diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,3886 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1098,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00546448087431694,
+      "grad_norm": 0.7586923837661743,
+      "learning_rate": 3.6363636363636366e-07,
+      "loss": 2.660745620727539,
+      "step": 2
+    },
+    {
+      "epoch": 0.01092896174863388,
+      "grad_norm": 1.0751519203186035,
+      "learning_rate": 1.090909090909091e-06,
+      "loss": 1.938146948814392,
+      "step": 4
+    },
+    {
+      "epoch": 0.01639344262295082,
+      "grad_norm": 0.8923681974411011,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 1.9057133197784424,
+      "step": 6
+    },
+    {
+      "epoch": 0.02185792349726776,
+      "grad_norm": 0.7297345399856567,
+      "learning_rate": 2.5454545454545456e-06,
+      "loss": 1.7945566177368164,
+      "step": 8
+    },
+    {
+      "epoch": 0.0273224043715847,
+      "grad_norm": 0.818263828754425,
+      "learning_rate": 3.272727272727273e-06,
+      "loss": 1.770993709564209,
+      "step": 10
+    },
+    {
+      "epoch": 0.03278688524590164,
+      "grad_norm": 0.4792977273464203,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.62543523311615,
+      "step": 12
+    },
+    {
+      "epoch": 0.03825136612021858,
+      "grad_norm": 0.4317499101161957,
+      "learning_rate": 4.727272727272728e-06,
+      "loss": 1.519499659538269,
+      "step": 14
+    },
+    {
+      "epoch": 0.04371584699453552,
+      "grad_norm": 0.3740278482437134,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 1.2777570486068726,
+      "step": 16
+    },
+    {
+      "epoch": 0.04918032786885246,
+      "grad_norm": 0.652406632900238,
+      "learning_rate": 6.181818181818182e-06,
+      "loss": 1.1676603555679321,
+      "step": 18
+    },
+    {
+      "epoch": 0.0546448087431694,
+      "grad_norm": 0.3864721953868866,
+      "learning_rate": 6.90909090909091e-06,
+      "loss": 1.4224069118499756,
+      "step": 20
+    },
+    {
+      "epoch": 0.060109289617486336,
+      "grad_norm": 0.541534960269928,
+      "learning_rate": 7.636363636363638e-06,
+      "loss": 1.3205312490463257,
+      "step": 22
+    },
+    {
+      "epoch": 0.06557377049180328,
+      "grad_norm": 0.44199395179748535,
+      "learning_rate": 8.363636363636365e-06,
+      "loss": 1.4163978099822998,
+      "step": 24
+    },
+    {
+      "epoch": 0.07103825136612021,
+      "grad_norm": 0.8410191535949707,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 1.598981261253357,
+      "step": 26
+    },
+    {
+      "epoch": 0.07650273224043716,
+      "grad_norm": 1.757845163345337,
+      "learning_rate": 9.81818181818182e-06,
+      "loss": 1.375367283821106,
+      "step": 28
+    },
+    {
+      "epoch": 0.08196721311475409,
+      "grad_norm": 0.46648648381233215,
+      "learning_rate": 1.0545454545454546e-05,
+      "loss": 1.2981315851211548,
+      "step": 30
+    },
+    {
+      "epoch": 0.08743169398907104,
+      "grad_norm": 0.24247297644615173,
+      "learning_rate": 1.1272727272727272e-05,
+      "loss": 1.2934328317642212,
+      "step": 32
+    },
+    {
+      "epoch": 0.09289617486338798,
+      "grad_norm": 0.7804842591285706,
+      "learning_rate": 1.2e-05,
+      "loss": 1.0114089250564575,
+      "step": 34
+    },
+    {
+      "epoch": 0.09836065573770492,
+      "grad_norm": 0.24668505787849426,
+      "learning_rate": 1.2727272727272728e-05,
+      "loss": 1.1152716875076294,
+      "step": 36
+    },
+    {
+      "epoch": 0.10382513661202186,
+      "grad_norm": 0.3483196198940277,
+      "learning_rate": 1.3454545454545455e-05,
+      "loss": 1.3653188943862915,
+      "step": 38
+    },
+    {
+      "epoch": 0.1092896174863388,
+      "grad_norm": 0.5274494886398315,
+      "learning_rate": 1.4181818181818183e-05,
+      "loss": 1.316364049911499,
+      "step": 40
+    },
+    {
+      "epoch": 0.11475409836065574,
+      "grad_norm": 1.7519822120666504,
+      "learning_rate": 1.4909090909090911e-05,
+      "loss": 1.0375714302062988,
+      "step": 42
+    },
+    {
+      "epoch": 0.12021857923497267,
+      "grad_norm": 0.2554769814014435,
+      "learning_rate": 1.563636363636364e-05,
+      "loss": 0.9050353169441223,
+      "step": 44
+    },
+    {
+      "epoch": 0.12568306010928962,
+      "grad_norm": 0.2430635243654251,
+      "learning_rate": 1.6363636363636366e-05,
+      "loss": 1.3280794620513916,
+      "step": 46
+    },
+    {
+      "epoch": 0.13114754098360656,
+      "grad_norm": 0.3220200836658478,
+      "learning_rate": 1.7090909090909092e-05,
+      "loss": 1.3129568099975586,
+      "step": 48
+    },
+    {
+      "epoch": 0.1366120218579235,
+      "grad_norm": 0.3611713647842407,
+      "learning_rate": 1.781818181818182e-05,
+      "loss": 1.2813783884048462,
+      "step": 50
+    },
+    {
+      "epoch": 0.14207650273224043,
+      "grad_norm": 0.3371615707874298,
+      "learning_rate": 1.8545454545454545e-05,
+      "loss": 1.2453889846801758,
+      "step": 52
+    },
+    {
+      "epoch": 0.14754098360655737,
+      "grad_norm": 3.5102477073669434,
+      "learning_rate": 1.9272727272727275e-05,
+      "loss": 0.9431768655776978,
+      "step": 54
+    },
+    {
+      "epoch": 0.15300546448087432,
+      "grad_norm": 0.3596407175064087,
+      "learning_rate": 2e-05,
+      "loss": 1.4184821844100952,
+      "step": 56
+    },
+    {
+      "epoch": 0.15846994535519127,
+      "grad_norm": 0.6473796367645264,
+      "learning_rate": 1.9998327792599505e-05,
+      "loss": 1.4325121641159058,
+      "step": 58
+    },
+    {
+      "epoch": 0.16393442622950818,
+      "grad_norm": 0.4482963979244232,
+      "learning_rate": 1.999331179179304e-05,
+      "loss": 1.3262360095977783,
+      "step": 60
+    },
+    {
+      "epoch": 0.16939890710382513,
+      "grad_norm": 0.45529159903526306,
+      "learning_rate": 1.9984953861534752e-05,
+      "loss": 1.505706787109375,
+      "step": 62
+    },
+    {
+      "epoch": 0.17486338797814208,
+      "grad_norm": 1.139926791191101,
+      "learning_rate": 1.997325710764527e-05,
+      "loss": 1.1403143405914307,
+      "step": 64
+    },
+    {
+      "epoch": 0.18032786885245902,
+      "grad_norm": 0.5250591039657593,
+      "learning_rate": 1.9958225876657575e-05,
+      "loss": 1.30501127243042,
+      "step": 66
+    },
+    {
+      "epoch": 0.18579234972677597,
+      "grad_norm": 1.2981220483779907,
+      "learning_rate": 1.9939865754201825e-05,
+      "loss": 1.2804065942764282,
+      "step": 68
+    },
+    {
+      "epoch": 0.1912568306010929,
+      "grad_norm": 0.49463051557540894,
+      "learning_rate": 1.9918183562929717e-05,
+      "loss": 1.309598445892334,
+      "step": 70
+    },
+    {
+      "epoch": 0.19672131147540983,
+      "grad_norm": 0.5240225195884705,
+      "learning_rate": 1.9893187359979183e-05,
+      "loss": 1.0792406797409058,
+      "step": 72
+    },
+    {
+      "epoch": 0.20218579234972678,
+      "grad_norm": 0.3060704469680786,
+      "learning_rate": 1.986488643398035e-05,
+      "loss": 1.3067054748535156,
+      "step": 74
+    },
+    {
+      "epoch": 0.20765027322404372,
+      "grad_norm": 0.6043591499328613,
+      "learning_rate": 1.9833291301603863e-05,
+      "loss": 1.307369589805603,
+      "step": 76
+    },
+    {
+      "epoch": 0.21311475409836064,
+      "grad_norm": 1.3378225564956665,
+      "learning_rate": 1.9798413703652867e-05,
+      "loss": 0.5469598174095154,
+      "step": 78
+    },
+    {
+      "epoch": 0.2185792349726776,
+      "grad_norm": 0.36606472730636597,
+      "learning_rate": 1.976026660070012e-05,
+      "loss": 0.5283974409103394,
+      "step": 80
+    },
+    {
+      "epoch": 0.22404371584699453,
+      "grad_norm": 0.4779462218284607,
+      "learning_rate": 1.9718864168271823e-05,
+      "loss": 1.3443689346313477,
+      "step": 82
+    },
+    {
+      "epoch": 0.22950819672131148,
+      "grad_norm": 0.3752780854701996,
+      "learning_rate": 1.9674221791579946e-05,
+      "loss": 0.729979395866394,
+      "step": 84
+    },
+    {
+      "epoch": 0.23497267759562843,
+      "grad_norm": 0.23232732713222504,
+      "learning_rate": 1.9626356059805085e-05,
+      "loss": 1.2514656782150269,
+      "step": 86
+    },
+    {
+      "epoch": 0.24043715846994534,
+      "grad_norm": 0.37319672107696533,
+      "learning_rate": 1.957528475993189e-05,
+      "loss": 1.2355018854141235,
+      "step": 88
+    },
+    {
+      "epoch": 0.2459016393442623,
+      "grad_norm": 1.1993027925491333,
+      "learning_rate": 1.952102687013938e-05,
+      "loss": 0.86878901720047,
+      "step": 90
+    },
+    {
+      "epoch": 0.25136612021857924,
+      "grad_norm": 0.5708634257316589,
+      "learning_rate": 1.946360255274863e-05,
+      "loss": 1.2946072816848755,
+      "step": 92
+    },
+    {
+      "epoch": 0.2568306010928962,
+      "grad_norm": 0.2725592851638794,
+      "learning_rate": 1.9403033146730424e-05,
+      "loss": 1.1185024976730347,
+      "step": 94
+    },
+    {
+      "epoch": 0.26229508196721313,
+      "grad_norm": 0.6162624955177307,
+      "learning_rate": 1.9339341159775647e-05,
+      "loss": 1.5296945571899414,
+      "step": 96
+    },
+    {
+      "epoch": 0.2677595628415301,
+      "grad_norm": 0.30999237298965454,
+      "learning_rate": 1.9272550259931398e-05,
+      "loss": 0.9915840029716492,
+      "step": 98
+    },
+    {
+      "epoch": 0.273224043715847,
+      "grad_norm": 0.27694171667099,
+      "learning_rate": 1.9202685266805896e-05,
+      "loss": 1.0551968812942505,
+      "step": 100
+    },
+    {
+      "epoch": 0.2786885245901639,
+      "grad_norm": 0.34108948707580566,
+      "learning_rate": 1.9129772142345484e-05,
+      "loss": 1.253743290901184,
+      "step": 102
+    },
+    {
+      "epoch": 0.28415300546448086,
+      "grad_norm": 0.3947954773902893,
+      "learning_rate": 1.9053837981187125e-05,
+      "loss": 0.9518170952796936,
+      "step": 104
+    },
+    {
+      "epoch": 0.2896174863387978,
+      "grad_norm": 0.3549712896347046,
+      "learning_rate": 1.897491100058998e-05,
+      "loss": 1.2458157539367676,
+      "step": 106
+    },
+    {
+      "epoch": 0.29508196721311475,
+      "grad_norm": 0.30054518580436707,
+      "learning_rate": 1.8893020529949838e-05,
+      "loss": 1.386643648147583,
+      "step": 108
+    },
+    {
+      "epoch": 0.3005464480874317,
+      "grad_norm": 1.3760162591934204,
+      "learning_rate": 1.880819699990027e-05,
+      "loss": 1.2258188724517822,
+      "step": 110
+    },
+    {
+      "epoch": 0.30601092896174864,
+      "grad_norm": 0.2765187621116638,
+      "learning_rate": 1.8720471931004526e-05,
+      "loss": 1.2596882581710815,
+      "step": 112
+    },
+    {
+      "epoch": 0.3114754098360656,
+      "grad_norm": 0.6589577794075012,
+      "learning_rate": 1.8629877922042485e-05,
+      "loss": 1.3348774909973145,
+      "step": 114
+    },
+    {
+      "epoch": 0.31693989071038253,
+      "grad_norm": 0.3075348436832428,
+      "learning_rate": 1.8536448637896866e-05,
+      "loss": 1.2452787160873413,
+      "step": 116
+    },
+    {
+      "epoch": 0.3224043715846995,
+      "grad_norm": 0.3071703016757965,
+      "learning_rate": 1.84402187970433e-05,
+      "loss": 1.0434746742248535,
+      "step": 118
+    },
+    {
+      "epoch": 0.32786885245901637,
+      "grad_norm": 2.645376682281494,
+      "learning_rate": 1.834122415864891e-05,
+      "loss": 1.2307820320129395,
+      "step": 120
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.5483799576759338,
+      "learning_rate": 1.8239501509284123e-05,
+      "loss": 1.3374857902526855,
+      "step": 122
+    },
+    {
+      "epoch": 0.33879781420765026,
+      "grad_norm": 0.3543969988822937,
+      "learning_rate": 1.8135088649252725e-05,
+      "loss": 1.030819296836853,
+      "step": 124
+    },
+    {
+      "epoch": 0.3442622950819672,
+      "grad_norm": 0.5614028573036194,
+      "learning_rate": 1.8028024378545224e-05,
+      "loss": 1.0396068096160889,
+      "step": 126
+    },
+    {
+      "epoch": 0.34972677595628415,
+      "grad_norm": 0.4731024503707886,
+      "learning_rate": 1.7918348482420692e-05,
+      "loss": 1.3047938346862793,
+      "step": 128
+    },
+    {
+      "epoch": 0.3551912568306011,
+      "grad_norm": 0.42884743213653564,
+      "learning_rate": 1.7806101716622486e-05,
+      "loss": 1.2398240566253662,
+      "step": 130
+    },
+    {
+      "epoch": 0.36065573770491804,
+      "grad_norm": 0.32270482182502747,
+      "learning_rate": 1.7691325792233378e-05,
+      "loss": 1.1840364933013916,
+      "step": 132
+    },
+    {
+      "epoch": 0.366120218579235,
+      "grad_norm": 0.8914838433265686,
+      "learning_rate": 1.7574063360175625e-05,
+      "loss": 1.0900051593780518,
+      "step": 134
+    },
+    {
+      "epoch": 0.37158469945355194,
+      "grad_norm": 0.4666799306869507,
+      "learning_rate": 1.745435799536183e-05,
+      "loss": 0.976508378982544,
+      "step": 136
+    },
+    {
+      "epoch": 0.3770491803278688,
+      "grad_norm": 0.3514363467693329,
+      "learning_rate": 1.7332254180502407e-05,
