diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5297 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9926958831341302,
+  "eval_steps": 500,
+  "global_step": 752,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0026542800265428003,
+      "grad_norm": 6.981875026564461,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.5142,
+      "step": 1
+    },
+    {
+      "epoch": 0.0053085600530856005,
+      "grad_norm": 7.3585790096388095,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.5399,
+      "step": 2
+    },
+    {
+      "epoch": 0.007962840079628402,
+      "grad_norm": 6.187456707748293,
+      "learning_rate": 3e-06,
+      "loss": 1.4331,
+      "step": 3
+    },
+    {
+      "epoch": 0.010617120106171201,
+      "grad_norm": 3.322201654326747,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.4069,
+      "step": 4
+    },
+    {
+      "epoch": 0.013271400132714002,
+      "grad_norm": 6.898750176066953,
+      "learning_rate": 5e-06,
+      "loss": 1.398,
+      "step": 5
+    },
+    {
+      "epoch": 0.015925680159256803,
+      "grad_norm": 9.84996090672252,
+      "learning_rate": 6e-06,
+      "loss": 1.4207,
+      "step": 6
+    },
+    {
+      "epoch": 0.0185799601857996,
+      "grad_norm": 5.911295227152632,
+      "learning_rate": 7e-06,
+      "loss": 1.396,
+      "step": 7
+    },
+    {
+      "epoch": 0.021234240212342402,
+      "grad_norm": 4.487244776192921,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.3842,
+      "step": 8
+    },
+    {
+      "epoch": 0.023888520238885203,
+      "grad_norm": 3.963029697998983,
+      "learning_rate": 9e-06,
+      "loss": 1.4094,
+      "step": 9
+    },
+    {
+      "epoch": 0.026542800265428004,
+      "grad_norm": 3.9061159778838777,
+      "learning_rate": 1e-05,
+      "loss": 1.3289,
+      "step": 10
+    },
+    {
+      "epoch": 0.029197080291970802,
+      "grad_norm": 4.06704445096823,
+      "learning_rate": 9.999988945517944e-06,
+      "loss": 1.3803,
+      "step": 11
+    },
+    {
+      "epoch": 0.03185136031851361,
+      "grad_norm": 2.604059841393431,
+      "learning_rate": 9.999955782120656e-06,
+      "loss": 1.3673,
+      "step": 12
+    },
+    {
+      "epoch": 0.034505640345056404,
+      "grad_norm": 2.257369986938731,
+      "learning_rate": 9.999900509954779e-06,
+      "loss": 1.3601,
+      "step": 13
+    },
+    {
+      "epoch": 0.0371599203715992,
+      "grad_norm": 2.1847487718443896,
+      "learning_rate": 9.999823129264712e-06,
+      "loss": 1.2897,
+      "step": 14
+    },
+    {
+      "epoch": 0.039814200398142006,
+      "grad_norm": 2.07192636064569,
+      "learning_rate": 9.99972364039262e-06,
+      "loss": 1.3908,
+      "step": 15
+    },
+    {
+      "epoch": 0.042468480424684804,
+      "grad_norm": 1.9501283234498499,
+      "learning_rate": 9.99960204377842e-06,
+      "loss": 1.3159,
+      "step": 16
+    },
+    {
+      "epoch": 0.0451227604512276,
+      "grad_norm": 2.493269875780714,
+      "learning_rate": 9.999458339959787e-06,
+      "loss": 1.294,
+      "step": 17
+    },
+    {
+      "epoch": 0.047777040477770406,
+      "grad_norm": 2.5949272964482057,
+      "learning_rate": 9.999292529572152e-06,
+      "loss": 1.3185,
+      "step": 18
+    },
+    {
+      "epoch": 0.050431320504313204,
+      "grad_norm": 2.2388099180255643,
+      "learning_rate": 9.99910461334869e-06,
+      "loss": 1.2132,
+      "step": 19
+    },
+    {
+      "epoch": 0.05308560053085601,
+      "grad_norm": 2.2899507956205762,
+      "learning_rate": 9.99889459212033e-06,
+      "loss": 1.293,
+      "step": 20
+    },
+    {
+      "epoch": 0.055739880557398806,
+      "grad_norm": 1.9172996859984803,
+      "learning_rate": 9.998662466815743e-06,
+      "loss": 1.2282,
+      "step": 21
+    },
+    {
+      "epoch": 0.058394160583941604,
+      "grad_norm": 1.695255684950505,
+      "learning_rate": 9.99840823846134e-06,
+      "loss": 1.2453,
+      "step": 22
+    },
+    {
+      "epoch": 0.06104844061048441,
+      "grad_norm": 1.9499380269432547,
+      "learning_rate": 9.998131908181262e-06,
+      "loss": 1.3126,
+      "step": 23
+    },
+    {
+      "epoch": 0.06370272063702721,
+      "grad_norm": 2.235039095134041,
+      "learning_rate": 9.997833477197386e-06,
+      "loss": 1.3,
+      "step": 24
+    },
+    {
+      "epoch": 0.06635700066357,
+      "grad_norm": 1.7768874556246268,
+      "learning_rate": 9.997512946829314e-06,
+      "loss": 1.3027,
+      "step": 25
+    },
+    {
+      "epoch": 0.06901128069011281,
+      "grad_norm": 1.8710962469378072,
+      "learning_rate": 9.997170318494362e-06,
+      "loss": 1.2571,
+      "step": 26
+    },
+    {
+      "epoch": 0.07166556071665561,
+      "grad_norm": 2.1786369089634734,
+      "learning_rate": 9.996805593707566e-06,
+      "loss": 1.2633,
+      "step": 27
+    },
+    {
+      "epoch": 0.0743198407431984,
+      "grad_norm": 2.3145346512057805,
+      "learning_rate": 9.996418774081658e-06,
+      "loss": 1.2439,
+      "step": 28
+    },
+    {
+      "epoch": 0.07697412076974121,
+      "grad_norm": 1.740845781272116,
+      "learning_rate": 9.996009861327077e-06,
+      "loss": 1.2437,
+      "step": 29
+    },
+    {
+      "epoch": 0.07962840079628401,
+      "grad_norm": 1.9183185283288997,
+      "learning_rate": 9.99557885725195e-06,
+      "loss": 1.333,
+      "step": 30
+    },
+    {
+      "epoch": 0.0822826808228268,
+      "grad_norm": 2.1688553875791987,
+      "learning_rate": 9.995125763762089e-06,
+      "loss": 1.3145,
+      "step": 31
+    },
+    {
+      "epoch": 0.08493696084936961,
+      "grad_norm": 1.9658120398634014,
+      "learning_rate": 9.994650582860978e-06,
+      "loss": 1.2682,
+      "step": 32
+    },
+    {
+      "epoch": 0.08759124087591241,
+      "grad_norm": 2.136505316782775,
+      "learning_rate": 9.994153316649769e-06,
+      "loss": 1.24,
+      "step": 33
+    },
+    {
+      "epoch": 0.0902455209024552,
+      "grad_norm": 1.670383957571605,
+      "learning_rate": 9.99363396732727e-06,
+      "loss": 1.2421,
+      "step": 34
+    },
+    {
+      "epoch": 0.09289980092899801,
+      "grad_norm": 1.9007693724974954,
+      "learning_rate": 9.993092537189936e-06,
+      "loss": 1.1936,
+      "step": 35
+    },
+    {
+      "epoch": 0.09555408095554081,
+      "grad_norm": 1.8427231589681057,
+      "learning_rate": 9.992529028631859e-06,
+      "loss": 1.2568,
+      "step": 36
+    },
+    {
+      "epoch": 0.0982083609820836,
+      "grad_norm": 1.9014135968523682,
+      "learning_rate": 9.991943444144758e-06,
+      "loss": 1.231,
+      "step": 37
+    },
+    {
+      "epoch": 0.10086264100862641,
+      "grad_norm": 1.714724530777744,
+      "learning_rate": 9.991335786317964e-06,
+      "loss": 1.2559,
+      "step": 38
+    },
+    {
+      "epoch": 0.10351692103516921,
+      "grad_norm": 1.9540837660082362,
+      "learning_rate": 9.990706057838417e-06,
+      "loss": 1.2583,
+      "step": 39
+    },
+    {
+      "epoch": 0.10617120106171202,
+      "grad_norm": 1.7120831927587263,
+      "learning_rate": 9.990054261490643e-06,
+      "loss": 1.2095,
+      "step": 40
+    },
+    {
+      "epoch": 0.10882548108825481,
+      "grad_norm": 1.7124302876215762,
+      "learning_rate": 9.989380400156752e-06,
+      "loss": 1.2361,
+      "step": 41
+    },
+    {
+      "epoch": 0.11147976111479761,
+      "grad_norm": 1.7683889355402804,
+      "learning_rate": 9.98868447681642e-06,
+      "loss": 1.2134,
+      "step": 42
+    },
+    {
+      "epoch": 0.11413404114134042,
+      "grad_norm": 1.8716677445605339,
+      "learning_rate": 9.987966494546873e-06,
+      "loss": 1.3081,
+      "step": 43
+    },
+    {
+      "epoch": 0.11678832116788321,
+      "grad_norm": 1.810504125507985,
+      "learning_rate": 9.987226456522884e-06,
+      "loss": 1.2789,
+      "step": 44
+    },
+    {
+      "epoch": 0.11944260119442601,
+      "grad_norm": 2.107999452852097,
+      "learning_rate": 9.986464366016743e-06,
+      "loss": 1.2965,
+      "step": 45
+    },
+    {
+      "epoch": 0.12209688122096882,
+      "grad_norm": 1.9463843496195974,
+      "learning_rate": 9.985680226398261e-06,
+      "loss": 1.2455,
+      "step": 46
+    },
+    {
+      "epoch": 0.12475116124751161,
+      "grad_norm": 1.8557616158057193,
+      "learning_rate": 9.984874041134738e-06,
+      "loss": 1.2432,
+      "step": 47
+    },
+    {
+      "epoch": 0.12740544127405443,
+      "grad_norm": 1.7060682110649106,
+      "learning_rate": 9.984045813790959e-06,
+      "loss": 1.1864,
+      "step": 48
+    },
+    {
+      "epoch": 0.1300597213005972,
+      "grad_norm": 1.9204675247056242,
+      "learning_rate": 9.983195548029173e-06,
+      "loss": 1.2525,
+      "step": 49
+    },
+    {
+      "epoch": 0.13271400132714,
+      "grad_norm": 1.8711916549013854,
+      "learning_rate": 9.98232324760908e-06,
+      "loss": 1.2836,
+      "step": 50
+    },
+    {
+      "epoch": 0.1353682813536828,
+      "grad_norm": 1.6842894937517436,
+      "learning_rate": 9.981428916387812e-06,
+      "loss": 1.183,
+      "step": 51
+    },
+    {
+      "epoch": 0.13802256138022562,
+      "grad_norm": 1.6306230130296617,
+      "learning_rate": 9.980512558319915e-06,
+      "loss": 1.2369,
+      "step": 52
+    },
+    {
+      "epoch": 0.14067684140676842,
+      "grad_norm": 1.566857719000752,
+      "learning_rate": 9.979574177457337e-06,
+      "loss": 1.1844,
+      "step": 53
+    },
+    {
+      "epoch": 0.14333112143331123,
+      "grad_norm": 2.0361026967903966,
+      "learning_rate": 9.978613777949401e-06,
+      "loss": 1.2064,
+      "step": 54
+    },
+    {
+      "epoch": 0.145985401459854,
+      "grad_norm": 1.8265620367248863,
+      "learning_rate": 9.977631364042796e-06,
+      "loss": 1.2432,
+      "step": 55
+    },
+    {
+      "epoch": 0.1486396814863968,
+      "grad_norm": 1.7838441388683621,
+      "learning_rate": 9.976626940081553e-06,
+      "loss": 1.2852,
+      "step": 56
+    },
+    {
+      "epoch": 0.1512939615129396,
+      "grad_norm": 1.7544005767854343,
+      "learning_rate": 9.975600510507025e-06,
+      "loss": 1.1735,
+      "step": 57
+    },
+    {
+      "epoch": 0.15394824153948242,
+      "grad_norm": 1.80732471729061,
+      "learning_rate": 9.974552079857873e-06,
+      "loss": 1.2198,
+      "step": 58
+    },
+    {
+      "epoch": 0.15660252156602522,
+      "grad_norm": 1.7204547120415132,
+      "learning_rate": 9.973481652770039e-06,
+      "loss": 1.2409,
+      "step": 59
+    },
+    {
+      "epoch": 0.15925680159256803,
+      "grad_norm": 1.7446267682486616,
+      "learning_rate": 9.972389233976729e-06,
+      "loss": 1.236,
+      "step": 60
+    },
+    {
+      "epoch": 0.1619110816191108,
+      "grad_norm": 1.8367615413386507,
+      "learning_rate": 9.971274828308396e-06,
+      "loss": 1.2333,
+      "step": 61
+    },
+    {
+      "epoch": 0.1645653616456536,
+      "grad_norm": 2.024298503175875,
+      "learning_rate": 9.970138440692706e-06,
+      "loss": 1.1798,
+      "step": 62
+    },
+    {
+      "epoch": 0.1672196416721964,
+      "grad_norm": 1.6639343310164172,
+      "learning_rate": 9.968980076154533e-06,
+      "loss": 1.2429,
+      "step": 63
+    },
+    {
+      "epoch": 0.16987392169873922,
+      "grad_norm": 1.7526879469365466,
+      "learning_rate": 9.967799739815925e-06,
+      "loss": 1.2448,
+      "step": 64
+    },
+    {
+      "epoch": 0.17252820172528202,
+      "grad_norm": 1.551089471898675,
+      "learning_rate": 9.966597436896085e-06,
+      "loss": 1.2221,
+      "step": 65
+    },
+    {
+      "epoch": 0.17518248175182483,
+      "grad_norm": 3.285115069392907,
+      "learning_rate": 9.965373172711343e-06,
+      "loss": 1.2576,
+      "step": 66
+    },
+    {
+      "epoch": 0.17783676177836763,
+      "grad_norm": 1.9429706042692902,
+      "learning_rate": 9.964126952675148e-06,
+      "loss": 1.2211,
+      "step": 67
+    },
+    {
+      "epoch": 0.1804910418049104,
+      "grad_norm": 1.6014290221743892,
+      "learning_rate": 9.962858782298023e-06,
+      "loss": 1.2105,
+      "step": 68
+    },
+    {
+      "epoch": 0.1831453218314532,
+      "grad_norm": 1.7597646377706344,
+      "learning_rate": 9.961568667187556e-06,
+      "loss": 1.2401,
+      "step": 69
+    },
+    {
+      "epoch": 0.18579960185799602,
+      "grad_norm": 1.6408652164338016,
+      "learning_rate": 9.960256613048367e-06,
+      "loss": 1.1577,
+      "step": 70
+    },
+    {
+      "epoch": 0.18845388188453882,
+      "grad_norm": 2.4349687274686334,
+      "learning_rate": 9.958922625682088e-06,
+      "loss": 1.193,
+      "step": 71
+    },
+    {
+      "epoch": 0.19110816191108163,
+      "grad_norm": 1.9037201335080784,
+      "learning_rate": 9.957566710987338e-06,
+      "loss": 1.1489,
+      "step": 72
+    },
+    {
+      "epoch": 0.19376244193762443,
+      "grad_norm": 2.1368452991743014,
+      "learning_rate": 9.956188874959686e-06,
+      "loss": 1.3215,
+      "step": 73
+    },
+    {
+      "epoch": 0.1964167219641672,
+      "grad_norm": 1.7342842074642177,
+      "learning_rate": 9.954789123691643e-06,
+      "loss": 1.2288,
+      "step": 74
+    },
+    {
+      "epoch": 0.19907100199071,
+      "grad_norm": 1.883985974459675,
+      "learning_rate": 9.953367463372615e-06,
+      "loss": 1.2349,
+      "step": 75
+    },
+    {
+      "epoch": 0.20172528201725282,
+      "grad_norm": 1.8151478617151462,
+      "learning_rate": 9.951923900288888e-06,
+      "loss": 1.1481,
+      "step": 76
+    },
+    {
+      "epoch": 0.20437956204379562,
+      "grad_norm": 1.8542475078063598,
+      "learning_rate": 9.950458440823602e-06,
+      "loss": 1.262,
+      "step": 77
+    },
+    {
+      "epoch": 0.20703384207033843,
+      "grad_norm": 1.77553753540162,
+      "learning_rate": 9.948971091456715e-06,
+      "loss": 1.1834,
+      "step": 78
+    },
+    {
+      "epoch": 0.20968812209688123,
+      "grad_norm": 1.8152497279053155,
+      "learning_rate": 9.947461858764978e-06,
+      "loss": 1.1749,
+      "step": 79
+    },
+    {
+      "epoch": 0.21234240212342403,
+      "grad_norm": 1.5929993680573362,
+      "learning_rate": 9.945930749421903e-06,
+      "loss": 1.2696,
+      "step": 80
+    },
+    {
+      "epoch": 0.2149966821499668,
+      "grad_norm": 2.1883175245684092,
+      "learning_rate": 9.944377770197741e-06,
+      "loss": 1.2375,
+      "step": 81
+    },
+    {
+      "epoch": 0.21765096217650962,
+      "grad_norm": 1.7556050567556294,
+      "learning_rate": 9.942802927959444e-06,
+      "loss": 1.2017,
+      "step": 82
+    },
+    {
+      "epoch": 0.22030524220305242,
+      "grad_norm": 1.7392066404895135,
+      "learning_rate": 9.941206229670634e-06,
+      "loss": 1.1788,
+      "step": 83
+    },
+    {
+      "epoch": 0.22295952222959523,
+      "grad_norm": 2.230563622877349,
+      "learning_rate": 9.939587682391587e-06,
+      "loss": 1.2629,
+      "step": 84
+    },
+    {
+      "epoch": 0.22561380225613803,
+      "grad_norm": 1.7257531276366218,
+      "learning_rate": 9.937947293279178e-06,
+      "loss": 1.1574,
+      "step": 85
+    },
+    {
+      "epoch": 0.22826808228268083,
+      "grad_norm": 3.7946522494948134,
+      "learning_rate": 9.93628506958687e-06,
+      "loss": 1.2539,
+      "step": 86
+    },
+    {
+      "epoch": 0.2309223623092236,
+      "grad_norm": 1.7746755901383093,
+      "learning_rate": 9.934601018664672e-06,
+      "loss": 1.1672,
+      "step": 87
+    },
+    {
+      "epoch": 0.23357664233576642,
+      "grad_norm": 1.6160184542663385,
+      "learning_rate": 9.932895147959106e-06,
+      "loss": 1.2047,
+      "step": 88
+    },
+    {
+      "epoch": 0.23623092236230922,
+      "grad_norm": 1.597818533914632,
+      "learning_rate": 9.931167465013182e-06,
+      "loss": 1.2087,
+      "step": 89
+    },
+    {
+      "epoch": 0.23888520238885202,
+      "grad_norm": 1.8200709191179871,
+      "learning_rate": 9.929417977466356e-06,
+      "loss": 1.2594,
+      "step": 90
+    },
+    {
+      "epoch": 0.24153948241539483,
+      "grad_norm": 1.5869876859286098,
+      "learning_rate": 9.927646693054498e-06,
+      "loss": 1.2923,
+      "step": 91
+    },
+    {
+      "epoch": 0.24419376244193763,
+      "grad_norm": 1.6678230174198274,
+      "learning_rate": 9.925853619609858e-06,
+      "loss": 1.1979,
+      "step": 92
+    },
+    {
+      "epoch": 0.2468480424684804,
+      "grad_norm": 1.7206885835934083,
+      "learning_rate": 9.924038765061042e-06,
+      "loss": 1.2248,
+      "step": 93
+    },
+    {
+      "epoch": 0.24950232249502322,
+      "grad_norm": 1.8965216866987153,
+      "learning_rate": 9.922202137432954e-06,
+      "loss": 1.1793,
+      "step": 94
