad_norm": 2.5653648376464844,
+      "learning_rate": 7.167712528706844e-06,
+      "loss": 3.8961,
+      "step": 673
+    },
+    {
+      "epoch": 0.7539149888143176,
+      "grad_norm": 1.9980870485305786,
+      "learning_rate": 7.106250047690588e-06,
+      "loss": 3.9753,
+      "step": 674
+    },
+    {
+      "epoch": 0.7550335570469798,
+      "grad_norm": 2.534268379211426,
+      "learning_rate": 7.045008532761366e-06,
+      "loss": 3.7062,
+      "step": 675
+    },
+    {
+      "epoch": 0.756152125279642,
+      "grad_norm": 1.909271001815796,
+      "learning_rate": 6.983988740177552e-06,
+      "loss": 3.9121,
+      "step": 676
+    },
+    {
+      "epoch": 0.7572706935123042,
+      "grad_norm": 1.9393848180770874,
+      "learning_rate": 6.923191423459482e-06,
+      "loss": 3.9378,
+      "step": 677
+    },
+    {
+      "epoch": 0.7583892617449665,
+      "grad_norm": 3.104290723800659,
+      "learning_rate": 6.862617333380214e-06,
+      "loss": 3.6173,
+      "step": 678
+    },
+    {
+      "epoch": 0.7595078299776287,
+      "grad_norm": 1.9617093801498413,
+      "learning_rate": 6.802267217956221e-06,
+      "loss": 3.8215,
+      "step": 679
+    },
+    {
+      "epoch": 0.7606263982102909,
+      "grad_norm": 2.2684578895568848,
+      "learning_rate": 6.742141822438144e-06,
+      "loss": 3.8073,
+      "step": 680
+    },
+    {
+      "epoch": 0.761744966442953,
+      "grad_norm": 4.458706855773926,
+      "learning_rate": 6.682241889301635e-06,
+      "loss": 3.5522,
+      "step": 681
+    },
+    {
+      "epoch": 0.7628635346756152,
+      "grad_norm": 1.9600862264633179,
+      "learning_rate": 6.622568158238126e-06,
+      "loss": 3.7883,
+      "step": 682
+    },
+    {
+      "epoch": 0.7639821029082774,
+      "grad_norm": 2.454009771347046,
+      "learning_rate": 6.563121366145758e-06,
+      "loss": 3.723,
+      "step": 683
+    },
+    {
+      "epoch": 0.7651006711409396,
+      "grad_norm": 1.8544883728027344,
+      "learning_rate": 6.503902247120239e-06,
+      "loss": 3.894,
+      "step": 684
+    },
+    {
+      "epoch": 0.7662192393736018,
+      "grad_norm": 2.5093746185302734,
+      "learning_rate": 6.444911532445782e-06,
+      "loss": 3.9987,
+      "step": 685
+    },
+    {
+      "epoch": 0.767337807606264,
+      "grad_norm": 3.8050105571746826,
+      "learning_rate": 6.386149950586098e-06,
+      "loss": 3.7126,
+      "step": 686
+    },
+    {
+      "epoch": 0.7684563758389261,
+      "grad_norm": 2.824566125869751,
+      "learning_rate": 6.327618227175389e-06,
+      "loss": 3.7332,
+      "step": 687
+    },
+    {
+      "epoch": 0.7695749440715883,
+      "grad_norm": 3.873337984085083,
+      "learning_rate": 6.269317085009363e-06,
+      "loss": 3.6191,
+      "step": 688
+    },
+    {
+      "epoch": 0.7706935123042505,
+      "grad_norm": 2.275029182434082,
+      "learning_rate": 6.211247244036339e-06,
+      "loss": 3.6649,
+      "step": 689
+    },
+    {
+      "epoch": 0.7718120805369127,
+      "grad_norm": 3.252362012863159,
+      "learning_rate": 6.153409421348358e-06,
+      "loss": 3.8387,
+      "step": 690
+    },
+    {
+      "epoch": 0.772930648769575,
+      "grad_norm": 5.253241539001465,
+      "learning_rate": 6.095804331172308e-06,
+      "loss": 3.503,
+      "step": 691
+    },
+    {
+      "epoch": 0.7740492170022372,
+      "grad_norm": 1.9518181085586548,
+      "learning_rate": 6.0384326848611225e-06,
+      "loss": 4.2226,
+      "step": 692
+    },
+    {
+      "epoch": 0.7751677852348994,
+      "grad_norm": 2.372791051864624,
+      "learning_rate": 5.981295190884961e-06,
+      "loss": 4.0139,
+      "step": 693
+    },
+    {
+      "epoch": 0.7762863534675615,
+      "grad_norm": 2.35370135307312,
+      "learning_rate": 5.924392554822519e-06,
+      "loss": 3.7916,
+      "step": 694
+    },
+    {
+      "epoch": 0.7774049217002237,
+      "grad_norm": 2.5694122314453125,
+      "learning_rate": 5.867725479352251e-06,
+      "loss": 3.9369,
+      "step": 695
+    },
+    {
+      "epoch": 0.7785234899328859,
+      "grad_norm": 2.143845558166504,
+      "learning_rate": 5.811294664243752e-06,
+      "loss": 4.0455,
+      "step": 696
+    },
+    {
+      "epoch": 0.7796420581655481,
+      "grad_norm": 1.9714043140411377,
+      "learning_rate": 5.755100806349076e-06,
+      "loss": 3.9658,
+      "step": 697
+    },
+    {
+      "epoch": 0.7807606263982103,
+      "grad_norm": 3.7511520385742188,
+      "learning_rate": 5.699144599594131e-06,
+      "loss": 3.4662,
+      "step": 698
+    },
+    {
+      "epoch": 0.7818791946308725,
+      "grad_norm": 2.6238203048706055,
+      "learning_rate": 5.643426734970139e-06,
+      "loss": 3.7648,
+      "step": 699
+    },
+    {
+      "epoch": 0.7829977628635347,
+      "grad_norm": 2.54085111618042,
+      "learning_rate": 5.587947900525093e-06,
+      "loss": 3.9747,
+      "step": 700
+    },
+    {
+      "epoch": 0.7841163310961968,
