diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,32508 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.7002356226546818,
+  "eval_steps": 500,
+  "global_step": 92826,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005817843325479244,
+      "grad_norm": 14.125,
+      "learning_rate": 9.998063627249894e-06,
+      "loss": 1.8693,
+      "step": 20
+    },
+    {
+      "epoch": 0.0011635686650958489,
+      "grad_norm": 11.625,
+      "learning_rate": 9.996124345577279e-06,
+      "loss": 1.767,
+      "step": 40
+    },
+    {
+      "epoch": 0.0017453529976437735,
+      "grad_norm": 13.125,
+      "learning_rate": 9.994185063904664e-06,
+      "loss": 1.7496,
+      "step": 60
+    },
+    {
+      "epoch": 0.0023271373301916977,
+      "grad_norm": 13.3125,
+      "learning_rate": 9.992245782232049e-06,
+      "loss": 1.7907,
+      "step": 80
+    },
+    {
+      "epoch": 0.0029089216627396224,
+      "grad_norm": 12.125,
+      "learning_rate": 9.990306500559434e-06,
+      "loss": 1.6286,
+      "step": 100
+    },
+    {
+      "epoch": 0.003490705995287547,
+      "grad_norm": 12.4375,
+      "learning_rate": 9.98836721888682e-06,
+      "loss": 1.8037,
+      "step": 120
+    },
+    {
+      "epoch": 0.004072490327835471,
+      "grad_norm": 12.5625,
+      "learning_rate": 9.986427937214204e-06,
+      "loss": 1.7154,
+      "step": 140
+    },
+    {
+      "epoch": 0.0046542746603833954,
+      "grad_norm": 14.4375,
+      "learning_rate": 9.98448865554159e-06,
+      "loss": 1.7231,
+      "step": 160
+    },
+    {
+      "epoch": 0.0052360589929313205,
+      "grad_norm": 9.5625,
+      "learning_rate": 9.982549373868974e-06,
+      "loss": 1.701,
+      "step": 180
+    },
+    {
+      "epoch": 0.005817843325479245,
+      "grad_norm": 12.625,
+      "learning_rate": 9.98061009219636e-06,
+      "loss": 1.7652,
+      "step": 200
+    },
+    {
+      "epoch": 0.006399627658027169,
+      "grad_norm": 12.6875,
+      "learning_rate": 9.978670810523745e-06,
+      "loss": 1.7116,
+      "step": 220
+    },
+    {
+      "epoch": 0.006981411990575094,
+      "grad_norm": 12.6875,
+      "learning_rate": 9.97673152885113e-06,
+      "loss": 1.6567,
+      "step": 240
+    },
+    {
+      "epoch": 0.007563196323123018,
+      "grad_norm": 14.8125,
+      "learning_rate": 9.974792247178515e-06,
+      "loss": 1.7739,
+      "step": 260
+    },
+    {
+      "epoch": 0.008144980655670942,
+      "grad_norm": 13.0625,
+      "learning_rate": 9.9728529655059e-06,
+      "loss": 1.9183,
+      "step": 280
+    },
+    {
+      "epoch": 0.008726764988218868,
+      "grad_norm": 14.375,
+      "learning_rate": 9.970913683833285e-06,
+      "loss": 1.7885,
+      "step": 300
+    },
+    {
+      "epoch": 0.009308549320766791,
+      "grad_norm": 13.3125,
+      "learning_rate": 9.96897440216067e-06,
+      "loss": 1.7945,
+      "step": 320
+    },
+    {
+      "epoch": 0.009890333653314716,
+      "grad_norm": 12.5625,
+      "learning_rate": 9.967035120488055e-06,
+      "loss": 1.7623,
+      "step": 340
+    },
+    {
+      "epoch": 0.010472117985862641,
+      "grad_norm": 13.375,
+      "learning_rate": 9.96509583881544e-06,
+      "loss": 1.6728,
+      "step": 360
+    },
+    {
+      "epoch": 0.011053902318410564,
+      "grad_norm": 13.3125,
+      "learning_rate": 9.963156557142825e-06,
+      "loss": 1.7254,
+      "step": 380
+    },
+    {
+      "epoch": 0.01163568665095849,
+      "grad_norm": 15.5,
+      "learning_rate": 9.96121727547021e-06,
+      "loss": 1.6798,
+      "step": 400
+    },
+    {
+      "epoch": 0.012217470983506415,
+      "grad_norm": 12.6875,
+      "learning_rate": 9.959277993797596e-06,
+      "loss": 1.7626,
+      "step": 420
+    },
+    {
+      "epoch": 0.012799255316054338,
+      "grad_norm": 13.0625,
+      "learning_rate": 9.95733871212498e-06,
+      "loss": 1.823,
+      "step": 440
+    },
+    {
+      "epoch": 0.013381039648602263,
+      "grad_norm": 11.75,
+      "learning_rate": 9.955399430452366e-06,
+      "loss": 1.688,
+      "step": 460
+    },
+    {
+      "epoch": 0.013962823981150188,
+      "grad_norm": 13.3125,
+      "learning_rate": 9.953460148779751e-06,
+      "loss": 1.762,
+      "step": 480
+    },
+    {
+      "epoch": 0.014544608313698111,
+      "grad_norm": 12.1875,
+      "learning_rate": 9.951520867107136e-06,
+      "loss": 1.6487,
+      "step": 500
+    },
+    {
+      "epoch": 0.015126392646246037,
+      "grad_norm": 15.0625,
+      "learning_rate": 9.949581585434521e-06,
+      "loss": 1.8255,
+      "step": 520
+    },
+    {
+      "epoch": 0.01570817697879396,
+      "grad_norm": 13.4375,
+      "learning_rate": 9.947642303761906e-06,
+      "loss": 1.7325,
+      "step": 540
+    },
+    {
+      "epoch": 0.016289961311341885,
+      "grad_norm": 15.25,
+      "learning_rate": 9.945703022089291e-06,
+      "loss": 1.8285,
+      "step": 560
+    },
+    {
+      "epoch": 0.01687174564388981,
+      "grad_norm": 12.375,
+      "learning_rate": 9.943763740416676e-06,
+      "loss": 1.7398,
+      "step": 580
+    },
+    {
+      "epoch": 0.017453529976437735,
+      "grad_norm": 16.625,
+      "learning_rate": 9.941824458744062e-06,
+      "loss": 1.7851,
+      "step": 600
+    },
+    {
+      "epoch": 0.01803531430898566,
+      "grad_norm": 11.875,
+      "learning_rate": 9.939885177071447e-06,
+      "loss": 1.7592,
+      "step": 620
+    },
+    {
+      "epoch": 0.018617098641533582,
+      "grad_norm": 16.875,
+      "learning_rate": 9.937945895398832e-06,
+      "loss": 1.7966,
+      "step": 640
+    },
+    {
+      "epoch": 0.019198882974081507,
+      "grad_norm": 10.9375,
+      "learning_rate": 9.936006613726217e-06,
+      "loss": 1.704,
+      "step": 660
+    },
+    {
+      "epoch": 0.019780667306629432,
+      "grad_norm": 18.5,
+      "learning_rate": 9.934067332053602e-06,
+      "loss": 1.7116,
+      "step": 680
+    },
+    {
+      "epoch": 0.020362451639177357,
+      "grad_norm": 14.6875,
+      "learning_rate": 9.932128050380987e-06,
+      "loss": 1.7695,
+      "step": 700
+    },
+    {
+      "epoch": 0.020944235971725282,
+      "grad_norm": 12.625,
+      "learning_rate": 9.930188768708372e-06,
+      "loss": 1.7622,
+      "step": 720
+    },
+    {
+      "epoch": 0.021526020304273207,
+      "grad_norm": 13.3125,
+      "learning_rate": 9.928249487035757e-06,
+      "loss": 1.7275,
+      "step": 740
+    },
+    {
+      "epoch": 0.02210780463682113,
+      "grad_norm": 12.5,
+      "learning_rate": 9.926310205363142e-06,
+      "loss": 1.6935,
+      "step": 760
+    },
+    {
+      "epoch": 0.022689588969369054,
+      "grad_norm": 12.75,
+      "learning_rate": 9.924370923690527e-06,
+      "loss": 1.8071,
+      "step": 780
+    },
+    {
+      "epoch": 0.02327137330191698,
+      "grad_norm": 13.9375,
+      "learning_rate": 9.922431642017912e-06,
+      "loss": 1.7624,
+      "step": 800
+    },
+    {
+      "epoch": 0.023853157634464904,
+      "grad_norm": 12.375,
+      "learning_rate": 9.920492360345298e-06,
+      "loss": 1.7581,
+      "step": 820
+    },
+    {
+      "epoch": 0.02443494196701283,
+      "grad_norm": 11.625,
+      "learning_rate": 9.918553078672683e-06,
+      "loss": 1.7039,
+      "step": 840
+    },
+    {
+      "epoch": 0.025016726299560754,
+      "grad_norm": 12.125,
+      "learning_rate": 9.916613797000068e-06,
+      "loss": 1.7564,
+      "step": 860
+    },
+    {
+      "epoch": 0.025598510632108676,
+      "grad_norm": 13.875,
+      "learning_rate": 9.914674515327453e-06,
+      "loss": 1.7778,
+      "step": 880
+    },
+    {
+      "epoch": 0.0261802949646566,
+      "grad_norm": 12.25,
+      "learning_rate": 9.912735233654838e-06,
+      "loss": 1.7182,
+      "step": 900
+    },
+    {
+      "epoch": 0.026762079297204526,
+      "grad_norm": 14.4375,
+      "learning_rate": 9.910795951982223e-06,
+      "loss": 1.6581,
+      "step": 920
+    },
+    {
+      "epoch": 0.02734386362975245,
+      "grad_norm": 13.5625,
+      "learning_rate": 9.908856670309608e-06,
+      "loss": 1.7495,
+      "step": 940
+    },
+    {
+      "epoch": 0.027925647962300376,
+      "grad_norm": 13.125,
+      "learning_rate": 9.906917388636993e-06,
+      "loss": 1.75,
+      "step": 960
+    },
+    {
+      "epoch": 0.0285074322948483,
+      "grad_norm": 13.4375,
+      "learning_rate": 9.904978106964378e-06,
+      "loss": 1.8321,
+      "step": 980
+    },
+    {
+      "epoch": 0.029089216627396223,
+      "grad_norm": 12.0625,
+      "learning_rate": 9.903038825291763e-06,
+      "loss": 1.7237,
+      "step": 1000
+    },
+    {
+      "epoch": 0.029671000959944148,
+      "grad_norm": 14.0625,
+      "learning_rate": 9.901099543619149e-06,
+      "loss": 1.7602,
+      "step": 1020
+    },
+    {
+      "epoch": 0.030252785292492073,
+      "grad_norm": 10.375,
+      "learning_rate": 9.899160261946534e-06,
+      "loss": 1.7526,
+      "step": 1040
+    },
+    {
+      "epoch": 0.030834569625039998,
+      "grad_norm": 12.0625,
+      "learning_rate": 9.897220980273919e-06,
+      "loss": 1.6564,
+      "step": 1060
+    },
+    {
+      "epoch": 0.03141635395758792,
+      "grad_norm": 14.0,
+      "learning_rate": 9.895281698601304e-06,
+      "loss": 1.7071,
+      "step": 1080
+    },
+    {
+      "epoch": 0.03199813829013585,
+      "grad_norm": 13.25,
+      "learning_rate": 9.893342416928689e-06,
+      "loss": 1.7021,
+      "step": 1100
+    },
+    {
+      "epoch": 0.03257992262268377,
+      "grad_norm": 14.5625,
+      "learning_rate": 9.891403135256074e-06,
+      "loss": 1.6877,
+      "step": 1120
+    },
+    {
+      "epoch": 0.0331617069552317,
+      "grad_norm": 12.5625,
+      "learning_rate": 9.88946385358346e-06,
+      "loss": 1.6961,
+      "step": 1140
+    },
+    {
+      "epoch": 0.03374349128777962,
+      "grad_norm": 13.0625,
+      "learning_rate": 9.887524571910844e-06,
+      "loss": 1.7913,
+      "step": 1160
+    },
+    {
+      "epoch": 0.03432527562032754,
+      "grad_norm": 13.125,
+      "learning_rate": 9.88558529023823e-06,
+      "loss": 1.7257,
+      "step": 1180
+    },
+    {
+      "epoch": 0.03490705995287547,
+      "grad_norm": 14.125,
+      "learning_rate": 9.883646008565614e-06,
+      "loss": 1.7231,
+      "step": 1200
+    },
+    {
+      "epoch": 0.03548884428542339,
+      "grad_norm": 12.75,
+      "learning_rate": 9.881706726893e-06,
+      "loss": 1.744,
+      "step": 1220
+    },
+    {
+      "epoch": 0.03607062861797132,
+      "grad_norm": 12.25,
+      "learning_rate": 9.879767445220385e-06,
+      "loss": 1.6365,
+      "step": 1240
+    },
+    {
+      "epoch": 0.03665241295051924,
+      "grad_norm": 9.375,
+      "learning_rate": 9.87782816354777e-06,
+      "loss": 1.7705,
+      "step": 1260
+    },
+    {
+      "epoch": 0.037234197283067164,
+      "grad_norm": 10.8125,
+      "learning_rate": 9.875888881875155e-06,
+      "loss": 1.7309,
+      "step": 1280
+    },
+    {
+      "epoch": 0.03781598161561509,
+      "grad_norm": 11.0,
+      "learning_rate": 9.87394960020254e-06,
+      "loss": 1.7818,
+      "step": 1300
+    },
+    {
+      "epoch": 0.038397765948163014,
+      "grad_norm": 13.6875,
+      "learning_rate": 9.872010318529925e-06,
+      "loss": 1.7461,
+      "step": 1320
+    },
+    {
+      "epoch": 0.03897955028071094,
+      "grad_norm": 13.8125,
+      "learning_rate": 9.87007103685731e-06,
+      "loss": 1.7633,
+      "step": 1340
+    },
+    {
+      "epoch": 0.039561334613258864,
+      "grad_norm": 12.25,
+      "learning_rate": 9.868131755184695e-06,
+      "loss": 1.7222,
+      "step": 1360
+    },
+    {
+      "epoch": 0.04014311894580679,
+      "grad_norm": 13.3125,
+      "learning_rate": 9.86619247351208e-06,
+      "loss": 1.7887,
+      "step": 1380
+    },
+    {
+      "epoch": 0.040724903278354714,
+      "grad_norm": 12.9375,
+      "learning_rate": 9.864253191839465e-06,
+      "loss": 1.7775,
+      "step": 1400
+    },
+    {
+      "epoch": 0.041306687610902636,
+      "grad_norm": 12.25,
+      "learning_rate": 9.86231391016685e-06,
+      "loss": 1.724,
+      "step": 1420
+    },
+    {
+      "epoch": 0.041888471943450564,
+      "grad_norm": 9.125,
+      "learning_rate": 9.860374628494236e-06,
+      "loss": 1.7322,
+      "step": 1440
+    },
+    {
+      "epoch": 0.042470256275998486,
+      "grad_norm": 13.125,
+      "learning_rate": 9.858435346821619e-06,
+      "loss": 1.7101,
+      "step": 1460
+    },
+    {
+      "epoch": 0.043052040608546414,
+      "grad_norm": 14.0,
+      "learning_rate": 9.856496065149004e-06,
+      "loss": 1.7176,
+      "step": 1480
+    },
+    {
+      "epoch": 0.043633824941094336,
+      "grad_norm": 14.6875,
+      "learning_rate": 9.85455678347639e-06,
+      "loss": 1.6656,
+      "step": 1500
+    },
+    {
+      "epoch": 0.04421560927364226,
+      "grad_norm": 12.5,
+      "learning_rate": 9.852617501803774e-06,
+      "loss": 1.7662,
+      "step": 1520
+    },
+    {
+      "epoch": 0.044797393606190186,
+      "grad_norm": 13.4375,
+      "learning_rate": 9.85067822013116e-06,
+      "loss": 1.6732,
+      "step": 1540
+    },
+    {
+      "epoch": 0.04537917793873811,
+      "grad_norm": 13.0,
+      "learning_rate": 9.848738938458545e-06,
+      "loss": 1.7837,
+      "step": 1560
+    },
+    {
+      "epoch": 0.045960962271286036,
+      "grad_norm": 11.8125,
+      "learning_rate": 9.84679965678593e-06,
+      "loss": 1.703,
+      "step": 1580
+    },
+    {
+      "epoch": 0.04654274660383396,
+      "grad_norm": 9.3125,
+      "learning_rate": 9.844860375113315e-06,
+      "loss": 1.7084,
+      "step": 1600
+    },
+    {
+      "epoch": 0.047124530936381887,
+      "grad_norm": 10.8125,
+      "learning_rate": 9.8429210934407e-06,
+      "loss": 1.795,
+      "step": 1620
+    },
+    {
+      "epoch": 0.04770631526892981,
+      "grad_norm": 10.75,
+      "learning_rate": 9.840981811768085e-06,
+      "loss": 1.8042,
+      "step": 1640
+    },
+    {
+      "epoch": 0.04828809960147773,
+      "grad_norm": 12.4375,
+      "learning_rate": 9.83904253009547e-06,
+      "loss": 1.6848,
+      "step": 1660
+    },
+    {
+      "epoch": 0.04886988393402566,
+      "grad_norm": 14.0,
+      "learning_rate": 9.837103248422855e-06,
+      "loss": 1.6867,
+      "step": 1680
+    },
+    {
+      "epoch": 0.04945166826657358,
+      "grad_norm": 13.9375,
+      "learning_rate": 9.83516396675024e-06,
+      "loss": 1.7605,
+      "step": 1700
+    },
+    {
+      "epoch": 0.05003345259912151,
+      "grad_norm": 13.4375,
+      "learning_rate": 9.833224685077625e-06,
+      "loss": 1.6898,
+      "step": 1720
+    },
+    {
+      "epoch": 0.05061523693166943,
+      "grad_norm": 11.5,
+      "learning_rate": 9.83128540340501e-06,
+      "loss": 1.723,
+      "step": 1740
+    },
+    {
+      "epoch": 0.05119702126421735,
+      "grad_norm": 14.5625,
+      "learning_rate": 9.829346121732395e-06,
+      "loss": 1.7505,
+      "step": 1760
+    },
+    {
+      "epoch": 0.05177880559676528,
+      "grad_norm": 12.1875,
+      "learning_rate": 9.82740684005978e-06,
+      "loss": 1.7424,
+      "step": 1780
+    },
+    {
+      "epoch": 0.0523605899293132,
+      "grad_norm": 13.0625,
+      "learning_rate": 9.825467558387166e-06,
+      "loss": 1.7796,
+      "step": 1800
+    },
+    {
+      "epoch": 0.05294237426186113,
+      "grad_norm": 11.3125,
+      "learning_rate": 9.82352827671455e-06,
+      "loss": 1.6866,
+      "step": 1820
+    },
+    {
+      "epoch": 0.05352415859440905,
+      "grad_norm": 11.375,
+      "learning_rate": 9.821588995041936e-06,
+      "loss": 1.812,
+      "step": 1840
+    },
+    {
+      "epoch": 0.054105942926956974,
+      "grad_norm": 13.625,
+      "learning_rate": 9.819649713369321e-06,
+      "loss": 1.7411,
+      "step": 1860
+    },
+    {
+      "epoch": 0.0546877272595049,
+      "grad_norm": 12.5625,
+      "learning_rate": 9.817710431696706e-06,
+      "loss": 1.7386,
+      "step": 1880
+    },
+    {
+      "epoch": 0.055269511592052824,
+      "grad_norm": 13.9375,
+      "learning_rate": 9.815771150024091e-06,
+      "loss": 1.7493,
+      "step": 1900
+    },
+    {
+      "epoch": 0.05585129592460075,
+      "grad_norm": 16.625,
+      "learning_rate": 9.813831868351476e-06,
+      "loss": 1.7573,
+      "step": 1920
+    },
+    {
+      "epoch": 0.056433080257148674,
+      "grad_norm": 15.3125,
+      "learning_rate": 9.811892586678861e-06,
+      "loss": 1.7363,
+      "step": 1940
+    },
+    {
+      "epoch": 0.0570148645896966,
+      "grad_norm": 13.75,
+      "learning_rate": 9.809953305006246e-06,
+      "loss": 1.7155,
+      "step": 1960
+    },
+    {
+      "epoch": 0.057596648922244524,
+      "grad_norm": 13.5,
+      "learning_rate": 9.808014023333632e-06,
+      "loss": 1.7193,
+      "step": 1980
+    },
+    {
+      "epoch": 0.058178433254792446,
+      "grad_norm": 14.5625,
+      "learning_rate": 9.806074741661017e-06,
+      "loss": 1.6673,
+      "step": 2000
+    },
+    {
+      "epoch": 0.058760217587340374,
+      "grad_norm": 12.375,
+      "learning_rate": 9.804135459988402e-06,
+      "loss": 1.7952,
+      "step": 2020
+    },
+    {
+      "epoch": 0.059342001919888296,
+      "grad_norm": 13.1875,
+      "learning_rate": 9.802196178315787e-06,
+      "loss": 1.7525,
+      "step": 2040
+    },
+    {
+      "epoch": 0.059923786252436224,
+      "grad_norm": 12.6875,
+      "learning_rate": 9.800256896643172e-06,
+      "loss": 1.7017,
+      "step": 2060
+    },
+    {
+      "epoch": 0.060505570584984146,
+      "grad_norm": 14.0625,
+      "learning_rate": 9.798317614970557e-06,
+      "loss": 1.6701,
+      "step": 2080
+    },
+    {
+      "epoch": 0.06108735491753207,
+      "grad_norm": 12.8125,
+      "learning_rate": 9.796378333297942e-06,
+      "loss": 1.7619,
+      "step": 2100
+    },
+    {
+      "epoch": 0.061669139250079996,
+      "grad_norm": 12.875,
+      "learning_rate": 9.794439051625327e-06,
+      "loss": 1.6994,
+      "step": 2120
+    },
+    {
+      "epoch": 0.06225092358262792,
+      "grad_norm": 13.6875,
+      "learning_rate": 9.792499769952712e-06,
+      "loss": 1.6774,
+      "step": 2140
+    },
+    {
+      "epoch": 0.06283270791517584,
+      "grad_norm": 10.6875,
+      "learning_rate": 9.790560488280097e-06,
+      "loss": 1.7265,
+      "step": 2160
+    },
+    {
+      "epoch": 0.06341449224772377,
+      "grad_norm": 11.375,
+      "learning_rate": 9.788621206607483e-06,
+      "loss": 1.749,
+      "step": 2180
+    },
+    {
+      "epoch": 0.0639962765802717,
+      "grad_norm": 11.0,
+      "learning_rate": 9.786681924934868e-06,
+      "loss": 1.7317,
+      "step": 2200
+    },
+    {
+      "epoch": 0.06457806091281962,
+      "grad_norm": 12.4375,
+      "learning_rate": 9.784742643262253e-06,
+      "loss": 1.7559,
+      "step": 2220
+    },
+    {
+      "epoch": 0.06515984524536754,
+      "grad_norm": 10.375,
+      "learning_rate": 9.782803361589638e-06,
+      "loss": 1.7727,
+      "step": 2240
+    },
+    {
+      "epoch": 0.06574162957791546,
+      "grad_norm": 14.0625,
+      "learning_rate": 9.780864079917023e-06,
+      "loss": 1.6845,
+      "step": 2260
+    },
+    {
+      "epoch": 0.0663234139104634,
+      "grad_norm": 12.8125,
+      "learning_rate": 9.778924798244408e-06,
+      "loss": 1.762,
+      "step": 2280
+    },
+    {
+      "epoch": 0.06690519824301132,
+      "grad_norm": 11.9375,
+      "learning_rate": 9.776985516571793e-06,
+      "loss": 1.7802,
+      "step": 2300
+    },
+    {
+      "epoch": 0.06748698257555924,
+      "grad_norm": 10.125,
+      "learning_rate": 9.775046234899178e-06,
+      "loss": 1.6035,
+      "step": 2320
+    },
+    {
+      "epoch": 0.06806876690810716,
+      "grad_norm": 11.4375,
+      "learning_rate": 9.773106953226563e-06,
+      "loss": 1.7954,
+      "step": 2340
+    },
+    {
+      "epoch": 0.06865055124065508,
+      "grad_norm": 13.4375,
+      "learning_rate": 9.771167671553948e-06,
+      "loss": 1.7277,
+      "step": 2360
+    },
+    {
+      "epoch": 0.06923233557320302,
+      "grad_norm": 14.0,
+      "learning_rate": 9.769228389881334e-06,
+      "loss": 1.7112,
+      "step": 2380
+    },
+    {
+      "epoch": 0.06981411990575094,
+      "grad_norm": 11.875,
+      "learning_rate": 9.767289108208719e-06,
+      "loss": 1.7412,
+      "step": 2400
+    },
+    {
+      "epoch": 0.07039590423829886,
+      "grad_norm": 14.6875,
+      "learning_rate": 9.765349826536104e-06,
+      "loss": 1.6549,
+      "step": 2420
+    },
+    {
+      "epoch": 0.07097768857084678,
+      "grad_norm": 13.125,
+      "learning_rate": 9.763410544863489e-06,
+      "loss": 1.7308,
+      "step": 2440
+    },
+    {
+      "epoch": 0.0715594729033947,
+      "grad_norm": 10.25,
+      "learning_rate": 9.761471263190874e-06,
+      "loss": 1.7989,
+      "step": 2460
+    },
+    {
+      "epoch": 0.07214125723594264,
+      "grad_norm": 12.8125,
+      "learning_rate": 9.759531981518259e-06,
+      "loss": 1.6674,
+      "step": 2480
+    },
+    {
+      "epoch": 0.07272304156849056,
+      "grad_norm": 12.125,
+      "learning_rate": 9.757592699845644e-06,
+      "loss": 1.7343,
+      "step": 2500
+    },
+    {
+      "epoch": 0.07330482590103848,
+      "grad_norm": 13.1875,
+      "learning_rate": 9.75565341817303e-06,
+      "loss": 1.7163,
+      "step": 2520
+    },
+    {
+      "epoch": 0.0738866102335864,
+      "grad_norm": 13.3125,
+      "learning_rate": 9.753714136500414e-06,
+      "loss": 1.7604,
+      "step": 2540
+    },
+    {
+      "epoch": 0.07446839456613433,
+      "grad_norm": 12.9375,
+      "learning_rate": 9.7517748548278e-06,
+      "loss": 1.73,
+      "step": 2560
+    },
+    {
+      "epoch": 0.07505017889868226,
+      "grad_norm": 12.625,
+      "learning_rate": 9.749835573155184e-06,
+      "loss": 1.721,
+      "step": 2580
+    },
+    {
+      "epoch": 0.07563196323123018,
+      "grad_norm": 11.6875,
+      "learning_rate": 9.74789629148257e-06,
+      "loss": 1.7146,
+      "step": 2600
+    },
+    {
+      "epoch": 0.0762137475637781,
+      "grad_norm": 15.3125,
+      "learning_rate": 9.745957009809955e-06,
+      "loss": 1.7791,
+      "step": 2620
+    },
+    {
+      "epoch": 0.07679553189632603,
+      "grad_norm": 16.0,
+      "learning_rate": 9.74401772813734e-06,
+      "loss": 1.7371,
+      "step": 2640
+    },
+    {
+      "epoch": 0.07737731622887396,
+      "grad_norm": 11.5625,
+      "learning_rate": 9.742078446464725e-06,
+      "loss": 1.7157,
+      "step": 2660
+    },
+    {
+      "epoch": 0.07795910056142188,
+      "grad_norm": 12.375,
+      "learning_rate": 9.74013916479211e-06,
+      "loss": 1.6902,
+      "step": 2680
+    },
+    {
+      "epoch": 0.0785408848939698,
+      "grad_norm": 13.1875,
+      "learning_rate": 9.738199883119495e-06,
+      "loss": 1.7734,
+      "step": 2700
+    },
+    {
+      "epoch": 0.07912266922651773,
+      "grad_norm": 13.5625,
+      "learning_rate": 9.73626060144688e-06,
+      "loss": 1.7238,
+      "step": 2720
+    },
+    {
+      "epoch": 0.07970445355906565,
+      "grad_norm": 10.9375,
+      "learning_rate": 9.734321319774265e-06,
+      "loss": 1.7213,
+      "step": 2740
+    },
+    {
+      "epoch": 0.08028623789161358,
+      "grad_norm": 12.5625,
+      "learning_rate": 9.73238203810165e-06,
+      "loss": 1.647,
+      "step": 2760
+    },
+    {
+      "epoch": 0.0808680222241615,
+      "grad_norm": 14.875,
+      "learning_rate": 9.730442756429035e-06,
+      "loss": 1.6843,
+      "step": 2780
+    },
+    {
+      "epoch": 0.08144980655670943,
+      "grad_norm": 13.0625,
+      "learning_rate": 9.72850347475642e-06,
+      "loss": 1.734,
+      "step": 2800
+    },
+    {
+      "epoch": 0.08203159088925735,
+      "grad_norm": 11.0,
+      "learning_rate": 9.726564193083806e-06,
+      "loss": 1.7499,
+      "step": 2820
+    },
+    {
+      "epoch": 0.08261337522180527,
+      "grad_norm": 16.75,
+      "learning_rate": 9.72462491141119e-06,
+      "loss": 1.7646,
+      "step": 2840
+    },
+    {
+      "epoch": 0.0831951595543532,
+      "grad_norm": 11.8125,
+      "learning_rate": 9.722685629738576e-06,
+      "loss": 1.7599,
+      "step": 2860
+    },
+    {
+      "epoch": 0.08377694388690113,
+      "grad_norm": 11.5,
+      "learning_rate": 9.720746348065961e-06,
+      "loss": 1.7405,
+      "step": 2880
+    },
+    {
+      "epoch": 0.08435872821944905,
+      "grad_norm": 11.375,
+      "learning_rate": 9.718807066393346e-06,
+      "loss": 1.6979,
+      "step": 2900
+    },
+    {
+      "epoch": 0.08494051255199697,
+      "grad_norm": 11.8125,
+      "learning_rate": 9.716867784720731e-06,
+      "loss": 1.7044,
+      "step": 2920
+    },
+    {
+      "epoch": 0.0855222968845449,
+      "grad_norm": 12.25,
+      "learning_rate": 9.714928503048116e-06,
+      "loss": 1.7057,
+      "step": 2940
+    },
+    {
+      "epoch": 0.08610408121709283,
+      "grad_norm": 12.3125,
+      "learning_rate": 9.712989221375501e-06,
+      "loss": 1.6687,
+      "step": 2960
+    },
+    {
+      "epoch": 0.08668586554964075,
+      "grad_norm": 12.5625,
+      "learning_rate": 9.711049939702886e-06,
+      "loss": 1.7973,
+      "step": 2980
+    },
+    {
+      "epoch": 0.08726764988218867,
+      "grad_norm": 11.4375,
+      "learning_rate": 9.709110658030272e-06,
+      "loss": 1.7279,
+      "step": 3000
+    },
+    {
+      "epoch": 0.0878494342147366,
+      "grad_norm": 12.4375,
+      "learning_rate": 9.707171376357657e-06,
+      "loss": 1.6971,
+      "step": 3020
+    },
+    {
+      "epoch": 0.08843121854728452,
+      "grad_norm": 11.125,
+      "learning_rate": 9.705232094685042e-06,
+      "loss": 1.679,
+      "step": 3040
+    },
+    {
+      "epoch": 0.08901300287983245,
+      "grad_norm": 14.875,
+      "learning_rate": 9.703292813012427e-06,
+      "loss": 1.699,
+      "step": 3060
+    },
+    {
+      "epoch": 0.08959478721238037,
+      "grad_norm": 11.875,
+      "learning_rate": 9.70135353133981e-06,
+      "loss": 1.803,
+      "step": 3080
+    },
+    {
+      "epoch": 0.0901765715449283,
+      "grad_norm": 13.75,
+      "learning_rate": 9.699414249667195e-06,
+      "loss": 1.6699,
+      "step": 3100
+    },
+    {
+      "epoch": 0.09075835587747622,
+      "grad_norm": 11.9375,
+      "learning_rate": 9.69747496799458e-06,
+      "loss": 1.7585,
+      "step": 3120
+    },
+    {
+      "epoch": 0.09134014021002414,
+      "grad_norm": 15.1875,
+      "learning_rate": 9.695535686321966e-06,
+      "loss": 1.7541,
+      "step": 3140
+    },
+    {
+      "epoch": 0.09192192454257207,
+      "grad_norm": 14.125,
+      "learning_rate": 9.69359640464935e-06,
+      "loss": 1.7145,
+      "step": 3160
+    },
+    {
+      "epoch": 0.09250370887512,
+      "grad_norm": 12.9375,
+      "learning_rate": 9.691657122976736e-06,
+      "loss": 1.7008,
+      "step": 3180
+    },
+    {
+      "epoch": 0.09308549320766792,
+      "grad_norm": 11.8125,
+      "learning_rate": 9.68971784130412e-06,
+      "loss": 1.6591,
+      "step": 3200
+    },
+    {
+      "epoch": 0.09366727754021584,
+      "grad_norm": 11.9375,
+      "learning_rate": 9.687778559631506e-06,
+      "loss": 1.6305,
+      "step": 3220
+    },
+    {
+      "epoch": 0.09424906187276377,
+      "grad_norm": 12.0625,
+      "learning_rate": 9.685839277958891e-06,
+      "loss": 1.7425,
+      "step": 3240
+    },
+    {
+      "epoch": 0.0948308462053117,
+      "grad_norm": 12.3125,
+      "learning_rate": 9.683899996286276e-06,
+      "loss": 1.7959,
+      "step": 3260
+    },
+    {
+      "epoch": 0.09541263053785962,
+      "grad_norm": 12.875,
+      "learning_rate": 9.681960714613661e-06,
+      "loss": 1.765,
+      "step": 3280
+    },
+    {
+      "epoch": 0.09599441487040754,
+      "grad_norm": 14.0625,
+      "learning_rate": 9.680021432941046e-06,
+      "loss": 1.7248,
+      "step": 3300
+    },
+    {
+      "epoch": 0.09657619920295546,
+      "grad_norm": 14.875,
+      "learning_rate": 9.678082151268431e-06,
+      "loss": 1.6914,
+      "step": 3320
+    },
+    {
+      "epoch": 0.0971579835355034,
+      "grad_norm": 12.4375,
+      "learning_rate": 9.676142869595816e-06,
+      "loss": 1.7039,
+      "step": 3340
+    },
+    {
+      "epoch": 0.09773976786805132,
+      "grad_norm": 11.8125,
+      "learning_rate": 9.674203587923202e-06,
+      "loss": 1.5978,
+      "step": 3360
+    },
+    {
+      "epoch": 0.09832155220059924,
+      "grad_norm": 12.625,
+      "learning_rate": 9.672264306250587e-06,
+      "loss": 1.6436,
+      "step": 3380
+    },
+    {
+      "epoch": 0.09890333653314716,
+      "grad_norm": 10.0625,
+      "learning_rate": 9.670325024577972e-06,
+      "loss": 1.7001,
+      "step": 3400
+    },
+    {
+      "epoch": 0.09948512086569508,
+      "grad_norm": 12.3125,
+      "learning_rate": 9.668385742905357e-06,
+      "loss": 1.7009,
+      "step": 3420
+    },
+    {
+      "epoch": 0.10006690519824302,
+      "grad_norm": 13.1875,
+      "learning_rate": 9.666446461232742e-06,
+      "loss": 1.7319,
+      "step": 3440
+    },
+    {
+      "epoch": 0.10064868953079094,
+      "grad_norm": 17.75,
+      "learning_rate": 9.664507179560127e-06,
+      "loss": 1.7174,
+      "step": 3460
+    },
+    {
+      "epoch": 0.10123047386333886,
+      "grad_norm": 10.5,
+      "learning_rate": 9.662567897887512e-06,
+      "loss": 1.7301,
+      "step": 3480
+    },
+    {
+      "epoch": 0.10181225819588678,
+      "grad_norm": 10.75,
+      "learning_rate": 9.660628616214897e-06,
+      "loss": 1.6914,
+      "step": 3500
+    },
+    {
+      "epoch": 0.1023940425284347,
+      "grad_norm": 12.75,
+      "learning_rate": 9.658689334542282e-06,
+      "loss": 1.6277,
+      "step": 3520
+    },
+    {
+      "epoch": 0.10297582686098264,
+      "grad_norm": 14.4375,
+      "learning_rate": 9.656750052869667e-06,
+      "loss": 1.6839,
+      "step": 3540
+    },
+    {
+      "epoch": 0.10355761119353056,
+      "grad_norm": 12.5625,
+      "learning_rate": 9.654810771197053e-06,
+      "loss": 1.7156,
+      "step": 3560
+    },
+    {
+      "epoch": 0.10413939552607848,
+      "grad_norm": 13.6875,
+      "learning_rate": 9.652871489524438e-06,
+      "loss": 1.6836,
+      "step": 3580
+    },
+    {
+      "epoch": 0.1047211798586264,
+      "grad_norm": 13.3125,
+      "learning_rate": 9.650932207851823e-06,
+      "loss": 1.7005,
+      "step": 3600
+    },
+    {
+      "epoch": 0.10530296419117433,
+      "grad_norm": 10.6875,
+      "learning_rate": 9.648992926179208e-06,
+      "loss": 1.7016,
+      "step": 3620
+    },
+    {
+      "epoch": 0.10588474852372226,
+      "grad_norm": 13.4375,
+      "learning_rate": 9.647053644506591e-06,
+      "loss": 1.6735,
+      "step": 3640
+    },
+    {
+      "epoch": 0.10646653285627018,
+      "grad_norm": 13.75,
+      "learning_rate": 9.645114362833976e-06,
+      "loss": 1.7366,
+      "step": 3660
+    },
+    {
+      "epoch": 0.1070483171888181,
+      "grad_norm": 12.75,
+      "learning_rate": 9.643175081161361e-06,
+      "loss": 1.6118,
+      "step": 3680
+    },
+    {
+      "epoch": 0.10763010152136603,
+      "grad_norm": 14.75,
+      "learning_rate": 9.641235799488747e-06,
+      "loss": 1.7359,
+      "step": 3700
+    },
+    {
+      "epoch": 0.10821188585391395,
+      "grad_norm": 13.0625,
+      "learning_rate": 9.639296517816132e-06,
+      "loss": 1.687,
+      "step": 3720
+    },
+    {
+      "epoch": 0.10879367018646188,
+      "grad_norm": 12.75,
+      "learning_rate": 9.637357236143517e-06,
+      "loss": 1.6533,
+      "step": 3740
+    },
+    {
+      "epoch": 0.1093754545190098,
+      "grad_norm": 13.9375,
+      "learning_rate": 9.635417954470902e-06,
+      "loss": 1.742,
+      "step": 3760
+    },
+    {
+      "epoch": 0.10995723885155773,
+      "grad_norm": 10.8125,
+      "learning_rate": 9.633478672798287e-06,
+      "loss": 1.6269,
+      "step": 3780
+    },
+    {
+      "epoch": 0.11053902318410565,
+      "grad_norm": 11.625,
+      "learning_rate": 9.631539391125672e-06,
+      "loss": 1.7434,
+      "step": 3800
+    },
+    {
+      "epoch": 0.11112080751665358,
+      "grad_norm": 12.1875,
+      "learning_rate": 9.629600109453057e-06,
+      "loss": 1.6067,
+      "step": 3820
+    },
+    {
+      "epoch": 0.1117025918492015,
+      "grad_norm": 14.375,
+      "learning_rate": 9.627660827780442e-06,
+      "loss": 1.6903,
+      "step": 3840
+    },
+    {
+      "epoch": 0.11228437618174943,
+      "grad_norm": 13.9375,
+      "learning_rate": 9.625721546107827e-06,
+      "loss": 1.6103,
+      "step": 3860
+    },
+    {
+      "epoch": 0.11286616051429735,
+      "grad_norm": 13.1875,
+      "learning_rate": 9.623782264435212e-06,
+      "loss": 1.7284,
+      "step": 3880
+    },
+    {
+      "epoch": 0.11344794484684527,
+      "grad_norm": 14.375,
+      "learning_rate": 9.621842982762598e-06,
+      "loss": 1.6879,
+      "step": 3900
+    },
+    {
+      "epoch": 0.1140297291793932,
+      "grad_norm": 13.9375,
+      "learning_rate": 9.619903701089983e-06,
+      "loss": 1.7217,
+      "step": 3920
+    },
+    {
+      "epoch": 0.11461151351194113,
+      "grad_norm": 10.125,
+      "learning_rate": 9.617964419417368e-06,
+      "loss": 1.6572,
+      "step": 3940
+    },
+    {
+      "epoch": 0.11519329784448905,
+      "grad_norm": 12.6875,
+      "learning_rate": 9.616025137744753e-06,
+      "loss": 1.7197,
+      "step": 3960
+    },
+    {
+      "epoch": 0.11577508217703697,
+      "grad_norm": 11.8125,
+      "learning_rate": 9.614085856072138e-06,
+      "loss": 1.7059,
+      "step": 3980
+    },
+    {
+      "epoch": 0.11635686650958489,
+      "grad_norm": 10.4375,
+      "learning_rate": 9.612146574399523e-06,
+      "loss": 1.6629,
+      "step": 4000
+    },
+    {
+      "epoch": 0.11693865084213283,
+      "grad_norm": 15.75,
+      "learning_rate": 9.610207292726908e-06,
+      "loss": 1.5862,
+      "step": 4020
+    },
+    {
+      "epoch": 0.11752043517468075,
+      "grad_norm": 12.9375,
+      "learning_rate": 9.608268011054293e-06,
+      "loss": 1.7666,
+      "step": 4040
+    },
+    {
+      "epoch": 0.11810221950722867,
+      "grad_norm": 11.1875,
+      "learning_rate": 9.606328729381678e-06,
+      "loss": 1.7658,
+      "step": 4060
+    },
+    {
+      "epoch": 0.11868400383977659,
+      "grad_norm": 12.1875,
+      "learning_rate": 9.604389447709063e-06,
+      "loss": 1.6754,
+      "step": 4080
+    },
+    {
+      "epoch": 0.11926578817232451,
+      "grad_norm": 12.625,
+      "learning_rate": 9.602450166036448e-06,
+      "loss": 1.7163,
+      "step": 4100
+    },
+    {
+      "epoch": 0.11984757250487245,
+      "grad_norm": 12.0,
+      "learning_rate": 9.600510884363834e-06,
+      "loss": 1.7055,
+      "step": 4120
+    },
+    {
+      "epoch": 0.12042935683742037,
+      "grad_norm": 13.8125,
+      "learning_rate": 9.598571602691219e-06,
+      "loss": 1.6927,
+      "step": 4140
+    },
+    {
+      "epoch": 0.12101114116996829,
+      "grad_norm": 11.375,
+      "learning_rate": 9.596632321018604e-06,
+      "loss": 1.6708,
+      "step": 4160
+    },
+    {
+      "epoch": 0.12159292550251621,
+      "grad_norm": 12.3125,
+      "learning_rate": 9.594693039345989e-06,
+      "loss": 1.6681,
+      "step": 4180
+    },
+    {
+      "epoch": 0.12217470983506414,
+      "grad_norm": 13.125,
+      "learning_rate": 9.592753757673374e-06,
+      "loss": 1.6725,
+      "step": 4200
+    },
+    {
+      "epoch": 0.12275649416761207,
+      "grad_norm": 15.0625,
+      "learning_rate": 9.590814476000759e-06,
+      "loss": 1.6429,
+      "step": 4220
+    },
+    {
+      "epoch": 0.12333827850015999,
+      "grad_norm": 12.625,
+      "learning_rate": 9.588875194328144e-06,
+      "loss": 1.67,
+      "step": 4240
+    },
+    {
+      "epoch": 0.12392006283270791,
+      "grad_norm": 10.9375,
+      "learning_rate": 9.58693591265553e-06,
+      "loss": 1.6725,
+      "step": 4260
+    },
+    {
+      "epoch": 0.12450184716525584,
+      "grad_norm": 12.1875,
+      "learning_rate": 9.584996630982914e-06,
+      "loss": 1.6875,
+      "step": 4280
+    },
+    {
+      "epoch": 0.12508363149780377,
+      "grad_norm": 12.125,
+      "learning_rate": 9.5830573493103e-06,
+      "loss": 1.6594,
+      "step": 4300
+    },
+    {
+      "epoch": 0.12566541583035168,
+      "grad_norm": 12.6875,
+      "learning_rate": 9.581118067637685e-06,
+      "loss": 1.6841,
+      "step": 4320
+    },
+    {
+      "epoch": 0.12624720016289961,
+      "grad_norm": 12.75,
+      "learning_rate": 9.57917878596507e-06,
+      "loss": 1.6427,
+      "step": 4340
+    },
+    {
+      "epoch": 0.12682898449544755,
+      "grad_norm": 13.625,
+      "learning_rate": 9.577239504292455e-06,
+      "loss": 1.6997,
+      "step": 4360
+    },
+    {
+      "epoch": 0.12741076882799546,
+      "grad_norm": 14.75,
+      "learning_rate": 9.57530022261984e-06,
+      "loss": 1.732,
+      "step": 4380
+    },
+    {
+      "epoch": 0.1279925531605434,
+      "grad_norm": 15.125,
+      "learning_rate": 9.573360940947225e-06,
+      "loss": 1.8135,
+      "step": 4400
+    },
+    {
+      "epoch": 0.1285743374930913,
+      "grad_norm": 11.75,
+      "learning_rate": 9.57142165927461e-06,
+      "loss": 1.6542,
+      "step": 4420
+    },
+    {
+      "epoch": 0.12915612182563924,
+      "grad_norm": 11.5,
+      "learning_rate": 9.569482377601995e-06,
+      "loss": 1.6981,
+      "step": 4440
+    },
+    {
+      "epoch": 0.12973790615818717,
+      "grad_norm": 11.5,
+      "learning_rate": 9.56754309592938e-06,
+      "loss": 1.6306,
+      "step": 4460
+    },
+    {
+      "epoch": 0.13031969049073508,
+      "grad_norm": 10.75,
+      "learning_rate": 9.565603814256765e-06,
+      "loss": 1.6223,
+      "step": 4480
+    },
+    {
+      "epoch": 0.13090147482328301,
+      "grad_norm": 13.1875,
+      "learning_rate": 9.56366453258415e-06,
+      "loss": 1.6891,
+      "step": 4500
+    },
+    {
+      "epoch": 0.13148325915583092,
+      "grad_norm": 17.125,
+      "learning_rate": 9.561725250911536e-06,
+      "loss": 1.691,
+      "step": 4520
+    },
+    {
+      "epoch": 0.13206504348837886,
+      "grad_norm": 13.6875,
+      "learning_rate": 9.55978596923892e-06,
+      "loss": 1.5271,
+      "step": 4540
+    },
+    {
+      "epoch": 0.1326468278209268,
+      "grad_norm": 11.375,
+      "learning_rate": 9.557846687566306e-06,
+      "loss": 1.6127,
+      "step": 4560
+    },
+    {
+      "epoch": 0.1332286121534747,
+      "grad_norm": 15.1875,
+      "learning_rate": 9.55590740589369e-06,
+      "loss": 1.7001,
+      "step": 4580
+    },
+    {
+      "epoch": 0.13381039648602264,
+      "grad_norm": 15.4375,
+      "learning_rate": 9.553968124221076e-06,
+      "loss": 1.6584,
+      "step": 4600
+    },
+    {
+      "epoch": 0.13439218081857054,
+      "grad_norm": 11.8125,
+      "learning_rate": 9.552028842548461e-06,
+      "loss": 1.6233,
+      "step": 4620
+    },
+    {
+      "epoch": 0.13497396515111848,
+      "grad_norm": 15.625,
+      "learning_rate": 9.550089560875846e-06,
+      "loss": 1.7007,
+      "step": 4640
+    },
+    {
+      "epoch": 0.13555574948366642,
+      "grad_norm": 12.9375,
+      "learning_rate": 9.548150279203231e-06,
+      "loss": 1.7278,
+      "step": 4660
+    },
+    {
+      "epoch": 0.13613753381621432,
+      "grad_norm": 12.6875,
+      "learning_rate": 9.546210997530616e-06,
+      "loss": 1.7739,
+      "step": 4680
+    },
+    {
+      "epoch": 0.13671931814876226,
+      "grad_norm": 13.4375,
+      "learning_rate": 9.544271715858001e-06,
+      "loss": 1.6601,
+      "step": 4700
+    },
+    {
+      "epoch": 0.13730110248131017,
+      "grad_norm": 15.0625,
+      "learning_rate": 9.542332434185387e-06,
+      "loss": 1.6969,
+      "step": 4720
+    },
+    {
+      "epoch": 0.1378828868138581,
+      "grad_norm": 14.4375,
+      "learning_rate": 9.540393152512772e-06,
+      "loss": 1.6679,
+      "step": 4740
+    },
+    {
+      "epoch": 0.13846467114640604,
+      "grad_norm": 12.375,
+      "learning_rate": 9.538453870840157e-06,
+      "loss": 1.6858,
+      "step": 4760
+    },
+    {
+      "epoch": 0.13904645547895395,
+      "grad_norm": 11.3125,
+      "learning_rate": 9.536514589167542e-06,
+      "loss": 1.6907,
+      "step": 4780
+    },
+    {
+      "epoch": 0.13962823981150188,
+      "grad_norm": 11.8125,
+      "learning_rate": 9.534575307494927e-06,
+      "loss": 1.6233,
+      "step": 4800
+    },
+    {
+      "epoch": 0.1402100241440498,
+      "grad_norm": 12.75,
+      "learning_rate": 9.532636025822312e-06,
+      "loss": 1.689,
+      "step": 4820
+    },
+    {
+      "epoch": 0.14079180847659772,
+      "grad_norm": 13.1875,
+      "learning_rate": 9.530696744149697e-06,
+      "loss": 1.6551,
+      "step": 4840
+    },
+    {
+      "epoch": 0.14137359280914566,
+      "grad_norm": 13.8125,
+      "learning_rate": 9.528757462477082e-06,
+      "loss": 1.6108,
+      "step": 4860
+    },
+    {
+      "epoch": 0.14195537714169357,
+      "grad_norm": 11.625,
+      "learning_rate": 9.526818180804467e-06,
+      "loss": 1.7142,
+      "step": 4880
+    },
+    {
+      "epoch": 0.1425371614742415,
+      "grad_norm": 12.75,
+      "learning_rate": 9.524878899131852e-06,
+      "loss": 1.6855,
+      "step": 4900
+    },
+    {
+      "epoch": 0.1431189458067894,
+      "grad_norm": 10.0,
+      "learning_rate": 9.522939617459237e-06,
+      "loss": 1.7138,
+      "step": 4920
+    },
+    {
+      "epoch": 0.14370073013933735,
+      "grad_norm": 11.75,
+      "learning_rate": 9.521000335786623e-06,
+      "loss": 1.7594,
+      "step": 4940
+    },
+    {
+      "epoch": 0.14428251447188528,
+      "grad_norm": 12.0625,
+      "learning_rate": 9.519061054114008e-06,
+      "loss": 1.6585,
+      "step": 4960
+    },
+    {
+      "epoch": 0.1448642988044332,
+      "grad_norm": 14.5625,
+      "learning_rate": 9.517121772441393e-06,
+      "loss": 1.7204,
+      "step": 4980
+    },
+    {
+      "epoch": 0.14544608313698112,
+      "grad_norm": 12.5,
+      "learning_rate": 9.515182490768778e-06,
+      "loss": 1.6766,
+      "step": 5000
+    },
+    {
+      "epoch": 0.14602786746952903,
+      "grad_norm": 13.125,
+      "learning_rate": 9.513243209096163e-06,
+      "loss": 1.7624,
+      "step": 5020
+    },
+    {
+      "epoch": 0.14660965180207697,
+      "grad_norm": 12.1875,
+      "learning_rate": 9.511303927423548e-06,
+      "loss": 1.7071,
+      "step": 5040
+    },
+    {
+      "epoch": 0.1471914361346249,
+      "grad_norm": 12.8125,
+      "learning_rate": 9.509364645750933e-06,
+      "loss": 1.7172,
+      "step": 5060
+    },
+    {
+      "epoch": 0.1477732204671728,
+      "grad_norm": 10.9375,
+      "learning_rate": 9.507425364078318e-06,
+      "loss": 1.7156,
+      "step": 5080
+    },
+    {
+      "epoch": 0.14835500479972075,
+      "grad_norm": 12.9375,
+      "learning_rate": 9.505486082405703e-06,
+      "loss": 1.6806,
+      "step": 5100
+    },
+    {
+      "epoch": 0.14893678913226865,
+      "grad_norm": 13.0,
+      "learning_rate": 9.503546800733088e-06,
+      "loss": 1.652,
+      "step": 5120
+    },
+    {
+      "epoch": 0.1495185734648166,
+      "grad_norm": 14.0625,
+      "learning_rate": 9.501607519060474e-06,
+      "loss": 1.6509,
+      "step": 5140
+    },
+    {
+      "epoch": 0.15010035779736453,
+      "grad_norm": 12.0,
+      "learning_rate": 9.499668237387859e-06,
+      "loss": 1.7384,
+      "step": 5160
+    },
+    {
+      "epoch": 0.15068214212991243,
+      "grad_norm": 13.25,
+      "learning_rate": 9.497728955715244e-06,
+      "loss": 1.6139,
+      "step": 5180
+    },
+    {
+      "epoch": 0.15126392646246037,
+      "grad_norm": 13.3125,
+      "learning_rate": 9.495789674042629e-06,
+      "loss": 1.708,
+      "step": 5200
+    },
+    {
+      "epoch": 0.1518457107950083,
+      "grad_norm": 15.6875,
+      "learning_rate": 9.493850392370014e-06,
+      "loss": 1.6423,
+      "step": 5220
+    },
+    {
+      "epoch": 0.1524274951275562,
+      "grad_norm": 12.9375,
+      "learning_rate": 9.491911110697399e-06,
+      "loss": 1.7107,
+      "step": 5240
+    },
+    {
+      "epoch": 0.15300927946010415,
+      "grad_norm": 12.9375,
+      "learning_rate": 9.489971829024784e-06,
+      "loss": 1.6481,
+      "step": 5260
+    },
+    {
+      "epoch": 0.15359106379265205,
+      "grad_norm": 12.875,
+      "learning_rate": 9.488032547352168e-06,
+      "loss": 1.5836,
+      "step": 5280
+    },
+    {
+      "epoch": 0.1541728481252,
+      "grad_norm": 10.6875,
+      "learning_rate": 9.486093265679553e-06,
+      "loss": 1.6645,
+      "step": 5300
+    },
+    {
+      "epoch": 0.15475463245774793,
+      "grad_norm": 11.875,
+      "learning_rate": 9.484153984006938e-06,
+      "loss": 1.7185,
+      "step": 5320
+    },
+    {
+      "epoch": 0.15533641679029583,
+      "grad_norm": 13.0625,
+      "learning_rate": 9.482214702334323e-06,
+      "loss": 1.5641,
+      "step": 5340
+    },
+    {
+      "epoch": 0.15591820112284377,
+      "grad_norm": 12.875,
+      "learning_rate": 9.480275420661708e-06,
+      "loss": 1.7268,
+      "step": 5360
+    },
+    {
+      "epoch": 0.15649998545539168,
+      "grad_norm": 9.9375,
+      "learning_rate": 9.478336138989093e-06,
+      "loss": 1.6341,
+      "step": 5380
+    },
+    {
+      "epoch": 0.1570817697879396,
+      "grad_norm": 15.8125,
+      "learning_rate": 9.476396857316478e-06,
+      "loss": 1.7263,
+      "step": 5400
+    },
+    {
+      "epoch": 0.15766355412048755,
+      "grad_norm": 12.9375,
+      "learning_rate": 9.474457575643863e-06,
+      "loss": 1.7495,
+      "step": 5420
+    },
+    {
+      "epoch": 0.15824533845303546,
+      "grad_norm": 14.9375,
+      "learning_rate": 9.472518293971248e-06,
+      "loss": 1.7751,
+      "step": 5440
+    },
+    {
+      "epoch": 0.1588271227855834,
+      "grad_norm": 12.375,
+      "learning_rate": 9.470579012298633e-06,
+      "loss": 1.6839,
+      "step": 5460
+    },
+    {
+      "epoch": 0.1594089071181313,
+      "grad_norm": 13.0,
+      "learning_rate": 9.468639730626019e-06,
+      "loss": 1.6549,
+      "step": 5480
+    },
+    {
+      "epoch": 0.15999069145067923,
+      "grad_norm": 12.125,
+      "learning_rate": 9.466700448953404e-06,
+      "loss": 1.6218,
+      "step": 5500
+    },
+    {
+      "epoch": 0.16057247578322717,
+      "grad_norm": 10.125,
+      "learning_rate": 9.464761167280789e-06,
+      "loss": 1.7006,
+      "step": 5520
+    },
+    {
+      "epoch": 0.16115426011577508,
+      "grad_norm": 14.25,
+      "learning_rate": 9.462821885608174e-06,
+      "loss": 1.7311,
+      "step": 5540
+    },
+    {
+      "epoch": 0.161736044448323,
+      "grad_norm": 11.125,
+      "learning_rate": 9.460882603935559e-06,
+      "loss": 1.6975,
+      "step": 5560
+    },
+    {
+      "epoch": 0.16231782878087092,
+      "grad_norm": 10.0625,
+      "learning_rate": 9.458943322262944e-06,
+      "loss": 1.6798,
+      "step": 5580
+    },
+    {
+      "epoch": 0.16289961311341886,
+      "grad_norm": 11.5625,
+      "learning_rate": 9.457004040590329e-06,
+      "loss": 1.7398,
+      "step": 5600
+    },
+    {
+      "epoch": 0.1634813974459668,
+      "grad_norm": 12.125,
+      "learning_rate": 9.455064758917714e-06,
+      "loss": 1.6346,
+      "step": 5620
+    },
+    {
+      "epoch": 0.1640631817785147,
+      "grad_norm": 16.375,
+      "learning_rate": 9.4531254772451e-06,
+      "loss": 1.6982,
+      "step": 5640
+    },
+    {
+      "epoch": 0.16464496611106264,
+      "grad_norm": 13.625,
+      "learning_rate": 9.451186195572484e-06,
+      "loss": 1.6089,
+      "step": 5660
+    },
+    {
+      "epoch": 0.16522675044361054,
+      "grad_norm": 12.125,
+      "learning_rate": 9.44924691389987e-06,
+      "loss": 1.5636,
+      "step": 5680
+    },
+    {
+      "epoch": 0.16580853477615848,
+      "grad_norm": 11.75,
+      "learning_rate": 9.447307632227255e-06,
+      "loss": 1.6292,
+      "step": 5700
+    },
+    {
+      "epoch": 0.1663903191087064,
+      "grad_norm": 11.8125,
+      "learning_rate": 9.44536835055464e-06,
+      "loss": 1.6882,
+      "step": 5720
+    },
+    {
+      "epoch": 0.16697210344125432,
+      "grad_norm": 12.125,
+      "learning_rate": 9.443429068882025e-06,
+      "loss": 1.5953,
+      "step": 5740
+    },
+    {
+      "epoch": 0.16755388777380226,
+      "grad_norm": 11.0,
+      "learning_rate": 9.44148978720941e-06,
+      "loss": 1.6087,
+      "step": 5760
+    },
+    {
+      "epoch": 0.16813567210635016,
+      "grad_norm": 12.5,
+      "learning_rate": 9.439550505536795e-06,
+      "loss": 1.6753,
+      "step": 5780
+    },
+    {
+      "epoch": 0.1687174564388981,
+      "grad_norm": 12.875,
+      "learning_rate": 9.43761122386418e-06,
+      "loss": 1.6255,
+      "step": 5800
+    },
+    {
+      "epoch": 0.16929924077144604,
+      "grad_norm": 13.0625,
+      "learning_rate": 9.435671942191565e-06,
+      "loss": 1.6765,
+      "step": 5820
+    },
+    {
+      "epoch": 0.16988102510399394,
+      "grad_norm": 9.6875,
+      "learning_rate": 9.43373266051895e-06,
+      "loss": 1.7139,
+      "step": 5840
+    },
+    {
+      "epoch": 0.17046280943654188,
+      "grad_norm": 10.6875,
+      "learning_rate": 9.431793378846335e-06,
+      "loss": 1.7313,
+      "step": 5860
+    },
+    {
+      "epoch": 0.1710445937690898,
+      "grad_norm": 12.8125,
+      "learning_rate": 9.42985409717372e-06,
+      "loss": 1.6744,
+      "step": 5880
+    },
+    {
+      "epoch": 0.17162637810163772,
+      "grad_norm": 11.875,
+      "learning_rate": 9.427914815501106e-06,
+      "loss": 1.6938,
+      "step": 5900
+    },
+    {
+      "epoch": 0.17220816243418566,
+      "grad_norm": 13.125,
+      "learning_rate": 9.42597553382849e-06,
+      "loss": 1.7012,
+      "step": 5920
+    },
+    {
+      "epoch": 0.17278994676673357,
+      "grad_norm": 10.625,
+      "learning_rate": 9.424036252155876e-06,
+      "loss": 1.6517,
+      "step": 5940
+    },
+    {
+      "epoch": 0.1733717310992815,
+      "grad_norm": 13.5625,
+      "learning_rate": 9.422096970483261e-06,
+      "loss": 1.7294,
+      "step": 5960
+    },
+    {
+      "epoch": 0.1739535154318294,
+      "grad_norm": 9.25,
+      "learning_rate": 9.420157688810646e-06,
+      "loss": 1.761,
+      "step": 5980
+    },
+    {
+      "epoch": 0.17453529976437734,
+      "grad_norm": 12.9375,
+      "learning_rate": 9.418218407138031e-06,
+      "loss": 1.7016,
+      "step": 6000
+    },
+    {
+      "epoch": 0.17511708409692528,
+      "grad_norm": 13.875,
+      "learning_rate": 9.416279125465416e-06,
+      "loss": 1.6872,
+      "step": 6020
+    },
+    {
+      "epoch": 0.1756988684294732,
+      "grad_norm": 12.0625,
+      "learning_rate": 9.414339843792801e-06,
+      "loss": 1.6448,
+      "step": 6040
+    },
+    {
+      "epoch": 0.17628065276202112,
+      "grad_norm": 14.6875,
+      "learning_rate": 9.412400562120186e-06,
+      "loss": 1.7946,
+      "step": 6060
+    },
+    {
+      "epoch": 0.17686243709456903,
+      "grad_norm": 11.875,
+      "learning_rate": 9.410461280447571e-06,
+      "loss": 1.5854,
+      "step": 6080
+    },
+    {
+      "epoch": 0.17744422142711697,
+      "grad_norm": 15.75,
+      "learning_rate": 9.408521998774957e-06,
+      "loss": 1.6513,
+      "step": 6100
+    },
+    {
+      "epoch": 0.1780260057596649,
+      "grad_norm": 12.0625,
+      "learning_rate": 9.406582717102342e-06,
+      "loss": 1.6085,
+      "step": 6120
+    },
+    {
+      "epoch": 0.1786077900922128,
+      "grad_norm": 14.25,
+      "learning_rate": 9.404643435429727e-06,
+      "loss": 1.7134,
+      "step": 6140
+    },
+    {
+      "epoch": 0.17918957442476074,
+      "grad_norm": 9.25,
+      "learning_rate": 9.402704153757112e-06,
+      "loss": 1.6351,
+      "step": 6160
+    },
+    {
+      "epoch": 0.17977135875730865,
+      "grad_norm": 12.375,
+      "learning_rate": 9.400764872084497e-06,
+      "loss": 1.6563,
+      "step": 6180
+    },
+    {
+      "epoch": 0.1803531430898566,
+      "grad_norm": 15.125,
+      "learning_rate": 9.398825590411882e-06,
+      "loss": 1.6515,
+      "step": 6200
+    },
+    {
+      "epoch": 0.18093492742240452,
+      "grad_norm": 12.8125,
+      "learning_rate": 9.396886308739267e-06,
+      "loss": 1.593,
+      "step": 6220
+    },
+    {
+      "epoch": 0.18151671175495243,
+      "grad_norm": 11.0625,
+      "learning_rate": 9.394947027066652e-06,
+      "loss": 1.6267,
+      "step": 6240
+    },
+    {
+      "epoch": 0.18209849608750037,
+      "grad_norm": 13.875,
+      "learning_rate": 9.393007745394037e-06,
+      "loss": 1.6304,
+      "step": 6260
+    },
+    {
+      "epoch": 0.18268028042004827,
+      "grad_norm": 11.375,
+      "learning_rate": 9.391068463721422e-06,
+      "loss": 1.7136,
+      "step": 6280
+    },
+    {
+      "epoch": 0.1832620647525962,
+      "grad_norm": 12.8125,
+      "learning_rate": 9.389129182048808e-06,
+      "loss": 1.7938,
+      "step": 6300
+    },
+    {
+      "epoch": 0.18384384908514415,
+      "grad_norm": 11.625,
+      "learning_rate": 9.387189900376193e-06,
+      "loss": 1.7028,
+      "step": 6320
+    },
+    {
+      "epoch": 0.18442563341769205,
+      "grad_norm": 11.25,
+      "learning_rate": 9.385250618703578e-06,
+      "loss": 1.6818,
+      "step": 6340
+    },
+    {
+      "epoch": 0.18500741775024,
+      "grad_norm": 12.25,
+      "learning_rate": 9.383311337030963e-06,
+      "loss": 1.668,
+      "step": 6360
+    },
+    {
+      "epoch": 0.18558920208278792,
+      "grad_norm": 19.125,
+      "learning_rate": 9.381372055358348e-06,
+      "loss": 1.6622,
+      "step": 6380
+    },
+    {
+      "epoch": 0.18617098641533583,
+      "grad_norm": 12.125,
+      "learning_rate": 9.379432773685733e-06,
+      "loss": 1.6649,
+      "step": 6400
+    },
+    {
+      "epoch": 0.18675277074788377,
+      "grad_norm": 12.3125,
+      "learning_rate": 9.377493492013118e-06,
+      "loss": 1.6025,
+      "step": 6420
+    },
+    {
+      "epoch": 0.18733455508043168,
+      "grad_norm": 11.4375,
+      "learning_rate": 9.375554210340503e-06,
+      "loss": 1.7254,
+      "step": 6440
+    },
+    {
+      "epoch": 0.1879163394129796,
+      "grad_norm": 13.9375,
+      "learning_rate": 9.373614928667888e-06,
+      "loss": 1.6749,
+      "step": 6460
+    },
+    {
+      "epoch": 0.18849812374552755,
+      "grad_norm": 16.375,
+      "learning_rate": 9.371675646995273e-06,
+      "loss": 1.7615,
+      "step": 6480
+    },
+    {
+      "epoch": 0.18907990807807545,
+      "grad_norm": 11.1875,
+      "learning_rate": 9.369736365322659e-06,
+      "loss": 1.582,
+      "step": 6500
+    },
+    {
+      "epoch": 0.1896616924106234,
+      "grad_norm": 14.4375,
+      "learning_rate": 9.367797083650044e-06,
+      "loss": 1.6986,
+      "step": 6520
+    },
+    {
+      "epoch": 0.1902434767431713,
+      "grad_norm": 15.25,
+      "learning_rate": 9.365857801977429e-06,
+      "loss": 1.6597,
+      "step": 6540
+    },
+    {
+      "epoch": 0.19082526107571923,
+      "grad_norm": 12.5,
+      "learning_rate": 9.363918520304814e-06,
+      "loss": 1.6914,
+      "step": 6560
+    },
+    {
+      "epoch": 0.19140704540826717,
+      "grad_norm": 13.8125,
+      "learning_rate": 9.361979238632199e-06,
+      "loss": 1.6375,
+      "step": 6580
+    },
+    {
+      "epoch": 0.19198882974081508,
+      "grad_norm": 12.125,
+      "learning_rate": 9.360039956959584e-06,
+      "loss": 1.6513,
+      "step": 6600
+    },
+    {
+      "epoch": 0.192570614073363,
+      "grad_norm": 13.625,
+      "learning_rate": 9.358100675286969e-06,
+      "loss": 1.6101,
+      "step": 6620
+    },
+    {
+      "epoch": 0.19315239840591092,
+      "grad_norm": 10.875,
+      "learning_rate": 9.356161393614354e-06,
+      "loss": 1.686,
+      "step": 6640
+    },
+    {
+      "epoch": 0.19373418273845885,
+      "grad_norm": 11.4375,
+      "learning_rate": 9.35422211194174e-06,
+      "loss": 1.7196,
+      "step": 6660
+    },
+    {
+      "epoch": 0.1943159670710068,
+      "grad_norm": 12.5,
+      "learning_rate": 9.352282830269124e-06,
+      "loss": 1.6264,
+      "step": 6680
+    },
+    {
+      "epoch": 0.1948977514035547,
+      "grad_norm": 12.375,
+      "learning_rate": 9.35034354859651e-06,
+      "loss": 1.6006,
+      "step": 6700
+    },
+    {
+      "epoch": 0.19547953573610263,
+      "grad_norm": 13.0625,
+      "learning_rate": 9.348404266923895e-06,
+      "loss": 1.6414,
+      "step": 6720
+    },
+    {
+      "epoch": 0.19606132006865054,
+      "grad_norm": 12.3125,
+      "learning_rate": 9.34646498525128e-06,
+      "loss": 1.622,
+      "step": 6740
+    },
+    {
+      "epoch": 0.19664310440119848,
+      "grad_norm": 15.9375,
+      "learning_rate": 9.344525703578665e-06,
+      "loss": 1.6411,
+      "step": 6760
+    },
+    {
+      "epoch": 0.1972248887337464,
+      "grad_norm": 13.0625,
+      "learning_rate": 9.34258642190605e-06,
+      "loss": 1.6437,
+      "step": 6780
+    },
+    {
+      "epoch": 0.19780667306629432,
+      "grad_norm": 14.75,
+      "learning_rate": 9.340647140233435e-06,
+      "loss": 1.7252,
+      "step": 6800
+    },
+    {
+      "epoch": 0.19838845739884226,
+      "grad_norm": 14.25,
+      "learning_rate": 9.33870785856082e-06,
+      "loss": 1.6502,
+      "step": 6820
+    },
+    {
+      "epoch": 0.19897024173139016,
+      "grad_norm": 13.0,
+      "learning_rate": 9.336768576888205e-06,
+      "loss": 1.7225,
+      "step": 6840
+    },
+    {
+      "epoch": 0.1995520260639381,
+      "grad_norm": 17.0,
+      "learning_rate": 9.33482929521559e-06,
+      "loss": 1.6959,
+      "step": 6860
+    },
+    {
+      "epoch": 0.20013381039648603,
+      "grad_norm": 12.1875,
+      "learning_rate": 9.332890013542975e-06,
+      "loss": 1.6273,
+      "step": 6880
+    },
+    {
+      "epoch": 0.20071559472903394,
+      "grad_norm": 11.875,
+      "learning_rate": 9.330950731870359e-06,
+      "loss": 1.7678,
+      "step": 6900
+    },
+    {
+      "epoch": 0.20129737906158188,
+      "grad_norm": 10.375,
+      "learning_rate": 9.329011450197744e-06,
+      "loss": 1.7142,
+      "step": 6920
+    },
+    {
+      "epoch": 0.20187916339412978,
+      "grad_norm": 13.375,
+      "learning_rate": 9.327072168525129e-06,
+      "loss": 1.6364,
+      "step": 6940
+    },
+    {
+      "epoch": 0.20246094772667772,
+      "grad_norm": 14.25,
+      "learning_rate": 9.325132886852514e-06,
+      "loss": 1.7133,
+      "step": 6960
+    },
+    {
+      "epoch": 0.20304273205922566,
+      "grad_norm": 13.375,
+      "learning_rate": 9.323193605179899e-06,
+      "loss": 1.7386,
+      "step": 6980
+    },
+    {
+      "epoch": 0.20362451639177356,
+      "grad_norm": 10.625,
+      "learning_rate": 9.321254323507284e-06,
+      "loss": 1.6479,
+      "step": 7000
+    },
+    {
+      "epoch": 0.2042063007243215,
+      "grad_norm": 12.8125,
+      "learning_rate": 9.31931504183467e-06,
+      "loss": 1.7419,
+      "step": 7020
+    },
+    {
+      "epoch": 0.2047880850568694,
+      "grad_norm": 14.5625,
+      "learning_rate": 9.317375760162054e-06,
+      "loss": 1.6332,
+      "step": 7040
+    },
+    {
+      "epoch": 0.20536986938941734,
+      "grad_norm": 15.3125,
+      "learning_rate": 9.31543647848944e-06,
+      "loss": 1.7959,
+      "step": 7060
+    },
+    {
+      "epoch": 0.20595165372196528,
+      "grad_norm": 13.9375,
+      "learning_rate": 9.313497196816825e-06,
+      "loss": 1.5956,
+      "step": 7080
+    },
+    {
+      "epoch": 0.20653343805451319,
+      "grad_norm": 9.4375,
+      "learning_rate": 9.31155791514421e-06,
+      "loss": 1.5998,
+      "step": 7100
+    },
+    {
+      "epoch": 0.20711522238706112,
+      "grad_norm": 12.5625,
+      "learning_rate": 9.309618633471595e-06,
+      "loss": 1.6106,
+      "step": 7120
+    },
+    {
+      "epoch": 0.20769700671960903,
+      "grad_norm": 11.375,
+      "learning_rate": 9.30767935179898e-06,
+      "loss": 1.6298,
+      "step": 7140
+    },
+    {
+      "epoch": 0.20827879105215696,
+      "grad_norm": 12.75,
+      "learning_rate": 9.305740070126365e-06,
+      "loss": 1.7399,
+      "step": 7160
+    },
+    {
+      "epoch": 0.2088605753847049,
+      "grad_norm": 14.3125,
+      "learning_rate": 9.30380078845375e-06,
+      "loss": 1.649,
+      "step": 7180
+    },
+    {
+      "epoch": 0.2094423597172528,
+      "grad_norm": 9.1875,
+      "learning_rate": 9.301861506781135e-06,
+      "loss": 1.6379,
+      "step": 7200
+    },
+    {
+      "epoch": 0.21002414404980074,
+      "grad_norm": 11.0,
+      "learning_rate": 9.29992222510852e-06,
+      "loss": 1.7574,
+      "step": 7220
+    },
+    {
+      "epoch": 0.21060592838234865,
+      "grad_norm": 10.9375,
+      "learning_rate": 9.297982943435905e-06,
+      "loss": 1.7173,
+      "step": 7240
+    },
+    {
+      "epoch": 0.21118771271489659,
+      "grad_norm": 13.0625,
+      "learning_rate": 9.29604366176329e-06,
+      "loss": 1.6358,
+      "step": 7260
+    },
+    {
+      "epoch": 0.21176949704744452,
+      "grad_norm": 11.375,
+      "learning_rate": 9.294104380090676e-06,
+      "loss": 1.677,
+      "step": 7280
+    },
+    {
+      "epoch": 0.21235128137999243,
+      "grad_norm": 12.8125,
+      "learning_rate": 9.29216509841806e-06,
+      "loss": 1.6471,
+      "step": 7300
+    },
+    {
+      "epoch": 0.21293306571254036,
+      "grad_norm": 11.625,
+      "learning_rate": 9.290225816745446e-06,
+      "loss": 1.7079,
+      "step": 7320
+    },
+    {
+      "epoch": 0.21351485004508827,
+      "grad_norm": 13.0,
+      "learning_rate": 9.288286535072831e-06,
+      "loss": 1.6672,
+      "step": 7340
+    },
+    {
+      "epoch": 0.2140966343776362,
+      "grad_norm": 12.75,
+      "learning_rate": 9.286347253400216e-06,
+      "loss": 1.512,
+      "step": 7360
+    },
+    {
+      "epoch": 0.21467841871018414,
+      "grad_norm": 13.1875,
+      "learning_rate": 9.284407971727601e-06,
+      "loss": 1.6741,
+      "step": 7380
+    },
+    {
+      "epoch": 0.21526020304273205,
+      "grad_norm": 11.8125,
+      "learning_rate": 9.282468690054986e-06,
+      "loss": 1.7275,
+      "step": 7400
+    },
+    {
+      "epoch": 0.21584198737528,
+      "grad_norm": 12.4375,
+      "learning_rate": 9.280529408382371e-06,
+      "loss": 1.6383,
+      "step": 7420
+    },
+    {
+      "epoch": 0.2164237717078279,
+      "grad_norm": 11.125,
+      "learning_rate": 9.278590126709756e-06,
+      "loss": 1.6778,
+      "step": 7440
+    },
+    {
+      "epoch": 0.21700555604037583,
+      "grad_norm": 12.875,
+      "learning_rate": 9.276650845037141e-06,
+      "loss": 1.6757,
+      "step": 7460
+    },
+    {
+      "epoch": 0.21758734037292377,
+      "grad_norm": 15.6875,
+      "learning_rate": 9.274711563364527e-06,
+      "loss": 1.7122,
+      "step": 7480
+    },
+    {
+      "epoch": 0.21816912470547167,
+      "grad_norm": 15.0625,
+      "learning_rate": 9.272772281691912e-06,
+      "loss": 1.7075,
+      "step": 7500
+    },
+    {
+      "epoch": 0.2187509090380196,
+      "grad_norm": 12.25,
+      "learning_rate": 9.270833000019297e-06,
+      "loss": 1.7087,
+      "step": 7520
+    },
+    {
+      "epoch": 0.21933269337056754,
+      "grad_norm": 12.375,
+      "learning_rate": 9.268893718346682e-06,
+      "loss": 1.6896,
+      "step": 7540
+    },
+    {
+      "epoch": 0.21991447770311545,
+      "grad_norm": 13.75,
+      "learning_rate": 9.266954436674067e-06,
+      "loss": 1.7717,
+      "step": 7560
+    },
+    {
+      "epoch": 0.2204962620356634,
+      "grad_norm": 12.25,
+      "learning_rate": 9.265015155001452e-06,
+      "loss": 1.5941,
+      "step": 7580
+    },
+    {
+      "epoch": 0.2210780463682113,
+      "grad_norm": 12.75,
+      "learning_rate": 9.263075873328837e-06,
+      "loss": 1.6637,
+      "step": 7600
+    },
+    {
+      "epoch": 0.22165983070075923,
+      "grad_norm": 13.6875,
+      "learning_rate": 9.261136591656222e-06,
+      "loss": 1.6377,
+      "step": 7620
+    },
+    {
+      "epoch": 0.22224161503330717,
+      "grad_norm": 12.3125,
+      "learning_rate": 9.259197309983607e-06,
+      "loss": 1.6165,
+      "step": 7640
+    },
+    {
+      "epoch": 0.22282339936585507,
+      "grad_norm": 11.875,
+      "learning_rate": 9.257258028310992e-06,
+      "loss": 1.6967,
+      "step": 7660
+    },
+    {
+      "epoch": 0.223405183698403,
+      "grad_norm": 13.8125,
+      "learning_rate": 9.255318746638378e-06,
+      "loss": 1.6089,
+      "step": 7680
+    },
+    {
+      "epoch": 0.22398696803095092,
+      "grad_norm": 12.75,
+      "learning_rate": 9.253379464965763e-06,
+      "loss": 1.6017,
+      "step": 7700
+    },
+    {
+      "epoch": 0.22456875236349885,
+      "grad_norm": 15.4375,
+      "learning_rate": 9.251440183293148e-06,
+      "loss": 1.6056,
+      "step": 7720
+    },
+    {
+      "epoch": 0.2251505366960468,
+      "grad_norm": 11.9375,
+      "learning_rate": 9.249500901620533e-06,
+      "loss": 1.614,
+      "step": 7740
+    },
+    {
+      "epoch": 0.2257323210285947,
+      "grad_norm": 13.8125,
+      "learning_rate": 9.247561619947918e-06,
+      "loss": 1.6657,
+      "step": 7760
+    },
+    {
+      "epoch": 0.22631410536114263,
+      "grad_norm": 12.5625,
+      "learning_rate": 9.245622338275303e-06,
+      "loss": 1.699,
+      "step": 7780
+    },
+    {
+      "epoch": 0.22689588969369054,
+      "grad_norm": 15.5,
+      "learning_rate": 9.243683056602688e-06,
+      "loss": 1.6335,
+      "step": 7800
+    },
+    {
+      "epoch": 0.22747767402623847,
+      "grad_norm": 14.6875,
+      "learning_rate": 9.241743774930073e-06,
+      "loss": 1.7117,
+      "step": 7820
+    },
+    {
+      "epoch": 0.2280594583587864,
+      "grad_norm": 10.375,
+      "learning_rate": 9.239804493257457e-06,
+      "loss": 1.6218,
+      "step": 7840
+    },
+    {
+      "epoch": 0.22864124269133432,
+      "grad_norm": 12.5,
+      "learning_rate": 9.237865211584842e-06,
+      "loss": 1.6866,
+      "step": 7860
+    },
+    {
+      "epoch": 0.22922302702388225,
+      "grad_norm": 12.5,
+      "learning_rate": 9.235925929912227e-06,
+      "loss": 1.6505,
+      "step": 7880
+    },
+    {
+      "epoch": 0.22980481135643016,
+      "grad_norm": 12.5625,
+      "learning_rate": 9.233986648239612e-06,
+      "loss": 1.6149,
+      "step": 7900
+    },
+    {
+      "epoch": 0.2303865956889781,
+      "grad_norm": 11.3125,
+      "learning_rate": 9.232047366566997e-06,
+      "loss": 1.7142,
+      "step": 7920
+    },
+    {
+      "epoch": 0.23096838002152603,
+      "grad_norm": 13.875,
+      "learning_rate": 9.230108084894382e-06,
+      "loss": 1.6732,
+      "step": 7940
+    },
+    {
+      "epoch": 0.23155016435407394,
+      "grad_norm": 12.25,
+      "learning_rate": 9.228168803221767e-06,
+      "loss": 1.797,
+      "step": 7960
+    },
+    {
+      "epoch": 0.23213194868662188,
+      "grad_norm": 11.8125,
+      "learning_rate": 9.226229521549152e-06,
+      "loss": 1.779,
+      "step": 7980
+    },
+    {
+      "epoch": 0.23271373301916978,
+      "grad_norm": 14.875,
+      "learning_rate": 9.224290239876537e-06,
+      "loss": 1.6473,
+      "step": 8000
+    },
+    {
+      "epoch": 0.23329551735171772,
+      "grad_norm": 10.625,
+      "learning_rate": 9.222350958203923e-06,
+      "loss": 1.6575,
+      "step": 8020
+    },
+    {
+      "epoch": 0.23387730168426565,
+      "grad_norm": 12.8125,
+      "learning_rate": 9.220411676531308e-06,
+      "loss": 1.6458,
+      "step": 8040
+    },
+    {
+      "epoch": 0.23445908601681356,
+      "grad_norm": 13.5,
+      "learning_rate": 9.218472394858693e-06,
+      "loss": 1.6488,
+      "step": 8060
+    },
+    {
+      "epoch": 0.2350408703493615,
+      "grad_norm": 11.5625,
+      "learning_rate": 9.216533113186078e-06,
+      "loss": 1.681,
+      "step": 8080
+    },
+    {
+      "epoch": 0.2356226546819094,
+      "grad_norm": 12.125,
+      "learning_rate": 9.214593831513463e-06,
+      "loss": 1.6335,
+      "step": 8100
+    },
+    {
+      "epoch": 0.23620443901445734,
+      "grad_norm": 8.625,
+      "learning_rate": 9.212654549840848e-06,
+      "loss": 1.6555,
+      "step": 8120
+    },
+    {
+      "epoch": 0.23678622334700528,
+      "grad_norm": 11.5625,
+      "learning_rate": 9.210715268168233e-06,
+      "loss": 1.6875,
+      "step": 8140
+    },
+    {
+      "epoch": 0.23736800767955318,
+      "grad_norm": 13.125,
+      "learning_rate": 9.208775986495618e-06,
+      "loss": 1.6405,
+      "step": 8160
+    },
+    {
+      "epoch": 0.23794979201210112,
+      "grad_norm": 14.1875,
+      "learning_rate": 9.206836704823003e-06,
+      "loss": 1.629,
+      "step": 8180
+    },
+    {
+      "epoch": 0.23853157634464903,
+      "grad_norm": 13.0625,
+      "learning_rate": 9.204897423150388e-06,
+      "loss": 1.6058,
+      "step": 8200
+    },
+    {
+      "epoch": 0.23911336067719696,
+      "grad_norm": 15.6875,
+      "learning_rate": 9.202958141477773e-06,
+      "loss": 1.6622,
+      "step": 8220
+    },
+    {
+      "epoch": 0.2396951450097449,
+      "grad_norm": 13.5,
+      "learning_rate": 9.201018859805159e-06,
+      "loss": 1.6367,
+      "step": 8240
+    },
+    {
+      "epoch": 0.2402769293422928,
+      "grad_norm": 15.375,
+      "learning_rate": 9.199079578132544e-06,
+      "loss": 1.6016,
+      "step": 8260
+    },
+    {
+      "epoch": 0.24085871367484074,
+      "grad_norm": 13.8125,
+      "learning_rate": 9.197140296459929e-06,
+      "loss": 1.6196,
+      "step": 8280
+    },
+    {
+      "epoch": 0.24144049800738865,
+      "grad_norm": 13.0625,
+      "learning_rate": 9.195201014787314e-06,
+      "loss": 1.6289,
+      "step": 8300
+    },
+    {
+      "epoch": 0.24202228233993658,
+      "grad_norm": 12.0,
+      "learning_rate": 9.193261733114699e-06,
+      "loss": 1.6877,
+      "step": 8320
+    },
+    {
+      "epoch": 0.24260406667248452,
+      "grad_norm": 13.25,
+      "learning_rate": 9.191322451442084e-06,
+      "loss": 1.7558,
+      "step": 8340
+    },
+    {
+      "epoch": 0.24318585100503243,
+      "grad_norm": 12.0,
+      "learning_rate": 9.18938316976947e-06,
+      "loss": 1.7199,
+      "step": 8360
+    },
+    {
+      "epoch": 0.24376763533758036,
+      "grad_norm": 12.25,
+      "learning_rate": 9.187443888096854e-06,
+      "loss": 1.6592,
+      "step": 8380
+    },
+    {
+      "epoch": 0.24434941967012827,
+      "grad_norm": 10.25,
+      "learning_rate": 9.18550460642424e-06,
+      "loss": 1.6278,
+      "step": 8400
+    },
+    {
+      "epoch": 0.2449312040026762,
+      "grad_norm": 13.8125,
+      "learning_rate": 9.183565324751624e-06,
+      "loss": 1.6542,
+      "step": 8420
+    },
+    {
+      "epoch": 0.24551298833522414,
+      "grad_norm": 13.8125,
+      "learning_rate": 9.18162604307901e-06,
+      "loss": 1.6557,
+      "step": 8440
+    },
+    {
+      "epoch": 0.24609477266777205,
+      "grad_norm": 12.625,
+      "learning_rate": 9.179686761406395e-06,
+      "loss": 1.7821,
+      "step": 8460
+    },
+    {
+      "epoch": 0.24667655700031998,
+      "grad_norm": 13.5625,
+      "learning_rate": 9.17774747973378e-06,
+      "loss": 1.5931,
+      "step": 8480
+    },
+    {
+      "epoch": 0.2472583413328679,
+      "grad_norm": 12.0,
+      "learning_rate": 9.175808198061165e-06,
+      "loss": 1.6858,
+      "step": 8500
+    },
+    {
+      "epoch": 0.24784012566541583,
+      "grad_norm": 12.375,
+      "learning_rate": 9.17386891638855e-06,
+      "loss": 1.667,
+      "step": 8520
+    },
+    {
+      "epoch": 0.24842190999796376,
+      "grad_norm": 12.875,
+      "learning_rate": 9.171929634715935e-06,
+      "loss": 1.6255,
+      "step": 8540
+    },
+    {
+      "epoch": 0.24900369433051167,
+      "grad_norm": 16.375,
+      "learning_rate": 9.16999035304332e-06,
+      "loss": 1.62,
+      "step": 8560
+    },
+    {
+      "epoch": 0.2495854786630596,
+      "grad_norm": 13.3125,
+      "learning_rate": 9.168051071370705e-06,
+      "loss": 1.6483,
+      "step": 8580
+    },
+    {
+      "epoch": 0.25016726299560754,
+      "grad_norm": 14.0,
+      "learning_rate": 9.16611178969809e-06,
+      "loss": 1.7872,
+      "step": 8600
+    },
+    {
+      "epoch": 0.25074904732815545,
+      "grad_norm": 14.0,
+      "learning_rate": 9.164172508025475e-06,
+      "loss": 1.6429,
+      "step": 8620
+    },
+    {
+      "epoch": 0.25133083166070336,
+      "grad_norm": 13.0,
+      "learning_rate": 9.16223322635286e-06,
+      "loss": 1.6614,
+      "step": 8640
+    },
+    {
+      "epoch": 0.2519126159932513,
+      "grad_norm": 12.25,
+      "learning_rate": 9.160293944680246e-06,
+      "loss": 1.7051,
+      "step": 8660
+    },
+    {
+      "epoch": 0.25249440032579923,
+      "grad_norm": 12.1875,
+      "learning_rate": 9.15835466300763e-06,
+      "loss": 1.6515,
+      "step": 8680
+    },
+    {
+      "epoch": 0.25307618465834714,
+      "grad_norm": 11.25,
+      "learning_rate": 9.156415381335016e-06,
+      "loss": 1.6462,
+      "step": 8700
+    },
+    {
+      "epoch": 0.2536579689908951,
+      "grad_norm": 12.0,
+      "learning_rate": 9.154476099662401e-06,
+      "loss": 1.6229,
+      "step": 8720
+    },
+    {
+      "epoch": 0.254239753323443,
+      "grad_norm": 13.625,
+      "learning_rate": 9.152536817989786e-06,
+      "loss": 1.6951,
+      "step": 8740
+    },
+    {
+      "epoch": 0.2548215376559909,
+      "grad_norm": 11.3125,
+      "learning_rate": 9.150597536317171e-06,
+      "loss": 1.6354,
+      "step": 8760
+    },
+    {
+      "epoch": 0.2554033219885388,
+      "grad_norm": 12.9375,
+      "learning_rate": 9.148658254644556e-06,
+      "loss": 1.6501,
+      "step": 8780
+    },
+    {
+      "epoch": 0.2559851063210868,
+      "grad_norm": 10.5,
+      "learning_rate": 9.146718972971941e-06,
+      "loss": 1.7109,
+      "step": 8800
+    },
+    {
+      "epoch": 0.2565668906536347,
+      "grad_norm": 12.6875,
+      "learning_rate": 9.144779691299326e-06,
+      "loss": 1.6589,
+      "step": 8820
+    },
+    {
+      "epoch": 0.2571486749861826,
+      "grad_norm": 13.625,
+      "learning_rate": 9.142840409626712e-06,
+      "loss": 1.7272,
+      "step": 8840
+    },
+    {
+      "epoch": 0.25773045931873056,
+      "grad_norm": 14.6875,
+      "learning_rate": 9.140901127954097e-06,
+      "loss": 1.6892,
+      "step": 8860
+    },
+    {
+      "epoch": 0.2583122436512785,
+      "grad_norm": 13.1875,
+      "learning_rate": 9.138961846281482e-06,
+      "loss": 1.6675,
+      "step": 8880
+    },
+    {
+      "epoch": 0.2588940279838264,
+      "grad_norm": 11.25,
+      "learning_rate": 9.137022564608867e-06,
+      "loss": 1.5748,
+      "step": 8900
+    },
+    {
+      "epoch": 0.25947581231637434,
+      "grad_norm": 13.6875,
+      "learning_rate": 9.135083282936252e-06,
+      "loss": 1.6906,
+      "step": 8920
+    },
+    {
+      "epoch": 0.26005759664892225,
+      "grad_norm": 13.4375,
+      "learning_rate": 9.133144001263637e-06,
+      "loss": 1.7507,
+      "step": 8940
+    },
+    {
+      "epoch": 0.26063938098147016,
+      "grad_norm": 13.25,
+      "learning_rate": 9.131204719591022e-06,
+      "loss": 1.7374,
+      "step": 8960
+    },
+    {
+      "epoch": 0.26122116531401807,
+      "grad_norm": 13.125,
+      "learning_rate": 9.129265437918407e-06,
+      "loss": 1.6141,
+      "step": 8980
+    },
+    {
+      "epoch": 0.26180294964656603,
+      "grad_norm": 11.375,
+      "learning_rate": 9.127326156245792e-06,
+      "loss": 1.6063,
+      "step": 9000
+    },
+    {
+      "epoch": 0.26238473397911394,
+      "grad_norm": 13.375,
+      "learning_rate": 9.125386874573177e-06,
+      "loss": 1.7222,
+      "step": 9020
+    },
+    {
+      "epoch": 0.26296651831166185,
+      "grad_norm": 14.0625,
+      "learning_rate": 9.123447592900562e-06,
+      "loss": 1.7228,
+      "step": 9040
+    },
+    {
+      "epoch": 0.2635483026442098,
+      "grad_norm": 12.5,
+      "learning_rate": 9.121508311227948e-06,
+      "loss": 1.6867,
+      "step": 9060
+    },
+    {
+      "epoch": 0.2641300869767577,
+      "grad_norm": 11.25,
+      "learning_rate": 9.119569029555333e-06,
+      "loss": 1.7106,
+      "step": 9080
+    },
+    {
+      "epoch": 0.2647118713093056,
+      "grad_norm": 13.0625,
+      "learning_rate": 9.117629747882716e-06,
+      "loss": 1.6625,
+      "step": 9100
+    },
+    {
+      "epoch": 0.2652936556418536,
+      "grad_norm": 12.625,
+      "learning_rate": 9.115690466210101e-06,
+      "loss": 1.6357,
+      "step": 9120
+    },
+    {
+      "epoch": 0.2658754399744015,
+      "grad_norm": 14.3125,
+      "learning_rate": 9.113751184537486e-06,
+      "loss": 1.6298,
+      "step": 9140
+    },
+    {
+      "epoch": 0.2664572243069494,
+      "grad_norm": 12.0,
+      "learning_rate": 9.111811902864871e-06,
+      "loss": 1.7246,
+      "step": 9160
+    },
+    {
+      "epoch": 0.2670390086394973,
+      "grad_norm": 10.625,
+      "learning_rate": 9.109872621192256e-06,
+      "loss": 1.658,
+      "step": 9180
+    },
+    {
+      "epoch": 0.2676207929720453,
+      "grad_norm": 12.875,
+      "learning_rate": 9.107933339519642e-06,
+      "loss": 1.6617,
+      "step": 9200
+    },
+    {
+      "epoch": 0.2682025773045932,
+      "grad_norm": 12.125,
+      "learning_rate": 9.105994057847027e-06,
+      "loss": 1.7339,
+      "step": 9220
+    },
+    {
+      "epoch": 0.2687843616371411,
+      "grad_norm": 15.3125,
+      "learning_rate": 9.104054776174412e-06,
+      "loss": 1.7189,
+      "step": 9240
+    },
+    {
+      "epoch": 0.26936614596968905,
+      "grad_norm": 12.6875,