+    },
+    {
+      "epoch": 0.252156602521566,
+      "grad_norm": 1.7827181222199764,
+      "learning_rate": 9.920343744846786e-06,
+      "loss": 1.2539,
+      "step": 95
+    },
+    {
+      "epoch": 0.25481088254810885,
+      "grad_norm": 1.704509646049322,
+      "learning_rate": 9.918463595519963e-06,
+      "loss": 1.1845,
+      "step": 96
+    },
+    {
+      "epoch": 0.25746516257465163,
+      "grad_norm": 1.8008684596562938,
+      "learning_rate": 9.916561697766114e-06,
+      "loss": 1.1873,
+      "step": 97
+    },
+    {
+      "epoch": 0.2601194426011944,
+      "grad_norm": 1.5268474470110187,
+      "learning_rate": 9.91463805999504e-06,
+      "loss": 1.1634,
+      "step": 98
+    },
+    {
+      "epoch": 0.26277372262773724,
+      "grad_norm": 1.69776670129652,
+      "learning_rate": 9.912692690712667e-06,
+      "loss": 1.2496,
+      "step": 99
+    },
+    {
+      "epoch": 0.26542800265428,
+      "grad_norm": 19.950768938401303,
+      "learning_rate": 9.910725598521014e-06,
+      "loss": 1.2266,
+      "step": 100
+    },
+    {
+      "epoch": 0.26808228268082285,
+      "grad_norm": 1.888189833523382,
+      "learning_rate": 9.908736792118157e-06,
+      "loss": 1.1783,
+      "step": 101
+    },
+    {
+      "epoch": 0.2707365627073656,
+      "grad_norm": 1.6553587447766995,
+      "learning_rate": 9.906726280298185e-06,
+      "loss": 1.1888,
+      "step": 102
+    },
+    {
+      "epoch": 0.2733908427339084,
+      "grad_norm": 1.7645503651144456,
+      "learning_rate": 9.904694071951167e-06,
+      "loss": 1.2331,
+      "step": 103
+    },
+    {
+      "epoch": 0.27604512276045123,
+      "grad_norm": 1.7561319773931536,
+      "learning_rate": 9.902640176063103e-06,
+      "loss": 1.2429,
+      "step": 104
+    },
+    {
+      "epoch": 0.278699402786994,
+      "grad_norm": 1.826781329048666,
+      "learning_rate": 9.900564601715898e-06,
+      "loss": 1.2053,
+      "step": 105
+    },
+    {
+      "epoch": 0.28135368281353684,
+      "grad_norm": 2.0178364653670777,
+      "learning_rate": 9.89846735808731e-06,
+      "loss": 1.1855,
+      "step": 106
+    },
+    {
+      "epoch": 0.2840079628400796,
+      "grad_norm": 2.1853732110604027,
+      "learning_rate": 9.896348454450918e-06,
+      "loss": 1.1514,
+      "step": 107
+    },
+    {
+      "epoch": 0.28666224286662245,
+      "grad_norm": 1.863102490412834,
+      "learning_rate": 9.894207900176074e-06,
+      "loss": 1.1582,
+      "step": 108
+    },
+    {
+      "epoch": 0.28931652289316523,
+      "grad_norm": 2.1558166021806504,
+      "learning_rate": 9.892045704727864e-06,
+      "loss": 1.2692,
+      "step": 109
+    },
+    {
+      "epoch": 0.291970802919708,
+      "grad_norm": 1.623855596114215,
+      "learning_rate": 9.889861877667071e-06,
+      "loss": 1.1406,
+      "step": 110
+    },
+    {
+      "epoch": 0.29462508294625084,
+      "grad_norm": 1.925359975573577,
+      "learning_rate": 9.887656428650123e-06,
+      "loss": 1.144,
+      "step": 111
+    },
+    {
+      "epoch": 0.2972793629727936,
+      "grad_norm": 2.0780064875742634,
+      "learning_rate": 9.885429367429062e-06,
+      "loss": 1.2095,
+      "step": 112
+    },
+    {
+      "epoch": 0.29993364299933645,
+      "grad_norm": 1.757836459981376,
+      "learning_rate": 9.883180703851488e-06,
+      "loss": 1.2129,
+      "step": 113
+    },
+    {
+      "epoch": 0.3025879230258792,
+      "grad_norm": 1.6497771335719753,
+      "learning_rate": 9.880910447860527e-06,
+      "loss": 1.1528,
+      "step": 114
+    },
+    {
+      "epoch": 0.30524220305242206,
+      "grad_norm": 1.9314161378924497,
+      "learning_rate": 9.878618609494781e-06,
+      "loss": 1.2038,
+      "step": 115
+    },
+    {
+      "epoch": 0.30789648307896483,
+      "grad_norm": 1.5945997988558909,
+      "learning_rate": 9.876305198888284e-06,
+      "loss": 1.1349,
+      "step": 116
+    },
+    {
+      "epoch": 0.3105507631055076,
+      "grad_norm": 1.7095400428823162,
+      "learning_rate": 9.873970226270458e-06,
+      "loss": 1.1543,
+      "step": 117
+    },
+    {
+      "epoch": 0.31320504313205044,
+      "grad_norm": 1.6150384960254696,
+      "learning_rate": 9.871613701966067e-06,
+      "loss": 1.1527,
+      "step": 118
+    },
+    {
+      "epoch": 0.3158593231585932,
+      "grad_norm": 3.288441610824325,
+      "learning_rate": 9.869235636395177e-06,
+      "loss": 1.2411,
+      "step": 119
+    },
+    {
+      "epoch": 0.31851360318513605,
+      "grad_norm": 1.6258023683537948,
+      "learning_rate": 9.866836040073099e-06,
+      "loss": 1.2002,
+      "step": 120
+    },
+    {
+      "epoch": 0.32116788321167883,
+      "grad_norm": 1.6467592688369062,
+      "learning_rate": 9.86441492361035e-06,
+      "loss": 1.2134,
+      "step": 121
+    },
+    {
+      "epoch": 0.3238221632382216,
+      "grad_norm": 1.5988307616959179,
+      "learning_rate": 9.861972297712606e-06,
+      "loss": 1.2259,
+      "step": 122
+    },
+    {
+      "epoch": 0.32647644326476444,
+      "grad_norm": 1.9915164437167947,
+      "learning_rate": 9.859508173180653e-06,
+      "loss": 1.2369,
+      "step": 123
+    },
+    {
+      "epoch": 0.3291307232913072,
+      "grad_norm": 1.751874113048822,
+      "learning_rate": 9.857022560910338e-06,
+      "loss": 1.1954,
+      "step": 124
+    },
+    {
+      "epoch": 0.33178500331785005,
+      "grad_norm": 1.589249974809787,
+      "learning_rate": 9.854515471892527e-06,
+      "loss": 1.1434,
+      "step": 125
+    },
+    {
+      "epoch": 0.3344392833443928,
+      "grad_norm": 1.6571603039493696,
+      "learning_rate": 9.851986917213044e-06,
+      "loss": 1.1276,
+      "step": 126
+    },
+    {
+      "epoch": 0.33709356337093566,
+      "grad_norm": 1.821753338127428,
+      "learning_rate": 9.849436908052636e-06,
+      "loss": 1.1889,
+      "step": 127
+    },
+    {
+      "epoch": 0.33974784339747843,
+      "grad_norm": 1.55634782143693,
+      "learning_rate": 9.846865455686915e-06,
+      "loss": 1.1833,
+      "step": 128
+    },
+    {
+      "epoch": 0.3424021234240212,
+      "grad_norm": 1.4585503339043484,
+      "learning_rate": 9.844272571486313e-06,
+      "loss": 1.1979,
+      "step": 129
+    },
+    {
+      "epoch": 0.34505640345056404,
+      "grad_norm": 1.6155812940652678,
+      "learning_rate": 9.84165826691602e-06,
+      "loss": 1.179,
+      "step": 130
+    },
+    {
+      "epoch": 0.3477106834771068,
+      "grad_norm": 1.609059975302855,
+      "learning_rate": 9.839022553535957e-06,
+      "loss": 1.2091,
+      "step": 131
+    },
+    {
+      "epoch": 0.35036496350364965,
+      "grad_norm": 1.6996931072095949,
+      "learning_rate": 9.836365443000697e-06,
+      "loss": 1.1223,
+      "step": 132
+    },
+    {
+      "epoch": 0.35301924353019243,
+      "grad_norm": 1.616355220759201,
+      "learning_rate": 9.833686947059436e-06,
+      "loss": 1.0918,
+      "step": 133
+    },
+    {
+      "epoch": 0.35567352355673526,
+      "grad_norm": 1.6096571582268207,
+      "learning_rate": 9.830987077555925e-06,
+      "loss": 1.1654,
+      "step": 134
+    },
+    {
+      "epoch": 0.35832780358327804,
+      "grad_norm": 1.565339921018465,
+      "learning_rate": 9.828265846428428e-06,
+      "loss": 1.1634,
+      "step": 135
+    },
+    {
+      "epoch": 0.3609820836098208,
+      "grad_norm": 1.546016830156871,
+      "learning_rate": 9.825523265709667e-06,
+      "loss": 1.1751,
+      "step": 136
+    },
+    {
+      "epoch": 0.36363636363636365,
+      "grad_norm": 2.0425258490960836,
+      "learning_rate": 9.822759347526766e-06,
+      "loss": 1.1841,
+      "step": 137
+    },
+    {
+      "epoch": 0.3662906436629064,
+      "grad_norm": 1.7147276492095496,
+      "learning_rate": 9.819974104101198e-06,
+      "loss": 1.2335,
+      "step": 138
+    },
+    {
+      "epoch": 0.36894492368944926,
+      "grad_norm": 1.5869898030324339,
+      "learning_rate": 9.817167547748729e-06,
+      "loss": 1.2584,
+      "step": 139
+    },
+    {
+      "epoch": 0.37159920371599203,
+      "grad_norm": 1.6478179096824475,
+      "learning_rate": 9.814339690879376e-06,
+      "loss": 1.1961,
+      "step": 140
+    },
+    {
+      "epoch": 0.37425348374253486,
+      "grad_norm": 1.5596319041645448,
+      "learning_rate": 9.811490545997331e-06,
+      "loss": 1.2046,
+      "step": 141
+    },
+    {
+      "epoch": 0.37690776376907764,
+      "grad_norm": 1.81548347845434,
+      "learning_rate": 9.808620125700925e-06,
+      "loss": 1.2137,
+      "step": 142
+    },
+    {
+      "epoch": 0.3795620437956204,
+      "grad_norm": 1.5700607431043994,
+      "learning_rate": 9.80572844268256e-06,
+      "loss": 1.2,
+      "step": 143
+    },
+    {
+      "epoch": 0.38221632382216325,
+      "grad_norm": 1.6383722320139935,
+      "learning_rate": 9.802815509728662e-06,
+      "loss": 1.1747,
+      "step": 144
+    },
+    {
+      "epoch": 0.384870603848706,
+      "grad_norm": 1.8125605110455933,
+      "learning_rate": 9.799881339719615e-06,
+      "loss": 1.1867,
+      "step": 145
+    },
+    {
+      "epoch": 0.38752488387524886,
+      "grad_norm": 1.7582804382886328,
+      "learning_rate": 9.796925945629711e-06,
+      "loss": 1.3162,
+      "step": 146
+    },
+    {
+      "epoch": 0.39017916390179164,
+      "grad_norm": 2.0734414221603665,
+      "learning_rate": 9.793949340527091e-06,
+      "loss": 1.234,
+      "step": 147
+    },
+    {
+      "epoch": 0.3928334439283344,
+      "grad_norm": 1.5379766795331946,
+      "learning_rate": 9.790951537573686e-06,
+      "loss": 1.1185,
+      "step": 148
+    },
+    {
+      "epoch": 0.39548772395487725,
+      "grad_norm": 1.6227118631483388,
+      "learning_rate": 9.787932550025158e-06,
+      "loss": 1.1523,
+      "step": 149
+    },
+    {
+      "epoch": 0.39814200398142,
+      "grad_norm": 1.5530246573576652,
+      "learning_rate": 9.784892391230847e-06,
+      "loss": 1.1405,
+      "step": 150
+    },
+    {
+      "epoch": 0.40079628400796286,
+      "grad_norm": 1.6380120481890832,
+      "learning_rate": 9.781831074633703e-06,
+      "loss": 1.2153,
+      "step": 151
+    },
+    {
+      "epoch": 0.40345056403450563,
+      "grad_norm": 1.9786402997029178,
+      "learning_rate": 9.778748613770234e-06,
+      "loss": 1.2213,
+      "step": 152
+    },
+    {
+      "epoch": 0.40610484406104846,
+      "grad_norm": 1.5864201917409944,
+      "learning_rate": 9.775645022270448e-06,
+      "loss": 1.1674,
+      "step": 153
+    },
+    {
+      "epoch": 0.40875912408759124,
+      "grad_norm": 1.6501859286504295,
+      "learning_rate": 9.772520313857777e-06,
+      "loss": 1.1805,
+      "step": 154
+    },
+    {
+      "epoch": 0.411413404114134,
+      "grad_norm": 1.5132303328319994,
+      "learning_rate": 9.769374502349038e-06,
+      "loss": 1.15,
+      "step": 155
+    },
+    {
+      "epoch": 0.41406768414067685,
+      "grad_norm": 1.6907615300646948,
+      "learning_rate": 9.766207601654356e-06,
+      "loss": 1.1848,
+      "step": 156
+    },
+    {
+      "epoch": 0.4167219641672196,
+      "grad_norm": 1.7001401085077357,
+      "learning_rate": 9.763019625777111e-06,
+      "loss": 1.2335,
+      "step": 157
+    },
+    {
+      "epoch": 0.41937624419376246,
+      "grad_norm": 1.7600942954439958,
+      "learning_rate": 9.759810588813872e-06,
+      "loss": 1.1743,
+      "step": 158
+    },
+    {
+      "epoch": 0.42203052422030524,
+      "grad_norm": 1.6515697771161784,
+      "learning_rate": 9.756580504954334e-06,
+      "loss": 1.2276,
+      "step": 159
+    },
+    {
+      "epoch": 0.42468480424684807,
+      "grad_norm": 1.6967375558942543,
+      "learning_rate": 9.753329388481261e-06,
+      "loss": 1.2082,
+      "step": 160
+    },
+    {
+      "epoch": 0.42733908427339085,
+      "grad_norm": 1.7152427713922846,
+      "learning_rate": 9.750057253770413e-06,
+      "loss": 1.1458,
+      "step": 161
+    },
+    {
+      "epoch": 0.4299933642999336,
+      "grad_norm": 1.625060781845651,
+      "learning_rate": 9.746764115290496e-06,
+      "loss": 1.2033,
+      "step": 162
+    },
+    {
+      "epoch": 0.43264764432647645,
+      "grad_norm": 1.7069397928194143,
+      "learning_rate": 9.743449987603082e-06,
+      "loss": 1.2342,
+      "step": 163
+    },
+    {
+      "epoch": 0.43530192435301923,
+      "grad_norm": 1.708629436384557,
+      "learning_rate": 9.740114885362562e-06,
+      "loss": 1.2442,
+      "step": 164
+    },
+    {
+      "epoch": 0.43795620437956206,
+      "grad_norm": 1.5857203714681123,
+      "learning_rate": 9.736758823316062e-06,
+      "loss": 1.2097,
+      "step": 165
+    },
+    {
+      "epoch": 0.44061048440610484,
+      "grad_norm": 1.657268677184339,
+      "learning_rate": 9.733381816303395e-06,
+      "loss": 1.1215,
+      "step": 166
+    },
+    {
+      "epoch": 0.4432647644326476,
+      "grad_norm": 1.4640436666626744,
+      "learning_rate": 9.729983879256988e-06,
+      "loss": 1.1646,
+      "step": 167
+    },
+    {
+      "epoch": 0.44591904445919045,
+      "grad_norm": 1.6268091054804499,
+      "learning_rate": 9.726565027201813e-06,
+      "loss": 1.2264,
+      "step": 168
+    },
+    {
+      "epoch": 0.4485733244857332,
+      "grad_norm": 1.5858930123997803,
+      "learning_rate": 9.723125275255325e-06,
+      "loss": 1.1661,
+      "step": 169
+    },
+    {
+      "epoch": 0.45122760451227606,
+      "grad_norm": 1.5759031230494174,
+      "learning_rate": 9.719664638627395e-06,
+      "loss": 1.1558,
+      "step": 170
+    },
+    {
+      "epoch": 0.45388188453881884,
+      "grad_norm": 1.7486351365651316,
+      "learning_rate": 9.716183132620242e-06,
+      "loss": 1.19,
+      "step": 171
+    },
+    {
+      "epoch": 0.45653616456536167,
+      "grad_norm": 1.7251231178841304,
+      "learning_rate": 9.712680772628365e-06,
+      "loss": 1.2261,
+      "step": 172
+    },
+    {
+      "epoch": 0.45919044459190445,
+      "grad_norm": 1.6118734678264717,
+      "learning_rate": 9.70915757413847e-06,
+      "loss": 1.2014,
+      "step": 173
+    },
+    {
+      "epoch": 0.4618447246184472,
+      "grad_norm": 1.5762577213086215,
+      "learning_rate": 9.705613552729416e-06,
+      "loss": 1.1487,
+      "step": 174
+    },
+    {
+      "epoch": 0.46449900464499005,
+      "grad_norm": 1.5672859542358526,
+      "learning_rate": 9.702048724072128e-06,
+      "loss": 1.1892,
+      "step": 175
+    },
+    {
+      "epoch": 0.46715328467153283,
+      "grad_norm": 1.616073022266597,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 1.1718,
+      "step": 176
+    },
+    {
+      "epoch": 0.46980756469807566,
+      "grad_norm": 1.605222482810264,
+      "learning_rate": 9.694856708156526e-06,
+      "loss": 1.1022,
+      "step": 177
+    },
+    {
+      "epoch": 0.47246184472461844,
+      "grad_norm": 1.483617625625729,
+      "learning_rate": 9.691229552699817e-06,
+      "loss": 1.1204,
+      "step": 178
+    },
+    {
+      "epoch": 0.4751161247511613,
+      "grad_norm": 1.6018473502205803,
+      "learning_rate": 9.68758165359794e-06,
+      "loss": 1.1816,
+      "step": 179
+    },
+    {
+      "epoch": 0.47777040477770405,
+      "grad_norm": 1.5779836150848479,
+      "learning_rate": 9.683913026981155e-06,
+      "loss": 1.1871,
+      "step": 180
+    },
+    {
+      "epoch": 0.4804246848042468,
+      "grad_norm": 1.6463102663610685,
+      "learning_rate": 9.680223689071364e-06,
+      "loss": 1.1139,
+      "step": 181
+    },
+    {
+      "epoch": 0.48307896483078966,
+      "grad_norm": 1.7091000919337074,
+      "learning_rate": 9.676513656182059e-06,
+      "loss": 1.1695,
+      "step": 182
+    },
+    {
+      "epoch": 0.48573324485733244,
+      "grad_norm": 1.633509337933534,
+      "learning_rate": 9.672782944718234e-06,
+      "loss": 1.1811,
+      "step": 183
+    },
+    {
+      "epoch": 0.48838752488387527,
+      "grad_norm": 1.5767561431519088,
+      "learning_rate": 9.669031571176322e-06,
+      "loss": 1.2062,
+      "step": 184
+    },
+    {
+      "epoch": 0.49104180491041804,
+      "grad_norm": 1.6306907003550404,
+      "learning_rate": 9.665259552144122e-06,
+      "loss": 1.1829,
+      "step": 185
+    },
+    {
+      "epoch": 0.4936960849369608,
+      "grad_norm": 1.517988528061533,
+      "learning_rate": 9.66146690430072e-06,
+      "loss": 1.2014,
+      "step": 186
+    },
+    {
+      "epoch": 0.49635036496350365,
+      "grad_norm": 1.598871387440831,
+      "learning_rate": 9.657653644416417e-06,
+      "loss": 1.1496,
+      "step": 187
+    },
+    {
+      "epoch": 0.49900464499004643,
+      "grad_norm": 2.400377785726973,
+      "learning_rate": 9.65381978935266e-06,
+      "loss": 1.1905,
+      "step": 188