+      "grad_norm": 2.562391519546509,
+      "learning_rate": 5.532708781355231e-06,
+      "loss": 3.7304,
+      "step": 701
+    },
+    {
+      "epoch": 0.785234899328859,
+      "grad_norm": 1.9780319929122925,
+      "learning_rate": 5.4777100595965994e-06,
+      "loss": 3.7713,
+      "step": 702
+    },
+    {
+      "epoch": 0.7863534675615212,
+      "grad_norm": 2.6395468711853027,
+      "learning_rate": 5.422952414416649e-06,
+      "loss": 3.6651,
+      "step": 703
+    },
+    {
+      "epoch": 0.7874720357941835,
+      "grad_norm": 2.2408347129821777,
+      "learning_rate": 5.368436522005815e-06,
+      "loss": 3.8067,
+      "step": 704
+    },
+    {
+      "epoch": 0.7885906040268457,
+      "grad_norm": 2.3998045921325684,
+      "learning_rate": 5.314163055569188e-06,
+      "loss": 3.9425,
+      "step": 705
+    },
+    {
+      "epoch": 0.7897091722595079,
+      "grad_norm": 1.973858118057251,
+      "learning_rate": 5.26013268531817e-06,
+      "loss": 4.0397,
+      "step": 706
+    },
+    {
+      "epoch": 0.79082774049217,
+      "grad_norm": 2.511659860610962,
+      "learning_rate": 5.206346078462249e-06,
+      "loss": 3.661,
+      "step": 707
+    },
+    {
+      "epoch": 0.7919463087248322,
+      "grad_norm": 2.8732688426971436,
+      "learning_rate": 5.1528038992007e-06,
+      "loss": 3.8967,
+      "step": 708
+    },
+    {
+      "epoch": 0.7930648769574944,
+      "grad_norm": 2.265322685241699,
+      "learning_rate": 5.099506808714452e-06,
+      "loss": 3.8421,
+      "step": 709
+    },
+    {
+      "epoch": 0.7941834451901566,
+      "grad_norm": 3.5834262371063232,
+      "learning_rate": 5.046455465157848e-06,
+      "loss": 4.1527,
+      "step": 710
+    },
+    {
+      "epoch": 0.7953020134228188,
+      "grad_norm": 3.056978225708008,
+      "learning_rate": 4.993650523650575e-06,
+      "loss": 3.4209,
+      "step": 711
+    },
+    {
+      "epoch": 0.796420581655481,
+      "grad_norm": 2.173541784286499,
+      "learning_rate": 4.941092636269554e-06,
+      "loss": 3.6813,
+      "step": 712
+    },
+    {
+      "epoch": 0.7975391498881432,
+      "grad_norm": 2.102139949798584,
+      "learning_rate": 4.888782452040885e-06,
+      "loss": 3.7805,
+      "step": 713
+    },
+    {
+      "epoch": 0.7986577181208053,
+      "grad_norm": 2.5020039081573486,
+      "learning_rate": 4.836720616931831e-06,
+      "loss": 3.7669,
+      "step": 714
+    },
+    {
+      "epoch": 0.7997762863534675,
+      "grad_norm": 2.2521355152130127,
+      "learning_rate": 4.784907773842839e-06,
+      "loss": 4.0918,
+      "step": 715
+    },
+    {
+      "epoch": 0.8008948545861297,
+      "grad_norm": 2.366150379180908,
+      "learning_rate": 4.733344562599623e-06,
+      "loss": 4.0847,
+      "step": 716
+    },
+    {
+      "epoch": 0.802013422818792,
+      "grad_norm": 3.198983907699585,
+      "learning_rate": 4.682031619945238e-06,
+      "loss": 3.5284,
+      "step": 717
+    },
+    {
+      "epoch": 0.8031319910514542,
+      "grad_norm": 1.931961178779602,
+      "learning_rate": 4.630969579532232e-06,
+      "loss": 3.8843,
+      "step": 718
+    },
+    {
+      "epoch": 0.8042505592841164,
+      "grad_norm": 2.4697344303131104,
+      "learning_rate": 4.580159071914794e-06,
+      "loss": 3.8388,
+      "step": 719
+    },
+    {
+      "epoch": 0.8053691275167785,
+      "grad_norm": 2.202491283416748,
+      "learning_rate": 4.529600724541022e-06,
+      "loss": 3.8412,
+      "step": 720
+    },
+    {
+      "epoch": 0.8064876957494407,
+      "grad_norm": 3.0058343410491943,
+      "learning_rate": 4.479295161745109e-06,
+      "loss": 4.1481,
+      "step": 721
+    },
+    {
+      "epoch": 0.8076062639821029,
+      "grad_norm": 2.1180286407470703,
+      "learning_rate": 4.4292430047396914e-06,
+      "loss": 3.8576,
+      "step": 722
+    },
+    {
+      "epoch": 0.8087248322147651,
+      "grad_norm": 2.6063430309295654,
+      "learning_rate": 4.379444871608124e-06,
+      "loss": 3.7986,
+      "step": 723
+    },
+    {
+      "epoch": 0.8098434004474273,
+      "grad_norm": 2.227067470550537,
+      "learning_rate": 4.329901377296902e-06,
+      "loss": 3.5418,
+      "step": 724
+    },
+    {
+      "epoch": 0.8109619686800895,
+      "grad_norm": 2.3542473316192627,
+      "learning_rate": 4.280613133608028e-06,
+      "loss": 3.7732,
+      "step": 725
+    },
+    {
+      "epoch": 0.8120805369127517,
+      "grad_norm": 3.143573045730591,
+      "learning_rate": 4.231580749191474e-06,
+      "loss": 4.1625,
+      "step": 726
+    },
+    {
+      "epoch": 0.8131991051454138,
+      "grad_norm": 2.3651700019836426,
+      "learning_rate": 4.182804829537654e-06,
+      "loss": 4.3148,
+      "step": 727
+    },
+    {
+      "epoch": 0.814317673378076,
+      "grad_norm": 2.07985258102417,
+      "learning_rate": 4.134285976969948e-06,
+      "loss": 3.8726,
+      "step": 728
+    },
+    {