+      "loss": 0.9537405967712402,
+      "step": 138
+    },
+    {
+      "epoch": 0.3825136612021858,
+      "grad_norm": 0.22689107060432434,
+      "learning_rate": 1.7207797289575777e-05,
+      "loss": 0.6181514859199524,
+      "step": 140
+    },
+    {
+      "epoch": 0.3879781420765027,
+      "grad_norm": 0.337587833404541,
+      "learning_rate": 1.708103357096728e-05,
+      "loss": 1.2152214050292969,
+      "step": 142
+    },
+    {
+      "epoch": 0.39344262295081966,
+      "grad_norm": 0.34758704900741577,
+      "learning_rate": 1.695201013028322e-05,
+      "loss": 1.2426707744598389,
+      "step": 144
+    },
+    {
+      "epoch": 0.3989071038251366,
+      "grad_norm": 0.2902460992336273,
+      "learning_rate": 1.6820774912846335e-05,
+      "loss": 1.2040138244628906,
+      "step": 146
+    },
+    {
+      "epoch": 0.40437158469945356,
+      "grad_norm": 0.3400241434574127,
+      "learning_rate": 1.668737668587926e-05,
+      "loss": 1.1996750831604004,
+      "step": 148
+    },
+    {
+      "epoch": 0.4098360655737705,
+      "grad_norm": 0.35784855484962463,
+      "learning_rate": 1.655186502038251e-05,
+      "loss": 1.1286766529083252,
+      "step": 150
+    },
+    {
+      "epoch": 0.41530054644808745,
+      "grad_norm": 0.2693675458431244,
+      "learning_rate": 1.641429027271384e-05,
+      "loss": 1.2409323453903198,
+      "step": 152
+    },
+    {
+      "epoch": 0.4207650273224044,
+      "grad_norm": 0.483333557844162,
+      "learning_rate": 1.6274703565875736e-05,
+      "loss": 1.2143176794052124,
+      "step": 154
+    },
+    {
+      "epoch": 0.4262295081967213,
+      "grad_norm": 0.4158734679222107,
+      "learning_rate": 1.613315677051801e-05,
+      "loss": 1.2641037702560425,
+      "step": 156
+    },
+    {
+      "epoch": 0.43169398907103823,
+      "grad_norm": 0.6228234171867371,
+      "learning_rate": 1.598970248566261e-05,
+      "loss": 0.9883160591125488,
+      "step": 158
+    },
+    {
+      "epoch": 0.4371584699453552,
+      "grad_norm": 0.3108808398246765,
+      "learning_rate": 1.5844394019157697e-05,
+      "loss": 1.1966590881347656,
+      "step": 160
+    },
+    {
+      "epoch": 0.4426229508196721,
+      "grad_norm": 0.4892992377281189,
+      "learning_rate": 1.5697285367868393e-05,
+      "loss": 0.6886841058731079,
+      "step": 162
+    },
+    {
+      "epoch": 0.44808743169398907,
+      "grad_norm": 0.3893924653530121,
+      "learning_rate": 1.5548431197611448e-05,
+      "loss": 1.5789108276367188,
+      "step": 164
+    },
+    {
+      "epoch": 0.453551912568306,
+      "grad_norm": 0.6205107569694519,
+      "learning_rate": 1.539788682284133e-05,
+      "loss": 0.724977970123291,
+      "step": 166
+    },
+    {
+      "epoch": 0.45901639344262296,
+      "grad_norm": 0.2539767324924469,
+      "learning_rate": 1.5245708186095275e-05,
+      "loss": 1.1216144561767578,
+      "step": 168
+    },
+    {
+      "epoch": 0.4644808743169399,
+      "grad_norm": 0.31149035692214966,
+      "learning_rate": 1.5091951837204973e-05,
+      "loss": 1.2175238132476807,
+      "step": 170
+    },
+    {
+      "epoch": 0.46994535519125685,
+      "grad_norm": 0.5983729958534241,
+      "learning_rate": 1.4936674912282525e-05,
+      "loss": 1.292568325996399,
+      "step": 172
+    },
+    {
+      "epoch": 0.47540983606557374,
+      "grad_norm": 0.41234031319618225,
+      "learning_rate": 1.4779935112488597e-05,
+      "loss": 1.404640793800354,
+      "step": 174
+    },
+    {
+      "epoch": 0.4808743169398907,
+      "grad_norm": 0.25759682059288025,
+      "learning_rate": 1.4621790682590556e-05,
+      "loss": 1.200560450553894,
+      "step": 176
+    },
+    {
+      "epoch": 0.48633879781420764,
+      "grad_norm": 0.8127933144569397,
+      "learning_rate": 1.4462300389318635e-05,
+      "loss": 1.161008358001709,
+      "step": 178
+    },
+    {
+      "epoch": 0.4918032786885246,
+      "grad_norm": 0.2617407739162445,
+      "learning_rate": 1.4301523499528099e-05,
+      "loss": 1.167149543762207,
+      "step": 180
+    },
+    {
+      "epoch": 0.4972677595628415,
+      "grad_norm": 0.38304898142814636,
+      "learning_rate": 1.4139519758175602e-05,
+      "loss": 1.154693841934204,
+      "step": 182
+    },
+    {
+      "epoch": 0.5027322404371585,
+      "grad_norm": 1.8618555068969727,
+      "learning_rate": 1.3976349366117861e-05,
+      "loss": 1.2942745685577393,
+      "step": 184
+    },
+    {
+      "epoch": 0.5081967213114754,
+      "grad_norm": 0.7506605386734009,
+      "learning_rate": 1.3812072957740898e-05,
+      "loss": 1.165328860282898,
+      "step": 186
+    },
+    {
+      "epoch": 0.5136612021857924,
+      "grad_norm": 0.716810405254364,
+      "learning_rate": 1.3646751578428231e-05,
+      "loss": 1.1603736877441406,
+      "step": 188
+    },
+    {
+      "epoch": 0.5191256830601093,
+      "grad_norm": 0.2964042127132416,
+      "learning_rate": 1.3480446661876295e-05,
+      "loss": 1.2188032865524292,
+      "step": 190
+    },
+    {
+      "epoch": 0.5245901639344263,
+      "grad_norm": 0.5146150588989258,
+      "learning_rate": 1.3313220007265572e-05,
+      "loss": 1.1854382753372192,
+      "step": 192
+    },
+    {
+      "epoch": 0.5300546448087432,
+      "grad_norm": 0.3248935341835022,
+      "learning_rate": 1.3145133756295936e-05,
+      "loss": 1.1913241147994995,
+      "step": 194
+    },
+    {
+      "epoch": 0.5355191256830601,
+      "grad_norm": 0.5341064929962158,
+      "learning_rate": 1.2976250370094668e-05,
+      "loss": 1.2624198198318481,
+      "step": 196
+    },
+    {
+      "epoch": 0.5409836065573771,
+      "grad_norm": 0.31245383620262146,
+      "learning_rate": 1.2806632606005822e-05,
+      "loss": 1.0798206329345703,
+      "step": 198
+    },
+    {
+      "epoch": 0.546448087431694,
+      "grad_norm": 0.46515926718711853,
+      "learning_rate": 1.2636343494269479e-05,
+      "loss": 1.2240967750549316,
+      "step": 200
+    },
+    {
+      "epoch": 0.5519125683060109,
+      "grad_norm": 0.4435195326805115,
+      "learning_rate": 1.2465446314599609e-05,
+      "loss": 1.0411522388458252,
+      "step": 202
+    },
+    {
+      "epoch": 0.5573770491803278,
+      "grad_norm": 0.45708101987838745,
+      "learning_rate": 1.2294004572669228e-05,
+      "loss": 0.869902491569519,
+      "step": 204
+    },
+    {
+      "epoch": 0.5628415300546448,
+      "grad_norm": 0.31885483860969543,
+      "learning_rate": 1.2122081976511581e-05,
+      "loss": 1.2103023529052734,
+      "step": 206
+    },
+    {
+      "epoch": 0.5683060109289617,
+      "grad_norm": 0.6482557058334351,
+      "learning_rate": 1.1949742412846142e-05,
+      "loss": 1.1392232179641724,
+      "step": 208
+    },
+    {
+      "epoch": 0.5737704918032787,
+      "grad_norm": 0.3855375349521637,
+      "learning_rate": 1.177704992333818e-05,
+      "loss": 1.1707613468170166,
+      "step": 210
+    },
+    {
+      "epoch": 0.5792349726775956,
+      "grad_norm": 0.326550155878067,
+      "learning_rate": 1.1604068680800809e-05,
+      "loss": 1.4717496633529663,
+      "step": 212
+    },
+    {
+      "epoch": 0.5846994535519126,
+      "grad_norm": 0.5021712183952332,
+      "learning_rate": 1.1430862965348224e-05,
+      "loss": 1.184004306793213,
+      "step": 214
+    },
+    {
+      "epoch": 0.5901639344262295,
+      "grad_norm": 0.9780371189117432,
+      "learning_rate": 1.1257497140509141e-05,
+      "loss": 1.53951895236969,
+      "step": 216
+    },
+    {
+      "epoch": 0.5956284153005464,
+      "grad_norm": 0.3760947585105896,
+      "learning_rate": 1.1084035629309176e-05,
+      "loss": 1.152180790901184,
+      "step": 218
+    },
+    {
+      "epoch": 0.6010928961748634,
+      "grad_norm": 0.3090932071208954,
+      "learning_rate": 1.0910542890331162e-05,
+      "loss": 1.22081458568573,
+      "step": 220
+    },
+    {
+      "epoch": 0.6065573770491803,
+      "grad_norm": 0.7166253328323364,
+      "learning_rate": 1.0737083393762213e-05,
+      "loss": 1.041406273841858,
+      "step": 222
+    },
+    {
+      "epoch": 0.6120218579234973,
+      "grad_norm": 1.7460894584655762,
+      "learning_rate": 1.0563721597436525e-05,
+      "loss": 0.9747711420059204,
+      "step": 224
+    },
+    {
+      "epoch": 0.6174863387978142,
+      "grad_norm": 0.7671729326248169,
+      "learning_rate": 1.039052192288271e-05,
+      "loss": 1.1867763996124268,
+      "step": 226
+    },
+    {
+      "epoch": 0.6229508196721312,
+      "grad_norm": 1.2053930759429932,
+      "learning_rate": 1.0217548731384677e-05,
+      "loss": 1.1788854598999023,
+      "step": 228
+    },
+    {
+      "epoch": 0.6284153005464481,
+      "grad_norm": 0.2746277153491974,
+      "learning_rate": 1.0044866300064842e-05,
+      "loss": 0.6711171865463257,
+      "step": 230
+    },
+    {
+      "epoch": 0.6338797814207651,
+      "grad_norm": 0.643056333065033,
+      "learning_rate": 9.872538797998672e-06,
+      "loss": 1.1947264671325684,
+      "step": 232
+    },
+    {
+      "epoch": 0.639344262295082,
+      "grad_norm": 10.281498908996582,
+      "learning_rate": 9.700630262369337e-06,
+      "loss": 0.8791699409484863,
+      "step": 234
+    },
+    {
+      "epoch": 0.644808743169399,
+      "grad_norm": 0.8242344260215759,
+      "learning_rate": 9.529204574671391e-06,
+      "loss": 0.8508569598197937,
+      "step": 236
+    },
+    {
+      "epoch": 0.6502732240437158,
+      "grad_norm": 0.825342059135437,
+      "learning_rate": 9.3583254369723e-06,
+      "loss": 1.25392746925354,
+      "step": 238
+    },
+    {
+      "epoch": 0.6557377049180327,
+      "grad_norm": 0.4026271104812622,
+      "learning_rate": 9.188056348240655e-06,
+      "loss": 1.1808273792266846,
+      "step": 240
+    },
+    {
+      "epoch": 0.6612021857923497,
+      "grad_norm": 0.5428042411804199,
+      "learning_rate": 9.018460580749842e-06,
+      "loss": 1.1250752210617065,
+      "step": 242
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.5019088387489319,
+      "learning_rate": 8.849601156565972e-06,
+      "loss": 0.9948179125785828,
+      "step": 244
+    },
+    {
+      "epoch": 0.6721311475409836,
+      "grad_norm": 0.32495370507240295,
+      "learning_rate": 8.68154082412877e-06,
+      "loss": 1.1509572267532349,
+      "step": 246
+    },
+    {
+      "epoch": 0.6775956284153005,
+      "grad_norm": 0.47498419880867004,
+      "learning_rate": 8.514342034934159e-06,
+      "loss": 1.2174043655395508,
+      "step": 248
+    },
+    {
+      "epoch": 0.6830601092896175,
+      "grad_norm": 0.3702382445335388,
+      "learning_rate": 8.348066920327163e-06,
+      "loss": 1.1846040487289429,
+      "step": 250
+    },
+    {
+      "epoch": 0.6885245901639344,
+      "grad_norm": 0.4739413857460022,
+      "learning_rate": 8.182777268413822e-06,
+      "loss": 1.0815660953521729,
+      "step": 252
+    },
+    {
+      "epoch": 0.6939890710382514,
+      "grad_norm": 0.3407984673976898,
+      "learning_rate": 8.018534501100611e-06,
+      "loss": 1.136920690536499,
+      "step": 254
+    },
+    {
+      "epoch": 0.6994535519125683,
+      "grad_norm": 0.3524422347545624,
+      "learning_rate": 7.855399651269982e-06,
+      "loss": 1.1331547498703003,
+      "step": 256
+    },
+    {
+      "epoch": 0.7049180327868853,
+      "grad_norm": 0.28807151317596436,
+      "learning_rate": 7.6934333401004e-06,
+      "loss": 1.0285248756408691,
+      "step": 258
+    },
+    {
+      "epoch": 0.7103825136612022,
+      "grad_norm": 0.3343144655227661,
+      "learning_rate": 7.53269575453947e-06,
+      "loss": 1.1325820684432983,
+      "step": 260
+    },
+    {
+      "epoch": 0.7158469945355191,
+      "grad_norm": 0.38119539618492126,
+      "learning_rate": 7.373246624938324e-06,
+      "loss": 1.2263553142547607,
+      "step": 262
+    },
+    {
+      "epoch": 0.7213114754098361,
+      "grad_norm": 0.312750905752182,
+      "learning_rate": 7.215145202855746e-06,
+      "loss": 1.1782656908035278,
+      "step": 264
+    },
+    {
+      "epoch": 0.726775956284153,
+      "grad_norm": 0.37758493423461914,
+      "learning_rate": 7.0584502390401865e-06,
+      "loss": 1.0094555616378784,
+      "step": 266
+    },
+    {
+      "epoch": 0.73224043715847,
+      "grad_norm": 0.5180346369743347,
+      "learning_rate": 6.903219961597891e-06,
+      "loss": 1.1325385570526123,
+      "step": 268
+    },
+    {
+      "epoch": 0.7377049180327869,
+      "grad_norm": 0.38575479388237,
+      "learning_rate": 6.7495120543552475e-06,
+      "loss": 1.1679610013961792,
+      "step": 270
+    },
+    {
+      "epoch": 0.7431693989071039,
+      "grad_norm": 0.3461146056652069,
+      "learning_rate": 6.59738363542336e-06,
+      "loss": 1.1273691654205322,
+      "step": 272
+    },
+    {
+      "epoch": 0.7486338797814208,
+      "grad_norm": 0.4262787401676178,
+      "learning_rate": 6.446891235972894e-06,
+      "loss": 1.1511529684066772,
+      "step": 274
+    },
+    {
+      "epoch": 0.7540983606557377,
+      "grad_norm": 0.2948058843612671,
+      "learning_rate": 6.298090779226977e-06,