+    },
+    {
+      "epoch": 0.5016589250165893,
+      "grad_norm": 1.5306038174802905,
+      "learning_rate": 9.649965356061961e-06,
+      "loss": 1.1225,
+      "step": 189
+    },
+    {
+      "epoch": 0.504313205043132,
+      "grad_norm": 1.7432637039460837,
+      "learning_rate": 9.646090361587828e-06,
+      "loss": 1.2338,
+      "step": 190
+    },
+    {
+      "epoch": 0.5069674850696748,
+      "grad_norm": 1.549174772320108,
+      "learning_rate": 9.642194823064679e-06,
+      "loss": 1.1395,
+      "step": 191
+    },
+    {
+      "epoch": 0.5096217650962177,
+      "grad_norm": 1.4556718082039433,
+      "learning_rate": 9.63827875771778e-06,
+      "loss": 1.1054,
+      "step": 192
+    },
+    {
+      "epoch": 0.5122760451227605,
+      "grad_norm": 1.546232476076245,
+      "learning_rate": 9.634342182863163e-06,
+      "loss": 1.1821,
+      "step": 193
+    },
+    {
+      "epoch": 0.5149303251493033,
+      "grad_norm": 1.6428065540768686,
+      "learning_rate": 9.630385115907545e-06,
+      "loss": 1.2078,
+      "step": 194
+    },
+    {
+      "epoch": 0.517584605175846,
+      "grad_norm": 1.5932949193165389,
+      "learning_rate": 9.626407574348258e-06,
+      "loss": 1.1646,
+      "step": 195
+    },
+    {
+      "epoch": 0.5202388852023888,
+      "grad_norm": 1.5803201555935116,
+      "learning_rate": 9.622409575773162e-06,
+      "loss": 1.166,
+      "step": 196
+    },
+    {
+      "epoch": 0.5228931652289317,
+      "grad_norm": 1.5292820314306055,
+      "learning_rate": 9.618391137860583e-06,
+      "loss": 1.2152,
+      "step": 197
+    },
+    {
+      "epoch": 0.5255474452554745,
+      "grad_norm": 1.3727930028186761,
+      "learning_rate": 9.614352278379217e-06,
+      "loss": 1.1402,
+      "step": 198
+    },
+    {
+      "epoch": 0.5282017252820173,
+      "grad_norm": 1.6819090165661312,
+      "learning_rate": 9.610293015188067e-06,
+      "loss": 1.1665,
+      "step": 199
+    },
+    {
+      "epoch": 0.53085600530856,
+      "grad_norm": 1.5006037012098021,
+      "learning_rate": 9.606213366236354e-06,
+      "loss": 1.1877,
+      "step": 200
+    },
+    {
+      "epoch": 0.5335102853351028,
+      "grad_norm": 1.6016624668408799,
+      "learning_rate": 9.60211334956344e-06,
+      "loss": 1.1498,
+      "step": 201
+    },
+    {
+      "epoch": 0.5361645653616457,
+      "grad_norm": 1.7368140744461305,
+      "learning_rate": 9.597992983298748e-06,
+      "loss": 1.1922,
+      "step": 202
+    },
+    {
+      "epoch": 0.5388188453881885,
+      "grad_norm": 1.6176251602621352,
+      "learning_rate": 9.593852285661684e-06,
+      "loss": 1.1459,
+      "step": 203
+    },
+    {
+      "epoch": 0.5414731254147312,
+      "grad_norm": 1.3750495471617235,
+      "learning_rate": 9.589691274961556e-06,
+      "loss": 1.0835,
+      "step": 204
+    },
+    {
+      "epoch": 0.544127405441274,
+      "grad_norm": 1.6906485599869903,
+      "learning_rate": 9.585509969597491e-06,
+      "loss": 1.22,
+      "step": 205
+    },
+    {
+      "epoch": 0.5467816854678168,
+      "grad_norm": 1.5439326128894457,
+      "learning_rate": 9.581308388058354e-06,
+      "loss": 1.1364,
+      "step": 206
+    },
+    {
+      "epoch": 0.5494359654943597,
+      "grad_norm": 1.5251041120197495,
+      "learning_rate": 9.577086548922671e-06,
+      "loss": 1.2201,
+      "step": 207
+    },
+    {
+      "epoch": 0.5520902455209025,
+      "grad_norm": 1.511712369802414,
+      "learning_rate": 9.572844470858537e-06,
+      "loss": 1.1091,
+      "step": 208
+    },
+    {
+      "epoch": 0.5547445255474452,
+      "grad_norm": 1.8573483808679467,
+      "learning_rate": 9.568582172623544e-06,
+      "loss": 1.2284,
+      "step": 209
+    },
+    {
+      "epoch": 0.557398805573988,
+      "grad_norm": 1.4309251806187955,
+      "learning_rate": 9.56429967306469e-06,
+      "loss": 1.1646,
+      "step": 210
+    },
+    {
+      "epoch": 0.5600530856005309,
+      "grad_norm": 1.6268260856080405,
+      "learning_rate": 9.559996991118304e-06,
+      "loss": 1.1812,
+      "step": 211
+    },
+    {
+      "epoch": 0.5627073656270737,
+      "grad_norm": 1.6752285964398912,
+      "learning_rate": 9.55567414580995e-06,
+      "loss": 1.19,
+      "step": 212
+    },
+    {
+      "epoch": 0.5653616456536165,
+      "grad_norm": 1.6202125494829664,
+      "learning_rate": 9.551331156254358e-06,
+      "loss": 1.2001,
+      "step": 213
+    },
+    {
+      "epoch": 0.5680159256801592,
+      "grad_norm": 1.4441208249265054,
+      "learning_rate": 9.546968041655326e-06,
+      "loss": 1.2011,
+      "step": 214
+    },
+    {
+      "epoch": 0.570670205706702,
+      "grad_norm": 1.4681168393734876,
+      "learning_rate": 9.542584821305643e-06,
+      "loss": 1.118,
+      "step": 215
+    },
+    {
+      "epoch": 0.5733244857332449,
+      "grad_norm": 1.67215223118757,
+      "learning_rate": 9.538181514587004e-06,
+      "loss": 1.1441,
+      "step": 216
+    },
+    {
+      "epoch": 0.5759787657597877,
+      "grad_norm": 1.840004210878956,
+      "learning_rate": 9.533758140969913e-06,
+      "loss": 1.1689,
+      "step": 217
+    },
+    {
+      "epoch": 0.5786330457863305,
+      "grad_norm": 2.0817799067244387,
+      "learning_rate": 9.529314720013618e-06,
+      "loss": 1.1879,
+      "step": 218
+    },
+    {
+      "epoch": 0.5812873258128732,
+      "grad_norm": 1.6384013753881452,
+      "learning_rate": 9.524851271366002e-06,
+      "loss": 1.1157,
+      "step": 219
+    },
+    {
+      "epoch": 0.583941605839416,
+      "grad_norm": 1.6847540459176993,
+      "learning_rate": 9.520367814763514e-06,
+      "loss": 1.1583,
+      "step": 220
+    },
+    {
+      "epoch": 0.5865958858659589,
+      "grad_norm": 1.553367758905212,
+      "learning_rate": 9.515864370031066e-06,
+      "loss": 1.0916,
+      "step": 221
+    },
+    {
+      "epoch": 0.5892501658925017,
+      "grad_norm": 1.6595661898312408,
+      "learning_rate": 9.511340957081957e-06,
+      "loss": 1.1912,
+      "step": 222
+    },
+    {
+      "epoch": 0.5919044459190445,
+      "grad_norm": 1.6816767854984012,
+      "learning_rate": 9.506797595917787e-06,
+      "loss": 1.1948,
+      "step": 223
+    },
+    {
+      "epoch": 0.5945587259455872,
+      "grad_norm": 1.4766094174812612,
+      "learning_rate": 9.502234306628354e-06,
+      "loss": 1.1607,
+      "step": 224
+    },
+    {
+      "epoch": 0.59721300597213,
+      "grad_norm": 1.5815513019760774,
+      "learning_rate": 9.49765110939158e-06,
+      "loss": 1.1248,
+      "step": 225
+    },
+    {
+      "epoch": 0.5998672859986729,
+      "grad_norm": 1.6485658910927394,
+      "learning_rate": 9.493048024473413e-06,
+      "loss": 1.2191,
+      "step": 226
+    },
+    {
+      "epoch": 0.6025215660252157,
+      "grad_norm": 1.424065848656427,
+      "learning_rate": 9.488425072227738e-06,
+      "loss": 1.2521,
+      "step": 227
+    },
+    {
+      "epoch": 0.6051758460517584,
+      "grad_norm": 1.4486333802405926,
+      "learning_rate": 9.483782273096295e-06,
+      "loss": 1.1734,
+      "step": 228
+    },
+    {
+      "epoch": 0.6078301260783012,
+      "grad_norm": 1.6817918601770532,
+      "learning_rate": 9.47911964760858e-06,
+      "loss": 1.1695,
+      "step": 229
+    },
+    {
+      "epoch": 0.6104844061048441,
+      "grad_norm": 1.6160290558732326,
+      "learning_rate": 9.474437216381756e-06,
+      "loss": 1.154,
+      "step": 230
+    },
+    {
+      "epoch": 0.6131386861313869,
+      "grad_norm": 1.4261795572898603,
+      "learning_rate": 9.469735000120564e-06,
+      "loss": 1.1544,
+      "step": 231
+    },
+    {
+      "epoch": 0.6157929661579297,
+      "grad_norm": 1.458151411666846,
+      "learning_rate": 9.46501301961723e-06,
+      "loss": 1.2065,
+      "step": 232
+    },
+    {
+      "epoch": 0.6184472461844724,
+      "grad_norm": 1.5627499060274408,
+      "learning_rate": 9.460271295751373e-06,
+      "loss": 1.1579,
+      "step": 233
+    },
+    {
+      "epoch": 0.6211015262110152,
+      "grad_norm": 1.86944032236805,
+      "learning_rate": 9.455509849489915e-06,
+      "loss": 1.1519,
+      "step": 234
+    },
+    {
+      "epoch": 0.6237558062375581,
+      "grad_norm": 1.979766174904363,
+      "learning_rate": 9.450728701886985e-06,
+      "loss": 1.2358,
+      "step": 235
+    },
+    {
+      "epoch": 0.6264100862641009,
+      "grad_norm": 1.5229843416844162,
+      "learning_rate": 9.445927874083825e-06,
+      "loss": 1.1207,
+      "step": 236
+    },
+    {
+      "epoch": 0.6290643662906437,
+      "grad_norm": 1.5916340950774943,
+      "learning_rate": 9.441107387308701e-06,
+      "loss": 1.2486,
+      "step": 237
+    },
+    {
+      "epoch": 0.6317186463171864,
+      "grad_norm": 1.4982052500691954,
+      "learning_rate": 9.436267262876808e-06,
+      "loss": 1.1445,
+      "step": 238
+    },
+    {
+      "epoch": 0.6343729263437292,
+      "grad_norm": 1.868028818978397,
+      "learning_rate": 9.431407522190176e-06,
+      "loss": 1.2215,
+      "step": 239
+    },
+    {
+      "epoch": 0.6370272063702721,
+      "grad_norm": 1.5000893386206633,
+      "learning_rate": 9.426528186737566e-06,
+      "loss": 1.1748,
+      "step": 240
+    },
+    {
+      "epoch": 0.6396814863968149,
+      "grad_norm": 1.6105517075622542,
+      "learning_rate": 9.421629278094394e-06,
+      "loss": 1.1444,
+      "step": 241
+    },
+    {
+      "epoch": 0.6423357664233577,
+      "grad_norm": 1.6245044582496362,
+      "learning_rate": 9.416710817922615e-06,
+      "loss": 1.2016,
+      "step": 242
+    },
+    {
+      "epoch": 0.6449900464499004,
+      "grad_norm": 1.582791773770731,
+      "learning_rate": 9.411772827970642e-06,
+      "loss": 1.1595,
+      "step": 243
+    },
+    {
+      "epoch": 0.6476443264764432,
+      "grad_norm": 1.5289298221123744,
+      "learning_rate": 9.406815330073244e-06,
+      "loss": 1.196,
+      "step": 244
+    },
+    {
+      "epoch": 0.6502986065029861,
+      "grad_norm": 1.494805179412693,
+      "learning_rate": 9.40183834615145e-06,
+      "loss": 1.119,
+      "step": 245
+    },
+    {
+      "epoch": 0.6529528865295289,
+      "grad_norm": 1.6857955705395817,
+      "learning_rate": 9.396841898212452e-06,
+      "loss": 1.1222,
+      "step": 246
+    },
+    {
+      "epoch": 0.6556071665560717,
+      "grad_norm": 1.465569644664737,
+      "learning_rate": 9.391826008349507e-06,
+      "loss": 1.1196,
+      "step": 247
+    },
+    {
+      "epoch": 0.6582614465826144,
+      "grad_norm": 1.6038700287536702,
+      "learning_rate": 9.38679069874184e-06,
+      "loss": 1.1596,
+      "step": 248
+    },
+    {
+      "epoch": 0.6609157266091573,
+      "grad_norm": 1.754259412074635,
+      "learning_rate": 9.381735991654547e-06,
+      "loss": 1.185,
+      "step": 249
+    },
+    {
+      "epoch": 0.6635700066357001,
+      "grad_norm": 1.5054976516017162,
+      "learning_rate": 9.376661909438496e-06,
+      "loss": 1.14,
+      "step": 250
+    },
+    {
+      "epoch": 0.6662242866622429,
+      "grad_norm": 1.6591136646331954,
+      "learning_rate": 9.371568474530228e-06,
+      "loss": 1.1453,
+      "step": 251
+    },
+    {
+      "epoch": 0.6688785666887856,
+      "grad_norm": 1.602614315373211,
+      "learning_rate": 9.366455709451857e-06,
+      "loss": 1.115,
+      "step": 252
+    },
+    {
+      "epoch": 0.6715328467153284,
+      "grad_norm": 1.3802344389470933,
+      "learning_rate": 9.36132363681097e-06,
+      "loss": 1.0926,
+      "step": 253
+    },
+    {
+      "epoch": 0.6741871267418713,
+      "grad_norm": 1.5028041314507699,
+      "learning_rate": 9.356172279300528e-06,
+      "loss": 1.1388,
+      "step": 254
+    },
+    {
+      "epoch": 0.6768414067684141,
+      "grad_norm": 1.4603385973006835,
+      "learning_rate": 9.35100165969877e-06,
+      "loss": 1.1261,
+      "step": 255
+    },
+    {
+      "epoch": 0.6794956867949569,
+      "grad_norm": 2.222737357031752,
+      "learning_rate": 9.3458118008691e-06,
+      "loss": 1.1181,
+      "step": 256
+    },
+    {
+      "epoch": 0.6821499668214996,
+      "grad_norm": 1.5628150576174966,
+      "learning_rate": 9.340602725760003e-06,
+      "loss": 1.1269,
+      "step": 257
+    },
+    {
+      "epoch": 0.6848042468480424,
+      "grad_norm": 1.7660936315398623,
+      "learning_rate": 9.335374457404928e-06,
+      "loss": 1.1567,
+      "step": 258
+    },
+    {
+      "epoch": 0.6874585268745853,
+      "grad_norm": 1.5095573241471834,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 1.1407,
+      "step": 259
+    },
+    {
+      "epoch": 0.6901128069011281,
+      "grad_norm": 1.4506359372228914,
+      "learning_rate": 9.324860433514888e-06,
+      "loss": 1.1668,
+      "step": 260
+    },
+    {
+      "epoch": 0.6927670869276709,
+      "grad_norm": 1.536882345986633,
+      "learning_rate": 9.319574724470756e-06,
+      "loss": 1.1581,
+      "step": 261
+    },
+    {
+      "epoch": 0.6954213669542136,
+      "grad_norm": 1.4356269422691534,
+      "learning_rate": 9.314269915162115e-06,
+      "loss": 1.1075,
+      "step": 262
+    },
+    {
+      "epoch": 0.6980756469807564,
+      "grad_norm": 1.373904876593965,
+      "learning_rate": 9.308946029045726e-06,
+      "loss": 1.1121,
+      "step": 263
+    },
+    {
+      "epoch": 0.7007299270072993,
+      "grad_norm": 1.5328812905843867,
+      "learning_rate": 9.303603089662717e-06,
+      "loss": 1.0921,
+      "step": 264
+    },
+    {
+      "epoch": 0.7033842070338421,
+      "grad_norm": 1.5072781837506157,
+      "learning_rate": 9.298241120638451e-06,
+      "loss": 1.1198,
+      "step": 265
+    },
+    {
+      "epoch": 0.7060384870603849,
+      "grad_norm": 1.5995295442728128,
+      "learning_rate": 9.292860145682451e-06,
+      "loss": 1.1472,
+      "step": 266
+    },
+    {
+      "epoch": 0.7086927670869276,
+      "grad_norm": 1.586589487215959,
+      "learning_rate": 9.287460188588272e-06,
+      "loss": 1.2081,
+      "step": 267
+    },
+    {
+      "epoch": 0.7113470471134705,
+      "grad_norm": 1.6738675413951511,
+      "learning_rate": 9.282041273233402e-06,
+      "loss": 1.1676,
+      "step": 268
+    },
+    {
+      "epoch": 0.7140013271400133,
+      "grad_norm": 1.5986869946296454,
+      "learning_rate": 9.276603423579164e-06,
+      "loss": 1.213,
+      "step": 269
+    },
+    {
+      "epoch": 0.7166556071665561,
+      "grad_norm": 1.5027119454217344,
+      "learning_rate": 9.271146663670605e-06,
+      "loss": 1.1622,
+      "step": 270
+    },
+    {
+      "epoch": 0.7193098871930989,
+      "grad_norm": 1.4752249291840163,
+      "learning_rate": 9.265671017636384e-06,
+      "loss": 1.0725,
+      "step": 271
+    },
+    {
+      "epoch": 0.7219641672196416,
+      "grad_norm": 1.6425492982199013,
+      "learning_rate": 9.260176509688673e-06,
+      "loss": 1.2088,
+      "step": 272
+    },
+    {
+      "epoch": 0.7246184472461845,
+      "grad_norm": 1.671119694405482,
+      "learning_rate": 9.254663164123052e-06,
+      "loss": 1.1584,
+      "step": 273
+    },
+    {
+      "epoch": 0.7272727272727273,
+      "grad_norm": 3.2839083971639016,
+      "learning_rate": 9.249131005318388e-06,
+      "loss": 1.0801,
+      "step": 274
+    },
+    {
+      "epoch": 0.7299270072992701,
+      "grad_norm": 1.590670276122513,
+      "learning_rate": 9.243580057736743e-06,
+      "loss": 1.1157,
+      "step": 275
+    },
+    {
+      "epoch": 0.7325812873258128,
+      "grad_norm": 1.4517652800533363,
+      "learning_rate": 9.238010345923257e-06,
+      "loss": 1.1446,
+      "step": 276
+    },
+    {
+      "epoch": 0.7352355673523556,
+      "grad_norm": 1.9696673043614277,
+      "learning_rate": 9.232421894506043e-06,
+      "loss": 1.1857,
+      "step": 277
+    },
+    {
+      "epoch": 0.7378898473788985,
+      "grad_norm": 1.4778960277561557,
+      "learning_rate": 9.226814728196072e-06,
+      "loss": 1.1397,
+      "step": 278
+    },
+    {
+      "epoch": 0.7405441274054413,
+      "grad_norm": 1.6498804570471097,
+      "learning_rate": 9.221188871787076e-06,
+      "loss": 1.1625,
+      "step": 279
+    },
+    {
+      "epoch": 0.7431984074319841,
+      "grad_norm": 1.5796993896804141,
+      "learning_rate": 9.215544350155423e-06,
+      "loss": 1.1459,
+      "step": 280
+    },
+    {
+      "epoch": 0.7458526874585268,
+      "grad_norm": 1.5226644568838132,
+      "learning_rate": 9.209881188260021e-06,
+      "loss": 1.1894,
+      "step": 281
+    },
+    {
+      "epoch": 0.7485069674850697,
+      "grad_norm": 1.6645552718061039,
+      "learning_rate": 9.204199411142196e-06,
+      "loss": 1.0811,
+      "step": 282
+    },
+    {
+      "epoch": 0.7511612475116125,