+      "epoch": 0.8154362416107382,
+      "grad_norm": 3.1206445693969727,
+      "learning_rate": 4.086024790637285e-06,
+      "loss": 3.6906,
+      "step": 729
+    },
+    {
+      "epoch": 0.8165548098434005,
+      "grad_norm": 2.517094850540161,
+      "learning_rate": 4.038021866506725e-06,
+      "loss": 4.0283,
+      "step": 730
+    },
+    {
+      "epoch": 0.8176733780760627,
+      "grad_norm": 2.244453191757202,
+      "learning_rate": 3.990277797356104e-06,
+      "loss": 3.959,
+      "step": 731
+    },
+    {
+      "epoch": 0.8187919463087249,
+      "grad_norm": 1.9631199836730957,
+      "learning_rate": 3.942793172766699e-06,
+      "loss": 4.2073,
+      "step": 732
+    },
+    {
+      "epoch": 0.819910514541387,
+      "grad_norm": 2.9783220291137695,
+      "learning_rate": 3.895568579115983e-06,
+      "loss": 3.7062,
+      "step": 733
+    },
+    {
+      "epoch": 0.8210290827740492,
+      "grad_norm": 2.1372721195220947,
+      "learning_rate": 3.848604599570338e-06,
+      "loss": 3.7815,
+      "step": 734
+    },
+    {
+      "epoch": 0.8221476510067114,
+      "grad_norm": 2.8807451725006104,
+      "learning_rate": 3.80190181407791e-06,
+      "loss": 3.639,
+      "step": 735
+    },
+    {
+      "epoch": 0.8232662192393736,
+      "grad_norm": 1.9709060192108154,
+      "learning_rate": 3.755460799361382e-06,
+      "loss": 3.8587,
+      "step": 736
+    },
+    {
+      "epoch": 0.8243847874720358,
+      "grad_norm": 2.44718861579895,
+      "learning_rate": 3.709282128910907e-06,
+      "loss": 3.9482,
+      "step": 737
+    },
+    {
+      "epoch": 0.825503355704698,
+      "grad_norm": 2.036914110183716,
+      "learning_rate": 3.6633663729770008e-06,
+      "loss": 3.9106,
+      "step": 738
+    },
+    {
+      "epoch": 0.8266219239373602,
+      "grad_norm": 2.3819079399108887,
+      "learning_rate": 3.61771409856351e-06,
+      "loss": 4.0078,
+      "step": 739
+    },
+    {
+      "epoch": 0.8277404921700223,
+      "grad_norm": 2.206578493118286,
+      "learning_rate": 3.572325869420587e-06,
+      "loss": 3.839,
+      "step": 740
+    },
+    {
+      "epoch": 0.8288590604026845,
+      "grad_norm": 2.039262294769287,
+      "learning_rate": 3.527202246037756e-06,
+      "loss": 3.8825,
+      "step": 741
+    },
+    {
+      "epoch": 0.8299776286353467,
+      "grad_norm": 2.297976493835449,
+      "learning_rate": 3.4823437856369794e-06,
+      "loss": 3.9343,
+      "step": 742
+    },
+    {
+      "epoch": 0.831096196868009,
+      "grad_norm": 2.6157867908477783,
+      "learning_rate": 3.4377510421657906e-06,
+      "loss": 3.7783,
+      "step": 743
+    },
+    {
+      "epoch": 0.8322147651006712,
+      "grad_norm": 2.9260687828063965,
+      "learning_rate": 3.393424566290421e-06,
+      "loss": 3.8069,
+      "step": 744
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 4.025168418884277,
+      "learning_rate": 3.3493649053890326e-06,
+      "loss": 3.743,
+      "step": 745
+    },
+    {
+      "epoch": 0.8344519015659956,
+      "grad_norm": 2.532195806503296,
+      "learning_rate": 3.3055726035449484e-06,
+      "loss": 3.6006,
+      "step": 746
+    },
+    {
+      "epoch": 0.8355704697986577,
+      "grad_norm": 3.2442009449005127,
+      "learning_rate": 3.2620482015399302e-06,
+      "loss": 3.7358,
+      "step": 747
+    },
+    {
+      "epoch": 0.8366890380313199,
+      "grad_norm": 4.921333312988281,
+      "learning_rate": 3.2187922368474952e-06,
+      "loss": 3.8525,
+      "step": 748
+    },
+    {
+      "epoch": 0.8378076062639821,
+      "grad_norm": 2.6702213287353516,
+      "learning_rate": 3.175805243626284e-06,
+      "loss": 3.8665,
+      "step": 749
+    },
+    {
+      "epoch": 0.8389261744966443,
+      "grad_norm": 4.156053066253662,
+      "learning_rate": 3.133087752713479e-06,
+      "loss": 4.2799,
+      "step": 750
+    },
+    {
+      "epoch": 0.8400447427293065,
+      "grad_norm": 2.0885608196258545,
+      "learning_rate": 3.0906402916182297e-06,
+      "loss": 3.8346,
+      "step": 751
+    },
+    {
+      "epoch": 0.8411633109619687,
+      "grad_norm": 2.58349871635437,
+      "learning_rate": 3.0484633845151488e-06,
+      "loss": 3.7313,
+      "step": 752
+    },
+    {
+      "epoch": 0.8422818791946308,
+      "grad_norm": 3.2410836219787598,
+      "learning_rate": 3.0065575522378264e-06,
+      "loss": 3.7262,
+      "step": 753
+    },
+    {
+      "epoch": 0.843400447427293,
+      "grad_norm": 1.6323789358139038,
+      "learning_rate": 2.9649233122724105e-06,
+      "loss": 3.9603,
+      "step": 754
+    },
+    {
+      "epoch": 0.8445190156599552,
+      "grad_norm": 1.915100336074829,
+      "learning_rate": 2.923561178751219e-06,
+      "loss": 4.0198,
+      "step": 755
+    },
+    {
+      "epoch": 0.8456375838926175,
+      "grad_norm": 2.1250874996185303,
+      "learning_rate": 2.88247166244639e-06,
+      "loss": 3.9345,
+      "step": 756
+    },
+    {