+      "learning_rate": 9.102115494501797e-06,
+      "loss": 1.6779,
+      "step": 9260
+    },
+    {
+      "epoch": 0.26994793030223696,
+      "grad_norm": 10.75,
+      "learning_rate": 9.100176212829182e-06,
+      "loss": 1.6748,
+      "step": 9280
+    },
+    {
+      "epoch": 0.27052971463478487,
+      "grad_norm": 13.875,
+      "learning_rate": 9.098236931156567e-06,
+      "loss": 1.6947,
+      "step": 9300
+    },
+    {
+      "epoch": 0.27111149896733283,
+      "grad_norm": 13.8125,
+      "learning_rate": 9.096297649483952e-06,
+      "loss": 1.6742,
+      "step": 9320
+    },
+    {
+      "epoch": 0.27169328329988074,
+      "grad_norm": 12.625,
+      "learning_rate": 9.094358367811337e-06,
+      "loss": 1.7339,
+      "step": 9340
+    },
+    {
+      "epoch": 0.27227506763242865,
+      "grad_norm": 14.1875,
+      "learning_rate": 9.092419086138722e-06,
+      "loss": 1.6263,
+      "step": 9360
+    },
+    {
+      "epoch": 0.2728568519649766,
+      "grad_norm": 11.75,
+      "learning_rate": 9.090479804466107e-06,
+      "loss": 1.6642,
+      "step": 9380
+    },
+    {
+      "epoch": 0.2734386362975245,
+      "grad_norm": 11.6875,
+      "learning_rate": 9.088540522793493e-06,
+      "loss": 1.684,
+      "step": 9400
+    },
+    {
+      "epoch": 0.2740204206300724,
+      "grad_norm": 13.125,
+      "learning_rate": 9.086601241120878e-06,
+      "loss": 1.6405,
+      "step": 9420
+    },
+    {
+      "epoch": 0.27460220496262033,
+      "grad_norm": 19.375,
+      "learning_rate": 9.084661959448263e-06,
+      "loss": 1.712,
+      "step": 9440
+    },
+    {
+      "epoch": 0.2751839892951683,
+      "grad_norm": 9.6875,
+      "learning_rate": 9.082722677775648e-06,
+      "loss": 1.7061,
+      "step": 9460
+    },
+    {
+      "epoch": 0.2757657736277162,
+      "grad_norm": 11.875,
+      "learning_rate": 9.080783396103033e-06,
+      "loss": 1.7192,
+      "step": 9480
+    },
+    {
+      "epoch": 0.2763475579602641,
+      "grad_norm": 11.875,
+      "learning_rate": 9.078844114430418e-06,
+      "loss": 1.6492,
+      "step": 9500
+    },
+    {
+      "epoch": 0.2769293422928121,
+      "grad_norm": 11.6875,
+      "learning_rate": 9.076904832757803e-06,
+      "loss": 1.714,
+      "step": 9520
+    },
+    {
+      "epoch": 0.27751112662536,
+      "grad_norm": 12.5,
+      "learning_rate": 9.074965551085188e-06,
+      "loss": 1.6707,
+      "step": 9540
+    },
+    {
+      "epoch": 0.2780929109579079,
+      "grad_norm": 12.8125,
+      "learning_rate": 9.073026269412573e-06,
+      "loss": 1.748,
+      "step": 9560
+    },
+    {
+      "epoch": 0.27867469529045585,
+      "grad_norm": 13.6875,
+      "learning_rate": 9.071086987739958e-06,
+      "loss": 1.6314,
+      "step": 9580
+    },
+    {
+      "epoch": 0.27925647962300376,
+      "grad_norm": 14.6875,
+      "learning_rate": 9.069147706067344e-06,
+      "loss": 1.7624,
+      "step": 9600
+    },
+    {
+      "epoch": 0.27983826395555167,
+      "grad_norm": 14.125,
+      "learning_rate": 9.067208424394729e-06,
+      "loss": 1.5784,
+      "step": 9620
+    },
+    {
+      "epoch": 0.2804200482880996,
+      "grad_norm": 13.75,
+      "learning_rate": 9.065269142722114e-06,
+      "loss": 1.6444,
+      "step": 9640
+    },
+    {
+      "epoch": 0.28100183262064754,
+      "grad_norm": 13.75,
+      "learning_rate": 9.063329861049499e-06,
+      "loss": 1.6215,
+      "step": 9660
+    },
+    {
+      "epoch": 0.28158361695319545,
+      "grad_norm": 12.3125,
+      "learning_rate": 9.061390579376884e-06,
+      "loss": 1.7362,
+      "step": 9680
+    },
+    {
+      "epoch": 0.28216540128574336,
+      "grad_norm": 11.125,
+      "learning_rate": 9.059451297704269e-06,
+      "loss": 1.6616,
+      "step": 9700
+    },
+    {
+      "epoch": 0.2827471856182913,
+      "grad_norm": 15.5,
+      "learning_rate": 9.057512016031654e-06,
+      "loss": 1.5841,
+      "step": 9720
+    },
+    {
+      "epoch": 0.2833289699508392,
+      "grad_norm": 13.0625,
+      "learning_rate": 9.05557273435904e-06,
+      "loss": 1.5656,
+      "step": 9740
+    },
+    {
+      "epoch": 0.28391075428338713,
+      "grad_norm": 14.375,
+      "learning_rate": 9.053633452686424e-06,
+      "loss": 1.6562,
+      "step": 9760
+    },
+    {
+      "epoch": 0.2844925386159351,
+      "grad_norm": 11.5625,
+      "learning_rate": 9.05169417101381e-06,
+      "loss": 1.7046,
+      "step": 9780
+    },
+    {
+      "epoch": 0.285074322948483,
+      "grad_norm": 9.875,
+      "learning_rate": 9.049754889341195e-06,
+      "loss": 1.6812,
+      "step": 9800
+    },
+    {
+      "epoch": 0.2856561072810309,
+      "grad_norm": 12.6875,
+      "learning_rate": 9.04781560766858e-06,
+      "loss": 1.6193,
+      "step": 9820
+    },
+    {
+      "epoch": 0.2862378916135788,
+      "grad_norm": 10.875,
+      "learning_rate": 9.045876325995965e-06,
+      "loss": 1.5845,
+      "step": 9840
+    },
+    {
+      "epoch": 0.2868196759461268,
+      "grad_norm": 10.625,
+      "learning_rate": 9.04393704432335e-06,
+      "loss": 1.7287,
+      "step": 9860
+    },
+    {
+      "epoch": 0.2874014602786747,
+      "grad_norm": 12.9375,
+      "learning_rate": 9.041997762650735e-06,
+      "loss": 1.6522,
+      "step": 9880
+    },
+    {
+      "epoch": 0.2879832446112226,
+      "grad_norm": 13.25,
+      "learning_rate": 9.04005848097812e-06,
+      "loss": 1.638,
+      "step": 9900
+    },
+    {
+      "epoch": 0.28856502894377056,
+      "grad_norm": 13.4375,
+      "learning_rate": 9.038119199305505e-06,
+      "loss": 1.726,
+      "step": 9920
+    },
+    {
+      "epoch": 0.28914681327631847,
+      "grad_norm": 14.4375,
+      "learning_rate": 9.03617991763289e-06,
+      "loss": 1.7284,
+      "step": 9940
+    },
+    {
+      "epoch": 0.2897285976088664,
+      "grad_norm": 14.3125,
+      "learning_rate": 9.034240635960275e-06,
+      "loss": 1.6854,
+      "step": 9960
+    },
+    {
+      "epoch": 0.29031038194141434,
+      "grad_norm": 15.125,
+      "learning_rate": 9.03230135428766e-06,
+      "loss": 1.7404,
+      "step": 9980
+    },
+    {
+      "epoch": 0.29089216627396225,
+      "grad_norm": 12.125,
+      "learning_rate": 9.030362072615045e-06,
+      "loss": 1.6527,
+      "step": 10000
+    },
+    {
+      "epoch": 0.29147395060651016,
+      "grad_norm": 13.0625,
+      "learning_rate": 9.02842279094243e-06,
+      "loss": 1.708,
+      "step": 10020
+    },
+    {
+      "epoch": 0.29205573493905806,
+      "grad_norm": 16.125,
+      "learning_rate": 9.026483509269816e-06,
+      "loss": 1.6867,
+      "step": 10040
+    },
+    {
+      "epoch": 0.29263751927160603,
+      "grad_norm": 13.5,
+      "learning_rate": 9.0245442275972e-06,
+      "loss": 1.6858,
+      "step": 10060
+    },
+    {
+      "epoch": 0.29321930360415394,
+      "grad_norm": 16.375,
+      "learning_rate": 9.022604945924586e-06,
+      "loss": 1.6886,
+      "step": 10080
+    },
+    {
+      "epoch": 0.29380108793670184,
+      "grad_norm": 13.3125,
+      "learning_rate": 9.020665664251971e-06,
+      "loss": 1.6543,
+      "step": 10100
+    },
+    {
+      "epoch": 0.2943828722692498,
+      "grad_norm": 13.25,
+      "learning_rate": 9.018726382579356e-06,
+      "loss": 1.6289,
+      "step": 10120
+    },
+    {
+      "epoch": 0.2949646566017977,
+      "grad_norm": 10.75,
+      "learning_rate": 9.016787100906741e-06,
+      "loss": 1.6614,
+      "step": 10140
+    },
+    {
+      "epoch": 0.2955464409343456,
+      "grad_norm": 13.3125,
+      "learning_rate": 9.014847819234126e-06,
+      "loss": 1.6518,
+      "step": 10160
+    },
+    {
+      "epoch": 0.2961282252668936,
+      "grad_norm": 12.75,
+      "learning_rate": 9.012908537561511e-06,
+      "loss": 1.6559,
+      "step": 10180
+    },
+    {
+      "epoch": 0.2967100095994415,
+      "grad_norm": 14.75,
+      "learning_rate": 9.010969255888896e-06,
+      "loss": 1.6365,
+      "step": 10200
+    },
+    {
+      "epoch": 0.2972917939319894,
+      "grad_norm": 12.5,
+      "learning_rate": 9.009029974216282e-06,
+      "loss": 1.7192,
+      "step": 10220
+    },
+    {
+      "epoch": 0.2978735782645373,
+      "grad_norm": 13.25,
+      "learning_rate": 9.007090692543667e-06,
+      "loss": 1.6718,
+      "step": 10240
+    },
+    {
+      "epoch": 0.29845536259708527,
+      "grad_norm": 10.5625,
+      "learning_rate": 9.005151410871052e-06,
+      "loss": 1.6284,
+      "step": 10260
+    },
+    {
+      "epoch": 0.2990371469296332,
+      "grad_norm": 13.3125,
+      "learning_rate": 9.003212129198437e-06,
+      "loss": 1.6932,
+      "step": 10280
+    },
+    {
+      "epoch": 0.2996189312621811,
+      "grad_norm": 11.0625,
+      "learning_rate": 9.001272847525822e-06,
+      "loss": 1.6853,
+      "step": 10300
+    },
+    {
+      "epoch": 0.30020071559472905,
+      "grad_norm": 12.25,
+      "learning_rate": 8.999333565853207e-06,
+      "loss": 1.6984,
+      "step": 10320
+    },
+    {
+      "epoch": 0.30078249992727696,
+      "grad_norm": 10.75,
+      "learning_rate": 8.997394284180592e-06,
+      "loss": 1.6524,
+      "step": 10340
+    },
+    {
+      "epoch": 0.30136428425982487,
+      "grad_norm": 12.75,
+      "learning_rate": 8.995455002507977e-06,
+      "loss": 1.6859,
+      "step": 10360
+    },
+    {
+      "epoch": 0.30194606859237283,
+      "grad_norm": 13.5,
+      "learning_rate": 8.993515720835362e-06,
+      "loss": 1.5609,
+      "step": 10380
+    },
+    {
+      "epoch": 0.30252785292492074,
+      "grad_norm": 11.8125,
+      "learning_rate": 8.991576439162747e-06,
+      "loss": 1.6768,
+      "step": 10400
+    },
+    {
+      "epoch": 0.30310963725746864,
+      "grad_norm": 12.5625,
+      "learning_rate": 8.989637157490133e-06,
+      "loss": 1.6377,
+      "step": 10420
+    },
+    {
+      "epoch": 0.3036914215900166,
+      "grad_norm": 11.0625,
+      "learning_rate": 8.987697875817518e-06,
+      "loss": 1.7098,
+      "step": 10440
+    },
+    {
+      "epoch": 0.3042732059225645,
+      "grad_norm": 11.6875,
+      "learning_rate": 8.985758594144903e-06,
+      "loss": 1.6467,
+      "step": 10460
+    },
+    {
+      "epoch": 0.3048549902551124,
+      "grad_norm": 15.5,
+      "learning_rate": 8.983819312472288e-06,
+      "loss": 1.6654,
+      "step": 10480
+    },
+    {
+      "epoch": 0.30543677458766033,
+      "grad_norm": 12.625,
+      "learning_rate": 8.981880030799673e-06,
+      "loss": 1.7498,
+      "step": 10500
+    },
+    {
+      "epoch": 0.3060185589202083,
+      "grad_norm": 15.75,
+      "learning_rate": 8.979940749127058e-06,
+      "loss": 1.6435,
+      "step": 10520
+    },
+    {
+      "epoch": 0.3066003432527562,
+      "grad_norm": 12.0625,
+      "learning_rate": 8.978001467454443e-06,
+      "loss": 1.6964,
+      "step": 10540
+    },
+    {
+      "epoch": 0.3071821275853041,
+      "grad_norm": 13.875,
+      "learning_rate": 8.976062185781828e-06,
+      "loss": 1.6594,
+      "step": 10560
+    },
+    {
+      "epoch": 0.3077639119178521,
+      "grad_norm": 13.1875,
+      "learning_rate": 8.974122904109213e-06,
+      "loss": 1.6804,
+      "step": 10580
+    },
+    {
+      "epoch": 0.3083456962504,
+      "grad_norm": 10.25,
+      "learning_rate": 8.972183622436598e-06,
+      "loss": 1.735,
+      "step": 10600
+    },
+    {
+      "epoch": 0.3089274805829479,
+      "grad_norm": 13.5625,
+      "learning_rate": 8.970244340763983e-06,
+      "loss": 1.6542,
+      "step": 10620
+    },
+    {
+      "epoch": 0.30950926491549585,
+      "grad_norm": 9.75,
+      "learning_rate": 8.968305059091369e-06,
+      "loss": 1.5944,
+      "step": 10640
+    },
+    {
+      "epoch": 0.31009104924804376,
+      "grad_norm": 14.125,
+      "learning_rate": 8.966365777418754e-06,
+      "loss": 1.653,
+      "step": 10660
+    },
+    {
+      "epoch": 0.31067283358059167,
+      "grad_norm": 13.125,
+      "learning_rate": 8.964426495746139e-06,
+      "loss": 1.6451,
+      "step": 10680
+    },
+    {
+      "epoch": 0.3112546179131396,
+      "grad_norm": 16.125,
+      "learning_rate": 8.962487214073524e-06,
+      "loss": 1.7036,
+      "step": 10700
+    },
+    {
+      "epoch": 0.31183640224568754,
+      "grad_norm": 11.875,
+      "learning_rate": 8.960547932400909e-06,
+      "loss": 1.6025,
+      "step": 10720
+    },
+    {
+      "epoch": 0.31241818657823545,
+      "grad_norm": 13.5,
+      "learning_rate": 8.958608650728292e-06,
+      "loss": 1.6561,
+      "step": 10740
+    },
+    {
+      "epoch": 0.31299997091078335,
+      "grad_norm": 11.9375,
+      "learning_rate": 8.956669369055677e-06,
+      "loss": 1.5292,
+      "step": 10760
+    },
+    {
+      "epoch": 0.3135817552433313,
+      "grad_norm": 10.5625,
+      "learning_rate": 8.954730087383063e-06,
+      "loss": 1.6855,
+      "step": 10780
+    },
+    {
+      "epoch": 0.3141635395758792,
+      "grad_norm": 11.4375,
+      "learning_rate": 8.952790805710448e-06,
+      "loss": 1.6713,
+      "step": 10800
+    },
+    {
+      "epoch": 0.31474532390842713,
+      "grad_norm": 16.625,
+      "learning_rate": 8.950851524037833e-06,
+      "loss": 1.7154,
+      "step": 10820
+    },
+    {
+      "epoch": 0.3153271082409751,
+      "grad_norm": 13.875,
+      "learning_rate": 8.948912242365218e-06,
+      "loss": 1.6844,
+      "step": 10840
+    },
+    {
+      "epoch": 0.315908892573523,
+      "grad_norm": 12.1875,
+      "learning_rate": 8.946972960692603e-06,
+      "loss": 1.6695,
+      "step": 10860
+    },
+    {
+      "epoch": 0.3164906769060709,
+      "grad_norm": 12.1875,
+      "learning_rate": 8.945033679019988e-06,
+      "loss": 1.7592,
+      "step": 10880
+    },
+    {
+      "epoch": 0.3170724612386188,
+      "grad_norm": 14.5625,
+      "learning_rate": 8.943094397347373e-06,
+      "loss": 1.5866,
+      "step": 10900
+    },
+    {
+      "epoch": 0.3176542455711668,
+      "grad_norm": 14.0,
+      "learning_rate": 8.941155115674758e-06,
+      "loss": 1.7119,
+      "step": 10920
+    },
+    {
+      "epoch": 0.3182360299037147,
+      "grad_norm": 12.375,
+      "learning_rate": 8.939215834002143e-06,
+      "loss": 1.6872,
+      "step": 10940
+    },
+    {
+      "epoch": 0.3188178142362626,
+      "grad_norm": 13.875,
+      "learning_rate": 8.937276552329528e-06,
+      "loss": 1.736,
+      "step": 10960
+    },
+    {
+      "epoch": 0.31939959856881056,
+      "grad_norm": 12.9375,
+      "learning_rate": 8.935337270656914e-06,
+      "loss": 1.7081,
+      "step": 10980
+    },
+    {
+      "epoch": 0.31998138290135847,
+      "grad_norm": 12.6875,
+      "learning_rate": 8.933397988984299e-06,
+      "loss": 1.7345,
+      "step": 11000
+    },
+    {
+      "epoch": 0.3205631672339064,
+      "grad_norm": 12.4375,
+      "learning_rate": 8.931458707311684e-06,
+      "loss": 1.7188,
+      "step": 11020
+    },
+    {
+      "epoch": 0.32114495156645434,
+      "grad_norm": 12.0,
+      "learning_rate": 8.929519425639069e-06,
+      "loss": 1.6173,
+      "step": 11040
+    },
+    {
+      "epoch": 0.32172673589900225,
+      "grad_norm": 15.75,
+      "learning_rate": 8.927580143966454e-06,
+      "loss": 1.6647,
+      "step": 11060
+    },
+    {
+      "epoch": 0.32230852023155016,
+      "grad_norm": 11.6875,
+      "learning_rate": 8.925640862293839e-06,
+      "loss": 1.6596,
+      "step": 11080
+    },
+    {
+      "epoch": 0.32289030456409806,
+      "grad_norm": 10.875,
+      "learning_rate": 8.923701580621224e-06,
+      "loss": 1.6738,
+      "step": 11100
+    },
+    {
+      "epoch": 0.323472088896646,
+      "grad_norm": 12.6875,
+      "learning_rate": 8.92176229894861e-06,
+      "loss": 1.7109,
+      "step": 11120
+    },
+    {
+      "epoch": 0.32405387322919393,
+      "grad_norm": 13.625,
+      "learning_rate": 8.919823017275994e-06,
+      "loss": 1.6099,
+      "step": 11140
+    },
+    {
+      "epoch": 0.32463565756174184,
+      "grad_norm": 13.125,
+      "learning_rate": 8.91788373560338e-06,
+      "loss": 1.5996,
+      "step": 11160
+    },
+    {
+      "epoch": 0.3252174418942898,
+      "grad_norm": 11.875,
+      "learning_rate": 8.915944453930765e-06,
+      "loss": 1.6787,
+      "step": 11180
+    },
+    {
+      "epoch": 0.3257992262268377,
+      "grad_norm": 11.9375,
+      "learning_rate": 8.91400517225815e-06,
+      "loss": 1.5997,
+      "step": 11200
+    },
+    {
+      "epoch": 0.3263810105593856,
+      "grad_norm": 12.5,
+      "learning_rate": 8.912065890585535e-06,
+      "loss": 1.7873,
+      "step": 11220
+    },
+    {
+      "epoch": 0.3269627948919336,
+      "grad_norm": 11.5,
+      "learning_rate": 8.91012660891292e-06,
+      "loss": 1.7129,
+      "step": 11240
+    },
+    {
+      "epoch": 0.3275445792244815,
+      "grad_norm": 15.4375,
+      "learning_rate": 8.908187327240305e-06,
+      "loss": 1.6616,
+      "step": 11260
+    },
+    {
+      "epoch": 0.3281263635570294,
+      "grad_norm": 13.0625,
+      "learning_rate": 8.90624804556769e-06,
+      "loss": 1.6324,
+      "step": 11280
+    },
+    {
+      "epoch": 0.3287081478895773,
+      "grad_norm": 11.875,
+      "learning_rate": 8.904308763895075e-06,
+      "loss": 1.62,
+      "step": 11300
+    },
+    {
+      "epoch": 0.32928993222212527,
+      "grad_norm": 21.75,
+      "learning_rate": 8.90236948222246e-06,
+      "loss": 1.7055,
+      "step": 11320
+    },
+    {
+      "epoch": 0.3298717165546732,
+      "grad_norm": 11.75,
+      "learning_rate": 8.900430200549845e-06,
+      "loss": 1.6479,
+      "step": 11340
+    },
+    {
+      "epoch": 0.3304535008872211,
+      "grad_norm": 11.875,
+      "learning_rate": 8.89849091887723e-06,
+      "loss": 1.7365,
+      "step": 11360
+    },
+    {
+      "epoch": 0.33103528521976905,
+      "grad_norm": 12.375,
+      "learning_rate": 8.896551637204616e-06,
+      "loss": 1.7232,
+      "step": 11380
+    },
+    {
+      "epoch": 0.33161706955231696,
+      "grad_norm": 9.5,
+      "learning_rate": 8.894612355532e-06,
+      "loss": 1.6588,
+      "step": 11400
+    },
+    {
+      "epoch": 0.33219885388486486,
+      "grad_norm": 13.0625,
+      "learning_rate": 8.892673073859386e-06,
+      "loss": 1.6273,
+      "step": 11420
+    },
+    {
+      "epoch": 0.3327806382174128,
+      "grad_norm": 12.0,
+      "learning_rate": 8.89073379218677e-06,
+      "loss": 1.6455,
+      "step": 11440
+    },
+    {
+      "epoch": 0.33336242254996074,
+      "grad_norm": 10.6875,
+      "learning_rate": 8.888794510514156e-06,
+      "loss": 1.8009,
+      "step": 11460
+    },
+    {
+      "epoch": 0.33394420688250864,
+      "grad_norm": 12.125,
+      "learning_rate": 8.886855228841541e-06,
+      "loss": 1.7571,
+      "step": 11480
+    },
+    {
+      "epoch": 0.3345259912150566,
+      "grad_norm": 11.0,
+      "learning_rate": 8.884915947168926e-06,
+      "loss": 1.7168,
+      "step": 11500
+    },
+    {
+      "epoch": 0.3351077755476045,
+      "grad_norm": 16.5,
+      "learning_rate": 8.882976665496311e-06,
+      "loss": 1.607,
+      "step": 11520
+    },
+    {
+      "epoch": 0.3356895598801524,
+      "grad_norm": 12.75,
+      "learning_rate": 8.881037383823696e-06,
+      "loss": 1.65,
+      "step": 11540
+    },
+    {
+      "epoch": 0.33627134421270033,
+      "grad_norm": 13.75,
+      "learning_rate": 8.879098102151081e-06,
+      "loss": 1.5926,
+      "step": 11560
+    },
+    {
+      "epoch": 0.3368531285452483,
+      "grad_norm": 13.625,
+      "learning_rate": 8.877158820478466e-06,
+      "loss": 1.6459,
+      "step": 11580
+    },
+    {
+      "epoch": 0.3374349128777962,
+      "grad_norm": 12.4375,
+      "learning_rate": 8.875219538805852e-06,
+      "loss": 1.5444,
+      "step": 11600
+    },
+    {
+      "epoch": 0.3380166972103441,
+      "grad_norm": 11.75,
+      "learning_rate": 8.873280257133237e-06,
+      "loss": 1.6608,
+      "step": 11620
+    },
+    {
+      "epoch": 0.33859848154289207,
+      "grad_norm": 13.5625,
+      "learning_rate": 8.871340975460622e-06,
+      "loss": 1.642,
+      "step": 11640
+    },
+    {
+      "epoch": 0.33918026587544,
+      "grad_norm": 13.1875,
+      "learning_rate": 8.869401693788007e-06,
+      "loss": 1.7346,
+      "step": 11660
+    },
+    {
+      "epoch": 0.3397620502079879,
+      "grad_norm": 12.6875,
+      "learning_rate": 8.867462412115392e-06,
+      "loss": 1.5846,
+      "step": 11680
+    },
+    {
+      "epoch": 0.34034383454053585,
+      "grad_norm": 13.1875,
+      "learning_rate": 8.865523130442777e-06,
+      "loss": 1.5694,
+      "step": 11700
+    },
+    {
+      "epoch": 0.34092561887308376,
+      "grad_norm": 13.875,
+      "learning_rate": 8.863583848770162e-06,
+      "loss": 1.5998,
+      "step": 11720
+    },
+    {
+      "epoch": 0.34150740320563167,
+      "grad_norm": 13.25,
+      "learning_rate": 8.861644567097547e-06,
+      "loss": 1.7103,
+      "step": 11740
+    },
+    {
+      "epoch": 0.3420891875381796,
+      "grad_norm": 12.125,
+      "learning_rate": 8.859705285424932e-06,
+      "loss": 1.616,
+      "step": 11760
+    },
+    {
+      "epoch": 0.34267097187072754,
+      "grad_norm": 12.3125,
+      "learning_rate": 8.857766003752317e-06,
+      "loss": 1.7097,
+      "step": 11780
+    },
+    {
+      "epoch": 0.34325275620327544,
+      "grad_norm": 15.3125,
+      "learning_rate": 8.855826722079703e-06,
+      "loss": 1.733,
+      "step": 11800
+    },
+    {
+      "epoch": 0.34383454053582335,
+      "grad_norm": 15.25,
+      "learning_rate": 8.853887440407088e-06,
+      "loss": 1.5843,
+      "step": 11820
+    },
+    {
+      "epoch": 0.3444163248683713,
+      "grad_norm": 11.9375,
+      "learning_rate": 8.851948158734473e-06,
+      "loss": 1.685,
+      "step": 11840
+    },
+    {
+      "epoch": 0.3449981092009192,
+      "grad_norm": 11.9375,
+      "learning_rate": 8.850008877061858e-06,
+      "loss": 1.5462,
+      "step": 11860
+    },
+    {
+      "epoch": 0.34557989353346713,
+      "grad_norm": 12.5,
+      "learning_rate": 8.848069595389243e-06,
+      "loss": 1.7058,
+      "step": 11880
+    },
+    {
+      "epoch": 0.3461616778660151,
+      "grad_norm": 10.8125,
+      "learning_rate": 8.846130313716628e-06,
+      "loss": 1.6387,
+      "step": 11900
+    },
+    {
+      "epoch": 0.346743462198563,
+      "grad_norm": 12.8125,
+      "learning_rate": 8.844191032044013e-06,
+      "loss": 1.6074,
+      "step": 11920
+    },
+    {
+      "epoch": 0.3473252465311109,
+      "grad_norm": 13.3125,
+      "learning_rate": 8.842251750371398e-06,
+      "loss": 1.6532,
+      "step": 11940
+    },
+    {
+      "epoch": 0.3479070308636588,
+      "grad_norm": 11.5625,
+      "learning_rate": 8.840312468698783e-06,
+      "loss": 1.6338,
+      "step": 11960
+    },
+    {
+      "epoch": 0.3484888151962068,
+      "grad_norm": 14.3125,
+      "learning_rate": 8.838373187026168e-06,
+      "loss": 1.7099,
+      "step": 11980
+    },
+    {
+      "epoch": 0.3490705995287547,
+      "grad_norm": 12.1875,
+      "learning_rate": 8.836433905353554e-06,
+      "loss": 1.6162,
+      "step": 12000
+    },
+    {
+      "epoch": 0.3496523838613026,
+      "grad_norm": 10.75,
+      "learning_rate": 8.834494623680939e-06,
+      "loss": 1.6681,
+      "step": 12020
+    },
+    {
+      "epoch": 0.35023416819385056,
+      "grad_norm": 9.8125,
+      "learning_rate": 8.832555342008324e-06,
+      "loss": 1.6138,
+      "step": 12040
+    },
+    {
+      "epoch": 0.35081595252639847,
+      "grad_norm": 14.9375,
+      "learning_rate": 8.830616060335709e-06,
+      "loss": 1.5801,
+      "step": 12060
+    },
+    {
+      "epoch": 0.3513977368589464,
+      "grad_norm": 13.6875,
+      "learning_rate": 8.828676778663092e-06,
+      "loss": 1.6548,
+      "step": 12080
+    },
+    {
+      "epoch": 0.35197952119149434,
+      "grad_norm": 12.3125,
+      "learning_rate": 8.826737496990477e-06,
+      "loss": 1.6574,
+      "step": 12100
+    },
+    {
+      "epoch": 0.35256130552404225,
+      "grad_norm": 12.125,
+      "learning_rate": 8.824798215317862e-06,
+      "loss": 1.6967,
+      "step": 12120
+    },
+    {
+      "epoch": 0.35314308985659015,
+      "grad_norm": 16.125,
+      "learning_rate": 8.822858933645248e-06,
+      "loss": 1.6683,
+      "step": 12140
+    },
+    {
+      "epoch": 0.35372487418913806,
+      "grad_norm": 12.0625,
+      "learning_rate": 8.820919651972633e-06,
+      "loss": 1.7342,
+      "step": 12160
+    },
+    {
+      "epoch": 0.354306658521686,
+      "grad_norm": 13.0625,
+      "learning_rate": 8.818980370300018e-06,
+      "loss": 1.7129,
+      "step": 12180
+    },
+    {
+      "epoch": 0.35488844285423393,
+      "grad_norm": 13.0,
+      "learning_rate": 8.817041088627403e-06,
+      "loss": 1.6718,
+      "step": 12200
+    },
+    {
+      "epoch": 0.35547022718678184,
+      "grad_norm": 15.125,
+      "learning_rate": 8.815101806954788e-06,
+      "loss": 1.6067,
+      "step": 12220
+    },
+    {
+      "epoch": 0.3560520115193298,
+      "grad_norm": 10.9375,
+      "learning_rate": 8.813162525282173e-06,
+      "loss": 1.6761,
+      "step": 12240
+    },
+    {
+      "epoch": 0.3566337958518777,
+      "grad_norm": 11.25,
+      "learning_rate": 8.811223243609558e-06,
+      "loss": 1.7204,
+      "step": 12260
+    },
+    {
+      "epoch": 0.3572155801844256,
+      "grad_norm": 14.4375,
+      "learning_rate": 8.809283961936943e-06,
+      "loss": 1.6675,
+      "step": 12280
+    },
+    {
+      "epoch": 0.3577973645169736,
+      "grad_norm": 12.0,
+      "learning_rate": 8.807344680264328e-06,
+      "loss": 1.6827,
+      "step": 12300
+    },
+    {
+      "epoch": 0.3583791488495215,
+      "grad_norm": 12.625,
+      "learning_rate": 8.805405398591713e-06,
+      "loss": 1.7405,
+      "step": 12320
+    },
+    {
+      "epoch": 0.3589609331820694,
+      "grad_norm": 14.5625,
+      "learning_rate": 8.803466116919098e-06,
+      "loss": 1.6716,
+      "step": 12340
+    },
+    {
+      "epoch": 0.3595427175146173,
+      "grad_norm": 12.4375,
+      "learning_rate": 8.801526835246484e-06,
+      "loss": 1.6382,
+      "step": 12360
+    },
+    {
+      "epoch": 0.36012450184716527,
+      "grad_norm": 11.375,
+      "learning_rate": 8.799587553573869e-06,
+      "loss": 1.7131,
+      "step": 12380
+    },
+    {
+      "epoch": 0.3607062861797132,
+      "grad_norm": 12.875,
+      "learning_rate": 8.797648271901254e-06,
+      "loss": 1.607,
+      "step": 12400
+    },
+    {
+      "epoch": 0.3612880705122611,
+      "grad_norm": 12.0625,
+      "learning_rate": 8.795708990228639e-06,
+      "loss": 1.7249,
+      "step": 12420
+    },
+    {
+      "epoch": 0.36186985484480905,
+      "grad_norm": 10.0625,
+      "learning_rate": 8.793769708556024e-06,
+      "loss": 1.7111,
+      "step": 12440
+    },
+    {
+      "epoch": 0.36245163917735695,
+      "grad_norm": 13.375,
+      "learning_rate": 8.791830426883409e-06,
+      "loss": 1.6574,
+      "step": 12460
+    },
+    {
+      "epoch": 0.36303342350990486,
+      "grad_norm": 11.5625,
+      "learning_rate": 8.789891145210794e-06,
+      "loss": 1.7416,
+      "step": 12480
+    },
+    {
+      "epoch": 0.3636152078424528,
+      "grad_norm": 11.0,
+      "learning_rate": 8.78795186353818e-06,
+      "loss": 1.6661,
+      "step": 12500
+    },
+    {
+      "epoch": 0.36419699217500073,
+      "grad_norm": 11.625,
+      "learning_rate": 8.786012581865564e-06,
+      "loss": 1.7245,
+      "step": 12520
+    },
+    {
+      "epoch": 0.36477877650754864,
+      "grad_norm": 15.25,
+      "learning_rate": 8.78407330019295e-06,
+      "loss": 1.6578,
+      "step": 12540
+    },
+    {
+      "epoch": 0.36536056084009655,
+      "grad_norm": 15.1875,
+      "learning_rate": 8.782134018520335e-06,
+      "loss": 1.708,
+      "step": 12560
+    },
+    {
+      "epoch": 0.3659423451726445,
+      "grad_norm": 12.375,
+      "learning_rate": 8.78019473684772e-06,
+      "loss": 1.7078,
+      "step": 12580
+    },
+    {
+      "epoch": 0.3665241295051924,
+      "grad_norm": 12.1875,
+      "learning_rate": 8.778255455175105e-06,
+      "loss": 1.6425,
+      "step": 12600
+    },
+    {
+      "epoch": 0.3671059138377403,
+      "grad_norm": 11.5,
+      "learning_rate": 8.77631617350249e-06,
+      "loss": 1.8068,
+      "step": 12620
+    },
+    {
+      "epoch": 0.3676876981702883,
+      "grad_norm": 12.4375,
+      "learning_rate": 8.774376891829875e-06,
+      "loss": 1.6618,
+      "step": 12640
+    },
+    {
+      "epoch": 0.3682694825028362,
+      "grad_norm": 12.25,
+      "learning_rate": 8.77243761015726e-06,
+      "loss": 1.5911,
+      "step": 12660
+    },
+    {
+      "epoch": 0.3688512668353841,
+      "grad_norm": 13.75,
+      "learning_rate": 8.770498328484645e-06,
+      "loss": 1.5994,
+      "step": 12680
+    },
+    {
+      "epoch": 0.36943305116793207,
+      "grad_norm": 12.75,
+      "learning_rate": 8.76855904681203e-06,
+      "loss": 1.6917,
+      "step": 12700
+    },
+    {
+      "epoch": 0.37001483550048,
+      "grad_norm": 10.0,
+      "learning_rate": 8.766619765139415e-06,
+      "loss": 1.6336,
+      "step": 12720
+    },
+    {
+      "epoch": 0.3705966198330279,
+      "grad_norm": 13.625,
+      "learning_rate": 8.7646804834668e-06,
+      "loss": 1.7117,
+      "step": 12740
+    },
+    {
+      "epoch": 0.37117840416557585,
+      "grad_norm": 12.5,
+      "learning_rate": 8.762741201794186e-06,
+      "loss": 1.6783,
+      "step": 12760
+    },
+    {
+      "epoch": 0.37176018849812376,
+      "grad_norm": 12.9375,
+      "learning_rate": 8.76080192012157e-06,
+      "loss": 1.6464,
+      "step": 12780
+    },
+    {
+      "epoch": 0.37234197283067166,
+      "grad_norm": 11.3125,
+      "learning_rate": 8.758862638448956e-06,
+      "loss": 1.7115,
+      "step": 12800
+    },
+    {
+      "epoch": 0.37292375716321957,
+      "grad_norm": 11.0625,
+      "learning_rate": 8.75692335677634e-06,
+      "loss": 1.6268,
+      "step": 12820
+    },
+    {
+      "epoch": 0.37350554149576753,
+      "grad_norm": 11.1875,
+      "learning_rate": 8.754984075103726e-06,
+      "loss": 1.5796,
+      "step": 12840
+    },
+    {
+      "epoch": 0.37408732582831544,
+      "grad_norm": 13.3125,
+      "learning_rate": 8.753044793431111e-06,
+      "loss": 1.6056,
+      "step": 12860
+    },
+    {
+      "epoch": 0.37466911016086335,
+      "grad_norm": 15.9375,
+      "learning_rate": 8.751105511758496e-06,
+      "loss": 1.6311,
+      "step": 12880
+    },
+    {
+      "epoch": 0.3752508944934113,
+      "grad_norm": 14.8125,
+      "learning_rate": 8.749166230085881e-06,
+      "loss": 1.6125,
+      "step": 12900
+    },
+    {
+      "epoch": 0.3758326788259592,
+      "grad_norm": 11.4375,
+      "learning_rate": 8.747226948413266e-06,
+      "loss": 1.6119,
+      "step": 12920
+    },
+    {
+      "epoch": 0.37641446315850713,
+      "grad_norm": 14.5625,
+      "learning_rate": 8.74528766674065e-06,
+      "loss": 1.6739,
+      "step": 12940
+    },
+    {
+      "epoch": 0.3769962474910551,
+      "grad_norm": 9.25,
+      "learning_rate": 8.743348385068035e-06,
+      "loss": 1.7119,
+      "step": 12960
+    },
+    {
+      "epoch": 0.377578031823603,
+      "grad_norm": 13.0,
+      "learning_rate": 8.74140910339542e-06,
+      "loss": 1.6802,
+      "step": 12980
+    },
+    {
+      "epoch": 0.3781598161561509,
+      "grad_norm": 12.1875,
+      "learning_rate": 8.739469821722805e-06,
+      "loss": 1.5522,
+      "step": 13000
+    },
+    {
+      "epoch": 0.3787416004886988,
+      "grad_norm": 16.375,
+      "learning_rate": 8.73753054005019e-06,
+      "loss": 1.6771,
+      "step": 13020
+    },
+    {
+      "epoch": 0.3793233848212468,
+      "grad_norm": 10.5625,
+      "learning_rate": 8.735591258377575e-06,
+      "loss": 1.6635,
+      "step": 13040
+    },
+    {
+      "epoch": 0.3799051691537947,
+      "grad_norm": 10.8125,
+      "learning_rate": 8.73365197670496e-06,
+      "loss": 1.6157,
+      "step": 13060
+    },
+    {
+      "epoch": 0.3804869534863426,
+      "grad_norm": 11.125,
+      "learning_rate": 8.731712695032345e-06,
+      "loss": 1.6574,
+      "step": 13080
+    },
+    {
+      "epoch": 0.38106873781889056,
+      "grad_norm": 15.8125,
+      "learning_rate": 8.72977341335973e-06,
+      "loss": 1.6298,
+      "step": 13100
+    },
+    {
+      "epoch": 0.38165052215143846,
+      "grad_norm": 13.0,
+      "learning_rate": 8.727834131687116e-06,
+      "loss": 1.749,
+      "step": 13120
+    },
+    {
+      "epoch": 0.3822323064839864,
+      "grad_norm": 16.625,
+      "learning_rate": 8.7258948500145e-06,
+      "loss": 1.736,
+      "step": 13140
+    },
+    {
+      "epoch": 0.38281409081653434,
+      "grad_norm": 11.875,
+      "learning_rate": 8.723955568341886e-06,
+      "loss": 1.6702,
+      "step": 13160
+    },
+    {
+      "epoch": 0.38339587514908224,
+      "grad_norm": 12.75,
+      "learning_rate": 8.722016286669271e-06,
+      "loss": 1.6749,
+      "step": 13180
+    },
+    {
+      "epoch": 0.38397765948163015,
+      "grad_norm": 12.875,
+      "learning_rate": 8.720077004996656e-06,
+      "loss": 1.7487,
+      "step": 13200
+    },
+    {
+      "epoch": 0.38455944381417806,
+      "grad_norm": 12.5625,
+      "learning_rate": 8.718137723324041e-06,
+      "loss": 1.6517,
+      "step": 13220
+    },
+    {
+      "epoch": 0.385141228146726,
+      "grad_norm": 11.5,
+      "learning_rate": 8.716198441651426e-06,
+      "loss": 1.6829,
+      "step": 13240
+    },
+    {
+      "epoch": 0.38572301247927393,
+      "grad_norm": 12.875,
+      "learning_rate": 8.714259159978811e-06,
+      "loss": 1.6287,
+      "step": 13260
+    },
+    {
+      "epoch": 0.38630479681182184,
+      "grad_norm": 13.1875,
+      "learning_rate": 8.712319878306196e-06,
+      "loss": 1.7027,
+      "step": 13280
+    },
+    {
+      "epoch": 0.3868865811443698,
+      "grad_norm": 12.75,
+      "learning_rate": 8.710380596633581e-06,
+      "loss": 1.762,
+      "step": 13300
+    },
+    {
+      "epoch": 0.3874683654769177,
+      "grad_norm": 15.3125,
+      "learning_rate": 8.708441314960967e-06,
+      "loss": 1.6645,
+      "step": 13320
+    },
+    {
+      "epoch": 0.3880501498094656,
+      "grad_norm": 14.0,
+      "learning_rate": 8.706502033288352e-06,
+      "loss": 1.6303,
+      "step": 13340
+    },
+    {
+      "epoch": 0.3886319341420136,
+      "grad_norm": 11.4375,
+      "learning_rate": 8.704562751615737e-06,
+      "loss": 1.6327,
+      "step": 13360
+    },
+    {
+      "epoch": 0.3892137184745615,
+      "grad_norm": 10.8125,
+      "learning_rate": 8.702623469943122e-06,
+      "loss": 1.7711,
+      "step": 13380
+    },
+    {
+      "epoch": 0.3897955028071094,
+      "grad_norm": 12.75,
+      "learning_rate": 8.700684188270507e-06,
+      "loss": 1.697,
+      "step": 13400
+    },
+    {
+      "epoch": 0.3903772871396573,
+      "grad_norm": 12.1875,
+      "learning_rate": 8.698744906597892e-06,
+      "loss": 1.6974,
+      "step": 13420
+    },
+    {
+      "epoch": 0.39095907147220527,
+      "grad_norm": 10.875,
+      "learning_rate": 8.696805624925277e-06,
+      "loss": 1.6802,
+      "step": 13440
+    },
+    {
+      "epoch": 0.3915408558047532,
+      "grad_norm": 7.59375,
+      "learning_rate": 8.694866343252662e-06,
+      "loss": 1.6845,
+      "step": 13460
+    },
+    {
+      "epoch": 0.3921226401373011,
+      "grad_norm": 12.9375,
+      "learning_rate": 8.692927061580047e-06,
+      "loss": 1.7635,
+      "step": 13480
+    },
+    {
+      "epoch": 0.39270442446984904,
+      "grad_norm": 13.5,
+      "learning_rate": 8.690987779907432e-06,
+      "loss": 1.6645,
+      "step": 13500
+    },
+    {
+      "epoch": 0.39328620880239695,
+      "grad_norm": 13.375,
+      "learning_rate": 8.689048498234818e-06,
+      "loss": 1.6436,
+      "step": 13520
+    },
+    {
+      "epoch": 0.39386799313494486,
+      "grad_norm": 11.75,
+      "learning_rate": 8.687109216562203e-06,
+      "loss": 1.6567,
+      "step": 13540
+    },
+    {
+      "epoch": 0.3944497774674928,
+      "grad_norm": 11.0625,
+      "learning_rate": 8.685169934889588e-06,
+      "loss": 1.6634,
+      "step": 13560
+    },
+    {
+      "epoch": 0.39503156180004073,
+      "grad_norm": 11.3125,
+      "learning_rate": 8.683230653216973e-06,
+      "loss": 1.6337,
+      "step": 13580
+    },
+    {
+      "epoch": 0.39561334613258864,
+      "grad_norm": 10.75,
+      "learning_rate": 8.681291371544358e-06,
+      "loss": 1.6324,
+      "step": 13600
+    },
+    {
+      "epoch": 0.39619513046513655,
+      "grad_norm": 12.0625,
+      "learning_rate": 8.679352089871743e-06,
+      "loss": 1.7301,
+      "step": 13620
+    },
+    {
+      "epoch": 0.3967769147976845,
+      "grad_norm": 12.1875,
+      "learning_rate": 8.677412808199128e-06,
+      "loss": 1.6491,
+      "step": 13640
+    },
+    {
+      "epoch": 0.3973586991302324,
+      "grad_norm": 10.5625,
+      "learning_rate": 8.675473526526513e-06,
+      "loss": 1.6356,
+      "step": 13660
+    },
+    {
+      "epoch": 0.3979404834627803,
+      "grad_norm": 13.5625,
+      "learning_rate": 8.673534244853898e-06,
+      "loss": 1.6638,
+      "step": 13680
+    },
+    {
+      "epoch": 0.3985222677953283,
+      "grad_norm": 12.5,
+      "learning_rate": 8.671594963181283e-06,
+      "loss": 1.6306,
+      "step": 13700
+    },
+    {
+      "epoch": 0.3991040521278762,
+      "grad_norm": 11.875,
+      "learning_rate": 8.669655681508669e-06,
+      "loss": 1.6401,
+      "step": 13720
+    },
+    {
+      "epoch": 0.3996858364604241,
+      "grad_norm": 14.1875,
+      "learning_rate": 8.667716399836054e-06,
+      "loss": 1.7083,
+      "step": 13740
+    },
+    {
+      "epoch": 0.40026762079297207,
+      "grad_norm": 11.0,
+      "learning_rate": 8.665777118163439e-06,
+      "loss": 1.6409,
+      "step": 13760
+    },
+    {
+      "epoch": 0.40084940512552,
+      "grad_norm": 12.625,
+      "learning_rate": 8.663837836490824e-06,
+      "loss": 1.6309,
+      "step": 13780
+    },
+    {
+      "epoch": 0.4014311894580679,
+      "grad_norm": 14.0625,
+      "learning_rate": 8.661898554818209e-06,
+      "loss": 1.6464,
+      "step": 13800
+    },
+    {
+      "epoch": 0.40201297379061585,
+      "grad_norm": 12.5625,
+      "learning_rate": 8.659959273145594e-06,
+      "loss": 1.6834,
+      "step": 13820
+    },
+    {
+      "epoch": 0.40259475812316375,
+      "grad_norm": 10.9375,
+      "learning_rate": 8.658019991472979e-06,
+      "loss": 1.5307,
+      "step": 13840
+    },
+    {
+      "epoch": 0.40317654245571166,
+      "grad_norm": 12.0,
+      "learning_rate": 8.656080709800364e-06,
+      "loss": 1.6924,
+      "step": 13860
+    },
+    {
+      "epoch": 0.40375832678825957,
+      "grad_norm": 14.0,
+      "learning_rate": 8.65414142812775e-06,
+      "loss": 1.687,
+      "step": 13880
+    },
+    {
+      "epoch": 0.40434011112080753,
+      "grad_norm": 12.5625,
+      "learning_rate": 8.652202146455134e-06,
+      "loss": 1.6729,
+      "step": 13900
+    },
+    {
+      "epoch": 0.40492189545335544,
+      "grad_norm": 10.6875,
+      "learning_rate": 8.65026286478252e-06,
+      "loss": 1.6525,
+      "step": 13920
+    },
+    {
+      "epoch": 0.40550367978590335,
+      "grad_norm": 11.8125,
+      "learning_rate": 8.648323583109905e-06,
+      "loss": 1.6706,
+      "step": 13940
+    },
+    {
+      "epoch": 0.4060854641184513,
+      "grad_norm": 16.25,
+      "learning_rate": 8.64638430143729e-06,
+      "loss": 1.7039,
+      "step": 13960
+    },
+    {
+      "epoch": 0.4066672484509992,
+      "grad_norm": 13.9375,
+      "learning_rate": 8.644445019764675e-06,
+      "loss": 1.6011,
+      "step": 13980
+    },
+    {
+      "epoch": 0.4072490327835471,
+      "grad_norm": 10.5625,
+      "learning_rate": 8.64250573809206e-06,
+      "loss": 1.6004,
+      "step": 14000
+    },
+    {
+      "epoch": 0.4078308171160951,
+      "grad_norm": 13.875,
+      "learning_rate": 8.640566456419445e-06,
+      "loss": 1.6513,
+      "step": 14020
+    },
+    {
+      "epoch": 0.408412601448643,
+      "grad_norm": 13.125,
+      "learning_rate": 8.63862717474683e-06,
+      "loss": 1.637,
+      "step": 14040
+    },
+    {
+      "epoch": 0.4089943857811909,
+      "grad_norm": 11.375,
+      "learning_rate": 8.636687893074215e-06,
+      "loss": 1.6971,
+      "step": 14060
+    },
+    {
+      "epoch": 0.4095761701137388,
+      "grad_norm": 11.375,
+      "learning_rate": 8.6347486114016e-06,
+      "loss": 1.6934,
+      "step": 14080
+    },
+    {
+      "epoch": 0.4101579544462868,
+      "grad_norm": 14.1875,
+      "learning_rate": 8.632809329728985e-06,
+      "loss": 1.6761,
+      "step": 14100
+    },
+    {
+      "epoch": 0.4107397387788347,
+      "grad_norm": 11.125,
+      "learning_rate": 8.63087004805637e-06,
+      "loss": 1.7236,
+      "step": 14120
+    },
+    {
+      "epoch": 0.4113215231113826,
+      "grad_norm": 12.8125,
+      "learning_rate": 8.628930766383756e-06,
+      "loss": 1.7004,
+      "step": 14140
+    },
+    {
+      "epoch": 0.41190330744393056,
+      "grad_norm": 11.625,
+      "learning_rate": 8.62699148471114e-06,
+      "loss": 1.5457,
+      "step": 14160
+    },
+    {
+      "epoch": 0.41248509177647846,
+      "grad_norm": 14.375,
+      "learning_rate": 8.625052203038526e-06,
+      "loss": 1.5415,
+      "step": 14180
+    },
+    {
+      "epoch": 0.41306687610902637,
+      "grad_norm": 10.5625,
+      "learning_rate": 8.623112921365911e-06,
+      "loss": 1.7406,
+      "step": 14200
+    },
+    {
+      "epoch": 0.41364866044157433,
+      "grad_norm": 19.625,
+      "learning_rate": 8.621173639693296e-06,
+      "loss": 1.675,
+      "step": 14220
+    },
+    {
+      "epoch": 0.41423044477412224,
+      "grad_norm": 12.875,
+      "learning_rate": 8.619234358020681e-06,
+      "loss": 1.7145,
+      "step": 14240
+    },
+    {
+      "epoch": 0.41481222910667015,
+      "grad_norm": 10.75,
+      "learning_rate": 8.617295076348066e-06,
+      "loss": 1.6529,
+      "step": 14260
+    },
+    {
+      "epoch": 0.41539401343921806,
+      "grad_norm": 11.625,
+      "learning_rate": 8.615355794675451e-06,
+      "loss": 1.644,
+      "step": 14280
+    },
+    {
+      "epoch": 0.415975797771766,
+      "grad_norm": 13.3125,
+      "learning_rate": 8.613416513002836e-06,
+      "loss": 1.6978,
+      "step": 14300
+    },
+    {
+      "epoch": 0.41655758210431393,
+      "grad_norm": 13.3125,
+      "learning_rate": 8.611477231330221e-06,
+      "loss": 1.683,
+      "step": 14320
+    },
+    {
+      "epoch": 0.41713936643686184,
+      "grad_norm": 11.25,
+      "learning_rate": 8.609537949657607e-06,
+      "loss": 1.6661,
+      "step": 14340
+    },
+    {
+      "epoch": 0.4177211507694098,
+      "grad_norm": 12.75,
+      "learning_rate": 8.607598667984992e-06,
+      "loss": 1.6891,
+      "step": 14360
+    },
+    {
+      "epoch": 0.4183029351019577,
+      "grad_norm": 13.5,
+      "learning_rate": 8.605659386312377e-06,
+      "loss": 1.6421,
+      "step": 14380
+    },
+    {
+      "epoch": 0.4188847194345056,
+      "grad_norm": 10.8125,
+      "learning_rate": 8.603720104639762e-06,
+      "loss": 1.6044,
+      "step": 14400
+    },
+    {
+      "epoch": 0.4194665037670536,
+      "grad_norm": 13.3125,
+      "learning_rate": 8.601780822967147e-06,
+      "loss": 1.6135,
+      "step": 14420
+    },
+    {
+      "epoch": 0.4200482880996015,
+      "grad_norm": 12.0,
+      "learning_rate": 8.599841541294532e-06,
+      "loss": 1.6754,
+      "step": 14440
+    },
+    {
+      "epoch": 0.4206300724321494,
+      "grad_norm": 13.25,
+      "learning_rate": 8.597902259621917e-06,
+      "loss": 1.6302,
+      "step": 14460
+    },
+    {
+      "epoch": 0.4212118567646973,
+      "grad_norm": 11.0625,
+      "learning_rate": 8.595962977949302e-06,
+      "loss": 1.5996,
+      "step": 14480
+    },
+    {
+      "epoch": 0.42179364109724526,
+      "grad_norm": 12.25,
+      "learning_rate": 8.594023696276687e-06,
+      "loss": 1.6721,
+      "step": 14500
+    },
+    {
+      "epoch": 0.42237542542979317,
+      "grad_norm": 11.8125,
+      "learning_rate": 8.592084414604072e-06,
+      "loss": 1.6192,
+      "step": 14520
+    },
+    {
+      "epoch": 0.4229572097623411,
+      "grad_norm": 14.0,
+      "learning_rate": 8.590145132931458e-06,
+      "loss": 1.5439,
+      "step": 14540
+    },
+    {
+      "epoch": 0.42353899409488904,
+      "grad_norm": 12.5,
+      "learning_rate": 8.588205851258841e-06,
+      "loss": 1.661,
+      "step": 14560
+    },
+    {
+      "epoch": 0.42412077842743695,
+      "grad_norm": 11.4375,
+      "learning_rate": 8.586266569586226e-06,
+      "loss": 1.6487,
+      "step": 14580
+    },
+    {
+      "epoch": 0.42470256275998486,
+      "grad_norm": 15.6875,
+      "learning_rate": 8.584327287913611e-06,
+      "loss": 1.6176,
+      "step": 14600
+    },
+    {
+      "epoch": 0.4252843470925328,
+      "grad_norm": 12.5,
+      "learning_rate": 8.582388006240996e-06,
+      "loss": 1.6658,
+      "step": 14620
+    },
+    {
+      "epoch": 0.42586613142508073,
+      "grad_norm": 12.4375,
+      "learning_rate": 8.580448724568381e-06,
+      "loss": 1.6275,
+      "step": 14640
+    },
+    {
+      "epoch": 0.42644791575762864,
+      "grad_norm": 13.4375,
+      "learning_rate": 8.578509442895766e-06,
+      "loss": 1.6152,
+      "step": 14660
+    },
+    {
+      "epoch": 0.42702970009017654,
+      "grad_norm": 11.5,
+      "learning_rate": 8.576570161223152e-06,
+      "loss": 1.6167,
+      "step": 14680
+    },
+    {
+      "epoch": 0.4276114844227245,
+      "grad_norm": 13.5,
+      "learning_rate": 8.574630879550537e-06,
+      "loss": 1.605,
+      "step": 14700
+    },
+    {
+      "epoch": 0.4281932687552724,
+      "grad_norm": 16.0,
+      "learning_rate": 8.572691597877922e-06,
+      "loss": 1.6955,
+      "step": 14720
+    },
+    {
+      "epoch": 0.4287750530878203,
+      "grad_norm": 13.6875,
+      "learning_rate": 8.570752316205307e-06,
+      "loss": 1.5291,
+      "step": 14740
+    },
+    {
+      "epoch": 0.4293568374203683,
+      "grad_norm": 12.125,
+      "learning_rate": 8.568813034532692e-06,
+      "loss": 1.7469,
+      "step": 14760
+    },
+    {
+      "epoch": 0.4299386217529162,
+      "grad_norm": 11.5625,
+      "learning_rate": 8.566873752860077e-06,
+      "loss": 1.7228,
+      "step": 14780
+    },
+    {
+      "epoch": 0.4305204060854641,
+      "grad_norm": 14.75,
+      "learning_rate": 8.564934471187462e-06,
+      "loss": 1.6517,
+      "step": 14800
+    },
+    {
+      "epoch": 0.43110219041801207,
+      "grad_norm": 11.0,
+      "learning_rate": 8.562995189514847e-06,
+      "loss": 1.5406,
+      "step": 14820
+    },
+    {
+      "epoch": 0.43168397475056,
+      "grad_norm": 12.25,
+      "learning_rate": 8.561055907842232e-06,
+      "loss": 1.6401,
+      "step": 14840
+    },
+    {
+      "epoch": 0.4322657590831079,
+      "grad_norm": 12.0625,
+      "learning_rate": 8.559116626169617e-06,
+      "loss": 1.6032,
+      "step": 14860
+    },
+    {
+      "epoch": 0.4328475434156558,
+      "grad_norm": 12.5,
+      "learning_rate": 8.557177344497002e-06,
+      "loss": 1.7027,
+      "step": 14880
+    },
+    {
+      "epoch": 0.43342932774820375,
+      "grad_norm": 11.8125,
+      "learning_rate": 8.555238062824388e-06,
+      "loss": 1.6566,
+      "step": 14900
+    },
+    {
+      "epoch": 0.43401111208075166,
+      "grad_norm": 13.1875,
+      "learning_rate": 8.553298781151773e-06,
+      "loss": 1.6975,
+      "step": 14920
+    },
+    {
+      "epoch": 0.43459289641329957,
+      "grad_norm": 11.8125,
+      "learning_rate": 8.551359499479158e-06,
+      "loss": 1.6332,
+      "step": 14940
+    },
+    {
+      "epoch": 0.43517468074584753,
+      "grad_norm": 12.5,
+      "learning_rate": 8.549420217806543e-06,
+      "loss": 1.6447,
+      "step": 14960
+    },
+    {
+      "epoch": 0.43575646507839544,
+      "grad_norm": 13.0625,
+      "learning_rate": 8.547480936133928e-06,
+      "loss": 1.7399,
+      "step": 14980
+    },
+    {
+      "epoch": 0.43633824941094335,
+      "grad_norm": 12.6875,
+      "learning_rate": 8.545541654461313e-06,
+      "loss": 1.7293,
+      "step": 15000
+    },
+    {
+      "epoch": 0.4369200337434913,
+      "grad_norm": 13.0625,
+      "learning_rate": 8.543602372788698e-06,
+      "loss": 1.6866,
+      "step": 15020
+    },
+    {
+      "epoch": 0.4375018180760392,
+      "grad_norm": 13.9375,
+      "learning_rate": 8.541663091116083e-06,
+      "loss": 1.6139,
+      "step": 15040
+    },
+    {
+      "epoch": 0.4380836024085871,
+      "grad_norm": 12.9375,
+      "learning_rate": 8.539723809443468e-06,
+      "loss": 1.6182,
+      "step": 15060
+    },
+    {
+      "epoch": 0.4386653867411351,
+      "grad_norm": 13.375,
+      "learning_rate": 8.537784527770853e-06,
+      "loss": 1.6955,
+      "step": 15080
+    },
+    {
+      "epoch": 0.439247171073683,
+      "grad_norm": 12.6875,
+      "learning_rate": 8.535845246098239e-06,
+      "loss": 1.5942,
+      "step": 15100
+    },
+    {
+      "epoch": 0.4398289554062309,
+      "grad_norm": 12.5,
+      "learning_rate": 8.533905964425624e-06,
+      "loss": 1.6648,
+      "step": 15120
+    },
+    {
+      "epoch": 0.4404107397387788,
+      "grad_norm": 12.25,
+      "learning_rate": 8.531966682753009e-06,
+      "loss": 1.6914,
+      "step": 15140
+    },
+    {
+      "epoch": 0.4409925240713268,
+      "grad_norm": 11.9375,
+      "learning_rate": 8.530027401080394e-06,
+      "loss": 1.6168,
+      "step": 15160
+    },
+    {
+      "epoch": 0.4415743084038747,
+      "grad_norm": 11.5625,
+      "learning_rate": 8.528088119407779e-06,
+      "loss": 1.7224,
+      "step": 15180
+    },
+    {
+      "epoch": 0.4421560927364226,
+      "grad_norm": 13.6875,
+      "learning_rate": 8.526148837735164e-06,
+      "loss": 1.6296,
+      "step": 15200
+    },
+    {
+      "epoch": 0.44273787706897055,
+      "grad_norm": 12.25,
+      "learning_rate": 8.524209556062549e-06,
+      "loss": 1.5624,
+      "step": 15220
+    },
+    {
+      "epoch": 0.44331966140151846,
+      "grad_norm": 12.5625,
+      "learning_rate": 8.522270274389934e-06,
+      "loss": 1.6387,
+      "step": 15240
+    },
+    {
+      "epoch": 0.44390144573406637,
+      "grad_norm": 14.375,
+      "learning_rate": 8.52033099271732e-06,
+      "loss": 1.6439,
+      "step": 15260
+    },
+    {
+      "epoch": 0.44448323006661433,
+      "grad_norm": 11.1875,
+      "learning_rate": 8.518391711044704e-06,
+      "loss": 1.697,
+      "step": 15280
+    },
+    {
+      "epoch": 0.44506501439916224,
+      "grad_norm": 13.4375,
+      "learning_rate": 8.51645242937209e-06,
+      "loss": 1.7002,
+      "step": 15300
+    },
+    {
+      "epoch": 0.44564679873171015,
+      "grad_norm": 13.875,
+      "learning_rate": 8.514513147699475e-06,
+      "loss": 1.6779,
+      "step": 15320
+    },
+    {
+      "epoch": 0.44622858306425806,
+      "grad_norm": 12.8125,
+      "learning_rate": 8.51257386602686e-06,
+      "loss": 1.6922,
+      "step": 15340
+    },
+    {
+      "epoch": 0.446810367396806,
+      "grad_norm": 11.3125,
+      "learning_rate": 8.510634584354245e-06,
+      "loss": 1.6771,
+      "step": 15360
+    },
+    {
+      "epoch": 0.4473921517293539,
+      "grad_norm": 14.1875,
+      "learning_rate": 8.50869530268163e-06,
+      "loss": 1.7192,
+      "step": 15380
+    },
+    {
+      "epoch": 0.44797393606190183,
+      "grad_norm": 15.25,
+      "learning_rate": 8.506756021009015e-06,
+      "loss": 1.6994,
+      "step": 15400
+    },
+    {
+      "epoch": 0.4485557203944498,
+      "grad_norm": 11.125,
+      "learning_rate": 8.5048167393364e-06,
+      "loss": 1.7224,
+      "step": 15420
+    },
+    {
+      "epoch": 0.4491375047269977,
+      "grad_norm": 13.5625,
+      "learning_rate": 8.502877457663785e-06,
+      "loss": 1.596,
+      "step": 15440
+    },
+    {
+      "epoch": 0.4497192890595456,
+      "grad_norm": 13.375,
+      "learning_rate": 8.50093817599117e-06,
+      "loss": 1.626,
+      "step": 15460
+    },
+    {
+      "epoch": 0.4503010733920936,
+      "grad_norm": 11.125,
+      "learning_rate": 8.498998894318555e-06,
+      "loss": 1.6653,
+      "step": 15480
+    },
+    {
+      "epoch": 0.4508828577246415,
+      "grad_norm": 11.375,
+      "learning_rate": 8.49705961264594e-06,
+      "loss": 1.6906,
+      "step": 15500
+    },
+    {
+      "epoch": 0.4514646420571894,
+      "grad_norm": 12.1875,
+      "learning_rate": 8.495120330973326e-06,
+      "loss": 1.5862,
+      "step": 15520
+    },
+    {
+      "epoch": 0.4520464263897373,
+      "grad_norm": 12.125,
+      "learning_rate": 8.49318104930071e-06,
+      "loss": 1.624,
+      "step": 15540
+    },
+    {
+      "epoch": 0.45262821072228526,
+      "grad_norm": 10.875,
+      "learning_rate": 8.491241767628096e-06,
+      "loss": 1.6233,
+      "step": 15560
+    },
+    {
+      "epoch": 0.45320999505483317,
+      "grad_norm": 11.0,
+      "learning_rate": 8.489302485955481e-06,
+      "loss": 1.6777,
+      "step": 15580
+    },
+    {
+      "epoch": 0.4537917793873811,
+      "grad_norm": 12.0,
+      "learning_rate": 8.487363204282866e-06,
+      "loss": 1.6539,
+      "step": 15600
+    },
+    {
+      "epoch": 0.45437356371992904,
+      "grad_norm": 12.25,
+      "learning_rate": 8.485423922610251e-06,
+      "loss": 1.6355,
+      "step": 15620
+    },
+    {
+      "epoch": 0.45495534805247695,
+      "grad_norm": 13.3125,
+      "learning_rate": 8.483484640937636e-06,
+      "loss": 1.7094,
+      "step": 15640
+    },
+    {
+      "epoch": 0.45553713238502486,
+      "grad_norm": 12.625,
+      "learning_rate": 8.481545359265021e-06,
+      "loss": 1.6191,
+      "step": 15660
+    },
+    {
+      "epoch": 0.4561189167175728,
+      "grad_norm": 12.75,
+      "learning_rate": 8.479606077592406e-06,
+      "loss": 1.6599,
+      "step": 15680
+    },
+    {
+      "epoch": 0.4567007010501207,
+      "grad_norm": 11.4375,
+      "learning_rate": 8.477666795919791e-06,
+      "loss": 1.6194,
+      "step": 15700
+    },
+    {
+      "epoch": 0.45728248538266864,
+      "grad_norm": 12.375,
+      "learning_rate": 8.475727514247177e-06,
+      "loss": 1.7066,
+      "step": 15720
+    },
+    {
+      "epoch": 0.45786426971521654,
+      "grad_norm": 13.5625,
+      "learning_rate": 8.473788232574562e-06,
+      "loss": 1.6937,
+      "step": 15740
+    },
+    {
+      "epoch": 0.4584460540477645,
+      "grad_norm": 10.625,
+      "learning_rate": 8.471848950901947e-06,
+      "loss": 1.6028,
+      "step": 15760
+    },
+    {
+      "epoch": 0.4590278383803124,
+      "grad_norm": 12.6875,
+      "learning_rate": 8.469909669229332e-06,
+      "loss": 1.7428,
+      "step": 15780
+    },
+    {
+      "epoch": 0.4596096227128603,
+      "grad_norm": 12.75,
+      "learning_rate": 8.467970387556717e-06,
+      "loss": 1.6265,
+      "step": 15800
+    },
+    {
+      "epoch": 0.4601914070454083,
+      "grad_norm": 13.5,
+      "learning_rate": 8.466031105884102e-06,
+      "loss": 1.7238,
+      "step": 15820
+    },
+    {
+      "epoch": 0.4607731913779562,
+      "grad_norm": 17.5,
+      "learning_rate": 8.464091824211487e-06,
+      "loss": 1.6624,
+      "step": 15840
+    },
+    {
+      "epoch": 0.4613549757105041,
+      "grad_norm": 8.875,
+      "learning_rate": 8.462152542538872e-06,
+      "loss": 1.6134,
+      "step": 15860
+    },
+    {
+      "epoch": 0.46193676004305206,
+      "grad_norm": 11.4375,
+      "learning_rate": 8.460213260866257e-06,
+      "loss": 1.6501,
+      "step": 15880
+    },
+    {
+      "epoch": 0.46251854437559997,
+      "grad_norm": 15.25,
+      "learning_rate": 8.458273979193642e-06,
+      "loss": 1.6335,
+      "step": 15900
+    },
+    {
+      "epoch": 0.4631003287081479,
+      "grad_norm": 13.25,
+      "learning_rate": 8.456334697521028e-06,
+      "loss": 1.6504,
+      "step": 15920
+    },
+    {
+      "epoch": 0.4636821130406958,
+      "grad_norm": 13.4375,
+      "learning_rate": 8.454395415848413e-06,
+      "loss": 1.5638,
+      "step": 15940
+    },
+    {
+      "epoch": 0.46426389737324375,
+      "grad_norm": 13.5625,
+      "learning_rate": 8.452456134175798e-06,
+      "loss": 1.7835,
+      "step": 15960
+    },
+    {
+      "epoch": 0.46484568170579166,
+      "grad_norm": 13.5625,
+      "learning_rate": 8.450516852503183e-06,
+      "loss": 1.6515,
+      "step": 15980
+    },
+    {
+      "epoch": 0.46542746603833957,
+      "grad_norm": 12.0,
+      "learning_rate": 8.448577570830568e-06,
+      "loss": 1.6011,
+      "step": 16000
+    },
+    {
+      "epoch": 0.46600925037088753,
+      "grad_norm": 14.6875,
+      "learning_rate": 8.446638289157953e-06,
+      "loss": 1.637,
+      "step": 16020
+    },
+    {
+      "epoch": 0.46659103470343544,
+      "grad_norm": 12.125,
+      "learning_rate": 8.444699007485338e-06,
+      "loss": 1.7068,
+      "step": 16040
+    },
+    {
+      "epoch": 0.46717281903598334,
+      "grad_norm": 12.875,
+      "learning_rate": 8.442759725812723e-06,
+      "loss": 1.6211,
+      "step": 16060
+    },
+    {
+      "epoch": 0.4677546033685313,
+      "grad_norm": 13.6875,
+      "learning_rate": 8.440820444140108e-06,
+      "loss": 1.5866,
+      "step": 16080
+    },
+    {
+      "epoch": 0.4683363877010792,
+      "grad_norm": 11.8125,
+      "learning_rate": 8.438881162467493e-06,
+      "loss": 1.721,
+      "step": 16100
+    },
+    {
+      "epoch": 0.4689181720336271,
+      "grad_norm": 10.9375,
+      "learning_rate": 8.436941880794879e-06,
+      "loss": 1.6362,
+      "step": 16120
+    },
+    {
+      "epoch": 0.4694999563661751,
+      "grad_norm": 12.1875,
+      "learning_rate": 8.435002599122264e-06,
+      "loss": 1.6794,
+      "step": 16140
+    },
+    {
+      "epoch": 0.470081740698723,
+      "grad_norm": 14.375,
+      "learning_rate": 8.433063317449649e-06,
+      "loss": 1.6712,
+      "step": 16160
+    },
+    {
+      "epoch": 0.4706635250312709,
+      "grad_norm": 14.5625,
+      "learning_rate": 8.431124035777034e-06,
+      "loss": 1.6059,
+      "step": 16180
+    },
+    {
+      "epoch": 0.4712453093638188,
+      "grad_norm": 10.8125,
+      "learning_rate": 8.429184754104417e-06,
+      "loss": 1.6212,
+      "step": 16200
+    },
+    {
+      "epoch": 0.4718270936963668,
+      "grad_norm": 13.9375,
+      "learning_rate": 8.427245472431802e-06,
+      "loss": 1.6766,
+      "step": 16220
+    },
+    {
+      "epoch": 0.4724088780289147,
+      "grad_norm": 11.625,
+      "learning_rate": 8.425306190759187e-06,
+      "loss": 1.6919,
+      "step": 16240
+    },
+    {
+      "epoch": 0.4729906623614626,
+      "grad_norm": 12.0,
+      "learning_rate": 8.423366909086573e-06,
+      "loss": 1.6914,
+      "step": 16260
+    },
+    {
+      "epoch": 0.47357244669401055,
+      "grad_norm": 11.5,
+      "learning_rate": 8.421427627413958e-06,
+      "loss": 1.7106,
+      "step": 16280
+    },
+    {
+      "epoch": 0.47415423102655846,
+      "grad_norm": 12.3125,
+      "learning_rate": 8.419488345741343e-06,
+      "loss": 1.6553,
+      "step": 16300
+    },
+    {
+      "epoch": 0.47473601535910637,
+      "grad_norm": 11.9375,
+      "learning_rate": 8.417549064068728e-06,
+      "loss": 1.6824,
+      "step": 16320
+    },
+    {
+      "epoch": 0.47531779969165433,
+      "grad_norm": 14.0,
+      "learning_rate": 8.415609782396113e-06,
+      "loss": 1.6005,
+      "step": 16340
+    },
+    {
+      "epoch": 0.47589958402420224,
+      "grad_norm": 12.8125,
+      "learning_rate": 8.413670500723498e-06,
+      "loss": 1.6092,
+      "step": 16360
+    },
+    {
+      "epoch": 0.47648136835675015,
+      "grad_norm": 21.25,
+      "learning_rate": 8.411731219050883e-06,
+      "loss": 1.6701,
+      "step": 16380
+    },
+    {
+      "epoch": 0.47706315268929805,
+      "grad_norm": 14.625,
+      "learning_rate": 8.409791937378268e-06,
+      "loss": 1.6229,
+      "step": 16400
+    },
+    {
+      "epoch": 0.477644937021846,
+      "grad_norm": 11.375,
+      "learning_rate": 8.407852655705653e-06,
+      "loss": 1.5859,
+      "step": 16420
+    },
+    {
+      "epoch": 0.4782267213543939,
+      "grad_norm": 12.0625,
+      "learning_rate": 8.405913374033038e-06,
+      "loss": 1.6195,
+      "step": 16440
+    },
+    {
+      "epoch": 0.47880850568694183,
+      "grad_norm": 10.6875,
+      "learning_rate": 8.403974092360423e-06,
+      "loss": 1.7007,
+      "step": 16460
+    },
+    {
+      "epoch": 0.4793902900194898,
+      "grad_norm": 14.125,
+      "learning_rate": 8.402034810687809e-06,
+      "loss": 1.6735,
+      "step": 16480
+    },
+    {
+      "epoch": 0.4799720743520377,
+      "grad_norm": 12.5,
+      "learning_rate": 8.400095529015194e-06,
+      "loss": 1.6816,
+      "step": 16500
+    },
+    {
+      "epoch": 0.4805538586845856,
+      "grad_norm": 10.875,
+      "learning_rate": 8.398156247342579e-06,
+      "loss": 1.6985,
+      "step": 16520
+    },
+    {
+      "epoch": 0.4811356430171336,
+      "grad_norm": 11.75,
+      "learning_rate": 8.396216965669964e-06,
+      "loss": 1.7448,
+      "step": 16540
+    },
+    {
+      "epoch": 0.4817174273496815,
+      "grad_norm": 13.375,
+      "learning_rate": 8.394277683997349e-06,
+      "loss": 1.6185,
+      "step": 16560
+    },
+    {
+      "epoch": 0.4822992116822294,
+      "grad_norm": 14.875,
+      "learning_rate": 8.392338402324734e-06,
+      "loss": 1.6197,
+      "step": 16580
+    },
+    {
+      "epoch": 0.4828809960147773,
+      "grad_norm": 13.75,
+      "learning_rate": 8.39039912065212e-06,
+      "loss": 1.6744,
+      "step": 16600
+    },
+    {
+      "epoch": 0.48346278034732526,
+      "grad_norm": 11.875,
+      "learning_rate": 8.388459838979504e-06,
+      "loss": 1.6772,
+      "step": 16620
+    },
+    {
+      "epoch": 0.48404456467987317,
+      "grad_norm": 13.625,
+      "learning_rate": 8.38652055730689e-06,
+      "loss": 1.67,
+      "step": 16640
+    },
+    {
+      "epoch": 0.4846263490124211,
+      "grad_norm": 13.0625,
+      "learning_rate": 8.384581275634274e-06,
+      "loss": 1.6667,
+      "step": 16660
+    },
+    {
+      "epoch": 0.48520813334496904,
+      "grad_norm": 14.3125,
+      "learning_rate": 8.38264199396166e-06,
+      "loss": 1.7014,
+      "step": 16680
+    },
+    {
+      "epoch": 0.48578991767751695,
+      "grad_norm": 14.375,
+      "learning_rate": 8.380702712289045e-06,
+      "loss": 1.6499,
+      "step": 16700
+    },
+    {
+      "epoch": 0.48637170201006485,
+      "grad_norm": 14.0625,
+      "learning_rate": 8.37876343061643e-06,
+      "loss": 1.6073,
+      "step": 16720
+    },
+    {
+      "epoch": 0.4869534863426128,
+      "grad_norm": 13.8125,
+      "learning_rate": 8.376824148943815e-06,
+      "loss": 1.6416,
+      "step": 16740
+    },
+    {
+      "epoch": 0.4875352706751607,
+      "grad_norm": 15.5625,
+      "learning_rate": 8.374884867271198e-06,
+      "loss": 1.6485,
+      "step": 16760
+    },
+    {
+      "epoch": 0.48811705500770863,
+      "grad_norm": 12.3125,
+      "learning_rate": 8.372945585598583e-06,
+      "loss": 1.7104,
+      "step": 16780
+    },
+    {
+      "epoch": 0.48869883934025654,
+      "grad_norm": 10.6875,
+      "learning_rate": 8.371006303925968e-06,
+      "loss": 1.5394,
+      "step": 16800
+    },
+    {
+      "epoch": 0.4892806236728045,
+      "grad_norm": 14.0,
+      "learning_rate": 8.369067022253354e-06,
+      "loss": 1.5488,
+      "step": 16820
+    },
+    {
+      "epoch": 0.4898624080053524,
+      "grad_norm": 12.625,
+      "learning_rate": 8.367127740580739e-06,
+      "loss": 1.5977,
+      "step": 16840
+    },
+    {
+      "epoch": 0.4904441923379003,
+      "grad_norm": 11.875,
+      "learning_rate": 8.365188458908124e-06,
+      "loss": 1.5608,
+      "step": 16860
+    },
+    {
+      "epoch": 0.4910259766704483,
+      "grad_norm": 13.3125,
+      "learning_rate": 8.363249177235509e-06,
+      "loss": 1.6741,
+      "step": 16880
+    },
+    {
+      "epoch": 0.4916077610029962,
+      "grad_norm": 13.3125,
+      "learning_rate": 8.361309895562894e-06,
+      "loss": 1.6442,
+      "step": 16900
+    },
+    {
+      "epoch": 0.4921895453355441,
+      "grad_norm": 12.25,
+      "learning_rate": 8.359370613890279e-06,
+      "loss": 1.6331,
+      "step": 16920
+    },
+    {
+      "epoch": 0.49277132966809206,
+      "grad_norm": 11.75,
+      "learning_rate": 8.357431332217664e-06,
+      "loss": 1.627,
+      "step": 16940
+    },
+    {
+      "epoch": 0.49335311400063997,
+      "grad_norm": 12.6875,
+      "learning_rate": 8.35549205054505e-06,
+      "loss": 1.6895,
+      "step": 16960
+    },
+    {
+      "epoch": 0.4939348983331879,
+      "grad_norm": 10.0625,
+      "learning_rate": 8.353552768872434e-06,
+      "loss": 1.6495,
+      "step": 16980
+    },
+    {
+      "epoch": 0.4945166826657358,
+      "grad_norm": 12.25,
+      "learning_rate": 8.35161348719982e-06,
+      "loss": 1.6889,
+      "step": 17000
+    },
+    {
+      "epoch": 0.49509846699828375,
+      "grad_norm": 14.25,
+      "learning_rate": 8.349674205527205e-06,
+      "loss": 1.6643,
+      "step": 17020
+    },
+    {
+      "epoch": 0.49568025133083166,
+      "grad_norm": 13.625,
+      "learning_rate": 8.34773492385459e-06,
+      "loss": 1.572,
+      "step": 17040
+    },
+    {
+      "epoch": 0.49626203566337956,
+      "grad_norm": 11.4375,
+      "learning_rate": 8.345795642181975e-06,
+      "loss": 1.6003,
+      "step": 17060
+    },
+    {
+      "epoch": 0.4968438199959275,
+      "grad_norm": 13.25,
+      "learning_rate": 8.34385636050936e-06,
+      "loss": 1.7059,
+      "step": 17080
+    },
+    {
+      "epoch": 0.49742560432847543,
+      "grad_norm": 12.75,
+      "learning_rate": 8.341917078836745e-06,
+      "loss": 1.5876,
+      "step": 17100
+    },
+    {
+      "epoch": 0.49800738866102334,
+      "grad_norm": 12.0,
+      "learning_rate": 8.33997779716413e-06,
+      "loss": 1.6523,
+      "step": 17120
+    },
+    {
+      "epoch": 0.4985891729935713,
+      "grad_norm": 12.8125,
+      "learning_rate": 8.338038515491515e-06,
+      "loss": 1.7046,
+      "step": 17140
+    },
+    {
+      "epoch": 0.4991709573261192,
+      "grad_norm": 11.4375,
+      "learning_rate": 8.3360992338189e-06,
+      "loss": 1.5631,
+      "step": 17160
+    },
+    {
+      "epoch": 0.4997527416586671,
+      "grad_norm": 13.3125,
+      "learning_rate": 8.334159952146285e-06,
+      "loss": 1.585,
+      "step": 17180
+    },
+    {
+      "epoch": 0.5003345259912151,
+      "grad_norm": 13.5,
+      "learning_rate": 8.33222067047367e-06,
+      "loss": 1.7361,
+      "step": 17200
+    },
+    {
+      "epoch": 0.5009163103237629,
+      "grad_norm": 14.875,
+      "learning_rate": 8.330281388801056e-06,
+      "loss": 1.6168,
+      "step": 17220
+    },
+    {
+      "epoch": 0.5014980946563109,
+      "grad_norm": 11.6875,
+      "learning_rate": 8.32834210712844e-06,
+      "loss": 1.6471,
+      "step": 17240
+    },
+    {
+      "epoch": 0.5020798789888589,
+      "grad_norm": 12.3125,
+      "learning_rate": 8.326402825455826e-06,
+      "loss": 1.6724,
+      "step": 17260
+    },
+    {
+      "epoch": 0.5026616633214067,
+      "grad_norm": 11.1875,
+      "learning_rate": 8.32446354378321e-06,
+      "loss": 1.6533,
+      "step": 17280
+    },
+    {
+      "epoch": 0.5032434476539547,
+      "grad_norm": 10.9375,
+      "learning_rate": 8.322524262110596e-06,
+      "loss": 1.5692,
+      "step": 17300
+    },
+    {
+      "epoch": 0.5038252319865026,
+      "grad_norm": 13.75,
+      "learning_rate": 8.320584980437981e-06,
+      "loss": 1.6431,
+      "step": 17320
+    },
+    {
+      "epoch": 0.5044070163190505,
+      "grad_norm": 12.1875,
+      "learning_rate": 8.318645698765366e-06,
+      "loss": 1.6197,
+      "step": 17340
+    },
+    {
+      "epoch": 0.5049888006515985,
+      "grad_norm": 14.5,
+      "learning_rate": 8.316706417092751e-06,
+      "loss": 1.6078,
+      "step": 17360
+    },
+    {
+      "epoch": 0.5055705849841464,
+      "grad_norm": 12.75,
+      "learning_rate": 8.314767135420136e-06,
+      "loss": 1.7085,
+      "step": 17380
+    },
+    {
+      "epoch": 0.5061523693166943,
+      "grad_norm": 14.9375,
+      "learning_rate": 8.312827853747521e-06,
+      "loss": 1.5984,
+      "step": 17400
+    },
+    {
+      "epoch": 0.5067341536492422,
+      "grad_norm": 10.75,
+      "learning_rate": 8.310888572074906e-06,
+      "loss": 1.5446,
+      "step": 17420
+    },
+    {
+      "epoch": 0.5073159379817902,
+      "grad_norm": 10.6875,
+      "learning_rate": 8.308949290402292e-06,
+      "loss": 1.6235,
+      "step": 17440
+    },
+    {
+      "epoch": 0.507897722314338,
+      "grad_norm": 14.4375,
+      "learning_rate": 8.307010008729677e-06,
+      "loss": 1.6053,
+      "step": 17460
+    },
+    {
+      "epoch": 0.508479506646886,
+      "grad_norm": 12.5625,
+      "learning_rate": 8.305070727057062e-06,
+      "loss": 1.5673,
+      "step": 17480
+    },
+    {
+      "epoch": 0.509061290979434,
+      "grad_norm": 12.125,
+      "learning_rate": 8.303131445384447e-06,
+      "loss": 1.6572,
+      "step": 17500
+    },
+    {
+      "epoch": 0.5096430753119818,
+      "grad_norm": 14.0625,
+      "learning_rate": 8.301192163711832e-06,
+      "loss": 1.6572,
+      "step": 17520
+    },
+    {
+      "epoch": 0.5102248596445298,
+      "grad_norm": 12.9375,
+      "learning_rate": 8.299252882039217e-06,
+      "loss": 1.6663,
+      "step": 17540
+    },
+    {
+      "epoch": 0.5108066439770776,
+      "grad_norm": 11.8125,
+      "learning_rate": 8.297313600366602e-06,
+      "loss": 1.6452,
+      "step": 17560
+    },
+    {
+      "epoch": 0.5113884283096256,
+      "grad_norm": 14.625,
+      "learning_rate": 8.295374318693987e-06,
+      "loss": 1.6668,
+      "step": 17580
+    },
+    {
+      "epoch": 0.5119702126421736,
+      "grad_norm": 10.8125,
+      "learning_rate": 8.293435037021372e-06,
+      "loss": 1.5655,
+      "step": 17600
+    },
+    {
+      "epoch": 0.5125519969747214,
+      "grad_norm": 12.6875,
+      "learning_rate": 8.291495755348757e-06,
+      "loss": 1.761,
+      "step": 17620
+    },
+    {
+      "epoch": 0.5131337813072694,
+      "grad_norm": 12.75,
+      "learning_rate": 8.289556473676143e-06,
+      "loss": 1.6808,
+      "step": 17640
+    },
+    {
+      "epoch": 0.5137155656398174,
+      "grad_norm": 16.0,
+      "learning_rate": 8.287617192003528e-06,
+      "loss": 1.6404,
+      "step": 17660
+    },
+    {
+      "epoch": 0.5142973499723652,
+      "grad_norm": 11.4375,
+      "learning_rate": 8.285677910330913e-06,
+      "loss": 1.7102,
+      "step": 17680
+    },
+    {
+      "epoch": 0.5148791343049132,
+      "grad_norm": 10.75,
+      "learning_rate": 8.283738628658298e-06,
+      "loss": 1.5917,
+      "step": 17700
+    },
+    {
+      "epoch": 0.5154609186374611,
+      "grad_norm": 13.375,
+      "learning_rate": 8.281799346985683e-06,
+      "loss": 1.6896,
+      "step": 17720
+    },
+    {
+      "epoch": 0.516042702970009,
+      "grad_norm": 12.3125,
+      "learning_rate": 8.279860065313068e-06,
+      "loss": 1.6722,
+      "step": 17740
+    },
+    {
+      "epoch": 0.516624487302557,
+      "grad_norm": 12.75,
+      "learning_rate": 8.277920783640453e-06,
+      "loss": 1.6643,
+      "step": 17760
+    },
+    {
+      "epoch": 0.5172062716351049,
+      "grad_norm": 10.25,
+      "learning_rate": 8.275981501967838e-06,
+      "loss": 1.6292,
+      "step": 17780
+    },
+    {
+      "epoch": 0.5177880559676528,
+      "grad_norm": 13.0,
+      "learning_rate": 8.274042220295223e-06,
+      "loss": 1.664,
+      "step": 17800
+    },
+    {
+      "epoch": 0.5183698403002007,
+      "grad_norm": 12.25,
+      "learning_rate": 8.272102938622608e-06,
+      "loss": 1.558,
+      "step": 17820
+    },
+    {
+      "epoch": 0.5189516246327487,
+      "grad_norm": 13.0625,
+      "learning_rate": 8.270163656949994e-06,
+      "loss": 1.6129,
+      "step": 17840
+    },
+    {
+      "epoch": 0.5195334089652965,
+      "grad_norm": 14.3125,
+      "learning_rate": 8.268224375277379e-06,
+      "loss": 1.6074,
+      "step": 17860
+    },
+    {
+      "epoch": 0.5201151932978445,
+      "grad_norm": 9.625,
+      "learning_rate": 8.266285093604764e-06,
+      "loss": 1.6244,
+      "step": 17880
+    },
+    {
+      "epoch": 0.5206969776303925,
+      "grad_norm": 12.625,
+      "learning_rate": 8.264345811932149e-06,
+      "loss": 1.595,
+      "step": 17900
+    },
+    {
+      "epoch": 0.5212787619629403,
+      "grad_norm": 11.0625,
+      "learning_rate": 8.262406530259534e-06,
+      "loss": 1.5696,
+      "step": 17920
+    },
+    {
+      "epoch": 0.5218605462954883,
+      "grad_norm": 11.1875,
+      "learning_rate": 8.260467248586919e-06,
+      "loss": 1.6314,
+      "step": 17940
+    },
+    {
+      "epoch": 0.5224423306280361,
+      "grad_norm": 11.75,
+      "learning_rate": 8.258527966914304e-06,
+      "loss": 1.6623,
+      "step": 17960
+    },
+    {
+      "epoch": 0.5230241149605841,
+      "grad_norm": 10.5,
+      "learning_rate": 8.25658868524169e-06,
+      "loss": 1.6871,
+      "step": 17980
+    },
+    {
+      "epoch": 0.5236058992931321,
+      "grad_norm": 12.375,
+      "learning_rate": 8.254649403569074e-06,
+      "loss": 1.6504,
+      "step": 18000
+    },
+    {
+      "epoch": 0.5241876836256799,
+      "grad_norm": 12.1875,
+      "learning_rate": 8.25271012189646e-06,
+      "loss": 1.5599,
+      "step": 18020
+    },
+    {
+      "epoch": 0.5247694679582279,
+      "grad_norm": 15.25,
+      "learning_rate": 8.250770840223844e-06,
+      "loss": 1.6048,
+      "step": 18040
+    },
+    {
+      "epoch": 0.5253512522907758,
+      "grad_norm": 14.3125,
+      "learning_rate": 8.24883155855123e-06,
+      "loss": 1.6068,
+      "step": 18060
+    },
+    {
+      "epoch": 0.5259330366233237,
+      "grad_norm": 11.625,
+      "learning_rate": 8.246892276878615e-06,
+      "loss": 1.6132,
+      "step": 18080
+    },
+    {
+      "epoch": 0.5265148209558717,
+      "grad_norm": 12.875,
+      "learning_rate": 8.244952995206e-06,
+      "loss": 1.5648,
+      "step": 18100
+    },
+    {
+      "epoch": 0.5270966052884196,
+      "grad_norm": 13.1875,
+      "learning_rate": 8.243013713533385e-06,
+      "loss": 1.6553,
+      "step": 18120
+    },
+    {
+      "epoch": 0.5276783896209675,
+      "grad_norm": 11.5,
+      "learning_rate": 8.24107443186077e-06,
+      "loss": 1.5852,
+      "step": 18140
+    },
+    {
+      "epoch": 0.5282601739535154,
+      "grad_norm": 13.625,
+      "learning_rate": 8.239135150188155e-06,
+      "loss": 1.6632,
+      "step": 18160
+    },
+    {
+      "epoch": 0.5288419582860634,
+      "grad_norm": 14.5625,
+      "learning_rate": 8.23719586851554e-06,
+      "loss": 1.6971,
+      "step": 18180
+    },
+    {
+      "epoch": 0.5294237426186112,
+      "grad_norm": 17.75,
+      "learning_rate": 8.235256586842925e-06,
+      "loss": 1.7631,
+      "step": 18200
+    },
+    {
+      "epoch": 0.5300055269511592,
+      "grad_norm": 14.625,
+      "learning_rate": 8.23331730517031e-06,
+      "loss": 1.6325,
+      "step": 18220
+    },
+    {
+      "epoch": 0.5305873112837072,
+      "grad_norm": 13.3125,
+      "learning_rate": 8.231378023497695e-06,
+      "loss": 1.6736,
+      "step": 18240
+    },
+    {
+      "epoch": 0.531169095616255,
+      "grad_norm": 12.625,
+      "learning_rate": 8.22943874182508e-06,
+      "loss": 1.6393,
+      "step": 18260
+    },
+    {
+      "epoch": 0.531750879948803,
+      "grad_norm": 14.875,
+      "learning_rate": 8.227499460152466e-06,
+      "loss": 1.6378,
+      "step": 18280
+    },
+    {
+      "epoch": 0.532332664281351,
+      "grad_norm": 14.75,
+      "learning_rate": 8.22556017847985e-06,
+      "loss": 1.6205,
+      "step": 18300
+    },
+    {
+      "epoch": 0.5329144486138988,
+      "grad_norm": 12.625,
+      "learning_rate": 8.223620896807236e-06,
+      "loss": 1.6166,
+      "step": 18320
+    },
+    {
+      "epoch": 0.5334962329464468,
+      "grad_norm": 13.5625,
+      "learning_rate": 8.221681615134621e-06,
+      "loss": 1.6742,
+      "step": 18340
+    },
+    {
+      "epoch": 0.5340780172789946,
+      "grad_norm": 13.25,
+      "learning_rate": 8.219742333462006e-06,
+      "loss": 1.6505,
+      "step": 18360
+    },
+    {
+      "epoch": 0.5346598016115426,
+      "grad_norm": 12.375,
+      "learning_rate": 8.21780305178939e-06,
+      "loss": 1.6826,
+      "step": 18380
+    },
+    {
+      "epoch": 0.5352415859440905,
+      "grad_norm": 12.125,
+      "learning_rate": 8.215863770116775e-06,
+      "loss": 1.657,
+      "step": 18400
+    },
+    {
+      "epoch": 0.5358233702766384,
+      "grad_norm": 14.0625,
+      "learning_rate": 8.21392448844416e-06,
+      "loss": 1.5712,
+      "step": 18420
+    },
+    {
+      "epoch": 0.5364051546091864,
+      "grad_norm": 11.6875,
+      "learning_rate": 8.211985206771545e-06,
+      "loss": 1.7179,
+      "step": 18440
+    },
+    {
+      "epoch": 0.5369869389417343,
+      "grad_norm": 13.5625,
+      "learning_rate": 8.21004592509893e-06,
+      "loss": 1.6749,
+      "step": 18460
+    },
+    {
+      "epoch": 0.5375687232742822,
+      "grad_norm": 13.0625,
+      "learning_rate": 8.208106643426315e-06,
+      "loss": 1.6932,
+      "step": 18480
+    },
+    {
+      "epoch": 0.5381505076068301,
+      "grad_norm": 8.8125,
+      "learning_rate": 8.2061673617537e-06,
+      "loss": 1.6385,
+      "step": 18500
+    },
+    {