+      "loss": 1.2013486623764038,
+      "step": 276
+    },
+    {
+      "epoch": 0.7595628415300546,
+      "grad_norm": 0.3147113621234894,
+      "learning_rate": 6.151037559680047e-06,
+      "loss": 1.1714566946029663,
+      "step": 278
+    },
+    {
+      "epoch": 0.7650273224043715,
+      "grad_norm": 0.4043079912662506,
+      "learning_rate": 6.005786222550319e-06,
+      "loss": 0.8109129071235657,
+      "step": 280
+    },
+    {
+      "epoch": 0.7704918032786885,
+      "grad_norm": 0.31658855080604553,
+      "learning_rate": 5.8623907434735515e-06,
+      "loss": 1.1680246591567993,
+      "step": 282
+    },
+    {
+      "epoch": 0.7759562841530054,
+      "grad_norm": 0.4184737801551819,
+      "learning_rate": 5.720904408445589e-06,
+      "loss": 1.159695029258728,
+      "step": 284
+    },
+    {
+      "epoch": 0.7814207650273224,
+      "grad_norm": 0.69264817237854,
+      "learning_rate": 5.581379794021202e-06,
+      "loss": 1.173396348953247,
+      "step": 286
+    },
+    {
+      "epoch": 0.7868852459016393,
+      "grad_norm": 0.4731365144252777,
+      "learning_rate": 5.443868747776579e-06,
+      "loss": 1.141501545906067,
+      "step": 288
+    },
+    {
+      "epoch": 0.7923497267759563,
+      "grad_norm": 0.39572659134864807,
+      "learning_rate": 5.308422369042644e-06,
+      "loss": 1.077953577041626,
+      "step": 290
+    },
+    {
+      "epoch": 0.7978142076502732,
+      "grad_norm": 0.4793470501899719,
+      "learning_rate": 5.175090989916483e-06,
+      "loss": 1.1118550300598145,
+      "step": 292
+    },
+    {
+      "epoch": 0.8032786885245902,
+      "grad_norm": 0.2662019729614258,
+      "learning_rate": 5.043924156557844e-06,
+      "loss": 1.1436463594436646,
+      "step": 294
+    },
+    {
+      "epoch": 0.8087431693989071,
+      "grad_norm": 0.4349708557128906,
+      "learning_rate": 4.914970610777725e-06,
+      "loss": 1.4781270027160645,
+      "step": 296
+    },
+    {
+      "epoch": 0.8142076502732241,
+      "grad_norm": 0.49073347449302673,
+      "learning_rate": 4.788278271925802e-06,
+      "loss": 0.8274983763694763,
+      "step": 298
+    },
+    {
+      "epoch": 0.819672131147541,
+      "grad_norm": 0.3969985842704773,
+      "learning_rate": 4.663894219083548e-06,
+      "loss": 1.5442594289779663,
+      "step": 300
+    },
+    {
+      "epoch": 0.825136612021858,
+      "grad_norm": 0.4534980356693268,
+      "learning_rate": 4.541864673569551e-06,
+      "loss": 1.1690313816070557,
+      "step": 302
+    },
+    {
+      "epoch": 0.8306010928961749,
+      "grad_norm": 0.554429292678833,
+      "learning_rate": 4.422234981763613e-06,
+      "loss": 0.7355527281761169,
+      "step": 304
+    },
+    {
+      "epoch": 0.8360655737704918,
+      "grad_norm": 1.9707756042480469,
+      "learning_rate": 4.305049598255946e-06,
+      "loss": 1.5113667249679565,
+      "step": 306
+    },
+    {
+      "epoch": 0.8415300546448088,
+      "grad_norm": 0.36261430382728577,
+      "learning_rate": 4.190352069327777e-06,
+      "loss": 0.8062604069709778,
+      "step": 308
+    },
+    {
+      "epoch": 0.8469945355191257,
+      "grad_norm": 0.3810434639453888,
+      "learning_rate": 4.078185016769484e-06,
+      "loss": 1.1614996194839478,
+      "step": 310
+    },
+    {
+      "epoch": 0.8524590163934426,
+      "grad_norm": 0.39326491951942444,
+      "learning_rate": 3.968590122042265e-06,
+      "loss": 1.177278757095337,
+      "step": 312
+    },
+    {
+      "epoch": 0.8579234972677595,
+      "grad_norm": 0.40779945254325867,
+      "learning_rate": 3.861608110789228e-06,
+      "loss": 1.1831945180892944,
+      "step": 314
+    },
+    {
+      "epoch": 0.8633879781420765,
+      "grad_norm": 0.9780954718589783,
+      "learning_rate": 3.757278737701697e-06,
+      "loss": 1.1853556632995605,
+      "step": 316
+    },
+    {
+      "epoch": 0.8688524590163934,
+      "grad_norm": 1.053917407989502,
+      "learning_rate": 3.6556407717462856e-06,
+      "loss": 1.0666594505310059,
+      "step": 318
+    },
+    {
+      "epoch": 0.8743169398907104,
+      "grad_norm": 0.37668082118034363,
+      "learning_rate": 3.5567319817582944e-06,
+      "loss": 1.272523045539856,
+      "step": 320
+    },
+    {
+      "epoch": 0.8797814207650273,
+      "grad_norm": 0.5615814328193665,
+      "learning_rate": 3.4605891224067423e-06,
+      "loss": 1.1502110958099365,
+      "step": 322
+    },
+    {
+      "epoch": 0.8852459016393442,
+      "grad_norm": 0.3317868709564209,
+      "learning_rate": 3.3672479205362764e-06,
+      "loss": 0.5082761645317078,
+      "step": 324
+    },
+    {
+      "epoch": 0.8907103825136612,
+      "grad_norm": 0.3119502365589142,
+      "learning_rate": 3.276743061891014e-06,
+      "loss": 1.153531551361084,
+      "step": 326
+    },
+    {
+      "epoch": 0.8961748633879781,
+      "grad_norm": 0.6352930068969727,
+      "learning_rate": 3.1891081782252726e-06,
+      "loss": 1.0034343004226685,
+      "step": 328
+    },
+    {
+      "epoch": 0.9016393442622951,
+      "grad_norm": 0.3119983673095703,
+      "learning_rate": 3.1043758348059384e-06,
+      "loss": 1.181221842765808,
+      "step": 330
+    },
+    {
+      "epoch": 0.907103825136612,
+      "grad_norm": 0.9960837960243225,
+      "learning_rate": 3.0225775183111784e-06,
+      "loss": 1.2305197715759277,
+      "step": 332
+    },
+    {
+      "epoch": 0.912568306010929,
+      "grad_norm": 0.45826148986816406,
+      "learning_rate": 2.943743625129917e-06,
+      "loss": 1.013006567955017,
+      "step": 334
+    },
+    {
+      "epoch": 0.9180327868852459,
+      "grad_norm": 0.4111308455467224,
+      "learning_rate": 2.867903450066513e-06,
+      "loss": 1.1280704736709595,
+      "step": 336
+    },
+    {
+      "epoch": 0.9234972677595629,
+      "grad_norm": 0.4843418300151825,
+      "learning_rate": 2.795085175454741e-06,
+      "loss": 1.1432486772537231,
+      "step": 338
+    },
+    {
+      "epoch": 0.9289617486338798,
+      "grad_norm": 0.2703630030155182,
+      "learning_rate": 2.7253158606851983e-06,
+      "loss": 1.1488100290298462,
+      "step": 340
+    },
+    {
+      "epoch": 0.9344262295081968,
+      "grad_norm": 0.32970309257507324,
+      "learning_rate": 2.6586214321499952e-06,
+      "loss": 1.187166690826416,
+      "step": 342
+    },
+    {
+      "epoch": 0.9398907103825137,
+      "grad_norm": 0.714718222618103,
+      "learning_rate": 2.5950266736084558e-06,
+      "loss": 1.1312419176101685,
+      "step": 344
+    },
+    {
+      "epoch": 0.9453551912568307,
+      "grad_norm": 0.5256864428520203,
+      "learning_rate": 2.5345552169774413e-06,
+      "loss": 1.5058574676513672,
+      "step": 346
+    },
+    {
+      "epoch": 0.9508196721311475,
+      "grad_norm": 0.6660139560699463,
+      "learning_rate": 2.477229533549685e-06,
+      "loss": 0.7294368743896484,
+      "step": 348
+    },
+    {
+      "epoch": 0.9562841530054644,
+      "grad_norm": 0.3410135805606842,
+      "learning_rate": 2.423070925643422e-06,
+      "loss": 0.7677943110466003,
+      "step": 350
+    },
+    {
+      "epoch": 0.9617486338797814,
+      "grad_norm": 0.3177313506603241,
+      "learning_rate": 2.372099518686416e-06,
+      "loss": 1.1427574157714844,
+      "step": 352
+    },
+    {
+      "epoch": 0.9672131147540983,
+      "grad_norm": 0.3382633626461029,
+      "learning_rate": 2.324334253737321e-06,
+      "loss": 0.8091766238212585,
+      "step": 354
+    },
+    {
+      "epoch": 0.9726775956284153,
+      "grad_norm": 0.31415683031082153,
+      "learning_rate": 2.2797928804471413e-06,
+      "loss": 1.0229932069778442,
+      "step": 356
+    },
+    {
+      "epoch": 0.9781420765027322,
+      "grad_norm": 0.33465299010276794,
+      "learning_rate": 2.2384919504634465e-06,
+      "loss": 0.7048438787460327,
+      "step": 358
+    },
+    {
+      "epoch": 0.9836065573770492,
+      "grad_norm": 0.7479493618011475,
+      "learning_rate": 2.2004468112797345e-06,
+      "loss": 1.145750880241394,
+      "step": 360
+    },
+    {
+      "epoch": 0.9890710382513661,
+      "grad_norm": 0.6303642988204956,
+      "learning_rate": 2.165671600532298e-06,
+      "loss": 0.9127320051193237,
+      "step": 362
+    },
+    {
+      "epoch": 0.994535519125683,
+      "grad_norm": 0.678534746170044,
+      "learning_rate": 2.134179240746638e-06,
+      "loss": 1.1293567419052124,
+      "step": 364
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.48102641105651855,
+      "learning_rate": 2.1059814345354434e-06,
+      "loss": 1.1180713176727295,
+      "step": 366
+    },
+    {
+      "epoch": 1.005464480874317,
+      "grad_norm": 0.26332399249076843,
+      "learning_rate": 2.0810886602498733e-06,
+      "loss": 0.9773157238960266,
+      "step": 368
+    },
+    {
+      "epoch": 1.010928961748634,
+      "grad_norm": 0.48079046607017517,
+      "learning_rate": 2.059510168085791e-06,
+      "loss": 1.077552318572998,
+      "step": 370
+    },
+    {
+      "epoch": 1.0163934426229508,
+      "grad_norm": 0.3228934407234192,
+      "learning_rate": 2.0412539766463697e-06,
+      "loss": 1.0045071840286255,
+      "step": 372
+    },
+    {
+      "epoch": 1.0218579234972678,
+      "grad_norm": 0.3166205883026123,
+      "learning_rate": 2.0263268699623746e-06,
+      "loss": 1.0588802099227905,
+      "step": 374
+    },
+    {
+      "epoch": 1.0273224043715847,
+      "grad_norm": 1.7906802892684937,
+      "learning_rate": 2.0147343949711965e-06,
+      "loss": 0.8532091975212097,
+      "step": 376
+    },
+    {
+      "epoch": 1.0327868852459017,
+      "grad_norm": 0.40486088395118713,
+      "learning_rate": 2.0064808594556066e-06,
+      "loss": 0.959248960018158,
+      "step": 378
+    },
+    {
+      "epoch": 1.0382513661202186,
+      "grad_norm": 0.5614176392555237,
+      "learning_rate": 2.0015693304429757e-06,
+      "loss": 0.6482217311859131,
+      "step": 380
+    },
+    {
+      "epoch": 1.0437158469945356,
+      "grad_norm": 0.9394697546958923,
+      "learning_rate": 2.000001633065562e-06,
+      "loss": 0.9249519109725952,
+      "step": 382
+    },
+    {
+      "epoch": 1.0491803278688525,
+      "grad_norm": 0.38094520568847656,
+      "learning_rate": 2.0017783498822896e-06,
+      "loss": 0.9705590605735779,
+      "step": 384
+    },
+    {
+      "epoch": 1.0546448087431695,
+      "grad_norm": 1.61484956741333,
+      "learning_rate": 2.006898820662268e-06,
+      "loss": 0.623333215713501,
+      "step": 386
+    },
+    {
+      "epoch": 1.0601092896174864,
+      "grad_norm": 0.5918154120445251,
+      "learning_rate": 2.0153611426301325e-06,
+      "loss": 0.5663283467292786,
+      "step": 388
+    },
+    {
+      "epoch": 1.0655737704918034,
+      "grad_norm": 0.5102803707122803,
+      "learning_rate": 2.027162171173126e-06,
+      "loss": 1.11468505859375,
+      "step": 390
+    },
+    {
+      "epoch": 1.0710382513661203,
+      "grad_norm": 0.4305904507637024,
+      "learning_rate": 2.0422975210096317e-06,
+      "loss": 0.9271813631057739,
+      "step": 392
+    },
+    {
+      "epoch": 1.0765027322404372,
+      "grad_norm": 0.36596590280532837,
+      "learning_rate": 2.0607615678187605e-06,
+      "loss": 0.6126598119735718,
+      "step": 394
+    },
+    {
+      "epoch": 1.0819672131147542,
+      "grad_norm": 0.41866230964660645,
+      "learning_rate": 2.082547450330353e-06,
+      "loss": 0.971298336982727,
+      "step": 396
+    },
+    {
+      "epoch": 1.0874316939890711,
+      "grad_norm": 0.4388103187084198,
+      "learning_rate": 2.1076470728746407e-06,
+      "loss": 0.9021328091621399,
+      "step": 398
+    },
+    {
+      "epoch": 1.092896174863388,
+      "grad_norm": 3.2539632320404053,
+      "learning_rate": 2.136051108390608e-06,
+      "loss": 1.0814286470413208,
+      "step": 400
+    },
+    {
+      "epoch": 1.098360655737705,
+      "grad_norm": 0.8472021222114563,
+      "learning_rate": 2.167749001891944e-06,
+      "loss": 0.9241260290145874,
+      "step": 402
+    },
+    {
+      "epoch": 1.1038251366120218,
+      "grad_norm": 0.5540516376495361,
+      "learning_rate": 2.202728974389296e-06,
+      "loss": 1.0775995254516602,
+      "step": 404
+    },
+    {
+      "epoch": 1.1092896174863387,
+      "grad_norm": 1.049932599067688,
+      "learning_rate": 2.240978027267357e-06,
+      "loss": 1.1052746772766113,
+      "step": 406
+    },
+    {
+      "epoch": 1.1147540983606556,
+      "grad_norm": 0.5960119962692261,
+      "learning_rate": 2.2824819471151736e-06,
+      "loss": 0.610589861869812,
+      "step": 408
+    },
+    {
+      "epoch": 1.1202185792349726,
+      "grad_norm": 0.3115493953227997,
+      "learning_rate": 2.327225311007878e-06,
+      "loss": 1.1588163375854492,
+      "step": 410
+    },
+    {
+      "epoch": 1.1256830601092895,
+      "grad_norm": 1.5775190591812134,
+      "learning_rate": 2.3751914922378623e-06,
+      "loss": 0.662068784236908,
+      "step": 412
+    },
+    {
+      "epoch": 1.1311475409836065,
+      "grad_norm": 0.49976834654808044,
+      "learning_rate": 2.4263626664932998e-06,