+      "epoch": 0.8467561521252797,
+      "grad_norm": 3.2116942405700684,
+      "learning_rate": 2.841655270763549e-06,
+      "loss": 3.6633,
+      "step": 757
+    },
+    {
+      "epoch": 0.8478747203579419,
+      "grad_norm": 2.246950626373291,
+      "learning_rate": 2.801112507735587e-06,
+      "loss": 3.9842,
+      "step": 758
+    },
+    {
+      "epoch": 0.8489932885906041,
+      "grad_norm": 1.9269198179244995,
+      "learning_rate": 2.760843874016403e-06,
+      "loss": 4.043,
+      "step": 759
+    },
+    {
+      "epoch": 0.8501118568232662,
+      "grad_norm": 3.6027183532714844,
+      "learning_rate": 2.720849866874739e-06,
+      "loss": 3.693,
+      "step": 760
+    },
+    {
+      "epoch": 0.8512304250559284,
+      "grad_norm": 2.76350474357605,
+      "learning_rate": 2.6811309801880208e-06,
+      "loss": 4.1343,
+      "step": 761
+    },
+    {
+      "epoch": 0.8523489932885906,
+      "grad_norm": 2.693434238433838,
+      "learning_rate": 2.6416877044362685e-06,
+      "loss": 4.3856,
+      "step": 762
+    },
+    {
+      "epoch": 0.8534675615212528,
+      "grad_norm": 2.068485736846924,
+      "learning_rate": 2.602520526696059e-06,
+      "loss": 4.0551,
+      "step": 763
+    },
+    {
+      "epoch": 0.854586129753915,
+      "grad_norm": 2.524249792098999,
+      "learning_rate": 2.563629930634487e-06,
+      "loss": 3.776,
+      "step": 764
+    },
+    {
+      "epoch": 0.8557046979865772,
+      "grad_norm": 2.40083909034729,
+      "learning_rate": 2.525016396503185e-06,
+      "loss": 3.8733,
+      "step": 765
+    },
+    {
+      "epoch": 0.8568232662192393,
+      "grad_norm": 2.9691314697265625,
+      "learning_rate": 2.4866804011324296e-06,
+      "loss": 3.7485,
+      "step": 766
+    },
+    {
+      "epoch": 0.8579418344519015,
+      "grad_norm": 3.3835642337799072,
+      "learning_rate": 2.448622417925214e-06,
+      "loss": 3.7207,
+      "step": 767
+    },
+    {
+      "epoch": 0.8590604026845637,
+      "grad_norm": 2.4943041801452637,
+      "learning_rate": 2.4108429168514245e-06,
+      "loss": 3.4865,
+      "step": 768
+    },
+    {
+      "epoch": 0.860178970917226,
+      "grad_norm": 2.2213940620422363,
+      "learning_rate": 2.3733423644420353e-06,
+      "loss": 3.9459,
+      "step": 769
+    },
+    {
+      "epoch": 0.8612975391498882,
+      "grad_norm": 2.157849073410034,
+      "learning_rate": 2.3361212237833356e-06,
+      "loss": 3.8979,
+      "step": 770
+    },
+    {
+      "epoch": 0.8624161073825504,
+      "grad_norm": 2.031329870223999,
+      "learning_rate": 2.2991799545112215e-06,
+      "loss": 3.9194,
+      "step": 771
+    },
+    {
+      "epoch": 0.8635346756152126,
+      "grad_norm": 3.01249098777771,
+      "learning_rate": 2.2625190128055168e-06,
+      "loss": 3.7412,
+      "step": 772
+    },
+    {
+      "epoch": 0.8646532438478747,
+      "grad_norm": 2.1558024883270264,
+      "learning_rate": 2.2261388513843515e-06,
+      "loss": 4.12,
+      "step": 773
+    },
+    {
+      "epoch": 0.8657718120805369,
+      "grad_norm": 3.150437593460083,
+      "learning_rate": 2.190039919498543e-06,
+      "loss": 3.8121,
+      "step": 774
+    },
+    {
+      "epoch": 0.8668903803131991,
+      "grad_norm": 2.319565534591675,
+      "learning_rate": 2.154222662926067e-06,
+      "loss": 3.8679,
+      "step": 775
+    },
+    {
+      "epoch": 0.8680089485458613,
+      "grad_norm": 2.686678171157837,
+      "learning_rate": 2.118687523966559e-06,
+      "loss": 4.2429,
+      "step": 776
+    },
+    {
+      "epoch": 0.8691275167785235,
+      "grad_norm": 2.2821829319000244,
+      "learning_rate": 2.0834349414358495e-06,
+      "loss": 3.9263,
+      "step": 777
+    },
+    {
+      "epoch": 0.8702460850111857,
+      "grad_norm": 2.5908377170562744,
+      "learning_rate": 2.048465350660522e-06,
+      "loss": 3.4955,
+      "step": 778
+    },
+    {
+      "epoch": 0.8713646532438478,
+      "grad_norm": 3.3101370334625244,
+      "learning_rate": 2.013779183472575e-06,
+      "loss": 4.0192,
+      "step": 779
+    },
+    {
+      "epoch": 0.87248322147651,
+      "grad_norm": 2.681217908859253,
+      "learning_rate": 1.9793768682040524e-06,
+      "loss": 3.6952,
+      "step": 780
+    },
+    {
+      "epoch": 0.8736017897091722,
+      "grad_norm": 2.5683112144470215,
+      "learning_rate": 1.9452588296817843e-06,
+      "loss": 3.9709,
+      "step": 781
+    },
+    {
+      "epoch": 0.8747203579418344,
+      "grad_norm": 3.1038596630096436,
+      "learning_rate": 1.911425489222127e-06,
+      "loss": 4.0497,
+      "step": 782
+    },
+    {
+      "epoch": 0.8758389261744967,
+      "grad_norm": 2.36666202545166,
+      "learning_rate": 1.8778772646257491e-06,
+      "loss": 4.0068,
+      "step": 783
+    },
+    {
+      "epoch": 0.8769574944071589,
+      "grad_norm": 2.2028393745422363,
+      "learning_rate": 1.8446145701724983e-06,
+      "loss": 3.6578,