+      "grad_norm": 1.6581847965929961,
+      "learning_rate": 9.198499043925591e-06,
+      "loss": 1.1706,
+      "step": 283
+    },
+    {
+      "epoch": 0.7538155275381553,
+      "grad_norm": 1.5270964606037345,
+      "learning_rate": 9.192780111816048e-06,
+      "loss": 1.1009,
+      "step": 284
+    },
+    {
+      "epoch": 0.7564698075646981,
+      "grad_norm": 1.6698962782227256,
+      "learning_rate": 9.1870426401015e-06,
+      "loss": 1.1708,
+      "step": 285
+    },
+    {
+      "epoch": 0.7591240875912408,
+      "grad_norm": 1.7012646465038568,
+      "learning_rate": 9.18128665415186e-06,
+      "loss": 1.1728,
+      "step": 286
+    },
+    {
+      "epoch": 0.7617783676177837,
+      "grad_norm": 1.4354980241800914,
+      "learning_rate": 9.175512179418903e-06,
+      "loss": 1.1138,
+      "step": 287
+    },
+    {
+      "epoch": 0.7644326476443265,
+      "grad_norm": 1.5648924104277102,
+      "learning_rate": 9.169719241436162e-06,
+      "loss": 1.0936,
+      "step": 288
+    },
+    {
+      "epoch": 0.7670869276708693,
+      "grad_norm": 1.535950564272176,
+      "learning_rate": 9.163907865818806e-06,
+      "loss": 1.0884,
+      "step": 289
+    },
+    {
+      "epoch": 0.769741207697412,
+      "grad_norm": 1.4657493870841045,
+      "learning_rate": 9.158078078263536e-06,
+      "loss": 1.0962,
+      "step": 290
+    },
+    {
+      "epoch": 0.7723954877239548,
+      "grad_norm": 1.5960566218254721,
+      "learning_rate": 9.152229904548464e-06,
+      "loss": 1.1003,
+      "step": 291
+    },
+    {
+      "epoch": 0.7750497677504977,
+      "grad_norm": 1.5026317273526155,
+      "learning_rate": 9.146363370533004e-06,
+      "loss": 1.1334,
+      "step": 292
+    },
+    {
+      "epoch": 0.7777040477770405,
+      "grad_norm": 1.4667451034506551,
+      "learning_rate": 9.14047850215775e-06,
+      "loss": 1.188,
+      "step": 293
+    },
+    {
+      "epoch": 0.7803583278035833,
+      "grad_norm": 2.5527846830656773,
+      "learning_rate": 9.134575325444377e-06,
+      "loss": 1.1489,
+      "step": 294
+    },
+    {
+      "epoch": 0.783012607830126,
+      "grad_norm": 1.5656317760690617,
+      "learning_rate": 9.128653866495504e-06,
+      "loss": 1.1049,
+      "step": 295
+    },
+    {
+      "epoch": 0.7856668878566688,
+      "grad_norm": 1.4532042000319447,
+      "learning_rate": 9.122714151494599e-06,
+      "loss": 1.1156,
+      "step": 296
+    },
+    {
+      "epoch": 0.7883211678832117,
+      "grad_norm": 1.4759483242959985,
+      "learning_rate": 9.116756206705848e-06,
+      "loss": 1.1396,
+      "step": 297
+    },
+    {
+      "epoch": 0.7909754479097545,
+      "grad_norm": 1.4531099151254951,
+      "learning_rate": 9.110780058474052e-06,
+      "loss": 1.1011,
+      "step": 298
+    },
+    {
+      "epoch": 0.7936297279362973,
+      "grad_norm": 1.509245001105786,
+      "learning_rate": 9.104785733224498e-06,
+      "loss": 1.1052,
+      "step": 299
+    },
+    {
+      "epoch": 0.79628400796284,
+      "grad_norm": 1.4742686115404562,
+      "learning_rate": 9.09877325746285e-06,
+      "loss": 1.1627,
+      "step": 300
+    },
+    {
+      "epoch": 0.7989382879893829,
+      "grad_norm": 1.4451227706627736,
+      "learning_rate": 9.092742657775031e-06,
+      "loss": 1.1118,
+      "step": 301
+    },
+    {
+      "epoch": 0.8015925680159257,
+      "grad_norm": 1.575230566769605,
+      "learning_rate": 9.086693960827106e-06,
+      "loss": 1.1625,
+      "step": 302
+    },
+    {
+      "epoch": 0.8042468480424685,
+      "grad_norm": 1.6679637319120473,
+      "learning_rate": 9.080627193365155e-06,
+      "loss": 1.1452,
+      "step": 303
+    },
+    {
+      "epoch": 0.8069011280690113,
+      "grad_norm": 1.4072750238146392,
+      "learning_rate": 9.07454238221517e-06,
+      "loss": 1.1121,
+      "step": 304
+    },
+    {
+      "epoch": 0.809555408095554,
+      "grad_norm": 1.399645387242144,
+      "learning_rate": 9.068439554282924e-06,
+      "loss": 1.1101,
+      "step": 305
+    },
+    {
+      "epoch": 0.8122096881220969,
+      "grad_norm": 1.9740369624876526,
+      "learning_rate": 9.06231873655386e-06,
+      "loss": 1.0986,
+      "step": 306
+    },
+    {
+      "epoch": 0.8148639681486397,
+      "grad_norm": 1.4581046261229995,
+      "learning_rate": 9.056179956092961e-06,
+      "loss": 1.1228,
+      "step": 307
+    },
+    {
+      "epoch": 0.8175182481751825,
+      "grad_norm": 2.628430909687979,
+      "learning_rate": 9.050023240044649e-06,
+      "loss": 1.0783,
+      "step": 308
+    },
+    {
+      "epoch": 0.8201725282017253,
+      "grad_norm": 1.6691124773863195,
+      "learning_rate": 9.043848615632643e-06,
+      "loss": 1.167,
+      "step": 309
+    },
+    {
+      "epoch": 0.822826808228268,
+      "grad_norm": 1.7459906965590473,
+      "learning_rate": 9.03765611015985e-06,
+      "loss": 1.2287,
+      "step": 310
+    },
+    {
+      "epoch": 0.8254810882548109,
+      "grad_norm": 1.5373249007323673,
+      "learning_rate": 9.031445751008252e-06,
+      "loss": 1.1446,
+      "step": 311
+    },
+    {
+      "epoch": 0.8281353682813537,
+      "grad_norm": 1.526522854497616,
+      "learning_rate": 9.025217565638766e-06,
+      "loss": 1.1609,
+      "step": 312
+    },
+    {
+      "epoch": 0.8307896483078965,
+      "grad_norm": 1.3715974716678416,
+      "learning_rate": 9.018971581591141e-06,
+      "loss": 1.1761,
+      "step": 313
+    },
+    {
+      "epoch": 0.8334439283344393,
+      "grad_norm": 1.733161587991312,
+      "learning_rate": 9.012707826483823e-06,
+      "loss": 1.1241,
+      "step": 314
+    },
+    {
+      "epoch": 0.836098208360982,
+      "grad_norm": 1.5851407690090333,
+      "learning_rate": 9.006426328013838e-06,
+      "loss": 1.1898,
+      "step": 315
+    },
+    {
+      "epoch": 0.8387524883875249,
+      "grad_norm": 1.492565448115301,
+      "learning_rate": 9.000127113956673e-06,
+      "loss": 1.1281,
+      "step": 316
+    },
+    {
+      "epoch": 0.8414067684140677,
+      "grad_norm": 1.4675427619453145,
+      "learning_rate": 8.993810212166147e-06,
+      "loss": 1.1078,
+      "step": 317
+    },
+    {
+      "epoch": 0.8440610484406105,
+      "grad_norm": 1.7806802137808329,
+      "learning_rate": 8.987475650574289e-06,
+      "loss": 1.1113,
+      "step": 318
+    },
+    {
+      "epoch": 0.8467153284671532,
+      "grad_norm": 1.7957085592461643,
+      "learning_rate": 8.98112345719122e-06,
+      "loss": 1.0371,
+      "step": 319
+    },
+    {
+      "epoch": 0.8493696084936961,
+      "grad_norm": 1.6891739001445774,
+      "learning_rate": 8.974753660105023e-06,
+      "loss": 1.1939,
+      "step": 320
+    },
+    {
+      "epoch": 0.8520238885202389,
+      "grad_norm": 1.361414937851007,
+      "learning_rate": 8.968366287481621e-06,
+      "loss": 1.0606,
+      "step": 321
+    },
+    {
+      "epoch": 0.8546781685467817,
+      "grad_norm": 1.5477011631255944,
+      "learning_rate": 8.961961367564652e-06,
+      "loss": 1.1343,
+      "step": 322
+    },
+    {
+      "epoch": 0.8573324485733245,
+      "grad_norm": 1.398038196798421,
+      "learning_rate": 8.955538928675343e-06,
+      "loss": 1.0537,
+      "step": 323
+    },
+    {
+      "epoch": 0.8599867285998672,
+      "grad_norm": 1.4829616588106211,
+      "learning_rate": 8.94909899921239e-06,
+      "loss": 1.1244,
+      "step": 324
+    },
+    {
+      "epoch": 0.8626410086264101,
+      "grad_norm": 1.458234865181319,
+      "learning_rate": 8.94264160765183e-06,
+      "loss": 1.0945,
+      "step": 325
+    },
+    {
+      "epoch": 0.8652952886529529,
+      "grad_norm": 1.48674147774638,
+      "learning_rate": 8.936166782546907e-06,
+      "loss": 1.0698,
+      "step": 326
+    },
+    {
+      "epoch": 0.8679495686794957,
+      "grad_norm": 1.3468414140104497,
+      "learning_rate": 8.929674552527956e-06,
+      "loss": 1.0428,
+      "step": 327
+    },
+    {
+      "epoch": 0.8706038487060385,
+      "grad_norm": 1.4927690225590464,
+      "learning_rate": 8.923164946302274e-06,
+      "loss": 1.1367,
+      "step": 328
+    },
+    {
+      "epoch": 0.8732581287325812,
+      "grad_norm": 1.3780023302624582,
+      "learning_rate": 8.91663799265399e-06,
+      "loss": 1.1048,
+      "step": 329
+    },
+    {
+      "epoch": 0.8759124087591241,
+      "grad_norm": 1.8364949062401694,
+      "learning_rate": 8.910093720443945e-06,
+      "loss": 1.1962,
+      "step": 330
+    },
+    {
+      "epoch": 0.8785666887856669,
+      "grad_norm": 1.6803377587803117,
+      "learning_rate": 8.903532158609548e-06,
+      "loss": 1.1919,
+      "step": 331
+    },
+    {
+      "epoch": 0.8812209688122097,
+      "grad_norm": 1.5621810302199315,
+      "learning_rate": 8.89695333616467e-06,
+      "loss": 1.1177,
+      "step": 332
+    },
+    {
+      "epoch": 0.8838752488387525,
+      "grad_norm": 1.4462262022449852,
+      "learning_rate": 8.890357282199504e-06,
+      "loss": 1.1321,
+      "step": 333
+    },
+    {
+      "epoch": 0.8865295288652952,
+      "grad_norm": 1.4286236136174415,
+      "learning_rate": 8.883744025880429e-06,
+      "loss": 1.1717,
+      "step": 334
+    },
+    {
+      "epoch": 0.8891838088918381,
+      "grad_norm": 1.4484748876813012,
+      "learning_rate": 8.877113596449895e-06,
+      "loss": 1.1004,
+      "step": 335
+    },
+    {
+      "epoch": 0.8918380889183809,
+      "grad_norm": 1.4164984401949983,
+      "learning_rate": 8.87046602322629e-06,
+      "loss": 1.079,
+      "step": 336
+    },
+    {
+      "epoch": 0.8944923689449237,
+      "grad_norm": 1.3708607011124272,
+      "learning_rate": 8.863801335603802e-06,
+      "loss": 1.133,
+      "step": 337
+    },
+    {
+      "epoch": 0.8971466489714665,
+      "grad_norm": 1.3626382714893748,
+      "learning_rate": 8.857119563052301e-06,
+      "loss": 1.0734,
+      "step": 338
+    },
+    {
+      "epoch": 0.8998009289980093,
+      "grad_norm": 1.5082034601534042,
+      "learning_rate": 8.850420735117202e-06,
+      "loss": 1.1691,
+      "step": 339
+    },
+    {
+      "epoch": 0.9024552090245521,
+      "grad_norm": 1.3234730893075355,
+      "learning_rate": 8.843704881419333e-06,
+      "loss": 1.046,
+      "step": 340
+    },
+    {
+      "epoch": 0.9051094890510949,
+      "grad_norm": 1.4896833219647911,
+      "learning_rate": 8.836972031654807e-06,
+      "loss": 1.1586,
+      "step": 341
+    },
+    {
+      "epoch": 0.9077637690776377,
+      "grad_norm": 1.3697029850159739,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 1.0817,
+      "step": 342
+    },
+    {
+      "epoch": 0.9104180491041804,
+      "grad_norm": 1.747564979115208,
+      "learning_rate": 8.823455463085873e-06,
+      "loss": 1.0905,
+      "step": 343
+    },
+    {
+      "epoch": 0.9130723291307233,
+      "grad_norm": 1.5649272934153584,
+      "learning_rate": 8.816671804048933e-06,
+      "loss": 1.0434,
+      "step": 344
+    },
+    {
+      "epoch": 0.9157266091572661,
+      "grad_norm": 1.4823250348157,
+      "learning_rate": 8.809871268480004e-06,
+      "loss": 1.0895,
+      "step": 345
+    },
+    {
+      "epoch": 0.9183808891838089,
+      "grad_norm": 1.4264959835661182,
+      "learning_rate": 8.803053886449644e-06,
+      "loss": 1.1502,
+      "step": 346
+    },
+    {
+      "epoch": 0.9210351692103517,
+      "grad_norm": 1.5424239648407791,
+      "learning_rate": 8.796219688102906e-06,
+      "loss": 1.0734,
+      "step": 347
+    },
+    {
+      "epoch": 0.9236894492368944,
+      "grad_norm": 1.594778792432936,
+      "learning_rate": 8.789368703659199e-06,
+      "loss": 1.06,
+      "step": 348
+    },
+    {
+      "epoch": 0.9263437292634373,
+      "grad_norm": 1.425756455063989,
+      "learning_rate": 8.782500963412156e-06,
+      "loss": 1.1091,
+      "step": 349
+    },
+    {
+      "epoch": 0.9289980092899801,
+      "grad_norm": 1.4480941030784251,
+      "learning_rate": 8.775616497729502e-06,
+      "loss": 1.1146,
+      "step": 350
+    },
+    {
+      "epoch": 0.9316522893165229,
+      "grad_norm": 1.9595578470904635,
+      "learning_rate": 8.768715337052918e-06,
+      "loss": 1.1353,
+      "step": 351
+    },
+    {
+      "epoch": 0.9343065693430657,
+      "grad_norm": 1.6462295827570508,
+      "learning_rate": 8.761797511897907e-06,
+      "loss": 1.1376,
+      "step": 352
+    },
+    {
+      "epoch": 0.9369608493696084,
+      "grad_norm": 1.393588576405631,
+      "learning_rate": 8.754863052853658e-06,
+      "loss": 1.1317,
+      "step": 353
+    },
+    {
+      "epoch": 0.9396151293961513,
+      "grad_norm": 2.230474529090937,
+      "learning_rate": 8.747911990582912e-06,
+      "loss": 1.1086,
+      "step": 354
+    },
+    {
+      "epoch": 0.9422694094226941,
+      "grad_norm": 1.809443765521074,
+      "learning_rate": 8.740944355821827e-06,
+      "loss": 1.1018,
+      "step": 355
+    },
+    {
+      "epoch": 0.9449236894492369,
+      "grad_norm": 1.6826959358419462,
+      "learning_rate": 8.733960179379842e-06,
+      "loss": 1.1766,
+      "step": 356
+    },
+    {
+      "epoch": 0.9475779694757797,
+      "grad_norm": 1.429793323082417,
+      "learning_rate": 8.726959492139535e-06,
+      "loss": 1.062,
+      "step": 357
+    },
+    {
+      "epoch": 0.9502322495023225,
+      "grad_norm": 1.3304241051942485,
+      "learning_rate": 8.719942325056496e-06,
+      "loss": 1.0864,
+      "step": 358
+    },
+    {
+      "epoch": 0.9528865295288653,
+      "grad_norm": 1.838527760485716,
+      "learning_rate": 8.712908709159183e-06,
+      "loss": 1.08,
+      "step": 359
+    },
+    {
+      "epoch": 0.9555408095554081,
+      "grad_norm": 1.8095644142003555,
+      "learning_rate": 8.70585867554879e-06,
+      "loss": 1.0622,
+      "step": 360
+    },
+    {
+      "epoch": 0.9581950895819509,
+      "grad_norm": 1.3961944428914481,
+      "learning_rate": 8.698792255399104e-06,
+      "loss": 1.1279,
+      "step": 361
+    },
+    {
+      "epoch": 0.9608493696084937,
+      "grad_norm": 1.4265196608054989,
+      "learning_rate": 8.691709479956373e-06,
+      "loss": 1.0786,
+      "step": 362
+    },
+    {
+      "epoch": 0.9635036496350365,
+      "grad_norm": 1.5175948559199692,
+      "learning_rate": 8.68461038053916e-06,
+      "loss": 1.1046,
+      "step": 363
+    },
+    {
+      "epoch": 0.9661579296615793,
+      "grad_norm": 1.5709878342434411,
+      "learning_rate": 8.67749498853821e-06,
+      "loss": 1.0947,
+      "step": 364
+    },
+    {
+      "epoch": 0.9688122096881221,
+      "grad_norm": 1.5372734019009258,
+      "learning_rate": 8.670363335416319e-06,
+      "loss": 1.0346,
+      "step": 365
+    },
+    {
+      "epoch": 0.9714664897146649,
+      "grad_norm": 1.3550031766754063,
+      "learning_rate": 8.663215452708173e-06,
+      "loss": 1.0868,
+      "step": 366
+    },
+    {
+      "epoch": 0.9741207697412076,
+      "grad_norm": 1.5040356499297907,
+      "learning_rate": 8.656051372020232e-06,
+      "loss": 1.1083,
+      "step": 367
+    },
+    {
+      "epoch": 0.9767750497677505,
+      "grad_norm": 1.5264462091802162,
+      "learning_rate": 8.648871125030576e-06,
+      "loss": 1.1647,
+      "step": 368
+    },
+    {
+      "epoch": 0.9794293297942933,
+      "grad_norm": 1.8183949324824284,
+      "learning_rate": 8.64167474348877e-06,
+      "loss": 1.0809,
+      "step": 369
+    },
+    {
+      "epoch": 0.9820836098208361,
+      "grad_norm": 2.6148655405710874,
+      "learning_rate": 8.634462259215719e-06,
+      "loss": 1.1195,
+      "step": 370
+    },
+    {
+      "epoch": 0.9847378898473789,
+      "grad_norm": 1.5140959417993884,
+      "learning_rate": 8.627233704103538e-06,
+      "loss": 1.0768,
+      "step": 371
+    },
+    {
+      "epoch": 0.9873921698739216,
+      "grad_norm": 1.3953146864224168,
+      "learning_rate": 8.619989110115398e-06,
+      "loss": 1.0998,
+      "step": 372
+    },
+    {
+      "epoch": 0.9900464499004645,
+      "grad_norm": 1.5342377987936564,
+      "learning_rate": 8.612728509285395e-06,
+      "loss": 1.1627,
+      "step": 373
+    },
+    {
+      "epoch": 0.9927007299270073,
+      "grad_norm": 1.6359257997310512,
+      "learning_rate": 8.6054519337184e-06,
+      "loss": 1.0947,
+      "step": 374
+    },
+    {
+      "epoch": 0.9953550099535501,
+      "grad_norm": 1.468781234700457,
+      "learning_rate": 8.59815941558992e-06,
+      "loss": 1.0958,
+      "step": 375
+    },
+    {
+      "epoch": 0.9980092899800929,
+      "grad_norm": 1.8585796860334978,
+      "learning_rate": 8.590850987145964e-06,
+      "loss": 1.1439,
+      "step": 376
+    },
+    {
+      "epoch": 1.0006635700066357,
+      "grad_norm": 1.4419076627140608,