+      "epoch": 0.5387322919393781,
+      "grad_norm": 12.125,
+      "learning_rate": 8.204228080081085e-06,
+      "loss": 1.683,
+      "step": 18520
+    },
+    {
+      "epoch": 0.539314076271926,
+      "grad_norm": 13.125,
+      "learning_rate": 8.20228879840847e-06,
+      "loss": 1.59,
+      "step": 18540
+    },
+    {
+      "epoch": 0.5398958606044739,
+      "grad_norm": 12.4375,
+      "learning_rate": 8.200349516735855e-06,
+      "loss": 1.5987,
+      "step": 18560
+    },
+    {
+      "epoch": 0.5404776449370219,
+      "grad_norm": 12.625,
+      "learning_rate": 8.19841023506324e-06,
+      "loss": 1.63,
+      "step": 18580
+    },
+    {
+      "epoch": 0.5410594292695697,
+      "grad_norm": 11.6875,
+      "learning_rate": 8.196470953390626e-06,
+      "loss": 1.698,
+      "step": 18600
+    },
+    {
+      "epoch": 0.5416412136021177,
+      "grad_norm": 12.1875,
+      "learning_rate": 8.19453167171801e-06,
+      "loss": 1.6323,
+      "step": 18620
+    },
+    {
+      "epoch": 0.5422229979346657,
+      "grad_norm": 11.5625,
+      "learning_rate": 8.192592390045396e-06,
+      "loss": 1.6039,
+      "step": 18640
+    },
+    {
+      "epoch": 0.5428047822672135,
+      "grad_norm": 10.3125,
+      "learning_rate": 8.19065310837278e-06,
+      "loss": 1.6931,
+      "step": 18660
+    },
+    {
+      "epoch": 0.5433865665997615,
+      "grad_norm": 10.875,
+      "learning_rate": 8.188713826700166e-06,
+      "loss": 1.6977,
+      "step": 18680
+    },
+    {
+      "epoch": 0.5439683509323094,
+      "grad_norm": 9.8125,
+      "learning_rate": 8.186774545027551e-06,
+      "loss": 1.5836,
+      "step": 18700
+    },
+    {
+      "epoch": 0.5445501352648573,
+      "grad_norm": 12.6875,
+      "learning_rate": 8.184835263354936e-06,
+      "loss": 1.6545,
+      "step": 18720
+    },
+    {
+      "epoch": 0.5451319195974053,
+      "grad_norm": 10.0,
+      "learning_rate": 8.182895981682321e-06,
+      "loss": 1.5713,
+      "step": 18740
+    },
+    {
+      "epoch": 0.5457137039299532,
+      "grad_norm": 11.25,
+      "learning_rate": 8.180956700009706e-06,
+      "loss": 1.5724,
+      "step": 18760
+    },
+    {
+      "epoch": 0.5462954882625011,
+      "grad_norm": 11.5,
+      "learning_rate": 8.179017418337091e-06,
+      "loss": 1.7166,
+      "step": 18780
+    },
+    {
+      "epoch": 0.546877272595049,
+      "grad_norm": 11.9375,
+      "learning_rate": 8.177078136664477e-06,
+      "loss": 1.735,
+      "step": 18800
+    },
+    {
+      "epoch": 0.5474590569275969,
+      "grad_norm": 12.0625,
+      "learning_rate": 8.175138854991862e-06,
+      "loss": 1.6193,
+      "step": 18820
+    },
+    {
+      "epoch": 0.5480408412601449,
+      "grad_norm": 10.875,
+      "learning_rate": 8.173199573319247e-06,
+      "loss": 1.6175,
+      "step": 18840
+    },
+    {
+      "epoch": 0.5486226255926928,
+      "grad_norm": 12.0625,
+      "learning_rate": 8.171260291646632e-06,
+      "loss": 1.748,
+      "step": 18860
+    },
+    {
+      "epoch": 0.5492044099252407,
+      "grad_norm": 15.9375,
+      "learning_rate": 8.169321009974017e-06,
+      "loss": 1.66,
+      "step": 18880
+    },
+    {
+      "epoch": 0.5497861942577886,
+      "grad_norm": 10.6875,
+      "learning_rate": 8.167381728301402e-06,
+      "loss": 1.7462,
+      "step": 18900
+    },
+    {
+      "epoch": 0.5503679785903366,
+      "grad_norm": 12.125,
+      "learning_rate": 8.165442446628787e-06,
+      "loss": 1.6313,
+      "step": 18920
+    },
+    {
+      "epoch": 0.5509497629228844,
+      "grad_norm": 10.8125,
+      "learning_rate": 8.163503164956172e-06,
+      "loss": 1.6117,
+      "step": 18940
+    },
+    {
+      "epoch": 0.5515315472554324,
+      "grad_norm": 10.875,
+      "learning_rate": 8.161563883283557e-06,
+      "loss": 1.56,
+      "step": 18960
+    },
+    {
+      "epoch": 0.5521133315879804,
+      "grad_norm": 12.875,
+      "learning_rate": 8.159624601610942e-06,
+      "loss": 1.6744,
+      "step": 18980
+    },
+    {
+      "epoch": 0.5526951159205282,
+      "grad_norm": 13.25,
+      "learning_rate": 8.157685319938327e-06,
+      "loss": 1.6503,
+      "step": 19000
+    },
+    {
+      "epoch": 0.5532769002530762,
+      "grad_norm": 10.125,
+      "learning_rate": 8.155746038265713e-06,
+      "loss": 1.6612,
+      "step": 19020
+    },
+    {
+      "epoch": 0.5538586845856242,
+      "grad_norm": 14.6875,
+      "learning_rate": 8.153806756593098e-06,
+      "loss": 1.6828,
+      "step": 19040
+    },
+    {
+      "epoch": 0.554440468918172,
+      "grad_norm": 14.75,
+      "learning_rate": 8.151867474920483e-06,
+      "loss": 1.6411,
+      "step": 19060
+    },
+    {
+      "epoch": 0.55502225325072,
+      "grad_norm": 10.75,
+      "learning_rate": 8.149928193247868e-06,
+      "loss": 1.614,
+      "step": 19080
+    },
+    {
+      "epoch": 0.5556040375832679,
+      "grad_norm": 12.25,
+      "learning_rate": 8.147988911575253e-06,
+      "loss": 1.5938,
+      "step": 19100
+    },
+    {
+      "epoch": 0.5561858219158158,
+      "grad_norm": 13.0,
+      "learning_rate": 8.146049629902638e-06,
+      "loss": 1.6811,
+      "step": 19120
+    },
+    {
+      "epoch": 0.5567676062483637,
+      "grad_norm": 11.375,
+      "learning_rate": 8.144110348230023e-06,
+      "loss": 1.6226,
+      "step": 19140
+    },
+    {
+      "epoch": 0.5573493905809117,
+      "grad_norm": 10.5625,
+      "learning_rate": 8.142171066557408e-06,
+      "loss": 1.6727,
+      "step": 19160
+    },
+    {
+      "epoch": 0.5579311749134596,
+      "grad_norm": 15.3125,
+      "learning_rate": 8.140231784884793e-06,
+      "loss": 1.6248,
+      "step": 19180
+    },
+    {
+      "epoch": 0.5585129592460075,
+      "grad_norm": 10.1875,
+      "learning_rate": 8.138292503212178e-06,
+      "loss": 1.6383,
+      "step": 19200
+    },
+    {
+      "epoch": 0.5590947435785554,
+      "grad_norm": 10.4375,
+      "learning_rate": 8.136353221539564e-06,
+      "loss": 1.6951,
+      "step": 19220
+    },
+    {
+      "epoch": 0.5596765279111033,
+      "grad_norm": 12.0625,
+      "learning_rate": 8.134413939866949e-06,
+      "loss": 1.722,
+      "step": 19240
+    },
+    {
+      "epoch": 0.5602583122436513,
+      "grad_norm": 14.625,
+      "learning_rate": 8.132474658194334e-06,
+      "loss": 1.6129,
+      "step": 19260
+    },
+    {
+      "epoch": 0.5608400965761992,
+      "grad_norm": 13.0,
+      "learning_rate": 8.130535376521719e-06,
+      "loss": 1.6803,
+      "step": 19280
+    },
+    {
+      "epoch": 0.5614218809087471,
+      "grad_norm": 12.75,
+      "learning_rate": 8.128596094849104e-06,
+      "loss": 1.6305,
+      "step": 19300
+    },
+    {
+      "epoch": 0.5620036652412951,
+      "grad_norm": 13.3125,
+      "learning_rate": 8.126656813176489e-06,
+      "loss": 1.5954,
+      "step": 19320
+    },
+    {
+      "epoch": 0.5625854495738429,
+      "grad_norm": 11.375,
+      "learning_rate": 8.124717531503874e-06,
+      "loss": 1.5627,
+      "step": 19340
+    },
+    {
+      "epoch": 0.5631672339063909,
+      "grad_norm": 12.0625,
+      "learning_rate": 8.12277824983126e-06,
+      "loss": 1.6935,
+      "step": 19360
+    },
+    {
+      "epoch": 0.5637490182389389,
+      "grad_norm": 11.25,
+      "learning_rate": 8.120838968158644e-06,
+      "loss": 1.6396,
+      "step": 19380
+    },
+    {
+      "epoch": 0.5643308025714867,
+      "grad_norm": 13.0625,
+      "learning_rate": 8.11889968648603e-06,
+      "loss": 1.594,
+      "step": 19400
+    },
+    {
+      "epoch": 0.5649125869040347,
+      "grad_norm": 10.8125,
+      "learning_rate": 8.116960404813415e-06,
+      "loss": 1.6131,
+      "step": 19420
+    },
+    {
+      "epoch": 0.5654943712365826,
+      "grad_norm": 12.125,
+      "learning_rate": 8.1150211231408e-06,
+      "loss": 1.6273,
+      "step": 19440
+    },
+    {
+      "epoch": 0.5660761555691305,
+      "grad_norm": 15.8125,
+      "learning_rate": 8.113081841468185e-06,
+      "loss": 1.7437,
+      "step": 19460
+    },
+    {
+      "epoch": 0.5666579399016785,
+      "grad_norm": 11.8125,
+      "learning_rate": 8.11114255979557e-06,
+      "loss": 1.7092,
+      "step": 19480
+    },
+    {
+      "epoch": 0.5672397242342264,
+      "grad_norm": 16.75,
+      "learning_rate": 8.109203278122955e-06,
+      "loss": 1.6279,
+      "step": 19500
+    },
+    {
+      "epoch": 0.5678215085667743,
+      "grad_norm": 13.75,
+      "learning_rate": 8.10726399645034e-06,
+      "loss": 1.7001,
+      "step": 19520
+    },
+    {
+      "epoch": 0.5684032928993222,
+      "grad_norm": 13.9375,
+      "learning_rate": 8.105324714777725e-06,
+      "loss": 1.694,
+      "step": 19540
+    },
+    {
+      "epoch": 0.5689850772318702,
+      "grad_norm": 14.6875,
+      "learning_rate": 8.10338543310511e-06,
+      "loss": 1.6262,
+      "step": 19560
+    },
+    {
+      "epoch": 0.569566861564418,
+      "grad_norm": 12.75,
+      "learning_rate": 8.101446151432495e-06,
+      "loss": 1.6449,
+      "step": 19580
+    },
+    {
+      "epoch": 0.570148645896966,
+      "grad_norm": 12.9375,
+      "learning_rate": 8.09950686975988e-06,
+      "loss": 1.6502,
+      "step": 19600
+    },
+    {
+      "epoch": 0.570730430229514,
+      "grad_norm": 16.0,
+      "learning_rate": 8.097567588087266e-06,
+      "loss": 1.6979,
+      "step": 19620
+    },
+    {
+      "epoch": 0.5713122145620618,
+      "grad_norm": 14.25,
+      "learning_rate": 8.09562830641465e-06,
+      "loss": 1.6853,
+      "step": 19640
+    },
+    {
+      "epoch": 0.5718939988946098,
+      "grad_norm": 11.75,
+      "learning_rate": 8.093689024742036e-06,
+      "loss": 1.7246,
+      "step": 19660
+    },
+    {
+      "epoch": 0.5724757832271576,
+      "grad_norm": 12.8125,
+      "learning_rate": 8.09174974306942e-06,
+      "loss": 1.6622,
+      "step": 19680
+    },
+    {
+      "epoch": 0.5730575675597056,
+      "grad_norm": 13.25,
+      "learning_rate": 8.089810461396806e-06,
+      "loss": 1.6474,
+      "step": 19700
+    },
+    {
+      "epoch": 0.5736393518922536,
+      "grad_norm": 13.125,
+      "learning_rate": 8.087871179724191e-06,
+      "loss": 1.6399,
+      "step": 19720
+    },
+    {
+      "epoch": 0.5742211362248014,
+      "grad_norm": 8.6875,
+      "learning_rate": 8.085931898051576e-06,
+      "loss": 1.5597,
+      "step": 19740
+    },
+    {
+      "epoch": 0.5748029205573494,
+      "grad_norm": 13.9375,
+      "learning_rate": 8.083992616378961e-06,
+      "loss": 1.6188,
+      "step": 19760
+    },
+    {
+      "epoch": 0.5753847048898973,
+      "grad_norm": 11.3125,
+      "learning_rate": 8.082053334706346e-06,
+      "loss": 1.6458,
+      "step": 19780
+    },
+    {
+      "epoch": 0.5759664892224452,
+      "grad_norm": 14.6875,
+      "learning_rate": 8.080114053033731e-06,
+      "loss": 1.6713,
+      "step": 19800
+    },
+    {
+      "epoch": 0.5765482735549932,
+      "grad_norm": 12.9375,
+      "learning_rate": 8.078174771361116e-06,
+      "loss": 1.6076,
+      "step": 19820
+    },
+    {
+      "epoch": 0.5771300578875411,
+      "grad_norm": 12.875,
+      "learning_rate": 8.076235489688502e-06,
+      "loss": 1.6367,
+      "step": 19840
+    },
+    {
+      "epoch": 0.577711842220089,
+      "grad_norm": 13.375,
+      "learning_rate": 8.074296208015887e-06,
+      "loss": 1.6454,
+      "step": 19860
+    },
+    {
+      "epoch": 0.5782936265526369,
+      "grad_norm": 14.625,
+      "learning_rate": 8.072356926343272e-06,
+      "loss": 1.7048,
+      "step": 19880
+    },
+    {
+      "epoch": 0.5788754108851849,
+      "grad_norm": 13.9375,
+      "learning_rate": 8.070417644670657e-06,
+      "loss": 1.5943,
+      "step": 19900
+    },
+    {
+      "epoch": 0.5794571952177328,
+      "grad_norm": 14.5625,
+      "learning_rate": 8.068478362998042e-06,
+      "loss": 1.6203,
+      "step": 19920
+    },
+    {
+      "epoch": 0.5800389795502807,
+      "grad_norm": 14.4375,
+      "learning_rate": 8.066539081325427e-06,
+      "loss": 1.712,
+      "step": 19940
+    },
+    {
+      "epoch": 0.5806207638828287,
+      "grad_norm": 13.4375,
+      "learning_rate": 8.064599799652812e-06,
+      "loss": 1.6642,
+      "step": 19960
+    },
+    {
+      "epoch": 0.5812025482153765,
+      "grad_norm": 11.75,
+      "learning_rate": 8.062660517980197e-06,
+      "loss": 1.5492,
+      "step": 19980
+    },
+    {
+      "epoch": 0.5817843325479245,
+      "grad_norm": 11.75,
+      "learning_rate": 8.060721236307582e-06,
+      "loss": 1.5491,
+      "step": 20000
+    },
+    {
+      "epoch": 0.5823661168804725,
+      "grad_norm": 10.0,
+      "learning_rate": 8.058781954634966e-06,
+      "loss": 1.6416,
+      "step": 20020
+    },
+    {
+      "epoch": 0.5829479012130203,
+      "grad_norm": 13.6875,
+      "learning_rate": 8.056842672962351e-06,
+      "loss": 1.6852,
+      "step": 20040
+    },
+    {
+      "epoch": 0.5835296855455683,
+      "grad_norm": 17.0,
+      "learning_rate": 8.054903391289736e-06,
+      "loss": 1.648,
+      "step": 20060
+    },
+    {
+      "epoch": 0.5841114698781161,
+      "grad_norm": 12.9375,
+      "learning_rate": 8.052964109617121e-06,
+      "loss": 1.7016,
+      "step": 20080
+    },
+    {
+      "epoch": 0.5846932542106641,
+      "grad_norm": 13.5,
+      "learning_rate": 8.051024827944506e-06,
+      "loss": 1.6324,
+      "step": 20100
+    },
+    {
+      "epoch": 0.5852750385432121,
+      "grad_norm": 13.0,
+      "learning_rate": 8.049085546271891e-06,
+      "loss": 1.649,
+      "step": 20120
+    },
+    {
+      "epoch": 0.5858568228757599,
+      "grad_norm": 10.8125,
+      "learning_rate": 8.047146264599276e-06,
+      "loss": 1.6909,
+      "step": 20140
+    },
+    {
+      "epoch": 0.5864386072083079,
+      "grad_norm": 14.0,
+      "learning_rate": 8.045206982926661e-06,
+      "loss": 1.6395,
+      "step": 20160
+    },
+    {
+      "epoch": 0.5870203915408558,
+      "grad_norm": 12.0625,
+      "learning_rate": 8.043267701254047e-06,
+      "loss": 1.713,
+      "step": 20180
+    },
+    {
+      "epoch": 0.5876021758734037,
+      "grad_norm": 12.75,
+      "learning_rate": 8.041328419581432e-06,
+      "loss": 1.606,
+      "step": 20200
+    },
+    {
+      "epoch": 0.5881839602059517,
+      "grad_norm": 12.8125,
+      "learning_rate": 8.039389137908817e-06,
+      "loss": 1.6471,
+      "step": 20220
+    },
+    {
+      "epoch": 0.5887657445384996,
+      "grad_norm": 14.75,
+      "learning_rate": 8.037449856236202e-06,
+      "loss": 1.7117,
+      "step": 20240
+    },
+    {
+      "epoch": 0.5893475288710475,
+      "grad_norm": 12.4375,
+      "learning_rate": 8.035510574563587e-06,
+      "loss": 1.5581,
+      "step": 20260
+    },
+    {
+      "epoch": 0.5899293132035954,
+      "grad_norm": 12.0,
+      "learning_rate": 8.033571292890972e-06,
+      "loss": 1.6135,
+      "step": 20280
+    },
+    {
+      "epoch": 0.5905110975361434,
+      "grad_norm": 11.25,
+      "learning_rate": 8.031632011218357e-06,
+      "loss": 1.6054,
+      "step": 20300
+    },
+    {
+      "epoch": 0.5910928818686912,
+      "grad_norm": 13.4375,
+      "learning_rate": 8.029692729545742e-06,
+      "loss": 1.771,
+      "step": 20320
+    },
+    {
+      "epoch": 0.5916746662012392,
+      "grad_norm": 11.25,
+      "learning_rate": 8.027753447873127e-06,
+      "loss": 1.6584,
+      "step": 20340
+    },
+    {
+      "epoch": 0.5922564505337872,
+      "grad_norm": 9.875,
+      "learning_rate": 8.025814166200512e-06,
+      "loss": 1.6999,
+      "step": 20360
+    },
+    {
+      "epoch": 0.592838234866335,
+      "grad_norm": 11.4375,
+      "learning_rate": 8.023874884527898e-06,
+      "loss": 1.6357,
+      "step": 20380
+    },
+    {
+      "epoch": 0.593420019198883,
+      "grad_norm": 11.8125,
+      "learning_rate": 8.021935602855283e-06,
+      "loss": 1.5729,
+      "step": 20400
+    },
+    {
+      "epoch": 0.594001803531431,
+      "grad_norm": 13.5,
+      "learning_rate": 8.019996321182668e-06,
+      "loss": 1.6684,
+      "step": 20420
+    },
+    {
+      "epoch": 0.5945835878639788,
+      "grad_norm": 11.875,
+      "learning_rate": 8.018057039510053e-06,
+      "loss": 1.6585,
+      "step": 20440
+    },
+    {
+      "epoch": 0.5951653721965268,
+      "grad_norm": 12.625,
+      "learning_rate": 8.016117757837438e-06,
+      "loss": 1.6137,
+      "step": 20460
+    },
+    {
+      "epoch": 0.5957471565290746,
+      "grad_norm": 10.75,
+      "learning_rate": 8.014178476164823e-06,
+      "loss": 1.647,
+      "step": 20480
+    },
+    {
+      "epoch": 0.5963289408616226,
+      "grad_norm": 12.5625,
+      "learning_rate": 8.012239194492208e-06,
+      "loss": 1.6484,
+      "step": 20500
+    },
+    {
+      "epoch": 0.5969107251941705,
+      "grad_norm": 14.0625,
+      "learning_rate": 8.010299912819593e-06,
+      "loss": 1.7112,
+      "step": 20520
+    },
+    {
+      "epoch": 0.5974925095267184,
+      "grad_norm": 9.75,
+      "learning_rate": 8.008360631146978e-06,
+      "loss": 1.6521,
+      "step": 20540
+    },
+    {
+      "epoch": 0.5980742938592664,
+      "grad_norm": 11.5,
+      "learning_rate": 8.006421349474363e-06,
+      "loss": 1.6941,
+      "step": 20560
+    },
+    {
+      "epoch": 0.5986560781918143,
+      "grad_norm": 11.125,
+      "learning_rate": 8.004482067801747e-06,
+      "loss": 1.7052,
+      "step": 20580
+    },
+    {
+      "epoch": 0.5992378625243622,
+      "grad_norm": 11.5,
+      "learning_rate": 8.002542786129132e-06,
+      "loss": 1.5777,
+      "step": 20600
+    },
+    {
+      "epoch": 0.5998196468569101,
+      "grad_norm": 9.8125,
+      "learning_rate": 8.000603504456517e-06,
+      "loss": 1.7117,
+      "step": 20620
+    },
+    {
+      "epoch": 0.6004014311894581,
+      "grad_norm": 13.1875,
+      "learning_rate": 7.998664222783902e-06,
+      "loss": 1.6213,
+      "step": 20640
+    },
+    {
+      "epoch": 0.600983215522006,
+      "grad_norm": 13.375,
+      "learning_rate": 7.996724941111287e-06,
+      "loss": 1.6577,
+      "step": 20660
+    },
+    {
+      "epoch": 0.6015649998545539,
+      "grad_norm": 13.4375,
+      "learning_rate": 7.994785659438672e-06,
+      "loss": 1.6816,
+      "step": 20680
+    },
+    {
+      "epoch": 0.6021467841871019,
+      "grad_norm": 12.5,
+      "learning_rate": 7.992846377766057e-06,
+      "loss": 1.536,
+      "step": 20700
+    },
+    {
+      "epoch": 0.6027285685196497,
+      "grad_norm": 11.5,
+      "learning_rate": 7.990907096093442e-06,
+      "loss": 1.5839,
+      "step": 20720
+    },
+    {
+      "epoch": 0.6033103528521977,
+      "grad_norm": 10.625,
+      "learning_rate": 7.988967814420828e-06,
+      "loss": 1.7227,
+      "step": 20740
+    },
+    {
+      "epoch": 0.6038921371847457,
+      "grad_norm": 14.0625,
+      "learning_rate": 7.987028532748213e-06,
+      "loss": 1.6775,
+      "step": 20760
+    },
+    {
+      "epoch": 0.6044739215172935,
+      "grad_norm": 12.8125,
+      "learning_rate": 7.985089251075598e-06,
+      "loss": 1.6013,
+      "step": 20780
+    },
+    {
+      "epoch": 0.6050557058498415,
+      "grad_norm": 14.3125,
+      "learning_rate": 7.983149969402983e-06,
+      "loss": 1.5694,
+      "step": 20800
+    },
+    {
+      "epoch": 0.6056374901823894,
+      "grad_norm": 13.875,
+      "learning_rate": 7.981210687730368e-06,
+      "loss": 1.5985,
+      "step": 20820
+    },
+    {
+      "epoch": 0.6062192745149373,
+      "grad_norm": 10.375,
+      "learning_rate": 7.979271406057753e-06,
+      "loss": 1.586,
+      "step": 20840
+    },
+    {
+      "epoch": 0.6068010588474853,
+      "grad_norm": 11.25,
+      "learning_rate": 7.977332124385138e-06,
+      "loss": 1.5524,
+      "step": 20860
+    },
+    {
+      "epoch": 0.6073828431800332,
+      "grad_norm": 12.4375,
+      "learning_rate": 7.975392842712523e-06,
+      "loss": 1.5534,
+      "step": 20880
+    },
+    {
+      "epoch": 0.6079646275125811,
+      "grad_norm": 12.75,
+      "learning_rate": 7.973453561039908e-06,
+      "loss": 1.649,
+      "step": 20900
+    },
+    {
+      "epoch": 0.608546411845129,
+      "grad_norm": 11.8125,
+      "learning_rate": 7.971514279367293e-06,
+      "loss": 1.6024,
+      "step": 20920
+    },
+    {
+      "epoch": 0.6091281961776769,
+      "grad_norm": 15.875,
+      "learning_rate": 7.969574997694679e-06,
+      "loss": 1.6998,
+      "step": 20940
+    },
+    {
+      "epoch": 0.6097099805102248,
+      "grad_norm": 15.0625,
+      "learning_rate": 7.967635716022064e-06,
+      "loss": 1.6045,
+      "step": 20960
+    },
+    {
+      "epoch": 0.6102917648427728,
+      "grad_norm": 11.0625,
+      "learning_rate": 7.965696434349449e-06,
+      "loss": 1.6288,
+      "step": 20980
+    },
+    {
+      "epoch": 0.6108735491753207,
+      "grad_norm": 13.5625,
+      "learning_rate": 7.963757152676834e-06,
+      "loss": 1.5685,
+      "step": 21000
+    },
+    {
+      "epoch": 0.6114553335078686,
+      "grad_norm": 12.0625,
+      "learning_rate": 7.961817871004219e-06,
+      "loss": 1.7005,
+      "step": 21020
+    },
+    {
+      "epoch": 0.6120371178404166,
+      "grad_norm": 16.125,
+      "learning_rate": 7.959878589331604e-06,
+      "loss": 1.58,
+      "step": 21040
+    },
+    {
+      "epoch": 0.6126189021729644,
+      "grad_norm": 11.5,
+      "learning_rate": 7.957939307658989e-06,
+      "loss": 1.6479,
+      "step": 21060
+    },
+    {
+      "epoch": 0.6132006865055124,
+      "grad_norm": 12.375,
+      "learning_rate": 7.956000025986374e-06,
+      "loss": 1.5816,
+      "step": 21080
+    },
+    {
+      "epoch": 0.6137824708380604,
+      "grad_norm": 11.1875,
+      "learning_rate": 7.95406074431376e-06,
+      "loss": 1.6428,
+      "step": 21100
+    },
+    {
+      "epoch": 0.6143642551706082,
+      "grad_norm": 11.25,
+      "learning_rate": 7.952121462641144e-06,
+      "loss": 1.6375,
+      "step": 21120
+    },
+    {
+      "epoch": 0.6149460395031562,
+      "grad_norm": 12.4375,
+      "learning_rate": 7.95018218096853e-06,
+      "loss": 1.6391,
+      "step": 21140
+    },
+    {
+      "epoch": 0.6155278238357041,
+      "grad_norm": 13.875,
+      "learning_rate": 7.948242899295915e-06,
+      "loss": 1.6711,
+      "step": 21160
+    },
+    {
+      "epoch": 0.616109608168252,
+      "grad_norm": 14.875,
+      "learning_rate": 7.9463036176233e-06,
+      "loss": 1.6958,
+      "step": 21180
+    },
+    {
+      "epoch": 0.6166913925008,
+      "grad_norm": 13.125,
+      "learning_rate": 7.944364335950685e-06,
+      "loss": 1.5907,
+      "step": 21200
+    },
+    {
+      "epoch": 0.6172731768333479,
+      "grad_norm": 12.4375,
+      "learning_rate": 7.94242505427807e-06,
+      "loss": 1.6955,
+      "step": 21220
+    },
+    {
+      "epoch": 0.6178549611658958,
+      "grad_norm": 12.1875,
+      "learning_rate": 7.940485772605455e-06,
+      "loss": 1.646,
+      "step": 21240
+    },
+    {
+      "epoch": 0.6184367454984437,
+      "grad_norm": 13.1875,
+      "learning_rate": 7.93854649093284e-06,
+      "loss": 1.7521,
+      "step": 21260
+    },
+    {
+      "epoch": 0.6190185298309917,
+      "grad_norm": 12.625,
+      "learning_rate": 7.936607209260225e-06,
+      "loss": 1.6426,
+      "step": 21280
+    },
+    {
+      "epoch": 0.6196003141635396,
+      "grad_norm": 13.125,
+      "learning_rate": 7.93466792758761e-06,
+      "loss": 1.5326,
+      "step": 21300
+    },
+    {
+      "epoch": 0.6201820984960875,
+      "grad_norm": 12.5,
+      "learning_rate": 7.932728645914995e-06,
+      "loss": 1.5717,
+      "step": 21320
+    },
+    {
+      "epoch": 0.6207638828286354,
+      "grad_norm": 12.0,
+      "learning_rate": 7.93078936424238e-06,
+      "loss": 1.5635,
+      "step": 21340
+    },
+    {
+      "epoch": 0.6213456671611833,
+      "grad_norm": 11.8125,
+      "learning_rate": 7.928850082569766e-06,
+      "loss": 1.7181,
+      "step": 21360
+    },
+    {
+      "epoch": 0.6219274514937313,
+      "grad_norm": 14.3125,
+      "learning_rate": 7.92691080089715e-06,
+      "loss": 1.6385,
+      "step": 21380
+    },
+    {
+      "epoch": 0.6225092358262792,
+      "grad_norm": 13.0,
+      "learning_rate": 7.924971519224536e-06,
+      "loss": 1.6784,
+      "step": 21400
+    },
+    {
+      "epoch": 0.6230910201588271,
+      "grad_norm": 15.5,
+      "learning_rate": 7.923032237551921e-06,
+      "loss": 1.7064,
+      "step": 21420
+    },
+    {
+      "epoch": 0.6236728044913751,
+      "grad_norm": 12.125,
+      "learning_rate": 7.921092955879306e-06,
+      "loss": 1.6372,
+      "step": 21440
+    },
+    {
+      "epoch": 0.6242545888239229,
+      "grad_norm": 11.75,
+      "learning_rate": 7.919153674206691e-06,
+      "loss": 1.7176,
+      "step": 21460
+    },
+    {
+      "epoch": 0.6248363731564709,
+      "grad_norm": 11.875,
+      "learning_rate": 7.917214392534076e-06,
+      "loss": 1.6417,
+      "step": 21480
+    },
+    {
+      "epoch": 0.6254181574890189,
+      "grad_norm": 14.375,
+      "learning_rate": 7.915275110861461e-06,
+      "loss": 1.5704,
+      "step": 21500
+    },
+    {
+      "epoch": 0.6259999418215667,
+      "grad_norm": 10.0625,
+      "learning_rate": 7.913335829188846e-06,
+      "loss": 1.5801,
+      "step": 21520
+    },
+    {
+      "epoch": 0.6265817261541147,
+      "grad_norm": 11.5625,
+      "learning_rate": 7.911396547516231e-06,
+      "loss": 1.6788,
+      "step": 21540
+    },
+    {
+      "epoch": 0.6271635104866626,
+      "grad_norm": 12.625,
+      "learning_rate": 7.909457265843617e-06,
+      "loss": 1.6491,
+      "step": 21560
+    },
+    {
+      "epoch": 0.6277452948192105,
+      "grad_norm": 13.3125,
+      "learning_rate": 7.907517984171002e-06,
+      "loss": 1.6589,
+      "step": 21580
+    },
+    {
+      "epoch": 0.6283270791517584,
+      "grad_norm": 11.9375,
+      "learning_rate": 7.905578702498387e-06,
+      "loss": 1.6245,
+      "step": 21600
+    },
+    {
+      "epoch": 0.6289088634843064,
+      "grad_norm": 11.5,
+      "learning_rate": 7.903639420825772e-06,
+      "loss": 1.6734,
+      "step": 21620
+    },
+    {
+      "epoch": 0.6294906478168543,
+      "grad_norm": 14.6875,
+      "learning_rate": 7.901700139153157e-06,
+      "loss": 1.6331,
+      "step": 21640
+    },
+    {
+      "epoch": 0.6300724321494022,
+      "grad_norm": 9.5625,
+      "learning_rate": 7.899760857480542e-06,
+      "loss": 1.6249,
+      "step": 21660
+    },
+    {
+      "epoch": 0.6306542164819502,
+      "grad_norm": 10.1875,
+      "learning_rate": 7.897821575807927e-06,
+      "loss": 1.6278,
+      "step": 21680
+    },
+    {
+      "epoch": 0.631236000814498,
+      "grad_norm": 13.5625,
+      "learning_rate": 7.895882294135312e-06,
+      "loss": 1.689,
+      "step": 21700
+    },
+    {
+      "epoch": 0.631817785147046,
+      "grad_norm": 11.9375,
+      "learning_rate": 7.893943012462697e-06,
+      "loss": 1.6893,
+      "step": 21720
+    },
+    {
+      "epoch": 0.6323995694795939,
+      "grad_norm": 11.5625,
+      "learning_rate": 7.892003730790082e-06,
+      "loss": 1.6971,
+      "step": 21740
+    },
+    {
+      "epoch": 0.6329813538121418,
+      "grad_norm": 11.6875,
+      "learning_rate": 7.890064449117468e-06,
+      "loss": 1.5925,
+      "step": 21760
+    },
+    {
+      "epoch": 0.6335631381446898,
+      "grad_norm": 12.125,
+      "learning_rate": 7.888125167444853e-06,
+      "loss": 1.6094,
+      "step": 21780
+    },
+    {
+      "epoch": 0.6341449224772376,
+      "grad_norm": 11.4375,
+      "learning_rate": 7.886185885772238e-06,
+      "loss": 1.6971,
+      "step": 21800
+    },
+    {
+      "epoch": 0.6347267068097856,
+      "grad_norm": 12.8125,
+      "learning_rate": 7.884246604099623e-06,
+      "loss": 1.6558,
+      "step": 21820
+    },
+    {
+      "epoch": 0.6353084911423336,
+      "grad_norm": 13.6875,
+      "learning_rate": 7.882307322427008e-06,
+      "loss": 1.6819,
+      "step": 21840
+    },
+    {
+      "epoch": 0.6358902754748814,
+      "grad_norm": 13.5625,
+      "learning_rate": 7.880368040754393e-06,
+      "loss": 1.5455,
+      "step": 21860
+    },
+    {
+      "epoch": 0.6364720598074294,
+      "grad_norm": 12.1875,
+      "learning_rate": 7.878428759081778e-06,
+      "loss": 1.6307,
+      "step": 21880
+    },
+    {
+      "epoch": 0.6370538441399773,
+      "grad_norm": 11.5625,
+      "learning_rate": 7.876489477409163e-06,
+      "loss": 1.6101,
+      "step": 21900
+    },
+    {
+      "epoch": 0.6376356284725252,
+      "grad_norm": 12.6875,
+      "learning_rate": 7.874550195736548e-06,
+      "loss": 1.5181,
+      "step": 21920
+    },
+    {
+      "epoch": 0.6382174128050732,
+      "grad_norm": 12.4375,
+      "learning_rate": 7.872610914063933e-06,
+      "loss": 1.6125,
+      "step": 21940
+    },
+    {
+      "epoch": 0.6387991971376211,
+      "grad_norm": 13.1875,
+      "learning_rate": 7.870671632391319e-06,
+      "loss": 1.6586,
+      "step": 21960
+    },
+    {
+      "epoch": 0.639380981470169,
+      "grad_norm": 12.375,
+      "learning_rate": 7.868732350718704e-06,
+      "loss": 1.6246,
+      "step": 21980
+    },
+    {
+      "epoch": 0.6399627658027169,
+      "grad_norm": 11.875,
+      "learning_rate": 7.866793069046089e-06,
+      "loss": 1.5835,
+      "step": 22000
+    },
+    {
+      "epoch": 0.6405445501352649,
+      "grad_norm": 12.5,
+      "learning_rate": 7.864853787373474e-06,
+      "loss": 1.6767,
+      "step": 22020
+    },
+    {
+      "epoch": 0.6411263344678128,
+      "grad_norm": 12.25,
+      "learning_rate": 7.862914505700859e-06,
+      "loss": 1.5982,
+      "step": 22040
+    },
+    {
+      "epoch": 0.6417081188003607,
+      "grad_norm": 10.1875,
+      "learning_rate": 7.860975224028244e-06,
+      "loss": 1.6858,
+      "step": 22060
+    },
+    {
+      "epoch": 0.6422899031329087,
+      "grad_norm": 13.75,
+      "learning_rate": 7.859035942355629e-06,
+      "loss": 1.6265,
+      "step": 22080
+    },
+    {
+      "epoch": 0.6428716874654565,
+      "grad_norm": 13.5625,
+      "learning_rate": 7.857096660683014e-06,
+      "loss": 1.6599,
+      "step": 22100
+    },
+    {
+      "epoch": 0.6434534717980045,
+      "grad_norm": 12.3125,
+      "learning_rate": 7.8551573790104e-06,
+      "loss": 1.6791,
+      "step": 22120
+    },
+    {
+      "epoch": 0.6440352561305525,
+      "grad_norm": 13.625,
+      "learning_rate": 7.853218097337784e-06,
+      "loss": 1.6365,
+      "step": 22140
+    },
+    {
+      "epoch": 0.6446170404631003,
+      "grad_norm": 13.8125,
+      "learning_rate": 7.85127881566517e-06,
+      "loss": 1.6622,
+      "step": 22160
+    },
+    {
+      "epoch": 0.6451988247956483,
+      "grad_norm": 12.625,
+      "learning_rate": 7.849339533992555e-06,
+      "loss": 1.6112,
+      "step": 22180
+    },
+    {
+      "epoch": 0.6457806091281961,
+      "grad_norm": 12.0625,
+      "learning_rate": 7.84740025231994e-06,
+      "loss": 1.7412,
+      "step": 22200
+    },
+    {
+      "epoch": 0.6463623934607441,
+      "grad_norm": 12.5625,
+      "learning_rate": 7.845460970647323e-06,
+      "loss": 1.6402,
+      "step": 22220
+    },
+    {
+      "epoch": 0.646944177793292,
+      "grad_norm": 13.5625,
+      "learning_rate": 7.843521688974708e-06,
+      "loss": 1.6437,
+      "step": 22240
+    },
+    {
+      "epoch": 0.6475259621258399,
+      "grad_norm": 10.9375,
+      "learning_rate": 7.841582407302093e-06,
+      "loss": 1.5291,
+      "step": 22260
+    },
+    {
+      "epoch": 0.6481077464583879,
+      "grad_norm": 13.125,
+      "learning_rate": 7.839643125629478e-06,
+      "loss": 1.6853,
+      "step": 22280
+    },
+    {
+      "epoch": 0.6486895307909358,
+      "grad_norm": 11.375,
+      "learning_rate": 7.837703843956863e-06,
+      "loss": 1.6956,
+      "step": 22300
+    },
+    {
+      "epoch": 0.6492713151234837,
+      "grad_norm": 15.875,
+      "learning_rate": 7.835764562284249e-06,
+      "loss": 1.5831,
+      "step": 22320
+    },
+    {
+      "epoch": 0.6498530994560316,
+      "grad_norm": 13.1875,
+      "learning_rate": 7.833825280611634e-06,
+      "loss": 1.6551,
+      "step": 22340
+    },
+    {
+      "epoch": 0.6504348837885796,
+      "grad_norm": 11.875,
+      "learning_rate": 7.831885998939019e-06,
+      "loss": 1.5802,
+      "step": 22360
+    },
+    {
+      "epoch": 0.6510166681211275,
+      "grad_norm": 13.125,
+      "learning_rate": 7.829946717266404e-06,
+      "loss": 1.6445,
+      "step": 22380
+    },
+    {
+      "epoch": 0.6515984524536754,
+      "grad_norm": 12.3125,
+      "learning_rate": 7.828007435593789e-06,
+      "loss": 1.624,
+      "step": 22400
+    },
+    {
+      "epoch": 0.6521802367862234,
+      "grad_norm": 10.5,
+      "learning_rate": 7.826068153921174e-06,
+      "loss": 1.5483,
+      "step": 22420
+    },
+    {
+      "epoch": 0.6527620211187712,
+      "grad_norm": 11.5,
+      "learning_rate": 7.82412887224856e-06,
+      "loss": 1.6606,
+      "step": 22440
+    },
+    {
+      "epoch": 0.6533438054513192,
+      "grad_norm": 13.0625,
+      "learning_rate": 7.822189590575944e-06,
+      "loss": 1.6794,
+      "step": 22460
+    },
+    {
+      "epoch": 0.6539255897838672,
+      "grad_norm": 12.5625,
+      "learning_rate": 7.82025030890333e-06,
+      "loss": 1.6568,
+      "step": 22480
+    },
+    {
+      "epoch": 0.654507374116415,
+      "grad_norm": 12.0,
+      "learning_rate": 7.818311027230714e-06,
+      "loss": 1.6155,
+      "step": 22500
+    },
+    {
+      "epoch": 0.655089158448963,
+      "grad_norm": 10.875,
+      "learning_rate": 7.8163717455581e-06,
+      "loss": 1.5812,
+      "step": 22520
+    },
+    {
+      "epoch": 0.655670942781511,
+      "grad_norm": 10.125,
+      "learning_rate": 7.814432463885485e-06,
+      "loss": 1.7043,
+      "step": 22540
+    },
+    {
+      "epoch": 0.6562527271140588,
+      "grad_norm": 10.8125,
+      "learning_rate": 7.81249318221287e-06,
+      "loss": 1.6885,
+      "step": 22560
+    },
+    {
+      "epoch": 0.6568345114466068,
+      "grad_norm": 13.5625,
+      "learning_rate": 7.810553900540255e-06,
+      "loss": 1.6724,
+      "step": 22580
+    },
+    {
+      "epoch": 0.6574162957791546,
+      "grad_norm": 14.3125,
+      "learning_rate": 7.80861461886764e-06,
+      "loss": 1.6988,
+      "step": 22600
+    },
+    {
+      "epoch": 0.6579980801117026,
+      "grad_norm": 14.0625,
+      "learning_rate": 7.806675337195025e-06,
+      "loss": 1.5984,
+      "step": 22620
+    },
+    {
+      "epoch": 0.6585798644442505,
+      "grad_norm": 13.6875,
+      "learning_rate": 7.80473605552241e-06,
+      "loss": 1.625,
+      "step": 22640
+    },
+    {
+      "epoch": 0.6591616487767984,
+      "grad_norm": 11.5625,
+      "learning_rate": 7.802796773849795e-06,
+      "loss": 1.6485,
+      "step": 22660
+    },
+    {
+      "epoch": 0.6597434331093464,
+      "grad_norm": 11.625,
+      "learning_rate": 7.80085749217718e-06,
+      "loss": 1.6211,
+      "step": 22680
+    },
+    {
+      "epoch": 0.6603252174418943,
+      "grad_norm": 11.9375,
+      "learning_rate": 7.798918210504565e-06,
+      "loss": 1.6608,
+      "step": 22700
+    },
+    {
+      "epoch": 0.6609070017744422,
+      "grad_norm": 12.0,
+      "learning_rate": 7.79697892883195e-06,
+      "loss": 1.7853,
+      "step": 22720
+    },
+    {
+      "epoch": 0.6614887861069901,
+      "grad_norm": 12.9375,
+      "learning_rate": 7.795039647159336e-06,
+      "loss": 1.6535,
+      "step": 22740
+    },
+    {
+      "epoch": 0.6620705704395381,
+      "grad_norm": 11.9375,
+      "learning_rate": 7.79310036548672e-06,
+      "loss": 1.6026,
+      "step": 22760
+    },
+    {
+      "epoch": 0.662652354772086,
+      "grad_norm": 11.5,
+      "learning_rate": 7.791161083814106e-06,
+      "loss": 1.6255,
+      "step": 22780
+    },
+    {
+      "epoch": 0.6632341391046339,
+      "grad_norm": 11.875,
+      "learning_rate": 7.789221802141491e-06,
+      "loss": 1.6588,
+      "step": 22800
+    },
+    {
+      "epoch": 0.6638159234371819,
+      "grad_norm": 11.875,
+      "learning_rate": 7.787282520468876e-06,
+      "loss": 1.5569,
+      "step": 22820
+    },
+    {
+      "epoch": 0.6643977077697297,
+      "grad_norm": 12.125,
+      "learning_rate": 7.785343238796261e-06,
+      "loss": 1.6181,
+      "step": 22840
+    },
+    {
+      "epoch": 0.6649794921022777,
+      "grad_norm": 15.0,
+      "learning_rate": 7.783403957123646e-06,
+      "loss": 1.5772,
+      "step": 22860
+    },
+    {
+      "epoch": 0.6655612764348257,
+      "grad_norm": 11.5,
+      "learning_rate": 7.781464675451031e-06,
+      "loss": 1.5771,
+      "step": 22880
+    },
+    {
+      "epoch": 0.6661430607673735,
+      "grad_norm": 13.8125,
+      "learning_rate": 7.779525393778416e-06,
+      "loss": 1.704,
+      "step": 22900
+    },
+    {
+      "epoch": 0.6667248450999215,
+      "grad_norm": 13.125,
+      "learning_rate": 7.777586112105802e-06,
+      "loss": 1.6354,
+      "step": 22920
+    },
+    {
+      "epoch": 0.6673066294324694,
+      "grad_norm": 12.0625,
+      "learning_rate": 7.775646830433187e-06,
+      "loss": 1.6523,
+      "step": 22940
+    },
+    {
+      "epoch": 0.6678884137650173,
+      "grad_norm": 12.5625,
+      "learning_rate": 7.773707548760572e-06,
+      "loss": 1.6739,
+      "step": 22960
+    },
+    {
+      "epoch": 0.6684701980975652,
+      "grad_norm": 11.0,
+      "learning_rate": 7.771768267087957e-06,
+      "loss": 1.5769,
+      "step": 22980
+    },
+    {
+      "epoch": 0.6690519824301132,
+      "grad_norm": 10.8125,
+      "learning_rate": 7.769828985415342e-06,
+      "loss": 1.6732,
+      "step": 23000
+    },
+    {
+      "epoch": 0.6696337667626611,
+      "grad_norm": 12.9375,
+      "learning_rate": 7.767889703742727e-06,
+      "loss": 1.601,
+      "step": 23020
+    },
+    {
+      "epoch": 0.670215551095209,
+      "grad_norm": 19.125,
+      "learning_rate": 7.765950422070112e-06,
+      "loss": 1.7081,
+      "step": 23040
+    },
+    {
+      "epoch": 0.6707973354277569,
+      "grad_norm": 10.0625,
+      "learning_rate": 7.764011140397497e-06,
+      "loss": 1.5929,
+      "step": 23060
+    },
+    {
+      "epoch": 0.6713791197603048,
+      "grad_norm": 11.6875,
+      "learning_rate": 7.762071858724882e-06,
+      "loss": 1.6029,
+      "step": 23080
+    },
+    {
+      "epoch": 0.6719609040928528,
+      "grad_norm": 11.9375,
+      "learning_rate": 7.760132577052267e-06,
+      "loss": 1.5392,
+      "step": 23100
+    },
+    {
+      "epoch": 0.6725426884254007,
+      "grad_norm": 12.125,
+      "learning_rate": 7.758193295379652e-06,
+      "loss": 1.5838,
+      "step": 23120
+    },
+    {
+      "epoch": 0.6731244727579486,
+      "grad_norm": 12.125,
+      "learning_rate": 7.756254013707038e-06,
+      "loss": 1.6709,
+      "step": 23140
+    },
+    {
+      "epoch": 0.6737062570904966,
+      "grad_norm": 11.0625,
+      "learning_rate": 7.754314732034423e-06,
+      "loss": 1.6044,
+      "step": 23160
+    },
+    {
+      "epoch": 0.6742880414230444,
+      "grad_norm": 14.4375,
+      "learning_rate": 7.752375450361808e-06,
+      "loss": 1.6394,
+      "step": 23180
+    },
+    {
+      "epoch": 0.6748698257555924,
+      "grad_norm": 12.4375,
+      "learning_rate": 7.750436168689193e-06,
+      "loss": 1.6326,
+      "step": 23200
+    },
+    {
+      "epoch": 0.6754516100881404,
+      "grad_norm": 14.1875,
+      "learning_rate": 7.748496887016578e-06,
+      "loss": 1.6891,
+      "step": 23220
+    },
+    {
+      "epoch": 0.6760333944206882,
+      "grad_norm": 12.25,
+      "learning_rate": 7.746557605343963e-06,
+      "loss": 1.6324,
+      "step": 23240
+    },
+    {
+      "epoch": 0.6766151787532362,
+      "grad_norm": 13.25,
+      "learning_rate": 7.744618323671348e-06,
+      "loss": 1.6257,
+      "step": 23260
+    },
+    {
+      "epoch": 0.6771969630857841,
+      "grad_norm": 16.0,
+      "learning_rate": 7.742679041998733e-06,
+      "loss": 1.6609,
+      "step": 23280
+    },
+    {
+      "epoch": 0.677778747418332,
+      "grad_norm": 12.875,
+      "learning_rate": 7.740739760326118e-06,
+      "loss": 1.6568,
+      "step": 23300
+    },
+    {
+      "epoch": 0.67836053175088,
+      "grad_norm": 12.25,
+      "learning_rate": 7.738800478653503e-06,
+      "loss": 1.6479,
+      "step": 23320
+    },
+    {
+      "epoch": 0.6789423160834279,
+      "grad_norm": 13.125,
+      "learning_rate": 7.736861196980889e-06,
+      "loss": 1.6478,
+      "step": 23340
+    },
+    {
+      "epoch": 0.6795241004159758,
+      "grad_norm": 14.375,
+      "learning_rate": 7.734921915308274e-06,
+      "loss": 1.6291,
+      "step": 23360
+    },
+    {
+      "epoch": 0.6801058847485237,
+      "grad_norm": 12.3125,
+      "learning_rate": 7.732982633635659e-06,
+      "loss": 1.5804,
+      "step": 23380
+    },
+    {
+      "epoch": 0.6806876690810717,
+      "grad_norm": 9.875,
+      "learning_rate": 7.731043351963044e-06,
+      "loss": 1.6415,
+      "step": 23400
+    },
+    {
+      "epoch": 0.6812694534136196,
+      "grad_norm": 14.8125,
+      "learning_rate": 7.729104070290429e-06,
+      "loss": 1.5866,
+      "step": 23420
+    },
+    {
+      "epoch": 0.6818512377461675,
+      "grad_norm": 13.25,
+      "learning_rate": 7.727164788617814e-06,
+      "loss": 1.583,
+      "step": 23440
+    },
+    {
+      "epoch": 0.6824330220787154,
+      "grad_norm": 13.6875,
+      "learning_rate": 7.725225506945199e-06,
+      "loss": 1.6845,
+      "step": 23460
+    },
+    {
+      "epoch": 0.6830148064112633,
+      "grad_norm": 11.75,
+      "learning_rate": 7.723286225272584e-06,
+      "loss": 1.7366,
+      "step": 23480
+    },
+    {
+      "epoch": 0.6835965907438113,
+      "grad_norm": 8.6875,
+      "learning_rate": 7.72134694359997e-06,
+      "loss": 1.6201,
+      "step": 23500
+    },
+    {
+      "epoch": 0.6841783750763591,
+      "grad_norm": 13.3125,
+      "learning_rate": 7.719407661927354e-06,
+      "loss": 1.613,
+      "step": 23520
+    },
+    {
+      "epoch": 0.6847601594089071,
+      "grad_norm": 12.375,
+      "learning_rate": 7.71746838025474e-06,
+      "loss": 1.6545,
+      "step": 23540
+    },
+    {
+      "epoch": 0.6853419437414551,
+      "grad_norm": 10.875,
+      "learning_rate": 7.715529098582125e-06,
+      "loss": 1.5382,
+      "step": 23560
+    },
+    {
+      "epoch": 0.6859237280740029,
+      "grad_norm": 14.25,
+      "learning_rate": 7.71358981690951e-06,
+      "loss": 1.6518,
+      "step": 23580
+    },
+    {
+      "epoch": 0.6865055124065509,
+      "grad_norm": 12.1875,
+      "learning_rate": 7.711650535236895e-06,
+      "loss": 1.6836,
+      "step": 23600
+    },
+    {
+      "epoch": 0.6870872967390989,
+      "grad_norm": 9.25,
+      "learning_rate": 7.70971125356428e-06,
+      "loss": 1.6406,
+      "step": 23620
+    },
+    {
+      "epoch": 0.6876690810716467,
+      "grad_norm": 11.6875,
+      "learning_rate": 7.707771971891665e-06,
+      "loss": 1.6658,
+      "step": 23640
+    },
+    {
+      "epoch": 0.6882508654041947,
+      "grad_norm": 14.6875,
+      "learning_rate": 7.70583269021905e-06,
+      "loss": 1.67,
+      "step": 23660
+    },
+    {
+      "epoch": 0.6888326497367426,
+      "grad_norm": 12.25,
+      "learning_rate": 7.703893408546435e-06,
+      "loss": 1.5416,
+      "step": 23680
+    },
+    {
+      "epoch": 0.6894144340692905,
+      "grad_norm": 13.3125,
+      "learning_rate": 7.70195412687382e-06,
+      "loss": 1.6719,
+      "step": 23700
+    },
+    {
+      "epoch": 0.6899962184018384,
+      "grad_norm": 13.0625,
+      "learning_rate": 7.700014845201205e-06,
+      "loss": 1.6852,
+      "step": 23720
+    },
+    {
+      "epoch": 0.6905780027343864,
+      "grad_norm": 15.0,
+      "learning_rate": 7.69807556352859e-06,
+      "loss": 1.5557,
+      "step": 23740
+    },
+    {
+      "epoch": 0.6911597870669343,
+      "grad_norm": 13.9375,
+      "learning_rate": 7.696136281855976e-06,
+      "loss": 1.6628,
+      "step": 23760
+    },
+    {
+      "epoch": 0.6917415713994822,
+      "grad_norm": 11.125,
+      "learning_rate": 7.69419700018336e-06,
+      "loss": 1.6792,
+      "step": 23780
+    },
+    {
+      "epoch": 0.6923233557320302,
+      "grad_norm": 12.75,
+      "learning_rate": 7.692257718510746e-06,
+      "loss": 1.6234,
+      "step": 23800
+    },
+    {
+      "epoch": 0.692905140064578,
+      "grad_norm": 16.5,
+      "learning_rate": 7.690318436838131e-06,
+      "loss": 1.5968,
+      "step": 23820
+    },
+    {
+      "epoch": 0.693486924397126,
+      "grad_norm": 13.5,
+      "learning_rate": 7.688379155165514e-06,
+      "loss": 1.658,
+      "step": 23840
+    },
+    {
+      "epoch": 0.6940687087296739,
+      "grad_norm": 11.1875,
+      "learning_rate": 7.6864398734929e-06,
+      "loss": 1.6739,
+      "step": 23860
+    },
+    {
+      "epoch": 0.6946504930622218,
+      "grad_norm": 15.125,
+      "learning_rate": 7.684500591820284e-06,
+      "loss": 1.6477,
+      "step": 23880
+    },
+    {
+      "epoch": 0.6952322773947698,
+      "grad_norm": 12.3125,
+      "learning_rate": 7.68256131014767e-06,
+      "loss": 1.5933,
+      "step": 23900
+    },
+    {
+      "epoch": 0.6958140617273176,
+      "grad_norm": 12.625,
+      "learning_rate": 7.680622028475055e-06,
+      "loss": 1.642,
+      "step": 23920
+    },
+    {
+      "epoch": 0.6963958460598656,
+      "grad_norm": 13.375,
+      "learning_rate": 7.67868274680244e-06,
+      "loss": 1.6607,
+      "step": 23940
+    },
+    {
+      "epoch": 0.6969776303924136,
+      "grad_norm": 15.625,
+      "learning_rate": 7.676743465129825e-06,
+      "loss": 1.6606,
+      "step": 23960
+    },
+    {
+      "epoch": 0.6975594147249614,
+      "grad_norm": 13.625,
+      "learning_rate": 7.67480418345721e-06,
+      "loss": 1.6539,
+      "step": 23980
+    },
+    {
+      "epoch": 0.6981411990575094,
+      "grad_norm": 14.0,
+      "learning_rate": 7.672864901784595e-06,
+      "loss": 1.5874,
+      "step": 24000
+    },
+    {
+      "epoch": 0.6987229833900573,
+      "grad_norm": 9.625,
+      "learning_rate": 7.67092562011198e-06,
+      "loss": 1.5782,
+      "step": 24020
+    },
+    {
+      "epoch": 0.6993047677226052,
+      "grad_norm": 15.625,
+      "learning_rate": 7.668986338439365e-06,
+      "loss": 1.6254,
+      "step": 24040
+    },
+    {
+      "epoch": 0.6998865520551532,
+      "grad_norm": 12.25,
+      "learning_rate": 7.66704705676675e-06,
+      "loss": 1.671,
+      "step": 24060
+    },
+    {
+      "epoch": 0.7004683363877011,
+      "grad_norm": 10.3125,
+      "learning_rate": 7.665107775094135e-06,
+      "loss": 1.6397,
+      "step": 24080
+    },
+    {
+      "epoch": 0.701050120720249,
+      "grad_norm": 11.0,
+      "learning_rate": 7.66316849342152e-06,
+      "loss": 1.741,
+      "step": 24100
+    },
+    {
+      "epoch": 0.7016319050527969,
+      "grad_norm": 12.0,
+      "learning_rate": 7.661229211748906e-06,
+      "loss": 1.5878,
+      "step": 24120
+    },
+    {
+      "epoch": 0.7022136893853449,
+      "grad_norm": 8.375,
+      "learning_rate": 7.65928993007629e-06,
+      "loss": 1.6567,
+      "step": 24140
+    },
+    {
+      "epoch": 0.7027954737178927,
+      "grad_norm": 11.8125,
+      "learning_rate": 7.657350648403676e-06,
+      "loss": 1.56,
+      "step": 24160
+    },
+    {
+      "epoch": 0.7033772580504407,
+      "grad_norm": 12.4375,
+      "learning_rate": 7.655411366731061e-06,
+      "loss": 1.6866,
+      "step": 24180
+    },
+    {
+      "epoch": 0.7039590423829887,
+      "grad_norm": 14.0625,
+      "learning_rate": 7.653472085058446e-06,
+      "loss": 1.7478,
+      "step": 24200
+    },
+    {
+      "epoch": 0.7045408267155365,
+      "grad_norm": 9.9375,
+      "learning_rate": 7.651532803385831e-06,
+      "loss": 1.6249,
+      "step": 24220
+    },
+    {
+      "epoch": 0.7051226110480845,
+      "grad_norm": 12.25,
+      "learning_rate": 7.649593521713216e-06,
+      "loss": 1.6672,
+      "step": 24240
+    },
+    {
+      "epoch": 0.7057043953806325,
+      "grad_norm": 9.8125,
+      "learning_rate": 7.647654240040601e-06,
+      "loss": 1.6609,
+      "step": 24260
+    },
+    {
+      "epoch": 0.7062861797131803,
+      "grad_norm": 12.6875,
+      "learning_rate": 7.645714958367986e-06,
+      "loss": 1.6397,
+      "step": 24280
+    },
+    {
+      "epoch": 0.7068679640457283,
+      "grad_norm": 10.75,
+      "learning_rate": 7.643775676695372e-06,
+      "loss": 1.5696,
+      "step": 24300
+    },
+    {
+      "epoch": 0.7074497483782761,
+      "grad_norm": 7.09375,
+      "learning_rate": 7.641836395022757e-06,
+      "loss": 1.6747,
+      "step": 24320
+    },
+    {
+      "epoch": 0.7080315327108241,
+      "grad_norm": 12.75,
+      "learning_rate": 7.639897113350142e-06,
+      "loss": 1.5875,
+      "step": 24340
+    },
+    {
+      "epoch": 0.708613317043372,
+      "grad_norm": 13.125,
+      "learning_rate": 7.637957831677527e-06,
+      "loss": 1.6379,
+      "step": 24360
+    },
+    {
+      "epoch": 0.7091951013759199,
+      "grad_norm": 10.125,
+      "learning_rate": 7.636018550004912e-06,
+      "loss": 1.6031,
+      "step": 24380
+    },
+    {
+      "epoch": 0.7097768857084679,
+      "grad_norm": 10.375,
+      "learning_rate": 7.634079268332297e-06,
+      "loss": 1.6378,
+      "step": 24400
+    },
+    {
+      "epoch": 0.7103586700410158,
+      "grad_norm": 12.8125,
+      "learning_rate": 7.632139986659682e-06,
+      "loss": 1.703,
+      "step": 24420
+    },
+    {
+      "epoch": 0.7109404543735637,
+      "grad_norm": 13.625,
+      "learning_rate": 7.630200704987067e-06,
+      "loss": 1.5996,
+      "step": 24440
+    },
+    {
+      "epoch": 0.7115222387061116,
+      "grad_norm": 10.3125,
+      "learning_rate": 7.628261423314452e-06,
+      "loss": 1.6742,
+      "step": 24460
+    },
+    {
+      "epoch": 0.7121040230386596,
+      "grad_norm": 11.3125,
+      "learning_rate": 7.626322141641837e-06,
+      "loss": 1.6109,
+      "step": 24480
+    },
+    {
+      "epoch": 0.7126858073712075,
+      "grad_norm": 14.125,
+      "learning_rate": 7.6243828599692225e-06,
+      "loss": 1.6273,
+      "step": 24500
+    },
+    {
+      "epoch": 0.7132675917037554,
+      "grad_norm": 11.875,
+      "learning_rate": 7.622443578296608e-06,
+      "loss": 1.5985,
+      "step": 24520
+    },
+    {
+      "epoch": 0.7138493760363034,
+      "grad_norm": 12.5,
+      "learning_rate": 7.620504296623993e-06,
+      "loss": 1.5652,
+      "step": 24540
+    },
+    {
+      "epoch": 0.7144311603688512,
+      "grad_norm": 13.3125,
+      "learning_rate": 7.618565014951378e-06,
+      "loss": 1.604,
+      "step": 24560
+    },
+    {
+      "epoch": 0.7150129447013992,
+      "grad_norm": 13.125,
+      "learning_rate": 7.616625733278763e-06,
+      "loss": 1.5017,
+      "step": 24580
+    },
+    {
+      "epoch": 0.7155947290339472,
+      "grad_norm": 14.0,
+      "learning_rate": 7.614686451606148e-06,
+      "loss": 1.7284,
+      "step": 24600
+    },
+    {
+      "epoch": 0.716176513366495,
+      "grad_norm": 11.3125,
+      "learning_rate": 7.612747169933533e-06,
+      "loss": 1.6873,
+      "step": 24620
+    },
+    {
+      "epoch": 0.716758297699043,
+      "grad_norm": 14.1875,
+      "learning_rate": 7.610807888260918e-06,
+      "loss": 1.7274,
+      "step": 24640
+    },
+    {
+      "epoch": 0.7173400820315909,
+      "grad_norm": 12.875,
+      "learning_rate": 7.608868606588303e-06,
+      "loss": 1.5654,
+      "step": 24660
+    },
+    {
+      "epoch": 0.7179218663641388,
+      "grad_norm": 10.8125,
+      "learning_rate": 7.606929324915688e-06,
+      "loss": 1.5104,
+      "step": 24680
+    },
+    {
+      "epoch": 0.7185036506966868,
+      "grad_norm": 12.3125,
+      "learning_rate": 7.6049900432430735e-06,
+      "loss": 1.6304,
+      "step": 24700
+    },
+    {
+      "epoch": 0.7190854350292346,
+      "grad_norm": 13.625,
+      "learning_rate": 7.603050761570459e-06,
+      "loss": 1.5513,
+      "step": 24720
+    },
+    {
+      "epoch": 0.7196672193617826,
+      "grad_norm": 11.25,
+      "learning_rate": 7.601111479897844e-06,
+      "loss": 1.6296,
+      "step": 24740
+    },
+    {
+      "epoch": 0.7202490036943305,
+      "grad_norm": 11.375,
+      "learning_rate": 7.599172198225229e-06,
+      "loss": 1.6554,
+      "step": 24760
+    },
+    {
+      "epoch": 0.7208307880268784,
+      "grad_norm": 13.4375,
+      "learning_rate": 7.597232916552614e-06,
+      "loss": 1.689,
+      "step": 24780
+    },
+    {
+      "epoch": 0.7214125723594264,
+      "grad_norm": 13.0625,
+      "learning_rate": 7.595293634879999e-06,
+      "loss": 1.6671,
+      "step": 24800
+    },
+    {
+      "epoch": 0.7219943566919743,
+      "grad_norm": 13.25,
+      "learning_rate": 7.593354353207382e-06,
+      "loss": 1.6171,
+      "step": 24820
+    },
+    {
+      "epoch": 0.7225761410245222,
+      "grad_norm": 13.375,
+      "learning_rate": 7.5914150715347675e-06,
+      "loss": 1.6282,
+      "step": 24840
+    },
+    {
+      "epoch": 0.7231579253570701,
+      "grad_norm": 12.1875,
+      "learning_rate": 7.5894757898621526e-06,
+      "loss": 1.6428,
+      "step": 24860
+    },
+    {
+      "epoch": 0.7237397096896181,
+      "grad_norm": 12.5,
+      "learning_rate": 7.587536508189538e-06,
+      "loss": 1.7236,
+      "step": 24880
+    },
+    {
+      "epoch": 0.724321494022166,
+      "grad_norm": 11.625,
+      "learning_rate": 7.585597226516923e-06,
+      "loss": 1.7371,
+      "step": 24900
+    },
+    {
+      "epoch": 0.7249032783547139,
+      "grad_norm": 11.5625,
+      "learning_rate": 7.583657944844308e-06,
+      "loss": 1.6998,
+      "step": 24920
+    },
+    {
+      "epoch": 0.7254850626872619,
+      "grad_norm": 13.1875,
+      "learning_rate": 7.581718663171693e-06,
+      "loss": 1.5971,
+      "step": 24940
+    },
+    {
+      "epoch": 0.7260668470198097,
+      "grad_norm": 17.25,
+      "learning_rate": 7.579779381499078e-06,
+      "loss": 1.7575,
+      "step": 24960
+    },
+    {
+      "epoch": 0.7266486313523577,
+      "grad_norm": 13.25,
+      "learning_rate": 7.577840099826463e-06,
+      "loss": 1.6709,
+      "step": 24980
+    },
+    {
+      "epoch": 0.7272304156849057,
+      "grad_norm": 14.4375,
+      "learning_rate": 7.575900818153848e-06,
+      "loss": 1.5984,
+      "step": 25000
+    },
+    {
+      "epoch": 0.7278122000174535,
+      "grad_norm": 15.0,
+      "learning_rate": 7.573961536481233e-06,
+      "loss": 1.7063,
+      "step": 25020
+    },
+    {
+      "epoch": 0.7283939843500015,
+      "grad_norm": 11.125,
+      "learning_rate": 7.5720222548086184e-06,
+      "loss": 1.5983,
+      "step": 25040
+    },
+    {
+      "epoch": 0.7289757686825494,
+      "grad_norm": 13.375,
+      "learning_rate": 7.5700829731360035e-06,
+      "loss": 1.516,
+      "step": 25060
+    },
+    {
+      "epoch": 0.7295575530150973,
+      "grad_norm": 12.0,
+      "learning_rate": 7.568143691463389e-06,
+      "loss": 1.6036,
+      "step": 25080
+    },
+    {
+      "epoch": 0.7301393373476452,
+      "grad_norm": 15.9375,
+      "learning_rate": 7.566204409790774e-06,
+      "loss": 1.6088,
+      "step": 25100
+    },
+    {
+      "epoch": 0.7307211216801931,
+      "grad_norm": 12.0,
+      "learning_rate": 7.564265128118159e-06,
+      "loss": 1.7613,
+      "step": 25120
+    },
+    {
+      "epoch": 0.7313029060127411,
+      "grad_norm": 12.125,
+      "learning_rate": 7.562325846445544e-06,
+      "loss": 1.638,
+      "step": 25140
+    },
+    {
+      "epoch": 0.731884690345289,
+      "grad_norm": 13.375,
+      "learning_rate": 7.560386564772929e-06,
+      "loss": 1.644,
+      "step": 25160
+    },
+    {
+      "epoch": 0.7324664746778369,
+      "grad_norm": 12.1875,
+      "learning_rate": 7.558447283100314e-06,
+      "loss": 1.6405,
+      "step": 25180
+    },
+    {
+      "epoch": 0.7330482590103848,
+      "grad_norm": 10.625,
+      "learning_rate": 7.556508001427699e-06,
+      "loss": 1.6012,
+      "step": 25200
+    },
+    {
+      "epoch": 0.7336300433429328,
+      "grad_norm": 10.9375,
+      "learning_rate": 7.554568719755084e-06,
+      "loss": 1.6877,
+      "step": 25220
+    },
+    {
+      "epoch": 0.7342118276754807,
+      "grad_norm": 12.0,
+      "learning_rate": 7.5526294380824694e-06,
+      "loss": 1.6231,
+      "step": 25240
+    },
+    {
+      "epoch": 0.7347936120080286,
+      "grad_norm": 13.6875,
+      "learning_rate": 7.5506901564098545e-06,
+      "loss": 1.6582,
+      "step": 25260
+    },
+    {
+      "epoch": 0.7353753963405766,
+      "grad_norm": 14.1875,
+      "learning_rate": 7.54875087473724e-06,
+      "loss": 1.7438,
+      "step": 25280
+    },
+    {
+      "epoch": 0.7359571806731244,
+      "grad_norm": 12.25,
+      "learning_rate": 7.546811593064625e-06,
+      "loss": 1.6458,
+      "step": 25300
+    },
+    {
+      "epoch": 0.7365389650056724,
+      "grad_norm": 15.5,
+      "learning_rate": 7.54487231139201e-06,
+      "loss": 1.5911,
+      "step": 25320
+    },
+    {
+      "epoch": 0.7371207493382204,
+      "grad_norm": 8.25,
+      "learning_rate": 7.542933029719395e-06,
+      "loss": 1.5868,
+      "step": 25340
+    },
+    {
+      "epoch": 0.7377025336707682,
+      "grad_norm": 15.8125,
+      "learning_rate": 7.54099374804678e-06,
+      "loss": 1.583,
+      "step": 25360
+    },
+    {
+      "epoch": 0.7382843180033162,
+      "grad_norm": 10.3125,
+      "learning_rate": 7.539054466374165e-06,
+      "loss": 1.6235,
+      "step": 25380
+    },
+    {
+      "epoch": 0.7388661023358641,
+      "grad_norm": 11.8125,
+      "learning_rate": 7.53711518470155e-06,
+      "loss": 1.5862,
+      "step": 25400
+    },
+    {
+      "epoch": 0.739447886668412,
+      "grad_norm": 13.125,
+      "learning_rate": 7.535175903028935e-06,
+      "loss": 1.6101,
+      "step": 25420
+    },
+    {
+      "epoch": 0.74002967100096,
+      "grad_norm": 12.875,
+      "learning_rate": 7.53323662135632e-06,
+      "loss": 1.6228,
+      "step": 25440
+    },
+    {
+      "epoch": 0.7406114553335079,
+      "grad_norm": 11.3125,
+      "learning_rate": 7.5312973396837055e-06,
+      "loss": 1.5888,
+      "step": 25460
+    },
+    {
+      "epoch": 0.7411932396660558,
+      "grad_norm": 14.0625,
+      "learning_rate": 7.529358058011091e-06,
+      "loss": 1.6804,
+      "step": 25480
+    },
+    {
+      "epoch": 0.7417750239986037,
+      "grad_norm": 13.375,
+      "learning_rate": 7.527418776338476e-06,
+      "loss": 1.6769,
+      "step": 25500
+    },
+    {
+      "epoch": 0.7423568083311517,
+      "grad_norm": 12.4375,
+      "learning_rate": 7.525479494665861e-06,
+      "loss": 1.5462,
+      "step": 25520
+    },
+    {
+      "epoch": 0.7429385926636995,
+      "grad_norm": 14.1875,
+      "learning_rate": 7.523540212993246e-06,
+      "loss": 1.7738,
+      "step": 25540
+    },
+    {
+      "epoch": 0.7435203769962475,
+      "grad_norm": 8.8125,
+      "learning_rate": 7.521600931320631e-06,
+      "loss": 1.6262,
+      "step": 25560
+    },
+    {
+      "epoch": 0.7441021613287954,
+      "grad_norm": 14.0,
+      "learning_rate": 7.519661649648016e-06,
+      "loss": 1.5742,
+      "step": 25580
+    },
+    {
+      "epoch": 0.7446839456613433,
+      "grad_norm": 11.0625,
+      "learning_rate": 7.517722367975401e-06,
+      "loss": 1.7015,
+      "step": 25600
+    },
+    {
+      "epoch": 0.7452657299938913,
+      "grad_norm": 12.125,
+      "learning_rate": 7.515783086302786e-06,
+      "loss": 1.7003,
+      "step": 25620
+    },
+    {
+      "epoch": 0.7458475143264391,
+      "grad_norm": 13.25,
+      "learning_rate": 7.5138438046301705e-06,
+      "loss": 1.6813,
+      "step": 25640
+    },
+    {
+      "epoch": 0.7464292986589871,
+      "grad_norm": 12.25,
+      "learning_rate": 7.511904522957556e-06,
+      "loss": 1.6657,
+      "step": 25660
+    },
+    {
+      "epoch": 0.7470110829915351,
+      "grad_norm": 13.625,
+      "learning_rate": 7.509965241284941e-06,
+      "loss": 1.6187,
+      "step": 25680
+    },
+    {
+      "epoch": 0.7475928673240829,
+      "grad_norm": 11.375,
+      "learning_rate": 7.508025959612326e-06,
+      "loss": 1.6316,
+      "step": 25700
+    },
+    {
+      "epoch": 0.7481746516566309,
+      "grad_norm": 12.0625,
+      "learning_rate": 7.506086677939711e-06,
+      "loss": 1.6577,
+      "step": 25720
+    },
+    {
+      "epoch": 0.7487564359891788,
+      "grad_norm": 12.5625,
+      "learning_rate": 7.504147396267096e-06,
+      "loss": 1.6205,
+      "step": 25740
+    },
+    {
+      "epoch": 0.7493382203217267,
+      "grad_norm": 15.125,
+      "learning_rate": 7.502208114594481e-06,
+      "loss": 1.6341,
+      "step": 25760
+    },
+    {
+      "epoch": 0.7499200046542747,
+      "grad_norm": 12.9375,
+      "learning_rate": 7.500268832921866e-06,
+      "loss": 1.6506,
+      "step": 25780
+    },
+    {
+      "epoch": 0.7505017889868226,
+      "grad_norm": 13.5625,
+      "learning_rate": 7.498329551249251e-06,
+      "loss": 1.7797,
+      "step": 25800
+    },
+    {
+      "epoch": 0.7510835733193705,
+      "grad_norm": 14.125,
+      "learning_rate": 7.496390269576636e-06,
+      "loss": 1.6324,
+      "step": 25820
+    },
+    {
+      "epoch": 0.7516653576519184,
+      "grad_norm": 11.5625,
+      "learning_rate": 7.4944509879040215e-06,
+      "loss": 1.5381,
+      "step": 25840
+    },
+    {
+      "epoch": 0.7522471419844664,
+      "grad_norm": 12.0,
+      "learning_rate": 7.492511706231407e-06,
+      "loss": 1.6465,
+      "step": 25860
+    },
+    {
+      "epoch": 0.7528289263170143,
+      "grad_norm": 12.3125,
+      "learning_rate": 7.490572424558792e-06,
+      "loss": 1.6397,
+      "step": 25880
+    },
+    {
+      "epoch": 0.7534107106495622,
+      "grad_norm": 12.5,
+      "learning_rate": 7.488633142886177e-06,
+      "loss": 1.6028,
+      "step": 25900
+    },
+    {
+      "epoch": 0.7539924949821102,
+      "grad_norm": 12.125,
+      "learning_rate": 7.486693861213562e-06,
+      "loss": 1.6078,
+      "step": 25920
+    },
+    {
+      "epoch": 0.754574279314658,
+      "grad_norm": 12.75,
+      "learning_rate": 7.484754579540947e-06,
+      "loss": 1.5693,
+      "step": 25940
+    },
+    {
+      "epoch": 0.755156063647206,
+      "grad_norm": 11.5625,
+      "learning_rate": 7.482815297868332e-06,
+      "loss": 1.656,
+      "step": 25960
+    },
+    {
+      "epoch": 0.7557378479797539,
+      "grad_norm": 10.5,
+      "learning_rate": 7.480876016195717e-06,
+      "loss": 1.6052,
+      "step": 25980
+    },
+    {
+      "epoch": 0.7563196323123018,
+      "grad_norm": 14.5625,
+      "learning_rate": 7.478936734523102e-06,
+      "loss": 1.5874,
+      "step": 26000
+    },
+    {
+      "epoch": 0.7569014166448498,
+      "grad_norm": 15.25,
+      "learning_rate": 7.476997452850487e-06,
+      "loss": 1.6457,
+      "step": 26020
+    },
+    {
+      "epoch": 0.7574832009773976,
+      "grad_norm": 11.0625,
+      "learning_rate": 7.4750581711778725e-06,
+      "loss": 1.5725,
+      "step": 26040
+    },
+    {
+      "epoch": 0.7580649853099456,
+      "grad_norm": 12.0,
+      "learning_rate": 7.4731188895052576e-06,
+      "loss": 1.6549,
+      "step": 26060
+    },
+    {
+      "epoch": 0.7586467696424936,
+      "grad_norm": 12.3125,
+      "learning_rate": 7.471179607832643e-06,
+      "loss": 1.6379,
+      "step": 26080
+    },
+    {
+      "epoch": 0.7592285539750414,
+      "grad_norm": 11.625,
+      "learning_rate": 7.469240326160028e-06,
+      "loss": 1.5308,
+      "step": 26100
+    },
+    {
+      "epoch": 0.7598103383075894,
+      "grad_norm": 11.25,
+      "learning_rate": 7.467301044487413e-06,
+      "loss": 1.6131,
+      "step": 26120
+    },
+    {
+      "epoch": 0.7603921226401373,
+      "grad_norm": 16.875,
+      "learning_rate": 7.465361762814798e-06,
+      "loss": 1.5696,
+      "step": 26140
+    },
+    {
+      "epoch": 0.7609739069726852,
+      "grad_norm": 10.625,
+      "learning_rate": 7.463422481142183e-06,
+      "loss": 1.635,
+      "step": 26160
+    },
+    {
+      "epoch": 0.7615556913052332,
+      "grad_norm": 10.8125,
+      "learning_rate": 7.461483199469568e-06,
+      "loss": 1.5774,
+      "step": 26180
+    },
+    {
+      "epoch": 0.7621374756377811,
+      "grad_norm": 12.8125,
+      "learning_rate": 7.459543917796953e-06,
+      "loss": 1.5712,
+      "step": 26200
+    },
+    {
+      "epoch": 0.762719259970329,
+      "grad_norm": 13.375,
+      "learning_rate": 7.457604636124338e-06,
+      "loss": 1.5626,
+      "step": 26220
+    },
+    {
+      "epoch": 0.7633010443028769,
+      "grad_norm": 12.3125,
+      "learning_rate": 7.4556653544517235e-06,
+      "loss": 1.6614,
+      "step": 26240
+    },
+    {
+      "epoch": 0.7638828286354249,
+      "grad_norm": 13.0,
+      "learning_rate": 7.4537260727791085e-06,
+      "loss": 1.5953,
+      "step": 26260
+    },
+    {
+      "epoch": 0.7644646129679727,
+      "grad_norm": 12.125,
+      "learning_rate": 7.451786791106494e-06,
+      "loss": 1.6212,
+      "step": 26280
+    },
+    {
+      "epoch": 0.7650463973005207,
+      "grad_norm": 14.8125,
+      "learning_rate": 7.449847509433879e-06,
+      "loss": 1.5182,
+      "step": 26300
+    },
+    {
+      "epoch": 0.7656281816330687,
+      "grad_norm": 11.6875,
+      "learning_rate": 7.447908227761264e-06,
+      "loss": 1.6819,
+      "step": 26320
+    },
+    {
+      "epoch": 0.7662099659656165,
+      "grad_norm": 12.6875,
+      "learning_rate": 7.445968946088649e-06,
+      "loss": 1.5642,
+      "step": 26340
+    },
+    {
+      "epoch": 0.7667917502981645,
+      "grad_norm": 15.0,
+      "learning_rate": 7.444029664416034e-06,
+      "loss": 1.6226,
+      "step": 26360
+    },
+    {
+      "epoch": 0.7673735346307123,
+      "grad_norm": 14.1875,
+      "learning_rate": 7.442090382743419e-06,
+      "loss": 1.611,
+      "step": 26380
+    },
+    {
+      "epoch": 0.7679553189632603,
+      "grad_norm": 13.6875,
+      "learning_rate": 7.440151101070804e-06,
+      "loss": 1.6904,
+      "step": 26400
+    },
+    {
+      "epoch": 0.7685371032958083,
+      "grad_norm": 12.625,
+      "learning_rate": 7.438211819398189e-06,
+      "loss": 1.6579,
+      "step": 26420
+    },
+    {
+      "epoch": 0.7691188876283561,
+      "grad_norm": 11.125,
+      "learning_rate": 7.4362725377255744e-06,
+      "loss": 1.584,
+      "step": 26440
+    },
+    {
+      "epoch": 0.7697006719609041,
+      "grad_norm": 13.8125,
+      "learning_rate": 7.434333256052959e-06,
+      "loss": 1.6818,
+      "step": 26460
+    },
+    {
+      "epoch": 0.770282456293452,
+      "grad_norm": 15.75,
+      "learning_rate": 7.432393974380344e-06,
+      "loss": 1.6551,
+      "step": 26480
+    },
+    {
+      "epoch": 0.7708642406259999,
+      "grad_norm": 12.75,
+      "learning_rate": 7.430454692707729e-06,
+      "loss": 1.672,
+      "step": 26500
+    },
+    {
+      "epoch": 0.7714460249585479,
+      "grad_norm": 11.875,
+      "learning_rate": 7.428515411035114e-06,
+      "loss": 1.6364,
+      "step": 26520
+    },
+    {
+      "epoch": 0.7720278092910958,
+      "grad_norm": 11.25,
+      "learning_rate": 7.426576129362499e-06,
+      "loss": 1.6867,
+      "step": 26540
+    },
+    {
+      "epoch": 0.7726095936236437,
+      "grad_norm": 12.375,
+      "learning_rate": 7.424636847689884e-06,
+      "loss": 1.6134,
+      "step": 26560
+    },
+    {
+      "epoch": 0.7731913779561916,
+      "grad_norm": 14.1875,
+      "learning_rate": 7.422697566017269e-06,
+      "loss": 1.5726,
+      "step": 26580
+    },
+    {
+      "epoch": 0.7737731622887396,
+      "grad_norm": 14.25,
+      "learning_rate": 7.420758284344654e-06,
+      "loss": 1.6205,
+      "step": 26600
+    },
+    {
+      "epoch": 0.7743549466212875,
+      "grad_norm": 10.875,
+      "learning_rate": 7.4188190026720395e-06,
+      "loss": 1.6047,
+      "step": 26620
+    },
+    {
+      "epoch": 0.7749367309538354,
+      "grad_norm": 13.0,
+      "learning_rate": 7.4168797209994246e-06,
+      "loss": 1.6801,
+      "step": 26640
+    },
+    {
+      "epoch": 0.7755185152863834,
+      "grad_norm": 12.75,
+      "learning_rate": 7.41494043932681e-06,
+      "loss": 1.7161,
+      "step": 26660
+    },
+    {
+      "epoch": 0.7761002996189312,
+      "grad_norm": 12.1875,
+      "learning_rate": 7.413001157654195e-06,
+      "loss": 1.6308,
+      "step": 26680
+    },
+    {
+      "epoch": 0.7766820839514792,
+      "grad_norm": 14.4375,
+      "learning_rate": 7.41106187598158e-06,
+      "loss": 1.5738,
+      "step": 26700
+    },
+    {
+      "epoch": 0.7772638682840272,
+      "grad_norm": 11.5,
+      "learning_rate": 7.409122594308965e-06,
+      "loss": 1.6099,
+      "step": 26720
+    },
+    {
+      "epoch": 0.777845652616575,
+      "grad_norm": 11.375,
+      "learning_rate": 7.40718331263635e-06,
+      "loss": 1.6528,
+      "step": 26740
+    },
+    {
+      "epoch": 0.778427436949123,
+      "grad_norm": 12.75,
+      "learning_rate": 7.405244030963735e-06,
+      "loss": 1.6658,
+      "step": 26760
+    },
+    {
+      "epoch": 0.7790092212816709,
+      "grad_norm": 11.5,
+      "learning_rate": 7.40330474929112e-06,
+      "loss": 1.6278,
+      "step": 26780
+    },
+    {
+      "epoch": 0.7795910056142188,
+      "grad_norm": 11.25,
+      "learning_rate": 7.401365467618505e-06,
+      "loss": 1.6695,
+      "step": 26800
+    },
+    {
+      "epoch": 0.7801727899467668,
+      "grad_norm": 15.0,
+      "learning_rate": 7.3994261859458904e-06,
+      "loss": 1.6491,
+      "step": 26820
+    },
+    {
+      "epoch": 0.7807545742793146,
+      "grad_norm": 13.25,
+      "learning_rate": 7.3974869042732755e-06,
+      "loss": 1.6186,
+      "step": 26840
+    },
+    {
+      "epoch": 0.7813363586118626,
+      "grad_norm": 12.4375,
+      "learning_rate": 7.395547622600661e-06,
+      "loss": 1.5884,
+      "step": 26860
+    },
+    {
+      "epoch": 0.7819181429444105,
+      "grad_norm": 13.25,
+      "learning_rate": 7.393608340928046e-06,
+      "loss": 1.6544,
+      "step": 26880
+    },
+    {
+      "epoch": 0.7824999272769584,
+      "grad_norm": 13.5,
+      "learning_rate": 7.391669059255431e-06,
+      "loss": 1.6644,
+      "step": 26900
+    },
+    {
+      "epoch": 0.7830817116095063,
+      "grad_norm": 9.5,
+      "learning_rate": 7.389729777582816e-06,
+      "loss": 1.6021,
+      "step": 26920
+    },
+    {
+      "epoch": 0.7836634959420543,
+      "grad_norm": 14.0625,
+      "learning_rate": 7.387790495910201e-06,
+      "loss": 1.5777,
+      "step": 26940
+    },
+    {
+      "epoch": 0.7842452802746022,
+      "grad_norm": 10.5,
+      "learning_rate": 7.385851214237586e-06,
+      "loss": 1.5951,
+      "step": 26960
+    },
+    {
+      "epoch": 0.7848270646071501,
+      "grad_norm": 12.6875,
+      "learning_rate": 7.383911932564971e-06,
+      "loss": 1.5828,
+      "step": 26980
+    },
+    {
+      "epoch": 0.7854088489396981,
+      "grad_norm": 14.875,
+      "learning_rate": 7.381972650892356e-06,
+      "loss": 1.7095,
+      "step": 27000
+    },
+    {
+      "epoch": 0.7859906332722459,
+      "grad_norm": 12.25,
+      "learning_rate": 7.380033369219741e-06,
+      "loss": 1.642,
+      "step": 27020
+    },
+    {
+      "epoch": 0.7865724176047939,
+      "grad_norm": 12.5625,
+      "learning_rate": 7.3780940875471265e-06,
+      "loss": 1.6102,
+      "step": 27040
+    },
+    {
+      "epoch": 0.7871542019373419,
+      "grad_norm": 13.1875,
+      "learning_rate": 7.376154805874512e-06,
+      "loss": 1.6857,
+      "step": 27060
+    },
+    {
+      "epoch": 0.7877359862698897,
+      "grad_norm": 11.9375,
+      "learning_rate": 7.374215524201897e-06,
+      "loss": 1.6778,
+      "step": 27080
+    },
+    {
+      "epoch": 0.7883177706024377,
+      "grad_norm": 15.375,
+      "learning_rate": 7.372276242529282e-06,
+      "loss": 1.6085,
+      "step": 27100
+    },
+    {
+      "epoch": 0.7888995549349856,
+      "grad_norm": 13.8125,
+      "learning_rate": 7.370336960856667e-06,
+      "loss": 1.6804,
+      "step": 27120
+    },
+    {
+      "epoch": 0.7894813392675335,
+      "grad_norm": 10.0625,
+      "learning_rate": 7.368397679184052e-06,
+      "loss": 1.6082,
+      "step": 27140
+    },
+    {
+      "epoch": 0.7900631236000815,
+      "grad_norm": 10.4375,
+      "learning_rate": 7.366458397511437e-06,
+      "loss": 1.6325,
+      "step": 27160
+    },
+    {
+      "epoch": 0.7906449079326294,
+      "grad_norm": 12.25,
+      "learning_rate": 7.364519115838822e-06,
+      "loss": 1.6147,
+      "step": 27180
+    },
+    {
+      "epoch": 0.7912266922651773,
+      "grad_norm": 11.0,
+      "learning_rate": 7.362579834166207e-06,
+      "loss": 1.6164,
+      "step": 27200
+    },
+    {
+      "epoch": 0.7918084765977252,
+      "grad_norm": 10.125,
+      "learning_rate": 7.360640552493592e-06,
+      "loss": 1.7442,
+      "step": 27220
+    },
+    {
+      "epoch": 0.7923902609302731,
+      "grad_norm": 12.125,
+      "learning_rate": 7.3587012708209775e-06,
+      "loss": 1.6424,
+      "step": 27240
+    },
+    {
+      "epoch": 0.7929720452628211,
+      "grad_norm": 11.5625,
+      "learning_rate": 7.356761989148362e-06,
+      "loss": 1.6828,
+      "step": 27260
+    },
+    {
+      "epoch": 0.793553829595369,
+      "grad_norm": 11.8125,
+      "learning_rate": 7.354822707475747e-06,
+      "loss": 1.6025,
+      "step": 27280
+    },
+    {
+      "epoch": 0.7941356139279169,
+      "grad_norm": 12.6875,
+      "learning_rate": 7.352883425803132e-06,
+      "loss": 1.6738,
+      "step": 27300
+    },
+    {
+      "epoch": 0.7947173982604648,
+      "grad_norm": 13.875,
+      "learning_rate": 7.350944144130517e-06,
+      "loss": 1.5983,
+      "step": 27320
+    },
+    {
+      "epoch": 0.7952991825930128,
+      "grad_norm": 10.75,
+      "learning_rate": 7.349004862457902e-06,
+      "loss": 1.6961,
+      "step": 27340
+    },
+    {
+      "epoch": 0.7958809669255607,
+      "grad_norm": 12.5625,
+      "learning_rate": 7.347065580785287e-06,
+      "loss": 1.6454,
+      "step": 27360
+    },
+    {
+      "epoch": 0.7964627512581086,
+      "grad_norm": 14.25,
+      "learning_rate": 7.345126299112672e-06,
+      "loss": 1.6563,
+      "step": 27380
+    },
+    {
+      "epoch": 0.7970445355906566,
+      "grad_norm": 14.125,
+      "learning_rate": 7.343187017440057e-06,
+      "loss": 1.6725,
+      "step": 27400
+    },
+    {
+      "epoch": 0.7976263199232044,
+      "grad_norm": 9.8125,
+      "learning_rate": 7.3412477357674425e-06,
+      "loss": 1.7038,
+      "step": 27420
+    },
+    {
+      "epoch": 0.7982081042557524,
+      "grad_norm": 12.125,
+      "learning_rate": 7.339308454094828e-06,
+      "loss": 1.6789,
+      "step": 27440
+    },
+    {
+      "epoch": 0.7987898885883004,
+      "grad_norm": 11.625,
+      "learning_rate": 7.337369172422213e-06,
+      "loss": 1.5602,
+      "step": 27460
+    },
+    {
+      "epoch": 0.7993716729208482,
+      "grad_norm": 11.125,
+      "learning_rate": 7.335429890749598e-06,
+      "loss": 1.6001,
+      "step": 27480
+    },
+    {
+      "epoch": 0.7999534572533962,
+      "grad_norm": 12.1875,
+      "learning_rate": 7.333490609076983e-06,
+      "loss": 1.6053,
+      "step": 27500
+    },
+    {
+      "epoch": 0.8005352415859441,
+      "grad_norm": 12.8125,
+      "learning_rate": 7.331551327404368e-06,
+      "loss": 1.5731,
+      "step": 27520
+    },
+    {
+      "epoch": 0.801117025918492,
+      "grad_norm": 18.25,
+      "learning_rate": 7.329612045731753e-06,
+      "loss": 1.6737,
+      "step": 27540
+    },
+    {
+      "epoch": 0.80169881025104,
+      "grad_norm": 13.8125,
+      "learning_rate": 7.327672764059138e-06,
+      "loss": 1.7213,
+      "step": 27560
+    },
+    {
+      "epoch": 0.8022805945835879,
+      "grad_norm": 11.25,
+      "learning_rate": 7.325733482386523e-06,
+      "loss": 1.699,
+      "step": 27580
+    },
+    {
+      "epoch": 0.8028623789161358,
+      "grad_norm": 16.75,
+      "learning_rate": 7.323794200713908e-06,
+      "loss": 1.655,
+      "step": 27600
+    },
+    {
+      "epoch": 0.8034441632486837,
+      "grad_norm": 13.125,
+      "learning_rate": 7.3218549190412935e-06,
+      "loss": 1.6234,
+      "step": 27620
+    },
+    {
+      "epoch": 0.8040259475812317,
+      "grad_norm": 14.4375,
+      "learning_rate": 7.319915637368679e-06,
+      "loss": 1.6227,
+      "step": 27640
+    },
+    {
+      "epoch": 0.8046077319137795,
+      "grad_norm": 10.25,
+      "learning_rate": 7.317976355696064e-06,
+      "loss": 1.6854,
+      "step": 27660
+    },
+    {
+      "epoch": 0.8051895162463275,
+      "grad_norm": 10.875,
+      "learning_rate": 7.316037074023449e-06,
+      "loss": 1.6491,
+      "step": 27680
+    },
+    {
+      "epoch": 0.8057713005788754,
+      "grad_norm": 11.75,
+      "learning_rate": 7.314097792350834e-06,
+      "loss": 1.5881,
+      "step": 27700
+    },
+    {
+      "epoch": 0.8063530849114233,
+      "grad_norm": 14.5,
+      "learning_rate": 7.312158510678219e-06,
+      "loss": 1.6704,
+      "step": 27720
+    },
+    {
+      "epoch": 0.8069348692439713,
+      "grad_norm": 13.5625,
+      "learning_rate": 7.310219229005604e-06,
+      "loss": 1.5579,
+      "step": 27740
+    },
+    {
+      "epoch": 0.8075166535765191,
+      "grad_norm": 14.6875,
+      "learning_rate": 7.308279947332989e-06,
+      "loss": 1.4641,
+      "step": 27760
+    },
+    {
+      "epoch": 0.8080984379090671,
+      "grad_norm": 14.25,