+      "step": 784
+    },
+    {
+      "epoch": 0.8780760626398211,
+      "grad_norm": 4.168335437774658,
+      "learning_rate": 1.8116378166162685e-06,
+      "loss": 4.1479,
+      "step": 785
+    },
+    {
+      "epoch": 0.8791946308724832,
+      "grad_norm": 2.6044557094573975,
+      "learning_rate": 1.778947411179932e-06,
+      "loss": 3.573,
+      "step": 786
+    },
+    {
+      "epoch": 0.8803131991051454,
+      "grad_norm": 1.6643822193145752,
+      "learning_rate": 1.7465437575502952e-06,
+      "loss": 3.8277,
+      "step": 787
+    },
+    {
+      "epoch": 0.8814317673378076,
+      "grad_norm": 2.9622278213500977,
+      "learning_rate": 1.7144272558731467e-06,
+      "loss": 3.6871,
+      "step": 788
+    },
+    {
+      "epoch": 0.8825503355704698,
+      "grad_norm": 2.0143864154815674,
+      "learning_rate": 1.6825983027482867e-06,
+      "loss": 3.9285,
+      "step": 789
+    },
+    {
+      "epoch": 0.883668903803132,
+      "grad_norm": 2.577881097793579,
+      "learning_rate": 1.6510572912246475e-06,
+      "loss": 3.6799,
+      "step": 790
+    },
+    {
+      "epoch": 0.8847874720357942,
+      "grad_norm": 2.5292065143585205,
+      "learning_rate": 1.6198046107954223e-06,
+      "loss": 3.9658,
+      "step": 791
+    },
+    {
+      "epoch": 0.8859060402684564,
+      "grad_norm": 2.8256359100341797,
+      "learning_rate": 1.5888406473932692e-06,
+      "loss": 3.7874,
+      "step": 792
+    },
+    {
+      "epoch": 0.8870246085011185,
+      "grad_norm": 4.170554161071777,
+      "learning_rate": 1.5581657833855512e-06,
+      "loss": 3.6907,
+      "step": 793
+    },
+    {
+      "epoch": 0.8881431767337807,
+      "grad_norm": 1.9856733083724976,
+      "learning_rate": 1.5277803975695864e-06,
+      "loss": 3.8945,
+      "step": 794
+    },
+    {
+      "epoch": 0.889261744966443,
+      "grad_norm": 2.3850162029266357,
+      "learning_rate": 1.497684865168006e-06,
+      "loss": 3.7603,
+      "step": 795
+    },
+    {
+      "epoch": 0.8903803131991052,
+      "grad_norm": 2.514587879180908,
+      "learning_rate": 1.4678795578240894e-06,
+      "loss": 3.9449,
+      "step": 796
+    },
+    {
+      "epoch": 0.8914988814317674,
+      "grad_norm": 3.2767693996429443,
+      "learning_rate": 1.4383648435972007e-06,
+      "loss": 3.9914,
+      "step": 797
+    },
+    {
+      "epoch": 0.8926174496644296,
+      "grad_norm": 2.3066697120666504,
+      "learning_rate": 1.4091410869582267e-06,
+      "loss": 3.7617,
+      "step": 798
+    },
+    {
+      "epoch": 0.8937360178970917,
+      "grad_norm": 2.5513951778411865,
+      "learning_rate": 1.3802086487850807e-06,
+      "loss": 3.7506,
+      "step": 799
+    },
+    {
+      "epoch": 0.8948545861297539,
+      "grad_norm": 3.390204429626465,
+      "learning_rate": 1.3515678863582431e-06,
+      "loss": 3.6553,
+      "step": 800
+    },
+    {
+      "epoch": 0.8959731543624161,
+      "grad_norm": 1.921186923980713,
+      "learning_rate": 1.3232191533563586e-06,
+      "loss": 3.7809,
+      "step": 801
+    },
+    {
+      "epoch": 0.8970917225950783,
+      "grad_norm": 2.1258394718170166,
+      "learning_rate": 1.2951627998518623e-06,
+      "loss": 3.8879,
+      "step": 802
+    },
+    {
+      "epoch": 0.8982102908277405,
+      "grad_norm": 3.5940604209899902,
+      "learning_rate": 1.267399172306663e-06,
+      "loss": 3.5702,
+      "step": 803
+    },
+    {
+      "epoch": 0.8993288590604027,
+      "grad_norm": 3.0253639221191406,
+      "learning_rate": 1.2399286135678423e-06,
+      "loss": 3.9817,
+      "step": 804
+    },
+    {
+      "epoch": 0.9004474272930649,
+      "grad_norm": 3.926387071609497,
+      "learning_rate": 1.212751462863454e-06,
+      "loss": 3.449,
+      "step": 805
+    },
+    {
+      "epoch": 0.901565995525727,
+      "grad_norm": 3.5187435150146484,
+      "learning_rate": 1.1858680557983171e-06,
+      "loss": 3.8943,
+      "step": 806
+    },
+    {
+      "epoch": 0.9026845637583892,
+      "grad_norm": 2.5346837043762207,
+      "learning_rate": 1.1592787243498631e-06,
+      "loss": 3.98,
+      "step": 807
+    },
+    {
+      "epoch": 0.9038031319910514,
+      "grad_norm": 2.5961804389953613,
+      "learning_rate": 1.132983796864054e-06,
+      "loss": 4.0311,
+      "step": 808
+    },
+    {
+      "epoch": 0.9049217002237137,
+      "grad_norm": 2.107836961746216,
+      "learning_rate": 1.10698359805132e-06,
+      "loss": 3.6457,
+      "step": 809
+    },
+    {
+      "epoch": 0.9060402684563759,
+      "grad_norm": 2.8412914276123047,
+      "learning_rate": 1.0812784489825507e-06,
+      "loss": 4.1803,
+      "step": 810
+    },
+    {
+      "epoch": 0.9071588366890381,
+      "grad_norm": 2.366673469543457,
+      "learning_rate": 1.0558686670851303e-06,
+      "loss": 3.9735,
+      "step": 811
+    },
+    {
+      "epoch": 0.9082774049217002,
+      "grad_norm": 2.198279857635498,
+      "learning_rate": 1.0307545661390139e-06,