+      "loss": 0.7565730214118958,
+      "step": 414
+    },
+    {
+      "epoch": 1.1366120218579234,
+      "grad_norm": 1.2116132974624634,
+      "learning_rate": 2.4807198184816817e-06,
+      "loss": 0.6296537518501282,
+      "step": 416
+    },
+    {
+      "epoch": 1.1420765027322404,
+      "grad_norm": 1.1844316720962524,
+      "learning_rate": 2.5382427489959373e-06,
+      "loss": 0.9881916046142578,
+      "step": 418
+    },
+    {
+      "epoch": 1.1475409836065573,
+      "grad_norm": 0.5167036652565002,
+      "learning_rate": 2.5989100824204876e-06,
+      "loss": 0.9343990683555603,
+      "step": 420
+    },
+    {
+      "epoch": 1.1530054644808743,
+      "grad_norm": 0.558880627155304,
+      "learning_rate": 2.662699274674462e-06,
+      "loss": 1.0110293626785278,
+      "step": 422
+    },
+    {
+      "epoch": 1.1584699453551912,
+      "grad_norm": 1.1780766248703003,
+      "learning_rate": 2.7295866215891107e-06,
+      "loss": 0.7126035690307617,
+      "step": 424
+    },
+    {
+      "epoch": 1.1639344262295082,
+      "grad_norm": 0.3370395004749298,
+      "learning_rate": 2.799547267716326e-06,
+      "loss": 0.6112310886383057,
+      "step": 426
+    },
+    {
+      "epoch": 1.169398907103825,
+      "grad_norm": 0.8769522905349731,
+      "learning_rate": 2.872555215564946e-06,
+      "loss": 1.1242718696594238,
+      "step": 428
+    },
+    {
+      "epoch": 1.174863387978142,
+      "grad_norm": 0.6202550530433655,
+      "learning_rate": 2.9485833352614895e-06,
+      "loss": 0.9957835078239441,
+      "step": 430
+    },
+    {
+      "epoch": 1.180327868852459,
+      "grad_norm": 0.48545965552330017,
+      "learning_rate": 3.027603374631647e-06,
+      "loss": 1.0449634790420532,
+      "step": 432
+    },
+    {
+      "epoch": 1.185792349726776,
+      "grad_norm": 0.49180424213409424,
+      "learning_rate": 3.1095859696988273e-06,
+      "loss": 0.7476254105567932,
+      "step": 434
+    },
+    {
+      "epoch": 1.1912568306010929,
+      "grad_norm": 0.4692288041114807,
+      "learning_rate": 3.1945006555958885e-06,
+      "loss": 1.060101866722107,
+      "step": 436
+    },
+    {
+      "epoch": 1.1967213114754098,
+      "grad_norm": 0.26823532581329346,
+      "learning_rate": 3.2823158778858976e-06,
+      "loss": 0.9638835191726685,
+      "step": 438
+    },
+    {
+      "epoch": 1.2021857923497268,
+      "grad_norm": 0.3634532392024994,
+      "learning_rate": 3.372999004287839e-06,
+      "loss": 0.8774067163467407,
+      "step": 440
+    },
+    {
+      "epoch": 1.2076502732240437,
+      "grad_norm": 0.2828170359134674,
+      "learning_rate": 3.4665163368028044e-06,
+      "loss": 1.0114994049072266,
+      "step": 442
+    },
+    {
+      "epoch": 1.2131147540983607,
+      "grad_norm": 0.4509800970554352,
+      "learning_rate": 3.562833124236238e-06,
+      "loss": 0.9758923053741455,
+      "step": 444
+    },
+    {
+      "epoch": 1.2185792349726776,
+      "grad_norm": 0.9190130829811096,
+      "learning_rate": 3.6619135751115325e-06,
+      "loss": 0.5354660153388977,
+      "step": 446
+    },
+    {
+      "epoch": 1.2240437158469946,
+      "grad_norm": 0.39201292395591736,
+      "learning_rate": 3.763720870970201e-06,
+      "loss": 0.9081273674964905,
+      "step": 448
+    },
+    {
+      "epoch": 1.2295081967213115,
+      "grad_norm": 0.6562516689300537,
+      "learning_rate": 3.86821718005367e-06,
+      "loss": 0.9579398036003113,
+      "step": 450
+    },
+    {
+      "epoch": 1.2349726775956285,
+      "grad_norm": 0.5844786763191223,
+      "learning_rate": 3.975363671361641e-06,
+      "loss": 0.9740945100784302,
+      "step": 452
+    },
+    {
+      "epoch": 1.2404371584699454,
+      "grad_norm": 0.6142482757568359,
+      "learning_rate": 4.0851205290817254e-06,
+      "loss": 0.34633180499076843,
+      "step": 454
+    },
+    {
+      "epoch": 1.2459016393442623,
+      "grad_norm": 0.7541310787200928,
+      "learning_rate": 4.197446967385105e-06,
+      "loss": 0.7628012895584106,
+      "step": 456
+    },
+    {
+      "epoch": 1.2513661202185793,
+      "grad_norm": 0.44816914200782776,
+      "learning_rate": 4.312301245582571e-06,
+      "loss": 0.7231332063674927,
+      "step": 458
+    },
+    {
+      "epoch": 1.2568306010928962,
+      "grad_norm": 0.9218317866325378,
+      "learning_rate": 4.429640683635466e-06,
+      "loss": 1.0563876628875732,
+      "step": 460
+    },
+    {
+      "epoch": 1.2622950819672132,
+      "grad_norm": 2.9262020587921143,
+      "learning_rate": 4.549421678015633e-06,
+      "loss": 0.6581028699874878,
+      "step": 462
+    },
+    {
+      "epoch": 1.2677595628415301,
+      "grad_norm": 0.4514838457107544,
+      "learning_rate": 4.671599717908582e-06,
+      "loss": 0.985448956489563,
+      "step": 464
+    },
+    {
+      "epoch": 1.273224043715847,
+      "grad_norm": 0.3441561758518219,
+      "learning_rate": 4.796129401753752e-06,
+      "loss": 1.0194132328033447,
+      "step": 466
+    },
+    {
+      "epoch": 1.278688524590164,
+      "grad_norm": 0.3873656690120697,
+      "learning_rate": 4.922964454115837e-06,
+      "loss": 0.9317984580993652,
+      "step": 468
+    },
+    {
+      "epoch": 1.2841530054644807,
+      "grad_norm": 0.5289583802223206,
+      "learning_rate": 5.0520577428807835e-06,
+      "loss": 0.9312957525253296,
+      "step": 470
+    },
+    {
+      "epoch": 1.289617486338798,
+      "grad_norm": 0.3908224403858185,
+      "learning_rate": 5.183361296770197e-06,
+      "loss": 1.2282073497772217,
+      "step": 472
+    },
+    {
+      "epoch": 1.2950819672131146,
+      "grad_norm": 0.4844124913215637,
+      "learning_rate": 5.316826323167505e-06,
+      "loss": 0.9655548334121704,
+      "step": 474
+    },
+    {
+      "epoch": 1.3005464480874318,
+      "grad_norm": 0.6156196594238281,
+      "learning_rate": 5.4524032262494175e-06,
+      "loss": 0.5244005918502808,
+      "step": 476
+    },
+    {
+      "epoch": 1.3060109289617485,
+      "grad_norm": 0.43006354570388794,
+      "learning_rate": 5.590041625415783e-06,
+      "loss": 1.0403926372528076,
+      "step": 478
+    },
+    {
+      "epoch": 1.3114754098360657,
+      "grad_norm": 0.2751167118549347,
+      "learning_rate": 5.7296903740111076e-06,
+      "loss": 1.0301437377929688,
+      "step": 480
+    },
+    {
+      "epoch": 1.3169398907103824,
+      "grad_norm": 0.8657461404800415,
+      "learning_rate": 5.87129757833077e-06,
+      "loss": 0.923163652420044,
+      "step": 482
+    },
+    {
+      "epoch": 1.3224043715846996,
+      "grad_norm": 0.6536735892295837,
+      "learning_rate": 6.014810616904747e-06,
+      "loss": 0.6813654899597168,
+      "step": 484
+    },
+    {
+      "epoch": 1.3278688524590163,
+      "grad_norm": 0.3719988167285919,
+      "learning_rate": 6.160176160051906e-06,
+      "loss": 0.958286702632904,
+      "step": 486
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.5692084431648254,
+      "learning_rate": 6.307340189697344e-06,
+      "loss": 1.0355274677276611,
+      "step": 488
+    },
+    {
+      "epoch": 1.3387978142076502,
+      "grad_norm": 0.5361304879188538,
+      "learning_rate": 6.456248019445626e-06,
+      "loss": 0.5718910694122314,
+      "step": 490
+    },
+    {
+      "epoch": 1.3442622950819672,
+      "grad_norm": 0.46025416254997253,
+      "learning_rate": 6.606844314902321e-06,
+      "loss": 1.0728013515472412,
+      "step": 492
+    },
+    {
+      "epoch": 1.349726775956284,
+      "grad_norm": 0.14679169654846191,
+      "learning_rate": 6.7590731142363915e-06,
+      "loss": 0.7085850238800049,
+      "step": 494
+    },
+    {
+      "epoch": 1.355191256830601,
+      "grad_norm": 0.4374118149280548,
+      "learning_rate": 6.912877848975638e-06,
+      "loss": 1.0255423784255981,
+      "step": 496
+    },
+    {
+      "epoch": 1.360655737704918,
+      "grad_norm": 1.178649663925171,
+      "learning_rate": 7.068201365027712e-06,
+      "loss": 0.8840912580490112,
+      "step": 498
+    },
+    {
+      "epoch": 1.366120218579235,
+      "grad_norm": 0.41059738397598267,
+      "learning_rate": 7.2249859439185875e-06,
+      "loss": 1.1463178396224976,
+      "step": 500
+    },
+    {
+      "epoch": 1.3715846994535519,
+      "grad_norm": 0.38757556676864624,
+      "learning_rate": 7.3831733242409285e-06,
+      "loss": 0.8971600532531738,
+      "step": 502
+    },
+    {
+      "epoch": 1.3770491803278688,
+      "grad_norm": 0.3446296155452728,
+      "learning_rate": 7.5427047233040485e-06,
+      "loss": 0.48697370290756226,
+      "step": 504
+    },
+    {
+      "epoch": 1.3825136612021858,
+      "grad_norm": 0.24354782700538635,
+      "learning_rate": 7.703520858977702e-06,
+      "loss": 0.7695150375366211,
+      "step": 506
+    },
+    {
+      "epoch": 1.3879781420765027,
+      "grad_norm": 0.39158889651298523,
+      "learning_rate": 7.865561971721389e-06,
+      "loss": 0.9563601016998291,
+      "step": 508
+    },
+    {
+      "epoch": 1.3934426229508197,
+      "grad_norm": 0.47387513518333435,
+      "learning_rate": 8.02876784679115e-06,
+      "loss": 0.97979336977005,
+      "step": 510
+    },
+    {
+      "epoch": 1.3989071038251366,
+      "grad_norm": 0.5868260860443115,
+      "learning_rate": 8.193077836615386e-06,
+      "loss": 0.9143297672271729,
+      "step": 512
+    },
+    {
+      "epoch": 1.4043715846994536,
+      "grad_norm": 0.3844831883907318,
+      "learning_rate": 8.35843088333168e-06,
+      "loss": 0.6998130679130554,
+      "step": 514
+    },
+    {
+      "epoch": 1.4098360655737705,
+      "grad_norm": 0.3410232663154602,
+      "learning_rate": 8.524765541475935e-06,
+      "loss": 0.7496625185012817,
+      "step": 516
+    },
+    {
+      "epoch": 1.4153005464480874,
+      "grad_norm": 0.3015901744365692,
+      "learning_rate": 8.692020000815627e-06,
+      "loss": 0.9690025448799133,
+      "step": 518
+    },
+    {
+      "epoch": 1.4207650273224044,
+      "grad_norm": 1.459832787513733,
+      "learning_rate": 8.860132109318622e-06,
+      "loss": 0.7310088872909546,
+      "step": 520
+    },
+    {
+      "epoch": 1.4262295081967213,
+      "grad_norm": 0.32720673084259033,
+      "learning_rate": 9.029039396248916e-06,
+      "loss": 1.0313293933868408,
+      "step": 522
+    },
+    {
+      "epoch": 1.4316939890710383,
+      "grad_norm": 0.7867904901504517,
+      "learning_rate": 9.198679095380924e-06,
+      "loss": 1.0015819072723389,
+      "step": 524
+    },
+    {
+      "epoch": 1.4371584699453552,
+      "grad_norm": 0.7222017049789429,
+      "learning_rate": 9.368988168323451e-06,
+      "loss": 1.0047054290771484,
+      "step": 526
+    },
+    {
+      "epoch": 1.4426229508196722,
+      "grad_norm": 0.5384648442268372,
+      "learning_rate": 9.539903327944926e-06,
+      "loss": 0.6519490480422974,
+      "step": 528
+    },
+    {
+      "epoch": 1.4480874316939891,
+      "grad_norm": 0.9112948179244995,
+      "learning_rate": 9.711361061890942e-06,
+      "loss": 0.7796599864959717,
+      "step": 530
+    },
+    {
+      "epoch": 1.453551912568306,
+      "grad_norm": 1.056015133857727,
+      "learning_rate": 9.8832976561856e-06,
+      "loss": 0.7928510308265686,
+      "step": 532
+    },
+    {
+      "epoch": 1.459016393442623,
+      "grad_norm": 0.4741556644439697,
+      "learning_rate": 1.0055649218907688e-05,
+      "loss": 1.0018107891082764,
+      "step": 534
+    },
+    {
+      "epoch": 1.46448087431694,
+      "grad_norm": 0.5680137872695923,
+      "learning_rate": 1.0228351703933075e-05,
+      "loss": 0.9849597811698914,
+      "step": 536
+    },
+    {
+      "epoch": 1.469945355191257,
+      "grad_norm": 0.31153297424316406,
+      "learning_rate": 1.0401340934734287e-05,
+      "loss": 1.0444822311401367,
+      "step": 538
+    },
+    {
+      "epoch": 1.4754098360655736,
+      "grad_norm": 0.26671940088272095,
+      "learning_rate": 1.0574552628228691e-05,
+      "loss": 1.0459693670272827,
+      "step": 540
+    },
+    {
+      "epoch": 1.4808743169398908,
+      "grad_norm": 0.48387154936790466,
+      "learning_rate": 1.0747922418666115e-05,
+      "loss": 1.0217854976654053,
+      "step": 542
+    },
+    {
+      "epoch": 1.4863387978142075,
+      "grad_norm": 0.41534528136253357,
+      "learning_rate": 1.0921385881547311e-05,
+      "loss": 0.8353682160377502,
+      "step": 544
+    },
+    {
+      "epoch": 1.4918032786885247,
+      "grad_norm": 1.1913517713546753,
+      "learning_rate": 1.1094878557564217e-05,
+      "loss": 1.092334270477295,
+      "step": 546
+    },
+    {
+      "epoch": 1.4972677595628414,
+      "grad_norm": 0.35618963837623596,
+      "learning_rate": 1.1268335976553098e-05,
+      "loss": 0.4362333118915558,
+      "step": 548
+    },
+    {
+      "epoch": 1.5027322404371586,
+      "grad_norm": 0.45524540543556213,
+      "learning_rate": 1.144169368145179e-05,
+      "loss": 1.027480125427246,
+      "step": 550
+    },
+    {
+      "epoch": 1.5081967213114753,
+      "grad_norm": 0.34574341773986816,
+      "learning_rate": 1.1614887252252076e-05,