+      "learning_rate": 7.306340665660374e-06,
+      "loss": 1.565,
+      "step": 27780
+    },
+    {
+      "epoch": 0.8086802222416151,
+      "grad_norm": 13.125,
+      "learning_rate": 7.304401383987759e-06,
+      "loss": 1.6324,
+      "step": 27800
+    },
+    {
+      "epoch": 0.8092620065741629,
+      "grad_norm": 13.0,
+      "learning_rate": 7.3024621023151445e-06,
+      "loss": 1.6607,
+      "step": 27820
+    },
+    {
+      "epoch": 0.8098437909067109,
+      "grad_norm": 12.75,
+      "learning_rate": 7.3005228206425296e-06,
+      "loss": 1.6168,
+      "step": 27840
+    },
+    {
+      "epoch": 0.8104255752392588,
+      "grad_norm": 12.125,
+      "learning_rate": 7.298583538969915e-06,
+      "loss": 1.6173,
+      "step": 27860
+    },
+    {
+      "epoch": 0.8110073595718067,
+      "grad_norm": 10.6875,
+      "learning_rate": 7.2966442572973e-06,
+      "loss": 1.6901,
+      "step": 27880
+    },
+    {
+      "epoch": 0.8115891439043547,
+      "grad_norm": 12.4375,
+      "learning_rate": 7.294704975624685e-06,
+      "loss": 1.6655,
+      "step": 27900
+    },
+    {
+      "epoch": 0.8121709282369026,
+      "grad_norm": 12.8125,
+      "learning_rate": 7.29276569395207e-06,
+      "loss": 1.5395,
+      "step": 27920
+    },
+    {
+      "epoch": 0.8127527125694505,
+      "grad_norm": 13.25,
+      "learning_rate": 7.290826412279455e-06,
+      "loss": 1.4986,
+      "step": 27940
+    },
+    {
+      "epoch": 0.8133344969019984,
+      "grad_norm": 11.6875,
+      "learning_rate": 7.28888713060684e-06,
+      "loss": 1.6638,
+      "step": 27960
+    },
+    {
+      "epoch": 0.8139162812345464,
+      "grad_norm": 11.5,
+      "learning_rate": 7.286947848934225e-06,
+      "loss": 1.5842,
+      "step": 27980
+    },
+    {
+      "epoch": 0.8144980655670943,
+      "grad_norm": 12.75,
+      "learning_rate": 7.28500856726161e-06,
+      "loss": 1.6284,
+      "step": 28000
+    },
+    {
+      "epoch": 0.8150798498996422,
+      "grad_norm": 10.25,
+      "learning_rate": 7.2830692855889954e-06,
+      "loss": 1.6638,
+      "step": 28020
+    },
+    {
+      "epoch": 0.8156616342321902,
+      "grad_norm": 13.125,
+      "learning_rate": 7.2811300039163805e-06,
+      "loss": 1.7014,
+      "step": 28040
+    },
+    {
+      "epoch": 0.816243418564738,
+      "grad_norm": 12.125,
+      "learning_rate": 7.279190722243766e-06,
+      "loss": 1.6751,
+      "step": 28060
+    },
+    {
+      "epoch": 0.816825202897286,
+      "grad_norm": 14.0,
+      "learning_rate": 7.27725144057115e-06,
+      "loss": 1.651,
+      "step": 28080
+    },
+    {
+      "epoch": 0.8174069872298338,
+      "grad_norm": 13.25,
+      "learning_rate": 7.275312158898535e-06,
+      "loss": 1.5778,
+      "step": 28100
+    },
+    {
+      "epoch": 0.8179887715623818,
+      "grad_norm": 15.1875,
+      "learning_rate": 7.27337287722592e-06,
+      "loss": 1.6771,
+      "step": 28120
+    },
+    {
+      "epoch": 0.8185705558949298,
+      "grad_norm": 12.375,
+      "learning_rate": 7.271433595553305e-06,
+      "loss": 1.6015,
+      "step": 28140
+    },
+    {
+      "epoch": 0.8191523402274776,
+      "grad_norm": 12.625,
+      "learning_rate": 7.26949431388069e-06,
+      "loss": 1.6831,
+      "step": 28160
+    },
+    {
+      "epoch": 0.8197341245600256,
+      "grad_norm": 9.8125,
+      "learning_rate": 7.267555032208075e-06,
+      "loss": 1.6934,
+      "step": 28180
+    },
+    {
+      "epoch": 0.8203159088925736,
+      "grad_norm": 13.9375,
+      "learning_rate": 7.2656157505354605e-06,
+      "loss": 1.6941,
+      "step": 28200
+    },
+    {
+      "epoch": 0.8208976932251214,
+      "grad_norm": 11.875,
+      "learning_rate": 7.2636764688628456e-06,
+      "loss": 1.5997,
+      "step": 28220
+    },
+    {
+      "epoch": 0.8214794775576694,
+      "grad_norm": 16.5,
+      "learning_rate": 7.261737187190231e-06,
+      "loss": 1.6822,
+      "step": 28240
+    },
+    {
+      "epoch": 0.8220612618902173,
+      "grad_norm": 11.375,
+      "learning_rate": 7.259797905517616e-06,
+      "loss": 1.4695,
+      "step": 28260
+    },
+    {
+      "epoch": 0.8226430462227652,
+      "grad_norm": 17.0,
+      "learning_rate": 7.257858623845001e-06,
+      "loss": 1.6311,
+      "step": 28280
+    },
+    {
+      "epoch": 0.8232248305553131,
+      "grad_norm": 10.5,
+      "learning_rate": 7.255919342172386e-06,
+      "loss": 1.6257,
+      "step": 28300
+    },
+    {
+      "epoch": 0.8238066148878611,
+      "grad_norm": 13.5,
+      "learning_rate": 7.253980060499771e-06,
+      "loss": 1.6623,
+      "step": 28320
+    },
+    {
+      "epoch": 0.824388399220409,
+      "grad_norm": 12.4375,
+      "learning_rate": 7.252040778827156e-06,
+      "loss": 1.7535,
+      "step": 28340
+    },
+    {
+      "epoch": 0.8249701835529569,
+      "grad_norm": 11.5,
+      "learning_rate": 7.250101497154541e-06,
+      "loss": 1.6129,
+      "step": 28360
+    },
+    {
+      "epoch": 0.8255519678855049,
+      "grad_norm": 13.375,
+      "learning_rate": 7.248162215481926e-06,
+      "loss": 1.6641,
+      "step": 28380
+    },
+    {
+      "epoch": 0.8261337522180527,
+      "grad_norm": 14.0625,
+      "learning_rate": 7.2462229338093114e-06,
+      "loss": 1.6582,
+      "step": 28400
+    },
+    {
+      "epoch": 0.8267155365506007,
+      "grad_norm": 10.75,
+      "learning_rate": 7.2442836521366965e-06,
+      "loss": 1.5584,
+      "step": 28420
+    },
+    {
+      "epoch": 0.8272973208831487,
+      "grad_norm": 12.625,
+      "learning_rate": 7.242344370464082e-06,
+      "loss": 1.6738,
+      "step": 28440
+    },
+    {
+      "epoch": 0.8278791052156965,
+      "grad_norm": 13.375,
+      "learning_rate": 7.240405088791467e-06,
+      "loss": 1.6583,
+      "step": 28460
+    },
+    {
+      "epoch": 0.8284608895482445,
+      "grad_norm": 13.75,
+      "learning_rate": 7.238465807118852e-06,
+      "loss": 1.5918,
+      "step": 28480
+    },
+    {
+      "epoch": 0.8290426738807923,
+      "grad_norm": 16.0,
+      "learning_rate": 7.236526525446237e-06,
+      "loss": 1.6061,
+      "step": 28500
+    },
+    {
+      "epoch": 0.8296244582133403,
+      "grad_norm": 12.0625,
+      "learning_rate": 7.234587243773622e-06,
+      "loss": 1.6719,
+      "step": 28520
+    },
+    {
+      "epoch": 0.8302062425458883,
+      "grad_norm": 13.0,
+      "learning_rate": 7.232647962101007e-06,
+      "loss": 1.5677,
+      "step": 28540
+    },
+    {
+      "epoch": 0.8307880268784361,
+      "grad_norm": 13.1875,
+      "learning_rate": 7.230708680428392e-06,
+      "loss": 1.6761,
+      "step": 28560
+    },
+    {
+      "epoch": 0.8313698112109841,
+      "grad_norm": 15.5,
+      "learning_rate": 7.228769398755777e-06,
+      "loss": 1.6747,
+      "step": 28580
+    },
+    {
+      "epoch": 0.831951595543532,
+      "grad_norm": 11.375,
+      "learning_rate": 7.226830117083162e-06,
+      "loss": 1.683,
+      "step": 28600
+    },
+    {
+      "epoch": 0.8325333798760799,
+      "grad_norm": 12.4375,
+      "learning_rate": 7.2248908354105475e-06,
+      "loss": 1.6348,
+      "step": 28620
+    },
+    {
+      "epoch": 0.8331151642086279,
+      "grad_norm": 21.625,
+      "learning_rate": 7.222951553737933e-06,
+      "loss": 1.5586,
+      "step": 28640
+    },
+    {
+      "epoch": 0.8336969485411758,
+      "grad_norm": 11.5,
+      "learning_rate": 7.221012272065318e-06,
+      "loss": 1.6519,
+      "step": 28660
+    },
+    {
+      "epoch": 0.8342787328737237,
+      "grad_norm": 11.0,
+      "learning_rate": 7.219072990392703e-06,
+      "loss": 1.6271,
+      "step": 28680
+    },
+    {
+      "epoch": 0.8348605172062716,
+      "grad_norm": 12.375,
+      "learning_rate": 7.217133708720088e-06,
+      "loss": 1.6389,
+      "step": 28700
+    },
+    {
+      "epoch": 0.8354423015388196,
+      "grad_norm": 11.9375,
+      "learning_rate": 7.215194427047473e-06,
+      "loss": 1.7004,
+      "step": 28720
+    },
+    {
+      "epoch": 0.8360240858713675,
+      "grad_norm": 13.0625,
+      "learning_rate": 7.213255145374858e-06,
+      "loss": 1.6365,
+      "step": 28740
+    },
+    {
+      "epoch": 0.8366058702039154,
+      "grad_norm": 12.9375,
+      "learning_rate": 7.211315863702243e-06,
+      "loss": 1.6538,
+      "step": 28760
+    },
+    {
+      "epoch": 0.8371876545364634,
+      "grad_norm": 10.625,
+      "learning_rate": 7.209376582029628e-06,
+      "loss": 1.6667,
+      "step": 28780
+    },
+    {
+      "epoch": 0.8377694388690112,
+      "grad_norm": 13.25,
+      "learning_rate": 7.207437300357013e-06,
+      "loss": 1.5666,
+      "step": 28800
+    },
+    {
+      "epoch": 0.8383512232015592,
+      "grad_norm": 14.4375,
+      "learning_rate": 7.2054980186843985e-06,
+      "loss": 1.6309,
+      "step": 28820
+    },
+    {
+      "epoch": 0.8389330075341072,
+      "grad_norm": 13.3125,
+      "learning_rate": 7.203558737011784e-06,
+      "loss": 1.622,
+      "step": 28840
+    },
+    {
+      "epoch": 0.839514791866655,
+      "grad_norm": 13.625,
+      "learning_rate": 7.201619455339169e-06,
+      "loss": 1.5475,
+      "step": 28860
+    },
+    {
+      "epoch": 0.840096576199203,
+      "grad_norm": 12.5,
+      "learning_rate": 7.199680173666554e-06,
+      "loss": 1.6515,
+      "step": 28880
+    },
+    {
+      "epoch": 0.8406783605317509,
+      "grad_norm": 10.25,
+      "learning_rate": 7.197740891993938e-06,
+      "loss": 1.5814,
+      "step": 28900
+    },
+    {
+      "epoch": 0.8412601448642988,
+      "grad_norm": 9.1875,
+      "learning_rate": 7.195801610321323e-06,
+      "loss": 1.6171,
+      "step": 28920
+    },
+    {
+      "epoch": 0.8418419291968467,
+      "grad_norm": 13.0,
+      "learning_rate": 7.193862328648708e-06,
+      "loss": 1.6931,
+      "step": 28940
+    },
+    {
+      "epoch": 0.8424237135293946,
+      "grad_norm": 13.875,
+      "learning_rate": 7.191923046976093e-06,
+      "loss": 1.6066,
+      "step": 28960
+    },
+    {
+      "epoch": 0.8430054978619426,
+      "grad_norm": 11.75,
+      "learning_rate": 7.189983765303478e-06,
+      "loss": 1.5293,
+      "step": 28980
+    },
+    {
+      "epoch": 0.8435872821944905,
+      "grad_norm": 13.5,
+      "learning_rate": 7.1880444836308635e-06,
+      "loss": 1.5974,
+      "step": 29000
+    },
+    {
+      "epoch": 0.8441690665270384,
+      "grad_norm": 12.25,
+      "learning_rate": 7.186105201958249e-06,
+      "loss": 1.6792,
+      "step": 29020
+    },
+    {
+      "epoch": 0.8447508508595863,
+      "grad_norm": 11.0625,
+      "learning_rate": 7.184165920285634e-06,
+      "loss": 1.6707,
+      "step": 29040
+    },
+    {
+      "epoch": 0.8453326351921343,
+      "grad_norm": 11.5,
+      "learning_rate": 7.182226638613018e-06,
+      "loss": 1.5823,
+      "step": 29060
+    },
+    {
+      "epoch": 0.8459144195246822,
+      "grad_norm": 14.875,
+      "learning_rate": 7.180287356940403e-06,
+      "loss": 1.6895,
+      "step": 29080
+    },
+    {
+      "epoch": 0.8464962038572301,
+      "grad_norm": 13.8125,
+      "learning_rate": 7.178348075267788e-06,
+      "loss": 1.5561,
+      "step": 29100
+    },
+    {
+      "epoch": 0.8470779881897781,
+      "grad_norm": 14.3125,
+      "learning_rate": 7.176408793595173e-06,
+      "loss": 1.6273,
+      "step": 29120
+    },
+    {
+      "epoch": 0.8476597725223259,
+      "grad_norm": 12.8125,
+      "learning_rate": 7.174469511922558e-06,
+      "loss": 1.6924,
+      "step": 29140
+    },
+    {
+      "epoch": 0.8482415568548739,
+      "grad_norm": 11.6875,
+      "learning_rate": 7.1725302302499434e-06,
+      "loss": 1.5519,
+      "step": 29160
+    },
+    {
+      "epoch": 0.8488233411874219,
+      "grad_norm": 12.0625,
+      "learning_rate": 7.1705909485773285e-06,
+      "loss": 1.7582,
+      "step": 29180
+    },
+    {
+      "epoch": 0.8494051255199697,
+      "grad_norm": 12.4375,
+      "learning_rate": 7.168651666904714e-06,
+      "loss": 1.6532,
+      "step": 29200
+    },
+    {
+      "epoch": 0.8499869098525177,
+      "grad_norm": 11.875,
+      "learning_rate": 7.166712385232099e-06,
+      "loss": 1.6266,
+      "step": 29220
+    },
+    {
+      "epoch": 0.8505686941850656,
+      "grad_norm": 13.1875,
+      "learning_rate": 7.164773103559484e-06,
+      "loss": 1.6186,
+      "step": 29240
+    },
+    {
+      "epoch": 0.8511504785176135,
+      "grad_norm": 11.0625,
+      "learning_rate": 7.162833821886869e-06,
+      "loss": 1.5315,
+      "step": 29260
+    },
+    {
+      "epoch": 0.8517322628501615,
+      "grad_norm": 12.375,
+      "learning_rate": 7.160894540214254e-06,
+      "loss": 1.5611,
+      "step": 29280
+    },
+    {
+      "epoch": 0.8523140471827094,
+      "grad_norm": 12.125,
+      "learning_rate": 7.158955258541639e-06,
+      "loss": 1.56,
+      "step": 29300
+    },
+    {
+      "epoch": 0.8528958315152573,
+      "grad_norm": 12.5625,
+      "learning_rate": 7.157015976869024e-06,
+      "loss": 1.6365,
+      "step": 29320
+    },
+    {
+      "epoch": 0.8534776158478052,
+      "grad_norm": 13.5,
+      "learning_rate": 7.155076695196409e-06,
+      "loss": 1.7178,
+      "step": 29340
+    },
+    {
+      "epoch": 0.8540594001803531,
+      "grad_norm": 15.1875,
+      "learning_rate": 7.153137413523794e-06,
+      "loss": 1.7602,
+      "step": 29360
+    },
+    {
+      "epoch": 0.854641184512901,
+      "grad_norm": 12.6875,
+      "learning_rate": 7.1511981318511795e-06,
+      "loss": 1.6393,
+      "step": 29380
+    },
+    {
+      "epoch": 0.855222968845449,
+      "grad_norm": 10.6875,
+      "learning_rate": 7.149258850178565e-06,
+      "loss": 1.7082,
+      "step": 29400
+    },
+    {
+      "epoch": 0.8558047531779969,
+      "grad_norm": 11.75,
+      "learning_rate": 7.14731956850595e-06,
+      "loss": 1.6048,
+      "step": 29420
+    },
+    {
+      "epoch": 0.8563865375105448,
+      "grad_norm": 15.1875,
+      "learning_rate": 7.145380286833335e-06,
+      "loss": 1.601,
+      "step": 29440
+    },
+    {
+      "epoch": 0.8569683218430928,
+      "grad_norm": 12.125,
+      "learning_rate": 7.143441005160719e-06,
+      "loss": 1.6411,
+      "step": 29460
+    },
+    {
+      "epoch": 0.8575501061756406,
+      "grad_norm": 14.9375,
+      "learning_rate": 7.141501723488104e-06,
+      "loss": 1.6465,
+      "step": 29480
+    },
+    {
+      "epoch": 0.8581318905081886,
+      "grad_norm": 12.4375,
+      "learning_rate": 7.139562441815489e-06,
+      "loss": 1.6785,
+      "step": 29500
+    },
+    {
+      "epoch": 0.8587136748407366,
+      "grad_norm": 12.5625,
+      "learning_rate": 7.137623160142874e-06,
+      "loss": 1.6321,
+      "step": 29520
+    },
+    {
+      "epoch": 0.8592954591732844,
+      "grad_norm": 12.9375,
+      "learning_rate": 7.1356838784702595e-06,
+      "loss": 1.6339,
+      "step": 29540
+    },
+    {
+      "epoch": 0.8598772435058324,
+      "grad_norm": 14.375,
+      "learning_rate": 7.1337445967976445e-06,
+      "loss": 1.6174,
+      "step": 29560
+    },
+    {
+      "epoch": 0.8604590278383804,
+      "grad_norm": 11.5,
+      "learning_rate": 7.13180531512503e-06,
+      "loss": 1.5598,
+      "step": 29580
+    },
+    {
+      "epoch": 0.8610408121709282,
+      "grad_norm": 8.625,
+      "learning_rate": 7.129866033452415e-06,
+      "loss": 1.6395,
+      "step": 29600
+    },
+    {
+      "epoch": 0.8616225965034762,
+      "grad_norm": 14.0625,
+      "learning_rate": 7.1279267517798e-06,
+      "loss": 1.779,
+      "step": 29620
+    },
+    {
+      "epoch": 0.8622043808360241,
+      "grad_norm": 12.375,
+      "learning_rate": 7.125987470107185e-06,
+      "loss": 1.6451,
+      "step": 29640
+    },
+    {
+      "epoch": 0.862786165168572,
+      "grad_norm": 14.0625,
+      "learning_rate": 7.12404818843457e-06,
+      "loss": 1.523,
+      "step": 29660
+    },
+    {
+      "epoch": 0.86336794950112,
+      "grad_norm": 12.4375,
+      "learning_rate": 7.122108906761955e-06,
+      "loss": 1.6644,
+      "step": 29680
+    },
+    {
+      "epoch": 0.8639497338336679,
+      "grad_norm": 13.75,
+      "learning_rate": 7.12016962508934e-06,
+      "loss": 1.6535,
+      "step": 29700
+    },
+    {
+      "epoch": 0.8645315181662158,
+      "grad_norm": 11.8125,
+      "learning_rate": 7.118230343416725e-06,
+      "loss": 1.6443,
+      "step": 29720
+    },
+    {
+      "epoch": 0.8651133024987637,
+      "grad_norm": 13.75,
+      "learning_rate": 7.1162910617441104e-06,
+      "loss": 1.5852,
+      "step": 29740
+    },
+    {
+      "epoch": 0.8656950868313116,
+      "grad_norm": 9.8125,
+      "learning_rate": 7.1143517800714955e-06,
+      "loss": 1.5577,
+      "step": 29760
+    },
+    {
+      "epoch": 0.8662768711638595,
+      "grad_norm": 14.8125,
+      "learning_rate": 7.112412498398881e-06,
+      "loss": 1.5347,
+      "step": 29780
+    },
+    {
+      "epoch": 0.8668586554964075,
+      "grad_norm": 11.1875,
+      "learning_rate": 7.110473216726266e-06,
+      "loss": 1.6938,
+      "step": 29800
+    },
+    {
+      "epoch": 0.8674404398289554,
+      "grad_norm": 12.1875,
+      "learning_rate": 7.108533935053651e-06,
+      "loss": 1.5838,
+      "step": 29820
+    },
+    {
+      "epoch": 0.8680222241615033,
+      "grad_norm": 12.1875,
+      "learning_rate": 7.106594653381036e-06,
+      "loss": 1.654,
+      "step": 29840
+    },
+    {
+      "epoch": 0.8686040084940513,
+      "grad_norm": 10.8125,
+      "learning_rate": 7.104655371708421e-06,
+      "loss": 1.6771,
+      "step": 29860
+    },
+    {
+      "epoch": 0.8691857928265991,
+      "grad_norm": 14.25,
+      "learning_rate": 7.102716090035806e-06,
+      "loss": 1.7066,
+      "step": 29880
+    },
+    {
+      "epoch": 0.8697675771591471,
+      "grad_norm": 12.1875,
+      "learning_rate": 7.100776808363191e-06,
+      "loss": 1.6345,
+      "step": 29900
+    },
+    {
+      "epoch": 0.8703493614916951,
+      "grad_norm": 13.375,
+      "learning_rate": 7.098837526690576e-06,
+      "loss": 1.5612,
+      "step": 29920
+    },
+    {
+      "epoch": 0.8709311458242429,
+      "grad_norm": 12.0625,
+      "learning_rate": 7.096898245017961e-06,
+      "loss": 1.7013,
+      "step": 29940
+    },
+    {
+      "epoch": 0.8715129301567909,
+      "grad_norm": 13.8125,
+      "learning_rate": 7.0949589633453465e-06,
+      "loss": 1.6588,
+      "step": 29960
+    },
+    {
+      "epoch": 0.8720947144893388,
+      "grad_norm": 11.125,
+      "learning_rate": 7.093019681672732e-06,
+      "loss": 1.6883,
+      "step": 29980
+    },
+    {
+      "epoch": 0.8726764988218867,
+      "grad_norm": 14.875,
+      "learning_rate": 7.091080400000117e-06,
+      "loss": 1.657,
+      "step": 30000
+    },
+    {
+      "epoch": 0.8732582831544347,
+      "grad_norm": 10.375,
+      "learning_rate": 7.089141118327502e-06,
+      "loss": 1.6111,
+      "step": 30020
+    },
+    {
+      "epoch": 0.8738400674869826,
+      "grad_norm": 10.5625,
+      "learning_rate": 7.087201836654887e-06,
+      "loss": 1.6375,
+      "step": 30040
+    },
+    {
+      "epoch": 0.8744218518195305,
+      "grad_norm": 11.0625,
+      "learning_rate": 7.085262554982272e-06,
+      "loss": 1.5596,
+      "step": 30060
+    },
+    {
+      "epoch": 0.8750036361520784,
+      "grad_norm": 9.6875,
+      "learning_rate": 7.083323273309657e-06,
+      "loss": 1.6095,
+      "step": 30080
+    },
+    {
+      "epoch": 0.8755854204846264,
+      "grad_norm": 14.4375,
+      "learning_rate": 7.081383991637042e-06,
+      "loss": 1.5589,
+      "step": 30100
+    },
+    {
+      "epoch": 0.8761672048171742,
+      "grad_norm": 12.125,
+      "learning_rate": 7.079444709964427e-06,
+      "loss": 1.5975,
+      "step": 30120
+    },
+    {
+      "epoch": 0.8767489891497222,
+      "grad_norm": 12.375,
+      "learning_rate": 7.077505428291812e-06,
+      "loss": 1.6044,
+      "step": 30140
+    },
+    {
+      "epoch": 0.8773307734822702,
+      "grad_norm": 10.6875,
+      "learning_rate": 7.0755661466191975e-06,
+      "loss": 1.6167,
+      "step": 30160
+    },
+    {
+      "epoch": 0.877912557814818,
+      "grad_norm": 12.0,
+      "learning_rate": 7.0736268649465826e-06,
+      "loss": 1.6748,
+      "step": 30180
+    },
+    {
+      "epoch": 0.878494342147366,
+      "grad_norm": 10.75,
+      "learning_rate": 7.071687583273968e-06,
+      "loss": 1.651,
+      "step": 30200
+    },
+    {
+      "epoch": 0.8790761264799138,
+      "grad_norm": 13.25,
+      "learning_rate": 7.069748301601353e-06,
+      "loss": 1.7351,
+      "step": 30220
+    },
+    {
+      "epoch": 0.8796579108124618,
+      "grad_norm": 12.5625,
+      "learning_rate": 7.067809019928738e-06,
+      "loss": 1.5913,
+      "step": 30240
+    },
+    {
+      "epoch": 0.8802396951450098,
+      "grad_norm": 16.5,
+      "learning_rate": 7.065869738256123e-06,
+      "loss": 1.5976,
+      "step": 30260
+    },
+    {
+      "epoch": 0.8808214794775576,
+      "grad_norm": 10.5625,
+      "learning_rate": 7.063930456583507e-06,
+      "loss": 1.712,
+      "step": 30280
+    },
+    {
+      "epoch": 0.8814032638101056,
+      "grad_norm": 8.625,
+      "learning_rate": 7.061991174910892e-06,
+      "loss": 1.7202,
+      "step": 30300
+    },
+    {
+      "epoch": 0.8819850481426535,
+      "grad_norm": 11.875,
+      "learning_rate": 7.060051893238277e-06,
+      "loss": 1.6646,
+      "step": 30320
+    },
+    {
+      "epoch": 0.8825668324752014,
+      "grad_norm": 11.4375,
+      "learning_rate": 7.0581126115656625e-06,
+      "loss": 1.6245,
+      "step": 30340
+    },
+    {
+      "epoch": 0.8831486168077494,
+      "grad_norm": 13.75,
+      "learning_rate": 7.056173329893048e-06,
+      "loss": 1.6251,
+      "step": 30360
+    },
+    {
+      "epoch": 0.8837304011402973,
+      "grad_norm": 12.625,
+      "learning_rate": 7.054234048220433e-06,
+      "loss": 1.6073,
+      "step": 30380
+    },
+    {
+      "epoch": 0.8843121854728452,
+      "grad_norm": 12.875,
+      "learning_rate": 7.052294766547818e-06,
+      "loss": 1.6597,
+      "step": 30400
+    },
+    {
+      "epoch": 0.8848939698053931,
+      "grad_norm": 13.875,
+      "learning_rate": 7.050355484875203e-06,
+      "loss": 1.5641,
+      "step": 30420
+    },
+    {
+      "epoch": 0.8854757541379411,
+      "grad_norm": 13.0625,
+      "learning_rate": 7.048416203202588e-06,
+      "loss": 1.5914,
+      "step": 30440
+    },
+    {
+      "epoch": 0.886057538470489,
+      "grad_norm": 11.875,
+      "learning_rate": 7.046476921529973e-06,
+      "loss": 1.6748,
+      "step": 30460
+    },
+    {
+      "epoch": 0.8866393228030369,
+      "grad_norm": 12.0625,
+      "learning_rate": 7.044537639857358e-06,
+      "loss": 1.6939,
+      "step": 30480
+    },
+    {
+      "epoch": 0.8872211071355849,
+      "grad_norm": 11.0,
+      "learning_rate": 7.042598358184743e-06,
+      "loss": 1.5961,
+      "step": 30500
+    },
+    {
+      "epoch": 0.8878028914681327,
+      "grad_norm": 12.625,
+      "learning_rate": 7.040659076512128e-06,
+      "loss": 1.6494,
+      "step": 30520
+    },
+    {
+      "epoch": 0.8883846758006807,
+      "grad_norm": 11.25,
+      "learning_rate": 7.0387197948395135e-06,
+      "loss": 1.6075,
+      "step": 30540
+    },
+    {
+      "epoch": 0.8889664601332287,
+      "grad_norm": 10.875,
+      "learning_rate": 7.0367805131668986e-06,
+      "loss": 1.6193,
+      "step": 30560
+    },
+    {
+      "epoch": 0.8895482444657765,
+      "grad_norm": 15.5625,
+      "learning_rate": 7.034841231494284e-06,
+      "loss": 1.5741,
+      "step": 30580
+    },
+    {
+      "epoch": 0.8901300287983245,
+      "grad_norm": 13.375,
+      "learning_rate": 7.032901949821669e-06,
+      "loss": 1.6317,
+      "step": 30600
+    },
+    {
+      "epoch": 0.8907118131308723,
+      "grad_norm": 12.1875,
+      "learning_rate": 7.030962668149054e-06,
+      "loss": 1.6391,
+      "step": 30620
+    },
+    {
+      "epoch": 0.8912935974634203,
+      "grad_norm": 12.25,
+      "learning_rate": 7.029023386476439e-06,
+      "loss": 1.6586,
+      "step": 30640
+    },
+    {
+      "epoch": 0.8918753817959683,
+      "grad_norm": 11.75,
+      "learning_rate": 7.027084104803824e-06,
+      "loss": 1.7167,
+      "step": 30660
+    },
+    {
+      "epoch": 0.8924571661285161,
+      "grad_norm": 14.0625,
+      "learning_rate": 7.025144823131209e-06,
+      "loss": 1.6195,
+      "step": 30680
+    },
+    {
+      "epoch": 0.8930389504610641,
+      "grad_norm": 11.1875,
+      "learning_rate": 7.023205541458594e-06,
+      "loss": 1.6202,
+      "step": 30700
+    },
+    {
+      "epoch": 0.893620734793612,
+      "grad_norm": 11.1875,
+      "learning_rate": 7.021266259785979e-06,
+      "loss": 1.7246,
+      "step": 30720
+    },
+    {
+      "epoch": 0.8942025191261599,
+      "grad_norm": 14.5,
+      "learning_rate": 7.0193269781133645e-06,
+      "loss": 1.6541,
+      "step": 30740
+    },
+    {
+      "epoch": 0.8947843034587079,
+      "grad_norm": 14.125,
+      "learning_rate": 7.0173876964407496e-06,
+      "loss": 1.7151,
+      "step": 30760
+    },
+    {
+      "epoch": 0.8953660877912558,
+      "grad_norm": 12.5625,
+      "learning_rate": 7.015448414768135e-06,
+      "loss": 1.6388,
+      "step": 30780
+    },
+    {
+      "epoch": 0.8959478721238037,
+      "grad_norm": 12.9375,
+      "learning_rate": 7.01350913309552e-06,
+      "loss": 1.6886,
+      "step": 30800
+    },
+    {
+      "epoch": 0.8965296564563516,
+      "grad_norm": 14.5,
+      "learning_rate": 7.011569851422905e-06,
+      "loss": 1.7274,
+      "step": 30820
+    },
+    {
+      "epoch": 0.8971114407888996,
+      "grad_norm": 11.375,
+      "learning_rate": 7.00963056975029e-06,
+      "loss": 1.5491,
+      "step": 30840
+    },
+    {
+      "epoch": 0.8976932251214474,
+      "grad_norm": 11.75,
+      "learning_rate": 7.007691288077675e-06,
+      "loss": 1.7166,
+      "step": 30860
+    },
+    {
+      "epoch": 0.8982750094539954,
+      "grad_norm": 11.0,
+      "learning_rate": 7.00575200640506e-06,
+      "loss": 1.4878,
+      "step": 30880
+    },
+    {
+      "epoch": 0.8988567937865434,
+      "grad_norm": 14.5,
+      "learning_rate": 7.003812724732445e-06,
+      "loss": 1.4887,
+      "step": 30900
+    },
+    {
+      "epoch": 0.8994385781190912,
+      "grad_norm": 12.375,
+      "learning_rate": 7.00187344305983e-06,
+      "loss": 1.6791,
+      "step": 30920
+    },
+    {
+      "epoch": 0.9000203624516392,
+      "grad_norm": 14.0,
+      "learning_rate": 6.9999341613872154e-06,
+      "loss": 1.6454,
+      "step": 30940
+    },
+    {
+      "epoch": 0.9006021467841872,
+      "grad_norm": 12.375,
+      "learning_rate": 6.9979948797146005e-06,
+      "loss": 1.5783,
+      "step": 30960
+    },
+    {
+      "epoch": 0.901183931116735,
+      "grad_norm": 13.5625,
+      "learning_rate": 6.996055598041986e-06,
+      "loss": 1.6508,
+      "step": 30980
+    },
+    {
+      "epoch": 0.901765715449283,
+      "grad_norm": 15.4375,
+      "learning_rate": 6.994116316369371e-06,
+      "loss": 1.6058,
+      "step": 31000
+    },
+    {
+      "epoch": 0.9023474997818309,
+      "grad_norm": 11.4375,
+      "learning_rate": 6.992177034696756e-06,
+      "loss": 1.6019,
+      "step": 31020
+    },
+    {
+      "epoch": 0.9029292841143788,
+      "grad_norm": 11.125,
+      "learning_rate": 6.990237753024141e-06,
+      "loss": 1.7452,
+      "step": 31040
+    },
+    {
+      "epoch": 0.9035110684469267,
+      "grad_norm": 11.6875,
+      "learning_rate": 6.988298471351526e-06,
+      "loss": 1.6808,
+      "step": 31060
+    },
+    {
+      "epoch": 0.9040928527794746,
+      "grad_norm": 13.125,
+      "learning_rate": 6.986359189678911e-06,
+      "loss": 1.6858,
+      "step": 31080
+    },
+    {
+      "epoch": 0.9046746371120226,
+      "grad_norm": 13.625,
+      "learning_rate": 6.984419908006295e-06,
+      "loss": 1.7046,
+      "step": 31100
+    },
+    {
+      "epoch": 0.9052564214445705,
+      "grad_norm": 12.1875,
+      "learning_rate": 6.9824806263336805e-06,
+      "loss": 1.6436,
+      "step": 31120
+    },
+    {
+      "epoch": 0.9058382057771184,
+      "grad_norm": 11.0,
+      "learning_rate": 6.9805413446610656e-06,
+      "loss": 1.7363,
+      "step": 31140
+    },
+    {
+      "epoch": 0.9064199901096663,
+      "grad_norm": 12.125,
+      "learning_rate": 6.978602062988451e-06,
+      "loss": 1.6176,
+      "step": 31160
+    },
+    {
+      "epoch": 0.9070017744422143,
+      "grad_norm": 12.3125,
+      "learning_rate": 6.976662781315836e-06,
+      "loss": 1.6335,
+      "step": 31180
+    },
+    {
+      "epoch": 0.9075835587747622,
+      "grad_norm": 11.8125,
+      "learning_rate": 6.974723499643221e-06,
+      "loss": 1.5611,
+      "step": 31200
+    },
+    {
+      "epoch": 0.9081653431073101,
+      "grad_norm": 10.6875,
+      "learning_rate": 6.972784217970606e-06,
+      "loss": 1.6538,
+      "step": 31220
+    },
+    {
+      "epoch": 0.9087471274398581,
+      "grad_norm": 12.3125,
+      "learning_rate": 6.970844936297991e-06,
+      "loss": 1.6281,
+      "step": 31240
+    },
+    {
+      "epoch": 0.9093289117724059,
+      "grad_norm": 12.1875,
+      "learning_rate": 6.968905654625376e-06,
+      "loss": 1.7004,
+      "step": 31260
+    },
+    {
+      "epoch": 0.9099106961049539,
+      "grad_norm": 12.6875,
+      "learning_rate": 6.966966372952761e-06,
+      "loss": 1.7177,
+      "step": 31280
+    },
+    {
+      "epoch": 0.9104924804375019,
+      "grad_norm": 11.1875,
+      "learning_rate": 6.965027091280146e-06,
+      "loss": 1.591,
+      "step": 31300
+    },
+    {
+      "epoch": 0.9110742647700497,
+      "grad_norm": 8.375,
+      "learning_rate": 6.9630878096075314e-06,
+      "loss": 1.604,
+      "step": 31320
+    },
+    {
+      "epoch": 0.9116560491025977,
+      "grad_norm": 12.125,
+      "learning_rate": 6.9611485279349165e-06,
+      "loss": 1.5641,
+      "step": 31340
+    },
+    {
+      "epoch": 0.9122378334351456,
+      "grad_norm": 14.125,
+      "learning_rate": 6.959209246262302e-06,
+      "loss": 1.6305,
+      "step": 31360
+    },
+    {
+      "epoch": 0.9128196177676935,
+      "grad_norm": 12.375,
+      "learning_rate": 6.957269964589687e-06,
+      "loss": 1.6007,
+      "step": 31380
+    },
+    {
+      "epoch": 0.9134014021002415,
+      "grad_norm": 14.0625,
+      "learning_rate": 6.955330682917072e-06,
+      "loss": 1.7345,
+      "step": 31400
+    },
+    {
+      "epoch": 0.9139831864327894,
+      "grad_norm": 14.3125,
+      "learning_rate": 6.953391401244457e-06,
+      "loss": 1.6524,
+      "step": 31420
+    },
+    {
+      "epoch": 0.9145649707653373,
+      "grad_norm": 12.875,
+      "learning_rate": 6.951452119571842e-06,
+      "loss": 1.5854,
+      "step": 31440
+    },
+    {
+      "epoch": 0.9151467550978852,
+      "grad_norm": 10.9375,
+      "learning_rate": 6.949512837899227e-06,
+      "loss": 1.5793,
+      "step": 31460
+    },
+    {
+      "epoch": 0.9157285394304331,
+      "grad_norm": 11.625,
+      "learning_rate": 6.947573556226612e-06,
+      "loss": 1.6801,
+      "step": 31480
+    },
+    {
+      "epoch": 0.916310323762981,
+      "grad_norm": 9.5,
+      "learning_rate": 6.945634274553997e-06,
+      "loss": 1.64,
+      "step": 31500
+    },
+    {
+      "epoch": 0.916892108095529,
+      "grad_norm": 11.8125,
+      "learning_rate": 6.943694992881382e-06,
+      "loss": 1.5614,
+      "step": 31520
+    },
+    {
+      "epoch": 0.9174738924280769,
+      "grad_norm": 14.1875,
+      "learning_rate": 6.9417557112087675e-06,
+      "loss": 1.6406,
+      "step": 31540
+    },
+    {
+      "epoch": 0.9180556767606248,
+      "grad_norm": 14.1875,
+      "learning_rate": 6.939816429536153e-06,
+      "loss": 1.6144,
+      "step": 31560
+    },
+    {
+      "epoch": 0.9186374610931728,
+      "grad_norm": 13.875,
+      "learning_rate": 6.937877147863538e-06,
+      "loss": 1.6121,
+      "step": 31580
+    },
+    {
+      "epoch": 0.9192192454257206,
+      "grad_norm": 11.5,
+      "learning_rate": 6.935937866190923e-06,
+      "loss": 1.5784,
+      "step": 31600
+    },
+    {
+      "epoch": 0.9198010297582686,
+      "grad_norm": 15.5,
+      "learning_rate": 6.933998584518308e-06,
+      "loss": 1.506,
+      "step": 31620
+    },
+    {
+      "epoch": 0.9203828140908166,
+      "grad_norm": 11.9375,
+      "learning_rate": 6.932059302845693e-06,
+      "loss": 1.6829,
+      "step": 31640
+    },
+    {
+      "epoch": 0.9209645984233644,
+      "grad_norm": 14.625,
+      "learning_rate": 6.930120021173078e-06,
+      "loss": 1.7056,
+      "step": 31660
+    },
+    {
+      "epoch": 0.9215463827559124,
+      "grad_norm": 11.9375,
+      "learning_rate": 6.928180739500463e-06,
+      "loss": 1.6337,
+      "step": 31680
+    },
+    {
+      "epoch": 0.9221281670884603,
+      "grad_norm": 10.0,
+      "learning_rate": 6.926241457827848e-06,
+      "loss": 1.6357,
+      "step": 31700
+    },
+    {
+      "epoch": 0.9227099514210082,
+      "grad_norm": 12.5,
+      "learning_rate": 6.924302176155233e-06,
+      "loss": 1.6074,
+      "step": 31720
+    },
+    {
+      "epoch": 0.9232917357535562,
+      "grad_norm": 12.625,
+      "learning_rate": 6.9223628944826185e-06,
+      "loss": 1.7342,
+      "step": 31740
+    },
+    {
+      "epoch": 0.9238735200861041,
+      "grad_norm": 14.0,
+      "learning_rate": 6.920423612810004e-06,
+      "loss": 1.712,
+      "step": 31760
+    },
+    {
+      "epoch": 0.924455304418652,
+      "grad_norm": 24.375,
+      "learning_rate": 6.918484331137389e-06,
+      "loss": 1.6576,
+      "step": 31780
+    },
+    {
+      "epoch": 0.9250370887511999,
+      "grad_norm": 10.875,
+      "learning_rate": 6.916545049464774e-06,
+      "loss": 1.6502,
+      "step": 31800
+    },
+    {
+      "epoch": 0.9256188730837479,
+      "grad_norm": 10.625,
+      "learning_rate": 6.914605767792159e-06,
+      "loss": 1.5984,
+      "step": 31820
+    },
+    {
+      "epoch": 0.9262006574162958,
+      "grad_norm": 11.5,
+      "learning_rate": 6.912666486119544e-06,
+      "loss": 1.5611,
+      "step": 31840
+    },
+    {
+      "epoch": 0.9267824417488437,
+      "grad_norm": 12.5625,
+      "learning_rate": 6.910727204446929e-06,
+      "loss": 1.5556,
+      "step": 31860
+    },
+    {
+      "epoch": 0.9273642260813916,
+      "grad_norm": 12.0,
+      "learning_rate": 6.908787922774314e-06,
+      "loss": 1.591,
+      "step": 31880
+    },
+    {
+      "epoch": 0.9279460104139395,
+      "grad_norm": 13.4375,
+      "learning_rate": 6.906848641101699e-06,
+      "loss": 1.6348,
+      "step": 31900
+    },
+    {
+      "epoch": 0.9285277947464875,
+      "grad_norm": 12.25,
+      "learning_rate": 6.9049093594290835e-06,
+      "loss": 1.6385,
+      "step": 31920
+    },
+    {
+      "epoch": 0.9291095790790354,
+      "grad_norm": 12.5625,
+      "learning_rate": 6.902970077756469e-06,
+      "loss": 1.7238,
+      "step": 31940
+    },
+    {
+      "epoch": 0.9296913634115833,
+      "grad_norm": 12.8125,
+      "learning_rate": 6.901030796083854e-06,
+      "loss": 1.5489,
+      "step": 31960
+    },
+    {
+      "epoch": 0.9302731477441313,
+      "grad_norm": 13.75,
+      "learning_rate": 6.899091514411239e-06,
+      "loss": 1.647,
+      "step": 31980
+    },
+    {
+      "epoch": 0.9308549320766791,
+      "grad_norm": 13.75,
+      "learning_rate": 6.897152232738624e-06,
+      "loss": 1.6397,
+      "step": 32000
+    },
+    {
+      "epoch": 0.9314367164092271,
+      "grad_norm": 13.25,
+      "learning_rate": 6.895212951066009e-06,
+      "loss": 1.6137,
+      "step": 32020
+    },
+    {
+      "epoch": 0.9320185007417751,
+      "grad_norm": 10.0,
+      "learning_rate": 6.893273669393394e-06,
+      "loss": 1.6826,
+      "step": 32040
+    },
+    {
+      "epoch": 0.9326002850743229,
+      "grad_norm": 13.9375,
+      "learning_rate": 6.891334387720779e-06,
+      "loss": 1.6708,
+      "step": 32060
+    },
+    {
+      "epoch": 0.9331820694068709,
+      "grad_norm": 10.9375,
+      "learning_rate": 6.889395106048164e-06,
+      "loss": 1.5909,
+      "step": 32080
+    },
+    {
+      "epoch": 0.9337638537394188,
+      "grad_norm": 13.0,
+      "learning_rate": 6.887455824375549e-06,
+      "loss": 1.5761,
+      "step": 32100
+    },
+    {
+      "epoch": 0.9343456380719667,
+      "grad_norm": 11.0625,
+      "learning_rate": 6.8855165427029345e-06,
+      "loss": 1.642,
+      "step": 32120
+    },
+    {
+      "epoch": 0.9349274224045147,
+      "grad_norm": 11.625,
+      "learning_rate": 6.88357726103032e-06,
+      "loss": 1.6688,
+      "step": 32140
+    },
+    {
+      "epoch": 0.9355092067370626,
+      "grad_norm": 13.125,
+      "learning_rate": 6.881637979357705e-06,
+      "loss": 1.6207,
+      "step": 32160
+    },
+    {
+      "epoch": 0.9360909910696105,
+      "grad_norm": 11.125,
+      "learning_rate": 6.87969869768509e-06,
+      "loss": 1.6017,
+      "step": 32180
+    },
+    {
+      "epoch": 0.9366727754021584,
+      "grad_norm": 11.625,
+      "learning_rate": 6.877759416012475e-06,
+      "loss": 1.6321,
+      "step": 32200
+    },
+    {
+      "epoch": 0.9372545597347064,
+      "grad_norm": 12.875,
+      "learning_rate": 6.87582013433986e-06,
+      "loss": 1.6127,
+      "step": 32220
+    },
+    {
+      "epoch": 0.9378363440672542,
+      "grad_norm": 14.875,
+      "learning_rate": 6.873880852667245e-06,
+      "loss": 1.7372,
+      "step": 32240
+    },
+    {
+      "epoch": 0.9384181283998022,
+      "grad_norm": 15.9375,
+      "learning_rate": 6.87194157099463e-06,
+      "loss": 1.6319,
+      "step": 32260
+    },
+    {
+      "epoch": 0.9389999127323502,
+      "grad_norm": 13.625,
+      "learning_rate": 6.870002289322015e-06,
+      "loss": 1.5551,
+      "step": 32280
+    },
+    {
+      "epoch": 0.939581697064898,
+      "grad_norm": 10.6875,
+      "learning_rate": 6.8680630076494e-06,
+      "loss": 1.6711,
+      "step": 32300
+    },
+    {
+      "epoch": 0.940163481397446,
+      "grad_norm": 12.1875,
+      "learning_rate": 6.8661237259767855e-06,
+      "loss": 1.5174,
+      "step": 32320
+    },
+    {
+      "epoch": 0.9407452657299938,
+      "grad_norm": 12.3125,
+      "learning_rate": 6.8641844443041706e-06,
+      "loss": 1.7003,
+      "step": 32340
+    },
+    {
+      "epoch": 0.9413270500625418,
+      "grad_norm": 12.0625,
+      "learning_rate": 6.862245162631556e-06,
+      "loss": 1.6688,
+      "step": 32360
+    },
+    {
+      "epoch": 0.9419088343950898,
+      "grad_norm": 11.8125,
+      "learning_rate": 6.860305880958941e-06,
+      "loss": 1.5936,
+      "step": 32380
+    },
+    {
+      "epoch": 0.9424906187276376,
+      "grad_norm": 10.0625,
+      "learning_rate": 6.858366599286326e-06,
+      "loss": 1.67,
+      "step": 32400
+    },
+    {
+      "epoch": 0.9430724030601856,
+      "grad_norm": 16.75,
+      "learning_rate": 6.856427317613711e-06,
+      "loss": 1.6339,
+      "step": 32420
+    },
+    {
+      "epoch": 0.9436541873927335,
+      "grad_norm": 13.875,
+      "learning_rate": 6.854488035941096e-06,
+      "loss": 1.6749,
+      "step": 32440
+    },
+    {
+      "epoch": 0.9442359717252814,
+      "grad_norm": 14.25,
+      "learning_rate": 6.852548754268481e-06,
+      "loss": 1.6833,
+      "step": 32460
+    },
+    {
+      "epoch": 0.9448177560578294,
+      "grad_norm": 11.1875,
+      "learning_rate": 6.850609472595866e-06,
+      "loss": 1.6442,
+      "step": 32480
+    },
+    {
+      "epoch": 0.9453995403903773,
+      "grad_norm": 12.875,
+      "learning_rate": 6.848670190923251e-06,
+      "loss": 1.5534,
+      "step": 32500
+    },
+    {
+      "epoch": 0.9459813247229252,
+      "grad_norm": 12.5,
+      "learning_rate": 6.8467309092506364e-06,
+      "loss": 1.5668,
+      "step": 32520
+    },
+    {
+      "epoch": 0.9465631090554731,
+      "grad_norm": 13.8125,
+      "learning_rate": 6.8447916275780215e-06,
+      "loss": 1.6358,
+      "step": 32540
+    },
+    {
+      "epoch": 0.9471448933880211,
+      "grad_norm": 13.5,
+      "learning_rate": 6.842852345905407e-06,
+      "loss": 1.6178,
+      "step": 32560
+    },
+    {
+      "epoch": 0.947726677720569,
+      "grad_norm": 12.9375,
+      "learning_rate": 6.840913064232792e-06,
+      "loss": 1.5604,
+      "step": 32580
+    },
+    {
+      "epoch": 0.9483084620531169,
+      "grad_norm": 12.3125,
+      "learning_rate": 6.838973782560177e-06,
+      "loss": 1.6273,
+      "step": 32600
+    },
+    {
+      "epoch": 0.9488902463856649,
+      "grad_norm": 11.25,
+      "learning_rate": 6.837034500887562e-06,
+      "loss": 1.629,
+      "step": 32620
+    },
+    {
+      "epoch": 0.9494720307182127,
+      "grad_norm": 11.1875,
+      "learning_rate": 6.835095219214947e-06,
+      "loss": 1.621,
+      "step": 32640
+    },
+    {
+      "epoch": 0.9500538150507607,
+      "grad_norm": 12.8125,
+      "learning_rate": 6.833155937542332e-06,
+      "loss": 1.6799,
+      "step": 32660
+    },
+    {
+      "epoch": 0.9506355993833087,
+      "grad_norm": 11.25,
+      "learning_rate": 6.831216655869717e-06,
+      "loss": 1.5411,
+      "step": 32680
+    },
+    {
+      "epoch": 0.9512173837158565,
+      "grad_norm": 12.1875,
+      "learning_rate": 6.829277374197102e-06,
+      "loss": 1.6182,
+      "step": 32700
+    },
+    {
+      "epoch": 0.9517991680484045,
+      "grad_norm": 11.1875,
+      "learning_rate": 6.8273380925244866e-06,
+      "loss": 1.5881,
+      "step": 32720
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 12.625,
+      "learning_rate": 6.825398810851872e-06,
+      "loss": 1.6739,
+      "step": 32740
+    },
+    {
+      "epoch": 0.9529627367135003,
+      "grad_norm": 12.125,
+      "learning_rate": 6.823459529179257e-06,
+      "loss": 1.6474,
+      "step": 32760
+    },
+    {
+      "epoch": 0.9535445210460483,
+      "grad_norm": 12.125,
+      "learning_rate": 6.821520247506642e-06,
+      "loss": 1.652,
+      "step": 32780
+    },
+    {
+      "epoch": 0.9541263053785961,
+      "grad_norm": 12.4375,
+      "learning_rate": 6.819580965834027e-06,
+      "loss": 1.6006,
+      "step": 32800
+    },
+    {
+      "epoch": 0.9547080897111441,
+      "grad_norm": 10.8125,
+      "learning_rate": 6.817641684161412e-06,
+      "loss": 1.5544,
+      "step": 32820
+    },
+    {
+      "epoch": 0.955289874043692,
+      "grad_norm": 11.1875,
+      "learning_rate": 6.815702402488797e-06,
+      "loss": 1.6969,
+      "step": 32840
+    },
+    {
+      "epoch": 0.9558716583762399,
+      "grad_norm": 12.25,
+      "learning_rate": 6.813763120816182e-06,
+      "loss": 1.5689,
+      "step": 32860
+    },
+    {
+      "epoch": 0.9564534427087878,
+      "grad_norm": 11.9375,
+      "learning_rate": 6.811823839143567e-06,
+      "loss": 1.6163,
+      "step": 32880
+    },
+    {
+      "epoch": 0.9570352270413358,
+      "grad_norm": 11.625,
+      "learning_rate": 6.8098845574709524e-06,
+      "loss": 1.5934,
+      "step": 32900
+    },
+    {
+      "epoch": 0.9576170113738837,
+      "grad_norm": 11.25,
+      "learning_rate": 6.8079452757983375e-06,
+      "loss": 1.6005,
+      "step": 32920
+    },
+    {
+      "epoch": 0.9581987957064316,
+      "grad_norm": 13.3125,
+      "learning_rate": 6.806005994125723e-06,
+      "loss": 1.6646,
+      "step": 32940
+    },
+    {
+      "epoch": 0.9587805800389796,
+      "grad_norm": 12.4375,
+      "learning_rate": 6.804066712453108e-06,
+      "loss": 1.5621,
+      "step": 32960
+    },
+    {
+      "epoch": 0.9593623643715274,
+      "grad_norm": 12.1875,
+      "learning_rate": 6.802127430780493e-06,
+      "loss": 1.5968,
+      "step": 32980
+    },
+    {
+      "epoch": 0.9599441487040754,
+      "grad_norm": 13.25,
+      "learning_rate": 6.800188149107878e-06,
+      "loss": 1.5637,
+      "step": 33000
+    },
+    {
+      "epoch": 0.9605259330366234,
+      "grad_norm": 16.125,
+      "learning_rate": 6.798248867435263e-06,
+      "loss": 1.5478,
+      "step": 33020
+    },
+    {
+      "epoch": 0.9611077173691712,
+      "grad_norm": 14.5,
+      "learning_rate": 6.796309585762648e-06,
+      "loss": 1.4745,
+      "step": 33040
+    },
+    {
+      "epoch": 0.9616895017017192,
+      "grad_norm": 11.0625,
+      "learning_rate": 6.794370304090033e-06,
+      "loss": 1.6644,
+      "step": 33060
+    },
+    {
+      "epoch": 0.9622712860342671,
+      "grad_norm": 12.3125,
+      "learning_rate": 6.792431022417418e-06,
+      "loss": 1.6051,
+      "step": 33080
+    },
+    {
+      "epoch": 0.962853070366815,
+      "grad_norm": 12.6875,
+      "learning_rate": 6.790491740744803e-06,
+      "loss": 1.5699,
+      "step": 33100
+    },
+    {
+      "epoch": 0.963434854699363,
+      "grad_norm": 12.8125,
+      "learning_rate": 6.7885524590721885e-06,
+      "loss": 1.6223,
+      "step": 33120
+    },
+    {
+      "epoch": 0.9640166390319108,
+      "grad_norm": 12.0,
+      "learning_rate": 6.786613177399574e-06,
+      "loss": 1.6234,
+      "step": 33140
+    },
+    {
+      "epoch": 0.9645984233644588,
+      "grad_norm": 12.625,
+      "learning_rate": 6.784673895726959e-06,
+      "loss": 1.6072,
+      "step": 33160
+    },
+    {
+      "epoch": 0.9651802076970067,
+      "grad_norm": 12.375,
+      "learning_rate": 6.782734614054344e-06,
+      "loss": 1.6224,
+      "step": 33180
+    },
+    {
+      "epoch": 0.9657619920295546,
+      "grad_norm": 12.0625,
+      "learning_rate": 6.780795332381729e-06,
+      "loss": 1.6087,
+      "step": 33200
+    },
+    {
+      "epoch": 0.9663437763621026,
+      "grad_norm": 12.9375,
+      "learning_rate": 6.778856050709114e-06,
+      "loss": 1.6673,
+      "step": 33220
+    },
+    {
+      "epoch": 0.9669255606946505,
+      "grad_norm": 14.0,
+      "learning_rate": 6.776916769036499e-06,
+      "loss": 1.6716,
+      "step": 33240
+    },
+    {
+      "epoch": 0.9675073450271984,
+      "grad_norm": 11.4375,
+      "learning_rate": 6.774977487363884e-06,
+      "loss": 1.6317,
+      "step": 33260
+    },
+    {
+      "epoch": 0.9680891293597463,
+      "grad_norm": 11.875,
+      "learning_rate": 6.773038205691269e-06,
+      "loss": 1.6376,
+      "step": 33280
+    },
+    {
+      "epoch": 0.9686709136922943,
+      "grad_norm": 13.75,
+      "learning_rate": 6.771098924018654e-06,
+      "loss": 1.5793,
+      "step": 33300
+    },
+    {
+      "epoch": 0.9692526980248422,
+      "grad_norm": 11.6875,
+      "learning_rate": 6.769159642346038e-06,
+      "loss": 1.5495,
+      "step": 33320
+    },
+    {
+      "epoch": 0.9698344823573901,
+      "grad_norm": 14.1875,
+      "learning_rate": 6.767220360673423e-06,
+      "loss": 1.6489,
+      "step": 33340
+    },
+    {
+      "epoch": 0.9704162666899381,
+      "grad_norm": 13.5,
+      "learning_rate": 6.765281079000808e-06,
+      "loss": 1.6244,
+      "step": 33360
+    },
+    {
+      "epoch": 0.9709980510224859,
+      "grad_norm": 12.1875,
+      "learning_rate": 6.763341797328193e-06,
+      "loss": 1.6047,
+      "step": 33380
+    },
+    {
+      "epoch": 0.9715798353550339,
+      "grad_norm": 14.125,
+      "learning_rate": 6.761402515655578e-06,
+      "loss": 1.6346,
+      "step": 33400
+    },
+    {
+      "epoch": 0.9721616196875819,
+      "grad_norm": 11.75,
+      "learning_rate": 6.759463233982963e-06,
+      "loss": 1.5593,
+      "step": 33420
+    },
+    {
+      "epoch": 0.9727434040201297,
+      "grad_norm": 13.4375,
+      "learning_rate": 6.757523952310348e-06,
+      "loss": 1.4955,
+      "step": 33440
+    },
+    {
+      "epoch": 0.9733251883526777,
+      "grad_norm": 13.1875,
+      "learning_rate": 6.7555846706377335e-06,
+      "loss": 1.583,
+      "step": 33460
+    },
+    {
+      "epoch": 0.9739069726852256,
+      "grad_norm": 13.625,
+      "learning_rate": 6.7536453889651186e-06,
+      "loss": 1.6102,
+      "step": 33480
+    },
+    {
+      "epoch": 0.9744887570177735,
+      "grad_norm": 12.125,
+      "learning_rate": 6.751706107292504e-06,
+      "loss": 1.6518,
+      "step": 33500
+    },
+    {
+      "epoch": 0.9750705413503215,
+      "grad_norm": 12.625,
+      "learning_rate": 6.749766825619889e-06,
+      "loss": 1.6665,
+      "step": 33520
+    },
+    {
+      "epoch": 0.9756523256828694,
+      "grad_norm": 12.9375,
+      "learning_rate": 6.747827543947274e-06,
+      "loss": 1.7361,
+      "step": 33540
+    },
+    {
+      "epoch": 0.9762341100154173,
+      "grad_norm": 10.9375,
+      "learning_rate": 6.745888262274659e-06,
+      "loss": 1.6369,
+      "step": 33560
+    },
+    {
+      "epoch": 0.9768158943479652,
+      "grad_norm": 16.75,
+      "learning_rate": 6.743948980602044e-06,
+      "loss": 1.5685,
+      "step": 33580
+    },
+    {
+      "epoch": 0.9773976786805131,
+      "grad_norm": 12.9375,
+      "learning_rate": 6.742009698929429e-06,
+      "loss": 1.6494,
+      "step": 33600
+    },
+    {
+      "epoch": 0.977979463013061,
+      "grad_norm": 11.625,
+      "learning_rate": 6.740070417256814e-06,
+      "loss": 1.6524,
+      "step": 33620
+    },
+    {
+      "epoch": 0.978561247345609,
+      "grad_norm": 13.0625,
+      "learning_rate": 6.738131135584199e-06,
+      "loss": 1.5863,
+      "step": 33640
+    },
+    {
+      "epoch": 0.9791430316781569,
+      "grad_norm": 11.25,
+      "learning_rate": 6.7361918539115844e-06,
+      "loss": 1.6313,
+      "step": 33660
+    },
+    {
+      "epoch": 0.9797248160107048,
+      "grad_norm": 16.125,
+      "learning_rate": 6.7342525722389695e-06,
+      "loss": 1.6042,
+      "step": 33680
+    },
+    {
+      "epoch": 0.9803066003432528,
+      "grad_norm": 13.3125,
+      "learning_rate": 6.732313290566355e-06,
+      "loss": 1.5964,
+      "step": 33700
+    },
+    {
+      "epoch": 0.9808883846758006,
+      "grad_norm": 12.5625,
+      "learning_rate": 6.73037400889374e-06,
+      "loss": 1.5968,
+      "step": 33720
+    },
+    {
+      "epoch": 0.9814701690083486,
+      "grad_norm": 13.4375,
+      "learning_rate": 6.728434727221125e-06,
+      "loss": 1.6331,
+      "step": 33740
+    },
+    {
+      "epoch": 0.9820519533408966,
+      "grad_norm": 11.6875,
+      "learning_rate": 6.72649544554851e-06,
+      "loss": 1.6312,
+      "step": 33760
+    },
+    {
+      "epoch": 0.9826337376734444,
+      "grad_norm": 13.375,
+      "learning_rate": 6.724556163875895e-06,
+      "loss": 1.6851,
+      "step": 33780
+    },
+    {
+      "epoch": 0.9832155220059924,
+      "grad_norm": 14.25,
+      "learning_rate": 6.72261688220328e-06,
+      "loss": 1.6674,
+      "step": 33800
+    },
+    {
+      "epoch": 0.9837973063385403,
+      "grad_norm": 12.75,
+      "learning_rate": 6.720677600530665e-06,
+      "loss": 1.5453,
+      "step": 33820
+    },
+    {
+      "epoch": 0.9843790906710882,
+      "grad_norm": 12.25,
+      "learning_rate": 6.71873831885805e-06,
+      "loss": 1.7205,
+      "step": 33840
+    },
+    {
+      "epoch": 0.9849608750036362,
+      "grad_norm": 11.4375,
+      "learning_rate": 6.7167990371854354e-06,
+      "loss": 1.672,
+      "step": 33860
+    },
+    {
+      "epoch": 0.9855426593361841,
+      "grad_norm": 12.125,
+      "learning_rate": 6.7148597555128205e-06,
+      "loss": 1.5611,
+      "step": 33880
+    },
+    {
+      "epoch": 0.986124443668732,
+      "grad_norm": 11.75,
+      "learning_rate": 6.712920473840206e-06,
+      "loss": 1.6928,
+      "step": 33900
+    },
+    {
+      "epoch": 0.9867062280012799,
+      "grad_norm": 11.25,
+      "learning_rate": 6.710981192167591e-06,
+      "loss": 1.6471,
+      "step": 33920
+    },
+    {
+      "epoch": 0.9872880123338279,
+      "grad_norm": 13.5,
+      "learning_rate": 6.709041910494976e-06,
+      "loss": 1.6177,
+      "step": 33940
+    },
+    {
+      "epoch": 0.9878697966663758,
+      "grad_norm": 15.125,
+      "learning_rate": 6.707102628822361e-06,
+      "loss": 1.6374,
+      "step": 33960
+    },
+    {
+      "epoch": 0.9884515809989237,
+      "grad_norm": 13.6875,
+      "learning_rate": 6.705163347149746e-06,
+      "loss": 1.5153,
+      "step": 33980
+    },
+    {
+      "epoch": 0.9890333653314716,
+      "grad_norm": 13.0,
+      "learning_rate": 6.703224065477131e-06,
+      "loss": 1.6144,
+      "step": 34000
+    },
+    {
+      "epoch": 0.9896151496640195,
+      "grad_norm": 11.875,
+      "learning_rate": 6.701284783804516e-06,
+      "loss": 1.5558,
+      "step": 34020
+    },
+    {
+      "epoch": 0.9901969339965675,
+      "grad_norm": 10.6875,
+      "learning_rate": 6.699345502131901e-06,
+      "loss": 1.6159,
+      "step": 34040
+    },
+    {
+      "epoch": 0.9907787183291153,
+      "grad_norm": 14.4375,
+      "learning_rate": 6.697406220459286e-06,
+      "loss": 1.6065,
+      "step": 34060
+    },
+    {
+      "epoch": 0.9913605026616633,
+      "grad_norm": 13.5,
+      "learning_rate": 6.6954669387866715e-06,
+      "loss": 1.5929,
+      "step": 34080
+    },
+    {
+      "epoch": 0.9919422869942113,
+      "grad_norm": 11.8125,
+      "learning_rate": 6.693527657114056e-06,
+      "loss": 1.6878,
+      "step": 34100
+    },
+    {
+      "epoch": 0.9925240713267591,
+      "grad_norm": 12.3125,
+      "learning_rate": 6.691588375441441e-06,
+      "loss": 1.5968,
+      "step": 34120
+    },
+    {
+      "epoch": 0.9931058556593071,
+      "grad_norm": 13.8125,
+      "learning_rate": 6.689649093768826e-06,
+      "loss": 1.6993,
+      "step": 34140
+    },
+    {
+      "epoch": 0.993687639991855,
+      "grad_norm": 14.0,
+      "learning_rate": 6.687709812096211e-06,
+      "loss": 1.6291,
+      "step": 34160
+    },
+    {
+      "epoch": 0.9942694243244029,
+      "grad_norm": 14.5,
+      "learning_rate": 6.685770530423596e-06,
+      "loss": 1.5992,
+      "step": 34180
+    },
+    {
+      "epoch": 0.9948512086569509,
+      "grad_norm": 13.4375,
+      "learning_rate": 6.683831248750981e-06,
+      "loss": 1.6653,
+      "step": 34200
+    },
+    {
+      "epoch": 0.9954329929894988,
+      "grad_norm": 12.75,
+      "learning_rate": 6.681891967078366e-06,
+      "loss": 1.6667,
+      "step": 34220
+    },
+    {
+      "epoch": 0.9960147773220467,
+      "grad_norm": 13.1875,
+      "learning_rate": 6.6799526854057514e-06,
+      "loss": 1.6231,
+      "step": 34240
+    },
+    {
+      "epoch": 0.9965965616545946,
+      "grad_norm": 11.1875,
+      "learning_rate": 6.6780134037331365e-06,
+      "loss": 1.7432,
+      "step": 34260
+    },
+    {
+      "epoch": 0.9971783459871426,
+      "grad_norm": 13.875,
+      "learning_rate": 6.676074122060522e-06,
+      "loss": 1.6409,
+      "step": 34280
+    },
+    {
+      "epoch": 0.9977601303196905,
+      "grad_norm": 12.8125,
+      "learning_rate": 6.674134840387907e-06,
+      "loss": 1.6375,
+      "step": 34300
+    },
+    {
+      "epoch": 0.9983419146522384,
+      "grad_norm": 12.0,
+      "learning_rate": 6.672195558715292e-06,
+      "loss": 1.6822,
+      "step": 34320
+    },
+    {
+      "epoch": 0.9989236989847864,
+      "grad_norm": 15.5625,
+      "learning_rate": 6.670256277042677e-06,
+      "loss": 1.5642,
+      "step": 34340
+    },
+    {
+      "epoch": 0.9995054833173342,
+      "grad_norm": 9.75,
+      "learning_rate": 6.668316995370062e-06,
+      "loss": 1.7238,
+      "step": 34360
+    },
+    {
+      "epoch": 1.0000872676498822,
+      "grad_norm": 10.125,
+      "learning_rate": 6.666377713697447e-06,
+      "loss": 1.6699,
+      "step": 34380
+    },
+    {
+      "epoch": 1.0006690519824302,
+      "grad_norm": 12.5,
+      "learning_rate": 6.664438432024832e-06,
+      "loss": 1.5624,
+      "step": 34400
+    },
+    {
+      "epoch": 1.0012508363149781,
+      "grad_norm": 10.5,
+      "learning_rate": 6.662499150352217e-06,
+      "loss": 1.5326,
+      "step": 34420
+    },
+    {
+      "epoch": 1.0018326206475259,
+      "grad_norm": 14.5,
+      "learning_rate": 6.660559868679602e-06,
+      "loss": 1.5176,
+      "step": 34440
+    },
+    {
+      "epoch": 1.0024144049800738,
+      "grad_norm": 10.6875,
+      "learning_rate": 6.6586205870069875e-06,
+      "loss": 1.518,
+      "step": 34460
+    },
+    {
+      "epoch": 1.0029961893126218,
+      "grad_norm": 13.5625,
+      "learning_rate": 6.656681305334373e-06,
+      "loss": 1.519,
+      "step": 34480
+    },
+    {
+      "epoch": 1.0035779736451698,
+      "grad_norm": 11.625,
+      "learning_rate": 6.654742023661758e-06,
+      "loss": 1.5261,
+      "step": 34500
+    },
+    {
+      "epoch": 1.0041597579777177,
+      "grad_norm": 18.625,
+      "learning_rate": 6.652802741989143e-06,
+      "loss": 1.5199,
+      "step": 34520
+    },
+    {
+      "epoch": 1.0047415423102657,
+      "grad_norm": 14.0625,
+      "learning_rate": 6.650863460316528e-06,
+      "loss": 1.4781,
+      "step": 34540
+    },
+    {
+      "epoch": 1.0053233266428134,
+      "grad_norm": 12.5625,
+      "learning_rate": 6.648924178643913e-06,
+      "loss": 1.5739,
+      "step": 34560
+    },
+    {
+      "epoch": 1.0059051109753614,
+      "grad_norm": 13.9375,
+      "learning_rate": 6.646984896971298e-06,
+      "loss": 1.4837,
+      "step": 34580
+    },
+    {
+      "epoch": 1.0064868953079094,
+      "grad_norm": 12.5,
+      "learning_rate": 6.645045615298683e-06,
+      "loss": 1.546,
+      "step": 34600
+    },
+    {
+      "epoch": 1.0070686796404573,
+      "grad_norm": 14.9375,
+      "learning_rate": 6.643106333626068e-06,
+      "loss": 1.4559,
+      "step": 34620
+    },
+    {
+      "epoch": 1.0076504639730053,
+      "grad_norm": 10.75,
+      "learning_rate": 6.641167051953453e-06,
+      "loss": 1.5901,
+      "step": 34640
+    },
+    {
+      "epoch": 1.008232248305553,
+      "grad_norm": 13.375,
+      "learning_rate": 6.6392277702808385e-06,
+      "loss": 1.5853,
+      "step": 34660
+    },
+    {
+      "epoch": 1.008814032638101,
+      "grad_norm": 10.75,
+      "learning_rate": 6.6372884886082236e-06,
+      "loss": 1.5529,
+      "step": 34680
+    },
+    {
+      "epoch": 1.009395816970649,
+      "grad_norm": 13.4375,
+      "learning_rate": 6.635349206935609e-06,
+      "loss": 1.5142,
+      "step": 34700
+    },
+    {
+      "epoch": 1.009977601303197,
+      "grad_norm": 16.375,
+      "learning_rate": 6.633409925262994e-06,
+      "loss": 1.5687,
+      "step": 34720
+    },
+    {
+      "epoch": 1.0105593856357449,
+      "grad_norm": 15.0,
+      "learning_rate": 6.631470643590379e-06,
+      "loss": 1.5669,
+      "step": 34740
+    },
+    {
+      "epoch": 1.0111411699682928,
+      "grad_norm": 17.625,
+      "learning_rate": 6.629531361917764e-06,
+      "loss": 1.4651,
+      "step": 34760
+    },
+    {
+      "epoch": 1.0117229543008406,
+      "grad_norm": 11.125,
+      "learning_rate": 6.627592080245149e-06,
+      "loss": 1.4856,
+      "step": 34780
+    },
+    {
+      "epoch": 1.0123047386333885,
+      "grad_norm": 11.8125,
+      "learning_rate": 6.625652798572534e-06,
+      "loss": 1.6137,
+      "step": 34800
+    },
+    {
+      "epoch": 1.0128865229659365,
+      "grad_norm": 12.8125,
+      "learning_rate": 6.623713516899919e-06,
+      "loss": 1.5333,
+      "step": 34820
+    },
+    {
+      "epoch": 1.0134683072984845,
+      "grad_norm": 13.1875,
+      "learning_rate": 6.621774235227304e-06,
+      "loss": 1.4606,
+      "step": 34840
+    },
+    {
+      "epoch": 1.0140500916310324,
+      "grad_norm": 13.9375,
+      "learning_rate": 6.6198349535546895e-06,
+      "loss": 1.4368,
+      "step": 34860
+    },
+    {
+      "epoch": 1.0146318759635804,
+      "grad_norm": 12.5625,
+      "learning_rate": 6.6178956718820745e-06,
+      "loss": 1.4579,
+      "step": 34880
+    },
+    {
+      "epoch": 1.0152136602961281,
+      "grad_norm": 11.625,
+      "learning_rate": 6.61595639020946e-06,
+      "loss": 1.4373,
+      "step": 34900
+    },
+    {
+      "epoch": 1.015795444628676,
+      "grad_norm": 14.9375,
+      "learning_rate": 6.614017108536844e-06,
+      "loss": 1.5626,
+      "step": 34920
+    },
+    {
+      "epoch": 1.016377228961224,
+      "grad_norm": 12.25,
+      "learning_rate": 6.612077826864229e-06,
+      "loss": 1.5653,
+      "step": 34940
+    },
+    {
+      "epoch": 1.016959013293772,
+      "grad_norm": 13.0,
+      "learning_rate": 6.610138545191614e-06,
+      "loss": 1.54,
+      "step": 34960
+    },
+    {
+      "epoch": 1.01754079762632,
+      "grad_norm": 11.375,
+      "learning_rate": 6.608199263518999e-06,
+      "loss": 1.4087,
+      "step": 34980
+    },
+    {
+      "epoch": 1.018122581958868,
+      "grad_norm": 12.5,
+      "learning_rate": 6.606259981846384e-06,
+      "loss": 1.5305,
+      "step": 35000
+    },
+    {
+      "epoch": 1.0187043662914157,
+      "grad_norm": 12.75,
+      "learning_rate": 6.604320700173769e-06,
+      "loss": 1.5814,
+      "step": 35020
+    },
+    {
+      "epoch": 1.0192861506239637,
+      "grad_norm": 11.3125,
+      "learning_rate": 6.6023814185011545e-06,
+      "loss": 1.4229,
+      "step": 35040
+    },
+    {
+      "epoch": 1.0198679349565116,
+      "grad_norm": 11.625,
+      "learning_rate": 6.60044213682854e-06,
+      "loss": 1.5016,
+      "step": 35060
+    },
+    {
+      "epoch": 1.0204497192890596,
+      "grad_norm": 13.125,
+      "learning_rate": 6.598502855155925e-06,
+      "loss": 1.4902,
+      "step": 35080
+    },
+    {
+      "epoch": 1.0210315036216076,
+      "grad_norm": 11.4375,
+      "learning_rate": 6.59656357348331e-06,
+      "loss": 1.5648,
+      "step": 35100
+    },
+    {
+      "epoch": 1.0216132879541553,
+      "grad_norm": 13.0,
+      "learning_rate": 6.594624291810695e-06,
+      "loss": 1.5088,
+      "step": 35120
+    },
+    {
+      "epoch": 1.0221950722867033,
+      "grad_norm": 13.125,
+      "learning_rate": 6.59268501013808e-06,
+      "loss": 1.5048,
+      "step": 35140
+    },
+    {
+      "epoch": 1.0227768566192512,
+      "grad_norm": 12.8125,
+      "learning_rate": 6.590745728465465e-06,
+      "loss": 1.5046,
+      "step": 35160
+    },
+    {
+      "epoch": 1.0233586409517992,
+      "grad_norm": 12.375,
+      "learning_rate": 6.58880644679285e-06,
+      "loss": 1.5097,
+      "step": 35180
+    },
+    {
+      "epoch": 1.0239404252843471,
+      "grad_norm": 12.625,
+      "learning_rate": 6.586867165120235e-06,
+      "loss": 1.5104,
+      "step": 35200
+    },
+    {
+      "epoch": 1.024522209616895,
+      "grad_norm": 14.25,
+      "learning_rate": 6.58492788344762e-06,
+      "loss": 1.5456,
+      "step": 35220
+    },
+    {
+      "epoch": 1.0251039939494428,
+      "grad_norm": 12.625,
+      "learning_rate": 6.5829886017750055e-06,
+      "loss": 1.5434,
+      "step": 35240
+    },
+    {
+      "epoch": 1.0256857782819908,
+      "grad_norm": 12.625,
+      "learning_rate": 6.5810493201023906e-06,
+      "loss": 1.6952,
+      "step": 35260
+    },
+    {
+      "epoch": 1.0262675626145388,
+      "grad_norm": 12.5,
+      "learning_rate": 6.579110038429776e-06,
+      "loss": 1.5075,
+      "step": 35280
+    },
+    {
+      "epoch": 1.0268493469470867,
+      "grad_norm": 14.0,
+      "learning_rate": 6.577170756757161e-06,
+      "loss": 1.5463,
+      "step": 35300
+    },
+    {
+      "epoch": 1.0274311312796347,
+      "grad_norm": 14.1875,
+      "learning_rate": 6.575231475084546e-06,
+      "loss": 1.5216,
+      "step": 35320
+    },
+    {
+      "epoch": 1.0280129156121827,
+      "grad_norm": 11.8125,
+      "learning_rate": 6.573292193411931e-06,
+      "loss": 1.4992,
+      "step": 35340
+    },
+    {
+      "epoch": 1.0285946999447304,
+      "grad_norm": 14.125,
+      "learning_rate": 6.571352911739316e-06,
+      "loss": 1.5322,
+      "step": 35360
+    },
+    {
+      "epoch": 1.0291764842772784,
+      "grad_norm": 12.75,
+      "learning_rate": 6.569413630066701e-06,
+      "loss": 1.5707,
+      "step": 35380
+    },
+    {
+      "epoch": 1.0297582686098263,
+      "grad_norm": 12.875,
+      "learning_rate": 6.567474348394086e-06,
+      "loss": 1.428,
+      "step": 35400
+    },
+    {
+      "epoch": 1.0303400529423743,
+      "grad_norm": 13.4375,
+      "learning_rate": 6.565535066721471e-06,
+      "loss": 1.5032,
+      "step": 35420
+    },
+    {
+      "epoch": 1.0309218372749223,
+      "grad_norm": 14.8125,
+      "learning_rate": 6.5635957850488564e-06,
+      "loss": 1.5676,
+      "step": 35440
+    },
+    {
+      "epoch": 1.0315036216074702,
+      "grad_norm": 13.375,
+      "learning_rate": 6.5616565033762415e-06,
+      "loss": 1.5432,
+      "step": 35460
+    },
+    {
+      "epoch": 1.032085405940018,
+      "grad_norm": 11.6875,
+      "learning_rate": 6.559717221703627e-06,
+      "loss": 1.5117,
+      "step": 35480
+    },
+    {
+      "epoch": 1.032667190272566,
+      "grad_norm": 12.625,
+      "learning_rate": 6.557777940031012e-06,
+      "loss": 1.4937,
+      "step": 35500
+    },
+    {
+      "epoch": 1.033248974605114,
+      "grad_norm": 11.75,
+      "learning_rate": 6.555838658358397e-06,
+      "loss": 1.5287,
+      "step": 35520
+    },
+    {
+      "epoch": 1.0338307589376619,
+      "grad_norm": 12.125,
+      "learning_rate": 6.553899376685782e-06,
+      "loss": 1.4799,
+      "step": 35540
+    },
+    {
+      "epoch": 1.0344125432702098,
+      "grad_norm": 13.0,
+      "learning_rate": 6.551960095013167e-06,
+      "loss": 1.5311,
+      "step": 35560
+    },
+    {
+      "epoch": 1.0349943276027576,
+      "grad_norm": 13.625,
+      "learning_rate": 6.550020813340552e-06,
+      "loss": 1.4129,
+      "step": 35580
+    },
+    {
+      "epoch": 1.0355761119353055,
+      "grad_norm": 10.625,
+      "learning_rate": 6.548081531667937e-06,
+      "loss": 1.4979,
+      "step": 35600
+    },
+    {
+      "epoch": 1.0361578962678535,
+      "grad_norm": 13.25,
+      "learning_rate": 6.546142249995322e-06,
+      "loss": 1.5552,
+      "step": 35620
+    },
+    {
+      "epoch": 1.0367396806004014,
+      "grad_norm": 12.375,
+      "learning_rate": 6.544202968322707e-06,
+      "loss": 1.5192,
+      "step": 35640
+    },
+    {
+      "epoch": 1.0373214649329494,
+      "grad_norm": 13.125,
+      "learning_rate": 6.5422636866500925e-06,
+      "loss": 1.5752,
+      "step": 35660
+    },
+    {
+      "epoch": 1.0379032492654974,
+      "grad_norm": 14.125,
+      "learning_rate": 6.540324404977478e-06,
+      "loss": 1.4708,
+      "step": 35680
+    },
+    {
+      "epoch": 1.0384850335980451,
+      "grad_norm": 12.625,
+      "learning_rate": 6.538385123304863e-06,
+      "loss": 1.5739,
+      "step": 35700
+    },
+    {
+      "epoch": 1.039066817930593,
+      "grad_norm": 12.1875,
+      "learning_rate": 6.536445841632248e-06,
+      "loss": 1.5387,
+      "step": 35720
+    },
+    {
+      "epoch": 1.039648602263141,
+      "grad_norm": 15.1875,
+      "learning_rate": 6.534506559959632e-06,
+      "loss": 1.4135,
+      "step": 35740
+    },
+    {
+      "epoch": 1.040230386595689,
+      "grad_norm": 12.5625,
+      "learning_rate": 6.532567278287017e-06,
+      "loss": 1.5351,
+      "step": 35760
+    },
+    {
+      "epoch": 1.040812170928237,
+      "grad_norm": 13.9375,
+      "learning_rate": 6.530627996614402e-06,
+      "loss": 1.5075,
+      "step": 35780
+    },
+    {
+      "epoch": 1.041393955260785,
+      "grad_norm": 11.8125,
+      "learning_rate": 6.528688714941787e-06,
+      "loss": 1.5084,
+      "step": 35800
+    },
+    {
+      "epoch": 1.0419757395933327,
+      "grad_norm": 11.25,
+      "learning_rate": 6.5267494332691724e-06,
+      "loss": 1.5374,
+      "step": 35820
+    },
+    {
+      "epoch": 1.0425575239258806,
+      "grad_norm": 13.4375,
+      "learning_rate": 6.5248101515965575e-06,
+      "loss": 1.5538,
+      "step": 35840
+    },
+    {
+      "epoch": 1.0431393082584286,
+      "grad_norm": 13.3125,
+      "learning_rate": 6.522870869923943e-06,
+      "loss": 1.5906,
+      "step": 35860
+    },
+    {
+      "epoch": 1.0437210925909766,
+      "grad_norm": 9.5625,
+      "learning_rate": 6.520931588251328e-06,
+      "loss": 1.4507,
+      "step": 35880
+    },
+    {
+      "epoch": 1.0443028769235245,
+      "grad_norm": 12.75,
+      "learning_rate": 6.518992306578713e-06,
+      "loss": 1.4307,
+      "step": 35900
+    },
+    {
+      "epoch": 1.0448846612560723,
+      "grad_norm": 11.375,
+      "learning_rate": 6.517053024906098e-06,
+      "loss": 1.6122,
+      "step": 35920
+    },
+    {
+      "epoch": 1.0454664455886202,
+      "grad_norm": 13.25,
+      "learning_rate": 6.515113743233483e-06,
+      "loss": 1.5159,
+      "step": 35940
+    },
+    {
+      "epoch": 1.0460482299211682,
+      "grad_norm": 13.4375,
+      "learning_rate": 6.513174461560868e-06,
+      "loss": 1.4978,
+      "step": 35960
+    },
+    {
+      "epoch": 1.0466300142537162,
+      "grad_norm": 10.9375,
+      "learning_rate": 6.511235179888253e-06,
+      "loss": 1.547,
+      "step": 35980
+    },
+    {
+      "epoch": 1.0472117985862641,
+      "grad_norm": 14.0,
+      "learning_rate": 6.509295898215638e-06,
+      "loss": 1.5661,
+      "step": 36000
+    },
+    {
+      "epoch": 1.047793582918812,
+      "grad_norm": 13.3125,
+      "learning_rate": 6.507356616543023e-06,
+      "loss": 1.5157,
+      "step": 36020
+    },
+    {
+      "epoch": 1.0483753672513598,
+      "grad_norm": 11.5625,
+      "learning_rate": 6.5054173348704085e-06,
+      "loss": 1.5278,
+      "step": 36040
+    },
+    {
+      "epoch": 1.0489571515839078,
+      "grad_norm": 11.125,
+      "learning_rate": 6.503478053197794e-06,
+      "loss": 1.4772,
+      "step": 36060
+    },
+    {
+      "epoch": 1.0495389359164558,
+      "grad_norm": 14.75,
+      "learning_rate": 6.501538771525179e-06,
+      "loss": 1.5065,
+      "step": 36080
+    },
+    {
+      "epoch": 1.0501207202490037,
+      "grad_norm": 11.625,
+      "learning_rate": 6.499599489852564e-06,
+      "loss": 1.49,
+      "step": 36100
+    },
+    {
+      "epoch": 1.0507025045815517,
+      "grad_norm": 12.8125,
+      "learning_rate": 6.497660208179949e-06,
+      "loss": 1.5775,
+      "step": 36120
+    },
+    {
+      "epoch": 1.0512842889140996,
+      "grad_norm": 11.8125,
+      "learning_rate": 6.495720926507334e-06,
+      "loss": 1.5412,
+      "step": 36140
+    },
+    {
+      "epoch": 1.0518660732466474,
+      "grad_norm": 15.125,
+      "learning_rate": 6.493781644834719e-06,
+      "loss": 1.5165,
+      "step": 36160
+    },
+    {
+      "epoch": 1.0524478575791953,
+      "grad_norm": 13.875,
+      "learning_rate": 6.491842363162104e-06,
+      "loss": 1.5471,
+      "step": 36180
+    },
+    {
+      "epoch": 1.0530296419117433,
+      "grad_norm": 15.5,
+      "learning_rate": 6.489903081489489e-06,
+      "loss": 1.4839,
+      "step": 36200
+    },
+    {
+      "epoch": 1.0536114262442913,
+      "grad_norm": 13.875,
+      "learning_rate": 6.487963799816874e-06,
+      "loss": 1.5373,
+      "step": 36220
+    },
+    {
+      "epoch": 1.0541932105768392,
+      "grad_norm": 12.8125,
+      "learning_rate": 6.4860245181442595e-06,
+      "loss": 1.5859,
+      "step": 36240
+    },
+    {
+      "epoch": 1.054774994909387,
+      "grad_norm": 14.5625,
+      "learning_rate": 6.484085236471645e-06,
+      "loss": 1.5399,
+      "step": 36260
+    },
+    {
+      "epoch": 1.055356779241935,
+      "grad_norm": 12.3125,
+      "learning_rate": 6.48214595479903e-06,
+      "loss": 1.5187,
+      "step": 36280
+    },
+    {
+      "epoch": 1.055938563574483,
+      "grad_norm": 11.1875,
+      "learning_rate": 6.480206673126415e-06,
+      "loss": 1.503,
+      "step": 36300
+    },
+    {
+      "epoch": 1.0565203479070309,
+      "grad_norm": 12.875,
+      "learning_rate": 6.4782673914538e-06,
+      "loss": 1.5074,
+      "step": 36320
+    },
+    {
+      "epoch": 1.0571021322395788,
+      "grad_norm": 13.875,
+      "learning_rate": 6.476328109781185e-06,
+      "loss": 1.5128,
+      "step": 36340
+    },
+    {
+      "epoch": 1.0576839165721268,
+      "grad_norm": 12.0,
+      "learning_rate": 6.47438882810857e-06,
+      "loss": 1.5213,
+      "step": 36360
+    },
+    {
+      "epoch": 1.0582657009046745,
+      "grad_norm": 11.6875,
+      "learning_rate": 6.472449546435955e-06,
+      "loss": 1.5648,
+      "step": 36380
+    },
+    {
+      "epoch": 1.0588474852372225,
+      "grad_norm": 13.5625,
+      "learning_rate": 6.47051026476334e-06,
+      "loss": 1.5782,
+      "step": 36400
+    },
+    {
+      "epoch": 1.0594292695697705,
+      "grad_norm": 14.625,
+      "learning_rate": 6.468570983090725e-06,
+      "loss": 1.6054,
+      "step": 36420
+    },
+    {
+      "epoch": 1.0600110539023184,
+      "grad_norm": 14.8125,
+      "learning_rate": 6.4666317014181105e-06,
+      "loss": 1.5262,
+      "step": 36440
+    },
+    {
+      "epoch": 1.0605928382348664,
+      "grad_norm": 15.6875,
+      "learning_rate": 6.4646924197454956e-06,
+      "loss": 1.5041,
+      "step": 36460
+    },
+    {
+      "epoch": 1.0611746225674143,
+      "grad_norm": 12.0,
+      "learning_rate": 6.462753138072881e-06,
+      "loss": 1.4741,
+      "step": 36480
+    },
+    {
+      "epoch": 1.061756406899962,
+      "grad_norm": 13.6875,
+      "learning_rate": 6.460813856400266e-06,
+      "loss": 1.4637,
+      "step": 36500
+    },
+    {
+      "epoch": 1.06233819123251,
+      "grad_norm": 12.25,
+      "learning_rate": 6.458874574727651e-06,
+      "loss": 1.5187,
+      "step": 36520
+    },
+    {
+      "epoch": 1.062919975565058,
+      "grad_norm": 12.9375,
+      "learning_rate": 6.456935293055036e-06,
+      "loss": 1.4297,
+      "step": 36540
+    },
+    {
+      "epoch": 1.063501759897606,
+      "grad_norm": 10.25,
+      "learning_rate": 6.45499601138242e-06,
+      "loss": 1.499,
+      "step": 36560
+    },
+    {
+      "epoch": 1.064083544230154,
+      "grad_norm": 11.125,
+      "learning_rate": 6.453056729709805e-06,
+      "loss": 1.5329,
+      "step": 36580
+    },
+    {
+      "epoch": 1.064665328562702,
+      "grad_norm": 12.0625,
+      "learning_rate": 6.45111744803719e-06,
+      "loss": 1.589,
+      "step": 36600
+    },
+    {
+      "epoch": 1.0652471128952496,
+      "grad_norm": 12.875,
+      "learning_rate": 6.4491781663645755e-06,
+      "loss": 1.4859,
+      "step": 36620
+    },
+    {
+      "epoch": 1.0658288972277976,
+      "grad_norm": 12.5625,
+      "learning_rate": 6.447238884691961e-06,
+      "loss": 1.4984,
+      "step": 36640
+    },
+    {
+      "epoch": 1.0664106815603456,
+      "grad_norm": 12.375,
+      "learning_rate": 6.445299603019346e-06,
+      "loss": 1.5674,
+      "step": 36660
+    },
+    {
+      "epoch": 1.0669924658928935,
+      "grad_norm": 12.3125,
+      "learning_rate": 6.443360321346731e-06,
+      "loss": 1.5434,
+      "step": 36680
+    },
+    {
+      "epoch": 1.0675742502254415,
+      "grad_norm": 10.4375,
+      "learning_rate": 6.441421039674116e-06,
+      "loss": 1.4657,
+      "step": 36700
+    },
+    {
+      "epoch": 1.0681560345579895,
+      "grad_norm": 11.9375,
+      "learning_rate": 6.439481758001501e-06,
+      "loss": 1.4707,
+      "step": 36720
+    },
+    {
+      "epoch": 1.0687378188905372,
+      "grad_norm": 14.6875,
+      "learning_rate": 6.437542476328886e-06,
+      "loss": 1.5764,
+      "step": 36740
+    },
+    {
+      "epoch": 1.0693196032230852,
+      "grad_norm": 12.6875,
+      "learning_rate": 6.435603194656271e-06,
+      "loss": 1.525,
+      "step": 36760
+    },
+    {
+      "epoch": 1.0699013875556331,
+      "grad_norm": 16.125,
+      "learning_rate": 6.433663912983656e-06,
+      "loss": 1.5138,
+      "step": 36780
+    },
+    {
+      "epoch": 1.070483171888181,
+      "grad_norm": 12.0625,
+      "learning_rate": 6.431724631311041e-06,
+      "loss": 1.506,
+      "step": 36800
+    },
+    {
+      "epoch": 1.071064956220729,
+      "grad_norm": 15.5625,
+      "learning_rate": 6.4297853496384265e-06,
+      "loss": 1.5367,
+      "step": 36820
+    },
+    {
+      "epoch": 1.0716467405532768,
+      "grad_norm": 15.4375,
+      "learning_rate": 6.4278460679658116e-06,
+      "loss": 1.4997,
+      "step": 36840
+    },
+    {
+      "epoch": 1.0722285248858248,
+      "grad_norm": 12.5,
+      "learning_rate": 6.425906786293197e-06,
+      "loss": 1.5785,
+      "step": 36860
+    },
+    {
+      "epoch": 1.0728103092183727,
+      "grad_norm": 14.0625,
+      "learning_rate": 6.423967504620582e-06,
+      "loss": 1.4386,
+      "step": 36880
+    },
+    {
+      "epoch": 1.0733920935509207,
+      "grad_norm": 14.75,
+      "learning_rate": 6.422028222947967e-06,
+      "loss": 1.4992,
+      "step": 36900
+    },
+    {
+      "epoch": 1.0739738778834687,
+      "grad_norm": 11.8125,
+      "learning_rate": 6.420088941275352e-06,
+      "loss": 1.5329,
+      "step": 36920
+    },
+    {
+      "epoch": 1.0745556622160166,
+      "grad_norm": 11.125,
+      "learning_rate": 6.418149659602737e-06,
+      "loss": 1.5295,
+      "step": 36940
+    },
+    {
+      "epoch": 1.0751374465485644,
+      "grad_norm": 11.625,
+      "learning_rate": 6.416210377930122e-06,
+      "loss": 1.6297,
+      "step": 36960
+    },
+    {
+      "epoch": 1.0757192308811123,
+      "grad_norm": 14.1875,
+      "learning_rate": 6.414271096257507e-06,
+      "loss": 1.5128,
+      "step": 36980
+    },
+    {
+      "epoch": 1.0763010152136603,
+      "grad_norm": 13.625,
+      "learning_rate": 6.412331814584892e-06,
+      "loss": 1.4886,
+      "step": 37000
+    },
+    {
+      "epoch": 1.0768827995462082,
+      "grad_norm": 14.125,
+      "learning_rate": 6.4103925329122774e-06,
+      "loss": 1.5572,
+      "step": 37020
+    },
+    {
+      "epoch": 1.0774645838787562,
+      "grad_norm": 11.875,
+      "learning_rate": 6.4084532512396625e-06,