+      "learning_rate": 8.583526680702888e-06,
+      "loss": 1.1053,
+      "step": 377
+    },
+    {
+      "epoch": 1.0033178500331785,
+      "grad_norm": 1.9598302676254797,
+      "learning_rate": 8.576186528647253e-06,
+      "loss": 1.1538,
+      "step": 378
+    },
+    {
+      "epoch": 1.00199203187251,
+      "grad_norm": 2.6817151626469315,
+      "learning_rate": 8.568830563435695e-06,
+      "loss": 0.8784,
+      "step": 379
+    },
+    {
+      "epoch": 1.00464807436919,
+      "grad_norm": 2.3656261104141683,
+      "learning_rate": 8.561458817594767e-06,
+      "loss": 0.8496,
+      "step": 380
+    },
+    {
+      "epoch": 1.0073041168658698,
+      "grad_norm": 3.236925490140009,
+      "learning_rate": 8.554071323720802e-06,
+      "loss": 0.835,
+      "step": 381
+    },
+    {
+      "epoch": 1.0099601593625498,
+      "grad_norm": 3.843007571124674,
+      "learning_rate": 8.546668114479769e-06,
+      "loss": 0.8405,
+      "step": 382
+    },
+    {
+      "epoch": 1.0126162018592297,
+      "grad_norm": 1.9111030291143845,
+      "learning_rate": 8.53924922260712e-06,
+      "loss": 0.8939,
+      "step": 383
+    },
+    {
+      "epoch": 1.0152722443559097,
+      "grad_norm": 2.377637900130101,
+      "learning_rate": 8.531814680907664e-06,
+      "loss": 0.8582,
+      "step": 384
+    },
+    {
+      "epoch": 1.0179282868525896,
+      "grad_norm": 2.0107910554499444,
+      "learning_rate": 8.5243645222554e-06,
+      "loss": 0.8601,
+      "step": 385
+    },
+    {
+      "epoch": 1.0205843293492696,
+      "grad_norm": 1.6935928768441335,
+      "learning_rate": 8.51689877959339e-06,
+      "loss": 0.8809,
+      "step": 386
+    },
+    {
+      "epoch": 1.0232403718459495,
+      "grad_norm": 1.704066721140233,
+      "learning_rate": 8.509417485933598e-06,
+      "loss": 0.8165,
+      "step": 387
+    },
+    {
+      "epoch": 1.0258964143426295,
+      "grad_norm": 1.6573154345299554,
+      "learning_rate": 8.501920674356755e-06,
+      "loss": 0.775,
+      "step": 388
+    },
+    {
+      "epoch": 1.0285524568393094,
+      "grad_norm": 1.6237544976562632,
+      "learning_rate": 8.494408378012208e-06,
+      "loss": 0.8115,
+      "step": 389
+    },
+    {
+      "epoch": 1.0312084993359893,
+      "grad_norm": 1.7621516061992955,
+      "learning_rate": 8.48688063011778e-06,
+      "loss": 0.867,
+      "step": 390
+    },
+    {
+      "epoch": 1.0338645418326693,
+      "grad_norm": 1.656679120365536,
+      "learning_rate": 8.479337463959607e-06,
+      "loss": 0.8387,
+      "step": 391
+    },
+    {
+      "epoch": 1.0365205843293492,
+      "grad_norm": 1.8462379110802485,
+      "learning_rate": 8.471778912892008e-06,
+      "loss": 0.7986,
+      "step": 392
+    },
+    {
+      "epoch": 1.0391766268260292,
+      "grad_norm": 1.5346458154712161,
+      "learning_rate": 8.46420501033733e-06,
+      "loss": 0.7449,
+      "step": 393
+    },
+    {
+      "epoch": 1.0418326693227091,
+      "grad_norm": 1.6072720048731164,
+      "learning_rate": 8.456615789785804e-06,
+      "loss": 0.789,
+      "step": 394
+    },
+    {
+      "epoch": 1.044488711819389,
+      "grad_norm": 1.6542954637743557,
+      "learning_rate": 8.449011284795389e-06,
+      "loss": 0.8418,
+      "step": 395
+    },
+    {
+      "epoch": 1.047144754316069,
+      "grad_norm": 1.6140250309056903,
+      "learning_rate": 8.441391528991629e-06,
+      "loss": 0.787,
+      "step": 396
+    },
+    {
+      "epoch": 1.049800796812749,
+      "grad_norm": 1.6969918225387495,
+      "learning_rate": 8.433756556067506e-06,
+      "loss": 0.7224,
+      "step": 397
+    },
+    {
+      "epoch": 1.052456839309429,
+      "grad_norm": 1.6791557339222636,
+      "learning_rate": 8.42610639978329e-06,
+      "loss": 0.7711,
+      "step": 398
+    },
+    {
+      "epoch": 1.0551128818061088,
+      "grad_norm": 1.8496221919094664,
+      "learning_rate": 8.418441093966387e-06,
+      "loss": 0.8002,
+      "step": 399
+    },
+    {
+      "epoch": 1.0577689243027888,
+      "grad_norm": 1.6730347681300224,
+      "learning_rate": 8.410760672511188e-06,
+      "loss": 0.7967,
+      "step": 400
+    },
+    {
+      "epoch": 1.0604249667994687,
+      "grad_norm": 1.6351550046155985,
+      "learning_rate": 8.403065169378932e-06,
+      "loss": 0.7733,
+      "step": 401
+    },
+    {
+      "epoch": 1.0630810092961487,
+      "grad_norm": 1.7116450424772824,
+      "learning_rate": 8.395354618597533e-06,
+      "loss": 0.7989,
+      "step": 402
+    },
+    {
+      "epoch": 1.0657370517928286,
+      "grad_norm": 1.7530198918200703,
+      "learning_rate": 8.387629054261454e-06,
+      "loss": 0.8113,
+      "step": 403
+    },
+    {
+      "epoch": 1.0683930942895086,
+      "grad_norm": 1.7205729697156094,
+      "learning_rate": 8.379888510531536e-06,
+      "loss": 0.7841,
+      "step": 404
+    },
+    {
+      "epoch": 1.0710491367861885,
+      "grad_norm": 1.5855647522801817,
+      "learning_rate": 8.37213302163486e-06,
+      "loss": 0.7764,
+      "step": 405
+    },
+    {
+      "epoch": 1.0737051792828685,
+      "grad_norm": 1.7069540659005535,
+      "learning_rate": 8.364362621864595e-06,
+      "loss": 0.7622,
+      "step": 406
+    },
+    {
+      "epoch": 1.0763612217795484,
+      "grad_norm": 2.16867793043578,
+      "learning_rate": 8.356577345579836e-06,
+      "loss": 0.809,
+      "step": 407
+    },
+    {
+      "epoch": 1.0790172642762283,
+      "grad_norm": 1.6510857902080103,
+      "learning_rate": 8.348777227205462e-06,
+      "loss": 0.8271,
+      "step": 408
+    },
+    {
+      "epoch": 1.0816733067729083,
+      "grad_norm": 1.6345528498547808,
+      "learning_rate": 8.34096230123198e-06,
+      "loss": 0.8222,
+      "step": 409
+    },
+    {
+      "epoch": 1.0843293492695882,
+      "grad_norm": 1.6586084882710281,
+      "learning_rate": 8.333132602215374e-06,
+      "loss": 0.8207,
+      "step": 410
+    },
+    {
+      "epoch": 1.0869853917662682,
+      "grad_norm": 1.7037280491243554,
+      "learning_rate": 8.325288164776952e-06,
+      "loss": 0.8023,
+      "step": 411
+    },
+    {
+      "epoch": 1.0896414342629481,
+      "grad_norm": 1.7599459329779006,
+      "learning_rate": 8.31742902360319e-06,
+      "loss": 0.8487,
+      "step": 412
+    },
+    {
+      "epoch": 1.092297476759628,
+      "grad_norm": 1.6713698600025593,
+      "learning_rate": 8.309555213445583e-06,
+      "loss": 0.7517,
+      "step": 413
+    },
+    {
+      "epoch": 1.094953519256308,
+      "grad_norm": 1.9144563268361539,
+      "learning_rate": 8.301666769120488e-06,
+      "loss": 0.7743,
+      "step": 414
+    },
+    {
+      "epoch": 1.097609561752988,
+      "grad_norm": 1.803684622272647,
+      "learning_rate": 8.29376372550897e-06,
+      "loss": 0.8916,
+      "step": 415
+    },
+    {
+      "epoch": 1.100265604249668,
+      "grad_norm": 1.7244484243906384,
+      "learning_rate": 8.28584611755665e-06,
+      "loss": 0.8751,
+      "step": 416
+    },
+    {
+      "epoch": 1.1029216467463479,
+      "grad_norm": 1.571447077389202,
+      "learning_rate": 8.277913980273556e-06,
+      "loss": 0.792,
+      "step": 417
+    },
+    {
+      "epoch": 1.1055776892430278,
+      "grad_norm": 1.6265661139125767,
+      "learning_rate": 8.269967348733947e-06,
+      "loss": 0.8271,
+      "step": 418
+    },
+    {
+      "epoch": 1.1082337317397077,
+      "grad_norm": 1.4867715710827545,
+      "learning_rate": 8.262006258076187e-06,
+      "loss": 0.7518,
+      "step": 419
+    },
+    {
+      "epoch": 1.1108897742363877,
+      "grad_norm": 1.9065325615287738,
+      "learning_rate": 8.25403074350257e-06,
+      "loss": 0.8217,
+      "step": 420
+    },
+    {
+      "epoch": 1.1135458167330676,
+      "grad_norm": 1.4856711522583446,
+      "learning_rate": 8.246040840279165e-06,
+      "loss": 0.7575,
+      "step": 421
+    },
+    {
+      "epoch": 1.1162018592297476,
+      "grad_norm": 1.6522511921393792,
+      "learning_rate": 8.238036583735673e-06,
+      "loss": 0.8373,
+      "step": 422
+    },
+    {
+      "epoch": 1.1188579017264275,
+      "grad_norm": 1.6086804575482818,
+      "learning_rate": 8.230018009265255e-06,
+      "loss": 0.7999,
+      "step": 423
+    },
+    {
+      "epoch": 1.1215139442231075,
+      "grad_norm": 1.7684786767218659,
+      "learning_rate": 8.221985152324385e-06,
+      "loss": 0.8025,
+      "step": 424
+    },
+    {
+      "epoch": 1.1241699867197874,
+      "grad_norm": 1.6725764333554063,
+      "learning_rate": 8.213938048432697e-06,
+      "loss": 0.8117,
+      "step": 425
+    },
+    {
+      "epoch": 1.1268260292164674,
+      "grad_norm": 1.7827929534114502,
+      "learning_rate": 8.205876733172813e-06,
+      "loss": 0.8309,
+      "step": 426
+    },
+    {
+      "epoch": 1.1294820717131473,
+      "grad_norm": 1.698478383799961,
+      "learning_rate": 8.197801242190204e-06,
+      "loss": 0.8268,
+      "step": 427
+    },
+    {
+      "epoch": 1.1321381142098272,
+      "grad_norm": 1.8118203307554201,
+      "learning_rate": 8.189711611193012e-06,
+      "loss": 0.849,
+      "step": 428
+    },
+    {
+      "epoch": 1.1347941567065072,
+      "grad_norm": 2.070219778475728,
+      "learning_rate": 8.181607875951911e-06,
+      "loss": 0.7663,
+      "step": 429
+    },
+    {
+      "epoch": 1.1374501992031871,
+      "grad_norm": 1.6850427736472506,
+      "learning_rate": 8.17349007229994e-06,
+      "loss": 0.7611,
+      "step": 430
+    },
+    {
+      "epoch": 1.140106241699867,
+      "grad_norm": 1.6749598308998042,
+      "learning_rate": 8.165358236132347e-06,
+      "loss": 0.8187,
+      "step": 431
+    },
+    {
+      "epoch": 1.1427622841965472,
+      "grad_norm": 1.6288782095132437,
+      "learning_rate": 8.157212403406424e-06,
+      "loss": 0.8636,
+      "step": 432
+    },
+    {
+      "epoch": 1.1454183266932272,
+      "grad_norm": 4.442514317024235,
+      "learning_rate": 8.149052610141357e-06,
+      "loss": 0.7602,
+      "step": 433
+    },
+    {
+      "epoch": 1.1480743691899071,
+      "grad_norm": 1.8429556263296754,
+      "learning_rate": 8.14087889241806e-06,
+      "loss": 0.8371,
+      "step": 434
+    },
+    {
+      "epoch": 1.150730411686587,
+      "grad_norm": 1.7045658311688014,
+      "learning_rate": 8.132691286379022e-06,
+      "loss": 0.8294,
+      "step": 435
+    },
+    {
+      "epoch": 1.153386454183267,
+      "grad_norm": 1.6060394678635606,
+      "learning_rate": 8.124489828228136e-06,
+      "loss": 0.7894,
+      "step": 436
+    },
+    {
+      "epoch": 1.156042496679947,
+      "grad_norm": 1.6620423492382332,
+      "learning_rate": 8.116274554230557e-06,
+      "loss": 0.8314,
+      "step": 437
+    },
+    {
+      "epoch": 1.158698539176627,
+      "grad_norm": 1.6318153684467354,
+      "learning_rate": 8.108045500712518e-06,
+      "loss": 0.7925,
+      "step": 438
+    },
+    {
+      "epoch": 1.1613545816733069,
+      "grad_norm": 1.549105187586498,
+      "learning_rate": 8.099802704061194e-06,
+      "loss": 0.7802,
+      "step": 439
+    },
+    {
+      "epoch": 1.1640106241699868,
+      "grad_norm": 1.5295943417332036,
+      "learning_rate": 8.091546200724521e-06,
+      "loss": 0.783,
+      "step": 440
+    },
+    {
+      "epoch": 1.1666666666666667,
+      "grad_norm": 1.6920608032596305,
+      "learning_rate": 8.083276027211049e-06,
+      "loss": 0.8102,
+      "step": 441
+    },
+    {
+      "epoch": 1.1693227091633467,
+      "grad_norm": 1.6969436680242431,
+      "learning_rate": 8.07499222008977e-06,
+      "loss": 0.8349,
+      "step": 442
+    },
+    {
+      "epoch": 1.1719787516600266,
+      "grad_norm": 1.667367336891202,
+      "learning_rate": 8.066694815989961e-06,
+      "loss": 0.8588,
+      "step": 443
+    },
+    {
+      "epoch": 1.1746347941567066,
+      "grad_norm": 1.5443285933176338,
+      "learning_rate": 8.058383851601027e-06,
+      "loss": 0.8374,
+      "step": 444
+    },
+    {
+      "epoch": 1.1772908366533865,
+      "grad_norm": 1.6330085580242453,
+      "learning_rate": 8.05005936367233e-06,
+      "loss": 0.8429,
+      "step": 445
+    },
+    {
+      "epoch": 1.1799468791500665,
+      "grad_norm": 1.5343417288603696,
+      "learning_rate": 8.041721389013029e-06,
+      "loss": 0.7969,
+      "step": 446
+    },
+    {
+      "epoch": 1.1826029216467464,
+      "grad_norm": 1.6583884933795217,
+      "learning_rate": 8.033369964491924e-06,
+      "loss": 0.8525,
+      "step": 447
+    },
+    {
+      "epoch": 1.1852589641434264,
+      "grad_norm": 1.716175663053838,
+      "learning_rate": 8.025005127037282e-06,
+      "loss": 0.7671,
+      "step": 448
+    },
+    {
+      "epoch": 1.1879150066401063,
+      "grad_norm": 1.6417405981278685,
+      "learning_rate": 8.016626913636681e-06,
+      "loss": 0.7946,
+      "step": 449
+    },
+    {
+      "epoch": 1.1905710491367862,
+      "grad_norm": 1.7766412095191615,
+      "learning_rate": 8.008235361336845e-06,
+      "loss": 0.7745,
+      "step": 450
+    },
+    {
+      "epoch": 1.1932270916334662,
+      "grad_norm": 1.6563705890201488,
+      "learning_rate": 7.999830507243478e-06,
+      "loss": 0.8702,
+      "step": 451
+    },
+    {
+      "epoch": 1.1958831341301461,
+      "grad_norm": 1.5442276292341202,
+      "learning_rate": 7.991412388521108e-06,
+      "loss": 0.7552,
+      "step": 452
+    },
+    {
+      "epoch": 1.198539176626826,
+      "grad_norm": 1.6952272228903122,
+      "learning_rate": 7.982981042392907e-06,
+      "loss": 0.8314,
+      "step": 453
+    },
+    {
+      "epoch": 1.201195219123506,
+      "grad_norm": 1.5514245470035095,
+      "learning_rate": 7.974536506140546e-06,
+      "loss": 0.8379,
+      "step": 454
+    },
+    {
+      "epoch": 1.203851261620186,
+      "grad_norm": 1.5461371054169675,
+      "learning_rate": 7.966078817104012e-06,
+      "loss": 0.8277,
+      "step": 455
+    },
+    {
+      "epoch": 1.206507304116866,
+      "grad_norm": 1.537281161736465,
+      "learning_rate": 7.957608012681452e-06,
+      "loss": 0.7524,
+      "step": 456
+    },
+    {
+      "epoch": 1.2091633466135459,
+      "grad_norm": 1.6260111327687876,
+      "learning_rate": 7.94912413032901e-06,
+      "loss": 0.7796,
+      "step": 457
+    },
+    {
+      "epoch": 1.2118193891102258,
+      "grad_norm": 1.5526207364437223,
+      "learning_rate": 7.940627207560655e-06,
+      "loss": 0.7348,
+      "step": 458
+    },
+    {
+      "epoch": 1.2144754316069057,
+      "grad_norm": 1.5637447605556818,
+      "learning_rate": 7.932117281948021e-06,
+      "loss": 0.8037,
+      "step": 459
+    },
+    {
+      "epoch": 1.2171314741035857,
+      "grad_norm": 1.6178024259327493,
+      "learning_rate": 7.923594391120237e-06,
+      "loss": 0.8831,
+      "step": 460
+    },
+    {
+      "epoch": 1.2197875166002656,
+      "grad_norm": 1.850653096718784,
+      "learning_rate": 7.915058572763757e-06,
+      "loss": 0.7854,
+      "step": 461
+    },
+    {
+      "epoch": 1.2224435590969456,
+      "grad_norm": 1.7078288767363847,
+      "learning_rate": 7.906509864622202e-06,
+      "loss": 0.8495,
+      "step": 462
+    },
+    {
+      "epoch": 1.2250996015936255,
+      "grad_norm": 1.5513381210763566,
+      "learning_rate": 7.897948304496189e-06,
+      "loss": 0.7137,
+      "step": 463
+    },
+    {
+      "epoch": 1.2277556440903055,
+      "grad_norm": 1.6755385049155458,
+      "learning_rate": 7.889373930243166e-06,
+      "loss": 0.8259,
+      "step": 464
+    },
+    {
+      "epoch": 1.2304116865869854,
+      "grad_norm": 1.6856311582426975,
+      "learning_rate": 7.880786779777233e-06,
+      "loss": 0.7716,
+      "step": 465
+    },
+    {
+      "epoch": 1.2330677290836654,
+      "grad_norm": 1.7089109714412376,
+      "learning_rate": 7.872186891068997e-06,
+      "loss": 0.8888,
+      "step": 466
+    },
+    {
+      "epoch": 1.2357237715803453,
+      "grad_norm": 1.6298087069461054,
+      "learning_rate": 7.86357430214538e-06,
+      "loss": 0.7837,
+      "step": 467
+    },
+    {
+      "epoch": 1.2383798140770252,
+      "grad_norm": 1.5968984282075225,
+      "learning_rate": 7.854949051089467e-06,
+      "loss": 0.7803,
+      "step": 468
+    },
+    {
+      "epoch": 1.2410358565737052,
+      "grad_norm": 1.7290872403697073,
+      "learning_rate": 7.846311176040331e-06,
+      "loss": 0.7977,
+      "step": 469
+    },
+    {
+      "epoch": 1.2436918990703851,
+      "grad_norm": 1.6477947347063955,
+      "learning_rate": 7.837660715192867e-06,
+      "loss": 0.8181,
+      "step": 470
+    },
+    {
+      "epoch": 1.246347941567065,
+      "grad_norm": 1.7746140005250393,
+      "learning_rate": 7.82899770679762e-06,