+      "loss": 0.9967020153999329,
+      "step": 552
+    },
+    {
+      "epoch": 1.5136612021857925,
+      "grad_norm": 0.33023980259895325,
+      "learning_rate": 1.1787852329938198e-05,
+      "loss": 1.0062674283981323,
+      "step": 554
+    },
+    {
+      "epoch": 1.5191256830601092,
+      "grad_norm": 0.41989317536354065,
+      "learning_rate": 1.1960524640402862e-05,
+      "loss": 1.0924804210662842,
+      "step": 556
+    },
+    {
+      "epoch": 1.5245901639344264,
+      "grad_norm": 0.4085758924484253,
+      "learning_rate": 1.2132840018331514e-05,
+      "loss": 0.802219033241272,
+      "step": 558
+    },
+    {
+      "epoch": 1.530054644808743,
+      "grad_norm": 0.5502913594245911,
+      "learning_rate": 1.2304734431046335e-05,
+      "loss": 0.5939708948135376,
+      "step": 560
+    },
+    {
+      "epoch": 1.5355191256830603,
+      "grad_norm": 0.8035076856613159,
+      "learning_rate": 1.2476144002300864e-05,
+      "loss": 0.915844738483429,
+      "step": 562
+    },
+    {
+      "epoch": 1.540983606557377,
+      "grad_norm": 0.35893988609313965,
+      "learning_rate": 1.264700503601655e-05,
+      "loss": 0.8058934807777405,
+      "step": 564
+    },
+    {
+      "epoch": 1.5464480874316942,
+      "grad_norm": 0.40697717666625977,
+      "learning_rate": 1.2817254039952253e-05,
+      "loss": 1.0083532333374023,
+      "step": 566
+    },
+    {
+      "epoch": 1.5519125683060109,
+      "grad_norm": 0.9251991510391235,
+      "learning_rate": 1.2986827749298138e-05,
+      "loss": 0.8175945281982422,
+      "step": 568
+    },
+    {
+      "epoch": 1.5573770491803278,
+      "grad_norm": 0.4043232500553131,
+      "learning_rate": 1.3155663150184942e-05,
+      "loss": 1.135609745979309,
+      "step": 570
+    },
+    {
+      "epoch": 1.5628415300546448,
+      "grad_norm": 0.41902995109558105,
+      "learning_rate": 1.3323697503100035e-05,
+      "loss": 0.9482666850090027,
+      "step": 572
+    },
+    {
+      "epoch": 1.5683060109289617,
+      "grad_norm": 0.30340319871902466,
+      "learning_rate": 1.3490868366201527e-05,
+      "loss": 1.0169366598129272,
+      "step": 574
+    },
+    {
+      "epoch": 1.5737704918032787,
+      "grad_norm": 0.36948615312576294,
+      "learning_rate": 1.3657113618521763e-05,
+      "loss": 0.7928745746612549,
+      "step": 576
+    },
+    {
+      "epoch": 1.5792349726775956,
+      "grad_norm": 0.4047572910785675,
+      "learning_rate": 1.3822371483051593e-05,
+      "loss": 1.0838329792022705,
+      "step": 578
+    },
+    {
+      "epoch": 1.5846994535519126,
+      "grad_norm": 0.16013656556606293,
+      "learning_rate": 1.3986580549696777e-05,
+      "loss": 0.5277568697929382,
+      "step": 580
+    },
+    {
+      "epoch": 1.5901639344262295,
+      "grad_norm": 0.30199313163757324,
+      "learning_rate": 1.4149679798098097e-05,
+      "loss": 0.7084992527961731,
+      "step": 582
+    },
+    {
+      "epoch": 1.5956284153005464,
+      "grad_norm": 0.751322865486145,
+      "learning_rate": 1.4311608620306626e-05,
+      "loss": 0.6402939558029175,
+      "step": 584
+    },
+    {
+      "epoch": 1.6010928961748634,
+      "grad_norm": 0.4171648621559143,
+      "learning_rate": 1.447230684330573e-05,
+      "loss": 0.9727059602737427,
+      "step": 586
+    },
+    {
+      "epoch": 1.6065573770491803,
+      "grad_norm": 0.4916969835758209,
+      "learning_rate": 1.4631714751371456e-05,
+      "loss": 0.9658234119415283,
+      "step": 588
+    },
+    {
+      "epoch": 1.6120218579234973,
+      "grad_norm": 0.7164318561553955,
+      "learning_rate": 1.4789773108263016e-05,
+      "loss": 0.9527289867401123,
+      "step": 590
+    },
+    {
+      "epoch": 1.6174863387978142,
+      "grad_norm": 0.4969192445278168,
+      "learning_rate": 1.4946423179235068e-05,
+      "loss": 1.0558418035507202,
+      "step": 592
+    },
+    {
+      "epoch": 1.6229508196721312,
+      "grad_norm": 3.1464731693267822,
+      "learning_rate": 1.5101606752863606e-05,
+      "loss": 0.8807100057601929,
+      "step": 594
+    },
+    {
+      "epoch": 1.6284153005464481,
+      "grad_norm": 0.38706547021865845,
+      "learning_rate": 1.5255266162677466e-05,
+      "loss": 0.46191394329071045,
+      "step": 596
+    },
+    {
+      "epoch": 1.633879781420765,
+      "grad_norm": 0.5958669185638428,
+      "learning_rate": 1.540734430858725e-05,
+      "loss": 1.1970001459121704,
+      "step": 598
+    },
+    {
+      "epoch": 1.639344262295082,
+      "grad_norm": 0.4041631519794464,
+      "learning_rate": 1.5557784678103852e-05,
+      "loss": 1.1198264360427856,
+      "step": 600
+    },
+    {
+      "epoch": 1.644808743169399,
+      "grad_norm": 0.27836379408836365,
+      "learning_rate": 1.5706531367338546e-05,
+      "loss": 0.9951107501983643,
+      "step": 602
+    },
+    {
+      "epoch": 1.650273224043716,
+      "grad_norm": 0.3826007544994354,
+      "learning_rate": 1.5853529101776985e-05,
+      "loss": 1.015580654144287,
+      "step": 604
+    },
+    {
+      "epoch": 1.6557377049180326,
+      "grad_norm": 0.3726518452167511,
+      "learning_rate": 1.5998723256819298e-05,
+      "loss": 1.165217638015747,
+      "step": 606
+    },
+    {
+      "epoch": 1.6612021857923498,
+      "grad_norm": 0.3667909502983093,
+      "learning_rate": 1.614205987807872e-05,
+      "loss": 0.8437522649765015,
+      "step": 608
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.27238333225250244,
+      "learning_rate": 1.628348570143105e-05,
+      "loss": 1.1401313543319702,
+      "step": 610
+    },
+    {
+      "epoch": 1.6721311475409837,
+      "grad_norm": 0.6837865710258484,
+      "learning_rate": 1.6422948172807745e-05,
+      "loss": 0.6602663993835449,
+      "step": 612
+    },
+    {
+      "epoch": 1.6775956284153004,
+      "grad_norm": 0.23932527005672455,
+      "learning_rate": 1.6560395467725086e-05,
+      "loss": 0.9976965188980103,
+      "step": 614
+    },
+    {
+      "epoch": 1.6830601092896176,
+      "grad_norm": 0.3285486400127411,
+      "learning_rate": 1.6695776510542253e-05,
+      "loss": 1.039743185043335,
+      "step": 616
+    },
+    {
+      "epoch": 1.6885245901639343,
+      "grad_norm": 0.25682809948921204,
+      "learning_rate": 1.6829040993441085e-05,
+      "loss": 0.5925134420394897,
+      "step": 618
+    },
+    {
+      "epoch": 1.6939890710382515,
+      "grad_norm": 0.8280110955238342,
+      "learning_rate": 1.696013939512057e-05,
+      "loss": 1.1452248096466064,
+      "step": 620
+    },
+    {
+      "epoch": 1.6994535519125682,
+      "grad_norm": 1.396665334701538,
+      "learning_rate": 1.7089022999199064e-05,
+      "loss": 1.2453051805496216,
+      "step": 622
+    },
+    {
+      "epoch": 1.7049180327868854,
+      "grad_norm": 0.29501214623451233,
+      "learning_rate": 1.7215643912317323e-05,
+      "loss": 0.7819148302078247,
+      "step": 624
+    },
+    {
+      "epoch": 1.710382513661202,
+      "grad_norm": 0.2633409798145294,
+      "learning_rate": 1.73399550819358e-05,
+      "loss": 0.681013286113739,
+      "step": 626
+    },
+    {
+      "epoch": 1.7158469945355193,
+      "grad_norm": 1.002079963684082,
+      "learning_rate": 1.746191031381943e-05,
+      "loss": 1.099534511566162,
+      "step": 628
+    },
+    {
+      "epoch": 1.721311475409836,
+      "grad_norm": 0.31711679697036743,
+      "learning_rate": 1.7581464289203475e-05,
+      "loss": 0.8212696313858032,
+      "step": 630
+    },
+    {
+      "epoch": 1.7267759562841531,
+      "grad_norm": 0.2709159851074219,
+      "learning_rate": 1.7698572581634083e-05,
+      "loss": 1.104303002357483,
+      "step": 632
+    },
+    {
+      "epoch": 1.7322404371584699,
+      "grad_norm": 0.3629956841468811,
+      "learning_rate": 1.781319167347718e-05,
+      "loss": 1.1676666736602783,
+      "step": 634
+    },
+    {
+      "epoch": 1.737704918032787,
+      "grad_norm": 0.4824794828891754,
+      "learning_rate": 1.7925278972089748e-05,
+      "loss": 1.005893349647522,
+      "step": 636
+    },
+    {
+      "epoch": 1.7431693989071038,
+      "grad_norm": 0.28539636731147766,
+      "learning_rate": 1.8034792825647287e-05,
+      "loss": 1.1372097730636597,
+      "step": 638
+    },
+    {
+      "epoch": 1.748633879781421,
+      "grad_norm": 0.5941718816757202,
+      "learning_rate": 1.8141692538621716e-05,
+      "loss": 0.9606054425239563,
+      "step": 640
+    },
+    {
+      "epoch": 1.7540983606557377,
+      "grad_norm": 0.23658542335033417,
+      "learning_rate": 1.8245938386903896e-05,
+      "loss": 1.1501632928848267,
+      "step": 642
+    },
+    {
+      "epoch": 1.7595628415300546,
+      "grad_norm": 0.31275659799575806,
+      "learning_rate": 1.8347491632565156e-05,
+      "loss": 1.0679880380630493,
+      "step": 644
+    },
+    {
+      "epoch": 1.7650273224043715,
+      "grad_norm": 0.31820371747016907,
+      "learning_rate": 1.8446314538252407e-05,
+      "loss": 0.765678882598877,
+      "step": 646
+    },
+    {
+      "epoch": 1.7704918032786885,
+      "grad_norm": 0.2747284173965454,
+      "learning_rate": 1.8542370381211374e-05,
+      "loss": 1.0073585510253906,
+      "step": 648
+    },
+    {
+      "epoch": 1.7759562841530054,
+      "grad_norm": 0.33899250626564026,
+      "learning_rate": 1.8635623466932843e-05,
+      "loss": 1.0795483589172363,
+      "step": 650
+    },
+    {
+      "epoch": 1.7814207650273224,
+      "grad_norm": 0.3018670976161957,
+      "learning_rate": 1.8726039142416796e-05,
+      "loss": 1.139539122581482,
+      "step": 652
+    },
+    {
+      "epoch": 1.7868852459016393,
+      "grad_norm": 0.40128761529922485,
+      "learning_rate": 1.881358380904954e-05,
+      "loss": 0.9823530316352844,
+      "step": 654
+    },
+    {
+      "epoch": 1.7923497267759563,
+      "grad_norm": 1.3912447690963745,
+      "learning_rate": 1.889822493508897e-05,
+      "loss": 0.9858295917510986,
+      "step": 656
+    },
+    {
+      "epoch": 1.7978142076502732,
+      "grad_norm": 0.6013797521591187,
+      "learning_rate": 1.897993106775346e-05,
+      "loss": 1.03643000125885,
+      "step": 658
+    },
+    {
+      "epoch": 1.8032786885245902,
+      "grad_norm": 0.6496340036392212,
+      "learning_rate": 1.9058671844909742e-05,
+      "loss": 1.0535352230072021,
+      "step": 660
+    },
+    {
+      "epoch": 1.8087431693989071,
+      "grad_norm": 0.25427573919296265,
+      "learning_rate": 1.9134418006355532e-05,
+      "loss": 1.0733699798583984,
+      "step": 662
+    },
+    {
+      "epoch": 1.814207650273224,
+      "grad_norm": 0.5028781890869141,
+      "learning_rate": 1.9207141404692667e-05,
+      "loss": 1.0931841135025024,
+      "step": 664
+    },
+    {
+      "epoch": 1.819672131147541,
+      "grad_norm": 0.33322790265083313,
+      "learning_rate": 1.927681501578672e-05,
+      "loss": 1.368470549583435,
+      "step": 666
+    },
+    {
+      "epoch": 1.825136612021858,
+      "grad_norm": 0.48202478885650635,
+      "learning_rate": 1.934341294880924e-05,
+      "loss": 1.1083478927612305,
+      "step": 668
+    },
+    {
+      "epoch": 1.830601092896175,
+      "grad_norm": 0.265699565410614,
+      "learning_rate": 1.9406910455858783e-05,
+      "loss": 0.7845261693000793,
+      "step": 670
+    },
+    {
+      "epoch": 1.8360655737704918,
+      "grad_norm": 0.9111495018005371,
+      "learning_rate": 1.9467283941157304e-05,
+      "loss": 1.3333193063735962,
+      "step": 672
+    },
+    {
+      "epoch": 1.8415300546448088,
+      "grad_norm": 0.41249316930770874,
+      "learning_rate": 1.952451096981838e-05,
+      "loss": 1.0828064680099487,
+      "step": 674
+    },
+    {
+      "epoch": 1.8469945355191257,
+      "grad_norm": 0.34011855721473694,
+      "learning_rate": 1.957857027618405e-05,
+      "loss": 1.128922700881958,
+      "step": 676
+    },
+    {
+      "epoch": 1.8524590163934427,
+      "grad_norm": 0.11010833084583282,
+      "learning_rate": 1.9629441771727166e-05,
+      "loss": 0.27124765515327454,
+      "step": 678
+    },
+    {
+      "epoch": 1.8579234972677594,
+      "grad_norm": 0.34342408180236816,
+      "learning_rate": 1.9677106552516317e-05,
+      "loss": 0.9879894256591797,
+      "step": 680
+    },
+    {
+      "epoch": 1.8633879781420766,
+      "grad_norm": 0.39624279737472534,
+      "learning_rate": 1.9721546906240577e-05,
+      "loss": 1.00787353515625,
+      "step": 682
+    },
+    {
+      "epoch": 1.8688524590163933,
+      "grad_norm": 0.3148435056209564,
+      "learning_rate": 1.976274631879142e-05,
+      "loss": 1.0767008066177368,
+      "step": 684
+    },
+    {
+      "epoch": 1.8743169398907105,
+      "grad_norm": 0.601940393447876,
+      "learning_rate": 1.9800689480399383e-05,
+      "loss": 1.0266121625900269,
+      "step": 686
+    },
+    {
+      "epoch": 1.8797814207650272,
+      "grad_norm": 0.5932305455207825,
+      "learning_rate": 1.9835362291323222e-05,
+      "loss": 1.14261794090271,
+      "step": 688
+    },
+    {
+      "epoch": 1.8852459016393444,
+      "grad_norm": 0.7920188903808594,
+      "learning_rate": 1.9866751867089363e-05,