+      "loss": 1.4906,
+      "step": 37040
+    },
+    {
+      "epoch": 1.0780463682113042,
+      "grad_norm": 12.8125,
+      "learning_rate": 6.406513969567048e-06,
+      "loss": 1.5079,
+      "step": 37060
+    },
+    {
+      "epoch": 1.078628152543852,
+      "grad_norm": 11.5,
+      "learning_rate": 6.404574687894433e-06,
+      "loss": 1.5033,
+      "step": 37080
+    },
+    {
+      "epoch": 1.0792099368763999,
+      "grad_norm": 15.875,
+      "learning_rate": 6.402635406221818e-06,
+      "loss": 1.5184,
+      "step": 37100
+    },
+    {
+      "epoch": 1.0797917212089478,
+      "grad_norm": 13.75,
+      "learning_rate": 6.400696124549203e-06,
+      "loss": 1.5089,
+      "step": 37120
+    },
+    {
+      "epoch": 1.0803735055414958,
+      "grad_norm": 13.3125,
+      "learning_rate": 6.398756842876588e-06,
+      "loss": 1.5207,
+      "step": 37140
+    },
+    {
+      "epoch": 1.0809552898740438,
+      "grad_norm": 13.25,
+      "learning_rate": 6.396817561203973e-06,
+      "loss": 1.5109,
+      "step": 37160
+    },
+    {
+      "epoch": 1.0815370742065915,
+      "grad_norm": 12.5,
+      "learning_rate": 6.394878279531358e-06,
+      "loss": 1.5187,
+      "step": 37180
+    },
+    {
+      "epoch": 1.0821188585391395,
+      "grad_norm": 13.9375,
+      "learning_rate": 6.392938997858743e-06,
+      "loss": 1.494,
+      "step": 37200
+    },
+    {
+      "epoch": 1.0827006428716874,
+      "grad_norm": 11.875,
+      "learning_rate": 6.390999716186128e-06,
+      "loss": 1.4747,
+      "step": 37220
+    },
+    {
+      "epoch": 1.0832824272042354,
+      "grad_norm": 10.8125,
+      "learning_rate": 6.3890604345135135e-06,
+      "loss": 1.4979,
+      "step": 37240
+    },
+    {
+      "epoch": 1.0838642115367834,
+      "grad_norm": 14.5,
+      "learning_rate": 6.387121152840899e-06,
+      "loss": 1.5181,
+      "step": 37260
+    },
+    {
+      "epoch": 1.0844459958693313,
+      "grad_norm": 19.75,
+      "learning_rate": 6.385181871168284e-06,
+      "loss": 1.5811,
+      "step": 37280
+    },
+    {
+      "epoch": 1.085027780201879,
+      "grad_norm": 10.9375,
+      "learning_rate": 6.383242589495669e-06,
+      "loss": 1.5239,
+      "step": 37300
+    },
+    {
+      "epoch": 1.085609564534427,
+      "grad_norm": 14.1875,
+      "learning_rate": 6.381303307823054e-06,
+      "loss": 1.4841,
+      "step": 37320
+    },
+    {
+      "epoch": 1.086191348866975,
+      "grad_norm": 12.5,
+      "learning_rate": 6.379364026150439e-06,
+      "loss": 1.539,
+      "step": 37340
+    },
+    {
+      "epoch": 1.086773133199523,
+      "grad_norm": 16.0,
+      "learning_rate": 6.377424744477823e-06,
+      "loss": 1.495,
+      "step": 37360
+    },
+    {
+      "epoch": 1.087354917532071,
+      "grad_norm": 13.8125,
+      "learning_rate": 6.375485462805208e-06,
+      "loss": 1.5655,
+      "step": 37380
+    },
+    {
+      "epoch": 1.0879367018646189,
+      "grad_norm": 13.5,
+      "learning_rate": 6.3735461811325934e-06,
+      "loss": 1.5101,
+      "step": 37400
+    },
+    {
+      "epoch": 1.0885184861971666,
+      "grad_norm": 12.5,
+      "learning_rate": 6.3716068994599785e-06,
+      "loss": 1.446,
+      "step": 37420
+    },
+    {
+      "epoch": 1.0891002705297146,
+      "grad_norm": 9.9375,
+      "learning_rate": 6.369667617787364e-06,
+      "loss": 1.4292,
+      "step": 37440
+    },
+    {
+      "epoch": 1.0896820548622626,
+      "grad_norm": 12.125,
+      "learning_rate": 6.367728336114749e-06,
+      "loss": 1.4819,
+      "step": 37460
+    },
+    {
+      "epoch": 1.0902638391948105,
+      "grad_norm": 13.75,
+      "learning_rate": 6.365789054442134e-06,
+      "loss": 1.4845,
+      "step": 37480
+    },
+    {
+      "epoch": 1.0908456235273585,
+      "grad_norm": 12.8125,
+      "learning_rate": 6.363849772769519e-06,
+      "loss": 1.5017,
+      "step": 37500
+    },
+    {
+      "epoch": 1.0914274078599062,
+      "grad_norm": 13.0,
+      "learning_rate": 6.361910491096904e-06,
+      "loss": 1.4634,
+      "step": 37520
+    },
+    {
+      "epoch": 1.0920091921924542,
+      "grad_norm": 12.0,
+      "learning_rate": 6.359971209424289e-06,
+      "loss": 1.4176,
+      "step": 37540
+    },
+    {
+      "epoch": 1.0925909765250021,
+      "grad_norm": 10.5,
+      "learning_rate": 6.358031927751673e-06,
+      "loss": 1.5057,
+      "step": 37560
+    },
+    {
+      "epoch": 1.09317276085755,
+      "grad_norm": 14.25,
+      "learning_rate": 6.3560926460790585e-06,
+      "loss": 1.4842,
+      "step": 37580
+    },
+    {
+      "epoch": 1.093754545190098,
+      "grad_norm": 12.1875,
+      "learning_rate": 6.3541533644064436e-06,
+      "loss": 1.5763,
+      "step": 37600
+    },
+    {
+      "epoch": 1.094336329522646,
+      "grad_norm": 14.6875,
+      "learning_rate": 6.352214082733829e-06,
+      "loss": 1.6831,
+      "step": 37620
+    },
+    {
+      "epoch": 1.094918113855194,
+      "grad_norm": 11.375,
+      "learning_rate": 6.350274801061214e-06,
+      "loss": 1.5598,
+      "step": 37640
+    },
+    {
+      "epoch": 1.0954998981877417,
+      "grad_norm": 13.375,
+      "learning_rate": 6.348335519388599e-06,
+      "loss": 1.4496,
+      "step": 37660
+    },
+    {
+      "epoch": 1.0960816825202897,
+      "grad_norm": 13.1875,
+      "learning_rate": 6.346396237715984e-06,
+      "loss": 1.5174,
+      "step": 37680
+    },
+    {
+      "epoch": 1.0966634668528377,
+      "grad_norm": 14.375,
+      "learning_rate": 6.344456956043369e-06,
+      "loss": 1.5247,
+      "step": 37700
+    },
+    {
+      "epoch": 1.0972452511853856,
+      "grad_norm": 13.5,
+      "learning_rate": 6.342517674370754e-06,
+      "loss": 1.5636,
+      "step": 37720
+    },
+    {
+      "epoch": 1.0978270355179336,
+      "grad_norm": 14.8125,
+      "learning_rate": 6.340578392698139e-06,
+      "loss": 1.5537,
+      "step": 37740
+    },
+    {
+      "epoch": 1.0984088198504813,
+      "grad_norm": 13.5,
+      "learning_rate": 6.338639111025524e-06,
+      "loss": 1.5261,
+      "step": 37760
+    },
+    {
+      "epoch": 1.0989906041830293,
+      "grad_norm": 12.0625,
+      "learning_rate": 6.3366998293529094e-06,
+      "loss": 1.5312,
+      "step": 37780
+    },
+    {
+      "epoch": 1.0995723885155773,
+      "grad_norm": 11.9375,
+      "learning_rate": 6.3347605476802945e-06,
+      "loss": 1.5345,
+      "step": 37800
+    },
+    {
+      "epoch": 1.1001541728481252,
+      "grad_norm": 12.6875,
+      "learning_rate": 6.33282126600768e-06,
+      "loss": 1.6277,
+      "step": 37820
+    },
+    {
+      "epoch": 1.1007359571806732,
+      "grad_norm": 13.625,
+      "learning_rate": 6.330881984335065e-06,
+      "loss": 1.5504,
+      "step": 37840
+    },
+    {
+      "epoch": 1.1013177415132211,
+      "grad_norm": 11.5625,
+      "learning_rate": 6.32894270266245e-06,
+      "loss": 1.4967,
+      "step": 37860
+    },
+    {
+      "epoch": 1.101899525845769,
+      "grad_norm": 12.125,
+      "learning_rate": 6.327003420989835e-06,
+      "loss": 1.5639,
+      "step": 37880
+    },
+    {
+      "epoch": 1.1024813101783169,
+      "grad_norm": 14.375,
+      "learning_rate": 6.32506413931722e-06,
+      "loss": 1.5901,
+      "step": 37900
+    },
+    {
+      "epoch": 1.1030630945108648,
+      "grad_norm": 13.375,
+      "learning_rate": 6.323124857644605e-06,
+      "loss": 1.5309,
+      "step": 37920
+    },
+    {
+      "epoch": 1.1036448788434128,
+      "grad_norm": 13.5,
+      "learning_rate": 6.321185575971989e-06,
+      "loss": 1.5142,
+      "step": 37940
+    },
+    {
+      "epoch": 1.1042266631759607,
+      "grad_norm": 12.875,
+      "learning_rate": 6.3192462942993745e-06,
+      "loss": 1.5386,
+      "step": 37960
+    },
+    {
+      "epoch": 1.1048084475085087,
+      "grad_norm": 12.5,
+      "learning_rate": 6.3173070126267596e-06,
+      "loss": 1.5329,
+      "step": 37980
+    },
+    {
+      "epoch": 1.1053902318410564,
+      "grad_norm": 12.125,
+      "learning_rate": 6.315367730954145e-06,
+      "loss": 1.5415,
+      "step": 38000
+    },
+    {
+      "epoch": 1.1059720161736044,
+      "grad_norm": 13.9375,
+      "learning_rate": 6.31342844928153e-06,
+      "loss": 1.5508,
+      "step": 38020
+    },
+    {
+      "epoch": 1.1065538005061524,
+      "grad_norm": 16.0,
+      "learning_rate": 6.311489167608915e-06,
+      "loss": 1.5365,
+      "step": 38040
+    },
+    {
+      "epoch": 1.1071355848387003,
+      "grad_norm": 12.0,
+      "learning_rate": 6.3095498859363e-06,
+      "loss": 1.5612,
+      "step": 38060
+    },
+    {
+      "epoch": 1.1077173691712483,
+      "grad_norm": 12.0,
+      "learning_rate": 6.307610604263685e-06,
+      "loss": 1.46,
+      "step": 38080
+    },
+    {
+      "epoch": 1.108299153503796,
+      "grad_norm": 12.4375,
+      "learning_rate": 6.30567132259107e-06,
+      "loss": 1.5787,
+      "step": 38100
+    },
+    {
+      "epoch": 1.108880937836344,
+      "grad_norm": 13.25,
+      "learning_rate": 6.303732040918455e-06,
+      "loss": 1.543,
+      "step": 38120
+    },
+    {
+      "epoch": 1.109462722168892,
+      "grad_norm": 12.375,
+      "learning_rate": 6.30179275924584e-06,
+      "loss": 1.4542,
+      "step": 38140
+    },
+    {
+      "epoch": 1.11004450650144,
+      "grad_norm": 12.625,
+      "learning_rate": 6.2998534775732255e-06,
+      "loss": 1.4819,
+      "step": 38160
+    },
+    {
+      "epoch": 1.110626290833988,
+      "grad_norm": 13.0,
+      "learning_rate": 6.2979141959006105e-06,
+      "loss": 1.5394,
+      "step": 38180
+    },
+    {
+      "epoch": 1.1112080751665359,
+      "grad_norm": 12.9375,
+      "learning_rate": 6.295974914227996e-06,
+      "loss": 1.6128,
+      "step": 38200
+    },
+    {
+      "epoch": 1.1117898594990836,
+      "grad_norm": 12.0,
+      "learning_rate": 6.294035632555381e-06,
+      "loss": 1.4606,
+      "step": 38220
+    },
+    {
+      "epoch": 1.1123716438316316,
+      "grad_norm": 14.75,
+      "learning_rate": 6.292096350882766e-06,
+      "loss": 1.4817,
+      "step": 38240
+    },
+    {
+      "epoch": 1.1129534281641795,
+      "grad_norm": 13.75,
+      "learning_rate": 6.290157069210151e-06,
+      "loss": 1.5162,
+      "step": 38260
+    },
+    {
+      "epoch": 1.1135352124967275,
+      "grad_norm": 12.4375,
+      "learning_rate": 6.288217787537536e-06,
+      "loss": 1.4722,
+      "step": 38280
+    },
+    {
+      "epoch": 1.1141169968292755,
+      "grad_norm": 11.75,
+      "learning_rate": 6.286278505864921e-06,
+      "loss": 1.4585,
+      "step": 38300
+    },
+    {
+      "epoch": 1.1146987811618234,
+      "grad_norm": 11.9375,
+      "learning_rate": 6.284339224192306e-06,
+      "loss": 1.5375,
+      "step": 38320
+    },
+    {
+      "epoch": 1.1152805654943712,
+      "grad_norm": 13.1875,
+      "learning_rate": 6.282399942519691e-06,
+      "loss": 1.5263,
+      "step": 38340
+    },
+    {
+      "epoch": 1.1158623498269191,
+      "grad_norm": 11.5625,
+      "learning_rate": 6.2804606608470764e-06,
+      "loss": 1.5259,
+      "step": 38360
+    },
+    {
+      "epoch": 1.116444134159467,
+      "grad_norm": 12.5,
+      "learning_rate": 6.2785213791744615e-06,
+      "loss": 1.4978,
+      "step": 38380
+    },
+    {
+      "epoch": 1.117025918492015,
+      "grad_norm": 13.4375,
+      "learning_rate": 6.276582097501847e-06,
+      "loss": 1.4903,
+      "step": 38400
+    },
+    {
+      "epoch": 1.117607702824563,
+      "grad_norm": 12.0,
+      "learning_rate": 6.274642815829232e-06,
+      "loss": 1.5991,
+      "step": 38420
+    },
+    {
+      "epoch": 1.1181894871571108,
+      "grad_norm": 16.625,
+      "learning_rate": 6.272703534156617e-06,
+      "loss": 1.5308,
+      "step": 38440
+    },
+    {
+      "epoch": 1.1187712714896587,
+      "grad_norm": 17.125,
+      "learning_rate": 6.270764252484002e-06,
+      "loss": 1.4371,
+      "step": 38460
+    },
+    {
+      "epoch": 1.1193530558222067,
+      "grad_norm": 13.25,
+      "learning_rate": 6.268824970811387e-06,
+      "loss": 1.5051,
+      "step": 38480
+    },
+    {
+      "epoch": 1.1199348401547546,
+      "grad_norm": 12.0,
+      "learning_rate": 6.266885689138772e-06,
+      "loss": 1.5281,
+      "step": 38500
+    },
+    {
+      "epoch": 1.1205166244873026,
+      "grad_norm": 14.125,
+      "learning_rate": 6.264946407466157e-06,
+      "loss": 1.6,
+      "step": 38520
+    },
+    {
+      "epoch": 1.1210984088198506,
+      "grad_norm": 12.5,
+      "learning_rate": 6.263007125793542e-06,
+      "loss": 1.4649,
+      "step": 38540
+    },
+    {
+      "epoch": 1.1216801931523983,
+      "grad_norm": 15.5625,
+      "learning_rate": 6.261067844120927e-06,
+      "loss": 1.5477,
+      "step": 38560
+    },
+    {
+      "epoch": 1.1222619774849463,
+      "grad_norm": 11.8125,
+      "learning_rate": 6.2591285624483125e-06,
+      "loss": 1.5681,
+      "step": 38580
+    },
+    {
+      "epoch": 1.1228437618174942,
+      "grad_norm": 12.0625,
+      "learning_rate": 6.257189280775698e-06,
+      "loss": 1.4782,
+      "step": 38600
+    },
+    {
+      "epoch": 1.1234255461500422,
+      "grad_norm": 14.625,
+      "learning_rate": 6.255249999103083e-06,
+      "loss": 1.5647,
+      "step": 38620
+    },
+    {
+      "epoch": 1.1240073304825902,
+      "grad_norm": 12.6875,
+      "learning_rate": 6.253310717430468e-06,
+      "loss": 1.4799,
+      "step": 38640
+    },
+    {
+      "epoch": 1.1245891148151381,
+      "grad_norm": 11.25,
+      "learning_rate": 6.251371435757853e-06,
+      "loss": 1.4928,
+      "step": 38660
+    },
+    {
+      "epoch": 1.1251708991476859,
+      "grad_norm": 12.6875,
+      "learning_rate": 6.249432154085238e-06,
+      "loss": 1.5306,
+      "step": 38680
+    },
+    {
+      "epoch": 1.1257526834802338,
+      "grad_norm": 13.75,
+      "learning_rate": 6.247492872412623e-06,
+      "loss": 1.4904,
+      "step": 38700
+    },
+    {
+      "epoch": 1.1263344678127818,
+      "grad_norm": 12.875,
+      "learning_rate": 6.245553590740008e-06,
+      "loss": 1.5334,
+      "step": 38720
+    },
+    {
+      "epoch": 1.1269162521453298,
+      "grad_norm": 13.375,
+      "learning_rate": 6.2436143090673924e-06,
+      "loss": 1.5545,
+      "step": 38740
+    },
+    {
+      "epoch": 1.1274980364778777,
+      "grad_norm": 14.3125,
+      "learning_rate": 6.2416750273947775e-06,
+      "loss": 1.5655,
+      "step": 38760
+    },
+    {
+      "epoch": 1.1280798208104255,
+      "grad_norm": 17.0,
+      "learning_rate": 6.239735745722163e-06,
+      "loss": 1.4399,
+      "step": 38780
+    },
+    {
+      "epoch": 1.1286616051429734,
+      "grad_norm": 13.125,
+      "learning_rate": 6.237796464049548e-06,
+      "loss": 1.5477,
+      "step": 38800
+    },
+    {
+      "epoch": 1.1292433894755214,
+      "grad_norm": 12.8125,
+      "learning_rate": 6.235857182376933e-06,
+      "loss": 1.5395,
+      "step": 38820
+    },
+    {
+      "epoch": 1.1298251738080693,
+      "grad_norm": 12.0,
+      "learning_rate": 6.233917900704318e-06,
+      "loss": 1.5085,
+      "step": 38840
+    },
+    {
+      "epoch": 1.1304069581406173,
+      "grad_norm": 17.375,
+      "learning_rate": 6.231978619031703e-06,
+      "loss": 1.5742,
+      "step": 38860
+    },
+    {
+      "epoch": 1.1309887424731653,
+      "grad_norm": 12.6875,
+      "learning_rate": 6.230039337359088e-06,
+      "loss": 1.4685,
+      "step": 38880
+    },
+    {
+      "epoch": 1.1315705268057132,
+      "grad_norm": 12.4375,
+      "learning_rate": 6.228100055686473e-06,
+      "loss": 1.593,
+      "step": 38900
+    },
+    {
+      "epoch": 1.132152311138261,
+      "grad_norm": 12.8125,
+      "learning_rate": 6.226160774013858e-06,
+      "loss": 1.5004,
+      "step": 38920
+    },
+    {
+      "epoch": 1.132734095470809,
+      "grad_norm": 13.9375,
+      "learning_rate": 6.224221492341243e-06,
+      "loss": 1.5268,
+      "step": 38940
+    },
+    {
+      "epoch": 1.133315879803357,
+      "grad_norm": 12.625,
+      "learning_rate": 6.2222822106686285e-06,
+      "loss": 1.5868,
+      "step": 38960
+    },
+    {
+      "epoch": 1.1338976641359049,
+      "grad_norm": 17.125,
+      "learning_rate": 6.220342928996014e-06,
+      "loss": 1.538,
+      "step": 38980
+    },
+    {
+      "epoch": 1.1344794484684528,
+      "grad_norm": 12.5625,
+      "learning_rate": 6.218403647323399e-06,
+      "loss": 1.4494,
+      "step": 39000
+    },
+    {
+      "epoch": 1.1350612328010006,
+      "grad_norm": 13.5,
+      "learning_rate": 6.216464365650784e-06,
+      "loss": 1.5108,
+      "step": 39020
+    },
+    {
+      "epoch": 1.1356430171335485,
+      "grad_norm": 14.25,
+      "learning_rate": 6.214525083978169e-06,
+      "loss": 1.4658,
+      "step": 39040
+    },
+    {
+      "epoch": 1.1362248014660965,
+      "grad_norm": 12.25,
+      "learning_rate": 6.212585802305554e-06,
+      "loss": 1.5474,
+      "step": 39060
+    },
+    {
+      "epoch": 1.1368065857986445,
+      "grad_norm": 13.75,
+      "learning_rate": 6.210646520632939e-06,
+      "loss": 1.5372,
+      "step": 39080
+    },
+    {
+      "epoch": 1.1373883701311924,
+      "grad_norm": 13.125,
+      "learning_rate": 6.208707238960324e-06,
+      "loss": 1.3963,
+      "step": 39100
+    },
+    {
+      "epoch": 1.1379701544637402,
+      "grad_norm": 14.0,
+      "learning_rate": 6.206767957287709e-06,
+      "loss": 1.527,
+      "step": 39120
+    },
+    {
+      "epoch": 1.1385519387962881,
+      "grad_norm": 12.4375,
+      "learning_rate": 6.204828675615094e-06,
+      "loss": 1.5576,
+      "step": 39140
+    },
+    {
+      "epoch": 1.139133723128836,
+      "grad_norm": 12.5625,
+      "learning_rate": 6.2028893939424795e-06,
+      "loss": 1.5017,
+      "step": 39160
+    },
+    {
+      "epoch": 1.139715507461384,
+      "grad_norm": 13.4375,
+      "learning_rate": 6.2009501122698646e-06,
+      "loss": 1.5192,
+      "step": 39180
+    },
+    {
+      "epoch": 1.140297291793932,
+      "grad_norm": 11.75,
+      "learning_rate": 6.19901083059725e-06,
+      "loss": 1.4632,
+      "step": 39200
+    },
+    {
+      "epoch": 1.14087907612648,
+      "grad_norm": 14.5625,
+      "learning_rate": 6.197071548924635e-06,
+      "loss": 1.5165,
+      "step": 39220
+    },
+    {
+      "epoch": 1.141460860459028,
+      "grad_norm": 13.5,
+      "learning_rate": 6.19513226725202e-06,
+      "loss": 1.5024,
+      "step": 39240
+    },
+    {
+      "epoch": 1.1420426447915757,
+      "grad_norm": 15.8125,
+      "learning_rate": 6.193192985579405e-06,
+      "loss": 1.4939,
+      "step": 39260
+    },
+    {
+      "epoch": 1.1426244291241237,
+      "grad_norm": 12.125,
+      "learning_rate": 6.19125370390679e-06,
+      "loss": 1.5697,
+      "step": 39280
+    },
+    {
+      "epoch": 1.1432062134566716,
+      "grad_norm": 14.4375,
+      "learning_rate": 6.189314422234175e-06,
+      "loss": 1.4966,
+      "step": 39300
+    },
+    {
+      "epoch": 1.1437879977892196,
+      "grad_norm": 13.0625,
+      "learning_rate": 6.18737514056156e-06,
+      "loss": 1.4836,
+      "step": 39320
+    },
+    {
+      "epoch": 1.1443697821217675,
+      "grad_norm": 10.4375,
+      "learning_rate": 6.185435858888945e-06,
+      "loss": 1.5408,
+      "step": 39340
+    },
+    {
+      "epoch": 1.1449515664543153,
+      "grad_norm": 13.6875,
+      "learning_rate": 6.1834965772163305e-06,
+      "loss": 1.3676,
+      "step": 39360
+    },
+    {
+      "epoch": 1.1455333507868632,
+      "grad_norm": 11.1875,
+      "learning_rate": 6.1815572955437156e-06,
+      "loss": 1.5173,
+      "step": 39380
+    },
+    {
+      "epoch": 1.1461151351194112,
+      "grad_norm": 10.6875,
+      "learning_rate": 6.179618013871101e-06,
+      "loss": 1.4561,
+      "step": 39400
+    },
+    {
+      "epoch": 1.1466969194519592,
+      "grad_norm": 11.1875,
+      "learning_rate": 6.177678732198486e-06,
+      "loss": 1.4594,
+      "step": 39420
+    },
+    {
+      "epoch": 1.1472787037845071,
+      "grad_norm": 15.375,
+      "learning_rate": 6.175739450525871e-06,
+      "loss": 1.5095,
+      "step": 39440
+    },
+    {
+      "epoch": 1.147860488117055,
+      "grad_norm": 13.5625,
+      "learning_rate": 6.173800168853256e-06,
+      "loss": 1.5298,
+      "step": 39460
+    },
+    {
+      "epoch": 1.1484422724496028,
+      "grad_norm": 12.75,
+      "learning_rate": 6.171860887180641e-06,
+      "loss": 1.4985,
+      "step": 39480
+    },
+    {
+      "epoch": 1.1490240567821508,
+      "grad_norm": 11.4375,
+      "learning_rate": 6.169921605508026e-06,
+      "loss": 1.5058,
+      "step": 39500
+    },
+    {
+      "epoch": 1.1496058411146988,
+      "grad_norm": 14.4375,
+      "learning_rate": 6.167982323835411e-06,
+      "loss": 1.5532,
+      "step": 39520
+    },
+    {
+      "epoch": 1.1501876254472467,
+      "grad_norm": 13.25,
+      "learning_rate": 6.166043042162796e-06,
+      "loss": 1.5766,
+      "step": 39540
+    },
+    {
+      "epoch": 1.1507694097797947,
+      "grad_norm": 15.5,
+      "learning_rate": 6.164103760490181e-06,
+      "loss": 1.5411,
+      "step": 39560
+    },
+    {
+      "epoch": 1.1513511941123427,
+      "grad_norm": 15.5625,
+      "learning_rate": 6.162164478817566e-06,
+      "loss": 1.6242,
+      "step": 39580
+    },
+    {
+      "epoch": 1.1519329784448904,
+      "grad_norm": 10.4375,
+      "learning_rate": 6.160225197144951e-06,
+      "loss": 1.4599,
+      "step": 39600
+    },
+    {
+      "epoch": 1.1525147627774384,
+      "grad_norm": 10.5,
+      "learning_rate": 6.158285915472336e-06,
+      "loss": 1.5049,
+      "step": 39620
+    },
+    {
+      "epoch": 1.1530965471099863,
+      "grad_norm": 13.875,
+      "learning_rate": 6.156346633799721e-06,
+      "loss": 1.5083,
+      "step": 39640
+    },
+    {
+      "epoch": 1.1536783314425343,
+      "grad_norm": 15.5625,
+      "learning_rate": 6.154407352127106e-06,
+      "loss": 1.578,
+      "step": 39660
+    },
+    {
+      "epoch": 1.1542601157750823,
+      "grad_norm": 13.3125,
+      "learning_rate": 6.152468070454491e-06,
+      "loss": 1.5239,
+      "step": 39680
+    },
+    {
+      "epoch": 1.15484190010763,
+      "grad_norm": 16.25,
+      "learning_rate": 6.150528788781876e-06,
+      "loss": 1.5221,
+      "step": 39700
+    },
+    {
+      "epoch": 1.155423684440178,
+      "grad_norm": 12.25,
+      "learning_rate": 6.148589507109261e-06,
+      "loss": 1.5173,
+      "step": 39720
+    },
+    {
+      "epoch": 1.156005468772726,
+      "grad_norm": 12.25,
+      "learning_rate": 6.1466502254366465e-06,
+      "loss": 1.5572,
+      "step": 39740
+    },
+    {
+      "epoch": 1.1565872531052739,
+      "grad_norm": 12.5625,
+      "learning_rate": 6.1447109437640316e-06,
+      "loss": 1.5101,
+      "step": 39760
+    },
+    {
+      "epoch": 1.1571690374378218,
+      "grad_norm": 12.6875,
+      "learning_rate": 6.142771662091417e-06,
+      "loss": 1.4834,
+      "step": 39780
+    },
+    {
+      "epoch": 1.1577508217703698,
+      "grad_norm": 11.0,
+      "learning_rate": 6.140832380418802e-06,
+      "loss": 1.42,
+      "step": 39800
+    },
+    {
+      "epoch": 1.1583326061029178,
+      "grad_norm": 12.875,
+      "learning_rate": 6.138893098746187e-06,
+      "loss": 1.5027,
+      "step": 39820
+    },
+    {
+      "epoch": 1.1589143904354655,
+      "grad_norm": 13.0625,
+      "learning_rate": 6.136953817073572e-06,
+      "loss": 1.4606,
+      "step": 39840
+    },
+    {
+      "epoch": 1.1594961747680135,
+      "grad_norm": 15.1875,
+      "learning_rate": 6.135014535400957e-06,
+      "loss": 1.4864,
+      "step": 39860
+    },
+    {
+      "epoch": 1.1600779591005614,
+      "grad_norm": 12.4375,
+      "learning_rate": 6.133075253728342e-06,
+      "loss": 1.5325,
+      "step": 39880
+    },
+    {
+      "epoch": 1.1606597434331094,
+      "grad_norm": 13.625,
+      "learning_rate": 6.131135972055727e-06,
+      "loss": 1.6125,
+      "step": 39900
+    },
+    {
+      "epoch": 1.1612415277656574,
+      "grad_norm": 13.5,
+      "learning_rate": 6.129196690383112e-06,
+      "loss": 1.4932,
+      "step": 39920
+    },
+    {
+      "epoch": 1.161823312098205,
+      "grad_norm": 15.75,
+      "learning_rate": 6.1272574087104974e-06,
+      "loss": 1.5064,
+      "step": 39940
+    },
+    {
+      "epoch": 1.162405096430753,
+      "grad_norm": 13.6875,
+      "learning_rate": 6.1253181270378825e-06,
+      "loss": 1.4966,
+      "step": 39960
+    },
+    {
+      "epoch": 1.162986880763301,
+      "grad_norm": 14.75,
+      "learning_rate": 6.123378845365268e-06,
+      "loss": 1.5413,
+      "step": 39980
+    },
+    {
+      "epoch": 1.163568665095849,
+      "grad_norm": 16.75,
+      "learning_rate": 6.121439563692653e-06,
+      "loss": 1.4682,
+      "step": 40000
+    },
+    {
+      "epoch": 1.164150449428397,
+      "grad_norm": 13.5625,
+      "learning_rate": 6.119500282020038e-06,
+      "loss": 1.4839,
+      "step": 40020
+    },
+    {
+      "epoch": 1.1647322337609447,
+      "grad_norm": 14.25,
+      "learning_rate": 6.117561000347423e-06,
+      "loss": 1.5623,
+      "step": 40040
+    },
+    {
+      "epoch": 1.1653140180934927,
+      "grad_norm": 12.25,
+      "learning_rate": 6.115621718674808e-06,
+      "loss": 1.4892,
+      "step": 40060
+    },
+    {
+      "epoch": 1.1658958024260406,
+      "grad_norm": 12.3125,
+      "learning_rate": 6.113682437002193e-06,
+      "loss": 1.5352,
+      "step": 40080
+    },
+    {
+      "epoch": 1.1664775867585886,
+      "grad_norm": 12.375,
+      "learning_rate": 6.111743155329578e-06,
+      "loss": 1.5495,
+      "step": 40100
+    },
+    {
+      "epoch": 1.1670593710911366,
+      "grad_norm": 13.375,
+      "learning_rate": 6.109803873656963e-06,
+      "loss": 1.4987,
+      "step": 40120
+    },
+    {
+      "epoch": 1.1676411554236845,
+      "grad_norm": 14.0625,
+      "learning_rate": 6.107864591984348e-06,
+      "loss": 1.5625,
+      "step": 40140
+    },
+    {
+      "epoch": 1.1682229397562325,
+      "grad_norm": 14.125,
+      "learning_rate": 6.1059253103117335e-06,
+      "loss": 1.5876,
+      "step": 40160
+    },
+    {
+      "epoch": 1.1688047240887802,
+      "grad_norm": 11.75,
+      "learning_rate": 6.103986028639119e-06,
+      "loss": 1.5188,
+      "step": 40180
+    },
+    {
+      "epoch": 1.1693865084213282,
+      "grad_norm": 14.4375,
+      "learning_rate": 6.102046746966504e-06,
+      "loss": 1.5524,
+      "step": 40200
+    },
+    {
+      "epoch": 1.1699682927538761,
+      "grad_norm": 16.375,
+      "learning_rate": 6.100107465293889e-06,
+      "loss": 1.488,
+      "step": 40220
+    },
+    {
+      "epoch": 1.1705500770864241,
+      "grad_norm": 12.5,
+      "learning_rate": 6.098168183621274e-06,
+      "loss": 1.4927,
+      "step": 40240
+    },
+    {
+      "epoch": 1.171131861418972,
+      "grad_norm": 10.6875,
+      "learning_rate": 6.096228901948659e-06,
+      "loss": 1.493,
+      "step": 40260
+    },
+    {
+      "epoch": 1.1717136457515198,
+      "grad_norm": 11.1875,
+      "learning_rate": 6.094289620276044e-06,
+      "loss": 1.4916,
+      "step": 40280
+    },
+    {
+      "epoch": 1.1722954300840678,
+      "grad_norm": 13.4375,
+      "learning_rate": 6.092350338603429e-06,
+      "loss": 1.5093,
+      "step": 40300
+    },
+    {
+      "epoch": 1.1728772144166157,
+      "grad_norm": 12.5625,
+      "learning_rate": 6.090411056930814e-06,
+      "loss": 1.4595,
+      "step": 40320
+    },
+    {
+      "epoch": 1.1734589987491637,
+      "grad_norm": 15.875,
+      "learning_rate": 6.088471775258199e-06,
+      "loss": 1.6267,
+      "step": 40340
+    },
+    {
+      "epoch": 1.1740407830817117,
+      "grad_norm": 13.875,
+      "learning_rate": 6.0865324935855845e-06,
+      "loss": 1.5344,
+      "step": 40360
+    },
+    {
+      "epoch": 1.1746225674142594,
+      "grad_norm": 14.3125,
+      "learning_rate": 6.084593211912969e-06,
+      "loss": 1.5923,
+      "step": 40380
+    },
+    {
+      "epoch": 1.1752043517468074,
+      "grad_norm": 13.5,
+      "learning_rate": 6.082653930240354e-06,
+      "loss": 1.4998,
+      "step": 40400
+    },
+    {
+      "epoch": 1.1757861360793553,
+      "grad_norm": 13.5625,
+      "learning_rate": 6.080714648567739e-06,
+      "loss": 1.4985,
+      "step": 40420
+    },
+    {
+      "epoch": 1.1763679204119033,
+      "grad_norm": 14.875,
+      "learning_rate": 6.078775366895124e-06,
+      "loss": 1.4674,
+      "step": 40440
+    },
+    {
+      "epoch": 1.1769497047444513,
+      "grad_norm": 13.6875,
+      "learning_rate": 6.076836085222509e-06,
+      "loss": 1.5184,
+      "step": 40460
+    },
+    {
+      "epoch": 1.1775314890769992,
+      "grad_norm": 14.5,
+      "learning_rate": 6.074896803549894e-06,
+      "loss": 1.4749,
+      "step": 40480
+    },
+    {
+      "epoch": 1.1781132734095472,
+      "grad_norm": 13.375,
+      "learning_rate": 6.072957521877279e-06,
+      "loss": 1.6339,
+      "step": 40500
+    },
+    {
+      "epoch": 1.178695057742095,
+      "grad_norm": 12.5625,
+      "learning_rate": 6.071018240204664e-06,
+      "loss": 1.4874,
+      "step": 40520
+    },
+    {
+      "epoch": 1.179276842074643,
+      "grad_norm": 12.0625,
+      "learning_rate": 6.0690789585320495e-06,
+      "loss": 1.5409,
+      "step": 40540
+    },
+    {
+      "epoch": 1.1798586264071909,
+      "grad_norm": 10.75,
+      "learning_rate": 6.067139676859435e-06,
+      "loss": 1.6283,
+      "step": 40560
+    },
+    {
+      "epoch": 1.1804404107397388,
+      "grad_norm": 12.0,
+      "learning_rate": 6.06520039518682e-06,
+      "loss": 1.5583,
+      "step": 40580
+    },
+    {
+      "epoch": 1.1810221950722868,
+      "grad_norm": 11.25,
+      "learning_rate": 6.063261113514205e-06,
+      "loss": 1.5938,
+      "step": 40600
+    },
+    {
+      "epoch": 1.1816039794048345,
+      "grad_norm": 12.4375,
+      "learning_rate": 6.06132183184159e-06,
+      "loss": 1.5019,
+      "step": 40620
+    },
+    {
+      "epoch": 1.1821857637373825,
+      "grad_norm": 13.875,
+      "learning_rate": 6.059382550168975e-06,
+      "loss": 1.585,
+      "step": 40640
+    },
+    {
+      "epoch": 1.1827675480699305,
+      "grad_norm": 11.0625,
+      "learning_rate": 6.05744326849636e-06,
+      "loss": 1.5273,
+      "step": 40660
+    },
+    {
+      "epoch": 1.1833493324024784,
+      "grad_norm": 10.5625,
+      "learning_rate": 6.055503986823745e-06,
+      "loss": 1.5024,
+      "step": 40680
+    },
+    {
+      "epoch": 1.1839311167350264,
+      "grad_norm": 14.8125,
+      "learning_rate": 6.05356470515113e-06,
+      "loss": 1.5544,
+      "step": 40700
+    },
+    {
+      "epoch": 1.1845129010675743,
+      "grad_norm": 12.9375,
+      "learning_rate": 6.051625423478515e-06,
+      "loss": 1.4653,
+      "step": 40720
+    },
+    {
+      "epoch": 1.185094685400122,
+      "grad_norm": 12.125,
+      "learning_rate": 6.0496861418059005e-06,
+      "loss": 1.5179,
+      "step": 40740
+    },
+    {
+      "epoch": 1.18567646973267,
+      "grad_norm": 11.4375,
+      "learning_rate": 6.047746860133286e-06,
+      "loss": 1.5144,
+      "step": 40760
+    },
+    {
+      "epoch": 1.186258254065218,
+      "grad_norm": 8.75,
+      "learning_rate": 6.045807578460671e-06,
+      "loss": 1.4804,
+      "step": 40780
+    },
+    {
+      "epoch": 1.186840038397766,
+      "grad_norm": 11.125,
+      "learning_rate": 6.043868296788056e-06,
+      "loss": 1.5792,
+      "step": 40800
+    },
+    {
+      "epoch": 1.187421822730314,
+      "grad_norm": 13.875,
+      "learning_rate": 6.041929015115441e-06,
+      "loss": 1.4975,
+      "step": 40820
+    },
+    {
+      "epoch": 1.188003607062862,
+      "grad_norm": 13.1875,
+      "learning_rate": 6.039989733442826e-06,
+      "loss": 1.5481,
+      "step": 40840
+    },
+    {
+      "epoch": 1.1885853913954096,
+      "grad_norm": 13.5,
+      "learning_rate": 6.038050451770211e-06,
+      "loss": 1.4825,
+      "step": 40860
+    },
+    {
+      "epoch": 1.1891671757279576,
+      "grad_norm": 12.8125,
+      "learning_rate": 6.036111170097596e-06,
+      "loss": 1.4976,
+      "step": 40880
+    },
+    {
+      "epoch": 1.1897489600605056,
+      "grad_norm": 13.5,
+      "learning_rate": 6.034171888424981e-06,
+      "loss": 1.4153,
+      "step": 40900
+    },
+    {
+      "epoch": 1.1903307443930535,
+      "grad_norm": 12.875,
+      "learning_rate": 6.032232606752366e-06,
+      "loss": 1.5186,
+      "step": 40920
+    },
+    {
+      "epoch": 1.1909125287256015,
+      "grad_norm": 14.6875,
+      "learning_rate": 6.0302933250797515e-06,
+      "loss": 1.5677,
+      "step": 40940
+    },
+    {
+      "epoch": 1.1914943130581492,
+      "grad_norm": 13.5625,
+      "learning_rate": 6.0283540434071366e-06,
+      "loss": 1.5399,
+      "step": 40960
+    },
+    {
+      "epoch": 1.1920760973906972,
+      "grad_norm": 11.875,
+      "learning_rate": 6.026414761734522e-06,
+      "loss": 1.543,
+      "step": 40980
+    },
+    {
+      "epoch": 1.1926578817232452,
+      "grad_norm": 12.625,
+      "learning_rate": 6.024475480061907e-06,
+      "loss": 1.5075,
+      "step": 41000
+    },
+    {
+      "epoch": 1.1932396660557931,
+      "grad_norm": 15.3125,
+      "learning_rate": 6.022536198389292e-06,
+      "loss": 1.519,
+      "step": 41020
+    },
+    {
+      "epoch": 1.193821450388341,
+      "grad_norm": 12.6875,
+      "learning_rate": 6.020596916716677e-06,
+      "loss": 1.4778,
+      "step": 41040
+    },
+    {
+      "epoch": 1.194403234720889,
+      "grad_norm": 11.4375,
+      "learning_rate": 6.018657635044062e-06,
+      "loss": 1.4882,
+      "step": 41060
+    },
+    {
+      "epoch": 1.194985019053437,
+      "grad_norm": 15.0625,
+      "learning_rate": 6.016718353371447e-06,
+      "loss": 1.5328,
+      "step": 41080
+    },
+    {
+      "epoch": 1.1955668033859848,
+      "grad_norm": 13.375,
+      "learning_rate": 6.014779071698832e-06,
+      "loss": 1.4973,
+      "step": 41100
+    },
+    {
+      "epoch": 1.1961485877185327,
+      "grad_norm": 14.875,
+      "learning_rate": 6.012839790026217e-06,
+      "loss": 1.5371,
+      "step": 41120
+    },
+    {
+      "epoch": 1.1967303720510807,
+      "grad_norm": 13.1875,
+      "learning_rate": 6.0109005083536024e-06,
+      "loss": 1.5427,
+      "step": 41140
+    },
+    {
+      "epoch": 1.1973121563836286,
+      "grad_norm": 13.0625,
+      "learning_rate": 6.0089612266809875e-06,
+      "loss": 1.4804,
+      "step": 41160
+    },
+    {
+      "epoch": 1.1978939407161766,
+      "grad_norm": 15.5625,
+      "learning_rate": 6.007021945008373e-06,
+      "loss": 1.5384,
+      "step": 41180
+    },
+    {
+      "epoch": 1.1984757250487243,
+      "grad_norm": 12.625,
+      "learning_rate": 6.005082663335757e-06,
+      "loss": 1.5363,
+      "step": 41200
+    },
+    {
+      "epoch": 1.1990575093812723,
+      "grad_norm": 13.0,
+      "learning_rate": 6.003143381663142e-06,
+      "loss": 1.5551,
+      "step": 41220
+    },
+    {
+      "epoch": 1.1996392937138203,
+      "grad_norm": 12.5,
+      "learning_rate": 6.001204099990527e-06,
+      "loss": 1.5575,
+      "step": 41240
+    },
+    {
+      "epoch": 1.2002210780463682,
+      "grad_norm": 13.0,
+      "learning_rate": 5.999264818317912e-06,
+      "loss": 1.586,
+      "step": 41260
+    },
+    {
+      "epoch": 1.2008028623789162,
+      "grad_norm": 12.8125,
+      "learning_rate": 5.997325536645297e-06,
+      "loss": 1.569,
+      "step": 41280
+    },
+    {
+      "epoch": 1.201384646711464,
+      "grad_norm": 12.1875,
+      "learning_rate": 5.995386254972682e-06,
+      "loss": 1.5458,
+      "step": 41300
+    },
+    {
+      "epoch": 1.201966431044012,
+      "grad_norm": 13.4375,
+      "learning_rate": 5.9934469733000675e-06,
+      "loss": 1.523,
+      "step": 41320
+    },
+    {
+      "epoch": 1.2025482153765599,
+      "grad_norm": 13.9375,
+      "learning_rate": 5.9915076916274526e-06,
+      "loss": 1.5145,
+      "step": 41340
+    },
+    {
+      "epoch": 1.2031299997091078,
+      "grad_norm": 12.875,
+      "learning_rate": 5.989568409954838e-06,
+      "loss": 1.5263,
+      "step": 41360
+    },
+    {
+      "epoch": 1.2037117840416558,
+      "grad_norm": 13.0,
+      "learning_rate": 5.987629128282223e-06,
+      "loss": 1.606,
+      "step": 41380
+    },
+    {
+      "epoch": 1.2042935683742038,
+      "grad_norm": 12.5,
+      "learning_rate": 5.985689846609608e-06,
+      "loss": 1.5428,
+      "step": 41400
+    },
+    {
+      "epoch": 1.2048753527067517,
+      "grad_norm": 12.75,
+      "learning_rate": 5.983750564936993e-06,
+      "loss": 1.4788,
+      "step": 41420
+    },
+    {
+      "epoch": 1.2054571370392995,
+      "grad_norm": 13.875,
+      "learning_rate": 5.981811283264378e-06,
+      "loss": 1.5216,
+      "step": 41440
+    },
+    {
+      "epoch": 1.2060389213718474,
+      "grad_norm": 15.0625,
+      "learning_rate": 5.979872001591763e-06,
+      "loss": 1.606,
+      "step": 41460
+    },
+    {
+      "epoch": 1.2066207057043954,
+      "grad_norm": 12.875,
+      "learning_rate": 5.977932719919148e-06,
+      "loss": 1.4665,
+      "step": 41480
+    },
+    {
+      "epoch": 1.2072024900369434,
+      "grad_norm": 13.5,
+      "learning_rate": 5.975993438246533e-06,
+      "loss": 1.5042,
+      "step": 41500
+    },
+    {
+      "epoch": 1.2077842743694913,
+      "grad_norm": 14.4375,
+      "learning_rate": 5.9740541565739184e-06,
+      "loss": 1.5428,
+      "step": 41520
+    },
+    {
+      "epoch": 1.208366058702039,
+      "grad_norm": 14.625,
+      "learning_rate": 5.9721148749013035e-06,
+      "loss": 1.5344,
+      "step": 41540
+    },
+    {
+      "epoch": 1.208947843034587,
+      "grad_norm": 14.4375,
+      "learning_rate": 5.970175593228689e-06,
+      "loss": 1.5693,
+      "step": 41560
+    },
+    {
+      "epoch": 1.209529627367135,
+      "grad_norm": 14.875,
+      "learning_rate": 5.968236311556074e-06,
+      "loss": 1.4785,
+      "step": 41580
+    },
+    {
+      "epoch": 1.210111411699683,
+      "grad_norm": 14.375,
+      "learning_rate": 5.966297029883459e-06,
+      "loss": 1.6008,
+      "step": 41600
+    },
+    {
+      "epoch": 1.210693196032231,
+      "grad_norm": 14.3125,
+      "learning_rate": 5.964357748210844e-06,
+      "loss": 1.4983,
+      "step": 41620
+    },
+    {
+      "epoch": 1.2112749803647787,
+      "grad_norm": 14.8125,
+      "learning_rate": 5.962418466538229e-06,
+      "loss": 1.5593,
+      "step": 41640
+    },
+    {
+      "epoch": 1.2118567646973266,
+      "grad_norm": 12.875,
+      "learning_rate": 5.960479184865614e-06,
+      "loss": 1.4954,
+      "step": 41660
+    },
+    {
+      "epoch": 1.2124385490298746,
+      "grad_norm": 13.1875,
+      "learning_rate": 5.958539903192999e-06,
+      "loss": 1.4851,
+      "step": 41680
+    },
+    {
+      "epoch": 1.2130203333624225,
+      "grad_norm": 13.5,
+      "learning_rate": 5.956600621520384e-06,
+      "loss": 1.4913,
+      "step": 41700
+    },
+    {
+      "epoch": 1.2136021176949705,
+      "grad_norm": 12.625,
+      "learning_rate": 5.954661339847769e-06,
+      "loss": 1.5791,
+      "step": 41720
+    },
+    {
+      "epoch": 1.2141839020275185,
+      "grad_norm": 12.625,
+      "learning_rate": 5.9527220581751545e-06,
+      "loss": 1.5684,
+      "step": 41740
+    },
+    {
+      "epoch": 1.2147656863600664,
+      "grad_norm": 12.75,
+      "learning_rate": 5.95078277650254e-06,
+      "loss": 1.5621,
+      "step": 41760
+    },
+    {
+      "epoch": 1.2153474706926142,
+      "grad_norm": 13.5,
+      "learning_rate": 5.948843494829925e-06,
+      "loss": 1.5606,
+      "step": 41780
+    },
+    {
+      "epoch": 1.2159292550251621,
+      "grad_norm": 13.3125,
+      "learning_rate": 5.946904213157308e-06,
+      "loss": 1.5274,
+      "step": 41800
+    },
+    {
+      "epoch": 1.21651103935771,
+      "grad_norm": 14.375,
+      "learning_rate": 5.944964931484693e-06,
+      "loss": 1.4346,
+      "step": 41820
+    },
+    {
+      "epoch": 1.217092823690258,
+      "grad_norm": 10.0,
+      "learning_rate": 5.943025649812078e-06,
+      "loss": 1.4874,
+      "step": 41840
+    },
+    {
+      "epoch": 1.217674608022806,
+      "grad_norm": 13.3125,
+      "learning_rate": 5.941086368139463e-06,
+      "loss": 1.5017,
+      "step": 41860
+    },
+    {
+      "epoch": 1.2182563923553538,
+      "grad_norm": 16.125,
+      "learning_rate": 5.9391470864668485e-06,
+      "loss": 1.4754,
+      "step": 41880
+    },
+    {
+      "epoch": 1.2188381766879017,
+      "grad_norm": 14.0625,
+      "learning_rate": 5.937207804794234e-06,
+      "loss": 1.5927,
+      "step": 41900
+    },
+    {
+      "epoch": 1.2194199610204497,
+      "grad_norm": 16.125,
+      "learning_rate": 5.935268523121619e-06,
+      "loss": 1.4845,
+      "step": 41920
+    },
+    {
+      "epoch": 1.2200017453529977,
+      "grad_norm": 13.3125,
+      "learning_rate": 5.933329241449004e-06,
+      "loss": 1.53,
+      "step": 41940
+    },
+    {
+      "epoch": 1.2205835296855456,
+      "grad_norm": 14.0625,
+      "learning_rate": 5.931389959776389e-06,
+      "loss": 1.5277,
+      "step": 41960
+    },
+    {
+      "epoch": 1.2211653140180936,
+      "grad_norm": 12.75,
+      "learning_rate": 5.929450678103774e-06,
+      "loss": 1.5369,
+      "step": 41980
+    },
+    {
+      "epoch": 1.2217470983506413,
+      "grad_norm": 12.3125,
+      "learning_rate": 5.927511396431159e-06,
+      "loss": 1.5688,
+      "step": 42000
+    },
+    {
+      "epoch": 1.2223288826831893,
+      "grad_norm": 12.3125,
+      "learning_rate": 5.925572114758544e-06,
+      "loss": 1.5271,
+      "step": 42020
+    },
+    {
+      "epoch": 1.2229106670157373,
+      "grad_norm": 11.9375,
+      "learning_rate": 5.923632833085929e-06,
+      "loss": 1.5179,
+      "step": 42040
+    },
+    {
+      "epoch": 1.2234924513482852,
+      "grad_norm": 14.9375,
+      "learning_rate": 5.921693551413314e-06,
+      "loss": 1.5587,
+      "step": 42060
+    },
+    {
+      "epoch": 1.2240742356808332,
+      "grad_norm": 13.1875,
+      "learning_rate": 5.9197542697406995e-06,
+      "loss": 1.5901,
+      "step": 42080
+    },
+    {
+      "epoch": 1.2246560200133811,
+      "grad_norm": 11.625,
+      "learning_rate": 5.9178149880680846e-06,
+      "loss": 1.5209,
+      "step": 42100
+    },
+    {
+      "epoch": 1.2252378043459289,
+      "grad_norm": 12.9375,
+      "learning_rate": 5.91587570639547e-06,
+      "loss": 1.5068,
+      "step": 42120
+    },
+    {
+      "epoch": 1.2258195886784768,
+      "grad_norm": 15.375,
+      "learning_rate": 5.913936424722855e-06,
+      "loss": 1.4771,
+      "step": 42140
+    },
+    {
+      "epoch": 1.2264013730110248,
+      "grad_norm": 13.8125,
+      "learning_rate": 5.91199714305024e-06,
+      "loss": 1.532,
+      "step": 42160
+    },
+    {
+      "epoch": 1.2269831573435728,
+      "grad_norm": 13.0625,
+      "learning_rate": 5.910057861377625e-06,
+      "loss": 1.5224,
+      "step": 42180
+    },
+    {
+      "epoch": 1.2275649416761207,
+      "grad_norm": 14.375,
+      "learning_rate": 5.90811857970501e-06,
+      "loss": 1.5713,
+      "step": 42200
+    },
+    {
+      "epoch": 1.2281467260086685,
+      "grad_norm": 13.125,
+      "learning_rate": 5.906179298032395e-06,
+      "loss": 1.6152,
+      "step": 42220
+    },
+    {
+      "epoch": 1.2287285103412164,
+      "grad_norm": 10.125,
+      "learning_rate": 5.90424001635978e-06,
+      "loss": 1.5322,
+      "step": 42240
+    },
+    {
+      "epoch": 1.2293102946737644,
+      "grad_norm": 11.9375,
+      "learning_rate": 5.902300734687165e-06,
+      "loss": 1.4988,
+      "step": 42260
+    },
+    {
+      "epoch": 1.2298920790063124,
+      "grad_norm": 17.375,
+      "learning_rate": 5.9003614530145504e-06,
+      "loss": 1.5966,
+      "step": 42280
+    },
+    {
+      "epoch": 1.2304738633388603,
+      "grad_norm": 10.5,
+      "learning_rate": 5.8984221713419355e-06,
+      "loss": 1.5006,
+      "step": 42300
+    },
+    {
+      "epoch": 1.2310556476714083,
+      "grad_norm": 12.6875,
+      "learning_rate": 5.896482889669321e-06,
+      "loss": 1.5911,
+      "step": 42320
+    },
+    {
+      "epoch": 1.2316374320039563,
+      "grad_norm": 14.9375,
+      "learning_rate": 5.894543607996706e-06,
+      "loss": 1.5205,
+      "step": 42340
+    },
+    {
+      "epoch": 1.232219216336504,
+      "grad_norm": 12.0625,
+      "learning_rate": 5.892604326324091e-06,
+      "loss": 1.4876,
+      "step": 42360
+    },
+    {
+      "epoch": 1.232801000669052,
+      "grad_norm": 13.5,
+      "learning_rate": 5.890665044651476e-06,
+      "loss": 1.5186,
+      "step": 42380
+    },
+    {
+      "epoch": 1.2333827850016,
+      "grad_norm": 12.875,
+      "learning_rate": 5.888725762978861e-06,
+      "loss": 1.4763,
+      "step": 42400
+    },
+    {
+      "epoch": 1.2339645693341479,
+      "grad_norm": 14.125,
+      "learning_rate": 5.886786481306246e-06,
+      "loss": 1.5486,
+      "step": 42420
+    },
+    {
+      "epoch": 1.2345463536666959,
+      "grad_norm": 12.875,
+      "learning_rate": 5.884847199633631e-06,
+      "loss": 1.4882,
+      "step": 42440
+    },
+    {
+      "epoch": 1.2351281379992436,
+      "grad_norm": 12.6875,
+      "learning_rate": 5.882907917961016e-06,
+      "loss": 1.5763,
+      "step": 42460
+    },
+    {
+      "epoch": 1.2357099223317916,
+      "grad_norm": 13.4375,
+      "learning_rate": 5.8809686362884014e-06,
+      "loss": 1.452,
+      "step": 42480
+    },
+    {
+      "epoch": 1.2362917066643395,
+      "grad_norm": 14.5625,
+      "learning_rate": 5.8790293546157865e-06,
+      "loss": 1.5535,
+      "step": 42500
+    },
+    {
+      "epoch": 1.2368734909968875,
+      "grad_norm": 14.8125,
+      "learning_rate": 5.877090072943172e-06,
+      "loss": 1.5531,
+      "step": 42520
+    },
+    {
+      "epoch": 1.2374552753294354,
+      "grad_norm": 12.375,
+      "learning_rate": 5.875150791270557e-06,
+      "loss": 1.4908,
+      "step": 42540
+    },
+    {
+      "epoch": 1.2380370596619832,
+      "grad_norm": 14.0,
+      "learning_rate": 5.873211509597942e-06,
+      "loss": 1.4919,
+      "step": 42560
+    },
+    {
+      "epoch": 1.2386188439945311,
+      "grad_norm": 12.9375,
+      "learning_rate": 5.871272227925326e-06,
+      "loss": 1.4404,
+      "step": 42580
+    },
+    {
+      "epoch": 1.2392006283270791,
+      "grad_norm": 13.9375,
+      "learning_rate": 5.869332946252711e-06,
+      "loss": 1.4845,
+      "step": 42600
+    },
+    {
+      "epoch": 1.239782412659627,
+      "grad_norm": 12.1875,
+      "learning_rate": 5.867393664580096e-06,
+      "loss": 1.5165,
+      "step": 42620
+    },
+    {
+      "epoch": 1.240364196992175,
+      "grad_norm": 13.0,
+      "learning_rate": 5.865454382907481e-06,
+      "loss": 1.5542,
+      "step": 42640
+    },
+    {
+      "epoch": 1.240945981324723,
+      "grad_norm": 12.375,
+      "learning_rate": 5.8635151012348665e-06,
+      "loss": 1.4909,
+      "step": 42660
+    },
+    {
+      "epoch": 1.241527765657271,
+      "grad_norm": 12.375,
+      "learning_rate": 5.8615758195622516e-06,
+      "loss": 1.5737,
+      "step": 42680
+    },
+    {
+      "epoch": 1.2421095499898187,
+      "grad_norm": 12.1875,
+      "learning_rate": 5.859636537889637e-06,
+      "loss": 1.5644,
+      "step": 42700
+    },
+    {
+      "epoch": 1.2426913343223667,
+      "grad_norm": 13.625,
+      "learning_rate": 5.857697256217022e-06,
+      "loss": 1.596,
+      "step": 42720
+    },
+    {
+      "epoch": 1.2432731186549146,
+      "grad_norm": 11.5625,
+      "learning_rate": 5.855757974544407e-06,
+      "loss": 1.4895,
+      "step": 42740
+    },
+    {
+      "epoch": 1.2438549029874626,
+      "grad_norm": 14.5,
+      "learning_rate": 5.853818692871792e-06,
+      "loss": 1.5701,
+      "step": 42760
+    },
+    {
+      "epoch": 1.2444366873200106,
+      "grad_norm": 14.9375,
+      "learning_rate": 5.851879411199177e-06,
+      "loss": 1.5158,
+      "step": 42780
+    },
+    {
+      "epoch": 1.2450184716525583,
+      "grad_norm": 14.8125,
+      "learning_rate": 5.849940129526562e-06,
+      "loss": 1.5914,
+      "step": 42800
+    },
+    {
+      "epoch": 1.2456002559851063,
+      "grad_norm": 11.5625,
+      "learning_rate": 5.848000847853947e-06,
+      "loss": 1.5527,
+      "step": 42820
+    },
+    {
+      "epoch": 1.2461820403176542,
+      "grad_norm": 13.625,
+      "learning_rate": 5.846061566181332e-06,
+      "loss": 1.5446,
+      "step": 42840
+    },
+    {
+      "epoch": 1.2467638246502022,
+      "grad_norm": 12.625,
+      "learning_rate": 5.8441222845087174e-06,
+      "loss": 1.4778,
+      "step": 42860
+    },
+    {
+      "epoch": 1.2473456089827502,
+      "grad_norm": 15.875,
+      "learning_rate": 5.8421830028361025e-06,
+      "loss": 1.6255,
+      "step": 42880
+    },
+    {
+      "epoch": 1.247927393315298,
+      "grad_norm": 14.125,
+      "learning_rate": 5.840243721163488e-06,
+      "loss": 1.5588,
+      "step": 42900
+    },
+    {
+      "epoch": 1.2485091776478459,
+      "grad_norm": 13.125,
+      "learning_rate": 5.838304439490873e-06,
+      "loss": 1.5103,
+      "step": 42920
+    },
+    {
+      "epoch": 1.2490909619803938,
+      "grad_norm": 14.9375,
+      "learning_rate": 5.836365157818258e-06,
+      "loss": 1.5243,
+      "step": 42940
+    },
+    {
+      "epoch": 1.2496727463129418,
+      "grad_norm": 13.125,
+      "learning_rate": 5.834425876145643e-06,
+      "loss": 1.5647,
+      "step": 42960
+    },
+    {
+      "epoch": 1.2502545306454897,
+      "grad_norm": 12.375,
+      "learning_rate": 5.832486594473028e-06,
+      "loss": 1.5204,
+      "step": 42980
+    },
+    {
+      "epoch": 1.2508363149780377,
+      "grad_norm": 11.1875,
+      "learning_rate": 5.830547312800413e-06,
+      "loss": 1.5199,
+      "step": 43000
+    },
+    {
+      "epoch": 1.2514180993105857,
+      "grad_norm": 13.1875,
+      "learning_rate": 5.828608031127798e-06,
+      "loss": 1.5245,
+      "step": 43020
+    },
+    {
+      "epoch": 1.2519998836431334,
+      "grad_norm": 13.3125,
+      "learning_rate": 5.826668749455183e-06,
+      "loss": 1.5144,
+      "step": 43040
+    },
+    {
+      "epoch": 1.2525816679756814,
+      "grad_norm": 13.4375,
+      "learning_rate": 5.824729467782568e-06,
+      "loss": 1.5476,
+      "step": 43060
+    },
+    {
+      "epoch": 1.2531634523082293,
+      "grad_norm": 13.625,
+      "learning_rate": 5.8227901861099535e-06,
+      "loss": 1.4843,
+      "step": 43080
+    },
+    {
+      "epoch": 1.2537452366407773,
+      "grad_norm": 14.125,
+      "learning_rate": 5.820850904437339e-06,
+      "loss": 1.4631,
+      "step": 43100
+    },
+    {
+      "epoch": 1.2543270209733253,
+      "grad_norm": 13.25,
+      "learning_rate": 5.818911622764724e-06,
+      "loss": 1.5837,
+      "step": 43120
+    },
+    {
+      "epoch": 1.254908805305873,
+      "grad_norm": 15.0625,
+      "learning_rate": 5.816972341092109e-06,
+      "loss": 1.4783,
+      "step": 43140
+    },
+    {
+      "epoch": 1.255490589638421,
+      "grad_norm": 11.625,
+      "learning_rate": 5.815033059419494e-06,
+      "loss": 1.5691,
+      "step": 43160
+    },
+    {
+      "epoch": 1.256072373970969,
+      "grad_norm": 12.8125,
+      "learning_rate": 5.813093777746879e-06,
+      "loss": 1.5289,
+      "step": 43180
+    },
+    {
+      "epoch": 1.256654158303517,
+      "grad_norm": 15.625,
+      "learning_rate": 5.811154496074264e-06,
+      "loss": 1.5228,
+      "step": 43200
+    },
+    {
+      "epoch": 1.2572359426360649,
+      "grad_norm": 15.9375,
+      "learning_rate": 5.809215214401649e-06,
+      "loss": 1.5698,
+      "step": 43220
+    },
+    {
+      "epoch": 1.2578177269686126,
+      "grad_norm": 13.6875,
+      "learning_rate": 5.807275932729034e-06,
+      "loss": 1.511,
+      "step": 43240
+    },
+    {
+      "epoch": 1.2583995113011608,
+      "grad_norm": 13.375,
+      "learning_rate": 5.805336651056419e-06,
+      "loss": 1.599,
+      "step": 43260
+    },
+    {
+      "epoch": 1.2589812956337085,
+      "grad_norm": 11.875,
+      "learning_rate": 5.8033973693838045e-06,
+      "loss": 1.4926,
+      "step": 43280
+    },
+    {
+      "epoch": 1.2595630799662565,
+      "grad_norm": 13.9375,
+      "learning_rate": 5.8014580877111896e-06,
+      "loss": 1.537,
+      "step": 43300
+    },
+    {
+      "epoch": 1.2601448642988045,
+      "grad_norm": 16.125,
+      "learning_rate": 5.799518806038575e-06,
+      "loss": 1.4502,
+      "step": 43320
+    },
+    {
+      "epoch": 1.2607266486313524,
+      "grad_norm": 14.25,
+      "learning_rate": 5.79757952436596e-06,
+      "loss": 1.4721,
+      "step": 43340
+    },
+    {
+      "epoch": 1.2613084329639004,
+      "grad_norm": 12.5,
+      "learning_rate": 5.795640242693345e-06,
+      "loss": 1.537,
+      "step": 43360
+    },
+    {
+      "epoch": 1.2618902172964481,
+      "grad_norm": 15.125,
+      "learning_rate": 5.79370096102073e-06,
+      "loss": 1.538,
+      "step": 43380
+    },
+    {
+      "epoch": 1.262472001628996,
+      "grad_norm": 14.0625,
+      "learning_rate": 5.791761679348114e-06,
+      "loss": 1.5307,
+      "step": 43400
+    },
+    {
+      "epoch": 1.263053785961544,
+      "grad_norm": 14.3125,
+      "learning_rate": 5.789822397675499e-06,
+      "loss": 1.5407,
+      "step": 43420
+    },
+    {
+      "epoch": 1.263635570294092,
+      "grad_norm": 11.5,
+      "learning_rate": 5.787883116002884e-06,
+      "loss": 1.5532,
+      "step": 43440
+    },
+    {
+      "epoch": 1.26421735462664,
+      "grad_norm": 11.5,
+      "learning_rate": 5.7859438343302695e-06,
+      "loss": 1.6063,
+      "step": 43460
+    },
+    {
+      "epoch": 1.2647991389591877,
+      "grad_norm": 13.5,
+      "learning_rate": 5.784004552657655e-06,
+      "loss": 1.5031,
+      "step": 43480
+    },
+    {
+      "epoch": 1.2653809232917357,
+      "grad_norm": 13.25,
+      "learning_rate": 5.78206527098504e-06,
+      "loss": 1.5268,
+      "step": 43500
+    },
+    {
+      "epoch": 1.2659627076242836,
+      "grad_norm": 12.0625,
+      "learning_rate": 5.780125989312425e-06,
+      "loss": 1.4692,
+      "step": 43520
+    },
+    {
+      "epoch": 1.2665444919568316,
+      "grad_norm": 14.25,
+      "learning_rate": 5.77818670763981e-06,
+      "loss": 1.4345,
+      "step": 43540
+    },
+    {
+      "epoch": 1.2671262762893796,
+      "grad_norm": 13.375,
+      "learning_rate": 5.776247425967195e-06,
+      "loss": 1.4452,
+      "step": 43560
+    },
+    {
+      "epoch": 1.2677080606219273,
+      "grad_norm": 9.8125,
+      "learning_rate": 5.77430814429458e-06,
+      "loss": 1.5405,
+      "step": 43580
+    },
+    {
+      "epoch": 1.2682898449544755,
+      "grad_norm": 13.5625,
+      "learning_rate": 5.772368862621965e-06,
+      "loss": 1.5249,
+      "step": 43600
+    },
+    {
+      "epoch": 1.2688716292870232,
+      "grad_norm": 15.5,
+      "learning_rate": 5.77042958094935e-06,
+      "loss": 1.499,
+      "step": 43620
+    },
+    {
+      "epoch": 1.2694534136195712,
+      "grad_norm": 13.0625,
+      "learning_rate": 5.768490299276735e-06,
+      "loss": 1.4628,
+      "step": 43640
+    },
+    {
+      "epoch": 1.2700351979521192,
+      "grad_norm": 13.125,
+      "learning_rate": 5.7665510176041205e-06,
+      "loss": 1.464,
+      "step": 43660
+    },
+    {
+      "epoch": 1.2706169822846671,
+      "grad_norm": 11.375,
+      "learning_rate": 5.764611735931506e-06,
+      "loss": 1.5615,
+      "step": 43680
+    },
+    {
+      "epoch": 1.271198766617215,
+      "grad_norm": 13.625,
+      "learning_rate": 5.762672454258891e-06,
+      "loss": 1.5503,
+      "step": 43700
+    },
+    {
+      "epoch": 1.2717805509497628,
+      "grad_norm": 11.25,
+      "learning_rate": 5.760733172586276e-06,
+      "loss": 1.3915,
+      "step": 43720
+    },
+    {
+      "epoch": 1.2723623352823108,
+      "grad_norm": 13.625,
+      "learning_rate": 5.758793890913661e-06,
+      "loss": 1.6215,
+      "step": 43740
+    },
+    {
+      "epoch": 1.2729441196148588,
+      "grad_norm": 16.5,
+      "learning_rate": 5.756854609241046e-06,
+      "loss": 1.5395,
+      "step": 43760
+    },
+    {
+      "epoch": 1.2735259039474067,
+      "grad_norm": 17.125,
+      "learning_rate": 5.754915327568431e-06,
+      "loss": 1.5404,
+      "step": 43780
+    },
+    {
+      "epoch": 1.2741076882799547,
+      "grad_norm": 13.5,
+      "learning_rate": 5.752976045895816e-06,
+      "loss": 1.4823,
+      "step": 43800
+    },
+    {
+      "epoch": 1.2746894726125024,
+      "grad_norm": 12.0625,
+      "learning_rate": 5.751036764223201e-06,
+      "loss": 1.5202,
+      "step": 43820
+    },
+    {
+      "epoch": 1.2752712569450504,
+      "grad_norm": 14.875,
+      "learning_rate": 5.749097482550586e-06,
+      "loss": 1.5736,
+      "step": 43840
+    },
+    {
+      "epoch": 1.2758530412775984,
+      "grad_norm": 12.5625,
+      "learning_rate": 5.7471582008779715e-06,
+      "loss": 1.4768,
+      "step": 43860
+    },
+    {
+      "epoch": 1.2764348256101463,
+      "grad_norm": 16.875,
+      "learning_rate": 5.7452189192053566e-06,
+      "loss": 1.5277,
+      "step": 43880
+    },
+    {
+      "epoch": 1.2770166099426943,
+      "grad_norm": 11.0625,
+      "learning_rate": 5.743279637532742e-06,
+      "loss": 1.4631,
+      "step": 43900
+    },
+    {
+      "epoch": 1.2775983942752422,
+      "grad_norm": 14.875,
+      "learning_rate": 5.741340355860127e-06,
+      "loss": 1.5205,
+      "step": 43920
+    },
+    {
+      "epoch": 1.2781801786077902,
+      "grad_norm": 13.0,
+      "learning_rate": 5.739401074187512e-06,
+      "loss": 1.5229,
+      "step": 43940
+    },
+    {
+      "epoch": 1.278761962940338,
+      "grad_norm": 12.5,
+      "learning_rate": 5.737461792514897e-06,
+      "loss": 1.5677,
+      "step": 43960
+    },
+    {
+      "epoch": 1.279343747272886,
+      "grad_norm": 13.5,
+      "learning_rate": 5.735522510842282e-06,
+      "loss": 1.4604,
+      "step": 43980
+    },
+    {
+      "epoch": 1.2799255316054339,
+      "grad_norm": 13.125,
+      "learning_rate": 5.733583229169667e-06,
+      "loss": 1.5695,
+      "step": 44000
+    },
+    {
+      "epoch": 1.2805073159379818,
+      "grad_norm": 14.375,
+      "learning_rate": 5.731643947497052e-06,
+      "loss": 1.5317,
+      "step": 44020
+    },
+    {
+      "epoch": 1.2810891002705298,
+      "grad_norm": 14.5,
+      "learning_rate": 5.729704665824437e-06,
+      "loss": 1.5426,
+      "step": 44040
+    },
+    {
+      "epoch": 1.2816708846030775,
+      "grad_norm": 15.5625,
+      "learning_rate": 5.7277653841518224e-06,
+      "loss": 1.6236,
+      "step": 44060
+    },
+    {
+      "epoch": 1.2822526689356255,
+      "grad_norm": 12.4375,
+      "learning_rate": 5.7258261024792075e-06,
+      "loss": 1.5514,
+      "step": 44080
+    },
+    {
+      "epoch": 1.2828344532681735,
+      "grad_norm": 13.4375,
+      "learning_rate": 5.723886820806593e-06,
+      "loss": 1.5181,
+      "step": 44100
+    },
+    {
+      "epoch": 1.2834162376007214,
+      "grad_norm": 13.3125,
+      "learning_rate": 5.721947539133978e-06,
+      "loss": 1.5044,
+      "step": 44120
+    },
+    {
+      "epoch": 1.2839980219332694,
+      "grad_norm": 11.75,
+      "learning_rate": 5.720008257461363e-06,
+      "loss": 1.5696,
+      "step": 44140
+    },
+    {
+      "epoch": 1.2845798062658171,
+      "grad_norm": 14.1875,
+      "learning_rate": 5.718068975788748e-06,
+      "loss": 1.4719,
+      "step": 44160
+    },
+    {
+      "epoch": 1.2851615905983653,
+      "grad_norm": 13.375,
+      "learning_rate": 5.716129694116133e-06,
+      "loss": 1.4986,
+      "step": 44180
+    },
+    {
+      "epoch": 1.285743374930913,
+      "grad_norm": 12.8125,
+      "learning_rate": 5.714190412443517e-06,
+      "loss": 1.5602,
+      "step": 44200
+    },
+    {
+      "epoch": 1.286325159263461,
+      "grad_norm": 10.875,
+      "learning_rate": 5.712251130770902e-06,
+      "loss": 1.5185,
+      "step": 44220
+    },
+    {
+      "epoch": 1.286906943596009,
+      "grad_norm": 12.0625,
+      "learning_rate": 5.7103118490982875e-06,
+      "loss": 1.4907,
+      "step": 44240
+    },
+    {
+      "epoch": 1.287488727928557,
+      "grad_norm": 12.0625,
+      "learning_rate": 5.7083725674256726e-06,
+      "loss": 1.4145,
+      "step": 44260
+    },
+    {
+      "epoch": 1.288070512261105,
+      "grad_norm": 14.375,
+      "learning_rate": 5.706433285753058e-06,
+      "loss": 1.5275,
+      "step": 44280
+    },
+    {
+      "epoch": 1.2886522965936527,
+      "grad_norm": 15.3125,
+      "learning_rate": 5.704494004080443e-06,
+      "loss": 1.5402,
+      "step": 44300
+    },
+    {
+      "epoch": 1.2892340809262006,
+      "grad_norm": 12.375,
+      "learning_rate": 5.702554722407828e-06,
+      "loss": 1.6014,
+      "step": 44320
+    },
+    {
+      "epoch": 1.2898158652587486,
+      "grad_norm": 13.75,
+      "learning_rate": 5.700615440735213e-06,
+      "loss": 1.6052,
+      "step": 44340
+    },
+    {
+      "epoch": 1.2903976495912965,
+      "grad_norm": 12.5625,
+      "learning_rate": 5.698676159062598e-06,
+      "loss": 1.4923,
+      "step": 44360
+    },
+    {
+      "epoch": 1.2909794339238445,
+      "grad_norm": 11.5625,
+      "learning_rate": 5.696736877389983e-06,
+      "loss": 1.4916,
+      "step": 44380
+    },
+    {
+      "epoch": 1.2915612182563923,
+      "grad_norm": 13.4375,
+      "learning_rate": 5.694797595717368e-06,
+      "loss": 1.5162,
+      "step": 44400
+    },
+    {
+      "epoch": 1.2921430025889402,
+      "grad_norm": 11.875,
+      "learning_rate": 5.692858314044753e-06,
+      "loss": 1.6063,
+      "step": 44420
+    },
+    {
+      "epoch": 1.2927247869214882,
+      "grad_norm": 11.6875,
+      "learning_rate": 5.6909190323721384e-06,
+      "loss": 1.6026,
+      "step": 44440
+    },
+    {
+      "epoch": 1.2933065712540361,
+      "grad_norm": 12.3125,
+      "learning_rate": 5.6889797506995235e-06,
+      "loss": 1.3727,
+      "step": 44460
+    },
+    {
+      "epoch": 1.293888355586584,
+      "grad_norm": 12.4375,
+      "learning_rate": 5.687040469026909e-06,
+      "loss": 1.517,
+      "step": 44480
+    },
+    {
+      "epoch": 1.2944701399191318,
+      "grad_norm": 11.3125,
+      "learning_rate": 5.685101187354294e-06,
+      "loss": 1.4526,
+      "step": 44500
+    },
+    {
+      "epoch": 1.29505192425168,
+      "grad_norm": 17.375,
+      "learning_rate": 5.683161905681679e-06,
+      "loss": 1.6621,
+      "step": 44520
+    },
+    {
+      "epoch": 1.2956337085842278,
+      "grad_norm": 14.9375,
+      "learning_rate": 5.681222624009064e-06,
+      "loss": 1.5943,
+      "step": 44540
+    },
+    {
+      "epoch": 1.2962154929167757,
+      "grad_norm": 12.4375,
+      "learning_rate": 5.679283342336449e-06,
+      "loss": 1.5046,
+      "step": 44560
+    },
+    {
+      "epoch": 1.2967972772493237,
+      "grad_norm": 14.75,
+      "learning_rate": 5.677344060663834e-06,
+      "loss": 1.5444,
+      "step": 44580
+    },
+    {
+      "epoch": 1.2973790615818717,
+      "grad_norm": 13.8125,
+      "learning_rate": 5.675404778991219e-06,
+      "loss": 1.5583,
+      "step": 44600
+    },
+    {
+      "epoch": 1.2979608459144196,
+      "grad_norm": 14.625,
+      "learning_rate": 5.673465497318604e-06,
+      "loss": 1.4507,
+      "step": 44620
+    },
+    {
+      "epoch": 1.2985426302469674,
+      "grad_norm": 11.0625,
+      "learning_rate": 5.671526215645989e-06,
+      "loss": 1.5596,
+      "step": 44640
+    },
+    {
+      "epoch": 1.2991244145795153,
+      "grad_norm": 13.25,
+      "learning_rate": 5.6695869339733745e-06,
+      "loss": 1.5583,
+      "step": 44660
+    },
+    {
+      "epoch": 1.2997061989120633,
+      "grad_norm": 38.75,
+      "learning_rate": 5.66764765230076e-06,
+      "loss": 1.5247,
+      "step": 44680
+    },
+    {
+      "epoch": 1.3002879832446113,
+      "grad_norm": 7.75,
+      "learning_rate": 5.665708370628145e-06,
+      "loss": 1.3917,
+      "step": 44700
+    },
+    {
+      "epoch": 1.3008697675771592,
+      "grad_norm": 15.0625,
+      "learning_rate": 5.66376908895553e-06,
+      "loss": 1.5193,
+      "step": 44720
+    },
+    {
+      "epoch": 1.301451551909707,
+      "grad_norm": 13.3125,
+      "learning_rate": 5.661829807282915e-06,
+      "loss": 1.5081,
+      "step": 44740
+    },
+    {
+      "epoch": 1.302033336242255,
+      "grad_norm": 14.75,
+      "learning_rate": 5.6598905256103e-06,
+      "loss": 1.4606,
+      "step": 44760
+    },
+    {
+      "epoch": 1.3026151205748029,
+      "grad_norm": 11.4375,
+      "learning_rate": 5.657951243937685e-06,
+      "loss": 1.5302,
+      "step": 44780
+    },
+    {
+      "epoch": 1.3031969049073509,
+      "grad_norm": 10.875,
+      "learning_rate": 5.65601196226507e-06,
+      "loss": 1.53,
+      "step": 44800
+    },
+    {
+      "epoch": 1.3037786892398988,
+      "grad_norm": 12.6875,
+      "learning_rate": 5.654072680592455e-06,
+      "loss": 1.5268,
+      "step": 44820
+    },
+    {
+      "epoch": 1.3043604735724468,
+      "grad_norm": 12.625,
+      "learning_rate": 5.65213339891984e-06,
+      "loss": 1.5246,
+      "step": 44840
+    },
+    {
+      "epoch": 1.3049422579049947,
+      "grad_norm": 13.9375,
+      "learning_rate": 5.6501941172472255e-06,
+      "loss": 1.5355,
+      "step": 44860
+    },
+    {
+      "epoch": 1.3055240422375425,
+      "grad_norm": 12.4375,
+      "learning_rate": 5.648254835574611e-06,
+      "loss": 1.5432,
+      "step": 44880
+    },
+    {
+      "epoch": 1.3061058265700904,
+      "grad_norm": 13.3125,
+      "learning_rate": 5.646315553901996e-06,
+      "loss": 1.5031,
+      "step": 44900
+    },
+    {
+      "epoch": 1.3066876109026384,
+      "grad_norm": 12.75,
+      "learning_rate": 5.644376272229381e-06,
+      "loss": 1.4694,
+      "step": 44920
+    },
+    {
+      "epoch": 1.3072693952351864,
+      "grad_norm": 13.5,
+      "learning_rate": 5.642436990556766e-06,
+      "loss": 1.5157,
+      "step": 44940
+    },
+    {
+      "epoch": 1.3078511795677343,
+      "grad_norm": 12.0625,
+      "learning_rate": 5.640497708884151e-06,
+      "loss": 1.461,
+      "step": 44960
+    },
+    {
+      "epoch": 1.308432963900282,
+      "grad_norm": 12.6875,
+      "learning_rate": 5.638558427211536e-06,
+      "loss": 1.5395,
+      "step": 44980
+    },
+    {
+      "epoch": 1.30901474823283,
+      "grad_norm": 12.0,
+      "learning_rate": 5.636619145538921e-06,
+      "loss": 1.5547,
+      "step": 45000
+    },
+    {
+      "epoch": 1.309596532565378,
+      "grad_norm": 10.1875,
+      "learning_rate": 5.634679863866305e-06,
+      "loss": 1.5409,
+      "step": 45020
+    },
+    {
+      "epoch": 1.310178316897926,
+      "grad_norm": 13.4375,
+      "learning_rate": 5.6327405821936905e-06,
+      "loss": 1.4595,
+      "step": 45040
+    },
+    {
+      "epoch": 1.310760101230474,
+      "grad_norm": 11.3125,
+      "learning_rate": 5.630801300521076e-06,
+      "loss": 1.4294,
+      "step": 45060
+    },
+    {
+      "epoch": 1.3113418855630217,
+      "grad_norm": 15.625,
+      "learning_rate": 5.628862018848461e-06,
+      "loss": 1.5729,
+      "step": 45080
+    },
+    {
+      "epoch": 1.3119236698955696,
+      "grad_norm": 12.875,
+      "learning_rate": 5.626922737175846e-06,
+      "loss": 1.4949,
+      "step": 45100
+    },
+    {
+      "epoch": 1.3125054542281176,
+      "grad_norm": 15.4375,
+      "learning_rate": 5.624983455503231e-06,
+      "loss": 1.5148,
+      "step": 45120
+    },
+    {
+      "epoch": 1.3130872385606656,
+      "grad_norm": 12.5,
+      "learning_rate": 5.623044173830616e-06,
+      "loss": 1.5115,
+      "step": 45140
+    },
+    {
+      "epoch": 1.3136690228932135,
+      "grad_norm": 14.25,
+      "learning_rate": 5.621104892158001e-06,
+      "loss": 1.4664,
+      "step": 45160
+    },
+    {
+      "epoch": 1.3142508072257615,
+      "grad_norm": 10.4375,
+      "learning_rate": 5.619165610485386e-06,
+      "loss": 1.5492,
+      "step": 45180
+    },
+    {
+      "epoch": 1.3148325915583094,
+      "grad_norm": 11.625,
+      "learning_rate": 5.617226328812771e-06,
+      "loss": 1.4522,
+      "step": 45200
+    },
+    {
+      "epoch": 1.3154143758908572,
+      "grad_norm": 12.1875,
+      "learning_rate": 5.615287047140156e-06,
+      "loss": 1.582,
+      "step": 45220
+    },
+    {
+      "epoch": 1.3159961602234052,
+      "grad_norm": 14.0625,
+      "learning_rate": 5.6133477654675415e-06,
+      "loss": 1.4966,
+      "step": 45240
+    },
+    {
+      "epoch": 1.3165779445559531,
+      "grad_norm": 11.9375,
+      "learning_rate": 5.611408483794927e-06,
+      "loss": 1.5052,
+      "step": 45260
+    },
+    {
+      "epoch": 1.317159728888501,
+      "grad_norm": 11.9375,
+      "learning_rate": 5.609469202122312e-06,
+      "loss": 1.558,
+      "step": 45280
+    },
+    {
+      "epoch": 1.317741513221049,
+      "grad_norm": 12.9375,
+      "learning_rate": 5.607529920449697e-06,
+      "loss": 1.4795,
+      "step": 45300
+    },
+    {
+      "epoch": 1.3183232975535968,
+      "grad_norm": 11.8125,
+      "learning_rate": 5.605590638777082e-06,
+      "loss": 1.4875,
+      "step": 45320
+    },
+    {
+      "epoch": 1.3189050818861447,
+      "grad_norm": 14.0625,
+      "learning_rate": 5.603651357104467e-06,
+      "loss": 1.4081,
+      "step": 45340
+    },
+    {
+      "epoch": 1.3194868662186927,
+      "grad_norm": 13.0625,
+      "learning_rate": 5.601712075431852e-06,
+      "loss": 1.6053,
+      "step": 45360
+    },
+    {
+      "epoch": 1.3200686505512407,
+      "grad_norm": 17.0,
+      "learning_rate": 5.599772793759237e-06,
+      "loss": 1.5608,
+      "step": 45380
+    },
+    {
+      "epoch": 1.3206504348837886,
+      "grad_norm": 12.375,
+      "learning_rate": 5.597833512086622e-06,
+      "loss": 1.5341,
+      "step": 45400
+    },
+    {
+      "epoch": 1.3212322192163364,
+      "grad_norm": 15.6875,
+      "learning_rate": 5.595894230414007e-06,
+      "loss": 1.5038,
+      "step": 45420
+    },
+    {
+      "epoch": 1.3218140035488846,
+      "grad_norm": 13.6875,
+      "learning_rate": 5.5939549487413925e-06,
+      "loss": 1.4798,
+      "step": 45440
+    },
+    {
+      "epoch": 1.3223957878814323,
+      "grad_norm": 14.375,
+      "learning_rate": 5.5920156670687776e-06,
+      "loss": 1.5253,
+      "step": 45460
+    },
+    {
+      "epoch": 1.3229775722139803,
+      "grad_norm": 16.125,
+      "learning_rate": 5.590076385396163e-06,
+      "loss": 1.4623,
+      "step": 45480
+    },
+    {
+      "epoch": 1.3235593565465282,
+      "grad_norm": 12.0,
+      "learning_rate": 5.588137103723548e-06,
+      "loss": 1.5533,
+      "step": 45500
+    },
+    {
+      "epoch": 1.3241411408790762,
+      "grad_norm": 11.375,
+      "learning_rate": 5.586197822050933e-06,
+      "loss": 1.5577,
+      "step": 45520
+    },
+    {
+      "epoch": 1.3247229252116242,
+      "grad_norm": 11.125,
+      "learning_rate": 5.584258540378318e-06,
+      "loss": 1.5427,
+      "step": 45540
+    },
+    {
+      "epoch": 1.325304709544172,
+      "grad_norm": 13.9375,
+      "learning_rate": 5.582319258705703e-06,
+      "loss": 1.5284,
+      "step": 45560
+    },
+    {
+      "epoch": 1.3258864938767199,
+      "grad_norm": 12.375,
+      "learning_rate": 5.580379977033088e-06,
+      "loss": 1.5029,
+      "step": 45580
+    },
+    {
+      "epoch": 1.3264682782092678,
+      "grad_norm": 12.1875,
+      "learning_rate": 5.578440695360473e-06,
+      "loss": 1.5527,
+      "step": 45600
+    },
+    {
+      "epoch": 1.3270500625418158,
+      "grad_norm": 12.6875,
+      "learning_rate": 5.576501413687858e-06,
+      "loss": 1.5544,
+      "step": 45620
+    },
+    {
+      "epoch": 1.3276318468743638,
+      "grad_norm": 14.1875,
+      "learning_rate": 5.5745621320152434e-06,
+      "loss": 1.5591,
+      "step": 45640
+    },
+    {
+      "epoch": 1.3282136312069115,
+      "grad_norm": 12.8125,
+      "learning_rate": 5.5726228503426285e-06,
+      "loss": 1.4167,
+      "step": 45660
+    },
+    {
+      "epoch": 1.3287954155394595,
+      "grad_norm": 13.375,
+      "learning_rate": 5.570683568670014e-06,
+      "loss": 1.4603,
+      "step": 45680
+    },
+    {
+      "epoch": 1.3293771998720074,
+      "grad_norm": 15.0625,
+      "learning_rate": 5.568744286997399e-06,
+      "loss": 1.5703,
+      "step": 45700
+    },
+    {
+      "epoch": 1.3299589842045554,
+      "grad_norm": 14.6875,
+      "learning_rate": 5.566805005324784e-06,
+      "loss": 1.5153,
+      "step": 45720
+    },
+    {
+      "epoch": 1.3305407685371033,
+      "grad_norm": 14.375,
+      "learning_rate": 5.564865723652169e-06,
+      "loss": 1.4395,
+      "step": 45740
+    },
+    {
+      "epoch": 1.331122552869651,
+      "grad_norm": 14.875,
+      "learning_rate": 5.562926441979554e-06,
+      "loss": 1.394,
+      "step": 45760
+    },
+    {
+      "epoch": 1.3317043372021993,
+      "grad_norm": 13.5625,
+      "learning_rate": 5.560987160306939e-06,
+      "loss": 1.501,
+      "step": 45780
+    },
+    {
+      "epoch": 1.332286121534747,
+      "grad_norm": 12.8125,
+      "learning_rate": 5.559047878634324e-06,
+      "loss": 1.4852,
+      "step": 45800
+    },
+    {
+      "epoch": 1.332867905867295,
+      "grad_norm": 12.25,
+      "learning_rate": 5.557108596961709e-06,
+      "loss": 1.5356,
+      "step": 45820
+    },
+    {
+      "epoch": 1.333449690199843,
+      "grad_norm": 13.25,
+      "learning_rate": 5.5551693152890936e-06,
+      "loss": 1.4918,
+      "step": 45840
+    },
+    {
+      "epoch": 1.334031474532391,
+      "grad_norm": 15.5,
+      "learning_rate": 5.553230033616479e-06,
+      "loss": 1.4625,
+      "step": 45860
+    },
+    {
+      "epoch": 1.3346132588649389,
+      "grad_norm": 13.0,
+      "learning_rate": 5.551290751943864e-06,
+      "loss": 1.4486,
+      "step": 45880
+    },
+    {
+      "epoch": 1.3351950431974866,
+      "grad_norm": 12.1875,
+      "learning_rate": 5.549351470271249e-06,
+      "loss": 1.4094,
+      "step": 45900
+    },
+    {
+      "epoch": 1.3357768275300346,
+      "grad_norm": 12.875,
+      "learning_rate": 5.547412188598634e-06,
+      "loss": 1.5693,
+      "step": 45920
+    },
+    {
+      "epoch": 1.3363586118625825,
+      "grad_norm": 11.0625,
+      "learning_rate": 5.545472906926019e-06,
+      "loss": 1.5614,
+      "step": 45940
+    },
+    {
+      "epoch": 1.3369403961951305,
+      "grad_norm": 14.25,
+      "learning_rate": 5.543533625253404e-06,
+      "loss": 1.5938,
+      "step": 45960
+    },
+    {
+      "epoch": 1.3375221805276785,
+      "grad_norm": 11.625,
+      "learning_rate": 5.541594343580789e-06,
+      "loss": 1.4579,
+      "step": 45980
+    },
+    {
+      "epoch": 1.3381039648602262,
+      "grad_norm": 14.875,
+      "learning_rate": 5.539655061908174e-06,
+      "loss": 1.5772,
+      "step": 46000
+    },
+    {
+      "epoch": 1.3386857491927742,
+      "grad_norm": 14.0,
+      "learning_rate": 5.5377157802355594e-06,
+      "loss": 1.6044,
+      "step": 46020
+    },
+    {
+      "epoch": 1.3392675335253221,
+      "grad_norm": 13.6875,
+      "learning_rate": 5.535776498562944e-06,
+      "loss": 1.5251,
+      "step": 46040
+    },
+    {
+      "epoch": 1.33984931785787,
+      "grad_norm": 13.125,
+      "learning_rate": 5.533837216890329e-06,
+      "loss": 1.5302,
+      "step": 46060
+    },
+    {
+      "epoch": 1.340431102190418,
+      "grad_norm": 11.9375,
+      "learning_rate": 5.531897935217714e-06,
+      "loss": 1.5406,
+      "step": 46080
+    },
+    {
+      "epoch": 1.341012886522966,
+      "grad_norm": 12.5,
+      "learning_rate": 5.529958653545099e-06,
+      "loss": 1.5536,
+      "step": 46100
+    },
+    {
+      "epoch": 1.341594670855514,
+      "grad_norm": 17.875,
+      "learning_rate": 5.528019371872484e-06,
+      "loss": 1.5169,
+      "step": 46120
+    },
+    {
+      "epoch": 1.3421764551880617,
+      "grad_norm": 16.125,
+      "learning_rate": 5.526080090199869e-06,
+      "loss": 1.5308,
+      "step": 46140
+    },
+    {
+      "epoch": 1.3427582395206097,
+      "grad_norm": 11.3125,
+      "learning_rate": 5.524140808527254e-06,
+      "loss": 1.61,
+      "step": 46160
+    },
+    {
+      "epoch": 1.3433400238531576,
+      "grad_norm": 13.875,
+      "learning_rate": 5.522201526854639e-06,
+      "loss": 1.5143,
+      "step": 46180
+    },
+    {
+      "epoch": 1.3439218081857056,
+      "grad_norm": 13.5,
+      "learning_rate": 5.5202622451820245e-06,
+      "loss": 1.5182,
+      "step": 46200
+    },
+    {
+      "epoch": 1.3445035925182536,
+      "grad_norm": 13.25,
+      "learning_rate": 5.5183229635094096e-06,
+      "loss": 1.49,
+      "step": 46220
+    },
+    {
+      "epoch": 1.3450853768508013,
+      "grad_norm": 13.75,
+      "learning_rate": 5.516383681836795e-06,
+      "loss": 1.4638,
+      "step": 46240
+    },
+    {
+      "epoch": 1.3456671611833493,
+      "grad_norm": 15.5,
+      "learning_rate": 5.51444440016418e-06,
+      "loss": 1.4815,
+      "step": 46260
+    },
+    {
+      "epoch": 1.3462489455158972,
+      "grad_norm": 13.0,
+      "learning_rate": 5.512505118491565e-06,
+      "loss": 1.4858,
+      "step": 46280
+    },
+    {
+      "epoch": 1.3468307298484452,
+      "grad_norm": 15.25,
+      "learning_rate": 5.51056583681895e-06,
+      "loss": 1.5412,