+      "loss": 3.7553,
+      "step": 812
+    },
+    {
+      "epoch": 0.9093959731543624,
+      "grad_norm": 2.628816604614258,
+      "learning_rate": 1.0059364562728518e-06,
+      "loss": 3.7233,
+      "step": 813
+    },
+    {
+      "epoch": 0.9105145413870246,
+      "grad_norm": 2.737968921661377,
+      "learning_rate": 9.81414643960174e-07,
+      "loss": 3.8348,
+      "step": 814
+    },
+    {
+      "epoch": 0.9116331096196868,
+      "grad_norm": 1.8730359077453613,
+      "learning_rate": 9.5718943201559e-07,
+      "loss": 3.946,
+      "step": 815
+    },
+    {
+      "epoch": 0.912751677852349,
+      "grad_norm": 2.630864143371582,
+      "learning_rate": 9.332611195910584e-07,
+      "loss": 3.9293,
+      "step": 816
+    },
+    {
+      "epoch": 0.9138702460850112,
+      "grad_norm": 2.5964674949645996,
+      "learning_rate": 9.09630002172182e-07,
+      "loss": 3.9155,
+      "step": 817
+    },
+    {
+      "epoch": 0.9149888143176734,
+      "grad_norm": 1.8273096084594727,
+      "learning_rate": 8.862963715745687e-07,
+      "loss": 3.9176,
+      "step": 818
+    },
+    {
+      "epoch": 0.9161073825503355,
+      "grad_norm": 2.2656948566436768,
+      "learning_rate": 8.632605159402341e-07,
+      "loss": 4.1211,
+      "step": 819
+    },
+    {
+      "epoch": 0.9172259507829977,
+      "grad_norm": 2.1946887969970703,
+      "learning_rate": 8.405227197340216e-07,
+      "loss": 4.3939,
+      "step": 820
+    },
+    {
+      "epoch": 0.9183445190156599,
+      "grad_norm": 2.742736577987671,
+      "learning_rate": 8.180832637401075e-07,
+      "loss": 3.5518,
+      "step": 821
+    },
+    {
+      "epoch": 0.9194630872483222,
+      "grad_norm": 2.608285665512085,
+      "learning_rate": 7.959424250585323e-07,
+      "loss": 3.6809,
+      "step": 822
+    },
+    {
+      "epoch": 0.9205816554809844,
+      "grad_norm": 2.192291498184204,
+      "learning_rate": 7.741004771017691e-07,
+      "loss": 4.1284,
+      "step": 823
+    },
+    {
+      "epoch": 0.9217002237136466,
+      "grad_norm": 2.3379416465759277,
+      "learning_rate": 7.525576895913655e-07,
+      "loss": 3.8505,
+      "step": 824
+    },
+    {
+      "epoch": 0.9228187919463087,
+      "grad_norm": 2.133988618850708,
+      "learning_rate": 7.313143285545832e-07,
+      "loss": 3.9045,
+      "step": 825
+    },
+    {
+      "epoch": 0.9239373601789709,
+      "grad_norm": 2.171081781387329,
+      "learning_rate": 7.10370656321141e-07,
+      "loss": 3.7649,
+      "step": 826
+    },
+    {
+      "epoch": 0.9250559284116331,
+      "grad_norm": 1.7111164331436157,
+      "learning_rate": 6.897269315199628e-07,
+      "loss": 3.6116,
+      "step": 827
+    },
+    {
+      "epoch": 0.9261744966442953,
+      "grad_norm": 2.5523173809051514,
+      "learning_rate": 6.69383409075991e-07,
+      "loss": 4.0026,
+      "step": 828
+    },
+    {
+      "epoch": 0.9272930648769575,
+      "grad_norm": 2.1570398807525635,
+      "learning_rate": 6.493403402070247e-07,
+      "loss": 3.8648,
+      "step": 829
+    },
+    {
+      "epoch": 0.9284116331096197,
+      "grad_norm": 1.9127180576324463,
+      "learning_rate": 6.295979724206313e-07,
+      "loss": 3.9427,
+      "step": 830
+    },
+    {
+      "epoch": 0.9295302013422819,
+      "grad_norm": 2.699845790863037,
+      "learning_rate": 6.101565495110817e-07,
+      "loss": 3.5363,
+      "step": 831
+    },
+    {
+      "epoch": 0.930648769574944,
+      "grad_norm": 3.6851859092712402,
+      "learning_rate": 5.910163115563471e-07,
+      "loss": 3.7037,
+      "step": 832
+    },
+    {
+      "epoch": 0.9317673378076062,
+      "grad_norm": 2.0086777210235596,
+      "learning_rate": 5.721774949151298e-07,
+      "loss": 3.8431,
+      "step": 833
+    },
+    {
+      "epoch": 0.9328859060402684,
+      "grad_norm": 1.9770376682281494,
+      "learning_rate": 5.536403322239369e-07,
+      "loss": 3.8387,
+      "step": 834
+    },
+    {
+      "epoch": 0.9340044742729307,
+      "grad_norm": 2.4417667388916016,
+      "learning_rate": 5.35405052394225e-07,
+      "loss": 3.8362,
+      "step": 835
+    },
+    {
+      "epoch": 0.9351230425055929,
+      "grad_norm": 2.8682196140289307,
+      "learning_rate": 5.17471880609563e-07,
+      "loss": 3.8365,
+      "step": 836
+    },
+    {
+      "epoch": 0.9362416107382551,
+      "grad_norm": 2.904750108718872,
+      "learning_rate": 4.998410383228458e-07,
+      "loss": 3.7828,
+      "step": 837
+    },
+    {
+      "epoch": 0.9373601789709173,
+      "grad_norm": 2.7965922355651855,
+      "learning_rate": 4.825127432535714e-07,
+      "loss": 3.5812,
+      "step": 838
+    },
+    {
+      "epoch": 0.9384787472035794,
+      "grad_norm": 1.820050835609436,
+      "learning_rate": 4.654872093851487e-07,
+      "loss": 4.0985,
+      "step": 839
+    },
+    {
+      "epoch": 0.9395973154362416,
+      "grad_norm": 2.777374029159546,