+      "loss": 1.1080957651138306,
+      "step": 690
+    },
+    {
+      "epoch": 1.890710382513661,
+      "grad_norm": 0.9028313755989075,
+      "learning_rate": 1.9894846543279838e-05,
+      "loss": 0.8868638277053833,
+      "step": 692
+    },
+    {
+      "epoch": 1.8961748633879782,
+      "grad_norm": 0.44754672050476074,
+      "learning_rate": 1.991963587986677e-05,
+      "loss": 1.16123366355896,
+      "step": 694
+    },
+    {
+      "epoch": 1.901639344262295,
+      "grad_norm": 0.3467576205730438,
+      "learning_rate": 1.9941110665091922e-05,
+      "loss": 1.111083745956421,
+      "step": 696
+    },
+    {
+      "epoch": 1.9071038251366121,
+      "grad_norm": 0.8681450486183167,
+      "learning_rate": 1.9959262918889774e-05,
+      "loss": 1.0020962953567505,
+      "step": 698
+    },
+    {
+      "epoch": 1.9125683060109289,
+      "grad_norm": 0.37495821714401245,
+      "learning_rate": 1.9974085895852973e-05,
+      "loss": 1.062408208847046,
+      "step": 700
+    },
+    {
+      "epoch": 1.918032786885246,
+      "grad_norm": 0.728013277053833,
+      "learning_rate": 1.99855740877389e-05,
+      "loss": 1.0249536037445068,
+      "step": 702
+    },
+    {
+      "epoch": 1.9234972677595628,
+      "grad_norm": 0.37011271715164185,
+      "learning_rate": 1.9993723225516553e-05,
+      "loss": 1.1557523012161255,
+      "step": 704
+    },
+    {
+      "epoch": 1.92896174863388,
+      "grad_norm": 0.5814104676246643,
+      "learning_rate": 1.9998530280952938e-05,
+      "loss": 1.1579149961471558,
+      "step": 706
+    },
+    {
+      "epoch": 1.9344262295081966,
+      "grad_norm": 0.2298351526260376,
+      "learning_rate": 1.9999993467738345e-05,
+      "loss": 1.1778736114501953,
+      "step": 708
+    },
+    {
+      "epoch": 1.9398907103825138,
+      "grad_norm": 0.38486695289611816,
+      "learning_rate": 1.9998112242150162e-05,
+      "loss": 0.9057983160018921,
+      "step": 710
+    },
+    {
+      "epoch": 1.9453551912568305,
+      "grad_norm": 0.8062682151794434,
+      "learning_rate": 1.999288730325491e-05,
+      "loss": 0.6416868567466736,
+      "step": 712
+    },
+    {
+      "epoch": 1.9508196721311475,
+      "grad_norm": 0.29092615842819214,
+      "learning_rate": 1.9984320592648474e-05,
+      "loss": 1.1394217014312744,
+      "step": 714
+    },
+    {
+      "epoch": 1.9562841530054644,
+      "grad_norm": 1.3473927974700928,
+      "learning_rate": 1.9972415293734607e-05,
+      "loss": 0.8599758148193359,
+      "step": 716
+    },
+    {
+      "epoch": 1.9617486338797814,
+      "grad_norm": 0.2769167423248291,
+      "learning_rate": 1.995717583054196e-05,
+      "loss": 1.1438926458358765,
+      "step": 718
+    },
+    {
+      "epoch": 1.9672131147540983,
+      "grad_norm": 0.41115716099739075,
+      "learning_rate": 1.9938607866080114e-05,
+      "loss": 1.095086693763733,
+      "step": 720
+    },
+    {
+      "epoch": 1.9726775956284153,
+      "grad_norm": 1.553356409072876,
+      "learning_rate": 1.991671830023521e-05,
+      "loss": 1.1327463388442993,
+      "step": 722
+    },
+    {
+      "epoch": 1.9781420765027322,
+      "grad_norm": 0.21956849098205566,
+      "learning_rate": 1.989151526720591e-05,
+      "loss": 1.078493595123291,
+      "step": 724
+    },
+    {
+      "epoch": 1.9836065573770492,
+      "grad_norm": 0.5972254276275635,
+      "learning_rate": 1.986300813248073e-05,
+      "loss": 1.0689141750335693,
+      "step": 726
+    },
+    {
+      "epoch": 1.989071038251366,
+      "grad_norm": 0.3301437497138977,
+      "learning_rate": 1.9831207489357825e-05,
+      "loss": 1.1041065454483032,
+      "step": 728
+    },
+    {
+      "epoch": 1.994535519125683,
+      "grad_norm": 0.9666980504989624,
+      "learning_rate": 1.979612515500847e-05,
+      "loss": 0.987746000289917,
+      "step": 730
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3877274692058563,
+      "learning_rate": 1.97577741660858e-05,
+      "loss": 1.1460107564926147,
+      "step": 732
+    },
+    {
+      "epoch": 2.0054644808743167,
+      "grad_norm": 0.25538286566734314,
+      "learning_rate": 1.9716168773880382e-05,
+      "loss": 0.6630098223686218,
+      "step": 734
+    },
+    {
+      "epoch": 2.010928961748634,
+      "grad_norm": 0.4147356450557709,
+      "learning_rate": 1.9671324439024374e-05,
+      "loss": 0.8690657615661621,
+      "step": 736
+    },
+    {
+      "epoch": 2.0163934426229506,
+      "grad_norm": 0.4897252023220062,
+      "learning_rate": 1.9623257825746357e-05,
+      "loss": 0.9130824208259583,
+      "step": 738
+    },
+    {
+      "epoch": 2.021857923497268,
+      "grad_norm": 0.4052647352218628,
+      "learning_rate": 1.9571986795678878e-05,
+      "loss": 0.8773000240325928,
+      "step": 740
+    },
+    {
+      "epoch": 2.0273224043715845,
+      "grad_norm": 1.3314751386642456,
+      "learning_rate": 1.951753040122102e-05,
+      "loss": 0.8227751851081848,
+      "step": 742
+    },
+    {
+      "epoch": 2.0327868852459017,
+      "grad_norm": 3.8880512714385986,
+      "learning_rate": 1.9459908878458532e-05,
+      "loss": 1.0034403800964355,
+      "step": 744
+    },
+    {
+      "epoch": 2.0382513661202184,
+      "grad_norm": 0.4356456995010376,
+      "learning_rate": 1.939914363964402e-05,
+      "loss": 0.8533210754394531,
+      "step": 746
+    },
+    {
+      "epoch": 2.0437158469945356,
+      "grad_norm": 0.41386353969573975,
+      "learning_rate": 1.9335257265240168e-05,
+      "loss": 0.9781957864761353,
+      "step": 748
+    },
+    {
+      "epoch": 2.0491803278688523,
+      "grad_norm": 2.5565185546875,
+      "learning_rate": 1.9268273495528768e-05,
+      "loss": 0.6280704736709595,
+      "step": 750
+    },
+    {
+      "epoch": 2.0546448087431695,
+      "grad_norm": 0.46574458479881287,
+      "learning_rate": 1.9198217221788806e-05,
+      "loss": 0.6308586597442627,
+      "step": 752
+    },
+    {
+      "epoch": 2.060109289617486,
+      "grad_norm": 2.530082941055298,
+      "learning_rate": 1.9125114477046807e-05,
+      "loss": 0.6913048028945923,
+      "step": 754
+    },
+    {
+      "epoch": 2.0655737704918034,
+      "grad_norm": 0.8439071178436279,
+      "learning_rate": 1.9048992426402947e-05,
+      "loss": 1.3085075616836548,
+      "step": 756
+    },
+    {
+      "epoch": 2.07103825136612,
+      "grad_norm": 1.0051873922348022,
+      "learning_rate": 1.896987935693643e-05,
+      "loss": 1.0452558994293213,
+      "step": 758
+    },
+    {
+      "epoch": 2.0765027322404372,
+      "grad_norm": 0.3603721261024475,
+      "learning_rate": 1.888780466719397e-05,
+      "loss": 0.9518926739692688,
+      "step": 760
+    },
+    {
+      "epoch": 2.081967213114754,
+      "grad_norm": 0.6717904210090637,
+      "learning_rate": 1.8802798856265254e-05,
+      "loss": 1.1386586427688599,
+      "step": 762
+    },
+    {
+      "epoch": 2.087431693989071,
+      "grad_norm": 1.6152324676513672,
+      "learning_rate": 1.8714893512449424e-05,
+      "loss": 0.8252012729644775,
+      "step": 764
+    },
+    {
+      "epoch": 2.092896174863388,
+      "grad_norm": 0.29463207721710205,
+      "learning_rate": 1.8624121301516808e-05,
+      "loss": 1.0371973514556885,
+      "step": 766
+    },
+    {
+      "epoch": 2.098360655737705,
+      "grad_norm": 0.48132818937301636,
+      "learning_rate": 1.853051595457026e-05,
+      "loss": 0.7042374610900879,
+      "step": 768
+    },
+    {
+      "epoch": 2.1038251366120218,
+      "grad_norm": 0.4673232436180115,
+      "learning_rate": 1.843411225551065e-05,
+      "loss": 0.9674187302589417,
+      "step": 770
+    },
+    {
+      "epoch": 2.109289617486339,
+      "grad_norm": 3.6729300022125244,
+      "learning_rate": 1.8334946028111088e-05,
+      "loss": 0.9104331135749817,
+      "step": 772
+    },
+    {
+      "epoch": 2.1147540983606556,
+      "grad_norm": 0.38922831416130066,
+      "learning_rate": 1.8233054122704765e-05,
+      "loss": 0.8225538730621338,
+      "step": 774
+    },
+    {
+      "epoch": 2.120218579234973,
+      "grad_norm": 0.29239991307258606,
+      "learning_rate": 1.8128474402491286e-05,
+      "loss": 0.8900973200798035,
+      "step": 776
+    },
+    {
+      "epoch": 2.1256830601092895,
+      "grad_norm": 1.132876992225647,
+      "learning_rate": 1.802124572946668e-05,
+      "loss": 0.6526266932487488,
+      "step": 778
+    },
+    {
+      "epoch": 2.1311475409836067,
+      "grad_norm": 0.3578123450279236,
+      "learning_rate": 1.791140794998219e-05,
+      "loss": 0.673201858997345,
+      "step": 780
+    },
+    {
+      "epoch": 2.1366120218579234,
+      "grad_norm": 0.305058091878891,
+      "learning_rate": 1.7799001879937294e-05,
+      "loss": 0.8488634824752808,
+      "step": 782
+    },
+    {
+      "epoch": 2.1420765027322406,
+      "grad_norm": 0.27180346846580505,
+      "learning_rate": 1.768406928961248e-05,
+      "loss": 1.0290989875793457,
+      "step": 784
+    },
+    {
+      "epoch": 2.1475409836065573,
+      "grad_norm": 0.129277765750885,
+      "learning_rate": 1.7566652888147328e-05,
+      "loss": 0.646569550037384,
+      "step": 786
+    },
+    {
+      "epoch": 2.1530054644808745,
+      "grad_norm": 7.248612403869629,
+      "learning_rate": 1.7446796307669725e-05,
+      "loss": 0.6828120350837708,
+      "step": 788
+    },
+    {
+      "epoch": 2.158469945355191,
+      "grad_norm": 0.5306554436683655,
+      "learning_rate": 1.732454408708209e-05,
+      "loss": 0.9795069098472595,
+      "step": 790
+    },
+    {
+      "epoch": 2.1639344262295084,
+      "grad_norm": 0.3850613236427307,
+      "learning_rate": 1.719994165551063e-05,
+      "loss": 0.611956000328064,
+      "step": 792
+    },
+    {
+      "epoch": 2.169398907103825,
+      "grad_norm": 2.014388084411621,
+      "learning_rate": 1.7073035315423838e-05,
+      "loss": 0.8363724946975708,
+      "step": 794
+    },
+    {
+      "epoch": 2.1748633879781423,
+      "grad_norm": 0.5221168398857117,
+      "learning_rate": 1.6943872225426396e-05,
+      "loss": 0.8072232007980347,
+      "step": 796
+    },
+    {
+      "epoch": 2.180327868852459,
+      "grad_norm": 0.2606959044933319,
+      "learning_rate": 1.6812500382734977e-05,
+      "loss": 0.8440264463424683,
+      "step": 798
+    },
+    {
+      "epoch": 2.185792349726776,
+      "grad_norm": 1.0706250667572021,
+      "learning_rate": 1.6678968605342348e-05,
+      "loss": 1.0057344436645508,
+      "step": 800
+    },
+    {
+      "epoch": 2.191256830601093,
+      "grad_norm": 0.4442984163761139,
+      "learning_rate": 1.6543326513876602e-05,
+      "loss": 0.948396623134613,
+      "step": 802
+    },
+    {
+      "epoch": 2.19672131147541,
+      "grad_norm": 0.4648842215538025,
+      "learning_rate": 1.6405624513162002e-05,
+      "loss": 1.0737464427947998,
+      "step": 804
+    },
+    {
+      "epoch": 2.202185792349727,
+      "grad_norm": 0.2643176019191742,
+      "learning_rate": 1.6265913773488456e-05,
+      "loss": 0.9078125357627869,
+      "step": 806
+    },
+    {
+      "epoch": 2.2076502732240435,
+      "grad_norm": 0.3123255670070648,
+      "learning_rate": 1.6124246211596606e-05,
+      "loss": 0.5129435062408447,
+      "step": 808
+    },
+    {
+      "epoch": 2.2131147540983607,
+      "grad_norm": 0.33814558386802673,
+      "learning_rate": 1.598067447138542e-05,
+      "loss": 0.6270028352737427,
+      "step": 810
+    },
+    {
+      "epoch": 2.2185792349726774,
+      "grad_norm": 0.2907693684101105,
+      "learning_rate": 1.5835251904349688e-05,
+      "loss": 0.8713023662567139,
+      "step": 812
+    },
+    {
+      "epoch": 2.2240437158469946,
+      "grad_norm": 0.32165563106536865,
+      "learning_rate": 1.5688032549754453e-05,
+      "loss": 0.7885038256645203,
+      "step": 814
+    },
+    {
+      "epoch": 2.2295081967213113,
+      "grad_norm": 0.33382898569107056,
+      "learning_rate": 1.553907111455401e-05,
+      "loss": 0.847852885723114,
+      "step": 816
+    },
+    {
+      "epoch": 2.2349726775956285,
+      "grad_norm": 0.31947633624076843,
+      "learning_rate": 1.538842295306264e-05,
+      "loss": 0.9458706378936768,
+      "step": 818
+    },
+    {
+      "epoch": 2.240437158469945,
+      "grad_norm": 0.444782018661499,
+      "learning_rate": 1.5236144046384917e-05,
+      "loss": 0.9850488305091858,
+      "step": 820
+    },
+    {
+      "epoch": 2.2459016393442623,
+      "grad_norm": 0.9171160459518433,
+      "learning_rate": 1.5082290981612987e-05,
+      "loss": 1.1842435598373413,
+      "step": 822
+    },
+    {
+      "epoch": 2.251366120218579,
+      "grad_norm": 0.29313865303993225,
+      "learning_rate": 1.4926920930798736e-05,
+      "loss": 0.8148974776268005,
+      "step": 824
+    },
+    {
+      "epoch": 2.2568306010928962,
+      "grad_norm": 1.1459228992462158,
+      "learning_rate": 1.4770091629708562e-05,
+      "loss": 0.6716075539588928,
+      "step": 826
+    },
+    {
+      "epoch": 2.262295081967213,
+      "grad_norm": 0.2907215654850006,
+      "learning_rate": 1.461186135636868e-05,