+      "loss": 0.8288,
+      "step": 471
+    },
+    {
+      "epoch": 1.249003984063745,
+      "grad_norm": 1.7127466701681282,
+      "learning_rate": 7.820322189160618e-06,
+      "loss": 0.7727,
+      "step": 472
+    },
+    {
+      "epoch": 1.251660026560425,
+      "grad_norm": 1.9553401579042835,
+      "learning_rate": 7.811634200643202e-06,
+      "loss": 0.8062,
+      "step": 473
+    },
+    {
+      "epoch": 1.254316069057105,
+      "grad_norm": 1.7316246001340312,
+      "learning_rate": 7.80293377966186e-06,
+      "loss": 0.7913,
+      "step": 474
+    },
+    {
+      "epoch": 1.2569721115537849,
+      "grad_norm": 1.6106273608789883,
+      "learning_rate": 7.794220964688048e-06,
+      "loss": 0.7915,
+      "step": 475
+    },
+    {
+      "epoch": 1.2596281540504648,
+      "grad_norm": 1.6018273021238745,
+      "learning_rate": 7.78549579424803e-06,
+      "loss": 0.8312,
+      "step": 476
+    },
+    {
+      "epoch": 1.2622841965471447,
+      "grad_norm": 1.5745527953605545,
+      "learning_rate": 7.776758306922703e-06,
+      "loss": 0.8125,
+      "step": 477
+    },
+    {
+      "epoch": 1.2649402390438247,
+      "grad_norm": 1.7593012652879958,
+      "learning_rate": 7.768008541347423e-06,
+      "loss": 0.8354,
+      "step": 478
+    },
+    {
+      "epoch": 1.2675962815405046,
+      "grad_norm": 1.5641289543545533,
+      "learning_rate": 7.759246536211843e-06,
+      "loss": 0.7744,
+      "step": 479
+    },
+    {
+      "epoch": 1.2702523240371846,
+      "grad_norm": 1.6174503354343208,
+      "learning_rate": 7.750472330259735e-06,
+      "loss": 0.8251,
+      "step": 480
+    },
+    {
+      "epoch": 1.2729083665338645,
+      "grad_norm": 1.6488155603473844,
+      "learning_rate": 7.741685962288817e-06,
+      "loss": 0.8155,
+      "step": 481
+    },
+    {
+      "epoch": 1.2755644090305445,
+      "grad_norm": 1.7502255456864444,
+      "learning_rate": 7.732887471150589e-06,
+      "loss": 0.8199,
+      "step": 482
+    },
+    {
+      "epoch": 1.2782204515272244,
+      "grad_norm": 1.6325894467431792,
+      "learning_rate": 7.72407689575016e-06,
+      "loss": 0.7949,
+      "step": 483
+    },
+    {
+      "epoch": 1.2808764940239044,
+      "grad_norm": 1.5450834054808966,
+      "learning_rate": 7.715254275046062e-06,
+      "loss": 0.7488,
+      "step": 484
+    },
+    {
+      "epoch": 1.2835325365205843,
+      "grad_norm": 1.7564206935012117,
+      "learning_rate": 7.7064196480501e-06,
+      "loss": 0.7563,
+      "step": 485
+    },
+    {
+      "epoch": 1.2861885790172642,
+      "grad_norm": 1.4967916410847861,
+      "learning_rate": 7.697573053827163e-06,
+      "loss": 0.7613,
+      "step": 486
+    },
+    {
+      "epoch": 1.2888446215139442,
+      "grad_norm": 1.6339851679991706,
+      "learning_rate": 7.688714531495061e-06,
+      "loss": 0.8494,
+      "step": 487
+    },
+    {
+      "epoch": 1.2915006640106241,
+      "grad_norm": 1.5726472413180292,
+      "learning_rate": 7.67984412022434e-06,
+      "loss": 0.811,
+      "step": 488
+    },
+    {
+      "epoch": 1.294156706507304,
+      "grad_norm": 1.704796493237578,
+      "learning_rate": 7.670961859238124e-06,
+      "loss": 0.758,
+      "step": 489
+    },
+    {
+      "epoch": 1.296812749003984,
+      "grad_norm": 1.8152659977035888,
+      "learning_rate": 7.66206778781193e-06,
+      "loss": 0.7954,
+      "step": 490
+    },
+    {
+      "epoch": 1.299468791500664,
+      "grad_norm": 1.617282796396521,
+      "learning_rate": 7.653161945273497e-06,
+      "loss": 0.7816,
+      "step": 491
+    },
+    {
+      "epoch": 1.302124833997344,
+      "grad_norm": 1.8207616949669707,
+      "learning_rate": 7.644244371002619e-06,
+      "loss": 0.8187,
+      "step": 492
+    },
+    {
+      "epoch": 1.3047808764940239,
+      "grad_norm": 1.7889230655898831,
+      "learning_rate": 7.635315104430959e-06,
+      "loss": 0.7913,
+      "step": 493
+    },
+    {
+      "epoch": 1.3074369189907038,
+      "grad_norm": 1.6026212175311343,
+      "learning_rate": 7.626374185041887e-06,
+      "loss": 0.8469,
+      "step": 494
+    },
+    {
+      "epoch": 1.3100929614873837,
+      "grad_norm": 1.6034434747796849,
+      "learning_rate": 7.617421652370293e-06,
+      "loss": 0.8067,
+      "step": 495
+    },
+    {
+      "epoch": 1.3127490039840637,
+      "grad_norm": 1.7116713587335526,
+      "learning_rate": 7.608457546002423e-06,
+      "loss": 0.8237,
+      "step": 496
+    },
+    {
+      "epoch": 1.3154050464807436,
+      "grad_norm": 1.6713572165788912,
+      "learning_rate": 7.599481905575699e-06,
+      "loss": 0.8205,
+      "step": 497
+    },
+    {
+      "epoch": 1.3180610889774236,
+      "grad_norm": 1.6492136689563734,
+      "learning_rate": 7.5904947707785434e-06,
+      "loss": 0.8062,
+      "step": 498
+    },
+    {
+      "epoch": 1.3207171314741035,
+      "grad_norm": 1.8633234143797999,
+      "learning_rate": 7.581496181350203e-06,
+      "loss": 0.8574,
+      "step": 499
+    },
+    {
+      "epoch": 1.3233731739707835,
+      "grad_norm": 1.5222850331019526,
+      "learning_rate": 7.572486177080576e-06,
+      "loss": 0.8052,
+      "step": 500
+    },
+    {
+      "epoch": 1.3260292164674634,
+      "grad_norm": 1.5407209608086607,
+      "learning_rate": 7.563464797810038e-06,
+      "loss": 0.8536,
+      "step": 501
+    },
+    {
+      "epoch": 1.3286852589641434,
+      "grad_norm": 1.5919884093082233,
+      "learning_rate": 7.554432083429253e-06,
+      "loss": 0.7941,
+      "step": 502
+    },
+    {
+      "epoch": 1.3313413014608233,
+      "grad_norm": 1.6480739901612147,
+      "learning_rate": 7.545388073879018e-06,
+      "loss": 0.8236,
+      "step": 503
+    },
+    {
+      "epoch": 1.3339973439575032,
+      "grad_norm": 1.6210798461631093,
+      "learning_rate": 7.536332809150066e-06,
+      "loss": 0.823,
+      "step": 504
+    },
+    {
+      "epoch": 1.3366533864541832,
+      "grad_norm": 1.5948934469455742,
+      "learning_rate": 7.527266329282905e-06,
+      "loss": 0.7437,
+      "step": 505
+    },
+    {
+      "epoch": 1.3393094289508631,
+      "grad_norm": 1.6266868717112688,
+      "learning_rate": 7.518188674367628e-06,
+      "loss": 0.8009,
+      "step": 506
+    },
+    {
+      "epoch": 1.341965471447543,
+      "grad_norm": 1.7075965106124282,
+      "learning_rate": 7.509099884543745e-06,
+      "loss": 0.7933,
+      "step": 507
+    },
+    {
+      "epoch": 1.3446215139442232,
+      "grad_norm": 1.8851878790913914,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.8752,
+      "step": 508
+    },
+    {
+      "epoch": 1.3472775564409032,
+      "grad_norm": 1.802756747375581,
+      "learning_rate": 7.490889060974202e-06,
+      "loss": 0.8339,
+      "step": 509
+    },
+    {
+      "epoch": 1.3499335989375831,
+      "grad_norm": 1.6221335431584782,
+      "learning_rate": 7.4817671077530295e-06,
+      "loss": 0.8079,
+      "step": 510
+    },
+    {
+      "epoch": 1.352589641434263,
+      "grad_norm": 1.6762124149564765,
+      "learning_rate": 7.4726341806718735e-06,
+      "loss": 0.7527,
+      "step": 511
+    },
+    {
+      "epoch": 1.355245683930943,
+      "grad_norm": 1.817888811411582,
+      "learning_rate": 7.463490320114646e-06,
+      "loss": 0.8421,
+      "step": 512
+    },
+    {
+      "epoch": 1.357901726427623,
+      "grad_norm": 1.6277711511409727,
+      "learning_rate": 7.454335566513603e-06,
+      "loss": 0.8531,
+      "step": 513
+    },
+    {
+      "epoch": 1.360557768924303,
+      "grad_norm": 1.7798649541264224,
+      "learning_rate": 7.445169960349167e-06,
+      "loss": 0.8855,
+      "step": 514
+    },
+    {
+      "epoch": 1.3632138114209829,
+      "grad_norm": 1.6603027624169684,
+      "learning_rate": 7.435993542149751e-06,
+      "loss": 0.8034,
+      "step": 515
+    },
+    {
+      "epoch": 1.3658698539176628,
+      "grad_norm": 1.9257358203569979,
+      "learning_rate": 7.426806352491575e-06,
+      "loss": 0.7991,
+      "step": 516
+    },
+    {
+      "epoch": 1.3685258964143427,
+      "grad_norm": 1.6083096418618006,
+      "learning_rate": 7.417608431998487e-06,
+      "loss": 0.8302,
+      "step": 517
+    },
+    {
+      "epoch": 1.3711819389110227,
+      "grad_norm": 1.6367347889439254,
+      "learning_rate": 7.408399821341787e-06,
+      "loss": 0.7769,
+      "step": 518
+    },
+    {
+      "epoch": 1.3738379814077026,
+      "grad_norm": 1.7083783084901518,
+      "learning_rate": 7.399180561240044e-06,
+      "loss": 0.796,
+      "step": 519
+    },
+    {
+      "epoch": 1.3764940239043826,
+      "grad_norm": 1.5960672101189293,
+      "learning_rate": 7.389950692458916e-06,
+      "loss": 0.8736,
+      "step": 520
+    },
+    {
+      "epoch": 1.3791500664010625,
+      "grad_norm": 1.7397680807457505,
+      "learning_rate": 7.38071025581097e-06,
+      "loss": 0.8015,
+      "step": 521
+    },
+    {
+      "epoch": 1.3818061088977425,
+      "grad_norm": 1.6440955438519538,
+      "learning_rate": 7.371459292155501e-06,
+      "loss": 0.8266,
+      "step": 522
+    },
+    {
+      "epoch": 1.3844621513944224,
+      "grad_norm": 1.7177209541893135,
+      "learning_rate": 7.362197842398355e-06,
+      "loss": 0.7857,
+      "step": 523
+    },
+    {
+      "epoch": 1.3871181938911024,
+      "grad_norm": 1.6619729313815175,
+      "learning_rate": 7.3529259474917455e-06,
+      "loss": 0.7885,
+      "step": 524
+    },
+    {
+      "epoch": 1.3897742363877823,
+      "grad_norm": 1.7199503184792022,
+      "learning_rate": 7.34364364843407e-06,
+      "loss": 0.845,
+      "step": 525
+    },
+    {
+      "epoch": 1.3924302788844622,
+      "grad_norm": 1.8306223658893233,
+      "learning_rate": 7.3343509862697295e-06,
+      "loss": 0.8368,
+      "step": 526
+    },
+    {
+      "epoch": 1.3950863213811422,
+      "grad_norm": 1.5439578957272115,
+      "learning_rate": 7.325048002088955e-06,
+      "loss": 0.8093,
+      "step": 527
+    },
+    {
+      "epoch": 1.3977423638778221,
+      "grad_norm": 1.9216923568783284,
+      "learning_rate": 7.315734737027612e-06,
+      "loss": 0.8178,
+      "step": 528
+    },
+    {
+      "epoch": 1.400398406374502,
+      "grad_norm": 1.5591840729803677,
+      "learning_rate": 7.30641123226703e-06,
+      "loss": 0.7905,
+      "step": 529
+    },
+    {
+      "epoch": 1.403054448871182,
+      "grad_norm": 1.5800929847807523,
+      "learning_rate": 7.297077529033814e-06,
+      "loss": 0.8103,
+      "step": 530
+    },
+    {
+      "epoch": 1.405710491367862,
+      "grad_norm": 1.7493518294241435,
+      "learning_rate": 7.287733668599669e-06,
+      "loss": 0.8348,
+      "step": 531
+    },
+    {
+      "epoch": 1.408366533864542,
+      "grad_norm": 1.9535948089272044,
+      "learning_rate": 7.278379692281209e-06,
+      "loss": 0.8116,
+      "step": 532
+    },
+    {
+      "epoch": 1.4110225763612219,
+      "grad_norm": 1.7032888435426474,
+      "learning_rate": 7.2690156414397775e-06,
+      "loss": 0.7952,
+      "step": 533
+    },
+    {
+      "epoch": 1.4136786188579018,
+      "grad_norm": 1.7632789452727808,
+      "learning_rate": 7.2596415574812695e-06,
+      "loss": 0.8484,
+      "step": 534
+    },
+    {
+      "epoch": 1.4163346613545817,
+      "grad_norm": 1.9343554806946162,
+      "learning_rate": 7.250257481855941e-06,
+      "loss": 0.7913,
+      "step": 535
+    },
+    {
+      "epoch": 1.4189907038512617,
+      "grad_norm": 1.5626080706093446,
+      "learning_rate": 7.24086345605823e-06,
+      "loss": 0.8043,
+      "step": 536
+    },
+    {
+      "epoch": 1.4216467463479416,
+      "grad_norm": 1.6881053251561386,
+      "learning_rate": 7.231459521626574e-06,
+      "loss": 0.7897,
+      "step": 537
+    },
+    {
+      "epoch": 1.4243027888446216,
+      "grad_norm": 1.6829030745040456,
+      "learning_rate": 7.22204572014322e-06,
+      "loss": 0.7077,
+      "step": 538
+    },
+    {
+      "epoch": 1.4269588313413015,
+      "grad_norm": 1.7582319600473906,
+      "learning_rate": 7.212622093234049e-06,
+      "loss": 0.7394,
+      "step": 539
+    },
+    {
+      "epoch": 1.4296148738379815,
+      "grad_norm": 1.82729749435436,
+      "learning_rate": 7.20318868256839e-06,
+      "loss": 0.7831,
+      "step": 540
+    },
+    {
+      "epoch": 1.4322709163346614,
+      "grad_norm": 1.8895948305148935,
+      "learning_rate": 7.193745529858827e-06,
+      "loss": 0.8085,
+      "step": 541
+    },
+    {
+      "epoch": 1.4349269588313414,
+      "grad_norm": 1.5870695996483368,
+      "learning_rate": 7.184292676861024e-06,
+      "loss": 0.7976,
+      "step": 542
+    },
+    {
+      "epoch": 1.4375830013280213,
+      "grad_norm": 1.661068432134725,
+      "learning_rate": 7.174830165373542e-06,
+      "loss": 0.7795,
+      "step": 543
+    },
+    {
+      "epoch": 1.4402390438247012,
+      "grad_norm": 1.7637781271023871,
+      "learning_rate": 7.165358037237644e-06,
+      "loss": 0.7797,
+      "step": 544
+    },
+    {
+      "epoch": 1.4428950863213812,
+      "grad_norm": 1.7573078579380121,
+      "learning_rate": 7.155876334337119e-06,
+      "loss": 0.7881,
+      "step": 545
+    },
+    {
+      "epoch": 1.4455511288180611,
+      "grad_norm": 1.788971177497177,
+      "learning_rate": 7.146385098598092e-06,
+      "loss": 0.7926,
+      "step": 546
+    },
+    {
+      "epoch": 1.448207171314741,
+      "grad_norm": 1.483941017753339,
+      "learning_rate": 7.136884371988844e-06,
+      "loss": 0.7945,
+      "step": 547
+    },
+    {
+      "epoch": 1.450863213811421,
+      "grad_norm": 2.158603581584702,
+      "learning_rate": 7.127374196519616e-06,
+      "loss": 0.8339,
+      "step": 548
+    },
+    {
+      "epoch": 1.453519256308101,
+      "grad_norm": 1.5625619640257642,
+      "learning_rate": 7.117854614242434e-06,
+      "loss": 0.7366,
+      "step": 549
+    },
+    {
+      "epoch": 1.456175298804781,
+      "grad_norm": 1.797618930907081,
+      "learning_rate": 7.10832566725092e-06,
+      "loss": 0.8613,
+      "step": 550
+    },
+    {
+      "epoch": 1.4588313413014609,
+      "grad_norm": 1.7910995750400152,
+      "learning_rate": 7.098787397680104e-06,
+      "loss": 0.8439,
+      "step": 551
+    },
+    {
+      "epoch": 1.4614873837981408,
+      "grad_norm": 1.60673403396422,
+      "learning_rate": 7.0892398477062375e-06,
+      "loss": 0.8891,
+      "step": 552
+    },
+    {
+      "epoch": 1.4641434262948207,
+      "grad_norm": 1.6427911933197976,
+      "learning_rate": 7.079683059546607e-06,
+      "loss": 0.8271,
+      "step": 553
+    },
+    {
+      "epoch": 1.4667994687915007,
+      "grad_norm": 1.6285500237390729,
+      "learning_rate": 7.0701170754593516e-06,
+      "loss": 0.8588,
+      "step": 554
+    },
+    {
+      "epoch": 1.4694555112881806,
+      "grad_norm": 1.718883251928568,
+      "learning_rate": 7.060541937743269e-06,
+      "loss": 0.794,
+      "step": 555
+    },
+    {
+      "epoch": 1.4721115537848606,
+      "grad_norm": 1.5365027160461944,
+      "learning_rate": 7.0509576887376375e-06,
+      "loss": 0.7856,
+      "step": 556
+    },
+    {
+      "epoch": 1.4747675962815405,
+      "grad_norm": 1.5936955094419631,
+      "learning_rate": 7.041364370822017e-06,
+      "loss": 0.7704,
+      "step": 557
+    },
+    {
+      "epoch": 1.4774236387782205,
+      "grad_norm": 1.5986404122720763,
+      "learning_rate": 7.031762026416074e-06,
+      "loss": 0.784,
+      "step": 558
+    },
+    {
+      "epoch": 1.4800796812749004,
+      "grad_norm": 1.6290145069786532,
+      "learning_rate": 7.022150697979385e-06,
+      "loss": 0.8711,
+      "step": 559
+    },
+    {
+      "epoch": 1.4827357237715804,
+      "grad_norm": 1.646088871212861,
+      "learning_rate": 7.0125304280112546e-06,
+      "loss": 0.7449,
+      "step": 560
+    },
+    {
+      "epoch": 1.4853917662682603,
+      "grad_norm": 1.628406730851986,
+      "learning_rate": 7.002901259050523e-06,
+      "loss": 0.8154,
+      "step": 561
+    },
+    {
+      "epoch": 1.4880478087649402,
+      "grad_norm": 1.585585277641927,
+      "learning_rate": 6.99326323367538e-06,
+      "loss": 0.8304,
+      "step": 562
+    },
+    {
+      "epoch": 1.4907038512616202,
+      "grad_norm": 1.5731659024948346,
+      "learning_rate": 6.983616394503177e-06,
+      "loss": 0.7599,
+      "step": 563
+    },
+    {
+      "epoch": 1.4933598937583001,
+      "grad_norm": 1.8337268519091863,
+      "learning_rate": 6.9739607841902365e-06,
+      "loss": 0.8634,
+      "step": 564
+    },
+    {
+      "epoch": 1.49601593625498,
+      "grad_norm": 1.6324940281109015,
+      "learning_rate": 6.96429644543167e-06,
+      "loss": 0.7786,