+      "step": 46300
+    },
+    {
+      "epoch": 1.3474125141809932,
+      "grad_norm": 13.375,
+      "learning_rate": 5.508626555146335e-06,
+      "loss": 1.5732,
+      "step": 46320
+    },
+    {
+      "epoch": 1.347994298513541,
+      "grad_norm": 14.1875,
+      "learning_rate": 5.50668727347372e-06,
+      "loss": 1.548,
+      "step": 46340
+    },
+    {
+      "epoch": 1.3485760828460889,
+      "grad_norm": 13.8125,
+      "learning_rate": 5.504747991801105e-06,
+      "loss": 1.4905,
+      "step": 46360
+    },
+    {
+      "epoch": 1.3491578671786368,
+      "grad_norm": 14.25,
+      "learning_rate": 5.50280871012849e-06,
+      "loss": 1.4526,
+      "step": 46380
+    },
+    {
+      "epoch": 1.3497396515111848,
+      "grad_norm": 11.0,
+      "learning_rate": 5.500869428455875e-06,
+      "loss": 1.4639,
+      "step": 46400
+    },
+    {
+      "epoch": 1.3503214358437328,
+      "grad_norm": 13.5625,
+      "learning_rate": 5.49893014678326e-06,
+      "loss": 1.5546,
+      "step": 46420
+    },
+    {
+      "epoch": 1.3509032201762807,
+      "grad_norm": 13.8125,
+      "learning_rate": 5.496990865110645e-06,
+      "loss": 1.5003,
+      "step": 46440
+    },
+    {
+      "epoch": 1.3514850045088287,
+      "grad_norm": 13.3125,
+      "learning_rate": 5.49505158343803e-06,
+      "loss": 1.5832,
+      "step": 46460
+    },
+    {
+      "epoch": 1.3520667888413764,
+      "grad_norm": 13.75,
+      "learning_rate": 5.493112301765415e-06,
+      "loss": 1.5994,
+      "step": 46480
+    },
+    {
+      "epoch": 1.3526485731739244,
+      "grad_norm": 13.625,
+      "learning_rate": 5.4911730200928e-06,
+      "loss": 1.4468,
+      "step": 46500
+    },
+    {
+      "epoch": 1.3532303575064724,
+      "grad_norm": 14.9375,
+      "learning_rate": 5.489233738420185e-06,
+      "loss": 1.4975,
+      "step": 46520
+    },
+    {
+      "epoch": 1.3538121418390203,
+      "grad_norm": 12.375,
+      "learning_rate": 5.48729445674757e-06,
+      "loss": 1.4866,
+      "step": 46540
+    },
+    {
+      "epoch": 1.3543939261715683,
+      "grad_norm": 13.375,
+      "learning_rate": 5.485355175074955e-06,
+      "loss": 1.5129,
+      "step": 46560
+    },
+    {
+      "epoch": 1.354975710504116,
+      "grad_norm": 12.9375,
+      "learning_rate": 5.4834158934023405e-06,
+      "loss": 1.5391,
+      "step": 46580
+    },
+    {
+      "epoch": 1.355557494836664,
+      "grad_norm": 12.8125,
+      "learning_rate": 5.4814766117297256e-06,
+      "loss": 1.5308,
+      "step": 46600
+    },
+    {
+      "epoch": 1.356139279169212,
+      "grad_norm": 13.25,
+      "learning_rate": 5.479537330057111e-06,
+      "loss": 1.5014,
+      "step": 46620
+    },
+    {
+      "epoch": 1.35672106350176,
+      "grad_norm": 13.5,
+      "learning_rate": 5.477598048384496e-06,
+      "loss": 1.5595,
+      "step": 46640
+    },
+    {
+      "epoch": 1.3573028478343079,
+      "grad_norm": 13.5,
+      "learning_rate": 5.475658766711881e-06,
+      "loss": 1.4568,
+      "step": 46660
+    },
+    {
+      "epoch": 1.3578846321668556,
+      "grad_norm": 13.3125,
+      "learning_rate": 5.473719485039266e-06,
+      "loss": 1.5577,
+      "step": 46680
+    },
+    {
+      "epoch": 1.3584664164994038,
+      "grad_norm": 12.6875,
+      "learning_rate": 5.471780203366651e-06,
+      "loss": 1.5837,
+      "step": 46700
+    },
+    {
+      "epoch": 1.3590482008319515,
+      "grad_norm": 12.6875,
+      "learning_rate": 5.469840921694036e-06,
+      "loss": 1.5158,
+      "step": 46720
+    },
+    {
+      "epoch": 1.3596299851644995,
+      "grad_norm": 12.4375,
+      "learning_rate": 5.467901640021421e-06,
+      "loss": 1.5101,
+      "step": 46740
+    },
+    {
+      "epoch": 1.3602117694970475,
+      "grad_norm": 12.75,
+      "learning_rate": 5.465962358348806e-06,
+      "loss": 1.488,
+      "step": 46760
+    },
+    {
+      "epoch": 1.3607935538295954,
+      "grad_norm": 22.5,
+      "learning_rate": 5.4640230766761915e-06,
+      "loss": 1.5384,
+      "step": 46780
+    },
+    {
+      "epoch": 1.3613753381621434,
+      "grad_norm": 13.0625,
+      "learning_rate": 5.4620837950035766e-06,
+      "loss": 1.5471,
+      "step": 46800
+    },
+    {
+      "epoch": 1.3619571224946911,
+      "grad_norm": 14.625,
+      "learning_rate": 5.460144513330962e-06,
+      "loss": 1.5273,
+      "step": 46820
+    },
+    {
+      "epoch": 1.362538906827239,
+      "grad_norm": 13.6875,
+      "learning_rate": 5.458205231658347e-06,
+      "loss": 1.5492,
+      "step": 46840
+    },
+    {
+      "epoch": 1.363120691159787,
+      "grad_norm": 11.625,
+      "learning_rate": 5.456265949985732e-06,
+      "loss": 1.4581,
+      "step": 46860
+    },
+    {
+      "epoch": 1.363702475492335,
+      "grad_norm": 14.6875,
+      "learning_rate": 5.454326668313117e-06,
+      "loss": 1.4906,
+      "step": 46880
+    },
+    {
+      "epoch": 1.364284259824883,
+      "grad_norm": 14.6875,
+      "learning_rate": 5.452387386640502e-06,
+      "loss": 1.399,
+      "step": 46900
+    },
+    {
+      "epoch": 1.3648660441574307,
+      "grad_norm": 13.0,
+      "learning_rate": 5.450448104967887e-06,
+      "loss": 1.4955,
+      "step": 46920
+    },
+    {
+      "epoch": 1.3654478284899787,
+      "grad_norm": 13.0625,
+      "learning_rate": 5.448508823295272e-06,
+      "loss": 1.5175,
+      "step": 46940
+    },
+    {
+      "epoch": 1.3660296128225267,
+      "grad_norm": 13.0,
+      "learning_rate": 5.446569541622657e-06,
+      "loss": 1.5245,
+      "step": 46960
+    },
+    {
+      "epoch": 1.3666113971550746,
+      "grad_norm": 13.0,
+      "learning_rate": 5.4446302599500424e-06,
+      "loss": 1.6267,
+      "step": 46980
+    },
+    {
+      "epoch": 1.3671931814876226,
+      "grad_norm": 12.5625,
+      "learning_rate": 5.4426909782774275e-06,
+      "loss": 1.5281,
+      "step": 47000
+    },
+    {
+      "epoch": 1.3677749658201703,
+      "grad_norm": 13.375,
+      "learning_rate": 5.440751696604813e-06,
+      "loss": 1.5513,
+      "step": 47020
+    },
+    {
+      "epoch": 1.3683567501527185,
+      "grad_norm": 14.125,
+      "learning_rate": 5.438812414932198e-06,
+      "loss": 1.483,
+      "step": 47040
+    },
+    {
+      "epoch": 1.3689385344852663,
+      "grad_norm": 14.8125,
+      "learning_rate": 5.436873133259583e-06,
+      "loss": 1.4988,
+      "step": 47060
+    },
+    {
+      "epoch": 1.3695203188178142,
+      "grad_norm": 12.0625,
+      "learning_rate": 5.434933851586968e-06,
+      "loss": 1.532,
+      "step": 47080
+    },
+    {
+      "epoch": 1.3701021031503622,
+      "grad_norm": 11.125,
+      "learning_rate": 5.432994569914353e-06,
+      "loss": 1.4573,
+      "step": 47100
+    },
+    {
+      "epoch": 1.3706838874829101,
+      "grad_norm": 13.9375,
+      "learning_rate": 5.431055288241738e-06,
+      "loss": 1.494,
+      "step": 47120
+    },
+    {
+      "epoch": 1.371265671815458,
+      "grad_norm": 15.125,
+      "learning_rate": 5.429116006569123e-06,
+      "loss": 1.4989,
+      "step": 47140
+    },
+    {
+      "epoch": 1.3718474561480059,
+      "grad_norm": 12.375,
+      "learning_rate": 5.427176724896508e-06,
+      "loss": 1.5388,
+      "step": 47160
+    },
+    {
+      "epoch": 1.3724292404805538,
+      "grad_norm": 14.625,
+      "learning_rate": 5.425237443223893e-06,
+      "loss": 1.4658,
+      "step": 47180
+    },
+    {
+      "epoch": 1.3730110248131018,
+      "grad_norm": 14.1875,
+      "learning_rate": 5.4232981615512785e-06,
+      "loss": 1.4219,
+      "step": 47200
+    },
+    {
+      "epoch": 1.3735928091456497,
+      "grad_norm": 15.875,
+      "learning_rate": 5.421358879878663e-06,
+      "loss": 1.555,
+      "step": 47220
+    },
+    {
+      "epoch": 1.3741745934781977,
+      "grad_norm": 17.25,
+      "learning_rate": 5.419419598206048e-06,
+      "loss": 1.4773,
+      "step": 47240
+    },
+    {
+      "epoch": 1.3747563778107454,
+      "grad_norm": 13.9375,
+      "learning_rate": 5.417480316533433e-06,
+      "loss": 1.5697,
+      "step": 47260
+    },
+    {
+      "epoch": 1.3753381621432934,
+      "grad_norm": 15.125,
+      "learning_rate": 5.415541034860818e-06,
+      "loss": 1.564,
+      "step": 47280
+    },
+    {
+      "epoch": 1.3759199464758414,
+      "grad_norm": 15.4375,
+      "learning_rate": 5.413601753188203e-06,
+      "loss": 1.4945,
+      "step": 47300
+    },
+    {
+      "epoch": 1.3765017308083893,
+      "grad_norm": 13.9375,
+      "learning_rate": 5.411662471515588e-06,
+      "loss": 1.4841,
+      "step": 47320
+    },
+    {
+      "epoch": 1.3770835151409373,
+      "grad_norm": 12.9375,
+      "learning_rate": 5.409723189842973e-06,
+      "loss": 1.512,
+      "step": 47340
+    },
+    {
+      "epoch": 1.3776652994734853,
+      "grad_norm": 12.125,
+      "learning_rate": 5.4077839081703584e-06,
+      "loss": 1.4335,
+      "step": 47360
+    },
+    {
+      "epoch": 1.3782470838060332,
+      "grad_norm": 11.9375,
+      "learning_rate": 5.4058446264977435e-06,
+      "loss": 1.5114,
+      "step": 47380
+    },
+    {
+      "epoch": 1.378828868138581,
+      "grad_norm": 10.9375,
+      "learning_rate": 5.403905344825129e-06,
+      "loss": 1.452,
+      "step": 47400
+    },
+    {
+      "epoch": 1.379410652471129,
+      "grad_norm": 11.375,
+      "learning_rate": 5.401966063152514e-06,
+      "loss": 1.5196,
+      "step": 47420
+    },
+    {
+      "epoch": 1.379992436803677,
+      "grad_norm": 12.25,
+      "learning_rate": 5.400026781479899e-06,
+      "loss": 1.6073,
+      "step": 47440
+    },
+    {
+      "epoch": 1.3805742211362249,
+      "grad_norm": 11.1875,
+      "learning_rate": 5.398087499807284e-06,
+      "loss": 1.5454,
+      "step": 47460
+    },
+    {
+      "epoch": 1.3811560054687728,
+      "grad_norm": 11.5625,
+      "learning_rate": 5.396148218134669e-06,
+      "loss": 1.5641,
+      "step": 47480
+    },
+    {
+      "epoch": 1.3817377898013206,
+      "grad_norm": 14.0,
+      "learning_rate": 5.394208936462054e-06,
+      "loss": 1.6132,
+      "step": 47500
+    },
+    {
+      "epoch": 1.3823195741338685,
+      "grad_norm": 13.0,
+      "learning_rate": 5.392269654789439e-06,
+      "loss": 1.4819,
+      "step": 47520
+    },
+    {
+      "epoch": 1.3829013584664165,
+      "grad_norm": 13.125,
+      "learning_rate": 5.390330373116824e-06,
+      "loss": 1.6146,
+      "step": 47540
+    },
+    {
+      "epoch": 1.3834831427989644,
+      "grad_norm": 13.875,
+      "learning_rate": 5.388391091444209e-06,
+      "loss": 1.5331,
+      "step": 47560
+    },
+    {
+      "epoch": 1.3840649271315124,
+      "grad_norm": 12.625,
+      "learning_rate": 5.3864518097715945e-06,
+      "loss": 1.6131,
+      "step": 47580
+    },
+    {
+      "epoch": 1.3846467114640602,
+      "grad_norm": 12.1875,
+      "learning_rate": 5.38451252809898e-06,
+      "loss": 1.4784,
+      "step": 47600
+    },
+    {
+      "epoch": 1.3852284957966081,
+      "grad_norm": 14.3125,
+      "learning_rate": 5.382573246426365e-06,
+      "loss": 1.5693,
+      "step": 47620
+    },
+    {
+      "epoch": 1.385810280129156,
+      "grad_norm": 13.6875,
+      "learning_rate": 5.38063396475375e-06,
+      "loss": 1.497,
+      "step": 47640
+    },
+    {
+      "epoch": 1.386392064461704,
+      "grad_norm": 12.875,
+      "learning_rate": 5.378694683081135e-06,
+      "loss": 1.4902,
+      "step": 47660
+    },
+    {
+      "epoch": 1.386973848794252,
+      "grad_norm": 13.875,
+      "learning_rate": 5.37675540140852e-06,
+      "loss": 1.4721,
+      "step": 47680
+    },
+    {
+      "epoch": 1.3875556331268,
+      "grad_norm": 13.3125,
+      "learning_rate": 5.374816119735905e-06,
+      "loss": 1.4737,
+      "step": 47700
+    },
+    {
+      "epoch": 1.388137417459348,
+      "grad_norm": 12.125,
+      "learning_rate": 5.37287683806329e-06,
+      "loss": 1.5667,
+      "step": 47720
+    },
+    {
+      "epoch": 1.3887192017918957,
+      "grad_norm": 11.9375,
+      "learning_rate": 5.370937556390675e-06,
+      "loss": 1.5925,
+      "step": 47740
+    },
+    {
+      "epoch": 1.3893009861244436,
+      "grad_norm": 15.0,
+      "learning_rate": 5.36899827471806e-06,
+      "loss": 1.5047,
+      "step": 47760
+    },
+    {
+      "epoch": 1.3898827704569916,
+      "grad_norm": 13.375,
+      "learning_rate": 5.3670589930454455e-06,
+      "loss": 1.5228,
+      "step": 47780
+    },
+    {
+      "epoch": 1.3904645547895396,
+      "grad_norm": 14.1875,
+      "learning_rate": 5.365119711372831e-06,
+      "loss": 1.4962,
+      "step": 47800
+    },
+    {
+      "epoch": 1.3910463391220875,
+      "grad_norm": 13.5,
+      "learning_rate": 5.363180429700216e-06,
+      "loss": 1.5033,
+      "step": 47820
+    },
+    {
+      "epoch": 1.3916281234546353,
+      "grad_norm": 15.0625,
+      "learning_rate": 5.361241148027601e-06,
+      "loss": 1.5003,
+      "step": 47840
+    },
+    {
+      "epoch": 1.3922099077871832,
+      "grad_norm": 12.75,
+      "learning_rate": 5.359301866354986e-06,
+      "loss": 1.54,
+      "step": 47860
+    },
+    {
+      "epoch": 1.3927916921197312,
+      "grad_norm": 12.9375,
+      "learning_rate": 5.357362584682371e-06,
+      "loss": 1.4891,
+      "step": 47880
+    },
+    {
+      "epoch": 1.3933734764522792,
+      "grad_norm": 15.125,
+      "learning_rate": 5.355423303009756e-06,
+      "loss": 1.4649,
+      "step": 47900
+    },
+    {
+      "epoch": 1.3939552607848271,
+      "grad_norm": 13.5625,
+      "learning_rate": 5.353484021337141e-06,
+      "loss": 1.5445,
+      "step": 47920
+    },
+    {
+      "epoch": 1.3945370451173749,
+      "grad_norm": 13.125,
+      "learning_rate": 5.351544739664526e-06,
+      "loss": 1.5875,
+      "step": 47940
+    },
+    {
+      "epoch": 1.395118829449923,
+      "grad_norm": 12.8125,
+      "learning_rate": 5.349605457991911e-06,
+      "loss": 1.5797,
+      "step": 47960
+    },
+    {
+      "epoch": 1.3957006137824708,
+      "grad_norm": 12.625,
+      "learning_rate": 5.3476661763192965e-06,
+      "loss": 1.5652,
+      "step": 47980
+    },
+    {
+      "epoch": 1.3962823981150188,
+      "grad_norm": 15.375,
+      "learning_rate": 5.3457268946466816e-06,
+      "loss": 1.5665,
+      "step": 48000
+    },
+    {
+      "epoch": 1.3968641824475667,
+      "grad_norm": 16.375,
+      "learning_rate": 5.343787612974067e-06,
+      "loss": 1.4943,
+      "step": 48020
+    },
+    {
+      "epoch": 1.3974459667801147,
+      "grad_norm": 11.5625,
+      "learning_rate": 5.341848331301451e-06,
+      "loss": 1.5187,
+      "step": 48040
+    },
+    {
+      "epoch": 1.3980277511126626,
+      "grad_norm": 13.125,
+      "learning_rate": 5.339909049628836e-06,
+      "loss": 1.5284,
+      "step": 48060
+    },
+    {
+      "epoch": 1.3986095354452104,
+      "grad_norm": 9.375,
+      "learning_rate": 5.337969767956221e-06,
+      "loss": 1.4807,
+      "step": 48080
+    },
+    {
+      "epoch": 1.3991913197777583,
+      "grad_norm": 13.125,
+      "learning_rate": 5.336030486283606e-06,
+      "loss": 1.4139,
+      "step": 48100
+    },
+    {
+      "epoch": 1.3997731041103063,
+      "grad_norm": 13.25,
+      "learning_rate": 5.334091204610991e-06,
+      "loss": 1.5286,
+      "step": 48120
+    },
+    {
+      "epoch": 1.4003548884428543,
+      "grad_norm": 12.0,
+      "learning_rate": 5.332151922938376e-06,
+      "loss": 1.4749,
+      "step": 48140
+    },
+    {
+      "epoch": 1.4009366727754022,
+      "grad_norm": 14.375,
+      "learning_rate": 5.3302126412657615e-06,
+      "loss": 1.5266,
+      "step": 48160
+    },
+    {
+      "epoch": 1.40151845710795,
+      "grad_norm": 12.375,
+      "learning_rate": 5.328273359593147e-06,
+      "loss": 1.5522,
+      "step": 48180
+    },
+    {
+      "epoch": 1.402100241440498,
+      "grad_norm": 15.375,
+      "learning_rate": 5.326334077920532e-06,
+      "loss": 1.502,
+      "step": 48200
+    },
+    {
+      "epoch": 1.402682025773046,
+      "grad_norm": 13.0,
+      "learning_rate": 5.324394796247917e-06,
+      "loss": 1.4469,
+      "step": 48220
+    },
+    {
+      "epoch": 1.4032638101055939,
+      "grad_norm": 12.0,
+      "learning_rate": 5.322455514575302e-06,
+      "loss": 1.6373,
+      "step": 48240
+    },
+    {
+      "epoch": 1.4038455944381418,
+      "grad_norm": 12.9375,
+      "learning_rate": 5.320516232902687e-06,
+      "loss": 1.4664,
+      "step": 48260
+    },
+    {
+      "epoch": 1.4044273787706896,
+      "grad_norm": 13.125,
+      "learning_rate": 5.318576951230072e-06,
+      "loss": 1.5461,
+      "step": 48280
+    },
+    {
+      "epoch": 1.4050091631032378,
+      "grad_norm": 15.9375,
+      "learning_rate": 5.316637669557457e-06,
+      "loss": 1.5507,
+      "step": 48300
+    },
+    {
+      "epoch": 1.4055909474357855,
+      "grad_norm": 13.375,
+      "learning_rate": 5.314698387884842e-06,
+      "loss": 1.5017,
+      "step": 48320
+    },
+    {
+      "epoch": 1.4061727317683335,
+      "grad_norm": 14.6875,
+      "learning_rate": 5.312759106212227e-06,
+      "loss": 1.5485,
+      "step": 48340
+    },
+    {
+      "epoch": 1.4067545161008814,
+      "grad_norm": 14.9375,
+      "learning_rate": 5.3108198245396125e-06,
+      "loss": 1.576,
+      "step": 48360
+    },
+    {
+      "epoch": 1.4073363004334294,
+      "grad_norm": 14.9375,
+      "learning_rate": 5.3088805428669976e-06,
+      "loss": 1.5244,
+      "step": 48380
+    },
+    {
+      "epoch": 1.4079180847659774,
+      "grad_norm": 15.6875,
+      "learning_rate": 5.306941261194383e-06,
+      "loss": 1.539,
+      "step": 48400
+    },
+    {
+      "epoch": 1.408499869098525,
+      "grad_norm": 11.875,
+      "learning_rate": 5.305001979521768e-06,
+      "loss": 1.5648,
+      "step": 48420
+    },
+    {
+      "epoch": 1.409081653431073,
+      "grad_norm": 12.75,
+      "learning_rate": 5.303062697849153e-06,
+      "loss": 1.5098,
+      "step": 48440
+    },
+    {
+      "epoch": 1.409663437763621,
+      "grad_norm": 12.125,
+      "learning_rate": 5.301123416176538e-06,
+      "loss": 1.4494,
+      "step": 48460
+    },
+    {
+      "epoch": 1.410245222096169,
+      "grad_norm": 13.0,
+      "learning_rate": 5.299184134503923e-06,
+      "loss": 1.5425,
+      "step": 48480
+    },
+    {
+      "epoch": 1.410827006428717,
+      "grad_norm": 13.75,
+      "learning_rate": 5.297244852831308e-06,
+      "loss": 1.4474,
+      "step": 48500
+    },
+    {
+      "epoch": 1.4114087907612647,
+      "grad_norm": 11.9375,
+      "learning_rate": 5.295305571158693e-06,
+      "loss": 1.4395,
+      "step": 48520
+    },
+    {
+      "epoch": 1.4119905750938126,
+      "grad_norm": 13.8125,
+      "learning_rate": 5.293366289486078e-06,
+      "loss": 1.5865,
+      "step": 48540
+    },
+    {
+      "epoch": 1.4125723594263606,
+      "grad_norm": 16.375,
+      "learning_rate": 5.2914270078134634e-06,
+      "loss": 1.4994,
+      "step": 48560
+    },
+    {
+      "epoch": 1.4131541437589086,
+      "grad_norm": 11.3125,
+      "learning_rate": 5.2894877261408485e-06,
+      "loss": 1.4271,
+      "step": 48580
+    },
+    {
+      "epoch": 1.4137359280914565,
+      "grad_norm": 11.6875,
+      "learning_rate": 5.287548444468234e-06,
+      "loss": 1.5099,
+      "step": 48600
+    },
+    {
+      "epoch": 1.4143177124240045,
+      "grad_norm": 13.5,
+      "learning_rate": 5.285609162795619e-06,
+      "loss": 1.5398,
+      "step": 48620
+    },
+    {
+      "epoch": 1.4148994967565525,
+      "grad_norm": 12.75,
+      "learning_rate": 5.283669881123004e-06,
+      "loss": 1.5669,
+      "step": 48640
+    },
+    {
+      "epoch": 1.4154812810891002,
+      "grad_norm": 12.0,
+      "learning_rate": 5.281730599450389e-06,
+      "loss": 1.5473,
+      "step": 48660
+    },
+    {
+      "epoch": 1.4160630654216482,
+      "grad_norm": 15.25,
+      "learning_rate": 5.279791317777774e-06,
+      "loss": 1.5169,
+      "step": 48680
+    },
+    {
+      "epoch": 1.4166448497541961,
+      "grad_norm": 12.25,
+      "learning_rate": 5.277852036105159e-06,
+      "loss": 1.5115,
+      "step": 48700
+    },
+    {
+      "epoch": 1.417226634086744,
+      "grad_norm": 13.75,
+      "learning_rate": 5.275912754432544e-06,
+      "loss": 1.5089,
+      "step": 48720
+    },
+    {
+      "epoch": 1.417808418419292,
+      "grad_norm": 14.875,
+      "learning_rate": 5.273973472759929e-06,
+      "loss": 1.5885,
+      "step": 48740
+    },
+    {
+      "epoch": 1.4183902027518398,
+      "grad_norm": 13.6875,
+      "learning_rate": 5.272034191087314e-06,
+      "loss": 1.5541,
+      "step": 48760
+    },
+    {
+      "epoch": 1.4189719870843878,
+      "grad_norm": 11.0,
+      "learning_rate": 5.2700949094146995e-06,
+      "loss": 1.532,
+      "step": 48780
+    },
+    {
+      "epoch": 1.4195537714169357,
+      "grad_norm": 13.625,
+      "learning_rate": 5.268155627742085e-06,
+      "loss": 1.5522,
+      "step": 48800
+    },
+    {
+      "epoch": 1.4201355557494837,
+      "grad_norm": 11.8125,
+      "learning_rate": 5.26621634606947e-06,
+      "loss": 1.4616,
+      "step": 48820
+    },
+    {
+      "epoch": 1.4207173400820317,
+      "grad_norm": 12.3125,
+      "learning_rate": 5.264277064396855e-06,
+      "loss": 1.5783,
+      "step": 48840
+    },
+    {
+      "epoch": 1.4212991244145794,
+      "grad_norm": 12.125,
+      "learning_rate": 5.262337782724239e-06,
+      "loss": 1.4606,
+      "step": 48860
+    },
+    {
+      "epoch": 1.4218809087471274,
+      "grad_norm": 13.0,
+      "learning_rate": 5.260398501051624e-06,
+      "loss": 1.5216,
+      "step": 48880
+    },
+    {
+      "epoch": 1.4224626930796753,
+      "grad_norm": 11.5,
+      "learning_rate": 5.258459219379009e-06,
+      "loss": 1.552,
+      "step": 48900
+    },
+    {
+      "epoch": 1.4230444774122233,
+      "grad_norm": 12.875,
+      "learning_rate": 5.256519937706394e-06,
+      "loss": 1.5649,
+      "step": 48920
+    },
+    {
+      "epoch": 1.4236262617447712,
+      "grad_norm": 16.125,
+      "learning_rate": 5.2545806560337794e-06,
+      "loss": 1.4982,
+      "step": 48940
+    },
+    {
+      "epoch": 1.4242080460773192,
+      "grad_norm": 12.1875,
+      "learning_rate": 5.2526413743611645e-06,
+      "loss": 1.5495,
+      "step": 48960
+    },
+    {
+      "epoch": 1.4247898304098672,
+      "grad_norm": 12.1875,
+      "learning_rate": 5.25070209268855e-06,
+      "loss": 1.5104,
+      "step": 48980
+    },
+    {
+      "epoch": 1.425371614742415,
+      "grad_norm": 11.9375,
+      "learning_rate": 5.248762811015935e-06,
+      "loss": 1.4486,
+      "step": 49000
+    },
+    {
+      "epoch": 1.4259533990749629,
+      "grad_norm": 11.625,
+      "learning_rate": 5.24682352934332e-06,
+      "loss": 1.4121,
+      "step": 49020
+    },
+    {
+      "epoch": 1.4265351834075108,
+      "grad_norm": 13.25,
+      "learning_rate": 5.244884247670705e-06,
+      "loss": 1.5312,
+      "step": 49040
+    },
+    {
+      "epoch": 1.4271169677400588,
+      "grad_norm": 13.875,
+      "learning_rate": 5.24294496599809e-06,
+      "loss": 1.4839,
+      "step": 49060
+    },
+    {
+      "epoch": 1.4276987520726068,
+      "grad_norm": 14.625,
+      "learning_rate": 5.241005684325475e-06,
+      "loss": 1.5612,
+      "step": 49080
+    },
+    {
+      "epoch": 1.4282805364051545,
+      "grad_norm": 14.8125,
+      "learning_rate": 5.23906640265286e-06,
+      "loss": 1.5692,
+      "step": 49100
+    },
+    {
+      "epoch": 1.4288623207377025,
+      "grad_norm": 15.125,
+      "learning_rate": 5.237127120980245e-06,
+      "loss": 1.5126,
+      "step": 49120
+    },
+    {
+      "epoch": 1.4294441050702504,
+      "grad_norm": 13.375,
+      "learning_rate": 5.23518783930763e-06,
+      "loss": 1.6138,
+      "step": 49140
+    },
+    {
+      "epoch": 1.4300258894027984,
+      "grad_norm": 13.0,
+      "learning_rate": 5.2332485576350155e-06,
+      "loss": 1.4812,
+      "step": 49160
+    },
+    {
+      "epoch": 1.4306076737353464,
+      "grad_norm": 13.5,
+      "learning_rate": 5.231309275962401e-06,
+      "loss": 1.5094,
+      "step": 49180
+    },
+    {
+      "epoch": 1.431189458067894,
+      "grad_norm": 13.25,
+      "learning_rate": 5.229369994289786e-06,
+      "loss": 1.5418,
+      "step": 49200
+    },
+    {
+      "epoch": 1.4317712424004423,
+      "grad_norm": 13.875,
+      "learning_rate": 5.227430712617171e-06,
+      "loss": 1.7129,
+      "step": 49220
+    },
+    {
+      "epoch": 1.43235302673299,
+      "grad_norm": 12.5625,
+      "learning_rate": 5.225491430944556e-06,
+      "loss": 1.459,
+      "step": 49240
+    },
+    {
+      "epoch": 1.432934811065538,
+      "grad_norm": 12.625,
+      "learning_rate": 5.223552149271941e-06,
+      "loss": 1.4942,
+      "step": 49260
+    },
+    {
+      "epoch": 1.433516595398086,
+      "grad_norm": 12.3125,
+      "learning_rate": 5.221612867599326e-06,
+      "loss": 1.5423,
+      "step": 49280
+    },
+    {
+      "epoch": 1.434098379730634,
+      "grad_norm": 11.75,
+      "learning_rate": 5.219673585926711e-06,
+      "loss": 1.4238,
+      "step": 49300
+    },
+    {
+      "epoch": 1.4346801640631819,
+      "grad_norm": 13.0625,
+      "learning_rate": 5.217734304254096e-06,
+      "loss": 1.4657,
+      "step": 49320
+    },
+    {
+      "epoch": 1.4352619483957296,
+      "grad_norm": 13.3125,
+      "learning_rate": 5.215795022581481e-06,
+      "loss": 1.4548,
+      "step": 49340
+    },
+    {
+      "epoch": 1.4358437327282776,
+      "grad_norm": 17.125,
+      "learning_rate": 5.2138557409088665e-06,
+      "loss": 1.5447,
+      "step": 49360
+    },
+    {
+      "epoch": 1.4364255170608256,
+      "grad_norm": 13.1875,
+      "learning_rate": 5.211916459236252e-06,
+      "loss": 1.485,
+      "step": 49380
+    },
+    {
+      "epoch": 1.4370073013933735,
+      "grad_norm": 9.8125,
+      "learning_rate": 5.209977177563637e-06,
+      "loss": 1.59,
+      "step": 49400
+    },
+    {
+      "epoch": 1.4375890857259215,
+      "grad_norm": 12.3125,
+      "learning_rate": 5.208037895891022e-06,
+      "loss": 1.4542,
+      "step": 49420
+    },
+    {
+      "epoch": 1.4381708700584692,
+      "grad_norm": 13.875,
+      "learning_rate": 5.206098614218407e-06,
+      "loss": 1.5553,
+      "step": 49440
+    },
+    {
+      "epoch": 1.4387526543910172,
+      "grad_norm": 13.0,
+      "learning_rate": 5.204159332545792e-06,
+      "loss": 1.5352,
+      "step": 49460
+    },
+    {
+      "epoch": 1.4393344387235651,
+      "grad_norm": 12.75,
+      "learning_rate": 5.202220050873177e-06,
+      "loss": 1.498,
+      "step": 49480
+    },
+    {
+      "epoch": 1.439916223056113,
+      "grad_norm": 12.0625,
+      "learning_rate": 5.200280769200562e-06,
+      "loss": 1.509,
+      "step": 49500
+    },
+    {
+      "epoch": 1.440498007388661,
+      "grad_norm": 13.125,
+      "learning_rate": 5.198341487527947e-06,
+      "loss": 1.4639,
+      "step": 49520
+    },
+    {
+      "epoch": 1.4410797917212088,
+      "grad_norm": 10.875,
+      "learning_rate": 5.196402205855332e-06,
+      "loss": 1.516,
+      "step": 49540
+    },
+    {
+      "epoch": 1.441661576053757,
+      "grad_norm": 14.75,
+      "learning_rate": 5.1944629241827175e-06,
+      "loss": 1.5425,
+      "step": 49560
+    },
+    {
+      "epoch": 1.4422433603863047,
+      "grad_norm": 12.5,
+      "learning_rate": 5.1925236425101026e-06,
+      "loss": 1.5151,
+      "step": 49580
+    },
+    {
+      "epoch": 1.4428251447188527,
+      "grad_norm": 12.125,
+      "learning_rate": 5.190584360837488e-06,
+      "loss": 1.4453,
+      "step": 49600
+    },
+    {
+      "epoch": 1.4434069290514007,
+      "grad_norm": 14.8125,
+      "learning_rate": 5.188645079164873e-06,
+      "loss": 1.5709,
+      "step": 49620
+    },
+    {
+      "epoch": 1.4439887133839486,
+      "grad_norm": 11.3125,
+      "learning_rate": 5.186705797492258e-06,
+      "loss": 1.5566,
+      "step": 49640
+    },
+    {
+      "epoch": 1.4445704977164966,
+      "grad_norm": 10.6875,
+      "learning_rate": 5.184766515819642e-06,
+      "loss": 1.5536,
+      "step": 49660
+    },
+    {
+      "epoch": 1.4451522820490443,
+      "grad_norm": 13.375,
+      "learning_rate": 5.182827234147027e-06,
+      "loss": 1.57,
+      "step": 49680
+    },
+    {
+      "epoch": 1.4457340663815923,
+      "grad_norm": 11.875,
+      "learning_rate": 5.180887952474412e-06,
+      "loss": 1.5094,
+      "step": 49700
+    },
+    {
+      "epoch": 1.4463158507141403,
+      "grad_norm": 12.625,
+      "learning_rate": 5.178948670801797e-06,
+      "loss": 1.5041,
+      "step": 49720
+    },
+    {
+      "epoch": 1.4468976350466882,
+      "grad_norm": 12.9375,
+      "learning_rate": 5.1770093891291825e-06,
+      "loss": 1.5361,
+      "step": 49740
+    },
+    {
+      "epoch": 1.4474794193792362,
+      "grad_norm": 13.4375,
+      "learning_rate": 5.175070107456568e-06,
+      "loss": 1.51,
+      "step": 49760
+    },
+    {
+      "epoch": 1.448061203711784,
+      "grad_norm": 12.8125,
+      "learning_rate": 5.173130825783953e-06,
+      "loss": 1.4887,
+      "step": 49780
+    },
+    {
+      "epoch": 1.448642988044332,
+      "grad_norm": 12.8125,
+      "learning_rate": 5.171191544111338e-06,
+      "loss": 1.5794,
+      "step": 49800
+    },
+    {
+      "epoch": 1.4492247723768799,
+      "grad_norm": 12.75,
+      "learning_rate": 5.169252262438723e-06,
+      "loss": 1.5677,
+      "step": 49820
+    },
+    {
+      "epoch": 1.4498065567094278,
+      "grad_norm": 14.0,
+      "learning_rate": 5.167312980766108e-06,
+      "loss": 1.5516,
+      "step": 49840
+    },
+    {
+      "epoch": 1.4503883410419758,
+      "grad_norm": 11.875,
+      "learning_rate": 5.165373699093493e-06,
+      "loss": 1.5398,
+      "step": 49860
+    },
+    {
+      "epoch": 1.4509701253745237,
+      "grad_norm": 12.5,
+      "learning_rate": 5.163434417420878e-06,
+      "loss": 1.5098,
+      "step": 49880
+    },
+    {
+      "epoch": 1.4515519097070717,
+      "grad_norm": 11.625,
+      "learning_rate": 5.161495135748263e-06,
+      "loss": 1.5095,
+      "step": 49900
+    },
+    {
+      "epoch": 1.4521336940396194,
+      "grad_norm": 12.75,
+      "learning_rate": 5.159555854075648e-06,
+      "loss": 1.5098,
+      "step": 49920
+    },
+    {
+      "epoch": 1.4527154783721674,
+      "grad_norm": 14.0,
+      "learning_rate": 5.1576165724030335e-06,
+      "loss": 1.4726,
+      "step": 49940
+    },
+    {
+      "epoch": 1.4532972627047154,
+      "grad_norm": 15.6875,
+      "learning_rate": 5.1556772907304186e-06,
+      "loss": 1.5539,
+      "step": 49960
+    },
+    {
+      "epoch": 1.4538790470372633,
+      "grad_norm": 13.0,
+      "learning_rate": 5.153738009057804e-06,
+      "loss": 1.5476,
+      "step": 49980
+    },
+    {
+      "epoch": 1.4544608313698113,
+      "grad_norm": 10.6875,
+      "learning_rate": 5.151798727385189e-06,
+      "loss": 1.5702,
+      "step": 50000
+    },
+    {
+      "epoch": 1.455042615702359,
+      "grad_norm": 12.9375,
+      "learning_rate": 5.149859445712574e-06,
+      "loss": 1.5263,
+      "step": 50020
+    },
+    {
+      "epoch": 1.455624400034907,
+      "grad_norm": 12.5625,
+      "learning_rate": 5.147920164039959e-06,
+      "loss": 1.5251,
+      "step": 50040
+    },
+    {
+      "epoch": 1.456206184367455,
+      "grad_norm": 11.1875,
+      "learning_rate": 5.145980882367344e-06,
+      "loss": 1.5967,
+      "step": 50060
+    },
+    {
+      "epoch": 1.456787968700003,
+      "grad_norm": 13.0625,
+      "learning_rate": 5.144041600694729e-06,
+      "loss": 1.4547,
+      "step": 50080
+    },
+    {
+      "epoch": 1.457369753032551,
+      "grad_norm": 12.875,
+      "learning_rate": 5.142102319022114e-06,
+      "loss": 1.5841,
+      "step": 50100
+    },
+    {
+      "epoch": 1.4579515373650986,
+      "grad_norm": 10.875,
+      "learning_rate": 5.140163037349499e-06,
+      "loss": 1.5134,
+      "step": 50120
+    },
+    {
+      "epoch": 1.4585333216976466,
+      "grad_norm": 17.625,
+      "learning_rate": 5.1382237556768844e-06,
+      "loss": 1.5292,
+      "step": 50140
+    },
+    {
+      "epoch": 1.4591151060301946,
+      "grad_norm": 12.0625,
+      "learning_rate": 5.1362844740042695e-06,
+      "loss": 1.486,
+      "step": 50160
+    },
+    {
+      "epoch": 1.4596968903627425,
+      "grad_norm": 17.25,
+      "learning_rate": 5.134345192331655e-06,
+      "loss": 1.5385,
+      "step": 50180
+    },
+    {
+      "epoch": 1.4602786746952905,
+      "grad_norm": 10.0,
+      "learning_rate": 5.13240591065904e-06,
+      "loss": 1.5024,
+      "step": 50200
+    },
+    {
+      "epoch": 1.4608604590278385,
+      "grad_norm": 13.75,
+      "learning_rate": 5.130466628986425e-06,
+      "loss": 1.529,
+      "step": 50220
+    },
+    {
+      "epoch": 1.4614422433603864,
+      "grad_norm": 14.875,
+      "learning_rate": 5.12852734731381e-06,
+      "loss": 1.4843,
+      "step": 50240
+    },
+    {
+      "epoch": 1.4620240276929342,
+      "grad_norm": 12.0625,
+      "learning_rate": 5.126588065641195e-06,
+      "loss": 1.506,
+      "step": 50260
+    },
+    {
+      "epoch": 1.4626058120254821,
+      "grad_norm": 12.6875,
+      "learning_rate": 5.12464878396858e-06,
+      "loss": 1.4215,
+      "step": 50280
+    },
+    {
+      "epoch": 1.46318759635803,
+      "grad_norm": 12.375,
+      "learning_rate": 5.1227095022959635e-06,
+      "loss": 1.4734,
+      "step": 50300
+    },
+    {
+      "epoch": 1.463769380690578,
+      "grad_norm": 15.5,
+      "learning_rate": 5.120770220623349e-06,
+      "loss": 1.6313,
+      "step": 50320
+    },
+    {
+      "epoch": 1.464351165023126,
+      "grad_norm": 13.4375,
+      "learning_rate": 5.118830938950734e-06,
+      "loss": 1.4654,
+      "step": 50340
+    },
+    {
+      "epoch": 1.4649329493556738,
+      "grad_norm": 14.0,
+      "learning_rate": 5.116891657278119e-06,
+      "loss": 1.471,
+      "step": 50360
+    },
+    {
+      "epoch": 1.4655147336882217,
+      "grad_norm": 11.0,
+      "learning_rate": 5.114952375605504e-06,
+      "loss": 1.5427,
+      "step": 50380
+    },
+    {
+      "epoch": 1.4660965180207697,
+      "grad_norm": 14.625,
+      "learning_rate": 5.113013093932889e-06,
+      "loss": 1.572,
+      "step": 50400
+    },
+    {
+      "epoch": 1.4666783023533176,
+      "grad_norm": 16.75,
+      "learning_rate": 5.111073812260274e-06,
+      "loss": 1.5277,
+      "step": 50420
+    },
+    {
+      "epoch": 1.4672600866858656,
+      "grad_norm": 9.6875,
+      "learning_rate": 5.109134530587659e-06,
+      "loss": 1.471,
+      "step": 50440
+    },
+    {
+      "epoch": 1.4678418710184133,
+      "grad_norm": 12.8125,
+      "learning_rate": 5.107195248915044e-06,
+      "loss": 1.5825,
+      "step": 50460
+    },
+    {
+      "epoch": 1.4684236553509615,
+      "grad_norm": 14.0,
+      "learning_rate": 5.105255967242429e-06,
+      "loss": 1.5129,
+      "step": 50480
+    },
+    {
+      "epoch": 1.4690054396835093,
+      "grad_norm": 12.9375,
+      "learning_rate": 5.1033166855698145e-06,
+      "loss": 1.4023,
+      "step": 50500
+    },
+    {
+      "epoch": 1.4695872240160572,
+      "grad_norm": 13.375,
+      "learning_rate": 5.1013774038972e-06,
+      "loss": 1.5123,
+      "step": 50520
+    },
+    {
+      "epoch": 1.4701690083486052,
+      "grad_norm": 11.6875,
+      "learning_rate": 5.099438122224585e-06,
+      "loss": 1.5481,
+      "step": 50540
+    },
+    {
+      "epoch": 1.4707507926811532,
+      "grad_norm": 16.125,
+      "learning_rate": 5.09749884055197e-06,
+      "loss": 1.4437,
+      "step": 50560
+    },
+    {
+      "epoch": 1.4713325770137011,
+      "grad_norm": 12.25,
+      "learning_rate": 5.095559558879355e-06,
+      "loss": 1.4811,
+      "step": 50580
+    },
+    {
+      "epoch": 1.4719143613462489,
+      "grad_norm": 13.6875,
+      "learning_rate": 5.09362027720674e-06,
+      "loss": 1.5479,
+      "step": 50600
+    },
+    {
+      "epoch": 1.4724961456787968,
+      "grad_norm": 14.375,
+      "learning_rate": 5.091680995534125e-06,
+      "loss": 1.5808,
+      "step": 50620
+    },
+    {
+      "epoch": 1.4730779300113448,
+      "grad_norm": 13.0625,
+      "learning_rate": 5.08974171386151e-06,
+      "loss": 1.6277,
+      "step": 50640
+    },
+    {
+      "epoch": 1.4736597143438928,
+      "grad_norm": 14.6875,
+      "learning_rate": 5.087802432188895e-06,
+      "loss": 1.5107,
+      "step": 50660
+    },
+    {
+      "epoch": 1.4742414986764407,
+      "grad_norm": 14.5625,
+      "learning_rate": 5.08586315051628e-06,
+      "loss": 1.595,
+      "step": 50680
+    },
+    {
+      "epoch": 1.4748232830089885,
+      "grad_norm": 13.9375,
+      "learning_rate": 5.0839238688436655e-06,
+      "loss": 1.5751,
+      "step": 50700
+    },
+    {
+      "epoch": 1.4754050673415364,
+      "grad_norm": 13.125,
+      "learning_rate": 5.0819845871710506e-06,
+      "loss": 1.4475,
+      "step": 50720
+    },
+    {
+      "epoch": 1.4759868516740844,
+      "grad_norm": 13.5625,
+      "learning_rate": 5.080045305498436e-06,
+      "loss": 1.5289,
+      "step": 50740
+    },
+    {
+      "epoch": 1.4765686360066324,
+      "grad_norm": 15.9375,
+      "learning_rate": 5.078106023825821e-06,
+      "loss": 1.497,
+      "step": 50760
+    },
+    {
+      "epoch": 1.4771504203391803,
+      "grad_norm": 12.5625,
+      "learning_rate": 5.076166742153206e-06,
+      "loss": 1.4242,
+      "step": 50780
+    },
+    {
+      "epoch": 1.477732204671728,
+      "grad_norm": 12.4375,
+      "learning_rate": 5.074227460480591e-06,
+      "loss": 1.5828,
+      "step": 50800
+    },
+    {
+      "epoch": 1.4783139890042762,
+      "grad_norm": 12.125,
+      "learning_rate": 5.072288178807976e-06,
+      "loss": 1.4836,
+      "step": 50820
+    },
+    {
+      "epoch": 1.478895773336824,
+      "grad_norm": 14.0,
+      "learning_rate": 5.070348897135361e-06,
+      "loss": 1.5094,
+      "step": 50840
+    },
+    {
+      "epoch": 1.479477557669372,
+      "grad_norm": 13.5,
+      "learning_rate": 5.068409615462746e-06,
+      "loss": 1.5181,
+      "step": 50860
+    },
+    {
+      "epoch": 1.48005934200192,
+      "grad_norm": 12.625,
+      "learning_rate": 5.066470333790131e-06,
+      "loss": 1.4571,
+      "step": 50880
+    },
+    {
+      "epoch": 1.4806411263344679,
+      "grad_norm": 13.625,
+      "learning_rate": 5.0645310521175165e-06,
+      "loss": 1.4324,
+      "step": 50900
+    },
+    {
+      "epoch": 1.4812229106670158,
+      "grad_norm": 11.9375,
+      "learning_rate": 5.0625917704449015e-06,
+      "loss": 1.576,
+      "step": 50920
+    },
+    {
+      "epoch": 1.4818046949995636,
+      "grad_norm": 15.125,
+      "learning_rate": 5.060652488772287e-06,
+      "loss": 1.5361,
+      "step": 50940
+    },
+    {
+      "epoch": 1.4823864793321115,
+      "grad_norm": 11.8125,
+      "learning_rate": 5.058713207099672e-06,
+      "loss": 1.5862,
+      "step": 50960
+    },
+    {
+      "epoch": 1.4829682636646595,
+      "grad_norm": 13.0625,
+      "learning_rate": 5.056773925427057e-06,
+      "loss": 1.5856,
+      "step": 50980
+    },
+    {
+      "epoch": 1.4835500479972075,
+      "grad_norm": 15.5,
+      "learning_rate": 5.054834643754442e-06,
+      "loss": 1.5927,
+      "step": 51000
+    },
+    {
+      "epoch": 1.4841318323297554,
+      "grad_norm": 12.6875,
+      "learning_rate": 5.052895362081827e-06,
+      "loss": 1.5097,
+      "step": 51020
+    },
+    {
+      "epoch": 1.4847136166623032,
+      "grad_norm": 14.375,
+      "learning_rate": 5.050956080409211e-06,
+      "loss": 1.4944,
+      "step": 51040
+    },
+    {
+      "epoch": 1.4852954009948511,
+      "grad_norm": 12.5625,
+      "learning_rate": 5.049016798736596e-06,
+      "loss": 1.572,
+      "step": 51060
+    },
+    {
+      "epoch": 1.485877185327399,
+      "grad_norm": 16.25,
+      "learning_rate": 5.0470775170639815e-06,
+      "loss": 1.5845,
+      "step": 51080
+    },
+    {
+      "epoch": 1.486458969659947,
+      "grad_norm": 13.625,
+      "learning_rate": 5.045138235391367e-06,
+      "loss": 1.5383,
+      "step": 51100
+    },
+    {
+      "epoch": 1.487040753992495,
+      "grad_norm": 16.0,
+      "learning_rate": 5.043198953718752e-06,
+      "loss": 1.5037,
+      "step": 51120
+    },
+    {
+      "epoch": 1.487622538325043,
+      "grad_norm": 14.3125,
+      "learning_rate": 5.041259672046137e-06,
+      "loss": 1.4387,
+      "step": 51140
+    },
+    {
+      "epoch": 1.488204322657591,
+      "grad_norm": 13.4375,
+      "learning_rate": 5.039320390373522e-06,
+      "loss": 1.4998,
+      "step": 51160
+    },
+    {
+      "epoch": 1.4887861069901387,
+      "grad_norm": 20.75,
+      "learning_rate": 5.037381108700907e-06,
+      "loss": 1.5381,
+      "step": 51180
+    },
+    {
+      "epoch": 1.4893678913226867,
+      "grad_norm": 10.8125,
+      "learning_rate": 5.035441827028292e-06,
+      "loss": 1.5232,
+      "step": 51200
+    },
+    {
+      "epoch": 1.4899496756552346,
+      "grad_norm": 13.875,
+      "learning_rate": 5.033502545355677e-06,
+      "loss": 1.5357,
+      "step": 51220
+    },
+    {
+      "epoch": 1.4905314599877826,
+      "grad_norm": 13.375,
+      "learning_rate": 5.031563263683062e-06,
+      "loss": 1.4933,
+      "step": 51240
+    },
+    {
+      "epoch": 1.4911132443203305,
+      "grad_norm": 13.375,
+      "learning_rate": 5.029623982010447e-06,
+      "loss": 1.4801,
+      "step": 51260
+    },
+    {
+      "epoch": 1.4916950286528783,
+      "grad_norm": 15.25,
+      "learning_rate": 5.0276847003378325e-06,
+      "loss": 1.5051,
+      "step": 51280
+    },
+    {
+      "epoch": 1.4922768129854262,
+      "grad_norm": 17.375,
+      "learning_rate": 5.0257454186652176e-06,
+      "loss": 1.4838,
+      "step": 51300
+    },
+    {
+      "epoch": 1.4928585973179742,
+      "grad_norm": 12.625,
+      "learning_rate": 5.023806136992603e-06,
+      "loss": 1.5299,
+      "step": 51320
+    },
+    {
+      "epoch": 1.4934403816505222,
+      "grad_norm": 11.75,
+      "learning_rate": 5.021866855319988e-06,
+      "loss": 1.4594,
+      "step": 51340
+    },
+    {
+      "epoch": 1.4940221659830701,
+      "grad_norm": 14.3125,
+      "learning_rate": 5.019927573647373e-06,
+      "loss": 1.5199,
+      "step": 51360
+    },
+    {
+      "epoch": 1.4946039503156179,
+      "grad_norm": 13.0,
+      "learning_rate": 5.017988291974758e-06,
+      "loss": 1.5827,
+      "step": 51380
+    },
+    {
+      "epoch": 1.4951857346481658,
+      "grad_norm": 11.3125,
+      "learning_rate": 5.016049010302143e-06,
+      "loss": 1.5119,
+      "step": 51400
+    },
+    {
+      "epoch": 1.4957675189807138,
+      "grad_norm": 14.4375,
+      "learning_rate": 5.014109728629528e-06,
+      "loss": 1.5574,
+      "step": 51420
+    },
+    {
+      "epoch": 1.4963493033132618,
+      "grad_norm": 11.1875,
+      "learning_rate": 5.012170446956913e-06,
+      "loss": 1.4693,
+      "step": 51440
+    },
+    {
+      "epoch": 1.4969310876458097,
+      "grad_norm": 13.8125,
+      "learning_rate": 5.010231165284298e-06,
+      "loss": 1.6051,
+      "step": 51460
+    },
+    {
+      "epoch": 1.4975128719783577,
+      "grad_norm": 13.3125,
+      "learning_rate": 5.0082918836116834e-06,
+      "loss": 1.5523,
+      "step": 51480
+    },
+    {
+      "epoch": 1.4980946563109057,
+      "grad_norm": 13.125,
+      "learning_rate": 5.0063526019390685e-06,
+      "loss": 1.5153,
+      "step": 51500
+    },
+    {
+      "epoch": 1.4986764406434534,
+      "grad_norm": 14.0625,
+      "learning_rate": 5.004413320266454e-06,
+      "loss": 1.492,
+      "step": 51520
+    },
+    {
+      "epoch": 1.4992582249760014,
+      "grad_norm": 11.5,
+      "learning_rate": 5.002474038593839e-06,
+      "loss": 1.5005,
+      "step": 51540
+    },
+    {
+      "epoch": 1.4998400093085493,
+      "grad_norm": 13.9375,
+      "learning_rate": 5.000534756921224e-06,
+      "loss": 1.6143,
+      "step": 51560
+    },
+    {
+      "epoch": 1.5004217936410973,
+      "grad_norm": 15.9375,
+      "learning_rate": 4.998595475248609e-06,
+      "loss": 1.5416,
+      "step": 51580
+    },
+    {
+      "epoch": 1.5010035779736453,
+      "grad_norm": 14.75,
+      "learning_rate": 4.996656193575994e-06,
+      "loss": 1.5004,
+      "step": 51600
+    },
+    {
+      "epoch": 1.501585362306193,
+      "grad_norm": 14.875,
+      "learning_rate": 4.994716911903379e-06,
+      "loss": 1.4522,
+      "step": 51620
+    },
+    {
+      "epoch": 1.5021671466387412,
+      "grad_norm": 13.0,
+      "learning_rate": 4.992777630230764e-06,
+      "loss": 1.5349,
+      "step": 51640
+    },
+    {
+      "epoch": 1.502748930971289,
+      "grad_norm": 14.3125,
+      "learning_rate": 4.990838348558149e-06,
+      "loss": 1.4802,
+      "step": 51660
+    },
+    {
+      "epoch": 1.5033307153038369,
+      "grad_norm": 11.0,
+      "learning_rate": 4.988899066885534e-06,
+      "loss": 1.6301,
+      "step": 51680
+    },
+    {
+      "epoch": 1.5039124996363848,
+      "grad_norm": 11.25,
+      "learning_rate": 4.9869597852129195e-06,
+      "loss": 1.49,
+      "step": 51700
+    },
+    {
+      "epoch": 1.5044942839689326,
+      "grad_norm": 14.0625,
+      "learning_rate": 4.985020503540305e-06,
+      "loss": 1.5538,
+      "step": 51720
+    },
+    {
+      "epoch": 1.5050760683014808,
+      "grad_norm": 12.3125,
+      "learning_rate": 4.98308122186769e-06,
+      "loss": 1.5761,
+      "step": 51740
+    },
+    {
+      "epoch": 1.5056578526340285,
+      "grad_norm": 14.125,
+      "learning_rate": 4.981141940195075e-06,
+      "loss": 1.4782,
+      "step": 51760
+    },
+    {
+      "epoch": 1.5062396369665765,
+      "grad_norm": 15.1875,
+      "learning_rate": 4.97920265852246e-06,
+      "loss": 1.4338,
+      "step": 51780
+    },
+    {
+      "epoch": 1.5068214212991244,
+      "grad_norm": 14.6875,
+      "learning_rate": 4.977263376849845e-06,
+      "loss": 1.5249,
+      "step": 51800
+    },
+    {
+      "epoch": 1.5074032056316722,
+      "grad_norm": 13.3125,
+      "learning_rate": 4.97532409517723e-06,
+      "loss": 1.5055,
+      "step": 51820
+    },
+    {
+      "epoch": 1.5079849899642204,
+      "grad_norm": 13.5625,
+      "learning_rate": 4.973384813504615e-06,
+      "loss": 1.5288,
+      "step": 51840
+    },
+    {
+      "epoch": 1.508566774296768,
+      "grad_norm": 11.9375,
+      "learning_rate": 4.9714455318319994e-06,
+      "loss": 1.5375,
+      "step": 51860
+    },
+    {
+      "epoch": 1.509148558629316,
+      "grad_norm": 14.75,
+      "learning_rate": 4.9695062501593845e-06,
+      "loss": 1.4812,
+      "step": 51880
+    },
+    {
+      "epoch": 1.509730342961864,
+      "grad_norm": 12.8125,
+      "learning_rate": 4.96756696848677e-06,
+      "loss": 1.5579,
+      "step": 51900
+    },
+    {
+      "epoch": 1.510312127294412,
+      "grad_norm": 13.125,
+      "learning_rate": 4.965627686814155e-06,
+      "loss": 1.6508,
+      "step": 51920
+    },
+    {
+      "epoch": 1.51089391162696,
+      "grad_norm": 14.375,
+      "learning_rate": 4.96368840514154e-06,
+      "loss": 1.469,
+      "step": 51940
+    },
+    {
+      "epoch": 1.5114756959595077,
+      "grad_norm": 15.4375,
+      "learning_rate": 4.961749123468925e-06,
+      "loss": 1.6221,
+      "step": 51960
+    },
+    {
+      "epoch": 1.5120574802920559,
+      "grad_norm": 14.0625,
+      "learning_rate": 4.95980984179631e-06,
+      "loss": 1.5103,
+      "step": 51980
+    },
+    {
+      "epoch": 1.5126392646246036,
+      "grad_norm": 12.5625,
+      "learning_rate": 4.957870560123695e-06,
+      "loss": 1.5693,
+      "step": 52000
+    },
+    {
+      "epoch": 1.5132210489571516,
+      "grad_norm": 14.6875,
+      "learning_rate": 4.95593127845108e-06,
+      "loss": 1.4608,
+      "step": 52020
+    },
+    {
+      "epoch": 1.5138028332896996,
+      "grad_norm": 13.375,
+      "learning_rate": 4.953991996778465e-06,
+      "loss": 1.533,
+      "step": 52040
+    },
+    {
+      "epoch": 1.5143846176222473,
+      "grad_norm": 13.75,
+      "learning_rate": 4.95205271510585e-06,
+      "loss": 1.5669,
+      "step": 52060
+    },
+    {
+      "epoch": 1.5149664019547955,
+      "grad_norm": 15.8125,
+      "learning_rate": 4.9501134334332355e-06,
+      "loss": 1.519,
+      "step": 52080
+    },
+    {
+      "epoch": 1.5155481862873432,
+      "grad_norm": 12.375,
+      "learning_rate": 4.948174151760621e-06,
+      "loss": 1.5564,
+      "step": 52100
+    },
+    {
+      "epoch": 1.5161299706198912,
+      "grad_norm": 12.6875,
+      "learning_rate": 4.946234870088006e-06,
+      "loss": 1.5168,
+      "step": 52120
+    },
+    {
+      "epoch": 1.5167117549524392,
+      "grad_norm": 12.25,
+      "learning_rate": 4.944295588415391e-06,
+      "loss": 1.593,
+      "step": 52140
+    },
+    {
+      "epoch": 1.517293539284987,
+      "grad_norm": 12.6875,
+      "learning_rate": 4.942356306742776e-06,
+      "loss": 1.6088,
+      "step": 52160
+    },
+    {
+      "epoch": 1.517875323617535,
+      "grad_norm": 12.125,
+      "learning_rate": 4.940417025070161e-06,
+      "loss": 1.4852,
+      "step": 52180
+    },
+    {
+      "epoch": 1.5184571079500828,
+      "grad_norm": 11.25,
+      "learning_rate": 4.938477743397546e-06,
+      "loss": 1.5165,
+      "step": 52200
+    },
+    {
+      "epoch": 1.5190388922826308,
+      "grad_norm": 13.375,
+      "learning_rate": 4.936538461724931e-06,
+      "loss": 1.5314,
+      "step": 52220
+    },
+    {
+      "epoch": 1.5196206766151787,
+      "grad_norm": 14.125,
+      "learning_rate": 4.934599180052316e-06,
+      "loss": 1.5572,
+      "step": 52240
+    },
+    {
+      "epoch": 1.5202024609477267,
+      "grad_norm": 12.6875,
+      "learning_rate": 4.932659898379701e-06,
+      "loss": 1.5518,
+      "step": 52260
+    },
+    {
+      "epoch": 1.5207842452802747,
+      "grad_norm": 11.1875,
+      "learning_rate": 4.9307206167070865e-06,
+      "loss": 1.5268,
+      "step": 52280
+    },
+    {
+      "epoch": 1.5213660296128224,
+      "grad_norm": 16.625,
+      "learning_rate": 4.928781335034472e-06,
+      "loss": 1.5436,
+      "step": 52300
+    },
+    {
+      "epoch": 1.5219478139453706,
+      "grad_norm": 13.9375,
+      "learning_rate": 4.926842053361857e-06,
+      "loss": 1.559,
+      "step": 52320
+    },
+    {
+      "epoch": 1.5225295982779183,
+      "grad_norm": 11.625,
+      "learning_rate": 4.924902771689242e-06,
+      "loss": 1.5251,
+      "step": 52340
+    },
+    {
+      "epoch": 1.5231113826104663,
+      "grad_norm": 12.4375,
+      "learning_rate": 4.922963490016627e-06,
+      "loss": 1.6352,
+      "step": 52360
+    },
+    {
+      "epoch": 1.5236931669430143,
+      "grad_norm": 14.875,
+      "learning_rate": 4.921024208344012e-06,
+      "loss": 1.4764,
+      "step": 52380
+    },
+    {
+      "epoch": 1.524274951275562,
+      "grad_norm": 13.1875,
+      "learning_rate": 4.919084926671397e-06,
+      "loss": 1.5555,
+      "step": 52400
+    },
+    {
+      "epoch": 1.5248567356081102,
+      "grad_norm": 12.25,
+      "learning_rate": 4.917145644998782e-06,
+      "loss": 1.5648,
+      "step": 52420
+    },
+    {
+      "epoch": 1.525438519940658,
+      "grad_norm": 13.0625,
+      "learning_rate": 4.915206363326167e-06,
+      "loss": 1.4756,
+      "step": 52440
+    },
+    {
+      "epoch": 1.526020304273206,
+      "grad_norm": 11.0,
+      "learning_rate": 4.913267081653552e-06,
+      "loss": 1.5144,
+      "step": 52460
+    },
+    {
+      "epoch": 1.5266020886057539,
+      "grad_norm": 13.375,
+      "learning_rate": 4.9113277999809375e-06,
+      "loss": 1.47,
+      "step": 52480
+    },
+    {
+      "epoch": 1.5271838729383018,
+      "grad_norm": 12.125,
+      "learning_rate": 4.9093885183083226e-06,
+      "loss": 1.4436,
+      "step": 52500
+    },
+    {
+      "epoch": 1.5277656572708498,
+      "grad_norm": 12.0625,
+      "learning_rate": 4.907449236635708e-06,
+      "loss": 1.5564,
+      "step": 52520
+    },
+    {
+      "epoch": 1.5283474416033975,
+      "grad_norm": 13.5625,
+      "learning_rate": 4.905509954963093e-06,
+      "loss": 1.4803,
+      "step": 52540
+    },
+    {
+      "epoch": 1.5289292259359455,
+      "grad_norm": 11.9375,
+      "learning_rate": 4.903570673290478e-06,
+      "loss": 1.4645,
+      "step": 52560
+    },
+    {
+      "epoch": 1.5295110102684935,
+      "grad_norm": 13.25,
+      "learning_rate": 4.901631391617863e-06,
+      "loss": 1.5036,
+      "step": 52580
+    },
+    {
+      "epoch": 1.5300927946010414,
+      "grad_norm": 13.5625,
+      "learning_rate": 4.899692109945248e-06,
+      "loss": 1.5826,
+      "step": 52600
+    },
+    {
+      "epoch": 1.5306745789335894,
+      "grad_norm": 13.4375,
+      "learning_rate": 4.897752828272633e-06,
+      "loss": 1.5535,
+      "step": 52620
+    },
+    {
+      "epoch": 1.5312563632661371,
+      "grad_norm": 13.0,
+      "learning_rate": 4.895813546600018e-06,
+      "loss": 1.5745,
+      "step": 52640
+    },
+    {
+      "epoch": 1.5318381475986853,
+      "grad_norm": 12.1875,
+      "learning_rate": 4.893874264927403e-06,
+      "loss": 1.5237,
+      "step": 52660
+    },
+    {
+      "epoch": 1.532419931931233,
+      "grad_norm": 12.375,
+      "learning_rate": 4.891934983254788e-06,
+      "loss": 1.505,
+      "step": 52680
+    },
+    {
+      "epoch": 1.533001716263781,
+      "grad_norm": 11.9375,
+      "learning_rate": 4.889995701582173e-06,
+      "loss": 1.486,
+      "step": 52700
+    },
+    {
+      "epoch": 1.533583500596329,
+      "grad_norm": 13.25,
+      "learning_rate": 4.888056419909558e-06,
+      "loss": 1.5246,
+      "step": 52720
+    },
+    {
+      "epoch": 1.5341652849288767,
+      "grad_norm": 13.5625,
+      "learning_rate": 4.886117138236943e-06,
+      "loss": 1.5384,
+      "step": 52740
+    },
+    {
+      "epoch": 1.534747069261425,
+      "grad_norm": 14.5,
+      "learning_rate": 4.884177856564328e-06,
+      "loss": 1.5441,
+      "step": 52760
+    },
+    {
+      "epoch": 1.5353288535939726,
+      "grad_norm": 13.25,
+      "learning_rate": 4.882238574891713e-06,
+      "loss": 1.5209,
+      "step": 52780
+    },
+    {
+      "epoch": 1.5359106379265206,
+      "grad_norm": 11.4375,
+      "learning_rate": 4.880299293219098e-06,
+      "loss": 1.5664,
+      "step": 52800
+    },
+    {
+      "epoch": 1.5364924222590686,
+      "grad_norm": 12.6875,
+      "learning_rate": 4.878360011546483e-06,
+      "loss": 1.459,
+      "step": 52820
+    },
+    {
+      "epoch": 1.5370742065916165,
+      "grad_norm": 15.6875,
+      "learning_rate": 4.876420729873868e-06,
+      "loss": 1.5468,
+      "step": 52840
+    },
+    {
+      "epoch": 1.5376559909241645,
+      "grad_norm": 15.1875,
+      "learning_rate": 4.8744814482012535e-06,
+      "loss": 1.5809,
+      "step": 52860
+    },
+    {
+      "epoch": 1.5382377752567122,
+      "grad_norm": 11.625,
+      "learning_rate": 4.8725421665286386e-06,
+      "loss": 1.5441,
+      "step": 52880
+    },
+    {
+      "epoch": 1.5388195595892604,
+      "grad_norm": 17.25,
+      "learning_rate": 4.870602884856024e-06,
+      "loss": 1.588,
+      "step": 52900
+    },
+    {
+      "epoch": 1.5394013439218082,
+      "grad_norm": 13.3125,
+      "learning_rate": 4.868663603183409e-06,
+      "loss": 1.4962,
+      "step": 52920
+    },
+    {
+      "epoch": 1.5399831282543561,
+      "grad_norm": 14.0625,
+      "learning_rate": 4.866724321510794e-06,
+      "loss": 1.494,
+      "step": 52940
+    },
+    {
+      "epoch": 1.540564912586904,
+      "grad_norm": 14.1875,
+      "learning_rate": 4.864785039838179e-06,
+      "loss": 1.5254,
+      "step": 52960
+    },
+    {
+      "epoch": 1.5411466969194518,
+      "grad_norm": 15.9375,
+      "learning_rate": 4.862845758165564e-06,
+      "loss": 1.5036,
+      "step": 52980
+    },
+    {
+      "epoch": 1.541728481252,
+      "grad_norm": 13.125,
+      "learning_rate": 4.860906476492949e-06,
+      "loss": 1.4845,
+      "step": 53000
+    },
+    {
+      "epoch": 1.5423102655845478,
+      "grad_norm": 12.625,
+      "learning_rate": 4.858967194820334e-06,
+      "loss": 1.5102,
+      "step": 53020
+    },
+    {
+      "epoch": 1.5428920499170957,
+      "grad_norm": 10.5625,
+      "learning_rate": 4.857027913147719e-06,
+      "loss": 1.5368,
+      "step": 53040
+    },
+    {
+      "epoch": 1.5434738342496437,
+      "grad_norm": 13.875,
+      "learning_rate": 4.8550886314751044e-06,
+      "loss": 1.5532,
+      "step": 53060
+    },
+    {
+      "epoch": 1.5440556185821914,
+      "grad_norm": 13.75,
+      "learning_rate": 4.8531493498024895e-06,
+      "loss": 1.5391,
+      "step": 53080
+    },
+    {
+      "epoch": 1.5446374029147396,
+      "grad_norm": 14.0,
+      "learning_rate": 4.851210068129875e-06,
+      "loss": 1.5926,
+      "step": 53100
+    },
+    {
+      "epoch": 1.5452191872472874,
+      "grad_norm": 12.5625,
+      "learning_rate": 4.84927078645726e-06,
+      "loss": 1.528,
+      "step": 53120
+    },
+    {
+      "epoch": 1.5458009715798353,
+      "grad_norm": 15.0,
+      "learning_rate": 4.847331504784645e-06,
+      "loss": 1.481,
+      "step": 53140
+    },
+    {
+      "epoch": 1.5463827559123833,
+      "grad_norm": 14.5625,
+      "learning_rate": 4.84539222311203e-06,
+      "loss": 1.5169,
+      "step": 53160
+    },
+    {
+      "epoch": 1.5469645402449312,
+      "grad_norm": 9.125,
+      "learning_rate": 4.843452941439415e-06,
+      "loss": 1.537,
+      "step": 53180
+    },
+    {
+      "epoch": 1.5475463245774792,
+      "grad_norm": 12.0625,
+      "learning_rate": 4.8415136597668e-06,
+      "loss": 1.4713,
+      "step": 53200
+    },
+    {
+      "epoch": 1.548128108910027,
+      "grad_norm": 15.375,
+      "learning_rate": 4.839574378094185e-06,
+      "loss": 1.5724,
+      "step": 53220
+    },
+    {
+      "epoch": 1.5487098932425751,
+      "grad_norm": 11.25,
+      "learning_rate": 4.83763509642157e-06,
+      "loss": 1.5441,
+      "step": 53240
+    },
+    {
+      "epoch": 1.5492916775751229,
+      "grad_norm": 16.5,
+      "learning_rate": 4.835695814748955e-06,
+      "loss": 1.5224,
+      "step": 53260
+    },
+    {
+      "epoch": 1.5498734619076708,
+      "grad_norm": 11.9375,
+      "learning_rate": 4.8337565330763405e-06,
+      "loss": 1.4652,
+      "step": 53280
+    },
+    {
+      "epoch": 1.5504552462402188,
+      "grad_norm": 13.125,
+      "learning_rate": 4.831817251403726e-06,
+      "loss": 1.5336,
+      "step": 53300
+    },
+    {
+      "epoch": 1.5510370305727665,
+      "grad_norm": 10.375,
+      "learning_rate": 4.829877969731111e-06,
+      "loss": 1.5128,
+      "step": 53320
+    },
+    {
+      "epoch": 1.5516188149053147,
+      "grad_norm": 12.8125,
+      "learning_rate": 4.827938688058496e-06,
+      "loss": 1.4405,
+      "step": 53340
+    },
+    {
+      "epoch": 1.5522005992378625,
+      "grad_norm": 11.5,
+      "learning_rate": 4.825999406385881e-06,
+      "loss": 1.581,
+      "step": 53360
+    },
+    {
+      "epoch": 1.5527823835704104,
+      "grad_norm": 12.625,
+      "learning_rate": 4.824060124713266e-06,
+      "loss": 1.5412,
+      "step": 53380
+    },
+    {
+      "epoch": 1.5533641679029584,
+      "grad_norm": 13.6875,
+      "learning_rate": 4.822120843040651e-06,
+      "loss": 1.4463,
+      "step": 53400
+    },
+    {
+      "epoch": 1.5539459522355061,
+      "grad_norm": 13.3125,
+      "learning_rate": 4.820181561368036e-06,
+      "loss": 1.5782,
+      "step": 53420
+    },
+    {
+      "epoch": 1.5545277365680543,
+      "grad_norm": 16.75,
+      "learning_rate": 4.818242279695421e-06,
+      "loss": 1.5875,
+      "step": 53440
+    },
+    {
+      "epoch": 1.555109520900602,
+      "grad_norm": 12.0625,
+      "learning_rate": 4.816302998022806e-06,
+      "loss": 1.5127,
+      "step": 53460
+    },
+    {
+      "epoch": 1.55569130523315,
+      "grad_norm": 10.5625,
+      "learning_rate": 4.814363716350191e-06,
+      "loss": 1.5307,
+      "step": 53480
+    },
+    {
+      "epoch": 1.556273089565698,
+      "grad_norm": 14.625,
+      "learning_rate": 4.812424434677576e-06,
+      "loss": 1.4338,
+      "step": 53500
+    },
+    {
+      "epoch": 1.556854873898246,
+      "grad_norm": 14.5625,
+      "learning_rate": 4.810485153004961e-06,
+      "loss": 1.5595,
+      "step": 53520
+    },
+    {
+      "epoch": 1.557436658230794,
+      "grad_norm": 12.75,
+      "learning_rate": 4.808545871332346e-06,
+      "loss": 1.6282,
+      "step": 53540
+    },
+    {
+      "epoch": 1.5580184425633417,
+      "grad_norm": 13.8125,
+      "learning_rate": 4.806606589659731e-06,
+      "loss": 1.4419,
+      "step": 53560
+    },
+    {
+      "epoch": 1.5586002268958898,
+      "grad_norm": 13.3125,
+      "learning_rate": 4.804667307987116e-06,
+      "loss": 1.4877,
+      "step": 53580
+    },
+    {
+      "epoch": 1.5591820112284376,
+      "grad_norm": 10.75,
+      "learning_rate": 4.802728026314501e-06,
+      "loss": 1.4213,
+      "step": 53600
+    },
+    {
+      "epoch": 1.5597637955609855,
+      "grad_norm": 13.5625,
+      "learning_rate": 4.800788744641886e-06,
+      "loss": 1.5356,
+      "step": 53620
+    },
+    {
+      "epoch": 1.5603455798935335,
+      "grad_norm": 13.125,
+      "learning_rate": 4.798849462969271e-06,
+      "loss": 1.4539,
+      "step": 53640
+    },
+    {
+      "epoch": 1.5609273642260812,
+      "grad_norm": 12.6875,
+      "learning_rate": 4.7969101812966565e-06,
+      "loss": 1.4881,
+      "step": 53660
+    },
+    {
+      "epoch": 1.5615091485586294,
+      "grad_norm": 13.375,
+      "learning_rate": 4.794970899624042e-06,
+      "loss": 1.4572,
+      "step": 53680
+    },
+    {
+      "epoch": 1.5620909328911772,
+      "grad_norm": 14.5625,
+      "learning_rate": 4.793031617951427e-06,
+      "loss": 1.4704,
+      "step": 53700
+    },
+    {
+      "epoch": 1.5626727172237251,
+      "grad_norm": 12.5625,
+      "learning_rate": 4.791092336278812e-06,
+      "loss": 1.5089,
+      "step": 53720
+    },
+    {
+      "epoch": 1.563254501556273,
+      "grad_norm": 14.5625,
+      "learning_rate": 4.789153054606197e-06,
+      "loss": 1.4816,
+      "step": 53740
+    },
+    {
+      "epoch": 1.563836285888821,
+      "grad_norm": 12.4375,
+      "learning_rate": 4.787213772933582e-06,
+      "loss": 1.5167,
+      "step": 53760
+    },
+    {
+      "epoch": 1.564418070221369,
+      "grad_norm": 13.1875,
+      "learning_rate": 4.785274491260966e-06,
+      "loss": 1.5706,
+      "step": 53780
+    },
+    {
+      "epoch": 1.5649998545539168,
+      "grad_norm": 13.4375,
+      "learning_rate": 4.783335209588351e-06,
+      "loss": 1.4895,
+      "step": 53800
+    },
+    {
+      "epoch": 1.5655816388864647,
+      "grad_norm": 14.0,
+      "learning_rate": 4.7813959279157364e-06,
+      "loss": 1.5177,
+      "step": 53820
+    },
+    {
+      "epoch": 1.5661634232190127,
+      "grad_norm": 13.125,
+      "learning_rate": 4.7794566462431215e-06,
+      "loss": 1.5508,
+      "step": 53840
+    },
+    {
+      "epoch": 1.5667452075515607,
+      "grad_norm": 13.875,
+      "learning_rate": 4.777517364570507e-06,
+      "loss": 1.6271,
+      "step": 53860
+    },
+    {
+      "epoch": 1.5673269918841086,
+      "grad_norm": 12.375,
+      "learning_rate": 4.775578082897892e-06,
+      "loss": 1.5082,
+      "step": 53880
+    },
+    {
+      "epoch": 1.5679087762166564,
+      "grad_norm": 12.8125,
+      "learning_rate": 4.773638801225277e-06,
+      "loss": 1.5444,
+      "step": 53900
+    },
+    {
+      "epoch": 1.5684905605492045,
+      "grad_norm": 12.4375,
+      "learning_rate": 4.771699519552662e-06,
+      "loss": 1.5745,
+      "step": 53920
+    },
+    {
+      "epoch": 1.5690723448817523,
+      "grad_norm": 13.25,
+      "learning_rate": 4.769760237880047e-06,
+      "loss": 1.5928,
+      "step": 53940
+    },
+    {
+      "epoch": 1.5696541292143003,
+      "grad_norm": 12.875,
+      "learning_rate": 4.767820956207432e-06,
+      "loss": 1.5923,
+      "step": 53960
+    },
+    {
+      "epoch": 1.5702359135468482,
+      "grad_norm": 11.8125,
+      "learning_rate": 4.765881674534817e-06,
+      "loss": 1.5671,
+      "step": 53980
+    },
+    {
+      "epoch": 1.570817697879396,
+      "grad_norm": 15.1875,
+      "learning_rate": 4.763942392862202e-06,
+      "loss": 1.5224,
+      "step": 54000
+    },
+    {
+      "epoch": 1.5713994822119441,
+      "grad_norm": 14.25,
+      "learning_rate": 4.762003111189587e-06,
+      "loss": 1.4886,
+      "step": 54020
+    },
+    {
+      "epoch": 1.5719812665444919,
+      "grad_norm": 11.8125,
+      "learning_rate": 4.7600638295169725e-06,
+      "loss": 1.4678,
+      "step": 54040
+    },
+    {
+      "epoch": 1.5725630508770398,
+      "grad_norm": 13.75,
+      "learning_rate": 4.758124547844358e-06,
+      "loss": 1.4967,
+      "step": 54060
+    },
+    {
+      "epoch": 1.5731448352095878,
+      "grad_norm": 13.375,
+      "learning_rate": 4.756185266171743e-06,
+      "loss": 1.5169,
+      "step": 54080
+    },
+    {
+      "epoch": 1.5737266195421358,
+      "grad_norm": 12.6875,
+      "learning_rate": 4.754245984499128e-06,
+      "loss": 1.4806,
+      "step": 54100
+    },
+    {
+      "epoch": 1.5743084038746837,
+      "grad_norm": 14.5,
+      "learning_rate": 4.752306702826513e-06,
+      "loss": 1.5104,
+      "step": 54120
+    },
+    {
+      "epoch": 1.5748901882072315,
+      "grad_norm": 11.6875,
+      "learning_rate": 4.750367421153898e-06,
+      "loss": 1.5267,
+      "step": 54140
+    },
+    {
+      "epoch": 1.5754719725397797,
+      "grad_norm": 10.8125,
+      "learning_rate": 4.748428139481283e-06,
+      "loss": 1.4444,
+      "step": 54160
+    },
+    {
+      "epoch": 1.5760537568723274,
+      "grad_norm": 14.75,
+      "learning_rate": 4.746488857808668e-06,
+      "loss": 1.5271,
+      "step": 54180
+    },
+    {
+      "epoch": 1.5766355412048754,
+      "grad_norm": 14.0,
+      "learning_rate": 4.744549576136053e-06,
+      "loss": 1.4962,
+      "step": 54200
+    },
+    {
+      "epoch": 1.5772173255374233,
+      "grad_norm": 12.6875,
+      "learning_rate": 4.742610294463438e-06,
+      "loss": 1.5658,
+      "step": 54220
+    },
+    {
+      "epoch": 1.577799109869971,
+      "grad_norm": 11.3125,
+      "learning_rate": 4.7406710127908235e-06,
+      "loss": 1.5937,
+      "step": 54240
+    },
+    {
+      "epoch": 1.5783808942025193,
+      "grad_norm": 13.5625,
+      "learning_rate": 4.738731731118209e-06,
+      "loss": 1.52,
+      "step": 54260
+    },
+    {
+      "epoch": 1.578962678535067,
+      "grad_norm": 15.1875,
+      "learning_rate": 4.736792449445594e-06,
+      "loss": 1.4848,
+      "step": 54280
+    },
+    {
+      "epoch": 1.579544462867615,
+      "grad_norm": 9.9375,
+      "learning_rate": 4.734853167772979e-06,
+      "loss": 1.4543,
+      "step": 54300
+    },
+    {
+      "epoch": 1.580126247200163,
+      "grad_norm": 13.625,
+      "learning_rate": 4.732913886100364e-06,
+      "loss": 1.4826,
+      "step": 54320
+    },
+    {
+      "epoch": 1.5807080315327107,
+      "grad_norm": 12.1875,
+      "learning_rate": 4.730974604427749e-06,
+      "loss": 1.5062,
+      "step": 54340
+    },
+    {
+      "epoch": 1.5812898158652589,
+      "grad_norm": 10.8125,
+      "learning_rate": 4.729035322755134e-06,
+      "loss": 1.4645,
+      "step": 54360
+    },
+    {
+      "epoch": 1.5818716001978066,
+      "grad_norm": 13.3125,
+      "learning_rate": 4.727096041082519e-06,
+      "loss": 1.5349,
+      "step": 54380
+    },
+    {
+      "epoch": 1.5824533845303546,
+      "grad_norm": 11.6875,
+      "learning_rate": 4.725156759409904e-06,
+      "loss": 1.5129,
+      "step": 54400
+    },
+    {
+      "epoch": 1.5830351688629025,
+      "grad_norm": 15.125,
+      "learning_rate": 4.723217477737289e-06,
+      "loss": 1.5458,
+      "step": 54420
+    },
+    {
+      "epoch": 1.5836169531954505,
+      "grad_norm": 11.375,
+      "learning_rate": 4.7212781960646745e-06,
+      "loss": 1.5131,
+      "step": 54440
+    },
+    {
+      "epoch": 1.5841987375279984,
+      "grad_norm": 10.9375,
+      "learning_rate": 4.7193389143920596e-06,
+      "loss": 1.6478,
+      "step": 54460
+    },
+    {
+      "epoch": 1.5847805218605462,
+      "grad_norm": 14.875,
+      "learning_rate": 4.717399632719445e-06,
+      "loss": 1.5146,
+      "step": 54480
+    },
+    {
+      "epoch": 1.5853623061930944,
+      "grad_norm": 12.0,
+      "learning_rate": 4.71546035104683e-06,
+      "loss": 1.5281,
+      "step": 54500
+    },
+    {
+      "epoch": 1.5859440905256421,
+      "grad_norm": 13.8125,
+      "learning_rate": 4.713521069374215e-06,
+      "loss": 1.575,
+      "step": 54520
+    },
+    {
+      "epoch": 1.58652587485819,
+      "grad_norm": 16.875,
+      "learning_rate": 4.7115817877016e-06,
+      "loss": 1.5869,
+      "step": 54540
+    },
+    {
+      "epoch": 1.587107659190738,
+      "grad_norm": 12.375,
+      "learning_rate": 4.709642506028985e-06,
+      "loss": 1.4953,
+      "step": 54560
+    },
+    {
+      "epoch": 1.5876894435232858,
+      "grad_norm": 10.875,
+      "learning_rate": 4.70770322435637e-06,
+      "loss": 1.5689,
+      "step": 54580
+    },
+    {
+      "epoch": 1.588271227855834,
+      "grad_norm": 17.25,
+      "learning_rate": 4.705763942683754e-06,
+      "loss": 1.5589,
+      "step": 54600
+    },
+    {
+      "epoch": 1.5888530121883817,
+      "grad_norm": 13.0,
+      "learning_rate": 4.7038246610111395e-06,
+      "loss": 1.5302,
+      "step": 54620
+    },
+    {
+      "epoch": 1.5894347965209297,
+      "grad_norm": 13.5625,
+      "learning_rate": 4.701885379338525e-06,
+      "loss": 1.3932,
+      "step": 54640
+    },
+    {
+      "epoch": 1.5900165808534776,
+      "grad_norm": 11.875,
+      "learning_rate": 4.69994609766591e-06,
+      "loss": 1.4435,
+      "step": 54660
+    },
+    {
+      "epoch": 1.5905983651860254,
+      "grad_norm": 13.125,
+      "learning_rate": 4.698006815993295e-06,
+      "loss": 1.5546,
+      "step": 54680
+    },
+    {
+      "epoch": 1.5911801495185736,
+      "grad_norm": 11.625,
+      "learning_rate": 4.69606753432068e-06,
+      "loss": 1.5133,
+      "step": 54700
+    },
+    {
+      "epoch": 1.5917619338511213,
+      "grad_norm": 14.1875,
+      "learning_rate": 4.694128252648065e-06,
+      "loss": 1.4977,
+      "step": 54720
+    },
+    {
+      "epoch": 1.5923437181836693,
+      "grad_norm": 14.6875,
+      "learning_rate": 4.69218897097545e-06,
+      "loss": 1.6043,
+      "step": 54740
+    },
+    {
+      "epoch": 1.5929255025162172,
+      "grad_norm": 15.0625,
+      "learning_rate": 4.690249689302835e-06,
+      "loss": 1.5224,
+      "step": 54760
+    },
+    {
+      "epoch": 1.5935072868487652,
+      "grad_norm": 16.5,
+      "learning_rate": 4.68831040763022e-06,
+      "loss": 1.5228,
+      "step": 54780
+    },
+    {
+      "epoch": 1.5940890711813132,
+      "grad_norm": 16.125,
+      "learning_rate": 4.686371125957605e-06,
+      "loss": 1.5182,
+      "step": 54800
+    },
+    {
+      "epoch": 1.594670855513861,
+      "grad_norm": 16.375,
+      "learning_rate": 4.6844318442849905e-06,
+      "loss": 1.4713,
+      "step": 54820
+    },
+    {
+      "epoch": 1.595252639846409,
+      "grad_norm": 13.5625,
+      "learning_rate": 4.6824925626123756e-06,
+      "loss": 1.589,
+      "step": 54840
+    },
+    {
+      "epoch": 1.5958344241789568,
+      "grad_norm": 12.8125,
+      "learning_rate": 4.680553280939761e-06,
+      "loss": 1.5047,
+      "step": 54860
+    },
+    {
+      "epoch": 1.5964162085115048,
+      "grad_norm": 15.1875,
+      "learning_rate": 4.678613999267146e-06,
+      "loss": 1.5411,
+      "step": 54880
+    },
+    {
+      "epoch": 1.5969979928440527,
+      "grad_norm": 15.4375,
+      "learning_rate": 4.676674717594531e-06,
+      "loss": 1.4919,
+      "step": 54900
+    },
+    {
+      "epoch": 1.5975797771766005,
+      "grad_norm": 16.0,
+      "learning_rate": 4.674735435921916e-06,
+      "loss": 1.6321,
+      "step": 54920
+    },
+    {
+      "epoch": 1.5981615615091487,
+      "grad_norm": 14.25,
+      "learning_rate": 4.672796154249301e-06,
+      "loss": 1.4955,
+      "step": 54940
+    },
+    {
+      "epoch": 1.5987433458416964,
+      "grad_norm": 11.4375,
+      "learning_rate": 4.670856872576686e-06,
+      "loss": 1.5507,
+      "step": 54960
+    },
+    {
+      "epoch": 1.5993251301742444,
+      "grad_norm": 9.8125,
+      "learning_rate": 4.668917590904071e-06,
+      "loss": 1.514,
+      "step": 54980
+    },
+    {
+      "epoch": 1.5999069145067923,
+      "grad_norm": 12.6875,
+      "learning_rate": 4.666978309231456e-06,
+      "loss": 1.4391,
+      "step": 55000
+    },
+    {
+      "epoch": 1.6004886988393403,
+      "grad_norm": 13.0625,
+      "learning_rate": 4.6650390275588414e-06,
+      "loss": 1.5631,
+      "step": 55020
+    },
+    {
+      "epoch": 1.6010704831718883,
+      "grad_norm": 11.625,
+      "learning_rate": 4.6630997458862265e-06,
+      "loss": 1.4999,
+      "step": 55040
+    },
+    {
+      "epoch": 1.601652267504436,
+      "grad_norm": 11.125,
+      "learning_rate": 4.661160464213612e-06,
+      "loss": 1.5002,
+      "step": 55060
+    },
+    {
+      "epoch": 1.602234051836984,
+      "grad_norm": 13.125,
+      "learning_rate": 4.659221182540997e-06,
+      "loss": 1.451,
+      "step": 55080
+    },
+    {
+      "epoch": 1.602815836169532,
+      "grad_norm": 10.5625,
+      "learning_rate": 4.657281900868382e-06,
+      "loss": 1.5922,
+      "step": 55100
+    },
+    {
+      "epoch": 1.60339762050208,
+      "grad_norm": 15.125,
+      "learning_rate": 4.655342619195767e-06,
+      "loss": 1.5512,
+      "step": 55120
+    },
+    {
+      "epoch": 1.6039794048346279,
+      "grad_norm": 13.0,
+      "learning_rate": 4.653403337523152e-06,
+      "loss": 1.5206,
+      "step": 55140
+    },
+    {
+      "epoch": 1.6045611891671756,
+      "grad_norm": 13.9375,
+      "learning_rate": 4.651464055850537e-06,
+      "loss": 1.5631,
+      "step": 55160
+    },
+    {
+      "epoch": 1.6051429734997238,
+      "grad_norm": 13.375,
+      "learning_rate": 4.649524774177922e-06,
+      "loss": 1.5608,
+      "step": 55180
+    },
+    {
+      "epoch": 1.6057247578322715,
+      "grad_norm": 13.6875,
+      "learning_rate": 4.647585492505307e-06,
+      "loss": 1.5139,
+      "step": 55200
+    },
+    {
+      "epoch": 1.6063065421648195,
+      "grad_norm": 13.9375,
+      "learning_rate": 4.645646210832692e-06,
+      "loss": 1.5417,
+      "step": 55220
+    },
+    {
+      "epoch": 1.6068883264973675,
+      "grad_norm": 15.0,
+      "learning_rate": 4.6437069291600775e-06,
+      "loss": 1.4508,
+      "step": 55240
+    },
+    {
+      "epoch": 1.6074701108299152,
+      "grad_norm": 13.125,
+      "learning_rate": 4.641767647487463e-06,
+      "loss": 1.5182,
+      "step": 55260
+    },
+    {
+      "epoch": 1.6080518951624634,
+      "grad_norm": 12.125,
+      "learning_rate": 4.639828365814848e-06,
+      "loss": 1.5126,
+      "step": 55280
+    },
+    {
+      "epoch": 1.6086336794950111,
+      "grad_norm": 14.625,
+      "learning_rate": 4.637889084142233e-06,
+      "loss": 1.5812,
+      "step": 55300
+    },
+    {
+      "epoch": 1.609215463827559,
+      "grad_norm": 14.75,
+      "learning_rate": 4.635949802469618e-06,
+      "loss": 1.5074,
+      "step": 55320
+    },
+    {
+      "epoch": 1.609797248160107,
+      "grad_norm": 12.75,
+      "learning_rate": 4.634010520797003e-06,
+      "loss": 1.4601,
+      "step": 55340
+    },
+    {
+      "epoch": 1.610379032492655,
+      "grad_norm": 12.4375,
+      "learning_rate": 4.632071239124388e-06,
+      "loss": 1.5758,
+      "step": 55360
+    },
+    {
+      "epoch": 1.610960816825203,
+      "grad_norm": 10.25,
+      "learning_rate": 4.630131957451773e-06,
+      "loss": 1.5303,
+      "step": 55380
+    },
+    {
+      "epoch": 1.6115426011577507,
+      "grad_norm": 14.3125,
+      "learning_rate": 4.6281926757791575e-06,
+      "loss": 1.506,
+      "step": 55400
+    },
+    {
+      "epoch": 1.612124385490299,
+      "grad_norm": 14.75,
+      "learning_rate": 4.6262533941065426e-06,
+      "loss": 1.4768,
+      "step": 55420
+    },
+    {
+      "epoch": 1.6127061698228466,
+      "grad_norm": 13.75,
+      "learning_rate": 4.624314112433928e-06,
+      "loss": 1.5796,
+      "step": 55440
+    },
+    {
+      "epoch": 1.6132879541553946,
+      "grad_norm": 13.4375,
+      "learning_rate": 4.622374830761313e-06,
+      "loss": 1.5388,
+      "step": 55460
+    },
+    {
+      "epoch": 1.6138697384879426,
+      "grad_norm": 14.5,
+      "learning_rate": 4.620435549088698e-06,
+      "loss": 1.5302,
+      "step": 55480
+    },
+    {
+      "epoch": 1.6144515228204903,
+      "grad_norm": 11.875,
+      "learning_rate": 4.618496267416083e-06,
+      "loss": 1.5086,
+      "step": 55500
+    },
+    {
+      "epoch": 1.6150333071530385,
+      "grad_norm": 12.125,
+      "learning_rate": 4.616556985743468e-06,
+      "loss": 1.5663,
+      "step": 55520
+    },
+    {
+      "epoch": 1.6156150914855862,
+      "grad_norm": 14.5,
+      "learning_rate": 4.614617704070853e-06,
+      "loss": 1.4878,
+      "step": 55540
+    },
+    {
+      "epoch": 1.6161968758181342,
+      "grad_norm": 15.9375,
+      "learning_rate": 4.612678422398238e-06,
+      "loss": 1.5529,
+      "step": 55560
+    },
+    {
+      "epoch": 1.6167786601506822,