+      "loss": 0.5930277109146118,
+      "step": 828
+    },
+    {
+      "epoch": 2.26775956284153,
+      "grad_norm": 0.36279723048210144,
+      "learning_rate": 1.4452288909408864e-05,
+      "loss": 0.803901195526123,
+      "step": 830
+    },
+    {
+      "epoch": 2.273224043715847,
+      "grad_norm": 0.987859308719635,
+      "learning_rate": 1.4291433586212831e-05,
+      "loss": 0.586320698261261,
+      "step": 832
+    },
+    {
+      "epoch": 2.278688524590164,
+      "grad_norm": 0.4347521662712097,
+      "learning_rate": 1.4129355160883216e-05,
+      "loss": 0.7422566413879395,
+      "step": 834
+    },
+    {
+      "epoch": 2.2841530054644807,
+      "grad_norm": 0.6859447360038757,
+      "learning_rate": 1.3966113862029429e-05,
+      "loss": 0.9895762205123901,
+      "step": 836
+    },
+    {
+      "epoch": 2.289617486338798,
+      "grad_norm": 1.1535533666610718,
+      "learning_rate": 1.3801770350386568e-05,
+      "loss": 0.9528064131736755,
+      "step": 838
+    },
+    {
+      "epoch": 2.2950819672131146,
+      "grad_norm": 0.333217978477478,
+      "learning_rate": 1.363638569627384e-05,
+      "loss": 0.6650771498680115,
+      "step": 840
+    },
+    {
+      "epoch": 2.300546448087432,
+      "grad_norm": 0.3299920856952667,
+      "learning_rate": 1.3470021356900696e-05,
+      "loss": 0.9850642681121826,
+      "step": 842
+    },
+    {
+      "epoch": 2.3060109289617485,
+      "grad_norm": 0.2731953263282776,
+      "learning_rate": 1.3302739153529252e-05,
+      "loss": 1.0595612525939941,
+      "step": 844
+    },
+    {
+      "epoch": 2.3114754098360657,
+      "grad_norm": 0.20923754572868347,
+      "learning_rate": 1.3134601248501366e-05,
+      "loss": 0.5519586801528931,
+      "step": 846
+    },
+    {
+      "epoch": 2.3169398907103824,
+      "grad_norm": 0.3694831132888794,
+      "learning_rate": 1.2965670122139071e-05,
+      "loss": 0.9783721566200256,
+      "step": 848
+    },
+    {
+      "epoch": 2.3224043715846996,
+      "grad_norm": 0.2181791067123413,
+      "learning_rate": 1.2796008549526752e-05,
+      "loss": 0.8303213119506836,
+      "step": 850
+    },
+    {
+      "epoch": 2.3278688524590163,
+      "grad_norm": 0.5311651229858398,
+      "learning_rate": 1.262567957718378e-05,
+      "loss": 0.7890332341194153,
+      "step": 852
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 0.6432619690895081,
+      "learning_rate": 1.2454746499636408e-05,
+      "loss": 0.9261720776557922,
+      "step": 854
+    },
+    {
+      "epoch": 2.33879781420765,
+      "grad_norm": 0.23040394484996796,
+      "learning_rate": 1.2283272835897359e-05,
+      "loss": 0.8952475190162659,
+      "step": 856
+    },
+    {
+      "epoch": 2.3442622950819674,
+      "grad_norm": 0.4025140404701233,
+      "learning_rate": 1.2111322305862088e-05,
+      "loss": 0.641585648059845,
+      "step": 858
+    },
+    {
+      "epoch": 2.349726775956284,
+      "grad_norm": 0.2443774938583374,
+      "learning_rate": 1.1938958806630322e-05,
+      "loss": 0.7296821475028992,
+      "step": 860
+    },
+    {
+      "epoch": 2.3551912568306013,
+      "grad_norm": 0.3356713652610779,
+      "learning_rate": 1.1766246388761841e-05,
+      "loss": 0.8933297395706177,
+      "step": 862
+    },
+    {
+      "epoch": 2.360655737704918,
+      "grad_norm": 0.40507715940475464,
+      "learning_rate": 1.1593249232475162e-05,
+      "loss": 0.5992748737335205,
+      "step": 864
+    },
+    {
+      "epoch": 2.366120218579235,
+      "grad_norm": 0.8139327168464661,
+      "learning_rate": 1.142003162379808e-05,
+      "loss": 1.2816228866577148,
+      "step": 866
+    },
+    {
+      "epoch": 2.371584699453552,
+      "grad_norm": 0.22665131092071533,
+      "learning_rate": 1.1246657930678817e-05,
+      "loss": 0.9113897085189819,
+      "step": 868
+    },
+    {
+      "epoch": 2.3770491803278686,
+      "grad_norm": 0.2706243097782135,
+      "learning_rate": 1.1073192579066867e-05,
+      "loss": 0.7394488453865051,
+      "step": 870
+    },
+    {
+      "epoch": 2.3825136612021858,
+      "grad_norm": 0.2607859671115875,
+      "learning_rate": 1.0899700028972169e-05,
+      "loss": 0.8141605257987976,
+      "step": 872
+    },
+    {
+      "epoch": 2.387978142076503,
+      "grad_norm": 0.3351480960845947,
+      "learning_rate": 1.072624475051166e-05,
+      "loss": 0.9203171730041504,
+      "step": 874
+    },
+    {
+      "epoch": 2.3934426229508197,
+      "grad_norm": 2.871534824371338,
+      "learning_rate": 1.055289119995206e-05,
+      "loss": 0.7342777848243713,
+      "step": 876
+    },
+    {
+      "epoch": 2.3989071038251364,
+      "grad_norm": 0.31686869263648987,
+      "learning_rate": 1.0379703795757853e-05,
+      "loss": 0.822813093662262,
+      "step": 878
+    },
+    {
+      "epoch": 2.4043715846994536,
+      "grad_norm": 0.3635275363922119,
+      "learning_rate": 1.0206746894653252e-05,
+      "loss": 1.01185941696167,
+      "step": 880
+    },
+    {
+      "epoch": 2.4098360655737707,
+      "grad_norm": 0.2960561215877533,
+      "learning_rate": 1.0034084767707164e-05,
+      "loss": 0.491107702255249,
+      "step": 882
+    },
+    {
+      "epoch": 2.4153005464480874,
+      "grad_norm": 0.32399389147758484,
+      "learning_rate": 9.861781576449879e-06,
+      "loss": 0.7038915157318115,
+      "step": 884
+    },
+    {
+      "epoch": 2.420765027322404,
+      "grad_norm": 0.3570184111595154,
+      "learning_rate": 9.689901349030646e-06,
+      "loss": 0.8117411136627197,
+      "step": 886
+    },
+    {
+      "epoch": 2.4262295081967213,
+      "grad_norm": 0.3321293890476227,
+      "learning_rate": 9.518507956424643e-06,
+      "loss": 0.6311563849449158,
+      "step": 888
+    },
+    {
+      "epoch": 2.431693989071038,
+      "grad_norm": 0.3352676331996918,
+      "learning_rate": 9.347665088698444e-06,
+      "loss": 0.5507946610450745,
+      "step": 890
+    },
+    {
+      "epoch": 2.4371584699453552,
+      "grad_norm": 0.7419760227203369,
+      "learning_rate": 9.177436231342623e-06,
+      "loss": 0.8605806827545166,
+      "step": 892
+    },
+    {
+      "epoch": 2.442622950819672,
+      "grad_norm": 0.3643116056919098,
+      "learning_rate": 9.00788464168054e-06,
+      "loss": 0.49560198187828064,
+      "step": 894
+    },
+    {
+      "epoch": 2.448087431693989,
+      "grad_norm": 0.32895293831825256,
+      "learning_rate": 8.839073325361751e-06,
+      "loss": 0.8086169958114624,
+      "step": 896
+    },
+    {
+      "epoch": 2.453551912568306,
+      "grad_norm": 0.9999696016311646,
+      "learning_rate": 8.67106501294902e-06,
+      "loss": 0.7644316554069519,
+      "step": 898
+    },
+    {
+      "epoch": 2.459016393442623,
+      "grad_norm": 0.32013651728630066,
+      "learning_rate": 8.503922136607536e-06,
+      "loss": 0.8786826133728027,
+      "step": 900
+    },
+    {
+      "epoch": 2.4644808743169397,
+      "grad_norm": 0.34828099608421326,
+      "learning_rate": 8.337706806905029e-06,
+      "loss": 0.9184771776199341,
+      "step": 902
+    },
+    {
+      "epoch": 2.469945355191257,
+      "grad_norm": 0.21218842267990112,
+      "learning_rate": 8.172480789731374e-06,
+      "loss": 0.7311652302742004,
+      "step": 904
+    },
+    {
+      "epoch": 2.4754098360655736,
+      "grad_norm": 0.5770183205604553,
+      "learning_rate": 8.00830548334625e-06,
+      "loss": 0.972994327545166,
+      "step": 906
+    },
+    {
+      "epoch": 2.480874316939891,
+      "grad_norm": 0.15764155983924866,
+      "learning_rate": 7.84524189556352e-06,
+      "loss": 0.4051420986652374,
+      "step": 908
+    },
+    {
+      "epoch": 2.4863387978142075,
+      "grad_norm": 0.23087617754936218,
+      "learning_rate": 7.68335062108057e-06,
+      "loss": 0.6997837424278259,
+      "step": 910
+    },
+    {
+      "epoch": 2.4918032786885247,
+      "grad_norm": 0.3418113887310028,
+      "learning_rate": 7.522691818961252e-06,
+      "loss": 0.42771270871162415,
+      "step": 912
+    },
+    {
+      "epoch": 2.4972677595628414,
+      "grad_norm": 0.38769078254699707,
+      "learning_rate": 7.3633251902806165e-06,
+      "loss": 0.9568927884101868,
+      "step": 914
+    },
+    {
+      "epoch": 2.5027322404371586,
+      "grad_norm": 0.30056893825531006,
+      "learning_rate": 7.205309955939983e-06,
+      "loss": 0.7710816264152527,
+      "step": 916
+    },
+    {
+      "epoch": 2.5081967213114753,
+      "grad_norm": 0.3079778850078583,
+      "learning_rate": 7.048704834660296e-06,
+      "loss": 0.7985804677009583,
+      "step": 918
+    },
+    {
+      "epoch": 2.5136612021857925,
+      "grad_norm": 0.29017210006713867,
+      "learning_rate": 6.8935680211621715e-06,
+      "loss": 0.695990264415741,
+      "step": 920
+    },
+    {
+      "epoch": 2.519125683060109,
+      "grad_norm": 0.24671095609664917,
+      "learning_rate": 6.739957164540634e-06,
+      "loss": 0.7687212228775024,
+      "step": 922
+    },
+    {
+      "epoch": 2.5245901639344264,
+      "grad_norm": 0.2934419512748718,
+      "learning_rate": 6.587929346842625e-06,
+      "loss": 0.5792368650436401,
+      "step": 924
+    },
+    {
+      "epoch": 2.530054644808743,
+      "grad_norm": 0.43741247057914734,
+      "learning_rate": 6.437541061855222e-06,
+      "loss": 0.550622820854187,
+      "step": 926
+    },
+    {
+      "epoch": 2.5355191256830603,
+      "grad_norm": 0.28486597537994385,
+      "learning_rate": 6.288848194112459e-06,
+      "loss": 0.6305705904960632,
+      "step": 928
+    },
+    {
+      "epoch": 2.540983606557377,
+      "grad_norm": 0.3551803529262543,
+      "learning_rate": 6.141905998128495e-06,
+      "loss": 0.9627759456634521,
+      "step": 930
+    },
+    {
+      "epoch": 2.546448087431694,
+      "grad_norm": 0.34825268387794495,
+      "learning_rate": 5.996769077865029e-06,
+      "loss": 0.6931189894676208,
+      "step": 932
+    },
+    {
+      "epoch": 2.551912568306011,
+      "grad_norm": 0.39621520042419434,
+      "learning_rate": 5.853491366440313e-06,
+      "loss": 0.8206515908241272,
+      "step": 934
+    },
+    {
+      "epoch": 2.557377049180328,
+      "grad_norm": 6.864598274230957,
+      "learning_rate": 5.712126106087557e-06,
+      "loss": 0.6882966160774231,
+      "step": 936
+    },
+    {
+      "epoch": 2.5628415300546448,
+      "grad_norm": 0.7489693760871887,
+      "learning_rate": 5.572725828369961e-06,
+      "loss": 0.5850929021835327,
+      "step": 938
+    },
+    {
+      "epoch": 2.5683060109289615,
+      "grad_norm": 0.5904379487037659,
+      "learning_rate": 5.4353423346599944e-06,
+      "loss": 1.0272624492645264,
+      "step": 940
+    },
+    {
+      "epoch": 2.5737704918032787,
+      "grad_norm": 0.33465638756752014,
+      "learning_rate": 5.30002667688986e-06,
+      "loss": 0.5956709980964661,
+      "step": 942
+    },
+    {
+      "epoch": 2.579234972677596,
+      "grad_norm": 0.49614542722702026,
+      "learning_rate": 5.1668291385804995e-06,
+      "loss": 0.9119790196418762,
+      "step": 944
+    },
+    {
+      "epoch": 2.5846994535519126,
+      "grad_norm": 0.18383480608463287,
+      "learning_rate": 5.03579921615621e-06,
+      "loss": 0.339104026556015,
+      "step": 946
+    },
+    {
+      "epoch": 2.5901639344262293,
+      "grad_norm": 0.39025598764419556,
+      "learning_rate": 4.906985600551651e-06,
+      "loss": 0.7046570777893066,
+      "step": 948
+    },
+    {
+      "epoch": 2.5956284153005464,
+      "grad_norm": 0.32544654607772827,
+      "learning_rate": 4.780436159118221e-06,
+      "loss": 0.7014875411987305,
+      "step": 950
+    },
+    {
+      "epoch": 2.6010928961748636,
+      "grad_norm": 0.4727187752723694,
+      "learning_rate": 4.656197917836474e-06,
+      "loss": 0.9602597951889038,
+      "step": 952
+    },
+    {
+      "epoch": 2.6065573770491803,
+      "grad_norm": 0.5159199237823486,
+      "learning_rate": 4.5343170438411885e-06,
+      "loss": 0.45191124081611633,
+      "step": 954
+    },
+    {
+      "epoch": 2.612021857923497,
+      "grad_norm": 0.42049863934516907,
+      "learning_rate": 4.414838828265581e-06,
+      "loss": 0.2999444603919983,
+      "step": 956
+    },
+    {
+      "epoch": 2.6174863387978142,
+      "grad_norm": 0.2952582538127899,
+      "learning_rate": 4.297807669411057e-06,
+      "loss": 0.518054187297821,
+      "step": 958
+    },
+    {
+      "epoch": 2.6229508196721314,
+      "grad_norm": 0.30971741676330566,
+      "learning_rate": 4.183267056248689e-06,
+      "loss": 0.763922929763794,
+      "step": 960
+    },
+    {
+      "epoch": 2.628415300546448,
+      "grad_norm": 0.3294529616832733,
+      "learning_rate": 4.071259552258709e-06,
+      "loss": 0.6879506707191467,
+      "step": 962
+    },
+    {
+      "epoch": 2.633879781420765,
+      "grad_norm": 0.4937688708305359,
+      "learning_rate": 3.961826779613801e-06,
+      "loss": 0.9857227802276611,
+      "step": 964
+    },
+    {
+      "epoch": 2.639344262295082,
+      "grad_norm": 0.3687383532524109,
+      "learning_rate": 3.85500940371226e-06,
+      "loss": 0.6435363292694092,