+      "step": 565
+    },
+    {
+      "epoch": 1.49867197875166,
+      "grad_norm": 1.6269379425991606,
+      "learning_rate": 6.954623420961179e-06,
+      "loss": 0.7474,
+      "step": 566
+    },
+    {
+      "epoch": 1.50132802124834,
+      "grad_norm": 1.5892146882859803,
+      "learning_rate": 6.944941753550877e-06,
+      "loss": 0.8374,
+      "step": 567
+    },
+    {
+      "epoch": 1.50398406374502,
+      "grad_norm": 1.7162255540973645,
+      "learning_rate": 6.9352514860110876e-06,
+      "loss": 0.845,
+      "step": 568
+    },
+    {
+      "epoch": 1.5066401062416999,
+      "grad_norm": 1.6706147544781347,
+      "learning_rate": 6.925552661190166e-06,
+      "loss": 0.7899,
+      "step": 569
+    },
+    {
+      "epoch": 1.5092961487383798,
+      "grad_norm": 1.6566681467322444,
+      "learning_rate": 6.915845321974309e-06,
+      "loss": 0.8416,
+      "step": 570
+    },
+    {
+      "epoch": 1.5119521912350598,
+      "grad_norm": 1.6291748860885436,
+      "learning_rate": 6.906129511287358e-06,
+      "loss": 0.8324,
+      "step": 571
+    },
+    {
+      "epoch": 1.5146082337317397,
+      "grad_norm": 2.00417453104611,
+      "learning_rate": 6.8964052720906175e-06,
+      "loss": 0.7287,
+      "step": 572
+    },
+    {
+      "epoch": 1.5172642762284196,
+      "grad_norm": 1.7178352113474786,
+      "learning_rate": 6.886672647382653e-06,
+      "loss": 0.7881,
+      "step": 573
+    },
+    {
+      "epoch": 1.5199203187250996,
+      "grad_norm": 1.894127341764823,
+      "learning_rate": 6.876931680199121e-06,
+      "loss": 0.8068,
+      "step": 574
+    },
+    {
+      "epoch": 1.5225763612217795,
+      "grad_norm": 1.6002909924709852,
+      "learning_rate": 6.867182413612556e-06,
+      "loss": 0.7499,
+      "step": 575
+    },
+    {
+      "epoch": 1.5252324037184595,
+      "grad_norm": 1.6013025100948648,
+      "learning_rate": 6.857424890732195e-06,
+      "loss": 0.7948,
+      "step": 576
+    },
+    {
+      "epoch": 1.5278884462151394,
+      "grad_norm": 1.7428611130277234,
+      "learning_rate": 6.847659154703785e-06,
+      "loss": 0.8532,
+      "step": 577
+    },
+    {
+      "epoch": 1.5305444887118194,
+      "grad_norm": 1.5972099589901563,
+      "learning_rate": 6.837885248709386e-06,
+      "loss": 0.7441,
+      "step": 578
+    },
+    {
+      "epoch": 1.5332005312084993,
+      "grad_norm": 1.7281335555363582,
+      "learning_rate": 6.8281032159671865e-06,
+      "loss": 0.8236,
+      "step": 579
+    },
+    {
+      "epoch": 1.5358565737051793,
+      "grad_norm": 1.6497765487254892,
+      "learning_rate": 6.818313099731308e-06,
+      "loss": 0.846,
+      "step": 580
+    },
+    {
+      "epoch": 1.5385126162018592,
+      "grad_norm": 2.751709704124397,
+      "learning_rate": 6.8085149432916155e-06,
+      "loss": 0.752,
+      "step": 581
+    },
+    {
+      "epoch": 1.5411686586985391,
+      "grad_norm": 1.8663852239561056,
+      "learning_rate": 6.798708789973527e-06,
+      "loss": 0.77,
+      "step": 582
+    },
+    {
+      "epoch": 1.543824701195219,
+      "grad_norm": 1.6698060547997404,
+      "learning_rate": 6.788894683137822e-06,
+      "loss": 0.8077,
+      "step": 583
+    },
+    {
+      "epoch": 1.546480743691899,
+      "grad_norm": 1.6116361837634285,
+      "learning_rate": 6.779072666180447e-06,
+      "loss": 0.8133,
+      "step": 584
+    },
+    {
+      "epoch": 1.549136786188579,
+      "grad_norm": 1.542072654470527,
+      "learning_rate": 6.769242782532324e-06,
+      "loss": 0.7846,
+      "step": 585
+    },
+    {
+      "epoch": 1.551792828685259,
+      "grad_norm": 1.4898452945341245,
+      "learning_rate": 6.759405075659165e-06,
+      "loss": 0.8475,
+      "step": 586
+    },
+    {
+      "epoch": 1.5544488711819389,
+      "grad_norm": 1.5173495908692989,
+      "learning_rate": 6.749559589061273e-06,
+      "loss": 0.7964,
+      "step": 587
+    },
+    {
+      "epoch": 1.5571049136786188,
+      "grad_norm": 1.8749975291685148,
+      "learning_rate": 6.739706366273346e-06,
+      "loss": 0.8505,
+      "step": 588
+    },
+    {
+      "epoch": 1.5597609561752988,
+      "grad_norm": 1.827884336582995,
+      "learning_rate": 6.7298454508642945e-06,
+      "loss": 0.8439,
+      "step": 589
+    },
+    {
+      "epoch": 1.5624169986719787,
+      "grad_norm": 1.6275292009586355,
+      "learning_rate": 6.7199768864370455e-06,
+      "loss": 0.7982,
+      "step": 590
+    },
+    {
+      "epoch": 1.5650730411686586,
+      "grad_norm": 1.5747624399878266,
+      "learning_rate": 6.710100716628345e-06,
+      "loss": 0.797,
+      "step": 591
+    },
+    {
+      "epoch": 1.5677290836653386,
+      "grad_norm": 1.5774705921149774,
+      "learning_rate": 6.700216985108568e-06,
+      "loss": 0.8065,
+      "step": 592
+    },
+    {
+      "epoch": 1.5703851261620185,
+      "grad_norm": 1.5778429721001461,
+      "learning_rate": 6.690325735581532e-06,
+      "loss": 0.8202,
+      "step": 593
+    },
+    {
+      "epoch": 1.5730411686586985,
+      "grad_norm": 1.7604471332843126,
+      "learning_rate": 6.680427011784292e-06,
+      "loss": 0.7897,
+      "step": 594
+    },
+    {
+      "epoch": 1.5756972111553784,
+      "grad_norm": 1.686769635907997,
+      "learning_rate": 6.6705208574869504e-06,
+      "loss": 0.7761,
+      "step": 595
+    },
+    {
+      "epoch": 1.5783532536520584,
+      "grad_norm": 1.5053126110771002,
+      "learning_rate": 6.660607316492471e-06,
+      "loss": 0.8438,
+      "step": 596
+    },
+    {
+      "epoch": 1.5810092961487383,
+      "grad_norm": 1.6763577131611167,
+      "learning_rate": 6.65068643263648e-06,
+      "loss": 0.8336,
+      "step": 597
+    },
+    {
+      "epoch": 1.5836653386454183,
+      "grad_norm": 1.6119056861498604,
+      "learning_rate": 6.640758249787067e-06,
+      "loss": 0.866,
+      "step": 598
+    },
+    {
+      "epoch": 1.5863213811420982,
+      "grad_norm": 1.6391451248424211,
+      "learning_rate": 6.630822811844604e-06,
+      "loss": 0.7924,
+      "step": 599
+    },
+    {
+      "epoch": 1.5889774236387781,
+      "grad_norm": 1.5817186839646116,
+      "learning_rate": 6.620880162741534e-06,
+      "loss": 0.7637,
+      "step": 600
+    },
+    {
+      "epoch": 1.591633466135458,
+      "grad_norm": 1.6537509811359954,
+      "learning_rate": 6.610930346442198e-06,
+      "loss": 0.8418,
+      "step": 601
+    },
+    {
+      "epoch": 1.594289508632138,
+      "grad_norm": 1.6834866329043456,
+      "learning_rate": 6.600973406942617e-06,
+      "loss": 0.8283,
+      "step": 602
+    },
+    {
+      "epoch": 1.596945551128818,
+      "grad_norm": 1.5806755355796025,
+      "learning_rate": 6.591009388270315e-06,
+      "loss": 0.8051,
+      "step": 603
+    },
+    {
+      "epoch": 1.599601593625498,
+      "grad_norm": 1.6625638337455775,
+      "learning_rate": 6.58103833448412e-06,
+      "loss": 0.7581,
+      "step": 604
+    },
+    {
+      "epoch": 1.6022576361221779,
+      "grad_norm": 1.634181305057464,
+      "learning_rate": 6.571060289673966e-06,
+      "loss": 0.7476,
+      "step": 605
+    },
+    {
+      "epoch": 1.6049136786188578,
+      "grad_norm": 34.53829269757645,
+      "learning_rate": 6.5610752979607e-06,
+      "loss": 0.7517,
+      "step": 606
+    },
+    {
+      "epoch": 1.6075697211155378,
+      "grad_norm": 1.8445764149564496,
+      "learning_rate": 6.551083403495885e-06,
+      "loss": 0.8535,
+      "step": 607
+    },
+    {
+      "epoch": 1.6102257636122177,
+      "grad_norm": 1.7840446085100523,
+      "learning_rate": 6.54108465046161e-06,
+      "loss": 0.815,
+      "step": 608
+    },
+    {
+      "epoch": 1.6128818061088976,
+      "grad_norm": 1.5939129613902245,
+      "learning_rate": 6.531079083070289e-06,
+      "loss": 0.8234,
+      "step": 609
+    },
+    {
+      "epoch": 1.6155378486055776,
+      "grad_norm": 1.6112292642729917,
+      "learning_rate": 6.521066745564467e-06,
+      "loss": 0.7916,
+      "step": 610
+    },
+    {
+      "epoch": 1.6181938911022575,
+      "grad_norm": 1.5131144042037576,
+      "learning_rate": 6.511047682216628e-06,
+      "loss": 0.781,
+      "step": 611
+    },
+    {
+      "epoch": 1.6208499335989375,
+      "grad_norm": 1.6311550231780976,
+      "learning_rate": 6.501021937328992e-06,
+      "loss": 0.769,
+      "step": 612
+    },
+    {
+      "epoch": 1.6235059760956174,
+      "grad_norm": 1.5393946017610958,
+      "learning_rate": 6.490989555233328e-06,
+      "loss": 0.7904,
+      "step": 613
+    },
+    {
+      "epoch": 1.6261620185922974,
+      "grad_norm": 1.6268714117788674,
+      "learning_rate": 6.480950580290751e-06,
+      "loss": 0.8433,
+      "step": 614
+    },
+    {
+      "epoch": 1.6288180610889773,
+      "grad_norm": 1.6330202880257099,
+      "learning_rate": 6.470905056891526e-06,
+      "loss": 0.7714,
+      "step": 615
+    },
+    {
+      "epoch": 1.6314741035856573,
+      "grad_norm": 1.5562395275992014,
+      "learning_rate": 6.460853029454879e-06,
+      "loss": 0.7867,
+      "step": 616
+    },
+    {
+      "epoch": 1.6341301460823372,
+      "grad_norm": 2.217671570613303,
+      "learning_rate": 6.450794542428791e-06,
+      "loss": 0.8091,
+      "step": 617
+    },
+    {
+      "epoch": 1.6367861885790171,
+      "grad_norm": 1.544981928227297,
+      "learning_rate": 6.440729640289809e-06,
+      "loss": 0.7813,
+      "step": 618
+    },
+    {
+      "epoch": 1.639442231075697,
+      "grad_norm": 1.4740653481229222,
+      "learning_rate": 6.4306583675428435e-06,
+      "loss": 0.7833,
+      "step": 619
+    },
+    {
+      "epoch": 1.642098273572377,
+      "grad_norm": 1.6073920238722603,
+      "learning_rate": 6.420580768720977e-06,
+      "loss": 0.838,
+      "step": 620
+    },
+    {
+      "epoch": 1.644754316069057,
+      "grad_norm": 1.612548032767887,
+      "learning_rate": 6.410496888385266e-06,
+      "loss": 0.7618,
+      "step": 621
+    },
+    {
+      "epoch": 1.647410358565737,
+      "grad_norm": 1.6152490679947544,
+      "learning_rate": 6.4004067711245366e-06,
+      "loss": 0.8393,
+      "step": 622
+    },
+    {
+      "epoch": 1.6500664010624169,
+      "grad_norm": 1.6599220046956429,
+      "learning_rate": 6.3903104615551956e-06,
+      "loss": 0.8162,
+      "step": 623
+    },
+    {
+      "epoch": 1.6527224435590968,
+      "grad_norm": 1.4931102060723278,
+      "learning_rate": 6.380208004321037e-06,
+      "loss": 0.7566,
+      "step": 624
+    },
+    {
+      "epoch": 1.6553784860557768,
+      "grad_norm": 1.5929400984170048,
+      "learning_rate": 6.370099444093032e-06,
+      "loss": 0.7796,
+      "step": 625
+    },
+    {
+      "epoch": 1.6580345285524567,
+      "grad_norm": 1.727388116576355,
+      "learning_rate": 6.359984825569138e-06,
+      "loss": 0.7898,
+      "step": 626
+    },
+    {
+      "epoch": 1.6606905710491366,
+      "grad_norm": 1.8183131880092849,
+      "learning_rate": 6.349864193474104e-06,
+      "loss": 0.8609,
+      "step": 627
+    },
+    {
+      "epoch": 1.6633466135458166,
+      "grad_norm": 1.8676431191653386,
+      "learning_rate": 6.3397375925592675e-06,
+      "loss": 0.8784,
+      "step": 628
+    },
+    {
+      "epoch": 1.6660026560424965,
+      "grad_norm": 1.6708110546891144,
+      "learning_rate": 6.32960506760236e-06,
+      "loss": 0.7962,
+      "step": 629
+    },
+    {
+      "epoch": 1.6686586985391765,
+      "grad_norm": 1.6511789078182308,
+      "learning_rate": 6.319466663407309e-06,
+      "loss": 0.8078,
+      "step": 630
+    },
+    {
+      "epoch": 1.6713147410358564,
+      "grad_norm": 1.6698570509044466,
+      "learning_rate": 6.309322424804034e-06,
+      "loss": 0.7975,
+      "step": 631
+    },
+    {
+      "epoch": 1.6739707835325364,
+      "grad_norm": 1.9330575842850526,
+      "learning_rate": 6.29917239664826e-06,
+      "loss": 0.7993,
+      "step": 632
+    },
+    {
+      "epoch": 1.6766268260292163,
+      "grad_norm": 1.647084849034996,
+      "learning_rate": 6.289016623821308e-06,
+      "loss": 0.8084,
+      "step": 633
+    },
+    {
+      "epoch": 1.6792828685258963,
+      "grad_norm": 1.6897703069676664,
+      "learning_rate": 6.2788551512299014e-06,
+      "loss": 0.8016,
+      "step": 634
+    },
+    {
+      "epoch": 1.6819389110225762,
+      "grad_norm": 1.5440808150197218,
+      "learning_rate": 6.268688023805965e-06,
+      "loss": 0.7948,
+      "step": 635
+    },
+    {
+      "epoch": 1.6845949535192561,
+      "grad_norm": 2.7749014353276165,
+      "learning_rate": 6.25851528650643e-06,
+      "loss": 0.7971,
+      "step": 636
+    },
+    {
+      "epoch": 1.687250996015936,
+      "grad_norm": 1.7083444476431595,
+      "learning_rate": 6.248336984313035e-06,
+      "loss": 0.8095,
+      "step": 637
+    },
+    {
+      "epoch": 1.6899070385126163,
+      "grad_norm": 1.5870927607624352,
+      "learning_rate": 6.2381531622321234e-06,
+      "loss": 0.7709,
+      "step": 638
+    },
+    {
+      "epoch": 1.6925630810092962,
+      "grad_norm": 1.6015762357015875,
+      "learning_rate": 6.227963865294444e-06,
+      "loss": 0.8129,
+      "step": 639
+    },
+    {
+      "epoch": 1.6952191235059761,
+      "grad_norm": 1.65706125860874,
+      "learning_rate": 6.2177691385549595e-06,
+      "loss": 0.8378,
+      "step": 640
+    },
+    {
+      "epoch": 1.697875166002656,
+      "grad_norm": 1.604007060138125,
+      "learning_rate": 6.207569027092642e-06,
+      "loss": 0.8319,
+      "step": 641
+    },
+    {
+      "epoch": 1.700531208499336,
+      "grad_norm": 1.6045171117881705,
+      "learning_rate": 6.1973635760102645e-06,
+      "loss": 0.77,
+      "step": 642
+    },
+    {
+      "epoch": 1.703187250996016,
+      "grad_norm": 1.5733954742264373,
+      "learning_rate": 6.18715283043422e-06,
+      "loss": 0.8035,
+      "step": 643
+    },
+    {
+      "epoch": 1.705843293492696,
+      "grad_norm": 1.8117301973869768,
+      "learning_rate": 6.1769368355143125e-06,
+      "loss": 0.8834,
+      "step": 644
+    },
+    {
+      "epoch": 1.7084993359893759,
+      "grad_norm": 1.7588216405278665,
+      "learning_rate": 6.166715636423552e-06,
+      "loss": 0.8235,
+      "step": 645
+    },
+    {
+      "epoch": 1.7111553784860558,
+      "grad_norm": 1.6509712115249005,
+      "learning_rate": 6.156489278357967e-06,
+      "loss": 0.8207,
+      "step": 646
+    },
+    {
+      "epoch": 1.7138114209827358,
+      "grad_norm": 1.7572298275421452,
+      "learning_rate": 6.14625780653639e-06,
+      "loss": 0.7946,
+      "step": 647
+    },
+    {
+      "epoch": 1.7164674634794157,
+      "grad_norm": 1.538832182813809,
+      "learning_rate": 6.136021266200271e-06,
+      "loss": 0.7535,
+      "step": 648
+    },
+    {
+      "epoch": 1.7191235059760956,
+      "grad_norm": 1.8298709899462002,
+      "learning_rate": 6.125779702613471e-06,
+      "loss": 0.8351,
+      "step": 649
+    },
+    {
+      "epoch": 1.7217795484727756,
+      "grad_norm": 1.756099401566392,
+      "learning_rate": 6.115533161062062e-06,
+      "loss": 0.8398,
+      "step": 650
+    },
+    {
+      "epoch": 1.7244355909694555,
+      "grad_norm": 1.7062669920402536,
+      "learning_rate": 6.105281686854129e-06,
+      "loss": 0.816,
+      "step": 651
+    },
+    {
+      "epoch": 1.7270916334661355,
+      "grad_norm": 1.6150556400750276,
+      "learning_rate": 6.0950253253195656e-06,
+      "loss": 0.8606,
+      "step": 652
+    },
+    {
+      "epoch": 1.7297476759628154,
+      "grad_norm": 1.7332370774628512,
+      "learning_rate": 6.084764121809878e-06,
+      "loss": 0.821,
+      "step": 653
+    },
+    {
+      "epoch": 1.7324037184594954,
+      "grad_norm": 2.0279637900537955,
+      "learning_rate": 6.074498121697983e-06,
+      "loss": 0.8049,
+      "step": 654
+    },
+    {
+      "epoch": 1.7350597609561753,
+      "grad_norm": 1.6258653064667319,
+      "learning_rate": 6.064227370378007e-06,
+      "loss": 0.7857,
+      "step": 655
+    },
+    {
+      "epoch": 1.7377158034528553,
+      "grad_norm": 1.6293060217669535,
+      "learning_rate": 6.053951913265083e-06,
+      "loss": 0.8198,
+      "step": 656
+    },
+    {
+      "epoch": 1.7403718459495352,
+      "grad_norm": 1.6077770915178562,
+      "learning_rate": 6.043671795795152e-06,
+      "loss": 0.8127,
+      "step": 657
+    },
+    {
+      "epoch": 1.7430278884462151,
+      "grad_norm": 1.622591828340667,
+      "learning_rate": 6.033387063424765e-06,
+      "loss": 0.7998,
+      "step": 658
+    },
+    {
+      "epoch": 1.745683930942895,
+      "grad_norm": 1.6328037190371787,
+      "learning_rate": 6.023097761630879e-06,
+      "loss": 0.8009,
+      "step": 659
+    },
+    {