+      "learning_rate": 4.487646469622464e-07,
+      "loss": 3.5268,
+      "step": 840
+    },
+    {
+      "epoch": 0.9407158836689038,
+      "grad_norm": 2.9808647632598877,
+      "learning_rate": 4.3234526248820686e-07,
+      "loss": 3.5723,
+      "step": 841
+    },
+    {
+      "epoch": 0.941834451901566,
+      "grad_norm": 3.188647985458374,
+      "learning_rate": 4.162292587224947e-07,
+      "loss": 3.3647,
+      "step": 842
+    },
+    {
+      "epoch": 0.9429530201342282,
+      "grad_norm": 2.2936878204345703,
+      "learning_rate": 4.0041683467819115e-07,
+      "loss": 3.887,
+      "step": 843
+    },
+    {
+      "epoch": 0.9440715883668904,
+      "grad_norm": 2.2591426372528076,
+      "learning_rate": 3.8490818561953414e-07,
+      "loss": 3.8009,
+      "step": 844
+    },
+    {
+      "epoch": 0.9451901565995525,
+      "grad_norm": 2.7897520065307617,
+      "learning_rate": 3.697035030595125e-07,
+      "loss": 3.556,
+      "step": 845
+    },
+    {
+      "epoch": 0.9463087248322147,
+      "grad_norm": 2.164191246032715,
+      "learning_rate": 3.548029747574927e-07,
+      "loss": 3.7152,
+      "step": 846
+    },
+    {
+      "epoch": 0.9474272930648769,
+      "grad_norm": 2.1192331314086914,
+      "learning_rate": 3.4020678471690934e-07,
+      "loss": 4.0369,
+      "step": 847
+    },
+    {
+      "epoch": 0.9485458612975392,
+      "grad_norm": 2.3934147357940674,
+      "learning_rate": 3.2591511318298686e-07,
+      "loss": 4.0403,
+      "step": 848
+    },
+    {
+      "epoch": 0.9496644295302014,
+      "grad_norm": 2.3561530113220215,
+      "learning_rate": 3.119281366405213e-07,
+      "loss": 3.7943,
+      "step": 849
+    },
+    {
+      "epoch": 0.9507829977628636,
+      "grad_norm": 2.1101136207580566,
+      "learning_rate": 2.9824602781168833e-07,
+      "loss": 3.9526,
+      "step": 850
+    },
+    {
+      "epoch": 0.9519015659955258,
+      "grad_norm": 2.3856959342956543,
+      "learning_rate": 2.8486895565391913e-07,
+      "loss": 3.7386,
+      "step": 851
+    },
+    {
+      "epoch": 0.9530201342281879,
+      "grad_norm": 2.2807648181915283,
+      "learning_rate": 2.7179708535781943e-07,
+      "loss": 3.6456,
+      "step": 852
+    },
+    {
+      "epoch": 0.9541387024608501,
+      "grad_norm": 2.5656046867370605,
+      "learning_rate": 2.590305783451097e-07,
+      "loss": 3.8164,
+      "step": 853
+    },
+    {
+      "epoch": 0.9552572706935123,
+      "grad_norm": 2.58715558052063,
+      "learning_rate": 2.4656959226665446e-07,
+      "loss": 4.0307,
+      "step": 854
+    },
+    {
+      "epoch": 0.9563758389261745,
+      "grad_norm": 2.5126988887786865,
+      "learning_rate": 2.3441428100050566e-07,
+      "loss": 3.7051,
+      "step": 855
+    },
+    {
+      "epoch": 0.9574944071588367,
+      "grad_norm": 2.549105644226074,
+      "learning_rate": 2.2256479464999313e-07,
+      "loss": 3.8441,
+      "step": 856
+    },
+    {
+      "epoch": 0.9586129753914989,
+      "grad_norm": 2.6297028064727783,
+      "learning_rate": 2.110212795418953e-07,
+      "loss": 3.4869,
+      "step": 857
+    },
+    {
+      "epoch": 0.959731543624161,
+      "grad_norm": 1.7067815065383911,
+      "learning_rate": 1.9978387822460197e-07,
+      "loss": 3.8855,
+      "step": 858
+    },
+    {
+      "epoch": 0.9608501118568232,
+      "grad_norm": 2.1829380989074707,
+      "learning_rate": 1.8885272946637944e-07,
+      "loss": 3.7166,
+      "step": 859
+    },
+    {
+      "epoch": 0.9619686800894854,
+      "grad_norm": 2.6756093502044678,
+      "learning_rate": 1.782279682536414e-07,
+      "loss": 3.706,
+      "step": 860
+    },
+    {
+      "epoch": 0.9630872483221476,
+      "grad_norm": 2.74208664894104,
+      "learning_rate": 1.679097257892892e-07,
+      "loss": 4.2839,
+      "step": 861
+    },
+    {
+      "epoch": 0.9642058165548099,
+      "grad_norm": 2.6451363563537598,
+      "learning_rate": 1.578981294910936e-07,
+      "loss": 3.8752,
+      "step": 862
+    },
+    {
+      "epoch": 0.9653243847874721,
+      "grad_norm": 2.704803705215454,
+      "learning_rate": 1.4819330299011004e-07,
+      "loss": 3.9167,
+      "step": 863
+    },
+    {
+      "epoch": 0.9664429530201343,
+      "grad_norm": 1.9471423625946045,
+      "learning_rate": 1.38795366129163e-07,
+      "loss": 3.9097,
+      "step": 864
+    },
+    {
+      "epoch": 0.9675615212527964,
+      "grad_norm": 2.0253708362579346,
+      "learning_rate": 1.2970443496136407e-07,
+      "loss": 3.8803,
+      "step": 865
+    },
+    {
+      "epoch": 0.9686800894854586,
+      "grad_norm": 1.8609451055526733,
+      "learning_rate": 1.2092062174867413e-07,
+      "loss": 3.8417,
+      "step": 866
+    },
+    {
+      "epoch": 0.9697986577181208,
+      "grad_norm": 2.4085533618927,
+      "learning_rate": 1.1244403496052658e-07,
+      "loss": 3.7451,
+      "step": 867
+    },
+    {
+      "epoch": 0.970917225950783,