+      "step": 966
+    },
+    {
+      "epoch": 2.644808743169399,
+      "grad_norm": 0.3145747184753418,
+      "learning_rate": 3.750847118066614e-06,
+      "loss": 0.6846204400062561,
+      "step": 968
+    },
+    {
+      "epoch": 2.650273224043716,
+      "grad_norm": 0.8533056974411011,
+      "learning_rate": 3.6493786295535234e-06,
+      "loss": 0.8020285367965698,
+      "step": 970
+    },
+    {
+      "epoch": 2.6557377049180326,
+      "grad_norm": 0.38222646713256836,
+      "learning_rate": 3.5506416440301885e-06,
+      "loss": 1.064414143562317,
+      "step": 972
+    },
+    {
+      "epoch": 2.66120218579235,
+      "grad_norm": 0.7065198421478271,
+      "learning_rate": 3.4546728523228067e-06,
+      "loss": 0.6177721619606018,
+      "step": 974
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.28936606645584106,
+      "learning_rate": 3.361507916592206e-06,
+      "loss": 0.8083776831626892,
+      "step": 976
+    },
+    {
+      "epoch": 2.6721311475409837,
+      "grad_norm": 0.21318885684013367,
+      "learning_rate": 3.271181457081715e-06,
+      "loss": 0.7788843512535095,
+      "step": 978
+    },
+    {
+      "epoch": 2.6775956284153004,
+      "grad_norm": 0.35676872730255127,
+      "learning_rate": 3.1837270392522456e-06,
+      "loss": 0.8598240613937378,
+      "step": 980
+    },
+    {
+      "epoch": 2.6830601092896176,
+      "grad_norm": 0.2766752243041992,
+      "learning_rate": 3.0991771613092686e-06,
+      "loss": 0.9411115050315857,
+      "step": 982
+    },
+    {
+      "epoch": 2.6885245901639343,
+      "grad_norm": 0.2711624801158905,
+      "learning_rate": 3.017563242126483e-06,
+      "loss": 0.6280143857002258,
+      "step": 984
+    },
+    {
+      "epoch": 2.6939890710382515,
+      "grad_norm": 0.2689661681652069,
+      "learning_rate": 2.9389156095704764e-06,
+      "loss": 0.9351664185523987,
+      "step": 986
+    },
+    {
+      "epoch": 2.699453551912568,
+      "grad_norm": 0.962471067905426,
+      "learning_rate": 2.8632634892308535e-06,
+      "loss": 0.8302200436592102,
+      "step": 988
+    },
+    {
+      "epoch": 2.7049180327868854,
+      "grad_norm": 0.39670348167419434,
+      "learning_rate": 2.7906349935599326e-06,
+      "loss": 0.8161414265632629,
+      "step": 990
+    },
+    {
+      "epoch": 2.710382513661202,
+      "grad_norm": 0.35174086689949036,
+      "learning_rate": 2.721057111426154e-06,
+      "loss": 0.5259367227554321,
+      "step": 992
+    },
+    {
+      "epoch": 2.7158469945355193,
+      "grad_norm": 0.2511337399482727,
+      "learning_rate": 2.6545556980849417e-06,
+      "loss": 0.9182239770889282,
+      "step": 994
+    },
+    {
+      "epoch": 2.721311475409836,
+      "grad_norm": 0.21024306118488312,
+      "learning_rate": 2.591155465570866e-06,
+      "loss": 0.8048747181892395,
+      "step": 996
+    },
+    {
+      "epoch": 2.726775956284153,
+      "grad_norm": 0.2689720690250397,
+      "learning_rate": 2.5308799735145813e-06,
+      "loss": 0.7127807140350342,
+      "step": 998
+    },
+    {
+      "epoch": 2.73224043715847,
+      "grad_norm": 0.7397425770759583,
+      "learning_rate": 2.473751620388069e-06,
+      "loss": 1.054977536201477,
+      "step": 1000
+    },
+    {
+      "epoch": 2.737704918032787,
+      "grad_norm": 0.28134724497795105,
+      "learning_rate": 2.419791635181301e-06,
+      "loss": 0.8681496381759644,
+      "step": 1002
+    },
+    {
+      "epoch": 2.7431693989071038,
+      "grad_norm": 0.22395454347133636,
+      "learning_rate": 2.369020069513521e-06,
+      "loss": 0.6971544027328491,
+      "step": 1004
+    },
+    {
+      "epoch": 2.748633879781421,
+      "grad_norm": 0.273333877325058,
+      "learning_rate": 2.3214557901820258e-06,
+      "loss": 0.35114356875419617,
+      "step": 1006
+    },
+    {
+      "epoch": 2.7540983606557377,
+      "grad_norm": 0.33767545223236084,
+      "learning_rate": 2.27711647215124e-06,
+      "loss": 0.9685407280921936,
+      "step": 1008
+    },
+    {
+      "epoch": 2.7595628415300544,
+      "grad_norm": 0.3212963938713074,
+      "learning_rate": 2.2360185919846593e-06,
+      "loss": 1.0383319854736328,
+      "step": 1010
+    },
+    {
+      "epoch": 2.7650273224043715,
+      "grad_norm": 0.4074728190898895,
+      "learning_rate": 2.1981774217221474e-06,
+      "loss": 0.7557249069213867,
+      "step": 1012
+    },
+    {
+      "epoch": 2.7704918032786887,
+      "grad_norm": 0.3645928204059601,
+      "learning_rate": 2.1636070232047966e-06,
+      "loss": 0.9979292154312134,
+      "step": 1014
+    },
+    {
+      "epoch": 2.7759562841530054,
+      "grad_norm": 0.5972633361816406,
+      "learning_rate": 2.1323202428495544e-06,
+      "loss": 0.7092854380607605,
+      "step": 1016
+    },
+    {
+      "epoch": 2.781420765027322,
+      "grad_norm": 0.33370643854141235,
+      "learning_rate": 2.104328706875452e-06,
+      "loss": 0.9313384890556335,
+      "step": 1018
+    },
+    {
+      "epoch": 2.7868852459016393,
+      "grad_norm": 0.4620024263858795,
+      "learning_rate": 2.079642816983293e-06,
+      "loss": 0.6462956666946411,
+      "step": 1020
+    },
+    {
+      "epoch": 2.7923497267759565,
+      "grad_norm": 0.3582730293273926,
+      "learning_rate": 2.0582717464903546e-06,
+      "loss": 0.8746689558029175,
+      "step": 1022
+    },
+    {
+      "epoch": 2.797814207650273,
+      "grad_norm": 0.618171751499176,
+      "learning_rate": 2.040223436921581e-06,
+      "loss": 0.4330008327960968,
+      "step": 1024
+    },
+    {
+      "epoch": 2.80327868852459,
+      "grad_norm": 0.33677467703819275,
+      "learning_rate": 2.025504595058489e-06,
+      "loss": 0.9515740275382996,
+      "step": 1026
+    },
+    {
+      "epoch": 2.808743169398907,
+      "grad_norm": 1.293221354484558,
+      "learning_rate": 2.0141206904469206e-06,
+      "loss": 0.6902921795845032,
+      "step": 1028
+    },
+    {
+      "epoch": 2.8142076502732243,
+      "grad_norm": 0.2411491721868515,
+      "learning_rate": 2.006075953364551e-06,
+      "loss": 0.7488702535629272,
+      "step": 1030
+    },
+    {
+      "epoch": 2.819672131147541,
+      "grad_norm": 0.27980777621269226,
+      "learning_rate": 2.0013733732489103e-06,
+      "loss": 0.7026211619377136,
+      "step": 1032
+    },
+    {
+      "epoch": 2.8251366120218577,
+      "grad_norm": 0.5924912095069885,
+      "learning_rate": 2.000014697586502e-06,
+      "loss": 0.7316262125968933,
+      "step": 1034
+    },
+    {
+      "epoch": 2.830601092896175,
+      "grad_norm": 0.7772489190101624,
+      "learning_rate": 2.0020004312634374e-06,
+      "loss": 1.0607091188430786,
+      "step": 1036
+    },
+    {
+      "epoch": 2.836065573770492,
+      "grad_norm": 0.23920530080795288,
+      "learning_rate": 2.0073298363778166e-06,
+      "loss": 0.7372963428497314,
+      "step": 1038
+    },
+    {
+      "epoch": 2.841530054644809,
+      "grad_norm": 0.32360178232192993,
+      "learning_rate": 2.016000932513934e-06,
+      "loss": 0.8663780093193054,
+      "step": 1040
+    },
+    {
+      "epoch": 2.8469945355191255,
+      "grad_norm": 0.2977835536003113,
+      "learning_rate": 2.0280104974782058e-06,
+      "loss": 0.8212502002716064,
+      "step": 1042
+    },
+    {
+      "epoch": 2.8524590163934427,
+      "grad_norm": 0.29036644101142883,
+      "learning_rate": 2.043354068496541e-06,
+      "loss": 0.7420258522033691,
+      "step": 1044
+    },
+    {
+      "epoch": 2.8579234972677594,
+      "grad_norm": 0.3778425455093384,
+      "learning_rate": 2.0620259438727168e-06,
+      "loss": 0.9037473797798157,
+      "step": 1046
+    },
+    {
+      "epoch": 2.8633879781420766,
+      "grad_norm": 0.11288527399301529,
+      "learning_rate": 2.084019185107135e-06,
+      "loss": 0.38631874322891235,
+      "step": 1048
+    },
+    {
+      "epoch": 2.8688524590163933,
+      "grad_norm": 0.42890432476997375,
+      "learning_rate": 2.1093256194751822e-06,
+      "loss": 0.787931501865387,
+      "step": 1050
+    },
+    {
+      "epoch": 2.8743169398907105,
+      "grad_norm": 0.3188338279724121,
+      "learning_rate": 2.137935843064233e-06,
+      "loss": 0.7466610074043274,
+      "step": 1052
+    },
+    {
+      "epoch": 2.879781420765027,
+      "grad_norm": 0.3111341595649719,
+      "learning_rate": 2.1698392242681502e-06,
+      "loss": 0.966293454170227,
+      "step": 1054
+    },
+    {
+      "epoch": 2.8852459016393444,
+      "grad_norm": 0.7533649802207947,
+      "learning_rate": 2.2050239077380097e-06,
+      "loss": 0.6824079155921936,
+      "step": 1056
+    },
+    {
+      "epoch": 2.890710382513661,
+      "grad_norm": 2.3769004344940186,
+      "learning_rate": 2.2434768187875723e-06,
+      "loss": 0.5979699492454529,
+      "step": 1058
+    },
+    {
+      "epoch": 2.8961748633879782,
+      "grad_norm": 0.33416691422462463,
+      "learning_rate": 2.285183668251853e-06,
+      "loss": 0.8557246327400208,
+      "step": 1060
+    },
+    {
+      "epoch": 2.901639344262295,
+      "grad_norm": 0.44861775636672974,
+      "learning_rate": 2.3301289577970028e-06,
+      "loss": 0.8509962558746338,
+      "step": 1062
+    },
+    {
+      "epoch": 2.907103825136612,
+      "grad_norm": 0.28698277473449707,
+      "learning_rate": 2.3782959856795113e-06,
+      "loss": 0.8196679353713989,
+      "step": 1064
+    },
+    {
+      "epoch": 2.912568306010929,
+      "grad_norm": 0.17325085401535034,
+      "learning_rate": 2.4296668529525998e-06,
+      "loss": 0.46036872267723083,
+      "step": 1066
+    },
+    {
+      "epoch": 2.918032786885246,
+      "grad_norm": 0.31135252118110657,
+      "learning_rate": 2.4842224701175147e-06,
+      "loss": 0.6244635581970215,
+      "step": 1068
+    },
+    {
+      "epoch": 2.9234972677595628,
+      "grad_norm": 0.5156043171882629,
+      "learning_rate": 2.541942564217196e-06,
+      "loss": 0.7163093090057373,
+      "step": 1070
+    },
+    {
+      "epoch": 2.92896174863388,
+      "grad_norm": 0.2627166509628296,
+      "learning_rate": 2.6028056863697506e-06,
+      "loss": 0.9640206694602966,
+      "step": 1072
+    },
+    {
+      "epoch": 2.9344262295081966,
+      "grad_norm": 0.42738401889801025,
+      "learning_rate": 2.6667892197388884e-06,
+      "loss": 0.5566335916519165,
+      "step": 1074
+    },
+    {
+      "epoch": 2.939890710382514,
+      "grad_norm": 0.6010952591896057,
+      "learning_rate": 2.7338693879383967e-06,
+      "loss": 0.48107990622520447,
+      "step": 1076
+    },
+    {
+      "epoch": 2.9453551912568305,
+      "grad_norm": 0.4749331474304199,
+      "learning_rate": 2.8040212638674506e-06,
+      "loss": 0.5274478793144226,
+      "step": 1078
+    },
+    {
+      "epoch": 2.9508196721311473,
+      "grad_norm": 0.38017284870147705,
+      "learning_rate": 2.877218778973578e-06,
+      "loss": 0.9354230165481567,
+      "step": 1080
+    },
+    {
+      "epoch": 2.9562841530054644,
+      "grad_norm": 0.40139836072921753,
+      "learning_rate": 2.9534347329398027e-06,
+      "loss": 0.5934174656867981,
+      "step": 1082
+    },
+    {
+      "epoch": 2.9617486338797816,
+      "grad_norm": 0.30364465713500977,
+      "learning_rate": 3.0326408037922827e-06,
+      "loss": 0.8147178292274475,
+      "step": 1084
+    },
+    {
+      "epoch": 2.9672131147540983,
+      "grad_norm": 0.1606474071741104,
+      "learning_rate": 3.1148075584248306e-06,
+      "loss": 0.5487402677536011,
+      "step": 1086
+    },
+    {
+      "epoch": 2.972677595628415,
+      "grad_norm": 0.5508518815040588,
+      "learning_rate": 3.199904463536296e-06,
+      "loss": 1.1423696279525757,
+      "step": 1088
+    },
+    {
+      "epoch": 2.978142076502732,
+      "grad_norm": 0.6574386954307556,
+      "learning_rate": 3.2878998969767954e-06,
+      "loss": 0.6705480813980103,
+      "step": 1090
+    },
+    {
+      "epoch": 2.9836065573770494,
+      "grad_norm": 0.4296954572200775,
+      "learning_rate": 3.378761159498547e-06,
+      "loss": 0.8128272891044617,
+      "step": 1092
+    },
+    {
+      "epoch": 2.989071038251366,
+      "grad_norm": 0.28330495953559875,
+      "learning_rate": 3.472454486906972e-06,
+      "loss": 0.9465227723121643,
+      "step": 1094
+    },
+    {
+      "epoch": 2.994535519125683,
+      "grad_norm": 0.679470956325531,
+      "learning_rate": 3.5689450626075132e-06,
+      "loss": 0.5051727890968323,
+      "step": 1096
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.3425956070423126,
+      "learning_rate": 3.668197030543573e-06,
+      "loss": 0.7532654404640198,
+      "step": 1098
+    },
+    {
+      "epoch": 3.0,
+      "step": 1098,
+      "total_flos": 4.931331991116186e+18,
+      "train_loss": 0.9669766378315854,
+      "train_runtime": 12176.252,
+      "train_samples_per_second": 5.411,
+      "train_steps_per_second": 0.09
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 1098,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 99999,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.931331991116186e+18,
+  "train_batch_size": 3,
+  "trial_name": null,
+  "trial_params": null
+}