+      "epoch": 1.748339973439575,
+      "grad_norm": 1.6308555840386694,
+      "learning_rate": 6.012803935910655e-06,
+      "loss": 0.761,
+      "step": 660
+    },
+    {
+      "epoch": 1.750996015936255,
+      "grad_norm": 1.4971140185122018,
+      "learning_rate": 6.002505631781257e-06,
+      "loss": 0.7743,
+      "step": 661
+    },
+    {
+      "epoch": 1.753652058432935,
+      "grad_norm": 1.5208500640814875,
+      "learning_rate": 5.9922028947796495e-06,
+      "loss": 0.6865,
+      "step": 662
+    },
+    {
+      "epoch": 1.7563081009296149,
+      "grad_norm": 1.6739817122782799,
+      "learning_rate": 5.9818957704624046e-06,
+      "loss": 0.8353,
+      "step": 663
+    },
+    {
+      "epoch": 1.7589641434262948,
+      "grad_norm": 1.6866866619757162,
+      "learning_rate": 5.971584304405489e-06,
+      "loss": 0.8435,
+      "step": 664
+    },
+    {
+      "epoch": 1.7616201859229748,
+      "grad_norm": 1.8112833554197036,
+      "learning_rate": 5.96126854220407e-06,
+      "loss": 0.7812,
+      "step": 665
+    },
+    {
+      "epoch": 1.7642762284196547,
+      "grad_norm": 1.6679412962431888,
+      "learning_rate": 5.95094852947231e-06,
+      "loss": 0.7911,
+      "step": 666
+    },
+    {
+      "epoch": 1.7669322709163346,
+      "grad_norm": 1.6011378182746292,
+      "learning_rate": 5.94062431184317e-06,
+      "loss": 0.7617,
+      "step": 667
+    },
+    {
+      "epoch": 1.7695883134130146,
+      "grad_norm": 1.845248071116995,
+      "learning_rate": 5.930295934968197e-06,
+      "loss": 0.8711,
+      "step": 668
+    },
+    {
+      "epoch": 1.7722443559096945,
+      "grad_norm": 1.725596928754398,
+      "learning_rate": 5.919963444517338e-06,
+      "loss": 0.8052,
+      "step": 669
+    },
+    {
+      "epoch": 1.7749003984063745,
+      "grad_norm": 1.9163181126771034,
+      "learning_rate": 5.909626886178721e-06,
+      "loss": 0.8538,
+      "step": 670
+    },
+    {
+      "epoch": 1.7775564409030544,
+      "grad_norm": 1.7009303732171024,
+      "learning_rate": 5.899286305658468e-06,
+      "loss": 0.8197,
+      "step": 671
+    },
+    {
+      "epoch": 1.7802124833997344,
+      "grad_norm": 1.5247343576304235,
+      "learning_rate": 5.888941748680484e-06,
+      "loss": 0.8159,
+      "step": 672
+    },
+    {
+      "epoch": 1.7828685258964143,
+      "grad_norm": 2.044707850741483,
+      "learning_rate": 5.878593260986256e-06,
+      "loss": 0.8172,
+      "step": 673
+    },
+    {
+      "epoch": 1.7855245683930943,
+      "grad_norm": 1.7418155389848087,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.8196,
+      "step": 674
+    },
+    {
+      "epoch": 1.7881806108897742,
+      "grad_norm": 1.4597724863288657,
+      "learning_rate": 5.857884676501721e-06,
+      "loss": 0.7647,
+      "step": 675
+    },
+    {
+      "epoch": 1.7908366533864541,
+      "grad_norm": 1.6458660814062083,
+      "learning_rate": 5.8475246712804845e-06,
+      "loss": 0.8045,
+      "step": 676
+    },
+    {
+      "epoch": 1.793492695883134,
+      "grad_norm": 1.6252499093239745,
+      "learning_rate": 5.83716091848074e-06,
+      "loss": 0.8091,
+      "step": 677
+    },
+    {
+      "epoch": 1.796148738379814,
+      "grad_norm": 1.6610545996985242,
+      "learning_rate": 5.8267934639288525e-06,
+      "loss": 0.8809,
+      "step": 678
+    },
+    {
+      "epoch": 1.798804780876494,
+      "grad_norm": 1.681945309800806,
+      "learning_rate": 5.816422353467562e-06,
+      "loss": 0.7976,
+      "step": 679
+    },
+    {
+      "epoch": 1.801460823373174,
+      "grad_norm": 1.7407405789509727,
+      "learning_rate": 5.80604763295577e-06,
+      "loss": 0.7653,
+      "step": 680
+    },
+    {
+      "epoch": 1.8041168658698539,
+      "grad_norm": 1.7577038472175832,
+      "learning_rate": 5.795669348268339e-06,
+      "loss": 0.7965,
+      "step": 681
+    },
+    {
+      "epoch": 1.8067729083665338,
+      "grad_norm": 1.7731919678676291,
+      "learning_rate": 5.785287545295895e-06,
+      "loss": 0.7594,
+      "step": 682
+    },
+    {
+      "epoch": 1.8094289508632138,
+      "grad_norm": 1.6141949293008162,
+      "learning_rate": 5.77490226994462e-06,
+      "loss": 0.801,
+      "step": 683
+    },
+    {
+      "epoch": 1.8120849933598937,
+      "grad_norm": 1.812626295914488,
+      "learning_rate": 5.76451356813605e-06,
+      "loss": 0.8163,
+      "step": 684
+    },
+    {
+      "epoch": 1.8147410358565739,
+      "grad_norm": 1.6162681395964988,
+      "learning_rate": 5.7541214858068705e-06,
+      "loss": 0.8397,
+      "step": 685
+    },
+    {
+      "epoch": 1.8173970783532538,
+      "grad_norm": 1.721558869882608,
+      "learning_rate": 5.743726068908717e-06,
+      "loss": 0.8043,
+      "step": 686
+    },
+    {
+      "epoch": 1.8200531208499338,
+      "grad_norm": 1.8032721689703735,
+      "learning_rate": 5.733327363407973e-06,
+      "loss": 0.7866,
+      "step": 687
+    },
+    {
+      "epoch": 1.8227091633466137,
+      "grad_norm": 1.6708160503371885,
+      "learning_rate": 5.722925415285555e-06,
+      "loss": 0.7858,
+      "step": 688
+    },
+    {
+      "epoch": 1.8253652058432936,
+      "grad_norm": 1.766193995933153,
+      "learning_rate": 5.712520270536723e-06,
+      "loss": 0.7798,
+      "step": 689
+    },
+    {
+      "epoch": 1.8280212483399736,
+      "grad_norm": 1.5356744931510473,
+      "learning_rate": 5.702111975170875e-06,
+      "loss": 0.7936,
+      "step": 690
+    },
+    {
+      "epoch": 1.8306772908366535,
+      "grad_norm": 1.732102227757507,
+      "learning_rate": 5.691700575211335e-06,
+      "loss": 0.788,
+      "step": 691
+    },
+    {
+      "epoch": 1.8333333333333335,
+      "grad_norm": 1.6592997418979596,
+      "learning_rate": 5.681286116695155e-06,
+      "loss": 0.8294,
+      "step": 692
+    },
+    {
+      "epoch": 1.8359893758300134,
+      "grad_norm": 1.9062250490573815,
+      "learning_rate": 5.670868645672916e-06,
+      "loss": 0.8372,
+      "step": 693
+    },
+    {
+      "epoch": 1.8386454183266934,
+      "grad_norm": 1.7172232578747761,
+      "learning_rate": 5.660448208208513e-06,
+      "loss": 0.8292,
+      "step": 694
+    },
+    {
+      "epoch": 1.8413014608233733,
+      "grad_norm": 1.8188926512319041,
+      "learning_rate": 5.650024850378964e-06,
+      "loss": 0.8221,
+      "step": 695
+    },
+    {
+      "epoch": 1.8439575033200533,
+      "grad_norm": 1.638642817866283,
+      "learning_rate": 5.6395986182741965e-06,
+      "loss": 0.8157,
+      "step": 696
+    },
+    {
+      "epoch": 1.8466135458167332,
+      "grad_norm": 1.617796881961262,
+      "learning_rate": 5.629169557996848e-06,
+      "loss": 0.7954,
+      "step": 697
+    },
+    {
+      "epoch": 1.8492695883134131,
+      "grad_norm": 1.7846585636780499,
+      "learning_rate": 5.618737715662067e-06,
+      "loss": 0.8045,
+      "step": 698
+    },
+    {
+      "epoch": 1.851925630810093,
+      "grad_norm": 1.52955631208176,
+      "learning_rate": 5.608303137397294e-06,
+      "loss": 0.7532,
+      "step": 699
+    },
+    {
+      "epoch": 1.854581673306773,
+      "grad_norm": 1.6372599632527154,
+      "learning_rate": 5.597865869342075e-06,
+      "loss": 0.7908,
+      "step": 700
+    },
+    {
+      "epoch": 1.857237715803453,
+      "grad_norm": 1.557033830346597,
+      "learning_rate": 5.5874259576478465e-06,
+      "loss": 0.8461,
+      "step": 701
+    },
+    {
+      "epoch": 1.859893758300133,
+      "grad_norm": 2.0636224479100647,
+      "learning_rate": 5.5769834484777344e-06,
+      "loss": 0.794,
+      "step": 702
+    },
+    {
+      "epoch": 1.8625498007968129,
+      "grad_norm": 1.5906609489633872,
+      "learning_rate": 5.566538388006351e-06,
+      "loss": 0.7641,
+      "step": 703
+    },
+    {
+      "epoch": 1.8652058432934928,
+      "grad_norm": 1.7691225304057758,
+      "learning_rate": 5.556090822419589e-06,
+      "loss": 0.8186,
+      "step": 704
+    },
+    {
+      "epoch": 1.8678618857901728,
+      "grad_norm": 1.6618983785480637,
+      "learning_rate": 5.54564079791442e-06,
+      "loss": 0.7933,
+      "step": 705
+    },
+    {
+      "epoch": 1.8705179282868527,
+      "grad_norm": 1.6226756802008044,
+      "learning_rate": 5.535188360698687e-06,
+      "loss": 0.8021,
+      "step": 706
+    },
+    {
+      "epoch": 1.8731739707835326,
+      "grad_norm": 2.178124602174882,
+      "learning_rate": 5.524733556990904e-06,
+      "loss": 0.8244,
+      "step": 707
+    },
+    {
+      "epoch": 1.8758300132802126,
+      "grad_norm": 1.7670614639798954,
+      "learning_rate": 5.514276433020044e-06,
+      "loss": 0.6896,
+      "step": 708
+    },
+    {
+      "epoch": 1.8784860557768925,
+      "grad_norm": 1.6798674141158796,
+      "learning_rate": 5.503817035025341e-06,
+      "loss": 0.8297,
+      "step": 709
+    },
+    {
+      "epoch": 1.8811420982735725,
+      "grad_norm": 1.6026510714619167,
+      "learning_rate": 5.493355409256091e-06,
+      "loss": 0.8391,
+      "step": 710
+    },
+    {
+      "epoch": 1.8837981407702524,
+      "grad_norm": 1.8164233494348743,
+      "learning_rate": 5.482891601971434e-06,
+      "loss": 0.8216,
+      "step": 711
+    },
+    {
+      "epoch": 1.8864541832669324,
+      "grad_norm": 2.225469822194458,
+      "learning_rate": 5.472425659440157e-06,
+      "loss": 0.838,
+      "step": 712
+    },
+    {
+      "epoch": 1.8891102257636123,
+      "grad_norm": 1.6986504060233345,
+      "learning_rate": 5.461957627940489e-06,
+      "loss": 0.7715,
+      "step": 713
+    },
+    {
+      "epoch": 1.8917662682602923,
+      "grad_norm": 1.853182152455261,
+      "learning_rate": 5.451487553759899e-06,
+      "loss": 0.7993,
+      "step": 714
+    },
+    {
+      "epoch": 1.8944223107569722,
+      "grad_norm": 1.6193780421833341,
+      "learning_rate": 5.441015483194883e-06,
+      "loss": 0.7837,
+      "step": 715
+    },
+    {
+      "epoch": 1.8970783532536521,
+      "grad_norm": 1.65935355108129,
+      "learning_rate": 5.43054146255077e-06,
+      "loss": 0.7968,
+      "step": 716
+    },
+    {
+      "epoch": 1.899734395750332,
+      "grad_norm": 1.7128452889705919,
+      "learning_rate": 5.420065538141507e-06,
+      "loss": 0.8091,
+      "step": 717
+    },
+    {
+      "epoch": 1.902390438247012,
+      "grad_norm": 1.7131843789521455,
+      "learning_rate": 5.409587756289462e-06,
+      "loss": 0.7841,
+      "step": 718
+    },
+    {
+      "epoch": 1.905046480743692,
+      "grad_norm": 1.8326844843438326,
+      "learning_rate": 5.399108163325217e-06,
+      "loss": 0.7998,
+      "step": 719
+    },
+    {
+      "epoch": 1.907702523240372,
+      "grad_norm": 1.5677093021811603,
+      "learning_rate": 5.388626805587361e-06,
+      "loss": 0.7657,
+      "step": 720
+    },
+    {
+      "epoch": 1.9103585657370519,
+      "grad_norm": 1.6608756530604105,
+      "learning_rate": 5.378143729422285e-06,
+      "loss": 0.8002,
+      "step": 721
+    },
+    {
+      "epoch": 1.9130146082337318,
+      "grad_norm": 1.5762455616412385,
+      "learning_rate": 5.367658981183979e-06,
+      "loss": 0.7799,
+      "step": 722
+    },
+    {
+      "epoch": 1.9156706507304118,
+      "grad_norm": 1.6223542306458065,
+      "learning_rate": 5.357172607233831e-06,
+      "loss": 0.7568,
+      "step": 723
+    },
+    {
+      "epoch": 1.9183266932270917,
+      "grad_norm": 1.6243360897116195,
+      "learning_rate": 5.346684653940408e-06,
+      "loss": 0.8361,
+      "step": 724
+    },
+    {
+      "epoch": 1.9209827357237717,
+      "grad_norm": 1.717966580725845,
+      "learning_rate": 5.3361951676792745e-06,
+      "loss": 0.8181,
+      "step": 725
+    },
+    {
+      "epoch": 1.9236387782204516,
+      "grad_norm": 1.6409767034053258,
+      "learning_rate": 5.325704194832759e-06,
+      "loss": 0.7991,
+      "step": 726
+    },
+    {
+      "epoch": 1.9262948207171315,
+      "grad_norm": 1.7093634730103173,
+      "learning_rate": 5.315211781789775e-06,
+      "loss": 0.7401,
+      "step": 727
+    },
+    {
+      "epoch": 1.9289508632138115,
+      "grad_norm": 1.8814933118849915,
+      "learning_rate": 5.304717974945596e-06,
+      "loss": 0.8212,
+      "step": 728
+    },
+    {
+      "epoch": 1.9316069057104914,
+      "grad_norm": 1.6022448134502887,
+      "learning_rate": 5.294222820701661e-06,
+      "loss": 0.7712,
+      "step": 729
+    },
+    {
+      "epoch": 1.9342629482071714,
+      "grad_norm": 1.9257961733760494,
+      "learning_rate": 5.2837263654653715e-06,
+      "loss": 0.7694,
+      "step": 730
+    },
+    {
+      "epoch": 1.9369189907038513,
+      "grad_norm": 1.55634864513684,
+      "learning_rate": 5.273228655649873e-06,
+      "loss": 0.7937,
+      "step": 731
+    },
+    {
+      "epoch": 1.9395750332005313,
+      "grad_norm": 1.65857937432083,
+      "learning_rate": 5.2627297376738674e-06,
+      "loss": 0.7309,
+      "step": 732
+    },
+    {
+      "epoch": 1.9422310756972112,
+      "grad_norm": 1.6472003637687052,
+      "learning_rate": 5.252229657961394e-06,
+      "loss": 0.8135,
+      "step": 733
+    },
+    {
+      "epoch": 1.9448871181938912,
+      "grad_norm": 1.6658259885988707,
+      "learning_rate": 5.24172846294163e-06,
+      "loss": 0.7817,
+      "step": 734
+    },
+    {
+      "epoch": 1.947543160690571,
+      "grad_norm": 1.7271975202258703,
+      "learning_rate": 5.231226199048682e-06,
+      "loss": 0.7704,
+      "step": 735
+    },
+    {
+      "epoch": 1.950199203187251,
+      "grad_norm": 1.5236007987102846,
+      "learning_rate": 5.2207229127213866e-06,
+      "loss": 0.8125,
+      "step": 736
+    },
+    {
+      "epoch": 1.952855245683931,
+      "grad_norm": 1.5827626707270386,
+      "learning_rate": 5.210218650403101e-06,
+      "loss": 0.8218,
+      "step": 737
+    },
+    {
+      "epoch": 1.955511288180611,
+      "grad_norm": 1.6150367316693914,
+      "learning_rate": 5.199713458541495e-06,
+      "loss": 0.7933,
+      "step": 738
+    },
+    {
+      "epoch": 1.9581673306772909,
+      "grad_norm": 1.6711992720608577,
+      "learning_rate": 5.189207383588353e-06,
+      "loss": 0.8075,
+      "step": 739
+    },
+    {
+      "epoch": 1.9608233731739708,
+      "grad_norm": 1.6593919600555247,
+      "learning_rate": 5.178700471999357e-06,
+      "loss": 0.7473,
+      "step": 740
+    },
+    {
+      "epoch": 1.9634794156706508,
+      "grad_norm": 1.7172984617619111,
+      "learning_rate": 5.168192770233901e-06,
+      "loss": 0.7953,
+      "step": 741
+    },
+    {
+      "epoch": 1.9661354581673307,
+      "grad_norm": 1.7140610596606283,
+      "learning_rate": 5.157684324754858e-06,
+      "loss": 0.883,
+      "step": 742
+    },
+    {
+      "epoch": 1.9687915006640107,
+      "grad_norm": 1.564534131388113,
+      "learning_rate": 5.1471751820284e-06,
+      "loss": 0.7756,
+      "step": 743
+    },
+    {
+      "epoch": 1.9714475431606906,
+      "grad_norm": 1.7708506721058221,
+      "learning_rate": 5.136665388523779e-06,
+      "loss": 0.7802,
+      "step": 744
+    },
+    {
+      "epoch": 1.9741035856573705,
+      "grad_norm": 1.7847297514234355,
+      "learning_rate": 5.126154990713123e-06,
+      "loss": 0.794,
+      "step": 745
+    },
+    {
+      "epoch": 1.9767596281540505,
+      "grad_norm": 1.6540427509570816,
+      "learning_rate": 5.115644035071234e-06,
+      "loss": 0.803,
+      "step": 746
+    },
+    {
+      "epoch": 1.9794156706507304,
+      "grad_norm": 1.5823166162412527,
+      "learning_rate": 5.1051325680753826e-06,
+      "loss": 0.6617,
+      "step": 747
+    },
+    {
+      "epoch": 1.9820717131474104,
+      "grad_norm": 1.6625490997353582,
+      "learning_rate": 5.094620636205096e-06,
+      "loss": 0.7835,
+      "step": 748
+    },
+    {
+      "epoch": 1.9847277556440903,
+      "grad_norm": 1.6216871637425778,
+      "learning_rate": 5.084108285941959e-06,
+      "loss": 0.816,
+      "step": 749
+    },
+    {
+      "epoch": 1.9873837981407703,
+      "grad_norm": 1.6590146206027645,
+      "learning_rate": 5.073595563769407e-06,
+      "loss": 0.8446,
+      "step": 750
+    },
+    {
+      "epoch": 1.9900398406374502,
+      "grad_norm": 1.614326585836819,
+      "learning_rate": 5.06308251617252e-06,
+      "loss": 0.7208,
+      "step": 751
+    },
+    {
+      "epoch": 1.9926958831341302,
+      "grad_norm": 1.6342275277779148,
+      "learning_rate": 5.052569189637813e-06,
+      "loss": 0.7641,
+      "step": 752
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1504,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 376,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 716175260516352.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}