+      "grad_norm": 3.3671045303344727,
+      "learning_rate": 1.042747792724702e-07,
+      "loss": 3.7212,
+      "step": 868
+    },
+    {
+      "epoch": 0.9720357941834452,
+      "grad_norm": 1.9560606479644775,
+      "learning_rate": 9.641295556489793e-08,
+      "loss": 3.9571,
+      "step": 869
+    },
+    {
+      "epoch": 0.9731543624161074,
+      "grad_norm": 1.984646201133728,
+      "learning_rate": 8.885866092178952e-08,
+      "loss": 3.9128,
+      "step": 870
+    },
+    {
+      "epoch": 0.9742729306487695,
+      "grad_norm": 2.1129088401794434,
+      "learning_rate": 8.161198862950692e-08,
+      "loss": 3.8933,
+      "step": 871
+    },
+    {
+      "epoch": 0.9753914988814317,
+      "grad_norm": 2.0977556705474854,
+      "learning_rate": 7.46730281756619e-08,
+      "loss": 3.9295,
+      "step": 872
+    },
+    {
+      "epoch": 0.9765100671140939,
+      "grad_norm": 2.7585601806640625,
+      "learning_rate": 6.804186524798362e-08,
+      "loss": 3.8923,
+      "step": 873
+    },
+    {
+      "epoch": 0.9776286353467561,
+      "grad_norm": 3.502300977706909,
+      "learning_rate": 6.171858173328604e-08,
+      "loss": 3.9105,
+      "step": 874
+    },
+    {
+      "epoch": 0.9787472035794184,
+      "grad_norm": 3.1645312309265137,
+      "learning_rate": 5.5703255716446637e-08,
+      "loss": 3.912,
+      "step": 875
+    },
+    {
+      "epoch": 0.9798657718120806,
+      "grad_norm": 3.4221417903900146,
+      "learning_rate": 4.999596147943486e-08,
+      "loss": 3.8685,
+      "step": 876
+    },
+    {
+      "epoch": 0.9809843400447428,
+      "grad_norm": 3.9628350734710693,
+      "learning_rate": 4.4596769500407366e-08,
+      "loss": 3.8902,
+      "step": 877
+    },
+    {
+      "epoch": 0.9821029082774049,
+      "grad_norm": 2.23947811126709,
+      "learning_rate": 3.9505746452830896e-08,
+      "loss": 3.7735,
+      "step": 878
+    },
+    {
+      "epoch": 0.9832214765100671,
+      "grad_norm": 3.1684134006500244,
+      "learning_rate": 3.4722955204652406e-08,
+      "loss": 3.7775,
+      "step": 879
+    },
+    {
+      "epoch": 0.9843400447427293,
+      "grad_norm": 2.4224395751953125,
+      "learning_rate": 3.02484548175469e-08,
+      "loss": 3.8073,
+      "step": 880
+    },
+    {
+      "epoch": 0.9854586129753915,
+      "grad_norm": 2.877962589263916,
+      "learning_rate": 2.6082300546154127e-08,
+      "loss": 3.9051,
+      "step": 881
+    },
+    {
+      "epoch": 0.9865771812080537,
+      "grad_norm": 1.9865648746490479,
+      "learning_rate": 2.2224543837423562e-08,
+      "loss": 3.8927,
+      "step": 882
+    },
+    {
+      "epoch": 0.9876957494407159,
+      "grad_norm": 2.3272900581359863,
+      "learning_rate": 1.8675232329967708e-08,
+      "loss": 3.9353,
+      "step": 883
+    },
+    {
+      "epoch": 0.9888143176733781,
+      "grad_norm": 1.9652239084243774,
+      "learning_rate": 1.5434409853473662e-08,
+      "loss": 3.8841,
+      "step": 884
+    },
+    {
+      "epoch": 0.9899328859060402,
+      "grad_norm": 3.6001975536346436,
+      "learning_rate": 1.2502116428161902e-08,
+      "loss": 4.0209,
+      "step": 885
+    },
+    {
+      "epoch": 0.9910514541387024,
+      "grad_norm": 2.0110864639282227,
+      "learning_rate": 9.878388264300541e-09,
+      "loss": 3.9221,
+      "step": 886
+    },
+    {
+      "epoch": 0.9921700223713646,
+      "grad_norm": 2.5183699131011963,
+      "learning_rate": 7.563257761744601e-09,
+      "loss": 3.6199,
+      "step": 887
+    },
+    {
+      "epoch": 0.9932885906040269,
+      "grad_norm": 2.3704769611358643,
+      "learning_rate": 5.55675350954743e-09,
+      "loss": 3.603,
+      "step": 888
+    },
+    {
+      "epoch": 0.9944071588366891,
+      "grad_norm": 2.055521249771118,
+      "learning_rate": 3.8589002855971e-09,
+      "loss": 3.906,
+      "step": 889
+    },
+    {
+      "epoch": 0.9955257270693513,
+      "grad_norm": 2.1967926025390625,
+      "learning_rate": 2.4697190563194305e-09,
+      "loss": 3.9161,
+      "step": 890
+    },
+    {
+      "epoch": 0.9966442953020134,
+      "grad_norm": 3.107943534851074,
+      "learning_rate": 1.3892269764198551e-09,
+      "loss": 3.9214,
+      "step": 891
+    },
+    {
+      "epoch": 0.9977628635346756,
+      "grad_norm": 2.653813123703003,
+      "learning_rate": 6.174373886586037e-10,
+      "loss": 3.7566,
+      "step": 892
+    },
+    {
+      "epoch": 0.9988814317673378,
+      "grad_norm": 2.262310743331909,
+      "learning_rate": 1.5435982370359813e-10,
+      "loss": 3.6915,
+      "step": 893
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.972426414489746,
+      "learning_rate": 0.0,
+      "loss": 4.5787,
+      "step": 894
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 894,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 5000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.4928497725151232e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}