+      "grad_norm": 13.125,
+      "learning_rate": 4.610739140725623e-06,
+      "loss": 1.3758,
+      "step": 55580
+    },
+    {
+      "epoch": 1.61736044448323,
+      "grad_norm": 14.8125,
+      "learning_rate": 4.6087998590530084e-06,
+      "loss": 1.4997,
+      "step": 55600
+    },
+    {
+      "epoch": 1.617942228815778,
+      "grad_norm": 13.75,
+      "learning_rate": 4.6068605773803935e-06,
+      "loss": 1.4679,
+      "step": 55620
+    },
+    {
+      "epoch": 1.6185240131483258,
+      "grad_norm": 18.875,
+      "learning_rate": 4.604921295707779e-06,
+      "loss": 1.5344,
+      "step": 55640
+    },
+    {
+      "epoch": 1.6191057974808738,
+      "grad_norm": 13.375,
+      "learning_rate": 4.602982014035164e-06,
+      "loss": 1.4957,
+      "step": 55660
+    },
+    {
+      "epoch": 1.6196875818134218,
+      "grad_norm": 14.3125,
+      "learning_rate": 4.601042732362548e-06,
+      "loss": 1.5895,
+      "step": 55680
+    },
+    {
+      "epoch": 1.6202693661459697,
+      "grad_norm": 12.0625,
+      "learning_rate": 4.599103450689933e-06,
+      "loss": 1.5243,
+      "step": 55700
+    },
+    {
+      "epoch": 1.6208511504785177,
+      "grad_norm": 13.8125,
+      "learning_rate": 4.597164169017318e-06,
+      "loss": 1.5176,
+      "step": 55720
+    },
+    {
+      "epoch": 1.6214329348110654,
+      "grad_norm": 13.9375,
+      "learning_rate": 4.595224887344703e-06,
+      "loss": 1.5591,
+      "step": 55740
+    },
+    {
+      "epoch": 1.6220147191436136,
+      "grad_norm": 12.75,
+      "learning_rate": 4.593285605672088e-06,
+      "loss": 1.5138,
+      "step": 55760
+    },
+    {
+      "epoch": 1.6225965034761614,
+      "grad_norm": 11.6875,
+      "learning_rate": 4.5913463239994735e-06,
+      "loss": 1.5448,
+      "step": 55780
+    },
+    {
+      "epoch": 1.6231782878087093,
+      "grad_norm": 13.8125,
+      "learning_rate": 4.5894070423268586e-06,
+      "loss": 1.5806,
+      "step": 55800
+    },
+    {
+      "epoch": 1.6237600721412573,
+      "grad_norm": 13.75,
+      "learning_rate": 4.587467760654244e-06,
+      "loss": 1.5211,
+      "step": 55820
+    },
+    {
+      "epoch": 1.624341856473805,
+      "grad_norm": 12.0,
+      "learning_rate": 4.585528478981629e-06,
+      "loss": 1.5163,
+      "step": 55840
+    },
+    {
+      "epoch": 1.6249236408063532,
+      "grad_norm": 12.8125,
+      "learning_rate": 4.583589197309014e-06,
+      "loss": 1.5664,
+      "step": 55860
+    },
+    {
+      "epoch": 1.625505425138901,
+      "grad_norm": 14.125,
+      "learning_rate": 4.581649915636399e-06,
+      "loss": 1.464,
+      "step": 55880
+    },
+    {
+      "epoch": 1.626087209471449,
+      "grad_norm": 12.9375,
+      "learning_rate": 4.579710633963784e-06,
+      "loss": 1.5305,
+      "step": 55900
+    },
+    {
+      "epoch": 1.6266689938039969,
+      "grad_norm": 13.4375,
+      "learning_rate": 4.577771352291169e-06,
+      "loss": 1.5297,
+      "step": 55920
+    },
+    {
+      "epoch": 1.6272507781365446,
+      "grad_norm": 15.75,
+      "learning_rate": 4.575832070618554e-06,
+      "loss": 1.4527,
+      "step": 55940
+    },
+    {
+      "epoch": 1.6278325624690928,
+      "grad_norm": 12.3125,
+      "learning_rate": 4.573892788945939e-06,
+      "loss": 1.5322,
+      "step": 55960
+    },
+    {
+      "epoch": 1.6284143468016405,
+      "grad_norm": 12.4375,
+      "learning_rate": 4.5719535072733244e-06,
+      "loss": 1.5373,
+      "step": 55980
+    },
+    {
+      "epoch": 1.6289961311341885,
+      "grad_norm": 14.25,
+      "learning_rate": 4.5700142256007095e-06,
+      "loss": 1.5091,
+      "step": 56000
+    },
+    {
+      "epoch": 1.6295779154667365,
+      "grad_norm": 13.625,
+      "learning_rate": 4.568074943928095e-06,
+      "loss": 1.5388,
+      "step": 56020
+    },
+    {
+      "epoch": 1.6301596997992844,
+      "grad_norm": 16.625,
+      "learning_rate": 4.56613566225548e-06,
+      "loss": 1.5081,
+      "step": 56040
+    },
+    {
+      "epoch": 1.6307414841318324,
+      "grad_norm": 11.625,
+      "learning_rate": 4.564196380582865e-06,
+      "loss": 1.4889,
+      "step": 56060
+    },
+    {
+      "epoch": 1.6313232684643801,
+      "grad_norm": 11.875,
+      "learning_rate": 4.56225709891025e-06,
+      "loss": 1.5055,
+      "step": 56080
+    },
+    {
+      "epoch": 1.6319050527969283,
+      "grad_norm": 13.6875,
+      "learning_rate": 4.560317817237635e-06,
+      "loss": 1.5741,
+      "step": 56100
+    },
+    {
+      "epoch": 1.632486837129476,
+      "grad_norm": 13.1875,
+      "learning_rate": 4.55837853556502e-06,
+      "loss": 1.4833,
+      "step": 56120
+    },
+    {
+      "epoch": 1.633068621462024,
+      "grad_norm": 14.75,
+      "learning_rate": 4.556439253892405e-06,
+      "loss": 1.4771,
+      "step": 56140
+    },
+    {
+      "epoch": 1.633650405794572,
+      "grad_norm": 13.0,
+      "learning_rate": 4.55449997221979e-06,
+      "loss": 1.4479,
+      "step": 56160
+    },
+    {
+      "epoch": 1.6342321901271197,
+      "grad_norm": 12.6875,
+      "learning_rate": 4.552560690547175e-06,
+      "loss": 1.5089,
+      "step": 56180
+    },
+    {
+      "epoch": 1.634813974459668,
+      "grad_norm": 13.5625,
+      "learning_rate": 4.5506214088745605e-06,
+      "loss": 1.5075,
+      "step": 56200
+    },
+    {
+      "epoch": 1.6353957587922157,
+      "grad_norm": 13.625,
+      "learning_rate": 4.548682127201946e-06,
+      "loss": 1.4791,
+      "step": 56220
+    },
+    {
+      "epoch": 1.6359775431247636,
+      "grad_norm": 15.375,
+      "learning_rate": 4.546742845529331e-06,
+      "loss": 1.5792,
+      "step": 56240
+    },
+    {
+      "epoch": 1.6365593274573116,
+      "grad_norm": 14.375,
+      "learning_rate": 4.544803563856716e-06,
+      "loss": 1.5874,
+      "step": 56260
+    },
+    {
+      "epoch": 1.6371411117898595,
+      "grad_norm": 11.6875,
+      "learning_rate": 4.542864282184101e-06,
+      "loss": 1.5322,
+      "step": 56280
+    },
+    {
+      "epoch": 1.6377228961224075,
+      "grad_norm": 11.6875,
+      "learning_rate": 4.540925000511486e-06,
+      "loss": 1.5733,
+      "step": 56300
+    },
+    {
+      "epoch": 1.6383046804549553,
+      "grad_norm": 12.8125,
+      "learning_rate": 4.538985718838871e-06,
+      "loss": 1.4755,
+      "step": 56320
+    },
+    {
+      "epoch": 1.6388864647875034,
+      "grad_norm": 14.125,
+      "learning_rate": 4.537046437166256e-06,
+      "loss": 1.4399,
+      "step": 56340
+    },
+    {
+      "epoch": 1.6394682491200512,
+      "grad_norm": 16.125,
+      "learning_rate": 4.535107155493641e-06,
+      "loss": 1.5363,
+      "step": 56360
+    },
+    {
+      "epoch": 1.6400500334525991,
+      "grad_norm": 13.0,
+      "learning_rate": 4.533167873821026e-06,
+      "loss": 1.539,
+      "step": 56380
+    },
+    {
+      "epoch": 1.640631817785147,
+      "grad_norm": 11.8125,
+      "learning_rate": 4.5312285921484115e-06,
+      "loss": 1.4282,
+      "step": 56400
+    },
+    {
+      "epoch": 1.6412136021176948,
+      "grad_norm": 12.6875,
+      "learning_rate": 4.529289310475797e-06,
+      "loss": 1.5218,
+      "step": 56420
+    },
+    {
+      "epoch": 1.641795386450243,
+      "grad_norm": 13.3125,
+      "learning_rate": 4.527350028803182e-06,
+      "loss": 1.5561,
+      "step": 56440
+    },
+    {
+      "epoch": 1.6423771707827908,
+      "grad_norm": 10.875,
+      "learning_rate": 4.525410747130567e-06,
+      "loss": 1.4688,
+      "step": 56460
+    },
+    {
+      "epoch": 1.6429589551153387,
+      "grad_norm": 16.375,
+      "learning_rate": 4.523471465457952e-06,
+      "loss": 1.5251,
+      "step": 56480
+    },
+    {
+      "epoch": 1.6435407394478867,
+      "grad_norm": 12.3125,
+      "learning_rate": 4.521532183785336e-06,
+      "loss": 1.5564,
+      "step": 56500
+    },
+    {
+      "epoch": 1.6441225237804344,
+      "grad_norm": 11.3125,
+      "learning_rate": 4.519592902112721e-06,
+      "loss": 1.5388,
+      "step": 56520
+    },
+    {
+      "epoch": 1.6447043081129826,
+      "grad_norm": 9.9375,
+      "learning_rate": 4.517653620440106e-06,
+      "loss": 1.5414,
+      "step": 56540
+    },
+    {
+      "epoch": 1.6452860924455304,
+      "grad_norm": 13.1875,
+      "learning_rate": 4.515714338767491e-06,
+      "loss": 1.5193,
+      "step": 56560
+    },
+    {
+      "epoch": 1.6458678767780783,
+      "grad_norm": 14.375,
+      "learning_rate": 4.5137750570948765e-06,
+      "loss": 1.5994,
+      "step": 56580
+    },
+    {
+      "epoch": 1.6464496611106263,
+      "grad_norm": 12.625,
+      "learning_rate": 4.511835775422262e-06,
+      "loss": 1.5887,
+      "step": 56600
+    },
+    {
+      "epoch": 1.6470314454431743,
+      "grad_norm": 16.75,
+      "learning_rate": 4.509896493749647e-06,
+      "loss": 1.5113,
+      "step": 56620
+    },
+    {
+      "epoch": 1.6476132297757222,
+      "grad_norm": 13.3125,
+      "learning_rate": 4.507957212077032e-06,
+      "loss": 1.6203,
+      "step": 56640
+    },
+    {
+      "epoch": 1.64819501410827,
+      "grad_norm": 11.6875,
+      "learning_rate": 4.506017930404417e-06,
+      "loss": 1.4855,
+      "step": 56660
+    },
+    {
+      "epoch": 1.6487767984408181,
+      "grad_norm": 14.375,
+      "learning_rate": 4.504078648731802e-06,
+      "loss": 1.4891,
+      "step": 56680
+    },
+    {
+      "epoch": 1.6493585827733659,
+      "grad_norm": 12.5625,
+      "learning_rate": 4.502139367059187e-06,
+      "loss": 1.5076,
+      "step": 56700
+    },
+    {
+      "epoch": 1.6499403671059139,
+      "grad_norm": 11.75,
+      "learning_rate": 4.500200085386572e-06,
+      "loss": 1.5899,
+      "step": 56720
+    },
+    {
+      "epoch": 1.6505221514384618,
+      "grad_norm": 13.25,
+      "learning_rate": 4.498260803713957e-06,
+      "loss": 1.5738,
+      "step": 56740
+    },
+    {
+      "epoch": 1.6511039357710096,
+      "grad_norm": 11.9375,
+      "learning_rate": 4.496321522041342e-06,
+      "loss": 1.5436,
+      "step": 56760
+    },
+    {
+      "epoch": 1.6516857201035577,
+      "grad_norm": 14.0,
+      "learning_rate": 4.4943822403687275e-06,
+      "loss": 1.5207,
+      "step": 56780
+    },
+    {
+      "epoch": 1.6522675044361055,
+      "grad_norm": 11.5,
+      "learning_rate": 4.492442958696113e-06,
+      "loss": 1.5748,
+      "step": 56800
+    },
+    {
+      "epoch": 1.6528492887686534,
+      "grad_norm": 15.5625,
+      "learning_rate": 4.490503677023498e-06,
+      "loss": 1.5621,
+      "step": 56820
+    },
+    {
+      "epoch": 1.6534310731012014,
+      "grad_norm": 12.375,
+      "learning_rate": 4.488564395350883e-06,
+      "loss": 1.5014,
+      "step": 56840
+    },
+    {
+      "epoch": 1.6540128574337492,
+      "grad_norm": 12.75,
+      "learning_rate": 4.486625113678268e-06,
+      "loss": 1.5741,
+      "step": 56860
+    },
+    {
+      "epoch": 1.6545946417662973,
+      "grad_norm": 14.1875,
+      "learning_rate": 4.484685832005653e-06,
+      "loss": 1.6574,
+      "step": 56880
+    },
+    {
+      "epoch": 1.655176426098845,
+      "grad_norm": 12.875,
+      "learning_rate": 4.482746550333038e-06,
+      "loss": 1.4918,
+      "step": 56900
+    },
+    {
+      "epoch": 1.655758210431393,
+      "grad_norm": 11.875,
+      "learning_rate": 4.480807268660423e-06,
+      "loss": 1.4969,
+      "step": 56920
+    },
+    {
+      "epoch": 1.656339994763941,
+      "grad_norm": 14.5,
+      "learning_rate": 4.478867986987808e-06,
+      "loss": 1.5432,
+      "step": 56940
+    },
+    {
+      "epoch": 1.656921779096489,
+      "grad_norm": 13.6875,
+      "learning_rate": 4.476928705315193e-06,
+      "loss": 1.5645,
+      "step": 56960
+    },
+    {
+      "epoch": 1.657503563429037,
+      "grad_norm": 15.25,
+      "learning_rate": 4.4749894236425785e-06,
+      "loss": 1.4698,
+      "step": 56980
+    },
+    {
+      "epoch": 1.6580853477615847,
+      "grad_norm": 13.5625,
+      "learning_rate": 4.4730501419699636e-06,
+      "loss": 1.5622,
+      "step": 57000
+    },
+    {
+      "epoch": 1.6586671320941329,
+      "grad_norm": 12.5625,
+      "learning_rate": 4.471110860297349e-06,
+      "loss": 1.5406,
+      "step": 57020
+    },
+    {
+      "epoch": 1.6592489164266806,
+      "grad_norm": 15.25,
+      "learning_rate": 4.469171578624734e-06,
+      "loss": 1.5354,
+      "step": 57040
+    },
+    {
+      "epoch": 1.6598307007592286,
+      "grad_norm": 13.1875,
+      "learning_rate": 4.467232296952119e-06,
+      "loss": 1.4825,
+      "step": 57060
+    },
+    {
+      "epoch": 1.6604124850917765,
+      "grad_norm": 13.6875,
+      "learning_rate": 4.465293015279504e-06,
+      "loss": 1.4245,
+      "step": 57080
+    },
+    {
+      "epoch": 1.6609942694243243,
+      "grad_norm": 13.625,
+      "learning_rate": 4.463353733606889e-06,
+      "loss": 1.5259,
+      "step": 57100
+    },
+    {
+      "epoch": 1.6615760537568725,
+      "grad_norm": 12.375,
+      "learning_rate": 4.461414451934274e-06,
+      "loss": 1.562,
+      "step": 57120
+    },
+    {
+      "epoch": 1.6621578380894202,
+      "grad_norm": 9.0625,
+      "learning_rate": 4.459475170261659e-06,
+      "loss": 1.552,
+      "step": 57140
+    },
+    {
+      "epoch": 1.6627396224219682,
+      "grad_norm": 15.4375,
+      "learning_rate": 4.457535888589044e-06,
+      "loss": 1.4998,
+      "step": 57160
+    },
+    {
+      "epoch": 1.6633214067545161,
+      "grad_norm": 15.0625,
+      "learning_rate": 4.4555966069164294e-06,
+      "loss": 1.569,
+      "step": 57180
+    },
+    {
+      "epoch": 1.6639031910870639,
+      "grad_norm": 12.75,
+      "learning_rate": 4.4536573252438145e-06,
+      "loss": 1.4864,
+      "step": 57200
+    },
+    {
+      "epoch": 1.664484975419612,
+      "grad_norm": 13.1875,
+      "learning_rate": 4.4517180435712e-06,
+      "loss": 1.482,
+      "step": 57220
+    },
+    {
+      "epoch": 1.6650667597521598,
+      "grad_norm": 13.0625,
+      "learning_rate": 4.449778761898585e-06,
+      "loss": 1.5329,
+      "step": 57240
+    },
+    {
+      "epoch": 1.6656485440847077,
+      "grad_norm": 17.0,
+      "learning_rate": 4.44783948022597e-06,
+      "loss": 1.497,
+      "step": 57260
+    },
+    {
+      "epoch": 1.6662303284172557,
+      "grad_norm": 16.5,
+      "learning_rate": 4.445900198553355e-06,
+      "loss": 1.5104,
+      "step": 57280
+    },
+    {
+      "epoch": 1.6668121127498037,
+      "grad_norm": 13.0,
+      "learning_rate": 4.44396091688074e-06,
+      "loss": 1.5361,
+      "step": 57300
+    },
+    {
+      "epoch": 1.6673938970823516,
+      "grad_norm": 14.5,
+      "learning_rate": 4.442021635208124e-06,
+      "loss": 1.4874,
+      "step": 57320
+    },
+    {
+      "epoch": 1.6679756814148994,
+      "grad_norm": 13.625,
+      "learning_rate": 4.440082353535509e-06,
+      "loss": 1.4985,
+      "step": 57340
+    },
+    {
+      "epoch": 1.6685574657474476,
+      "grad_norm": 14.1875,
+      "learning_rate": 4.4381430718628945e-06,
+      "loss": 1.5271,
+      "step": 57360
+    },
+    {
+      "epoch": 1.6691392500799953,
+      "grad_norm": 11.8125,
+      "learning_rate": 4.4362037901902796e-06,
+      "loss": 1.5104,
+      "step": 57380
+    },
+    {
+      "epoch": 1.6697210344125433,
+      "grad_norm": 12.125,
+      "learning_rate": 4.434264508517665e-06,
+      "loss": 1.4442,
+      "step": 57400
+    },
+    {
+      "epoch": 1.6703028187450912,
+      "grad_norm": 14.6875,
+      "learning_rate": 4.43232522684505e-06,
+      "loss": 1.5881,
+      "step": 57420
+    },
+    {
+      "epoch": 1.670884603077639,
+      "grad_norm": 12.0,
+      "learning_rate": 4.430385945172435e-06,
+      "loss": 1.4902,
+      "step": 57440
+    },
+    {
+      "epoch": 1.6714663874101872,
+      "grad_norm": 12.4375,
+      "learning_rate": 4.42844666349982e-06,
+      "loss": 1.5443,
+      "step": 57460
+    },
+    {
+      "epoch": 1.672048171742735,
+      "grad_norm": 10.9375,
+      "learning_rate": 4.426507381827205e-06,
+      "loss": 1.5009,
+      "step": 57480
+    },
+    {
+      "epoch": 1.6726299560752829,
+      "grad_norm": 11.625,
+      "learning_rate": 4.42456810015459e-06,
+      "loss": 1.4822,
+      "step": 57500
+    },
+    {
+      "epoch": 1.6732117404078308,
+      "grad_norm": 12.25,
+      "learning_rate": 4.422628818481975e-06,
+      "loss": 1.5753,
+      "step": 57520
+    },
+    {
+      "epoch": 1.6737935247403788,
+      "grad_norm": 13.9375,
+      "learning_rate": 4.42068953680936e-06,
+      "loss": 1.4683,
+      "step": 57540
+    },
+    {
+      "epoch": 1.6743753090729268,
+      "grad_norm": 11.4375,
+      "learning_rate": 4.4187502551367454e-06,
+      "loss": 1.5026,
+      "step": 57560
+    },
+    {
+      "epoch": 1.6749570934054745,
+      "grad_norm": 16.25,
+      "learning_rate": 4.4168109734641305e-06,
+      "loss": 1.5711,
+      "step": 57580
+    },
+    {
+      "epoch": 1.6755388777380227,
+      "grad_norm": 12.75,
+      "learning_rate": 4.414871691791516e-06,
+      "loss": 1.5626,
+      "step": 57600
+    },
+    {
+      "epoch": 1.6761206620705704,
+      "grad_norm": 14.1875,
+      "learning_rate": 4.412932410118901e-06,
+      "loss": 1.5723,
+      "step": 57620
+    },
+    {
+      "epoch": 1.6767024464031184,
+      "grad_norm": 13.0,
+      "learning_rate": 4.410993128446286e-06,
+      "loss": 1.5563,
+      "step": 57640
+    },
+    {
+      "epoch": 1.6772842307356663,
+      "grad_norm": 15.875,
+      "learning_rate": 4.409053846773671e-06,
+      "loss": 1.5386,
+      "step": 57660
+    },
+    {
+      "epoch": 1.677866015068214,
+      "grad_norm": 15.9375,
+      "learning_rate": 4.407114565101056e-06,
+      "loss": 1.494,
+      "step": 57680
+    },
+    {
+      "epoch": 1.6784477994007623,
+      "grad_norm": 11.3125,
+      "learning_rate": 4.405175283428441e-06,
+      "loss": 1.5629,
+      "step": 57700
+    },
+    {
+      "epoch": 1.67902958373331,
+      "grad_norm": 12.5625,
+      "learning_rate": 4.403236001755825e-06,
+      "loss": 1.4474,
+      "step": 57720
+    },
+    {
+      "epoch": 1.679611368065858,
+      "grad_norm": 11.75,
+      "learning_rate": 4.4012967200832105e-06,
+      "loss": 1.6147,
+      "step": 57740
+    },
+    {
+      "epoch": 1.680193152398406,
+      "grad_norm": 14.4375,
+      "learning_rate": 4.3993574384105956e-06,
+      "loss": 1.4958,
+      "step": 57760
+    },
+    {
+      "epoch": 1.6807749367309537,
+      "grad_norm": 14.625,
+      "learning_rate": 4.397418156737981e-06,
+      "loss": 1.5962,
+      "step": 57780
+    },
+    {
+      "epoch": 1.6813567210635019,
+      "grad_norm": 15.4375,
+      "learning_rate": 4.395478875065366e-06,
+      "loss": 1.5277,
+      "step": 57800
+    },
+    {
+      "epoch": 1.6819385053960496,
+      "grad_norm": 13.5,
+      "learning_rate": 4.393539593392751e-06,
+      "loss": 1.5227,
+      "step": 57820
+    },
+    {
+      "epoch": 1.6825202897285976,
+      "grad_norm": 14.4375,
+      "learning_rate": 4.391600311720136e-06,
+      "loss": 1.5584,
+      "step": 57840
+    },
+    {
+      "epoch": 1.6831020740611455,
+      "grad_norm": 12.875,
+      "learning_rate": 4.389661030047521e-06,
+      "loss": 1.6491,
+      "step": 57860
+    },
+    {
+      "epoch": 1.6836838583936935,
+      "grad_norm": 12.4375,
+      "learning_rate": 4.387721748374906e-06,
+      "loss": 1.5882,
+      "step": 57880
+    },
+    {
+      "epoch": 1.6842656427262415,
+      "grad_norm": 13.3125,
+      "learning_rate": 4.385782466702291e-06,
+      "loss": 1.567,
+      "step": 57900
+    },
+    {
+      "epoch": 1.6848474270587892,
+      "grad_norm": 13.0625,
+      "learning_rate": 4.383843185029676e-06,
+      "loss": 1.5049,
+      "step": 57920
+    },
+    {
+      "epoch": 1.6854292113913374,
+      "grad_norm": 13.375,
+      "learning_rate": 4.3819039033570614e-06,
+      "loss": 1.5021,
+      "step": 57940
+    },
+    {
+      "epoch": 1.6860109957238851,
+      "grad_norm": 12.9375,
+      "learning_rate": 4.3799646216844465e-06,
+      "loss": 1.6091,
+      "step": 57960
+    },
+    {
+      "epoch": 1.686592780056433,
+      "grad_norm": 12.0625,
+      "learning_rate": 4.378025340011832e-06,
+      "loss": 1.5639,
+      "step": 57980
+    },
+    {
+      "epoch": 1.687174564388981,
+      "grad_norm": 14.875,
+      "learning_rate": 4.376086058339217e-06,
+      "loss": 1.5785,
+      "step": 58000
+    },
+    {
+      "epoch": 1.6877563487215288,
+      "grad_norm": 14.375,
+      "learning_rate": 4.374146776666602e-06,
+      "loss": 1.5313,
+      "step": 58020
+    },
+    {
+      "epoch": 1.688338133054077,
+      "grad_norm": 13.5,
+      "learning_rate": 4.372207494993987e-06,
+      "loss": 1.5673,
+      "step": 58040
+    },
+    {
+      "epoch": 1.6889199173866247,
+      "grad_norm": 11.9375,
+      "learning_rate": 4.370268213321372e-06,
+      "loss": 1.5271,
+      "step": 58060
+    },
+    {
+      "epoch": 1.6895017017191727,
+      "grad_norm": 12.375,
+      "learning_rate": 4.368328931648757e-06,
+      "loss": 1.5138,
+      "step": 58080
+    },
+    {
+      "epoch": 1.6900834860517207,
+      "grad_norm": 14.5625,
+      "learning_rate": 4.366389649976142e-06,
+      "loss": 1.5513,
+      "step": 58100
+    },
+    {
+      "epoch": 1.6906652703842684,
+      "grad_norm": 12.625,
+      "learning_rate": 4.364450368303527e-06,
+      "loss": 1.4659,
+      "step": 58120
+    },
+    {
+      "epoch": 1.6912470547168166,
+      "grad_norm": 17.125,
+      "learning_rate": 4.362511086630912e-06,
+      "loss": 1.5365,
+      "step": 58140
+    },
+    {
+      "epoch": 1.6918288390493643,
+      "grad_norm": 14.5625,
+      "learning_rate": 4.3605718049582975e-06,
+      "loss": 1.5765,
+      "step": 58160
+    },
+    {
+      "epoch": 1.6924106233819123,
+      "grad_norm": 14.8125,
+      "learning_rate": 4.358632523285683e-06,
+      "loss": 1.5635,
+      "step": 58180
+    },
+    {
+      "epoch": 1.6929924077144602,
+      "grad_norm": 16.25,
+      "learning_rate": 4.356693241613068e-06,
+      "loss": 1.4967,
+      "step": 58200
+    },
+    {
+      "epoch": 1.6935741920470082,
+      "grad_norm": 13.0,
+      "learning_rate": 4.354753959940453e-06,
+      "loss": 1.5344,
+      "step": 58220
+    },
+    {
+      "epoch": 1.6941559763795562,
+      "grad_norm": 10.25,
+      "learning_rate": 4.352814678267838e-06,
+      "loss": 1.5465,
+      "step": 58240
+    },
+    {
+      "epoch": 1.694737760712104,
+      "grad_norm": 12.0625,
+      "learning_rate": 4.350875396595223e-06,
+      "loss": 1.5161,
+      "step": 58260
+    },
+    {
+      "epoch": 1.695319545044652,
+      "grad_norm": 12.5625,
+      "learning_rate": 4.348936114922608e-06,
+      "loss": 1.5355,
+      "step": 58280
+    },
+    {
+      "epoch": 1.6959013293771998,
+      "grad_norm": 13.8125,
+      "learning_rate": 4.346996833249993e-06,
+      "loss": 1.4878,
+      "step": 58300
+    },
+    {
+      "epoch": 1.6964831137097478,
+      "grad_norm": 11.9375,
+      "learning_rate": 4.345057551577378e-06,
+      "loss": 1.5294,
+      "step": 58320
+    },
+    {
+      "epoch": 1.6970648980422958,
+      "grad_norm": 16.125,
+      "learning_rate": 4.343118269904763e-06,
+      "loss": 1.4918,
+      "step": 58340
+    },
+    {
+      "epoch": 1.6976466823748435,
+      "grad_norm": 14.6875,
+      "learning_rate": 4.3411789882321485e-06,
+      "loss": 1.4957,
+      "step": 58360
+    },
+    {
+      "epoch": 1.6982284667073917,
+      "grad_norm": 14.875,
+      "learning_rate": 4.339239706559534e-06,
+      "loss": 1.5475,
+      "step": 58380
+    },
+    {
+      "epoch": 1.6988102510399394,
+      "grad_norm": 14.75,
+      "learning_rate": 4.337300424886919e-06,
+      "loss": 1.4932,
+      "step": 58400
+    },
+    {
+      "epoch": 1.6993920353724874,
+      "grad_norm": 12.8125,
+      "learning_rate": 4.335361143214303e-06,
+      "loss": 1.5463,
+      "step": 58420
+    },
+    {
+      "epoch": 1.6999738197050354,
+      "grad_norm": 11.6875,
+      "learning_rate": 4.333421861541688e-06,
+      "loss": 1.5259,
+      "step": 58440
+    },
+    {
+      "epoch": 1.700555604037583,
+      "grad_norm": 10.8125,
+      "learning_rate": 4.331482579869073e-06,
+      "loss": 1.5044,
+      "step": 58460
+    },
+    {
+      "epoch": 1.7011373883701313,
+      "grad_norm": 17.875,
+      "learning_rate": 4.329543298196458e-06,
+      "loss": 1.5168,
+      "step": 58480
+    },
+    {
+      "epoch": 1.701719172702679,
+      "grad_norm": 11.5625,
+      "learning_rate": 4.327604016523843e-06,
+      "loss": 1.6243,
+      "step": 58500
+    },
+    {
+      "epoch": 1.702300957035227,
+      "grad_norm": 13.375,
+      "learning_rate": 4.3256647348512284e-06,
+      "loss": 1.4814,
+      "step": 58520
+    },
+    {
+      "epoch": 1.702882741367775,
+      "grad_norm": 13.125,
+      "learning_rate": 4.3237254531786135e-06,
+      "loss": 1.4382,
+      "step": 58540
+    },
+    {
+      "epoch": 1.703464525700323,
+      "grad_norm": 18.25,
+      "learning_rate": 4.321786171505999e-06,
+      "loss": 1.5173,
+      "step": 58560
+    },
+    {
+      "epoch": 1.7040463100328709,
+      "grad_norm": 16.0,
+      "learning_rate": 4.319846889833384e-06,
+      "loss": 1.54,
+      "step": 58580
+    },
+    {
+      "epoch": 1.7046280943654186,
+      "grad_norm": 12.5625,
+      "learning_rate": 4.317907608160769e-06,
+      "loss": 1.4458,
+      "step": 58600
+    },
+    {
+      "epoch": 1.7052098786979668,
+      "grad_norm": 11.8125,
+      "learning_rate": 4.315968326488154e-06,
+      "loss": 1.5298,
+      "step": 58620
+    },
+    {
+      "epoch": 1.7057916630305145,
+      "grad_norm": 13.625,
+      "learning_rate": 4.314029044815539e-06,
+      "loss": 1.4869,
+      "step": 58640
+    },
+    {
+      "epoch": 1.7063734473630625,
+      "grad_norm": 13.0625,
+      "learning_rate": 4.312089763142924e-06,
+      "loss": 1.5136,
+      "step": 58660
+    },
+    {
+      "epoch": 1.7069552316956105,
+      "grad_norm": 17.25,
+      "learning_rate": 4.310150481470309e-06,
+      "loss": 1.5782,
+      "step": 58680
+    },
+    {
+      "epoch": 1.7075370160281582,
+      "grad_norm": 12.25,
+      "learning_rate": 4.308211199797694e-06,
+      "loss": 1.5273,
+      "step": 58700
+    },
+    {
+      "epoch": 1.7081188003607064,
+      "grad_norm": 12.4375,
+      "learning_rate": 4.306271918125079e-06,
+      "loss": 1.5192,
+      "step": 58720
+    },
+    {
+      "epoch": 1.7087005846932541,
+      "grad_norm": 13.6875,
+      "learning_rate": 4.3043326364524645e-06,
+      "loss": 1.4448,
+      "step": 58740
+    },
+    {
+      "epoch": 1.709282369025802,
+      "grad_norm": 14.75,
+      "learning_rate": 4.30239335477985e-06,
+      "loss": 1.6084,
+      "step": 58760
+    },
+    {
+      "epoch": 1.70986415335835,
+      "grad_norm": 12.9375,
+      "learning_rate": 4.300454073107235e-06,
+      "loss": 1.497,
+      "step": 58780
+    },
+    {
+      "epoch": 1.710445937690898,
+      "grad_norm": 13.0625,
+      "learning_rate": 4.29851479143462e-06,
+      "loss": 1.5252,
+      "step": 58800
+    },
+    {
+      "epoch": 1.711027722023446,
+      "grad_norm": 12.8125,
+      "learning_rate": 4.296575509762005e-06,
+      "loss": 1.4394,
+      "step": 58820
+    },
+    {
+      "epoch": 1.7116095063559937,
+      "grad_norm": 14.0625,
+      "learning_rate": 4.29463622808939e-06,
+      "loss": 1.5077,
+      "step": 58840
+    },
+    {
+      "epoch": 1.712191290688542,
+      "grad_norm": 13.375,
+      "learning_rate": 4.292696946416775e-06,
+      "loss": 1.5944,
+      "step": 58860
+    },
+    {
+      "epoch": 1.7127730750210897,
+      "grad_norm": 12.4375,
+      "learning_rate": 4.29075766474416e-06,
+      "loss": 1.5142,
+      "step": 58880
+    },
+    {
+      "epoch": 1.7133548593536376,
+      "grad_norm": 14.0,
+      "learning_rate": 4.288818383071545e-06,
+      "loss": 1.5766,
+      "step": 58900
+    },
+    {
+      "epoch": 1.7139366436861856,
+      "grad_norm": 13.625,
+      "learning_rate": 4.28687910139893e-06,
+      "loss": 1.4333,
+      "step": 58920
+    },
+    {
+      "epoch": 1.7145184280187333,
+      "grad_norm": 14.125,
+      "learning_rate": 4.2849398197263155e-06,
+      "loss": 1.553,
+      "step": 58940
+    },
+    {
+      "epoch": 1.7151002123512815,
+      "grad_norm": 11.625,
+      "learning_rate": 4.2830005380537006e-06,
+      "loss": 1.5117,
+      "step": 58960
+    },
+    {
+      "epoch": 1.7156819966838293,
+      "grad_norm": 13.125,
+      "learning_rate": 4.281061256381086e-06,
+      "loss": 1.4692,
+      "step": 58980
+    },
+    {
+      "epoch": 1.7162637810163772,
+      "grad_norm": 13.3125,
+      "learning_rate": 4.279121974708471e-06,
+      "loss": 1.5303,
+      "step": 59000
+    },
+    {
+      "epoch": 1.7168455653489252,
+      "grad_norm": 13.5,
+      "learning_rate": 4.277182693035856e-06,
+      "loss": 1.503,
+      "step": 59020
+    },
+    {
+      "epoch": 1.717427349681473,
+      "grad_norm": 11.5625,
+      "learning_rate": 4.275243411363241e-06,
+      "loss": 1.5862,
+      "step": 59040
+    },
+    {
+      "epoch": 1.718009134014021,
+      "grad_norm": 12.4375,
+      "learning_rate": 4.273304129690626e-06,
+      "loss": 1.4501,
+      "step": 59060
+    },
+    {
+      "epoch": 1.7185909183465689,
+      "grad_norm": 13.25,
+      "learning_rate": 4.271364848018011e-06,
+      "loss": 1.4513,
+      "step": 59080
+    },
+    {
+      "epoch": 1.7191727026791168,
+      "grad_norm": 12.0,
+      "learning_rate": 4.269425566345396e-06,
+      "loss": 1.5399,
+      "step": 59100
+    },
+    {
+      "epoch": 1.7197544870116648,
+      "grad_norm": 14.125,
+      "learning_rate": 4.267486284672781e-06,
+      "loss": 1.395,
+      "step": 59120
+    },
+    {
+      "epoch": 1.7203362713442127,
+      "grad_norm": 15.3125,
+      "learning_rate": 4.2655470030001664e-06,
+      "loss": 1.5789,
+      "step": 59140
+    },
+    {
+      "epoch": 1.7209180556767607,
+      "grad_norm": 11.9375,
+      "learning_rate": 4.2636077213275515e-06,
+      "loss": 1.4762,
+      "step": 59160
+    },
+    {
+      "epoch": 1.7214998400093084,
+      "grad_norm": 10.375,
+      "learning_rate": 4.261668439654937e-06,
+      "loss": 1.5355,
+      "step": 59180
+    },
+    {
+      "epoch": 1.7220816243418566,
+      "grad_norm": 14.0,
+      "learning_rate": 4.259729157982322e-06,
+      "loss": 1.4571,
+      "step": 59200
+    },
+    {
+      "epoch": 1.7226634086744044,
+      "grad_norm": 12.625,
+      "learning_rate": 4.257789876309707e-06,
+      "loss": 1.4757,
+      "step": 59220
+    },
+    {
+      "epoch": 1.7232451930069523,
+      "grad_norm": 11.875,
+      "learning_rate": 4.255850594637091e-06,
+      "loss": 1.5687,
+      "step": 59240
+    },
+    {
+      "epoch": 1.7238269773395003,
+      "grad_norm": 13.625,
+      "learning_rate": 4.253911312964476e-06,
+      "loss": 1.5222,
+      "step": 59260
+    },
+    {
+      "epoch": 1.724408761672048,
+      "grad_norm": 12.9375,
+      "learning_rate": 4.251972031291861e-06,
+      "loss": 1.5058,
+      "step": 59280
+    },
+    {
+      "epoch": 1.7249905460045962,
+      "grad_norm": 13.125,
+      "learning_rate": 4.250032749619246e-06,
+      "loss": 1.4596,
+      "step": 59300
+    },
+    {
+      "epoch": 1.725572330337144,
+      "grad_norm": 14.5625,
+      "learning_rate": 4.2480934679466315e-06,
+      "loss": 1.5361,
+      "step": 59320
+    },
+    {
+      "epoch": 1.726154114669692,
+      "grad_norm": 15.0625,
+      "learning_rate": 4.2461541862740166e-06,
+      "loss": 1.4849,
+      "step": 59340
+    },
+    {
+      "epoch": 1.72673589900224,
+      "grad_norm": 13.3125,
+      "learning_rate": 4.244214904601402e-06,
+      "loss": 1.5622,
+      "step": 59360
+    },
+    {
+      "epoch": 1.7273176833347876,
+      "grad_norm": 13.25,
+      "learning_rate": 4.242275622928787e-06,
+      "loss": 1.5752,
+      "step": 59380
+    },
+    {
+      "epoch": 1.7278994676673358,
+      "grad_norm": 14.0625,
+      "learning_rate": 4.240336341256172e-06,
+      "loss": 1.5047,
+      "step": 59400
+    },
+    {
+      "epoch": 1.7284812519998836,
+      "grad_norm": 10.1875,
+      "learning_rate": 4.238397059583557e-06,
+      "loss": 1.4494,
+      "step": 59420
+    },
+    {
+      "epoch": 1.7290630363324315,
+      "grad_norm": 11.625,
+      "learning_rate": 4.236457777910942e-06,
+      "loss": 1.3919,
+      "step": 59440
+    },
+    {
+      "epoch": 1.7296448206649795,
+      "grad_norm": 11.125,
+      "learning_rate": 4.234518496238327e-06,
+      "loss": 1.4939,
+      "step": 59460
+    },
+    {
+      "epoch": 1.7302266049975275,
+      "grad_norm": 15.125,
+      "learning_rate": 4.232579214565712e-06,
+      "loss": 1.5392,
+      "step": 59480
+    },
+    {
+      "epoch": 1.7308083893300754,
+      "grad_norm": 17.25,
+      "learning_rate": 4.230639932893097e-06,
+      "loss": 1.5289,
+      "step": 59500
+    },
+    {
+      "epoch": 1.7313901736626232,
+      "grad_norm": 13.0,
+      "learning_rate": 4.2287006512204825e-06,
+      "loss": 1.5658,
+      "step": 59520
+    },
+    {
+      "epoch": 1.7319719579951713,
+      "grad_norm": 7.28125,
+      "learning_rate": 4.2267613695478675e-06,
+      "loss": 1.4307,
+      "step": 59540
+    },
+    {
+      "epoch": 1.732553742327719,
+      "grad_norm": 13.3125,
+      "learning_rate": 4.224822087875253e-06,
+      "loss": 1.5388,
+      "step": 59560
+    },
+    {
+      "epoch": 1.733135526660267,
+      "grad_norm": 15.5,
+      "learning_rate": 4.222882806202638e-06,
+      "loss": 1.5646,
+      "step": 59580
+    },
+    {
+      "epoch": 1.733717310992815,
+      "grad_norm": 14.75,
+      "learning_rate": 4.220943524530023e-06,
+      "loss": 1.5158,
+      "step": 59600
+    },
+    {
+      "epoch": 1.7342990953253627,
+      "grad_norm": 12.9375,
+      "learning_rate": 4.219004242857408e-06,
+      "loss": 1.5818,
+      "step": 59620
+    },
+    {
+      "epoch": 1.734880879657911,
+      "grad_norm": 14.875,
+      "learning_rate": 4.217064961184793e-06,
+      "loss": 1.4712,
+      "step": 59640
+    },
+    {
+      "epoch": 1.7354626639904587,
+      "grad_norm": 10.8125,
+      "learning_rate": 4.215125679512178e-06,
+      "loss": 1.5987,
+      "step": 59660
+    },
+    {
+      "epoch": 1.7360444483230066,
+      "grad_norm": 13.8125,
+      "learning_rate": 4.213186397839563e-06,
+      "loss": 1.5238,
+      "step": 59680
+    },
+    {
+      "epoch": 1.7366262326555546,
+      "grad_norm": 11.5625,
+      "learning_rate": 4.211247116166948e-06,
+      "loss": 1.4808,
+      "step": 59700
+    },
+    {
+      "epoch": 1.7372080169881026,
+      "grad_norm": 13.0,
+      "learning_rate": 4.2093078344943334e-06,
+      "loss": 1.5405,
+      "step": 59720
+    },
+    {
+      "epoch": 1.7377898013206505,
+      "grad_norm": 11.375,
+      "learning_rate": 4.2073685528217185e-06,
+      "loss": 1.49,
+      "step": 59740
+    },
+    {
+      "epoch": 1.7383715856531983,
+      "grad_norm": 14.0,
+      "learning_rate": 4.205429271149104e-06,
+      "loss": 1.5166,
+      "step": 59760
+    },
+    {
+      "epoch": 1.7389533699857462,
+      "grad_norm": 14.4375,
+      "learning_rate": 4.203489989476489e-06,
+      "loss": 1.5367,
+      "step": 59780
+    },
+    {
+      "epoch": 1.7395351543182942,
+      "grad_norm": 11.5,
+      "learning_rate": 4.201550707803874e-06,
+      "loss": 1.5738,
+      "step": 59800
+    },
+    {
+      "epoch": 1.7401169386508422,
+      "grad_norm": 15.75,
+      "learning_rate": 4.199611426131259e-06,
+      "loss": 1.4617,
+      "step": 59820
+    },
+    {
+      "epoch": 1.7406987229833901,
+      "grad_norm": 15.875,
+      "learning_rate": 4.197672144458643e-06,
+      "loss": 1.6239,
+      "step": 59840
+    },
+    {
+      "epoch": 1.7412805073159379,
+      "grad_norm": 12.5,
+      "learning_rate": 4.195732862786028e-06,
+      "loss": 1.5498,
+      "step": 59860
+    },
+    {
+      "epoch": 1.741862291648486,
+      "grad_norm": 12.25,
+      "learning_rate": 4.193793581113413e-06,
+      "loss": 1.5468,
+      "step": 59880
+    },
+    {
+      "epoch": 1.7424440759810338,
+      "grad_norm": 13.8125,
+      "learning_rate": 4.1918542994407985e-06,
+      "loss": 1.4817,
+      "step": 59900
+    },
+    {
+      "epoch": 1.7430258603135818,
+      "grad_norm": 14.75,
+      "learning_rate": 4.1899150177681836e-06,
+      "loss": 1.549,
+      "step": 59920
+    },
+    {
+      "epoch": 1.7436076446461297,
+      "grad_norm": 12.375,
+      "learning_rate": 4.187975736095569e-06,
+      "loss": 1.5096,
+      "step": 59940
+    },
+    {
+      "epoch": 1.7441894289786775,
+      "grad_norm": 14.3125,
+      "learning_rate": 4.186036454422954e-06,
+      "loss": 1.5118,
+      "step": 59960
+    },
+    {
+      "epoch": 1.7447712133112256,
+      "grad_norm": 14.8125,
+      "learning_rate": 4.184097172750339e-06,
+      "loss": 1.6053,
+      "step": 59980
+    },
+    {
+      "epoch": 1.7453529976437734,
+      "grad_norm": 15.4375,
+      "learning_rate": 4.182157891077724e-06,
+      "loss": 1.4679,
+      "step": 60000
+    },
+    {
+      "epoch": 1.7459347819763213,
+      "grad_norm": 10.4375,
+      "learning_rate": 4.180218609405109e-06,
+      "loss": 1.5594,
+      "step": 60020
+    },
+    {
+      "epoch": 1.7465165663088693,
+      "grad_norm": 14.0,
+      "learning_rate": 4.178279327732494e-06,
+      "loss": 1.5039,
+      "step": 60040
+    },
+    {
+      "epoch": 1.7470983506414173,
+      "grad_norm": 13.125,
+      "learning_rate": 4.176340046059879e-06,
+      "loss": 1.5033,
+      "step": 60060
+    },
+    {
+      "epoch": 1.7476801349739652,
+      "grad_norm": 14.1875,
+      "learning_rate": 4.174400764387264e-06,
+      "loss": 1.5387,
+      "step": 60080
+    },
+    {
+      "epoch": 1.748261919306513,
+      "grad_norm": 12.25,
+      "learning_rate": 4.1724614827146494e-06,
+      "loss": 1.5421,
+      "step": 60100
+    },
+    {
+      "epoch": 1.7488437036390612,
+      "grad_norm": 13.625,
+      "learning_rate": 4.1705222010420345e-06,
+      "loss": 1.5557,
+      "step": 60120
+    },
+    {
+      "epoch": 1.749425487971609,
+      "grad_norm": 13.0625,
+      "learning_rate": 4.16858291936942e-06,
+      "loss": 1.529,
+      "step": 60140
+    },
+    {
+      "epoch": 1.7500072723041569,
+      "grad_norm": 13.5625,
+      "learning_rate": 4.166643637696805e-06,
+      "loss": 1.59,
+      "step": 60160
+    },
+    {
+      "epoch": 1.7505890566367048,
+      "grad_norm": 14.75,
+      "learning_rate": 4.16470435602419e-06,
+      "loss": 1.5466,
+      "step": 60180
+    },
+    {
+      "epoch": 1.7511708409692526,
+      "grad_norm": 14.375,
+      "learning_rate": 4.162765074351575e-06,
+      "loss": 1.5484,
+      "step": 60200
+    },
+    {
+      "epoch": 1.7517526253018008,
+      "grad_norm": 13.1875,
+      "learning_rate": 4.16082579267896e-06,
+      "loss": 1.5077,
+      "step": 60220
+    },
+    {
+      "epoch": 1.7523344096343485,
+      "grad_norm": 14.6875,
+      "learning_rate": 4.158886511006345e-06,
+      "loss": 1.5449,
+      "step": 60240
+    },
+    {
+      "epoch": 1.7529161939668965,
+      "grad_norm": 12.9375,
+      "learning_rate": 4.15694722933373e-06,
+      "loss": 1.5163,
+      "step": 60260
+    },
+    {
+      "epoch": 1.7534979782994444,
+      "grad_norm": 12.0625,
+      "learning_rate": 4.155007947661115e-06,
+      "loss": 1.4794,
+      "step": 60280
+    },
+    {
+      "epoch": 1.7540797626319922,
+      "grad_norm": 13.625,
+      "learning_rate": 4.1530686659885e-06,
+      "loss": 1.4792,
+      "step": 60300
+    },
+    {
+      "epoch": 1.7546615469645404,
+      "grad_norm": 14.5625,
+      "learning_rate": 4.1511293843158855e-06,
+      "loss": 1.554,
+      "step": 60320
+    },
+    {
+      "epoch": 1.755243331297088,
+      "grad_norm": 13.625,
+      "learning_rate": 4.14919010264327e-06,
+      "loss": 1.5754,
+      "step": 60340
+    },
+    {
+      "epoch": 1.755825115629636,
+      "grad_norm": 12.8125,
+      "learning_rate": 4.147250820970655e-06,
+      "loss": 1.5289,
+      "step": 60360
+    },
+    {
+      "epoch": 1.756406899962184,
+      "grad_norm": 10.6875,
+      "learning_rate": 4.14531153929804e-06,
+      "loss": 1.3856,
+      "step": 60380
+    },
+    {
+      "epoch": 1.756988684294732,
+      "grad_norm": 13.4375,
+      "learning_rate": 4.143372257625425e-06,
+      "loss": 1.5795,
+      "step": 60400
+    },
+    {
+      "epoch": 1.75757046862728,
+      "grad_norm": 13.8125,
+      "learning_rate": 4.14143297595281e-06,
+      "loss": 1.5497,
+      "step": 60420
+    },
+    {
+      "epoch": 1.7581522529598277,
+      "grad_norm": 11.0625,
+      "learning_rate": 4.139493694280195e-06,
+      "loss": 1.5081,
+      "step": 60440
+    },
+    {
+      "epoch": 1.7587340372923759,
+      "grad_norm": 13.25,
+      "learning_rate": 4.13755441260758e-06,
+      "loss": 1.5566,
+      "step": 60460
+    },
+    {
+      "epoch": 1.7593158216249236,
+      "grad_norm": 12.8125,
+      "learning_rate": 4.1356151309349654e-06,
+      "loss": 1.4932,
+      "step": 60480
+    },
+    {
+      "epoch": 1.7598976059574716,
+      "grad_norm": 13.4375,
+      "learning_rate": 4.1336758492623505e-06,
+      "loss": 1.5285,
+      "step": 60500
+    },
+    {
+      "epoch": 1.7604793902900195,
+      "grad_norm": 14.4375,
+      "learning_rate": 4.131736567589736e-06,
+      "loss": 1.5651,
+      "step": 60520
+    },
+    {
+      "epoch": 1.7610611746225673,
+      "grad_norm": 13.8125,
+      "learning_rate": 4.129797285917121e-06,
+      "loss": 1.5124,
+      "step": 60540
+    },
+    {
+      "epoch": 1.7616429589551155,
+      "grad_norm": 13.9375,
+      "learning_rate": 4.127858004244506e-06,
+      "loss": 1.4658,
+      "step": 60560
+    },
+    {
+      "epoch": 1.7622247432876632,
+      "grad_norm": 17.5,
+      "learning_rate": 4.125918722571891e-06,
+      "loss": 1.4134,
+      "step": 60580
+    },
+    {
+      "epoch": 1.7628065276202112,
+      "grad_norm": 12.375,
+      "learning_rate": 4.123979440899276e-06,
+      "loss": 1.5309,
+      "step": 60600
+    },
+    {
+      "epoch": 1.7633883119527591,
+      "grad_norm": 13.4375,
+      "learning_rate": 4.122040159226661e-06,
+      "loss": 1.529,
+      "step": 60620
+    },
+    {
+      "epoch": 1.7639700962853069,
+      "grad_norm": 12.6875,
+      "learning_rate": 4.120100877554046e-06,
+      "loss": 1.5403,
+      "step": 60640
+    },
+    {
+      "epoch": 1.764551880617855,
+      "grad_norm": 13.1875,
+      "learning_rate": 4.118161595881431e-06,
+      "loss": 1.4753,
+      "step": 60660
+    },
+    {
+      "epoch": 1.7651336649504028,
+      "grad_norm": 18.125,
+      "learning_rate": 4.116222314208816e-06,
+      "loss": 1.5695,
+      "step": 60680
+    },
+    {
+      "epoch": 1.7657154492829508,
+      "grad_norm": 13.25,
+      "learning_rate": 4.1142830325362015e-06,
+      "loss": 1.5104,
+      "step": 60700
+    },
+    {
+      "epoch": 1.7662972336154987,
+      "grad_norm": 12.6875,
+      "learning_rate": 4.112343750863587e-06,
+      "loss": 1.5038,
+      "step": 60720
+    },
+    {
+      "epoch": 1.7668790179480467,
+      "grad_norm": 12.875,
+      "learning_rate": 4.110404469190972e-06,
+      "loss": 1.4563,
+      "step": 60740
+    },
+    {
+      "epoch": 1.7674608022805947,
+      "grad_norm": 13.0,
+      "learning_rate": 4.108465187518357e-06,
+      "loss": 1.5103,
+      "step": 60760
+    },
+    {
+      "epoch": 1.7680425866131424,
+      "grad_norm": 12.4375,
+      "learning_rate": 4.106525905845742e-06,
+      "loss": 1.4324,
+      "step": 60780
+    },
+    {
+      "epoch": 1.7686243709456906,
+      "grad_norm": 14.0625,
+      "learning_rate": 4.104586624173127e-06,
+      "loss": 1.4796,
+      "step": 60800
+    },
+    {
+      "epoch": 1.7692061552782383,
+      "grad_norm": 13.3125,
+      "learning_rate": 4.102647342500512e-06,
+      "loss": 1.5313,
+      "step": 60820
+    },
+    {
+      "epoch": 1.7697879396107863,
+      "grad_norm": 13.0,
+      "learning_rate": 4.100708060827897e-06,
+      "loss": 1.5113,
+      "step": 60840
+    },
+    {
+      "epoch": 1.7703697239433343,
+      "grad_norm": 13.5,
+      "learning_rate": 4.098768779155282e-06,
+      "loss": 1.5719,
+      "step": 60860
+    },
+    {
+      "epoch": 1.770951508275882,
+      "grad_norm": 15.25,
+      "learning_rate": 4.096829497482667e-06,
+      "loss": 1.5592,
+      "step": 60880
+    },
+    {
+      "epoch": 1.7715332926084302,
+      "grad_norm": 14.5,
+      "learning_rate": 4.0948902158100525e-06,
+      "loss": 1.5779,
+      "step": 60900
+    },
+    {
+      "epoch": 1.772115076940978,
+      "grad_norm": 13.5625,
+      "learning_rate": 4.092950934137438e-06,
+      "loss": 1.5096,
+      "step": 60920
+    },
+    {
+      "epoch": 1.7726968612735259,
+      "grad_norm": 14.0625,
+      "learning_rate": 4.091011652464823e-06,
+      "loss": 1.5509,
+      "step": 60940
+    },
+    {
+      "epoch": 1.7732786456060738,
+      "grad_norm": 14.3125,
+      "learning_rate": 4.089072370792208e-06,
+      "loss": 1.5349,
+      "step": 60960
+    },
+    {
+      "epoch": 1.7738604299386218,
+      "grad_norm": 14.4375,
+      "learning_rate": 4.087133089119593e-06,
+      "loss": 1.5832,
+      "step": 60980
+    },
+    {
+      "epoch": 1.7744422142711698,
+      "grad_norm": 13.5,
+      "learning_rate": 4.085193807446978e-06,
+      "loss": 1.5272,
+      "step": 61000
+    },
+    {
+      "epoch": 1.7750239986037175,
+      "grad_norm": 11.375,
+      "learning_rate": 4.083254525774363e-06,
+      "loss": 1.4585,
+      "step": 61020
+    },
+    {
+      "epoch": 1.7756057829362655,
+      "grad_norm": 13.125,
+      "learning_rate": 4.081315244101748e-06,
+      "loss": 1.3838,
+      "step": 61040
+    },
+    {
+      "epoch": 1.7761875672688134,
+      "grad_norm": 12.6875,
+      "learning_rate": 4.079375962429133e-06,
+      "loss": 1.6207,
+      "step": 61060
+    },
+    {
+      "epoch": 1.7767693516013614,
+      "grad_norm": 13.6875,
+      "learning_rate": 4.077436680756518e-06,
+      "loss": 1.4856,
+      "step": 61080
+    },
+    {
+      "epoch": 1.7773511359339094,
+      "grad_norm": 13.25,
+      "learning_rate": 4.0754973990839035e-06,
+      "loss": 1.5682,
+      "step": 61100
+    },
+    {
+      "epoch": 1.777932920266457,
+      "grad_norm": 14.0625,
+      "learning_rate": 4.0735581174112886e-06,
+      "loss": 1.577,
+      "step": 61120
+    },
+    {
+      "epoch": 1.7785147045990053,
+      "grad_norm": 14.8125,
+      "learning_rate": 4.071618835738673e-06,
+      "loss": 1.3515,
+      "step": 61140
+    },
+    {
+      "epoch": 1.779096488931553,
+      "grad_norm": 13.125,
+      "learning_rate": 4.069679554066058e-06,
+      "loss": 1.5357,
+      "step": 61160
+    },
+    {
+      "epoch": 1.779678273264101,
+      "grad_norm": 12.0,
+      "learning_rate": 4.067740272393443e-06,
+      "loss": 1.4878,
+      "step": 61180
+    },
+    {
+      "epoch": 1.780260057596649,
+      "grad_norm": 13.9375,
+      "learning_rate": 4.065800990720828e-06,
+      "loss": 1.5168,
+      "step": 61200
+    },
+    {
+      "epoch": 1.7808418419291967,
+      "grad_norm": 13.5625,
+      "learning_rate": 4.063861709048213e-06,
+      "loss": 1.5326,
+      "step": 61220
+    },
+    {
+      "epoch": 1.7814236262617449,
+      "grad_norm": 14.25,
+      "learning_rate": 4.061922427375598e-06,
+      "loss": 1.4947,
+      "step": 61240
+    },
+    {
+      "epoch": 1.7820054105942926,
+      "grad_norm": 13.25,
+      "learning_rate": 4.059983145702983e-06,
+      "loss": 1.559,
+      "step": 61260
+    },
+    {
+      "epoch": 1.7825871949268406,
+      "grad_norm": 14.6875,
+      "learning_rate": 4.0580438640303685e-06,
+      "loss": 1.5615,
+      "step": 61280
+    },
+    {
+      "epoch": 1.7831689792593886,
+      "grad_norm": 13.6875,
+      "learning_rate": 4.056104582357754e-06,
+      "loss": 1.6296,
+      "step": 61300
+    },
+    {
+      "epoch": 1.7837507635919365,
+      "grad_norm": 15.0,
+      "learning_rate": 4.054165300685139e-06,
+      "loss": 1.4333,
+      "step": 61320
+    },
+    {
+      "epoch": 1.7843325479244845,
+      "grad_norm": 13.1875,
+      "learning_rate": 4.052226019012524e-06,
+      "loss": 1.6036,
+      "step": 61340
+    },
+    {
+      "epoch": 1.7849143322570322,
+      "grad_norm": 12.625,
+      "learning_rate": 4.050286737339909e-06,
+      "loss": 1.5461,
+      "step": 61360
+    },
+    {
+      "epoch": 1.7854961165895804,
+      "grad_norm": 14.25,
+      "learning_rate": 4.048347455667294e-06,
+      "loss": 1.5809,
+      "step": 61380
+    },
+    {
+      "epoch": 1.7860779009221281,
+      "grad_norm": 12.75,
+      "learning_rate": 4.046408173994679e-06,
+      "loss": 1.523,
+      "step": 61400
+    },
+    {
+      "epoch": 1.786659685254676,
+      "grad_norm": 12.625,
+      "learning_rate": 4.044468892322064e-06,
+      "loss": 1.4993,
+      "step": 61420
+    },
+    {
+      "epoch": 1.787241469587224,
+      "grad_norm": 11.375,
+      "learning_rate": 4.042529610649449e-06,
+      "loss": 1.4793,
+      "step": 61440
+    },
+    {
+      "epoch": 1.7878232539197718,
+      "grad_norm": 13.625,
+      "learning_rate": 4.040590328976834e-06,
+      "loss": 1.5591,
+      "step": 61460
+    },
+    {
+      "epoch": 1.78840503825232,
+      "grad_norm": 14.6875,
+      "learning_rate": 4.0386510473042195e-06,
+      "loss": 1.6111,
+      "step": 61480
+    },
+    {
+      "epoch": 1.7889868225848677,
+      "grad_norm": 17.625,
+      "learning_rate": 4.0367117656316046e-06,
+      "loss": 1.503,
+      "step": 61500
+    },
+    {
+      "epoch": 1.7895686069174157,
+      "grad_norm": 15.625,
+      "learning_rate": 4.03477248395899e-06,
+      "loss": 1.5823,
+      "step": 61520
+    },
+    {
+      "epoch": 1.7901503912499637,
+      "grad_norm": 14.375,
+      "learning_rate": 4.032833202286375e-06,
+      "loss": 1.5218,
+      "step": 61540
+    },
+    {
+      "epoch": 1.7907321755825114,
+      "grad_norm": 12.875,
+      "learning_rate": 4.03089392061376e-06,
+      "loss": 1.4664,
+      "step": 61560
+    },
+    {
+      "epoch": 1.7913139599150596,
+      "grad_norm": 16.375,
+      "learning_rate": 4.028954638941145e-06,
+      "loss": 1.4916,
+      "step": 61580
+    },
+    {
+      "epoch": 1.7918957442476073,
+      "grad_norm": 14.75,
+      "learning_rate": 4.02701535726853e-06,
+      "loss": 1.5355,
+      "step": 61600
+    },
+    {
+      "epoch": 1.7924775285801553,
+      "grad_norm": 15.1875,
+      "learning_rate": 4.025076075595915e-06,
+      "loss": 1.5685,
+      "step": 61620
+    },
+    {
+      "epoch": 1.7930593129127033,
+      "grad_norm": 14.875,
+      "learning_rate": 4.0231367939233e-06,
+      "loss": 1.6289,
+      "step": 61640
+    },
+    {
+      "epoch": 1.7936410972452512,
+      "grad_norm": 12.8125,
+      "learning_rate": 4.021197512250685e-06,
+      "loss": 1.4561,
+      "step": 61660
+    },
+    {
+      "epoch": 1.7942228815777992,
+      "grad_norm": 12.0625,
+      "learning_rate": 4.0192582305780704e-06,
+      "loss": 1.4562,
+      "step": 61680
+    },
+    {
+      "epoch": 1.794804665910347,
+      "grad_norm": 13.5,
+      "learning_rate": 4.0173189489054555e-06,
+      "loss": 1.4413,
+      "step": 61700
+    },
+    {
+      "epoch": 1.7953864502428951,
+      "grad_norm": 15.25,
+      "learning_rate": 4.015379667232841e-06,
+      "loss": 1.4938,
+      "step": 61720
+    },
+    {
+      "epoch": 1.7959682345754429,
+      "grad_norm": 11.1875,
+      "learning_rate": 4.013440385560226e-06,
+      "loss": 1.5027,
+      "step": 61740
+    },
+    {
+      "epoch": 1.7965500189079908,
+      "grad_norm": 13.875,
+      "learning_rate": 4.011501103887611e-06,
+      "loss": 1.5781,
+      "step": 61760
+    },
+    {
+      "epoch": 1.7971318032405388,
+      "grad_norm": 13.8125,
+      "learning_rate": 4.009561822214996e-06,
+      "loss": 1.4809,
+      "step": 61780
+    },
+    {
+      "epoch": 1.7977135875730865,
+      "grad_norm": 12.1875,
+      "learning_rate": 4.007622540542381e-06,
+      "loss": 1.5166,
+      "step": 61800
+    },
+    {
+      "epoch": 1.7982953719056347,
+      "grad_norm": 7.1875,
+      "learning_rate": 4.005683258869766e-06,
+      "loss": 1.4717,
+      "step": 61820
+    },
+    {
+      "epoch": 1.7988771562381825,
+      "grad_norm": 11.0,
+      "learning_rate": 4.003743977197151e-06,
+      "loss": 1.5472,
+      "step": 61840
+    },
+    {
+      "epoch": 1.7994589405707304,
+      "grad_norm": 13.25,
+      "learning_rate": 4.001804695524536e-06,
+      "loss": 1.4978,
+      "step": 61860
+    },
+    {
+      "epoch": 1.8000407249032784,
+      "grad_norm": 11.0,
+      "learning_rate": 3.999865413851921e-06,
+      "loss": 1.505,
+      "step": 61880
+    },
+    {
+      "epoch": 1.8006225092358261,
+      "grad_norm": 17.0,
+      "learning_rate": 3.9979261321793065e-06,
+      "loss": 1.5679,
+      "step": 61900
+    },
+    {
+      "epoch": 1.8012042935683743,
+      "grad_norm": 11.375,
+      "learning_rate": 3.995986850506692e-06,
+      "loss": 1.4906,
+      "step": 61920
+    },
+    {
+      "epoch": 1.801786077900922,
+      "grad_norm": 12.6875,
+      "learning_rate": 3.994047568834077e-06,
+      "loss": 1.4992,
+      "step": 61940
+    },
+    {
+      "epoch": 1.80236786223347,
+      "grad_norm": 10.8125,
+      "learning_rate": 3.992108287161461e-06,
+      "loss": 1.5971,
+      "step": 61960
+    },
+    {
+      "epoch": 1.802949646566018,
+      "grad_norm": 12.875,
+      "learning_rate": 3.990169005488846e-06,
+      "loss": 1.4525,
+      "step": 61980
+    },
+    {
+      "epoch": 1.803531430898566,
+      "grad_norm": 12.3125,
+      "learning_rate": 3.988229723816231e-06,
+      "loss": 1.5309,
+      "step": 62000
+    },
+    {
+      "epoch": 1.804113215231114,
+      "grad_norm": 12.8125,
+      "learning_rate": 3.986290442143616e-06,
+      "loss": 1.5509,
+      "step": 62020
+    },
+    {
+      "epoch": 1.8046949995636616,
+      "grad_norm": 11.0,
+      "learning_rate": 3.984351160471001e-06,
+      "loss": 1.5824,
+      "step": 62040
+    },
+    {
+      "epoch": 1.8052767838962098,
+      "grad_norm": 14.0,
+      "learning_rate": 3.9824118787983864e-06,
+      "loss": 1.5651,
+      "step": 62060
+    },
+    {
+      "epoch": 1.8058585682287576,
+      "grad_norm": 12.1875,
+      "learning_rate": 3.9804725971257715e-06,
+      "loss": 1.5661,
+      "step": 62080
+    },
+    {
+      "epoch": 1.8064403525613055,
+      "grad_norm": 11.3125,
+      "learning_rate": 3.978533315453157e-06,
+      "loss": 1.5185,
+      "step": 62100
+    },
+    {
+      "epoch": 1.8070221368938535,
+      "grad_norm": 12.0,
+      "learning_rate": 3.976594033780542e-06,
+      "loss": 1.6406,
+      "step": 62120
+    },
+    {
+      "epoch": 1.8076039212264012,
+      "grad_norm": 13.25,
+      "learning_rate": 3.974654752107927e-06,
+      "loss": 1.5302,
+      "step": 62140
+    },
+    {
+      "epoch": 1.8081857055589494,
+      "grad_norm": 13.5625,
+      "learning_rate": 3.972715470435312e-06,
+      "loss": 1.5314,
+      "step": 62160
+    },
+    {
+      "epoch": 1.8087674898914972,
+      "grad_norm": 14.3125,
+      "learning_rate": 3.970776188762697e-06,
+      "loss": 1.6031,
+      "step": 62180
+    },
+    {
+      "epoch": 1.8093492742240451,
+      "grad_norm": 12.0,
+      "learning_rate": 3.968836907090082e-06,
+      "loss": 1.4817,
+      "step": 62200
+    },
+    {
+      "epoch": 1.809931058556593,
+      "grad_norm": 13.8125,
+      "learning_rate": 3.966897625417467e-06,
+      "loss": 1.585,
+      "step": 62220
+    },
+    {
+      "epoch": 1.810512842889141,
+      "grad_norm": 13.75,
+      "learning_rate": 3.9649583437448515e-06,
+      "loss": 1.513,
+      "step": 62240
+    },
+    {
+      "epoch": 1.811094627221689,
+      "grad_norm": 10.8125,
+      "learning_rate": 3.9630190620722366e-06,
+      "loss": 1.5597,
+      "step": 62260
+    },
+    {
+      "epoch": 1.8116764115542368,
+      "grad_norm": 11.375,
+      "learning_rate": 3.961079780399622e-06,
+      "loss": 1.5187,
+      "step": 62280
+    },
+    {
+      "epoch": 1.8122581958867847,
+      "grad_norm": 13.6875,
+      "learning_rate": 3.959140498727007e-06,
+      "loss": 1.4977,
+      "step": 62300
+    },
+    {
+      "epoch": 1.8128399802193327,
+      "grad_norm": 11.125,
+      "learning_rate": 3.957201217054392e-06,
+      "loss": 1.5414,
+      "step": 62320
+    },
+    {
+      "epoch": 1.8134217645518806,
+      "grad_norm": 13.0,
+      "learning_rate": 3.955261935381777e-06,
+      "loss": 1.5543,
+      "step": 62340
+    },
+    {
+      "epoch": 1.8140035488844286,
+      "grad_norm": 11.9375,
+      "learning_rate": 3.953322653709162e-06,
+      "loss": 1.4898,
+      "step": 62360
+    },
+    {
+      "epoch": 1.8145853332169763,
+      "grad_norm": 12.9375,
+      "learning_rate": 3.951383372036547e-06,
+      "loss": 1.4591,
+      "step": 62380
+    },
+    {
+      "epoch": 1.8151671175495245,
+      "grad_norm": 12.75,
+      "learning_rate": 3.949444090363932e-06,
+      "loss": 1.5138,
+      "step": 62400
+    },
+    {
+      "epoch": 1.8157489018820723,
+      "grad_norm": 14.0,
+      "learning_rate": 3.947504808691317e-06,
+      "loss": 1.5751,
+      "step": 62420
+    },
+    {
+      "epoch": 1.8163306862146202,
+      "grad_norm": 11.125,
+      "learning_rate": 3.9455655270187024e-06,
+      "loss": 1.4689,
+      "step": 62440
+    },
+    {
+      "epoch": 1.8169124705471682,
+      "grad_norm": 12.8125,
+      "learning_rate": 3.9436262453460875e-06,
+      "loss": 1.5339,
+      "step": 62460
+    },
+    {
+      "epoch": 1.817494254879716,
+      "grad_norm": 10.0,
+      "learning_rate": 3.941686963673473e-06,
+      "loss": 1.5708,
+      "step": 62480
+    },
+    {
+      "epoch": 1.8180760392122641,
+      "grad_norm": 12.6875,
+      "learning_rate": 3.939747682000858e-06,
+      "loss": 1.5428,
+      "step": 62500
+    },
+    {
+      "epoch": 1.8186578235448119,
+      "grad_norm": 13.0625,
+      "learning_rate": 3.937808400328243e-06,
+      "loss": 1.4831,
+      "step": 62520
+    },
+    {
+      "epoch": 1.8192396078773598,
+      "grad_norm": 13.3125,
+      "learning_rate": 3.935869118655628e-06,
+      "loss": 1.5751,
+      "step": 62540
+    },
+    {
+      "epoch": 1.8198213922099078,
+      "grad_norm": 13.0625,
+      "learning_rate": 3.933929836983013e-06,
+      "loss": 1.4526,
+      "step": 62560
+    },
+    {
+      "epoch": 1.8204031765424558,
+      "grad_norm": 13.25,
+      "learning_rate": 3.931990555310398e-06,
+      "loss": 1.4694,
+      "step": 62580
+    },
+    {
+      "epoch": 1.8209849608750037,
+      "grad_norm": 12.9375,
+      "learning_rate": 3.930051273637783e-06,
+      "loss": 1.5837,
+      "step": 62600
+    },
+    {
+      "epoch": 1.8215667452075515,
+      "grad_norm": 13.6875,
+      "learning_rate": 3.928111991965168e-06,
+      "loss": 1.5205,
+      "step": 62620
+    },
+    {
+      "epoch": 1.8221485295400996,
+      "grad_norm": 13.3125,
+      "learning_rate": 3.926172710292553e-06,
+      "loss": 1.4634,
+      "step": 62640
+    },
+    {
+      "epoch": 1.8227303138726474,
+      "grad_norm": 11.6875,
+      "learning_rate": 3.9242334286199385e-06,
+      "loss": 1.593,
+      "step": 62660
+    },
+    {
+      "epoch": 1.8233120982051954,
+      "grad_norm": 13.9375,
+      "learning_rate": 3.922294146947324e-06,
+      "loss": 1.4876,
+      "step": 62680
+    },
+    {
+      "epoch": 1.8238938825377433,
+      "grad_norm": 15.5625,
+      "learning_rate": 3.920354865274709e-06,
+      "loss": 1.5392,
+      "step": 62700
+    },
+    {
+      "epoch": 1.824475666870291,
+      "grad_norm": 9.25,
+      "learning_rate": 3.918415583602094e-06,
+      "loss": 1.5972,
+      "step": 62720
+    },
+    {
+      "epoch": 1.8250574512028392,
+      "grad_norm": 12.8125,
+      "learning_rate": 3.916476301929479e-06,
+      "loss": 1.4216,
+      "step": 62740
+    },
+    {
+      "epoch": 1.825639235535387,
+      "grad_norm": 14.3125,
+      "learning_rate": 3.914537020256864e-06,
+      "loss": 1.5259,
+      "step": 62760
+    },
+    {
+      "epoch": 1.826221019867935,
+      "grad_norm": 16.0,
+      "learning_rate": 3.912597738584249e-06,
+      "loss": 1.5084,
+      "step": 62780
+    },
+    {
+      "epoch": 1.826802804200483,
+      "grad_norm": 12.6875,
+      "learning_rate": 3.910658456911634e-06,
+      "loss": 1.583,
+      "step": 62800
+    },
+    {
+      "epoch": 1.8273845885330307,
+      "grad_norm": 13.375,
+      "learning_rate": 3.908719175239019e-06,
+      "loss": 1.5648,
+      "step": 62820
+    },
+    {
+      "epoch": 1.8279663728655788,
+      "grad_norm": 13.875,
+      "learning_rate": 3.906779893566404e-06,
+      "loss": 1.5481,
+      "step": 62840
+    },
+    {
+      "epoch": 1.8285481571981266,
+      "grad_norm": 13.3125,
+      "learning_rate": 3.9048406118937895e-06,
+      "loss": 1.6228,
+      "step": 62860
+    },
+    {
+      "epoch": 1.8291299415306745,
+      "grad_norm": 15.625,
+      "learning_rate": 3.902901330221175e-06,
+      "loss": 1.4922,
+      "step": 62880
+    },
+    {
+      "epoch": 1.8297117258632225,
+      "grad_norm": 15.75,
+      "learning_rate": 3.90096204854856e-06,
+      "loss": 1.5099,
+      "step": 62900
+    },
+    {
+      "epoch": 1.8302935101957705,
+      "grad_norm": 11.6875,
+      "learning_rate": 3.899022766875945e-06,
+      "loss": 1.5234,
+      "step": 62920
+    },
+    {
+      "epoch": 1.8308752945283184,
+      "grad_norm": 13.5625,
+      "learning_rate": 3.89708348520333e-06,
+      "loss": 1.5024,
+      "step": 62940
+    },
+    {
+      "epoch": 1.8314570788608662,
+      "grad_norm": 13.5625,
+      "learning_rate": 3.895144203530715e-06,
+      "loss": 1.5681,
+      "step": 62960
+    },
+    {
+      "epoch": 1.8320388631934144,
+      "grad_norm": 10.6875,
+      "learning_rate": 3.8932049218581e-06,
+      "loss": 1.5527,
+      "step": 62980
+    },
+    {
+      "epoch": 1.832620647525962,
+      "grad_norm": 13.6875,
+      "learning_rate": 3.891265640185485e-06,
+      "loss": 1.4973,
+      "step": 63000
+    },
+    {
+      "epoch": 1.83320243185851,
+      "grad_norm": 13.9375,
+      "learning_rate": 3.88932635851287e-06,
+      "loss": 1.5763,
+      "step": 63020
+    },
+    {
+      "epoch": 1.833784216191058,
+      "grad_norm": 12.3125,
+      "learning_rate": 3.887387076840255e-06,
+      "loss": 1.5511,
+      "step": 63040
+    },
+    {
+      "epoch": 1.8343660005236058,
+      "grad_norm": 14.0625,
+      "learning_rate": 3.88544779516764e-06,
+      "loss": 1.5727,
+      "step": 63060
+    },
+    {
+      "epoch": 1.834947784856154,
+      "grad_norm": 21.375,
+      "learning_rate": 3.883508513495025e-06,
+      "loss": 1.4631,
+      "step": 63080
+    },
+    {
+      "epoch": 1.8355295691887017,
+      "grad_norm": 15.0,
+      "learning_rate": 3.88156923182241e-06,
+      "loss": 1.5165,
+      "step": 63100
+    },
+    {
+      "epoch": 1.8361113535212497,
+      "grad_norm": 12.9375,
+      "learning_rate": 3.879629950149795e-06,
+      "loss": 1.5668,
+      "step": 63120
+    },
+    {
+      "epoch": 1.8366931378537976,
+      "grad_norm": 17.125,
+      "learning_rate": 3.87769066847718e-06,
+      "loss": 1.5744,
+      "step": 63140
+    },
+    {
+      "epoch": 1.8372749221863454,
+      "grad_norm": 13.9375,
+      "learning_rate": 3.875751386804565e-06,
+      "loss": 1.5037,
+      "step": 63160
+    },
+    {
+      "epoch": 1.8378567065188935,
+      "grad_norm": 14.0625,
+      "learning_rate": 3.87381210513195e-06,
+      "loss": 1.5699,
+      "step": 63180
+    },
+    {
+      "epoch": 1.8384384908514413,
+      "grad_norm": 15.875,
+      "learning_rate": 3.871872823459335e-06,
+      "loss": 1.5373,
+      "step": 63200
+    },
+    {
+      "epoch": 1.8390202751839893,
+      "grad_norm": 11.25,
+      "learning_rate": 3.86993354178672e-06,
+      "loss": 1.5129,
+      "step": 63220
+    },
+    {
+      "epoch": 1.8396020595165372,
+      "grad_norm": 14.5625,
+      "learning_rate": 3.8679942601141055e-06,
+      "loss": 1.499,
+      "step": 63240
+    },
+    {
+      "epoch": 1.8401838438490852,
+      "grad_norm": 13.625,
+      "learning_rate": 3.866054978441491e-06,
+      "loss": 1.5906,
+      "step": 63260
+    },
+    {
+      "epoch": 1.8407656281816331,
+      "grad_norm": 12.75,
+      "learning_rate": 3.864115696768876e-06,
+      "loss": 1.466,
+      "step": 63280
+    },
+    {
+      "epoch": 1.8413474125141809,
+      "grad_norm": 12.25,
+      "learning_rate": 3.862176415096261e-06,
+      "loss": 1.5761,
+      "step": 63300
+    },
+    {
+      "epoch": 1.841929196846729,
+      "grad_norm": 13.75,
+      "learning_rate": 3.860237133423646e-06,
+      "loss": 1.4375,
+      "step": 63320
+    },
+    {
+      "epoch": 1.8425109811792768,
+      "grad_norm": 11.5,
+      "learning_rate": 3.858297851751031e-06,
+      "loss": 1.4889,
+      "step": 63340
+    },
+    {
+      "epoch": 1.8430927655118248,
+      "grad_norm": 14.125,
+      "learning_rate": 3.856358570078416e-06,
+      "loss": 1.5614,
+      "step": 63360
+    },
+    {
+      "epoch": 1.8436745498443727,
+      "grad_norm": 13.6875,
+      "learning_rate": 3.854419288405801e-06,
+      "loss": 1.5795,
+      "step": 63380
+    },
+    {
+      "epoch": 1.8442563341769205,
+      "grad_norm": 16.375,
+      "learning_rate": 3.852480006733186e-06,
+      "loss": 1.5741,
+      "step": 63400
+    },
+    {
+      "epoch": 1.8448381185094687,
+      "grad_norm": 11.4375,
+      "learning_rate": 3.850540725060571e-06,
+      "loss": 1.5355,
+      "step": 63420
+    },
+    {
+      "epoch": 1.8454199028420164,
+      "grad_norm": 12.3125,
+      "learning_rate": 3.8486014433879565e-06,
+      "loss": 1.4689,
+      "step": 63440
+    },
+    {
+      "epoch": 1.8460016871745644,
+      "grad_norm": 12.3125,
+      "learning_rate": 3.8466621617153416e-06,
+      "loss": 1.5845,
+      "step": 63460
+    },
+    {
+      "epoch": 1.8465834715071123,
+      "grad_norm": 13.0625,
+      "learning_rate": 3.844722880042727e-06,
+      "loss": 1.5011,
+      "step": 63480
+    },
+    {
+      "epoch": 1.8471652558396603,
+      "grad_norm": 14.6875,
+      "learning_rate": 3.842783598370112e-06,
+      "loss": 1.5878,
+      "step": 63500
+    },
+    {
+      "epoch": 1.8477470401722083,
+      "grad_norm": 11.8125,
+      "learning_rate": 3.840844316697497e-06,
+      "loss": 1.536,
+      "step": 63520
+    },
+    {
+      "epoch": 1.848328824504756,
+      "grad_norm": 13.625,
+      "learning_rate": 3.838905035024882e-06,
+      "loss": 1.619,
+      "step": 63540
+    },
+    {
+      "epoch": 1.848910608837304,
+      "grad_norm": 14.8125,
+      "learning_rate": 3.836965753352267e-06,
+      "loss": 1.5491,
+      "step": 63560
+    },
+    {
+      "epoch": 1.849492393169852,
+      "grad_norm": 14.6875,
+      "learning_rate": 3.835026471679652e-06,
+      "loss": 1.4322,
+      "step": 63580
+    },
+    {
+      "epoch": 1.8500741775023999,
+      "grad_norm": 10.125,
+      "learning_rate": 3.833087190007037e-06,
+      "loss": 1.5235,
+      "step": 63600
+    },
+    {
+      "epoch": 1.8506559618349478,
+      "grad_norm": 10.375,
+      "learning_rate": 3.831147908334422e-06,
+      "loss": 1.4756,
+      "step": 63620
+    },
+    {
+      "epoch": 1.8512377461674956,
+      "grad_norm": 14.1875,
+      "learning_rate": 3.8292086266618074e-06,
+      "loss": 1.5428,
+      "step": 63640
+    },
+    {
+      "epoch": 1.8518195305000438,
+      "grad_norm": 12.8125,
+      "learning_rate": 3.8272693449891925e-06,
+      "loss": 1.4258,
+      "step": 63660
+    },
+    {
+      "epoch": 1.8524013148325915,
+      "grad_norm": 14.4375,
+      "learning_rate": 3.825330063316578e-06,
+      "loss": 1.5605,
+      "step": 63680
+    },
+    {
+      "epoch": 1.8529830991651395,
+      "grad_norm": 11.9375,
+      "learning_rate": 3.823390781643963e-06,
+      "loss": 1.6128,
+      "step": 63700
+    },
+    {
+      "epoch": 1.8535648834976874,
+      "grad_norm": 14.125,
+      "learning_rate": 3.821451499971348e-06,
+      "loss": 1.4939,
+      "step": 63720
+    },
+    {
+      "epoch": 1.8541466678302352,
+      "grad_norm": 14.125,
+      "learning_rate": 3.819512218298733e-06,
+      "loss": 1.5789,
+      "step": 63740
+    },
+    {
+      "epoch": 1.8547284521627834,
+      "grad_norm": 13.6875,
+      "learning_rate": 3.817572936626118e-06,
+      "loss": 1.5197,
+      "step": 63760
+    },
+    {
+      "epoch": 1.855310236495331,
+      "grad_norm": 13.9375,
+      "learning_rate": 3.815633654953503e-06,
+      "loss": 1.5028,
+      "step": 63780
+    },
+    {
+      "epoch": 1.855892020827879,
+      "grad_norm": 11.125,
+      "learning_rate": 3.813694373280888e-06,
+      "loss": 1.5733,
+      "step": 63800
+    },
+    {
+      "epoch": 1.856473805160427,
+      "grad_norm": 13.375,
+      "learning_rate": 3.811755091608273e-06,
+      "loss": 1.4623,
+      "step": 63820
+    },
+    {
+      "epoch": 1.857055589492975,
+      "grad_norm": 11.75,
+      "learning_rate": 3.809815809935658e-06,
+      "loss": 1.5711,
+      "step": 63840
+    },
+    {
+      "epoch": 1.857637373825523,
+      "grad_norm": 13.625,
+      "learning_rate": 3.807876528263043e-06,
+      "loss": 1.5969,
+      "step": 63860
+    },
+    {
+      "epoch": 1.8582191581580707,
+      "grad_norm": 13.0625,
+      "learning_rate": 3.805937246590428e-06,
+      "loss": 1.593,
+      "step": 63880
+    },
+    {
+      "epoch": 1.858800942490619,
+      "grad_norm": 14.375,
+      "learning_rate": 3.8039979649178133e-06,
+      "loss": 1.5419,
+      "step": 63900
+    },
+    {
+      "epoch": 1.8593827268231666,
+      "grad_norm": 12.375,
+      "learning_rate": 3.8020586832451984e-06,
+      "loss": 1.5262,
+      "step": 63920
+    },
+    {
+      "epoch": 1.8599645111557146,
+      "grad_norm": 14.125,
+      "learning_rate": 3.8001194015725835e-06,
+      "loss": 1.5379,
+      "step": 63940
+    },
+    {
+      "epoch": 1.8605462954882626,
+      "grad_norm": 14.0,
+      "learning_rate": 3.7981801198999686e-06,
+      "loss": 1.4929,
+      "step": 63960
+    },
+    {
+      "epoch": 1.8611280798208103,
+      "grad_norm": 12.5,
+      "learning_rate": 3.7962408382273537e-06,
+      "loss": 1.5153,
+      "step": 63980
+    },
+    {
+      "epoch": 1.8617098641533585,
+      "grad_norm": 14.3125,
+      "learning_rate": 3.7943015565547388e-06,
+      "loss": 1.4638,
+      "step": 64000
+    },
+    {
+      "epoch": 1.8622916484859062,
+      "grad_norm": 13.1875,
+      "learning_rate": 3.792362274882124e-06,
+      "loss": 1.5608,
+      "step": 64020
+    },
+    {
+      "epoch": 1.8628734328184542,
+      "grad_norm": 14.0,
+      "learning_rate": 3.790422993209509e-06,
+      "loss": 1.5902,
+      "step": 64040
+    },
+    {
+      "epoch": 1.8634552171510022,
+      "grad_norm": 13.125,
+      "learning_rate": 3.788483711536894e-06,
+      "loss": 1.4654,
+      "step": 64060
+    },
+    {
+      "epoch": 1.86403700148355,
+      "grad_norm": 13.625,
+      "learning_rate": 3.7865444298642783e-06,
+      "loss": 1.5492,
+      "step": 64080
+    },
+    {
+      "epoch": 1.864618785816098,
+      "grad_norm": 13.5625,
+      "learning_rate": 3.7846051481916634e-06,
+      "loss": 1.5703,
+      "step": 64100
+    },
+    {
+      "epoch": 1.8652005701486458,
+      "grad_norm": 12.4375,
+      "learning_rate": 3.7826658665190485e-06,
+      "loss": 1.4844,
+      "step": 64120
+    },
+    {
+      "epoch": 1.8657823544811938,
+      "grad_norm": 13.75,
+      "learning_rate": 3.7807265848464336e-06,
+      "loss": 1.5887,
+      "step": 64140
+    },
+    {
+      "epoch": 1.8663641388137417,
+      "grad_norm": 15.25,
+      "learning_rate": 3.7787873031738187e-06,
+      "loss": 1.5731,
+      "step": 64160
+    },
+    {
+      "epoch": 1.8669459231462897,
+      "grad_norm": 15.875,
+      "learning_rate": 3.776848021501204e-06,
+      "loss": 1.5551,
+      "step": 64180
+    },
+    {
+      "epoch": 1.8675277074788377,
+      "grad_norm": 14.125,
+      "learning_rate": 3.774908739828589e-06,
+      "loss": 1.5833,
+      "step": 64200
+    },
+    {
+      "epoch": 1.8681094918113854,
+      "grad_norm": 13.9375,
+      "learning_rate": 3.772969458155974e-06,
+      "loss": 1.5592,
+      "step": 64220
+    },
+    {
+      "epoch": 1.8686912761439336,
+      "grad_norm": 11.1875,
+      "learning_rate": 3.771030176483359e-06,
+      "loss": 1.4486,
+      "step": 64240
+    },
+    {
+      "epoch": 1.8692730604764813,
+      "grad_norm": 12.4375,
+      "learning_rate": 3.769090894810744e-06,
+      "loss": 1.5575,
+      "step": 64260
+    },
+    {
+      "epoch": 1.8698548448090293,
+      "grad_norm": 14.0625,
+      "learning_rate": 3.7671516131381293e-06,
+      "loss": 1.4613,
+      "step": 64280
+    },
+    {
+      "epoch": 1.8704366291415773,
+      "grad_norm": 13.375,
+      "learning_rate": 3.7652123314655144e-06,
+      "loss": 1.5542,
+      "step": 64300
+    },
+    {
+      "epoch": 1.871018413474125,
+      "grad_norm": 14.4375,
+      "learning_rate": 3.7632730497928995e-06,
+      "loss": 1.449,
+      "step": 64320
+    },
+    {
+      "epoch": 1.8716001978066732,
+      "grad_norm": 14.25,
+      "learning_rate": 3.7613337681202846e-06,
+      "loss": 1.5446,
+      "step": 64340
+    },
+    {
+      "epoch": 1.872181982139221,
+      "grad_norm": 12.0625,
+      "learning_rate": 3.7593944864476693e-06,
+      "loss": 1.5131,
+      "step": 64360
+    },
+    {
+      "epoch": 1.872763766471769,
+      "grad_norm": 13.375,
+      "learning_rate": 3.7574552047750544e-06,
+      "loss": 1.4918,
+      "step": 64380
+    },
+    {
+      "epoch": 1.8733455508043169,
+      "grad_norm": 14.25,
+      "learning_rate": 3.7555159231024395e-06,
+      "loss": 1.5377,
+      "step": 64400
+    },
+    {
+      "epoch": 1.8739273351368646,
+      "grad_norm": 14.125,
+      "learning_rate": 3.7535766414298246e-06,
+      "loss": 1.5263,
+      "step": 64420
+    },
+    {
+      "epoch": 1.8745091194694128,
+      "grad_norm": 12.75,
+      "learning_rate": 3.7516373597572097e-06,
+      "loss": 1.5512,
+      "step": 64440
+    },
+    {
+      "epoch": 1.8750909038019605,
+      "grad_norm": 12.5625,
+      "learning_rate": 3.7496980780845948e-06,
+      "loss": 1.5967,
+      "step": 64460
+    },
+    {
+      "epoch": 1.8756726881345085,
+      "grad_norm": 12.875,
+      "learning_rate": 3.74775879641198e-06,
+      "loss": 1.5346,
+      "step": 64480
+    },
+    {
+      "epoch": 1.8762544724670565,
+      "grad_norm": 12.1875,
+      "learning_rate": 3.745819514739365e-06,
+      "loss": 1.5279,
+      "step": 64500
+    },
+    {
+      "epoch": 1.8768362567996044,
+      "grad_norm": 14.3125,
+      "learning_rate": 3.74388023306675e-06,
+      "loss": 1.5636,
+      "step": 64520
+    },
+    {
+      "epoch": 1.8774180411321524,
+      "grad_norm": 13.625,
+      "learning_rate": 3.741940951394135e-06,
+      "loss": 1.5338,
+      "step": 64540
+    },
+    {
+      "epoch": 1.8779998254647001,
+      "grad_norm": 12.1875,
+      "learning_rate": 3.7400016697215202e-06,
+      "loss": 1.5769,
+      "step": 64560
+    },
+    {
+      "epoch": 1.8785816097972483,
+      "grad_norm": 13.5,
+      "learning_rate": 3.7380623880489053e-06,
+      "loss": 1.5317,
+      "step": 64580
+    },
+    {
+      "epoch": 1.879163394129796,
+      "grad_norm": 14.0625,
+      "learning_rate": 3.7361231063762904e-06,
+      "loss": 1.4931,
+      "step": 64600
+    },
+    {
+      "epoch": 1.879745178462344,
+      "grad_norm": 12.5625,
+      "learning_rate": 3.7341838247036755e-06,
+      "loss": 1.5744,
+      "step": 64620
+    },
+    {
+      "epoch": 1.880326962794892,
+      "grad_norm": 11.1875,
+      "learning_rate": 3.7322445430310606e-06,
+      "loss": 1.5326,
+      "step": 64640
+    },
+    {
+      "epoch": 1.8809087471274397,
+      "grad_norm": 12.5,
+      "learning_rate": 3.7303052613584457e-06,
+      "loss": 1.5492,
+      "step": 64660
+    },
+    {
+      "epoch": 1.881490531459988,
+      "grad_norm": 9.9375,
+      "learning_rate": 3.728365979685831e-06,
+      "loss": 1.4694,
+      "step": 64680
+    },
+    {
+      "epoch": 1.8820723157925356,
+      "grad_norm": 14.0,
+      "learning_rate": 3.726426698013216e-06,
+      "loss": 1.541,
+      "step": 64700
+    },
+    {
+      "epoch": 1.8826541001250836,
+      "grad_norm": 13.9375,
+      "learning_rate": 3.724487416340601e-06,
+      "loss": 1.5134,
+      "step": 64720
+    },
+    {
+      "epoch": 1.8832358844576316,
+      "grad_norm": 15.1875,
+      "learning_rate": 3.722548134667986e-06,
+      "loss": 1.5849,
+      "step": 64740
+    },
+    {
+      "epoch": 1.8838176687901795,
+      "grad_norm": 14.0,
+      "learning_rate": 3.720608852995371e-06,
+      "loss": 1.4249,
+      "step": 64760
+    },
+    {
+      "epoch": 1.8843994531227275,
+      "grad_norm": 13.0,
+      "learning_rate": 3.718669571322756e-06,
+      "loss": 1.5384,
+      "step": 64780
+    },
+    {
+      "epoch": 1.8849812374552752,
+      "grad_norm": 13.6875,
+      "learning_rate": 3.716730289650141e-06,
+      "loss": 1.5399,
+      "step": 64800
+    },
+    {
+      "epoch": 1.8855630217878232,
+      "grad_norm": 13.0,
+      "learning_rate": 3.714791007977526e-06,
+      "loss": 1.5114,
+      "step": 64820
+    },
+    {
+      "epoch": 1.8861448061203712,
+      "grad_norm": 11.4375,
+      "learning_rate": 3.712851726304911e-06,
+      "loss": 1.522,
+      "step": 64840
+    },
+    {
+      "epoch": 1.8867265904529191,
+      "grad_norm": 16.125,
+      "learning_rate": 3.7109124446322963e-06,
+      "loss": 1.4984,
+      "step": 64860
+    },
+    {
+      "epoch": 1.887308374785467,
+      "grad_norm": 14.5625,
+      "learning_rate": 3.7089731629596814e-06,
+      "loss": 1.4274,
+      "step": 64880
+    },
+    {
+      "epoch": 1.8878901591180148,
+      "grad_norm": 11.8125,
+      "learning_rate": 3.7070338812870665e-06,
+      "loss": 1.556,
+      "step": 64900
+    },
+    {
+      "epoch": 1.888471943450563,
+      "grad_norm": 13.0625,
+      "learning_rate": 3.7050945996144516e-06,
+      "loss": 1.5648,
+      "step": 64920
+    },
+    {
+      "epoch": 1.8890537277831108,
+      "grad_norm": 15.25,
+      "learning_rate": 3.7031553179418367e-06,
+      "loss": 1.6345,
+      "step": 64940
+    },
+    {
+      "epoch": 1.8896355121156587,
+      "grad_norm": 13.25,
+      "learning_rate": 3.7012160362692218e-06,
+      "loss": 1.5228,
+      "step": 64960
+    },
+    {
+      "epoch": 1.8902172964482067,
+      "grad_norm": 14.4375,
+      "learning_rate": 3.699276754596607e-06,
+      "loss": 1.4963,
+      "step": 64980
+    },
+    {
+      "epoch": 1.8907990807807544,
+      "grad_norm": 13.6875,
+      "learning_rate": 3.697337472923992e-06,
+      "loss": 1.5216,
+      "step": 65000
+    },
+    {
+      "epoch": 1.8913808651133026,
+      "grad_norm": 14.5625,
+      "learning_rate": 3.695398191251377e-06,
+      "loss": 1.4629,
+      "step": 65020
+    },
+    {
+      "epoch": 1.8919626494458504,
+      "grad_norm": 14.8125,
+      "learning_rate": 3.693458909578762e-06,
+      "loss": 1.4844,
+      "step": 65040
+    },
+    {
+      "epoch": 1.8925444337783983,
+      "grad_norm": 11.75,
+      "learning_rate": 3.6915196279061473e-06,
+      "loss": 1.6662,
+      "step": 65060
+    },
+    {
+      "epoch": 1.8931262181109463,
+      "grad_norm": 16.75,
+      "learning_rate": 3.6895803462335323e-06,
+      "loss": 1.5558,
+      "step": 65080
+    },
+    {
+      "epoch": 1.8937080024434942,
+      "grad_norm": 14.8125,
+      "learning_rate": 3.6876410645609174e-06,
+      "loss": 1.5334,
+      "step": 65100
+    },
+    {
+      "epoch": 1.8942897867760422,
+      "grad_norm": 13.0,
+      "learning_rate": 3.6857017828883025e-06,
+      "loss": 1.5759,
+      "step": 65120
+    },
+    {
+      "epoch": 1.89487157110859,
+      "grad_norm": 10.4375,
+      "learning_rate": 3.6837625012156876e-06,
+      "loss": 1.5059,
+      "step": 65140
+    },
+    {
+      "epoch": 1.8954533554411381,
+      "grad_norm": 11.8125,
+      "learning_rate": 3.6818232195430727e-06,
+      "loss": 1.5883,
+      "step": 65160
+    },
+    {
+      "epoch": 1.8960351397736859,
+      "grad_norm": 12.375,
+      "learning_rate": 3.6798839378704574e-06,
+      "loss": 1.5496,
+      "step": 65180
+    },
+    {
+      "epoch": 1.8966169241062338,
+      "grad_norm": 13.125,
+      "learning_rate": 3.6779446561978425e-06,
+      "loss": 1.4951,
+      "step": 65200
+    },
+    {
+      "epoch": 1.8971987084387818,
+      "grad_norm": 13.9375,
+      "learning_rate": 3.6760053745252276e-06,
+      "loss": 1.4773,
+      "step": 65220
+    },
+    {
+      "epoch": 1.8977804927713295,
+      "grad_norm": 14.25,
+      "learning_rate": 3.6740660928526127e-06,
+      "loss": 1.5271,
+      "step": 65240
+    },
+    {
+      "epoch": 1.8983622771038777,
+      "grad_norm": 12.6875,
+      "learning_rate": 3.672126811179998e-06,
+      "loss": 1.5227,
+      "step": 65260
+    },
+    {
+      "epoch": 1.8989440614364255,
+      "grad_norm": 16.5,
+      "learning_rate": 3.670187529507383e-06,
+      "loss": 1.567,
+      "step": 65280
+    },
+    {
+      "epoch": 1.8995258457689734,
+      "grad_norm": 14.625,
+      "learning_rate": 3.668248247834768e-06,
+      "loss": 1.5158,
+      "step": 65300
+    },
+    {
+      "epoch": 1.9001076301015214,
+      "grad_norm": 13.4375,
+      "learning_rate": 3.666308966162153e-06,
+      "loss": 1.5731,
+      "step": 65320
+    },
+    {
+      "epoch": 1.9006894144340691,
+      "grad_norm": 12.625,
+      "learning_rate": 3.664369684489538e-06,
+      "loss": 1.633,
+      "step": 65340
+    },
+    {
+      "epoch": 1.9012711987666173,
+      "grad_norm": 14.0625,
+      "learning_rate": 3.6624304028169233e-06,
+      "loss": 1.466,
+      "step": 65360
+    },
+    {
+      "epoch": 1.901852983099165,
+      "grad_norm": 13.6875,
+      "learning_rate": 3.6604911211443084e-06,
+      "loss": 1.5367,
+      "step": 65380
+    },
+    {
+      "epoch": 1.902434767431713,
+      "grad_norm": 11.125,
+      "learning_rate": 3.6585518394716935e-06,
+      "loss": 1.5736,
+      "step": 65400
+    },
+    {
+      "epoch": 1.903016551764261,
+      "grad_norm": 17.75,
+      "learning_rate": 3.6566125577990786e-06,
+      "loss": 1.5669,
+      "step": 65420
+    },
+    {
+      "epoch": 1.903598336096809,
+      "grad_norm": 14.6875,
+      "learning_rate": 3.6546732761264637e-06,
+      "loss": 1.5658,
+      "step": 65440
+    },
+    {
+      "epoch": 1.904180120429357,
+      "grad_norm": 12.6875,
+      "learning_rate": 3.6527339944538488e-06,
+      "loss": 1.5102,
+      "step": 65460
+    },
+    {
+      "epoch": 1.9047619047619047,
+      "grad_norm": 12.0,
+      "learning_rate": 3.650794712781234e-06,
+      "loss": 1.4242,
+      "step": 65480
+    },
+    {
+      "epoch": 1.9053436890944528,
+      "grad_norm": 14.8125,
+      "learning_rate": 3.648855431108619e-06,
+      "loss": 1.58,
+      "step": 65500
+    },
+    {
+      "epoch": 1.9059254734270006,
+      "grad_norm": 15.875,
+      "learning_rate": 3.646916149436004e-06,
+      "loss": 1.4414,
+      "step": 65520
+    },
+    {
+      "epoch": 1.9065072577595485,
+      "grad_norm": 12.1875,
+      "learning_rate": 3.644976867763389e-06,
+      "loss": 1.5314,
+      "step": 65540
+    },
+    {
+      "epoch": 1.9070890420920965,
+      "grad_norm": 14.625,
+      "learning_rate": 3.6430375860907743e-06,
+      "loss": 1.5462,
+      "step": 65560
+    },
+    {
+      "epoch": 1.9076708264246443,
+      "grad_norm": 12.75,
+      "learning_rate": 3.641098304418159e-06,
+      "loss": 1.4362,
+      "step": 65580
+    },
+    {
+      "epoch": 1.9082526107571924,
+      "grad_norm": 15.4375,
+      "learning_rate": 3.639159022745544e-06,
+      "loss": 1.6175,
+      "step": 65600
+    },
+    {
+      "epoch": 1.9088343950897402,
+      "grad_norm": 12.8125,
+      "learning_rate": 3.637219741072929e-06,
+      "loss": 1.4896,
+      "step": 65620
+    },
+    {
+      "epoch": 1.9094161794222881,
+      "grad_norm": 14.25,
+      "learning_rate": 3.6352804594003142e-06,
+      "loss": 1.5441,
+      "step": 65640
+    },
+    {
+      "epoch": 1.909997963754836,
+      "grad_norm": 12.625,
+      "learning_rate": 3.6333411777276993e-06,
+      "loss": 1.5906,
+      "step": 65660
+    },
+    {
+      "epoch": 1.9105797480873838,
+      "grad_norm": 13.6875,
+      "learning_rate": 3.6314018960550844e-06,
+      "loss": 1.5146,
+      "step": 65680
+    },
+    {
+      "epoch": 1.911161532419932,
+      "grad_norm": 14.0625,
+      "learning_rate": 3.6294626143824695e-06,
+      "loss": 1.4349,
+      "step": 65700
+    },
+    {
+      "epoch": 1.9117433167524798,
+      "grad_norm": 15.5625,
+      "learning_rate": 3.6275233327098546e-06,
+      "loss": 1.5319,
+      "step": 65720
+    },
+    {
+      "epoch": 1.9123251010850277,
+      "grad_norm": 14.5,
+      "learning_rate": 3.6255840510372397e-06,
+      "loss": 1.4693,
+      "step": 65740
+    },
+    {
+      "epoch": 1.9129068854175757,
+      "grad_norm": 15.5625,
+      "learning_rate": 3.623644769364625e-06,
+      "loss": 1.6036,
+      "step": 65760
+    },
+    {
+      "epoch": 1.9134886697501237,
+      "grad_norm": 13.8125,
+      "learning_rate": 3.62170548769201e-06,
+      "loss": 1.5009,
+      "step": 65780
+    },
+    {
+      "epoch": 1.9140704540826716,
+      "grad_norm": 15.3125,
+      "learning_rate": 3.619766206019395e-06,
+      "loss": 1.5315,
+      "step": 65800
+    },
+    {
+      "epoch": 1.9146522384152194,
+      "grad_norm": 11.875,
+      "learning_rate": 3.61782692434678e-06,
+      "loss": 1.6201,
+      "step": 65820
+    },
+    {
+      "epoch": 1.9152340227477676,
+      "grad_norm": 8.375,
+      "learning_rate": 3.615887642674165e-06,
+      "loss": 1.5182,
+      "step": 65840
+    },
+    {
+      "epoch": 1.9158158070803153,
+      "grad_norm": 13.9375,
+      "learning_rate": 3.6139483610015503e-06,
+      "loss": 1.5305,
+      "step": 65860
+    },
+    {
+      "epoch": 1.9163975914128633,
+      "grad_norm": 11.25,
+      "learning_rate": 3.6120090793289354e-06,
+      "loss": 1.4922,
+      "step": 65880
+    },
+    {
+      "epoch": 1.9169793757454112,
+      "grad_norm": 13.375,
+      "learning_rate": 3.6100697976563205e-06,
+      "loss": 1.5899,
+      "step": 65900
+    },
+    {
+      "epoch": 1.917561160077959,
+      "grad_norm": 14.4375,
+      "learning_rate": 3.6081305159837056e-06,
+      "loss": 1.4647,
+      "step": 65920
+    },
+    {
+      "epoch": 1.9181429444105071,
+      "grad_norm": 12.25,
+      "learning_rate": 3.6061912343110907e-06,
+      "loss": 1.6362,
+      "step": 65940
+    },
+    {
+      "epoch": 1.9187247287430549,
+      "grad_norm": 13.9375,
+      "learning_rate": 3.604251952638476e-06,
+      "loss": 1.5538,
+      "step": 65960
+    },
+    {
+      "epoch": 1.9193065130756028,
+      "grad_norm": 13.0,
+      "learning_rate": 3.602312670965861e-06,
+      "loss": 1.484,
+      "step": 65980
+    },
+    {
+      "epoch": 1.9198882974081508,
+      "grad_norm": 17.125,
+      "learning_rate": 3.6003733892932456e-06,
+      "loss": 1.4788,
+      "step": 66000
+    },
+    {
+      "epoch": 1.9204700817406988,
+      "grad_norm": 14.375,
+      "learning_rate": 3.5984341076206307e-06,
+      "loss": 1.5489,
+      "step": 66020
+    },
+    {
+      "epoch": 1.9210518660732467,
+      "grad_norm": 12.875,
+      "learning_rate": 3.5964948259480158e-06,
+      "loss": 1.461,
+      "step": 66040
+    },
+    {
+      "epoch": 1.9216336504057945,
+      "grad_norm": 13.25,
+      "learning_rate": 3.594555544275401e-06,
+      "loss": 1.5596,
+      "step": 66060
+    },
+    {
+      "epoch": 1.9222154347383424,
+      "grad_norm": 12.0625,
+      "learning_rate": 3.592616262602786e-06,
+      "loss": 1.5128,
+      "step": 66080
+    },
+    {
+      "epoch": 1.9227972190708904,
+      "grad_norm": 18.25,
+      "learning_rate": 3.590676980930171e-06,
+      "loss": 1.4885,
+      "step": 66100
+    },
+    {
+      "epoch": 1.9233790034034384,
+      "grad_norm": 12.375,
+      "learning_rate": 3.588737699257556e-06,
+      "loss": 1.496,
+      "step": 66120
+    },
+    {
+      "epoch": 1.9239607877359863,
+      "grad_norm": 15.25,
+      "learning_rate": 3.5867984175849412e-06,
+      "loss": 1.5943,
+      "step": 66140
+    },
+    {
+      "epoch": 1.924542572068534,
+      "grad_norm": 13.75,
+      "learning_rate": 3.5848591359123263e-06,
+      "loss": 1.5122,
+      "step": 66160
+    },
+    {
+      "epoch": 1.9251243564010823,
+      "grad_norm": 11.6875,
+      "learning_rate": 3.5829198542397114e-06,
+      "loss": 1.5545,
+      "step": 66180
+    },
+    {
+      "epoch": 1.92570614073363,
+      "grad_norm": 12.0,
+      "learning_rate": 3.580980572567096e-06,
+      "loss": 1.5628,
+      "step": 66200
+    },
+    {
+      "epoch": 1.926287925066178,
+      "grad_norm": 12.9375,
+      "learning_rate": 3.579041290894481e-06,
+      "loss": 1.4236,
+      "step": 66220
+    },
+    {
+      "epoch": 1.926869709398726,
+      "grad_norm": 15.3125,
+      "learning_rate": 3.5771020092218663e-06,
+      "loss": 1.5351,
+      "step": 66240
+    },
+    {
+      "epoch": 1.9274514937312737,
+      "grad_norm": 11.0625,
+      "learning_rate": 3.5751627275492514e-06,
+      "loss": 1.5499,
+      "step": 66260
+    },
+    {
+      "epoch": 1.9280332780638219,
+      "grad_norm": 13.1875,
+      "learning_rate": 3.573223445876636e-06,
+      "loss": 1.5883,
+      "step": 66280
+    },
+    {
+      "epoch": 1.9286150623963696,
+      "grad_norm": 12.0625,
+      "learning_rate": 3.571284164204021e-06,
+      "loss": 1.4578,
+      "step": 66300
+    },
+    {
+      "epoch": 1.9291968467289176,
+      "grad_norm": 11.5,
+      "learning_rate": 3.5693448825314063e-06,
+      "loss": 1.4527,
+      "step": 66320
+    },
+    {
+      "epoch": 1.9297786310614655,
+      "grad_norm": 13.8125,
+      "learning_rate": 3.5674056008587914e-06,
+      "loss": 1.5393,
+      "step": 66340
+    },
+    {
+      "epoch": 1.9303604153940135,
+      "grad_norm": 11.3125,
+      "learning_rate": 3.5654663191861765e-06,
+      "loss": 1.5419,
+      "step": 66360
+    },
+    {
+      "epoch": 1.9309421997265614,
+      "grad_norm": 13.875,
+      "learning_rate": 3.5635270375135616e-06,
+      "loss": 1.5344,
+      "step": 66380
+    },
+    {
+      "epoch": 1.9315239840591092,
+      "grad_norm": 12.625,
+      "learning_rate": 3.5615877558409467e-06,
+      "loss": 1.5528,
+      "step": 66400
+    },
+    {
+      "epoch": 1.9321057683916574,
+      "grad_norm": 13.5625,
+      "learning_rate": 3.5596484741683318e-06,
+      "loss": 1.4706,
+      "step": 66420
+    },
+    {
+      "epoch": 1.9326875527242051,
+      "grad_norm": 13.0,
+      "learning_rate": 3.557709192495717e-06,
+      "loss": 1.5307,
+      "step": 66440
+    },
+    {
+      "epoch": 1.933269337056753,
+      "grad_norm": 14.875,
+      "learning_rate": 3.555769910823102e-06,
+      "loss": 1.5119,
+      "step": 66460
+    },
+    {
+      "epoch": 1.933851121389301,
+      "grad_norm": 11.875,
+      "learning_rate": 3.553830629150487e-06,
+      "loss": 1.474,
+      "step": 66480
+    },
+    {
+      "epoch": 1.9344329057218488,
+      "grad_norm": 14.5,
+      "learning_rate": 3.551891347477872e-06,
+      "loss": 1.5309,
+      "step": 66500
+    },
+    {
+      "epoch": 1.935014690054397,
+      "grad_norm": 12.125,
+      "learning_rate": 3.5499520658052573e-06,
+      "loss": 1.509,
+      "step": 66520
+    },
+    {
+      "epoch": 1.9355964743869447,
+      "grad_norm": 14.5625,
+      "learning_rate": 3.5480127841326423e-06,
+      "loss": 1.5758,
+      "step": 66540
+    },
+    {
+      "epoch": 1.9361782587194927,
+      "grad_norm": 15.3125,
+      "learning_rate": 3.5460735024600274e-06,
+      "loss": 1.4817,
+      "step": 66560
+    },
+    {
+      "epoch": 1.9367600430520406,
+      "grad_norm": 15.125,
+      "learning_rate": 3.5441342207874125e-06,
+      "loss": 1.4302,
+      "step": 66580
+    },
+    {
+      "epoch": 1.9373418273845884,
+      "grad_norm": 13.5,
+      "learning_rate": 3.5421949391147976e-06,
+      "loss": 1.5171,
+      "step": 66600
+    },
+    {
+      "epoch": 1.9379236117171366,
+      "grad_norm": 13.5,
+      "learning_rate": 3.5402556574421827e-06,
+      "loss": 1.5666,
+      "step": 66620
+    },
+    {
+      "epoch": 1.9385053960496843,
+      "grad_norm": 13.5,
+      "learning_rate": 3.538316375769568e-06,
+      "loss": 1.4887,
+      "step": 66640
+    },
+    {
+      "epoch": 1.9390871803822323,
+      "grad_norm": 13.0625,
+      "learning_rate": 3.536377094096953e-06,
+      "loss": 1.4681,
+      "step": 66660
+    },
+    {
+      "epoch": 1.9396689647147802,
+      "grad_norm": 12.9375,
+      "learning_rate": 3.5344378124243376e-06,
+      "loss": 1.4435,
+      "step": 66680
+    },
+    {
+      "epoch": 1.9402507490473282,
+      "grad_norm": 15.0625,
+      "learning_rate": 3.5324985307517227e-06,
+      "loss": 1.4985,
+      "step": 66700
+    },
+    {
+      "epoch": 1.9408325333798762,
+      "grad_norm": 12.5,
+      "learning_rate": 3.530559249079108e-06,
+      "loss": 1.4723,
+      "step": 66720
+    },
+    {
+      "epoch": 1.941414317712424,
+      "grad_norm": 11.4375,
+      "learning_rate": 3.528619967406493e-06,
+      "loss": 1.5144,
+      "step": 66740
+    },
+    {
+      "epoch": 1.941996102044972,
+      "grad_norm": 15.3125,
+      "learning_rate": 3.526680685733878e-06,
+      "loss": 1.5583,
+      "step": 66760
+    },
+    {
+      "epoch": 1.9425778863775198,
+      "grad_norm": 15.0,
+      "learning_rate": 3.524741404061263e-06,
+      "loss": 1.5899,
+      "step": 66780
+    },
+    {
+      "epoch": 1.9431596707100678,
+      "grad_norm": 13.5,
+      "learning_rate": 3.522802122388648e-06,
+      "loss": 1.5051,
+      "step": 66800
+    },
+    {
+      "epoch": 1.9437414550426158,
+      "grad_norm": 14.6875,
+      "learning_rate": 3.5208628407160333e-06,
+      "loss": 1.5192,
+      "step": 66820
+    },
+    {
+      "epoch": 1.9443232393751635,
+      "grad_norm": 11.0,
+      "learning_rate": 3.5189235590434184e-06,
+      "loss": 1.5179,
+      "step": 66840
+    },
+    {
+      "epoch": 1.9449050237077117,
+      "grad_norm": 13.4375,
+      "learning_rate": 3.5169842773708035e-06,
+      "loss": 1.4844,
+      "step": 66860
+    },
+    {
+      "epoch": 1.9454868080402594,
+      "grad_norm": 13.4375,
+      "learning_rate": 3.5150449956981886e-06,
+      "loss": 1.4773,
+      "step": 66880
+    },
+    {
+      "epoch": 1.9460685923728074,
+      "grad_norm": 15.4375,
+      "learning_rate": 3.5131057140255737e-06,
+      "loss": 1.4834,
+      "step": 66900
+    },
+    {
+      "epoch": 1.9466503767053553,
+      "grad_norm": 13.4375,
+      "learning_rate": 3.5111664323529588e-06,
+      "loss": 1.4458,
+      "step": 66920
+    },
+    {
+      "epoch": 1.947232161037903,
+      "grad_norm": 13.6875,
+      "learning_rate": 3.509227150680344e-06,
+      "loss": 1.3961,
+      "step": 66940
+    },
+    {
+      "epoch": 1.9478139453704513,
+      "grad_norm": 18.625,
+      "learning_rate": 3.507287869007729e-06,
+      "loss": 1.4644,
+      "step": 66960
+    },
+    {
+      "epoch": 1.948395729702999,
+      "grad_norm": 15.125,
+      "learning_rate": 3.505348587335114e-06,
+      "loss": 1.4645,
+      "step": 66980
+    },
+    {
+      "epoch": 1.948977514035547,
+      "grad_norm": 12.25,
+      "learning_rate": 3.503409305662499e-06,
+      "loss": 1.5278,
+      "step": 67000
+    },
+    {
+      "epoch": 1.949559298368095,
+      "grad_norm": 13.5625,
+      "learning_rate": 3.5014700239898843e-06,
+      "loss": 1.5145,
+      "step": 67020
+    },
+    {
+      "epoch": 1.950141082700643,
+      "grad_norm": 14.25,
+      "learning_rate": 3.4995307423172694e-06,
+      "loss": 1.4483,
+      "step": 67040
+    },
+    {
+      "epoch": 1.9507228670331909,
+      "grad_norm": 14.8125,
+      "learning_rate": 3.4975914606446545e-06,
+      "loss": 1.5122,
+      "step": 67060
+    },
+    {
+      "epoch": 1.9513046513657386,
+      "grad_norm": 12.3125,
+      "learning_rate": 3.4956521789720396e-06,
+      "loss": 1.4802,
+      "step": 67080
+    },
+    {
+      "epoch": 1.9518864356982868,
+      "grad_norm": 12.625,
+      "learning_rate": 3.4937128972994242e-06,
+      "loss": 1.5416,
+      "step": 67100
+    },
+    {
+      "epoch": 1.9524682200308345,
+      "grad_norm": 12.875,
+      "learning_rate": 3.4917736156268093e-06,
+      "loss": 1.5825,
+      "step": 67120
+    },
+    {
+      "epoch": 1.9530500043633825,
+      "grad_norm": 11.1875,
+      "learning_rate": 3.4898343339541944e-06,
+      "loss": 1.4703,
+      "step": 67140
+    },
+    {
+      "epoch": 1.9536317886959305,
+      "grad_norm": 14.3125,
+      "learning_rate": 3.4878950522815795e-06,
+      "loss": 1.5021,
+      "step": 67160
+    },
+    {
+      "epoch": 1.9542135730284782,
+      "grad_norm": 13.125,
+      "learning_rate": 3.4859557706089646e-06,
+      "loss": 1.5426,
+      "step": 67180
+    },
+    {
+      "epoch": 1.9547953573610264,
+      "grad_norm": 13.125,
+      "learning_rate": 3.4840164889363497e-06,
+      "loss": 1.6451,
+      "step": 67200
+    },
+    {
+      "epoch": 1.9553771416935741,
+      "grad_norm": 11.9375,
+      "learning_rate": 3.482077207263735e-06,
+      "loss": 1.4643,
+      "step": 67220
+    },
+    {
+      "epoch": 1.955958926026122,
+      "grad_norm": 11.875,
+      "learning_rate": 3.48013792559112e-06,
+      "loss": 1.4222,
+      "step": 67240
+    },
+    {
+      "epoch": 1.95654071035867,
+      "grad_norm": 13.4375,
+      "learning_rate": 3.478198643918505e-06,
+      "loss": 1.5199,
+      "step": 67260
+    },
+    {
+      "epoch": 1.957122494691218,
+      "grad_norm": 14.75,
+      "learning_rate": 3.47625936224589e-06,
+      "loss": 1.52,
+      "step": 67280
+    },
+    {
+      "epoch": 1.957704279023766,
+      "grad_norm": 12.8125,
+      "learning_rate": 3.474320080573275e-06,
+      "loss": 1.5264,
+      "step": 67300
+    },
+    {
+      "epoch": 1.9582860633563137,
+      "grad_norm": 15.6875,
+      "learning_rate": 3.4723807989006603e-06,
+      "loss": 1.4517,
+      "step": 67320
+    },
+    {
+      "epoch": 1.9588678476888617,
+      "grad_norm": 13.3125,
+      "learning_rate": 3.4704415172280454e-06,
+      "loss": 1.4373,
+      "step": 67340
+    },
+    {
+      "epoch": 1.9594496320214096,
+      "grad_norm": 11.6875,
+      "learning_rate": 3.4685022355554305e-06,
+      "loss": 1.4871,
+      "step": 67360
+    },
+    {
+      "epoch": 1.9600314163539576,
+      "grad_norm": 13.625,
+      "learning_rate": 3.4665629538828156e-06,
+      "loss": 1.456,
+      "step": 67380
+    },
+    {
+      "epoch": 1.9606132006865056,
+      "grad_norm": 17.25,
+      "learning_rate": 3.4646236722102007e-06,
+      "loss": 1.6015,
+      "step": 67400
+    },
+    {
+      "epoch": 1.9611949850190533,
+      "grad_norm": 13.5,
+      "learning_rate": 3.462684390537586e-06,
+      "loss": 1.439,
+      "step": 67420
+    },
+    {
+      "epoch": 1.9617767693516015,
+      "grad_norm": 17.125,
+      "learning_rate": 3.460745108864971e-06,
+      "loss": 1.4945,
+      "step": 67440
+    },
+    {
+      "epoch": 1.9623585536841492,
+      "grad_norm": 14.0625,
+      "learning_rate": 3.458805827192356e-06,
+      "loss": 1.4674,
+      "step": 67460
+    },
+    {
+      "epoch": 1.9629403380166972,
+      "grad_norm": 12.3125,
+      "learning_rate": 3.456866545519741e-06,
+      "loss": 1.5322,
+      "step": 67480
+    },
+    {
+      "epoch": 1.9635221223492452,
+      "grad_norm": 15.0625,
+      "learning_rate": 3.4549272638471258e-06,
+      "loss": 1.472,
+      "step": 67500
+    },
+    {
+      "epoch": 1.964103906681793,
+      "grad_norm": 11.125,
+      "learning_rate": 3.452987982174511e-06,
+      "loss": 1.5724,
+      "step": 67520
+    },
+    {
+      "epoch": 1.964685691014341,
+      "grad_norm": 14.25,
+      "learning_rate": 3.451048700501896e-06,
+      "loss": 1.6187,
+      "step": 67540
+    },
+    {
+      "epoch": 1.9652674753468888,
+      "grad_norm": 12.625,
+      "learning_rate": 3.449109418829281e-06,
+      "loss": 1.4714,
+      "step": 67560
+    },
+    {
+      "epoch": 1.9658492596794368,
+      "grad_norm": 12.25,
+      "learning_rate": 3.447170137156666e-06,
+      "loss": 1.4059,
+      "step": 67580
+    },
+    {
+      "epoch": 1.9664310440119848,
+      "grad_norm": 13.0625,
+      "learning_rate": 3.4452308554840512e-06,
+      "loss": 1.5153,
+      "step": 67600
+    },
+    {
+      "epoch": 1.9670128283445327,
+      "grad_norm": 14.125,
+      "learning_rate": 3.4432915738114363e-06,
+      "loss": 1.4906,
+      "step": 67620
+    },
+    {
+      "epoch": 1.9675946126770807,
+      "grad_norm": 14.375,
+      "learning_rate": 3.4413522921388214e-06,
+      "loss": 1.5138,
+      "step": 67640
+    },
+    {
+      "epoch": 1.9681763970096284,
+      "grad_norm": 12.9375,
+      "learning_rate": 3.4394130104662065e-06,
+      "loss": 1.4947,
+      "step": 67660
+    },
+    {
+      "epoch": 1.9687581813421766,
+      "grad_norm": 13.75,
+      "learning_rate": 3.4374737287935916e-06,
+      "loss": 1.502,
+      "step": 67680
+    },
+    {
+      "epoch": 1.9693399656747244,
+      "grad_norm": 14.6875,
+      "learning_rate": 3.4355344471209767e-06,
+      "loss": 1.6162,
+      "step": 67700
+    },
+    {
+      "epoch": 1.9699217500072723,
+      "grad_norm": 14.9375,
+      "learning_rate": 3.433595165448362e-06,
+      "loss": 1.5033,
+      "step": 67720
+    },
+    {
+      "epoch": 1.9705035343398203,
+      "grad_norm": 13.8125,
+      "learning_rate": 3.431655883775747e-06,
+      "loss": 1.516,
+      "step": 67740
+    },
+    {
+      "epoch": 1.971085318672368,
+      "grad_norm": 13.125,
+      "learning_rate": 3.429716602103132e-06,
+      "loss": 1.5823,
+      "step": 67760
+    },
+    {
+      "epoch": 1.9716671030049162,
+      "grad_norm": 15.4375,
+      "learning_rate": 3.427777320430517e-06,
+      "loss": 1.5057,
+      "step": 67780
+    },
+    {
+      "epoch": 1.972248887337464,
+      "grad_norm": 13.75,
+      "learning_rate": 3.4258380387579022e-06,
+      "loss": 1.4956,
+      "step": 67800
+    },
+    {
+      "epoch": 1.972830671670012,
+      "grad_norm": 16.375,
+      "learning_rate": 3.4238987570852873e-06,
+      "loss": 1.4952,
+      "step": 67820
+    },
+    {
+      "epoch": 1.9734124560025599,
+      "grad_norm": 14.75,
+      "learning_rate": 3.4219594754126724e-06,
+      "loss": 1.5201,
+      "step": 67840
+    },
+    {
+      "epoch": 1.9739942403351076,
+      "grad_norm": 13.4375,
+      "learning_rate": 3.4200201937400575e-06,
+      "loss": 1.576,
+      "step": 67860
+    },
+    {
+      "epoch": 1.9745760246676558,
+      "grad_norm": 12.5625,
+      "learning_rate": 3.4180809120674426e-06,
+      "loss": 1.5281,
+      "step": 67880
+    },
+    {
+      "epoch": 1.9751578090002035,
+      "grad_norm": 12.1875,
+      "learning_rate": 3.4161416303948273e-06,
+      "loss": 1.4996,
+      "step": 67900
+    },
+    {
+      "epoch": 1.9757395933327515,
+      "grad_norm": 11.8125,
+      "learning_rate": 3.4142023487222124e-06,
+      "loss": 1.5158,
+      "step": 67920
+    },
+    {
+      "epoch": 1.9763213776652995,
+      "grad_norm": 14.3125,
+      "learning_rate": 3.4122630670495975e-06,
+      "loss": 1.4658,
+      "step": 67940
+    },
+    {
+      "epoch": 1.9769031619978474,
+      "grad_norm": 13.4375,
+      "learning_rate": 3.4103237853769826e-06,
+      "loss": 1.46,
+      "step": 67960
+    },
+    {
+      "epoch": 1.9774849463303954,
+      "grad_norm": 17.875,
+      "learning_rate": 3.4083845037043677e-06,
+      "loss": 1.5656,
+      "step": 67980
+    },
+    {
+      "epoch": 1.9780667306629431,
+      "grad_norm": 12.4375,
+      "learning_rate": 3.4064452220317528e-06,
+      "loss": 1.4976,
+      "step": 68000
+    },
+    {
+      "epoch": 1.9786485149954913,
+      "grad_norm": 14.1875,
+      "learning_rate": 3.404505940359138e-06,
+      "loss": 1.4997,
+      "step": 68020
+    },
+    {
+      "epoch": 1.979230299328039,
+      "grad_norm": 17.0,
+      "learning_rate": 3.402566658686523e-06,
+      "loss": 1.4584,
+      "step": 68040
+    },
+    {
+      "epoch": 1.979812083660587,
+      "grad_norm": 14.625,
+      "learning_rate": 3.400627377013908e-06,
+      "loss": 1.5321,
+      "step": 68060
+    },
+    {
+      "epoch": 1.980393867993135,
+      "grad_norm": 12.1875,
+      "learning_rate": 3.398688095341293e-06,
+      "loss": 1.3941,
+      "step": 68080
+    },
+    {
+      "epoch": 1.9809756523256827,
+      "grad_norm": 12.1875,
+      "learning_rate": 3.3967488136686783e-06,
+      "loss": 1.5388,
+      "step": 68100
+    },
+    {
+      "epoch": 1.981557436658231,
+      "grad_norm": 14.0,
+      "learning_rate": 3.3948095319960634e-06,
+      "loss": 1.5299,
+      "step": 68120
+    },
+    {
+      "epoch": 1.9821392209907787,
+      "grad_norm": 12.6875,
+      "learning_rate": 3.3928702503234485e-06,
+      "loss": 1.4172,
+      "step": 68140
+    },
+    {
+      "epoch": 1.9827210053233266,
+      "grad_norm": 15.5625,
+      "learning_rate": 3.3909309686508335e-06,
+      "loss": 1.4967,
+      "step": 68160
+    },
+    {
+      "epoch": 1.9833027896558746,
+      "grad_norm": 14.4375,
+      "learning_rate": 3.3889916869782186e-06,
+      "loss": 1.4214,
+      "step": 68180
+    },
+    {
+      "epoch": 1.9838845739884223,
+      "grad_norm": 12.75,
+      "learning_rate": 3.3870524053056037e-06,
+      "loss": 1.5038,
+      "step": 68200
+    },
+    {
+      "epoch": 1.9844663583209705,
+      "grad_norm": 13.75,
+      "learning_rate": 3.385113123632989e-06,
+      "loss": 1.568,
+      "step": 68220
+    },
+    {
+      "epoch": 1.9850481426535183,
+      "grad_norm": 14.3125,
+      "learning_rate": 3.383173841960374e-06,
+      "loss": 1.5062,
+      "step": 68240
+    },
+    {
+      "epoch": 1.9856299269860662,
+      "grad_norm": 11.6875,
+      "learning_rate": 3.381234560287759e-06,
+      "loss": 1.4475,
+      "step": 68260
+    },
+    {
+      "epoch": 1.9862117113186142,
+      "grad_norm": 13.0,
+      "learning_rate": 3.379295278615144e-06,
+      "loss": 1.4878,
+      "step": 68280
+    },
+    {
+      "epoch": 1.9867934956511621,
+      "grad_norm": 12.375,
+      "learning_rate": 3.3773559969425292e-06,
+      "loss": 1.5715,
+      "step": 68300
+    },
+    {
+      "epoch": 1.98737527998371,
+      "grad_norm": 15.0,
+      "learning_rate": 3.375416715269914e-06,
+      "loss": 1.4876,
+      "step": 68320
+    },
+    {
+      "epoch": 1.9879570643162578,
+      "grad_norm": 16.5,
+      "learning_rate": 3.3734774335972986e-06,
+      "loss": 1.4497,
+      "step": 68340
+    },
+    {
+      "epoch": 1.988538848648806,
+      "grad_norm": 15.4375,
+      "learning_rate": 3.3715381519246837e-06,
+      "loss": 1.4886,
+      "step": 68360
+    },
+    {
+      "epoch": 1.9891206329813538,
+      "grad_norm": 11.6875,
+      "learning_rate": 3.3695988702520688e-06,
+      "loss": 1.4935,
+      "step": 68380
+    },
+    {
+      "epoch": 1.9897024173139017,
+      "grad_norm": 13.5,
+      "learning_rate": 3.367659588579454e-06,
+      "loss": 1.4687,
+      "step": 68400
+    },
+    {
+      "epoch": 1.9902842016464497,
+      "grad_norm": 14.0,
+      "learning_rate": 3.365720306906839e-06,
+      "loss": 1.5841,
+      "step": 68420
+    },
+    {
+      "epoch": 1.9908659859789974,
+      "grad_norm": 12.125,
+      "learning_rate": 3.363781025234224e-06,
+      "loss": 1.5813,
+      "step": 68440
+    },
+    {
+      "epoch": 1.9914477703115456,
+      "grad_norm": 13.875,
+      "learning_rate": 3.361841743561609e-06,
+      "loss": 1.5345,
+      "step": 68460
+    },
+    {
+      "epoch": 1.9920295546440934,
+      "grad_norm": 13.0,
+      "learning_rate": 3.3599024618889943e-06,
+      "loss": 1.5137,
+      "step": 68480
+    },
+    {
+      "epoch": 1.9926113389766413,
+      "grad_norm": 13.4375,
+      "learning_rate": 3.3579631802163794e-06,
+      "loss": 1.4971,
+      "step": 68500
+    },
+    {
+      "epoch": 1.9931931233091893,
+      "grad_norm": 14.375,
+      "learning_rate": 3.3560238985437645e-06,
+      "loss": 1.4585,
+      "step": 68520
+    },
+    {
+      "epoch": 1.9937749076417373,
+      "grad_norm": 9.375,
+      "learning_rate": 3.3540846168711496e-06,
+      "loss": 1.4688,
+      "step": 68540
+    },
+    {
+      "epoch": 1.9943566919742852,
+      "grad_norm": 14.6875,
+      "learning_rate": 3.3521453351985347e-06,
+      "loss": 1.5335,
+      "step": 68560
+    },
+    {
+      "epoch": 1.994938476306833,
+      "grad_norm": 13.3125,
+      "learning_rate": 3.3502060535259197e-06,
+      "loss": 1.4753,
+      "step": 68580
+    },
+    {
+      "epoch": 1.995520260639381,
+      "grad_norm": 14.6875,
+      "learning_rate": 3.3482667718533044e-06,
+      "loss": 1.5276,
+      "step": 68600
+    },
+    {
+      "epoch": 1.996102044971929,
+      "grad_norm": 10.625,
+      "learning_rate": 3.3463274901806895e-06,
+      "loss": 1.589,
+      "step": 68620
+    },
+    {
+      "epoch": 1.9966838293044769,
+      "grad_norm": 12.3125,
+      "learning_rate": 3.3443882085080746e-06,
+      "loss": 1.448,
+      "step": 68640
+    },
+    {
+      "epoch": 1.9972656136370248,
+      "grad_norm": 11.3125,
+      "learning_rate": 3.3424489268354597e-06,
+      "loss": 1.5355,
+      "step": 68660
+    },
+    {
+      "epoch": 1.9978473979695726,
+      "grad_norm": 11.4375,
+      "learning_rate": 3.340509645162845e-06,
+      "loss": 1.462,
+      "step": 68680
+    },
+    {
+      "epoch": 1.9984291823021207,
+      "grad_norm": 14.375,
+      "learning_rate": 3.33857036349023e-06,
+      "loss": 1.47,
+      "step": 68700
+    },
+    {
+      "epoch": 1.9990109666346685,
+      "grad_norm": 12.6875,
+      "learning_rate": 3.336631081817615e-06,
+      "loss": 1.4833,
+      "step": 68720
+    },
+    {
+      "epoch": 1.9995927509672164,
+      "grad_norm": 12.6875,
+      "learning_rate": 3.334691800145e-06,
+      "loss": 1.5627,
+      "step": 68740
+    },
+    {
+      "epoch": 2.0001745352997644,
+      "grad_norm": 12.6875,
+      "learning_rate": 3.332752518472385e-06,
+      "loss": 1.4391,
+      "step": 68760
+    },
+    {
+      "epoch": 2.000756319632312,
+      "grad_norm": 16.375,
+      "learning_rate": 3.3308132367997703e-06,
+      "loss": 1.5977,
+      "step": 68780
+    },
+    {
+      "epoch": 2.0013381039648603,
+      "grad_norm": 12.875,
+      "learning_rate": 3.3288739551271554e-06,
+      "loss": 1.4505,
+      "step": 68800
+    },
+    {
+      "epoch": 2.001919888297408,
+      "grad_norm": 12.75,
+      "learning_rate": 3.3269346734545405e-06,
+      "loss": 1.5455,
+      "step": 68820
+    },
+    {
+      "epoch": 2.0025016726299563,
+      "grad_norm": 12.9375,
+      "learning_rate": 3.3249953917819256e-06,
+      "loss": 1.4766,
+      "step": 68840
+    },
+    {
+      "epoch": 2.003083456962504,
+      "grad_norm": 13.8125,
+      "learning_rate": 3.3230561101093107e-06,
+      "loss": 1.6074,
+      "step": 68860
+    },
+    {
+      "epoch": 2.0036652412950517,
+      "grad_norm": 14.3125,
+      "learning_rate": 3.3211168284366958e-06,
+      "loss": 1.4583,
+      "step": 68880
+    },
+    {
+      "epoch": 2.0042470256276,
+      "grad_norm": 14.8125,
+      "learning_rate": 3.319177546764081e-06,
+      "loss": 1.5819,
+      "step": 68900
+    },
+    {
+      "epoch": 2.0048288099601477,
+      "grad_norm": 12.5,
+      "learning_rate": 3.317238265091466e-06,
+      "loss": 1.4428,
+      "step": 68920
+    },
+    {
+      "epoch": 2.005410594292696,
+      "grad_norm": 13.0625,
+      "learning_rate": 3.315298983418851e-06,
+      "loss": 1.513,
+      "step": 68940
+    },
+    {
+      "epoch": 2.0059923786252436,
+      "grad_norm": 12.5,
+      "learning_rate": 3.313359701746236e-06,
+      "loss": 1.4663,
+      "step": 68960
+    },
+    {
+      "epoch": 2.0065741629577913,
+      "grad_norm": 11.375,
+      "learning_rate": 3.3114204200736213e-06,
+      "loss": 1.5026,
+      "step": 68980
+    },
+    {
+      "epoch": 2.0071559472903395,
+      "grad_norm": 14.9375,
+      "learning_rate": 3.309481138401006e-06,
+      "loss": 1.5613,
+      "step": 69000
+    },
+    {
+      "epoch": 2.0077377316228873,
+      "grad_norm": 13.625,
+      "learning_rate": 3.307541856728391e-06,
+      "loss": 1.4829,
+      "step": 69020
+    },
+    {
+      "epoch": 2.0083195159554355,
+      "grad_norm": 12.1875,
+      "learning_rate": 3.305602575055776e-06,
+      "loss": 1.4124,
+      "step": 69040
+    },
+    {
+      "epoch": 2.008901300287983,
+      "grad_norm": 18.625,
+      "learning_rate": 3.3036632933831612e-06,
+      "loss": 1.4377,
+      "step": 69060
+    },
+    {
+      "epoch": 2.0094830846205314,
+      "grad_norm": 14.3125,
+      "learning_rate": 3.3017240117105463e-06,
+      "loss": 1.5308,
+      "step": 69080
+    },
+    {
+      "epoch": 2.010064868953079,
+      "grad_norm": 16.625,
+      "learning_rate": 3.2997847300379314e-06,
+      "loss": 1.4723,
+      "step": 69100
+    },
+    {
+      "epoch": 2.010646653285627,
+      "grad_norm": 17.25,
+      "learning_rate": 3.2978454483653165e-06,
+      "loss": 1.5134,
+      "step": 69120
+    },
+    {
+      "epoch": 2.011228437618175,
+      "grad_norm": 15.125,
+      "learning_rate": 3.2959061666927016e-06,
+      "loss": 1.4609,
+      "step": 69140
+    },
+    {
+      "epoch": 2.011810221950723,
+      "grad_norm": 13.625,
+      "learning_rate": 3.2939668850200867e-06,
+      "loss": 1.5012,
+      "step": 69160
+    },
+    {
+      "epoch": 2.012392006283271,
+      "grad_norm": 16.75,
+      "learning_rate": 3.292027603347472e-06,
+      "loss": 1.5458,
+      "step": 69180
+    },
+    {
+      "epoch": 2.0129737906158187,
+      "grad_norm": 15.0,
+      "learning_rate": 3.290088321674857e-06,
+      "loss": 1.5486,
+      "step": 69200
+    },
+    {
+      "epoch": 2.0135555749483665,
+      "grad_norm": 17.625,
+      "learning_rate": 3.288149040002242e-06,
+      "loss": 1.4424,
+      "step": 69220
+    },
+    {
+      "epoch": 2.0141373592809146,
+      "grad_norm": 14.4375,
+      "learning_rate": 3.286209758329627e-06,
+      "loss": 1.5176,
+      "step": 69240
+    },
+    {
+      "epoch": 2.0147191436134624,
+      "grad_norm": 11.3125,
+      "learning_rate": 3.2842704766570122e-06,
+      "loss": 1.5829,
+      "step": 69260
+    },
+    {
+      "epoch": 2.0153009279460106,
+      "grad_norm": 16.625,
+      "learning_rate": 3.2823311949843973e-06,
+      "loss": 1.4862,
+      "step": 69280
+    },
+    {
+      "epoch": 2.0158827122785583,
+      "grad_norm": 15.4375,
+      "learning_rate": 3.2803919133117824e-06,
+      "loss": 1.4982,
+      "step": 69300
+    },
+    {
+      "epoch": 2.016464496611106,
+      "grad_norm": 11.8125,
+      "learning_rate": 3.2784526316391675e-06,
+      "loss": 1.5102,
+      "step": 69320
+    },
+    {
+      "epoch": 2.0170462809436542,
+      "grad_norm": 13.625,
+      "learning_rate": 3.2765133499665526e-06,
+      "loss": 1.488,
+      "step": 69340
+    },
+    {
+      "epoch": 2.017628065276202,
+      "grad_norm": 14.9375,
+      "learning_rate": 3.2745740682939377e-06,
+      "loss": 1.5736,
+      "step": 69360
+    },
+    {
+      "epoch": 2.01820984960875,
+      "grad_norm": 16.25,
+      "learning_rate": 3.272634786621323e-06,
+      "loss": 1.519,
+      "step": 69380
+    },
+    {
+      "epoch": 2.018791633941298,
+      "grad_norm": 14.25,
+      "learning_rate": 3.270695504948708e-06,
+      "loss": 1.4977,
+      "step": 69400
+    },
+    {
+      "epoch": 2.019373418273846,
+      "grad_norm": 15.25,
+      "learning_rate": 3.2687562232760926e-06,
+      "loss": 1.4406,
+      "step": 69420
+    },
+    {
+      "epoch": 2.019955202606394,
+      "grad_norm": 14.625,
+      "learning_rate": 3.2668169416034777e-06,
+      "loss": 1.5004,
+      "step": 69440
+    },
+    {
+      "epoch": 2.0205369869389416,
+      "grad_norm": 12.6875,
+      "learning_rate": 3.2648776599308628e-06,
+      "loss": 1.4982,
+      "step": 69460
+    },
+    {
+      "epoch": 2.0211187712714898,
+      "grad_norm": 15.3125,
+      "learning_rate": 3.262938378258248e-06,
+      "loss": 1.535,
+      "step": 69480
+    },
+    {
+      "epoch": 2.0217005556040375,
+      "grad_norm": 14.8125,
+      "learning_rate": 3.260999096585633e-06,
+      "loss": 1.5355,
+      "step": 69500
+    },
+    {
+      "epoch": 2.0222823399365857,
+      "grad_norm": 11.3125,
+      "learning_rate": 3.259059814913018e-06,
+      "loss": 1.5343,
+      "step": 69520
+    },
+    {
+      "epoch": 2.0228641242691334,
+      "grad_norm": 12.875,
+      "learning_rate": 3.257120533240403e-06,
+      "loss": 1.503,
+      "step": 69540
+    },
+    {
+      "epoch": 2.023445908601681,
+      "grad_norm": 14.1875,
+      "learning_rate": 3.2551812515677883e-06,
+      "loss": 1.5212,
+      "step": 69560
+    },
+    {
+      "epoch": 2.0240276929342293,
+      "grad_norm": 13.875,
+      "learning_rate": 3.2532419698951734e-06,
+      "loss": 1.4834,
+      "step": 69580
+    },
+    {
+      "epoch": 2.024609477266777,
+      "grad_norm": 12.8125,
+      "learning_rate": 3.2513026882225584e-06,
+      "loss": 1.5213,
+      "step": 69600
+    },
+    {
+      "epoch": 2.0251912615993253,
+      "grad_norm": 13.5,
+      "learning_rate": 3.2493634065499435e-06,
+      "loss": 1.5377,
+      "step": 69620
+    },
+    {
+      "epoch": 2.025773045931873,
+      "grad_norm": 13.5,
+      "learning_rate": 3.2474241248773286e-06,
+      "loss": 1.4808,
+      "step": 69640
+    },
+    {
+      "epoch": 2.026354830264421,
+      "grad_norm": 14.75,
+      "learning_rate": 3.2454848432047137e-06,
+      "loss": 1.4268,
+      "step": 69660
+    },
+    {
+      "epoch": 2.026936614596969,
+      "grad_norm": 13.125,
+      "learning_rate": 3.243545561532099e-06,
+      "loss": 1.5044,
+      "step": 69680
+    },
+    {
+      "epoch": 2.0275183989295167,
+      "grad_norm": 14.4375,
+      "learning_rate": 3.241606279859484e-06,
+      "loss": 1.52,
+      "step": 69700
+    },
+    {
+      "epoch": 2.028100183262065,
+      "grad_norm": 12.8125,
+      "learning_rate": 3.239666998186869e-06,
+      "loss": 1.4395,
+      "step": 69720
+    },
+    {
+      "epoch": 2.0286819675946126,
+      "grad_norm": 12.25,
+      "learning_rate": 3.237727716514254e-06,
+      "loss": 1.3982,
+      "step": 69740
+    },
+    {
+      "epoch": 2.029263751927161,
+      "grad_norm": 13.0,
+      "learning_rate": 3.2357884348416392e-06,
+      "loss": 1.4692,
+      "step": 69760
+    },
+    {
+      "epoch": 2.0298455362597085,
+      "grad_norm": 11.625,
+      "learning_rate": 3.2338491531690243e-06,
+      "loss": 1.412,
+      "step": 69780
+    },
+    {
+      "epoch": 2.0304273205922563,
+      "grad_norm": 13.875,
+      "learning_rate": 3.2319098714964094e-06,
+      "loss": 1.5024,
+      "step": 69800
+    },
+    {
+      "epoch": 2.0310091049248045,
+      "grad_norm": 10.6875,
+      "learning_rate": 3.229970589823794e-06,
+      "loss": 1.5416,
+      "step": 69820
+    },
+    {
+      "epoch": 2.031590889257352,
+      "grad_norm": 13.0,
+      "learning_rate": 3.228031308151179e-06,
+      "loss": 1.4766,
+      "step": 69840
+    },
+    {
+      "epoch": 2.0321726735899004,
+      "grad_norm": 15.4375,
+      "learning_rate": 3.2260920264785643e-06,
+      "loss": 1.4842,
+      "step": 69860
+    },
+    {
+      "epoch": 2.032754457922448,
+      "grad_norm": 12.25,
+      "learning_rate": 3.2241527448059494e-06,
+      "loss": 1.483,
+      "step": 69880
+    },
+    {
+      "epoch": 2.033336242254996,
+      "grad_norm": 10.75,
+      "learning_rate": 3.2222134631333345e-06,
+      "loss": 1.4638,
+      "step": 69900
+    },
+    {
+      "epoch": 2.033918026587544,
+      "grad_norm": 12.3125,
+      "learning_rate": 3.2202741814607196e-06,
+      "loss": 1.5076,
+      "step": 69920
+    },
+    {
+      "epoch": 2.034499810920092,
+      "grad_norm": 12.9375,
+      "learning_rate": 3.2183348997881047e-06,
+      "loss": 1.4741,
+      "step": 69940
+    },
+    {
+      "epoch": 2.03508159525264,
+      "grad_norm": 12.375,
+      "learning_rate": 3.2163956181154898e-06,
+      "loss": 1.4623,
+      "step": 69960
+    },
+    {
+      "epoch": 2.0356633795851877,
+      "grad_norm": 15.25,
+      "learning_rate": 3.214456336442875e-06,
+      "loss": 1.6044,
+      "step": 69980
+    },
+    {
+      "epoch": 2.036245163917736,
+      "grad_norm": 13.4375,
+      "learning_rate": 3.21251705477026e-06,
+      "loss": 1.4792,
+      "step": 70000
+    },
+    {
+      "epoch": 2.0368269482502837,
+      "grad_norm": 13.375,
+      "learning_rate": 3.210577773097645e-06,
+      "loss": 1.5712,
+      "step": 70020
+    },
+    {
+      "epoch": 2.0374087325828314,
+      "grad_norm": 11.25,
+      "learning_rate": 3.20863849142503e-06,
+      "loss": 1.4752,
+      "step": 70040
+    },
+    {
+      "epoch": 2.0379905169153796,
+      "grad_norm": 14.3125,
+      "learning_rate": 3.2066992097524153e-06,
+      "loss": 1.5129,
+      "step": 70060
+    },
+    {
+      "epoch": 2.0385723012479273,
+      "grad_norm": 13.25,
+      "learning_rate": 3.2047599280798004e-06,
+      "loss": 1.5771,
+      "step": 70080
+    },
+    {
+      "epoch": 2.0391540855804755,
+      "grad_norm": 12.4375,
+      "learning_rate": 3.2028206464071855e-06,
+      "loss": 1.4345,
+      "step": 70100
+    },
+    {
+      "epoch": 2.0397358699130232,
+      "grad_norm": 12.625,
+      "learning_rate": 3.2008813647345706e-06,
+      "loss": 1.4438,
+      "step": 70120
+    },
+    {
+      "epoch": 2.040317654245571,
+      "grad_norm": 12.75,
+      "learning_rate": 3.1989420830619557e-06,
+      "loss": 1.5879,
+      "step": 70140
+    },
+    {
+      "epoch": 2.040899438578119,
+      "grad_norm": 14.125,
+      "learning_rate": 3.1970028013893408e-06,
+      "loss": 1.5241,
+      "step": 70160
+    },
+    {
+      "epoch": 2.041481222910667,
+      "grad_norm": 14.3125,
+      "learning_rate": 3.195063519716726e-06,
+      "loss": 1.6103,
+      "step": 70180
+    },
+    {
+      "epoch": 2.042063007243215,
+      "grad_norm": 14.125,
+      "learning_rate": 3.193124238044111e-06,
+      "loss": 1.4937,
+      "step": 70200
+    },
+    {
+      "epoch": 2.042644791575763,
+      "grad_norm": 12.0,
+      "learning_rate": 3.1911849563714956e-06,
+      "loss": 1.47,
+      "step": 70220
+    },
+    {
+      "epoch": 2.0432265759083106,
+      "grad_norm": 13.75,
+      "learning_rate": 3.1892456746988807e-06,
+      "loss": 1.5478,
+      "step": 70240
+    },
+    {
+      "epoch": 2.0438083602408588,
+      "grad_norm": 14.6875,
+      "learning_rate": 3.187306393026266e-06,
+      "loss": 1.4608,
+      "step": 70260
+    },
+    {
+      "epoch": 2.0443901445734065,
+      "grad_norm": 15.8125,
+      "learning_rate": 3.185367111353651e-06,
+      "loss": 1.545,
+      "step": 70280
+    },
+    {
+      "epoch": 2.0449719289059547,
+      "grad_norm": 13.25,
+      "learning_rate": 3.183427829681036e-06,
+      "loss": 1.5346,
+      "step": 70300
+    },
+    {
+      "epoch": 2.0455537132385024,
+      "grad_norm": 13.875,
+      "learning_rate": 3.181488548008421e-06,
+      "loss": 1.4697,
+      "step": 70320
+    },
+    {
+      "epoch": 2.0461354975710506,
+      "grad_norm": 14.5625,
+      "learning_rate": 3.179549266335806e-06,
+      "loss": 1.4163,
+      "step": 70340
+    },
+    {
+      "epoch": 2.0467172819035984,
+      "grad_norm": 15.6875,
+      "learning_rate": 3.1776099846631913e-06,
+      "loss": 1.4977,
+      "step": 70360
+    },
+    {
+      "epoch": 2.047299066236146,
+      "grad_norm": 15.5,
+      "learning_rate": 3.1756707029905764e-06,
+      "loss": 1.6234,
+      "step": 70380
+    },
+    {
+      "epoch": 2.0478808505686943,
+      "grad_norm": 14.375,
+      "learning_rate": 3.1737314213179615e-06,
+      "loss": 1.4976,
+      "step": 70400
+    },
+    {
+      "epoch": 2.048462634901242,
+      "grad_norm": 13.875,
+      "learning_rate": 3.1717921396453466e-06,
+      "loss": 1.5693,
+      "step": 70420
+    },
+    {
+      "epoch": 2.04904441923379,
+      "grad_norm": 12.5625,
+      "learning_rate": 3.1698528579727317e-06,
+      "loss": 1.5112,
+      "step": 70440
+    },
+    {
+      "epoch": 2.049626203566338,
+      "grad_norm": 13.5,
+      "learning_rate": 3.1679135763001164e-06,
+      "loss": 1.4604,
+      "step": 70460
+    },
+    {
+      "epoch": 2.0502079878988857,
+      "grad_norm": 13.625,
+      "learning_rate": 3.1659742946275015e-06,
+      "loss": 1.4615,
+      "step": 70480
+    },
+    {
+      "epoch": 2.050789772231434,
+      "grad_norm": 13.875,
+      "learning_rate": 3.164035012954886e-06,
+      "loss": 1.5743,
+      "step": 70500
+    },
+    {
+      "epoch": 2.0513715565639816,
+      "grad_norm": 13.375,
+      "learning_rate": 3.1620957312822712e-06,
+      "loss": 1.4923,
+      "step": 70520
+    },
+    {
+      "epoch": 2.05195334089653,
+      "grad_norm": 13.0,
+      "learning_rate": 3.1601564496096563e-06,
+      "loss": 1.4632,
+      "step": 70540
+    },
+    {
+      "epoch": 2.0525351252290776,
+      "grad_norm": 14.625,
+      "learning_rate": 3.1582171679370414e-06,
+      "loss": 1.5149,
+      "step": 70560
+    },
+    {
+      "epoch": 2.0531169095616253,
+      "grad_norm": 14.0625,
+      "learning_rate": 3.1562778862644265e-06,
+      "loss": 1.525,
+      "step": 70580
+    },
+    {
+      "epoch": 2.0536986938941735,
+      "grad_norm": 14.0,
+      "learning_rate": 3.1543386045918116e-06,
+      "loss": 1.5286,
+      "step": 70600
+    },
+    {
+      "epoch": 2.054280478226721,
+      "grad_norm": 14.25,
+      "learning_rate": 3.1523993229191967e-06,
+      "loss": 1.4856,
+      "step": 70620
+    },
+    {
+      "epoch": 2.0548622625592694,
+      "grad_norm": 14.4375,
+      "learning_rate": 3.150460041246582e-06,
+      "loss": 1.4911,
+      "step": 70640
+    },
+    {
+      "epoch": 2.055444046891817,
+      "grad_norm": 13.5,
+      "learning_rate": 3.148520759573967e-06,
+      "loss": 1.5228,
+      "step": 70660
+    },
+    {
+      "epoch": 2.0560258312243653,
+      "grad_norm": 12.6875,
+      "learning_rate": 3.146581477901352e-06,
+      "loss": 1.4383,
+      "step": 70680
+    },
+    {
+      "epoch": 2.056607615556913,
+      "grad_norm": 12.8125,
+      "learning_rate": 3.144642196228737e-06,
+      "loss": 1.4915,
+      "step": 70700
+    },
+    {
+      "epoch": 2.057189399889461,
+      "grad_norm": 13.0,
+      "learning_rate": 3.1427029145561222e-06,
+      "loss": 1.4782,
+      "step": 70720
+    },
+    {
+      "epoch": 2.057771184222009,
+      "grad_norm": 11.4375,
+      "learning_rate": 3.1407636328835073e-06,
+      "loss": 1.485,
+      "step": 70740
+    },
+    {
+      "epoch": 2.0583529685545567,
+      "grad_norm": 14.25,
+      "learning_rate": 3.1388243512108924e-06,
+      "loss": 1.6041,
+      "step": 70760
+    },
+    {
+      "epoch": 2.058934752887105,
+      "grad_norm": 13.1875,
+      "learning_rate": 3.1368850695382775e-06,
+      "loss": 1.4376,
+      "step": 70780
+    },
+    {
+      "epoch": 2.0595165372196527,
+      "grad_norm": 13.8125,
+      "learning_rate": 3.1349457878656626e-06,
+      "loss": 1.4167,
+      "step": 70800
+    },
+    {
+      "epoch": 2.0600983215522004,
+      "grad_norm": 13.6875,
+      "learning_rate": 3.1330065061930477e-06,
+      "loss": 1.4912,
+      "step": 70820
+    },
+    {
+      "epoch": 2.0606801058847486,
+      "grad_norm": 12.9375,
+      "learning_rate": 3.131067224520433e-06,
+      "loss": 1.5176,
+      "step": 70840
+    },
+    {
+      "epoch": 2.0612618902172963,
+      "grad_norm": 13.1875,
+      "learning_rate": 3.129127942847818e-06,
+      "loss": 1.4853,
+      "step": 70860
+    },
+    {
+      "epoch": 2.0618436745498445,
+      "grad_norm": 13.4375,
+      "learning_rate": 3.127188661175203e-06,
+      "loss": 1.4644,
+      "step": 70880
+    },
+    {
+      "epoch": 2.0624254588823923,
+      "grad_norm": 11.0625,
+      "learning_rate": 3.125249379502588e-06,
+      "loss": 1.4847,
+      "step": 70900
+    },
+    {
+      "epoch": 2.0630072432149404,
+      "grad_norm": 12.5,
+      "learning_rate": 3.1233100978299728e-06,
+      "loss": 1.5436,
+      "step": 70920
+    },
+    {
+      "epoch": 2.063589027547488,
+      "grad_norm": 12.875,
+      "learning_rate": 3.121370816157358e-06,
+      "loss": 1.4474,
+      "step": 70940
+    },
+    {
+      "epoch": 2.064170811880036,
+      "grad_norm": 11.625,
+      "learning_rate": 3.119431534484743e-06,
+      "loss": 1.5005,
+      "step": 70960
+    },
+    {
+      "epoch": 2.064752596212584,
+      "grad_norm": 11.8125,
+      "learning_rate": 3.117492252812128e-06,
+      "loss": 1.5289,
+      "step": 70980
+    },
+    {
+      "epoch": 2.065334380545132,
+      "grad_norm": 12.625,
+      "learning_rate": 3.115552971139513e-06,
+      "loss": 1.4181,
+      "step": 71000
+    },
+    {
+      "epoch": 2.06591616487768,
+      "grad_norm": 10.5,
+      "learning_rate": 3.1136136894668983e-06,
+      "loss": 1.405,
+      "step": 71020
+    },
+    {
+      "epoch": 2.066497949210228,
+      "grad_norm": 12.8125,
+      "learning_rate": 3.1116744077942834e-06,
+      "loss": 1.522,
+      "step": 71040
+    },
+    {
+      "epoch": 2.0670797335427755,
+      "grad_norm": 15.8125,
+      "learning_rate": 3.1097351261216684e-06,
+      "loss": 1.5006,
+      "step": 71060
+    },
+    {
+      "epoch": 2.0676615178753237,
+      "grad_norm": 12.3125,
+      "learning_rate": 3.1077958444490535e-06,
+      "loss": 1.529,
+      "step": 71080
+    },
+    {
+      "epoch": 2.0682433022078714,
+      "grad_norm": 14.125,
+      "learning_rate": 3.1058565627764386e-06,
+      "loss": 1.4701,
+      "step": 71100
+    },
+    {
+      "epoch": 2.0688250865404196,
+      "grad_norm": 16.0,
+      "learning_rate": 3.1039172811038237e-06,
+      "loss": 1.5067,
+      "step": 71120
+    },
+    {
+      "epoch": 2.0694068708729674,
+      "grad_norm": 12.0,
+      "learning_rate": 3.101977999431209e-06,
+      "loss": 1.4489,
+      "step": 71140
+    },
+    {
+      "epoch": 2.069988655205515,
+      "grad_norm": 12.125,
+      "learning_rate": 3.100038717758594e-06,
+      "loss": 1.4791,
+      "step": 71160
+    },
+    {
+      "epoch": 2.0705704395380633,
+      "grad_norm": 14.25,
+      "learning_rate": 3.098099436085979e-06,
+      "loss": 1.4841,
+      "step": 71180
+    },
+    {
+      "epoch": 2.071152223870611,
+      "grad_norm": 11.1875,
+      "learning_rate": 3.096160154413364e-06,
+      "loss": 1.4563,
+      "step": 71200
+    },
+    {
+      "epoch": 2.0717340082031592,
+      "grad_norm": 14.125,
+      "learning_rate": 3.0942208727407492e-06,
+      "loss": 1.4879,
+      "step": 71220
+    },
+    {
+      "epoch": 2.072315792535707,
+      "grad_norm": 14.875,
+      "learning_rate": 3.0922815910681343e-06,
+      "loss": 1.4111,
+      "step": 71240
+    },
+    {
+      "epoch": 2.0728975768682547,
+      "grad_norm": 14.5625,
+      "learning_rate": 3.0903423093955194e-06,
+      "loss": 1.5508,
+      "step": 71260
+    },
+    {
+      "epoch": 2.073479361200803,
+      "grad_norm": 14.875,
+      "learning_rate": 3.0884030277229045e-06,
+      "loss": 1.4548,
+      "step": 71280
+    },
+    {
+      "epoch": 2.0740611455333506,
+      "grad_norm": 15.0625,
+      "learning_rate": 3.0864637460502896e-06,
+      "loss": 1.5113,
+      "step": 71300
+    },
+    {
+      "epoch": 2.074642929865899,
+      "grad_norm": 13.4375,
+      "learning_rate": 3.0845244643776743e-06,
+      "loss": 1.4911,
+      "step": 71320
+    },
+    {
+      "epoch": 2.0752247141984466,
+      "grad_norm": 12.1875,
+      "learning_rate": 3.0825851827050594e-06,
+      "loss": 1.5903,
+      "step": 71340
+    },
+    {
+      "epoch": 2.0758064985309947,
+      "grad_norm": 14.0,
+      "learning_rate": 3.0806459010324445e-06,
+      "loss": 1.4759,
+      "step": 71360
+    },
+    {
+      "epoch": 2.0763882828635425,
+      "grad_norm": 13.0,
+      "learning_rate": 3.0787066193598296e-06,
+      "loss": 1.493,
+      "step": 71380
+    },
+    {
+      "epoch": 2.0769700671960902,
+      "grad_norm": 14.125,
+      "learning_rate": 3.0767673376872147e-06,
+      "loss": 1.4959,
+      "step": 71400
+    },
+    {
+      "epoch": 2.0775518515286384,
+      "grad_norm": 14.4375,
+      "learning_rate": 3.0748280560145998e-06,
+      "loss": 1.4924,
+      "step": 71420
+    },
+    {
+      "epoch": 2.078133635861186,
+      "grad_norm": 12.625,
+      "learning_rate": 3.072888774341985e-06,
+      "loss": 1.4623,
+      "step": 71440
+    },
+    {
+      "epoch": 2.0787154201937343,
+      "grad_norm": 15.375,
+      "learning_rate": 3.07094949266937e-06,
+      "loss": 1.4677,
+      "step": 71460
+    },
+    {
+      "epoch": 2.079297204526282,
+      "grad_norm": 13.6875,
+      "learning_rate": 3.069010210996755e-06,
+      "loss": 1.4981,
+      "step": 71480
+    },
+    {
+      "epoch": 2.07987898885883,
+      "grad_norm": 13.875,
+      "learning_rate": 3.06707092932414e-06,
+      "loss": 1.4813,
+      "step": 71500
+    },
+    {
+      "epoch": 2.080460773191378,
+      "grad_norm": 12.1875,
+      "learning_rate": 3.0651316476515253e-06,
+      "loss": 1.454,
+      "step": 71520
+    },
+    {
+      "epoch": 2.0810425575239258,
+      "grad_norm": 11.5,
+      "learning_rate": 3.0631923659789104e-06,
+      "loss": 1.5398,
+      "step": 71540
+    },
+    {
+      "epoch": 2.081624341856474,
+      "grad_norm": 13.9375,
+      "learning_rate": 3.0612530843062955e-06,
+      "loss": 1.5193,
+      "step": 71560
+    },
+    {
+      "epoch": 2.0822061261890217,
+      "grad_norm": 11.875,
+      "learning_rate": 3.0593138026336806e-06,
+      "loss": 1.485,
+      "step": 71580
+    },
+    {
+      "epoch": 2.08278791052157,
+      "grad_norm": 12.5625,
+      "learning_rate": 3.0573745209610657e-06,
+      "loss": 1.4731,
+      "step": 71600
+    },
+    {
+      "epoch": 2.0833696948541176,
+      "grad_norm": 13.4375,
+      "learning_rate": 3.0554352392884508e-06,
+      "loss": 1.4991,
+      "step": 71620
+    },
+    {
+      "epoch": 2.0839514791866653,
+      "grad_norm": 12.8125,
+      "learning_rate": 3.053495957615836e-06,
+      "loss": 1.5984,
+      "step": 71640
+    },
+    {
+      "epoch": 2.0845332635192135,
+      "grad_norm": 18.25,
+      "learning_rate": 3.051556675943221e-06,
+      "loss": 1.4401,
+      "step": 71660
+    },
+    {
+      "epoch": 2.0851150478517613,
+      "grad_norm": 15.0,
+      "learning_rate": 3.049617394270606e-06,
+      "loss": 1.4982,
+      "step": 71680
+    },
+    {
+      "epoch": 2.0856968321843095,
+      "grad_norm": 12.125,
+      "learning_rate": 3.047678112597991e-06,
+      "loss": 1.442,
+      "step": 71700
+    },
+    {
+      "epoch": 2.086278616516857,
+      "grad_norm": 14.9375,
+      "learning_rate": 3.0457388309253762e-06,
+      "loss": 1.4435,
+      "step": 71720
+    },
+    {
+      "epoch": 2.086860400849405,
+      "grad_norm": 12.875,
+      "learning_rate": 3.043799549252761e-06,
+      "loss": 1.4565,
+      "step": 71740
+    },
+    {
+      "epoch": 2.087442185181953,
+      "grad_norm": 16.75,
+      "learning_rate": 3.041860267580146e-06,
+      "loss": 1.5337,
+      "step": 71760
+    },
+    {
+      "epoch": 2.088023969514501,
+      "grad_norm": 16.125,
+      "learning_rate": 3.039920985907531e-06,
+      "loss": 1.5205,
+      "step": 71780
+    },
+    {
+      "epoch": 2.088605753847049,
+      "grad_norm": 13.5,
+      "learning_rate": 3.037981704234916e-06,
+      "loss": 1.4813,
+      "step": 71800
+    },
+    {
+      "epoch": 2.089187538179597,
+      "grad_norm": 15.1875,
+      "learning_rate": 3.0360424225623013e-06,
+      "loss": 1.5243,
+      "step": 71820
+    },
+    {
+      "epoch": 2.0897693225121445,
+      "grad_norm": 14.875,
+      "learning_rate": 3.0341031408896864e-06,
+      "loss": 1.4947,
+      "step": 71840
+    },
+    {
+      "epoch": 2.0903511068446927,
+      "grad_norm": 14.5625,
+      "learning_rate": 3.0321638592170715e-06,
+      "loss": 1.4967,
+      "step": 71860
+    },
+    {
+      "epoch": 2.0909328911772405,
+      "grad_norm": 11.875,
+      "learning_rate": 3.0302245775444566e-06,
+      "loss": 1.527,
+      "step": 71880
+    },
+    {
+      "epoch": 2.0915146755097886,
+      "grad_norm": 13.625,
+      "learning_rate": 3.0282852958718417e-06,
+      "loss": 1.4975,
+      "step": 71900
+    },
+    {
+      "epoch": 2.0920964598423364,
+      "grad_norm": 14.875,
+      "learning_rate": 3.026346014199227e-06,
+      "loss": 1.535,
+      "step": 71920
+    },
+    {
+      "epoch": 2.0926782441748846,
+      "grad_norm": 16.375,
+      "learning_rate": 3.024406732526612e-06,
+      "loss": 1.5116,
+      "step": 71940
+    },
+    {
+      "epoch": 2.0932600285074323,
+      "grad_norm": 14.5625,
+      "learning_rate": 3.022467450853997e-06,
+      "loss": 1.5866,
+      "step": 71960
+    },
+    {
+      "epoch": 2.09384181283998,
+      "grad_norm": 12.75,
+      "learning_rate": 3.020528169181382e-06,
+      "loss": 1.468,
+      "step": 71980
+    },
+    {
+      "epoch": 2.0944235971725282,
+      "grad_norm": 11.3125,
+      "learning_rate": 3.018588887508767e-06,
+      "loss": 1.5545,
+      "step": 72000
+    },
+    {
+      "epoch": 2.095005381505076,
+      "grad_norm": 11.6875,
+      "learning_rate": 3.0166496058361523e-06,
+      "loss": 1.5561,
+      "step": 72020
+    },
+    {
+      "epoch": 2.095587165837624,
+      "grad_norm": 16.5,
+      "learning_rate": 3.0147103241635374e-06,
+      "loss": 1.4659,
+      "step": 72040
+    },
+    {
+      "epoch": 2.096168950170172,
+      "grad_norm": 12.875,
+      "learning_rate": 3.0127710424909225e-06,
+      "loss": 1.5107,
+      "step": 72060
+    },
+    {
+      "epoch": 2.0967507345027196,
+      "grad_norm": 12.375,
+      "learning_rate": 3.0108317608183076e-06,
+      "loss": 1.4283,
+      "step": 72080
+    },
+    {
+      "epoch": 2.097332518835268,
+      "grad_norm": 12.25,
+      "learning_rate": 3.0088924791456927e-06,
+      "loss": 1.5606,
+      "step": 72100
+    },
+    {
+      "epoch": 2.0979143031678156,
+      "grad_norm": 11.6875,
+      "learning_rate": 3.0069531974730778e-06,
+      "loss": 1.5268,
+      "step": 72120
+    },
+    {
+      "epoch": 2.0984960875003638,
+      "grad_norm": 13.375,
+      "learning_rate": 3.0050139158004624e-06,
+      "loss": 1.502,
+      "step": 72140
+    },
+    {
+      "epoch": 2.0990778718329115,
+      "grad_norm": 14.375,
+      "learning_rate": 3.0030746341278475e-06,
+      "loss": 1.5983,
+      "step": 72160
+    },
+    {
+      "epoch": 2.0996596561654597,
+      "grad_norm": 13.5,
+      "learning_rate": 3.0011353524552326e-06,
+      "loss": 1.586,
+      "step": 72180
+    },
+    {
+      "epoch": 2.1002414404980074,
+      "grad_norm": 14.625,
+      "learning_rate": 2.9991960707826177e-06,
+      "loss": 1.4499,
+      "step": 72200
+    },
+    {
+      "epoch": 2.100823224830555,
+      "grad_norm": 13.1875,
+      "learning_rate": 2.997256789110003e-06,
+      "loss": 1.5428,
+      "step": 72220
+    },
+    {
+      "epoch": 2.1014050091631034,
+      "grad_norm": 15.4375,
+      "learning_rate": 2.995317507437388e-06,
+      "loss": 1.4782,
+      "step": 72240
+    },
+    {
+      "epoch": 2.101986793495651,
+      "grad_norm": 12.875,
+      "learning_rate": 2.993378225764773e-06,
+      "loss": 1.4888,
+      "step": 72260
+    },
+    {
+      "epoch": 2.1025685778281993,
+      "grad_norm": 11.25,
+      "learning_rate": 2.991438944092158e-06,
+      "loss": 1.521,
+      "step": 72280
+    },
+    {
+      "epoch": 2.103150362160747,
+      "grad_norm": 14.625,
+      "learning_rate": 2.9894996624195432e-06,
+      "loss": 1.5041,
+      "step": 72300
+    },
+    {
+      "epoch": 2.1037321464932948,
+      "grad_norm": 12.6875,
+      "learning_rate": 2.9875603807469283e-06,
+      "loss": 1.4956,
+      "step": 72320
+    },
+    {
+      "epoch": 2.104313930825843,
+      "grad_norm": 12.75,
+      "learning_rate": 2.9856210990743134e-06,
+      "loss": 1.4933,
+      "step": 72340
+    },
+    {
+      "epoch": 2.1048957151583907,
+      "grad_norm": 12.8125,
+      "learning_rate": 2.9836818174016985e-06,
+      "loss": 1.5132,
+      "step": 72360
+    },
+    {
+      "epoch": 2.105477499490939,
+      "grad_norm": 11.75,
+      "learning_rate": 2.9817425357290836e-06,
+      "loss": 1.527,
+      "step": 72380
+    },
+    {
+      "epoch": 2.1060592838234866,
+      "grad_norm": 13.8125,
+      "learning_rate": 2.9798032540564687e-06,
+      "loss": 1.4684,
+      "step": 72400
+    },
+    {
+      "epoch": 2.1066410681560344,
+      "grad_norm": 12.0,
+      "learning_rate": 2.977863972383854e-06,
+      "loss": 1.5328,
+      "step": 72420
+    },
+    {
+      "epoch": 2.1072228524885825,
+      "grad_norm": 14.6875,
+      "learning_rate": 2.975924690711239e-06,
+      "loss": 1.4832,
+      "step": 72440
+    },
+    {
+      "epoch": 2.1078046368211303,
+      "grad_norm": 14.75,
+      "learning_rate": 2.973985409038624e-06,
+      "loss": 1.4761,
+      "step": 72460
+    },
+    {
+      "epoch": 2.1083864211536785,
+      "grad_norm": 14.125,
+      "learning_rate": 2.972046127366009e-06,
+      "loss": 1.5431,
+      "step": 72480
+    },
+    {
+      "epoch": 2.108968205486226,
+      "grad_norm": 15.5,
+      "learning_rate": 2.970106845693394e-06,
+      "loss": 1.5485,
+      "step": 72500
+    },
+    {
+      "epoch": 2.109549989818774,
+      "grad_norm": 11.5625,
+      "learning_rate": 2.9681675640207793e-06,
+      "loss": 1.4798,
+      "step": 72520
+    },
+    {
+      "epoch": 2.110131774151322,
+      "grad_norm": 15.75,
+      "learning_rate": 2.966228282348164e-06,
+      "loss": 1.5801,
+      "step": 72540
+    },
+    {
+      "epoch": 2.11071355848387,
+      "grad_norm": 13.0,
+      "learning_rate": 2.964289000675549e-06,
+      "loss": 1.4505,
+      "step": 72560
+    },
+    {
+      "epoch": 2.111295342816418,
+      "grad_norm": 13.5625,
+      "learning_rate": 2.9623497190029337e-06,
+      "loss": 1.4387,
+      "step": 72580
+    },
+    {
+      "epoch": 2.111877127148966,
+      "grad_norm": 14.75,
+      "learning_rate": 2.960410437330319e-06,
+      "loss": 1.4776,
+      "step": 72600
+    },
+    {
+      "epoch": 2.112458911481514,
+      "grad_norm": 14.75,
+      "learning_rate": 2.958471155657704e-06,
+      "loss": 1.4937,
+      "step": 72620
+    },
+    {
+      "epoch": 2.1130406958140617,
+      "grad_norm": 12.6875,
+      "learning_rate": 2.956531873985089e-06,
+      "loss": 1.5073,
+      "step": 72640
+    },
+    {
+      "epoch": 2.1136224801466095,
+      "grad_norm": 14.375,
+      "learning_rate": 2.954592592312474e-06,
+      "loss": 1.4216,
+      "step": 72660
+    },
+    {
+      "epoch": 2.1142042644791577,
+      "grad_norm": 14.875,
+      "learning_rate": 2.9526533106398592e-06,
+      "loss": 1.4707,
+      "step": 72680
+    },
+    {
+      "epoch": 2.1147860488117054,
+      "grad_norm": 15.0625,
+      "learning_rate": 2.9507140289672443e-06,
+      "loss": 1.4336,
+      "step": 72700
+    },
+    {
+      "epoch": 2.1153678331442536,
+      "grad_norm": 15.125,
+      "learning_rate": 2.9487747472946294e-06,
+      "loss": 1.4724,
+      "step": 72720
+    },
+    {
+      "epoch": 2.1159496174768013,
+      "grad_norm": 15.375,
+      "learning_rate": 2.9468354656220145e-06,
+      "loss": 1.4837,
+      "step": 72740
+    },
+    {
+      "epoch": 2.116531401809349,
+      "grad_norm": 15.0625,
+      "learning_rate": 2.9448961839493996e-06,
+      "loss": 1.433,
+      "step": 72760
+    },
+    {
+      "epoch": 2.1171131861418973,
+      "grad_norm": 13.5,
+      "learning_rate": 2.9429569022767847e-06,
+      "loss": 1.5166,
+      "step": 72780
+    },
+    {
+      "epoch": 2.117694970474445,
+      "grad_norm": 14.625,
+      "learning_rate": 2.94101762060417e-06,
+      "loss": 1.4642,
+      "step": 72800
+    },
+    {
+      "epoch": 2.118276754806993,
+      "grad_norm": 13.25,
+      "learning_rate": 2.939078338931555e-06,
+      "loss": 1.4825,
+      "step": 72820
+    },
+    {
+      "epoch": 2.118858539139541,
+      "grad_norm": 13.0625,
+      "learning_rate": 2.9371390572589396e-06,
+      "loss": 1.4745,
+      "step": 72840
+    },
+    {
+      "epoch": 2.119440323472089,
+      "grad_norm": 10.625,
+      "learning_rate": 2.9351997755863247e-06,
+      "loss": 1.4926,
+      "step": 72860
+    },
+    {
+      "epoch": 2.120022107804637,
+      "grad_norm": 12.3125,
+      "learning_rate": 2.9332604939137098e-06,
+      "loss": 1.4766,
+      "step": 72880
+    },
+    {
+      "epoch": 2.1206038921371846,
+      "grad_norm": 12.25,
+      "learning_rate": 2.931321212241095e-06,
+      "loss": 1.4237,
+      "step": 72900
+    },
+    {
+      "epoch": 2.1211856764697328,
+      "grad_norm": 13.625,
+      "learning_rate": 2.92938193056848e-06,
+      "loss": 1.4522,
+      "step": 72920
+    },
+    {
+      "epoch": 2.1217674608022805,
+      "grad_norm": 11.0625,
+      "learning_rate": 2.927442648895865e-06,
+      "loss": 1.4988,
+      "step": 72940
+    },
+    {
+      "epoch": 2.1223492451348287,
+      "grad_norm": 12.3125,
+      "learning_rate": 2.92550336722325e-06,
+      "loss": 1.5132,
+      "step": 72960
+    },
+    {
+      "epoch": 2.1229310294673764,
+      "grad_norm": 14.75,
+      "learning_rate": 2.9235640855506353e-06,
+      "loss": 1.4619,
+      "step": 72980
+    },
+    {
+      "epoch": 2.123512813799924,
+      "grad_norm": 12.1875,
+      "learning_rate": 2.9216248038780204e-06,
+      "loss": 1.4034,
+      "step": 73000
+    },
+    {
+      "epoch": 2.1240945981324724,
+      "grad_norm": 14.6875,
+      "learning_rate": 2.9196855222054055e-06,
+      "loss": 1.5266,
+      "step": 73020
+    },
+    {
+      "epoch": 2.12467638246502,
+      "grad_norm": 14.5625,
+      "learning_rate": 2.9177462405327906e-06,
+      "loss": 1.5114,
+      "step": 73040
+    },
+    {
+      "epoch": 2.1252581667975683,
+      "grad_norm": 14.3125,
+      "learning_rate": 2.9158069588601757e-06,
+      "loss": 1.4811,
+      "step": 73060
+    },
+    {
+      "epoch": 2.125839951130116,
+      "grad_norm": 17.25,
+      "learning_rate": 2.9138676771875608e-06,
+      "loss": 1.5085,
+      "step": 73080
+    },
+    {
+      "epoch": 2.1264217354626638,
+      "grad_norm": 14.1875,
+      "learning_rate": 2.911928395514946e-06,
+      "loss": 1.4951,
+      "step": 73100
+    },
+    {
+      "epoch": 2.127003519795212,
+      "grad_norm": 12.8125,
+      "learning_rate": 2.909989113842331e-06,
+      "loss": 1.4602,
+      "step": 73120
+    },
+    {
+      "epoch": 2.1275853041277597,
+      "grad_norm": 14.0,
+      "learning_rate": 2.908049832169716e-06,
+      "loss": 1.4742,
+      "step": 73140
+    },
+    {
+      "epoch": 2.128167088460308,
+      "grad_norm": 14.4375,
+      "learning_rate": 2.906110550497101e-06,
+      "loss": 1.5108,
+      "step": 73160
+    },
+    {
+      "epoch": 2.1287488727928556,
+      "grad_norm": 13.3125,
+      "learning_rate": 2.9041712688244862e-06,
+      "loss": 1.5393,
+      "step": 73180
+    },
+    {
+      "epoch": 2.129330657125404,
+      "grad_norm": 12.625,
+      "learning_rate": 2.9022319871518713e-06,
+      "loss": 1.4763,
+      "step": 73200
+    },
+    {
+      "epoch": 2.1299124414579516,
+      "grad_norm": 12.5,
+      "learning_rate": 2.9002927054792564e-06,
+      "loss": 1.4555,
+      "step": 73220
+    },
+    {
+      "epoch": 2.1304942257904993,
+      "grad_norm": 13.375,
+      "learning_rate": 2.898353423806641e-06,
+      "loss": 1.5222,
+      "step": 73240
+    },
+    {
+      "epoch": 2.1310760101230475,
+      "grad_norm": 15.5,
+      "learning_rate": 2.896414142134026e-06,
+      "loss": 1.5048,
+      "step": 73260
+    },
+    {
+      "epoch": 2.1316577944555952,
+      "grad_norm": 15.625,
+      "learning_rate": 2.8944748604614113e-06,
+      "loss": 1.5817,
+      "step": 73280
+    },
+    {
+      "epoch": 2.1322395787881434,
+      "grad_norm": 14.125,
+      "learning_rate": 2.8925355787887964e-06,
+      "loss": 1.4889,
+      "step": 73300
+    },
+    {
+      "epoch": 2.132821363120691,
+      "grad_norm": 14.1875,
+      "learning_rate": 2.8905962971161815e-06,
+      "loss": 1.5541,
+      "step": 73320
+    },
+    {
+      "epoch": 2.133403147453239,
+      "grad_norm": 13.875,
+      "learning_rate": 2.8886570154435666e-06,
+      "loss": 1.5182,
+      "step": 73340
+    },
+    {
+      "epoch": 2.133984931785787,
+      "grad_norm": 13.25,
+      "learning_rate": 2.8867177337709517e-06,
+      "loss": 1.5519,
+      "step": 73360
+    },
+    {
+      "epoch": 2.134566716118335,
+      "grad_norm": 14.375,
+      "learning_rate": 2.884778452098337e-06,
+      "loss": 1.4621,
+      "step": 73380
+    },
+    {
+      "epoch": 2.135148500450883,
+      "grad_norm": 10.0,
+      "learning_rate": 2.882839170425722e-06,
+      "loss": 1.441,
+      "step": 73400
+    },
+    {
+      "epoch": 2.1357302847834307,
+      "grad_norm": 12.9375,
+      "learning_rate": 2.880899888753107e-06,
+      "loss": 1.4363,
+      "step": 73420
+    },
+    {
+      "epoch": 2.136312069115979,
+      "grad_norm": 15.875,
+      "learning_rate": 2.878960607080492e-06,
+      "loss": 1.5074,
+      "step": 73440
+    },
+    {
+      "epoch": 2.1368938534485267,
+      "grad_norm": 14.5,
+      "learning_rate": 2.877021325407877e-06,
+      "loss": 1.4954,
+      "step": 73460
+    },
+    {
+      "epoch": 2.1374756377810744,
+      "grad_norm": 17.375,
+      "learning_rate": 2.8750820437352623e-06,
+      "loss": 1.6136,
+      "step": 73480
+    },
+    {
+      "epoch": 2.1380574221136226,
+      "grad_norm": 12.9375,
+      "learning_rate": 2.8731427620626474e-06,
+      "loss": 1.4908,
+      "step": 73500
+    },
+    {
+      "epoch": 2.1386392064461703,
+      "grad_norm": 13.8125,
+      "learning_rate": 2.8712034803900325e-06,
+      "loss": 1.4283,
+      "step": 73520
+    },
+    {
+      "epoch": 2.1392209907787185,
+      "grad_norm": 12.375,
+      "learning_rate": 2.8692641987174176e-06,
+      "loss": 1.4775,
+      "step": 73540
+    },
+    {
+      "epoch": 2.1398027751112663,
+      "grad_norm": 14.1875,
+      "learning_rate": 2.8673249170448027e-06,
+      "loss": 1.562,
+      "step": 73560
+    },
+    {
+      "epoch": 2.140384559443814,
+      "grad_norm": 14.0,
+      "learning_rate": 2.8653856353721878e-06,
+      "loss": 1.4407,
+      "step": 73580
+    },
+    {
+      "epoch": 2.140966343776362,
+      "grad_norm": 12.5625,
+      "learning_rate": 2.863446353699573e-06,
+      "loss": 1.5493,
+      "step": 73600
+    },
+    {
+      "epoch": 2.14154812810891,
+      "grad_norm": 12.25,
+      "learning_rate": 2.861507072026958e-06,
+      "loss": 1.4486,
+      "step": 73620
+    },
+    {
+      "epoch": 2.142129912441458,
+      "grad_norm": 14.75,
+      "learning_rate": 2.8595677903543426e-06,
+      "loss": 1.5861,
+      "step": 73640
+    },
+    {
+      "epoch": 2.142711696774006,
+      "grad_norm": 12.375,
+      "learning_rate": 2.8576285086817277e-06,
+      "loss": 1.5289,
+      "step": 73660
+    },
+    {
+      "epoch": 2.1432934811065536,
+      "grad_norm": 14.0,
+      "learning_rate": 2.855689227009113e-06,
+      "loss": 1.5793,
+      "step": 73680
+    },
+    {
+      "epoch": 2.143875265439102,
+      "grad_norm": 13.5625,
+      "learning_rate": 2.853749945336498e-06,
+      "loss": 1.481,
+      "step": 73700
+    },
+    {
+      "epoch": 2.1444570497716495,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.851810663663883e-06,
+      "loss": 1.4722,
+      "step": 73720
+    },
+    {
+      "epoch": 2.1450388341041977,
+      "grad_norm": 12.1875,
+      "learning_rate": 2.849871381991268e-06,
+      "loss": 1.3964,
+      "step": 73740
+    },
+    {
+      "epoch": 2.1456206184367455,
+      "grad_norm": 17.0,
+      "learning_rate": 2.8479321003186532e-06,
+      "loss": 1.4954,
+      "step": 73760
+    },
+    {
+      "epoch": 2.146202402769293,
+      "grad_norm": 12.8125,
+      "learning_rate": 2.8459928186460383e-06,
+      "loss": 1.5448,
+      "step": 73780
+    },
+    {
+      "epoch": 2.1467841871018414,
+      "grad_norm": 13.125,
+      "learning_rate": 2.8440535369734234e-06,
+      "loss": 1.5075,
+      "step": 73800
+    },
+    {
+      "epoch": 2.147365971434389,
+      "grad_norm": 14.125,
+      "learning_rate": 2.8421142553008085e-06,
+      "loss": 1.485,
+      "step": 73820
+    },
+    {
+      "epoch": 2.1479477557669373,
+      "grad_norm": 12.9375,
+      "learning_rate": 2.8401749736281936e-06,
+      "loss": 1.5189,
+      "step": 73840
+    },
+    {
+      "epoch": 2.148529540099485,
+      "grad_norm": 13.5,
+      "learning_rate": 2.8382356919555787e-06,
+      "loss": 1.5812,
+      "step": 73860
+    },
+    {
+      "epoch": 2.1491113244320332,
+      "grad_norm": 13.125,
+      "learning_rate": 2.836296410282964e-06,
+      "loss": 1.499,
+      "step": 73880
+    },
+    {
+      "epoch": 2.149693108764581,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.834357128610349e-06,
+      "loss": 1.4625,
+      "step": 73900
+    },
+    {
+      "epoch": 2.1502748930971287,
+      "grad_norm": 13.1875,
+      "learning_rate": 2.832417846937734e-06,
+      "loss": 1.4651,
+      "step": 73920
+    },
+    {
+      "epoch": 2.150856677429677,
+      "grad_norm": 10.4375,
+      "learning_rate": 2.830478565265119e-06,
+      "loss": 1.5115,
+      "step": 73940
+    },
+    {
+      "epoch": 2.1514384617622246,
+      "grad_norm": 12.1875,
+      "learning_rate": 2.828539283592504e-06,
+      "loss": 1.4763,
+      "step": 73960
+    },
+    {
+      "epoch": 2.152020246094773,
+      "grad_norm": 11.625,
+      "learning_rate": 2.8266000019198893e-06,
+      "loss": 1.5031,
+      "step": 73980
+    },
+    {
+      "epoch": 2.1526020304273206,
+      "grad_norm": 14.6875,
+      "learning_rate": 2.8246607202472744e-06,
+      "loss": 1.5051,
+      "step": 74000
+    },
+    {
+      "epoch": 2.1531838147598683,
+      "grad_norm": 14.875,
+      "learning_rate": 2.8227214385746595e-06,
+      "loss": 1.4666,
+      "step": 74020
+    },
+    {
+      "epoch": 2.1537655990924165,
+      "grad_norm": 12.25,
+      "learning_rate": 2.8207821569020446e-06,
+      "loss": 1.4348,
+      "step": 74040
+    },
+    {
+      "epoch": 2.1543473834249642,
+      "grad_norm": 13.0,
+      "learning_rate": 2.8188428752294293e-06,
+      "loss": 1.4955,
+      "step": 74060
+    },
+    {
+      "epoch": 2.1549291677575124,
+      "grad_norm": 16.25,
+      "learning_rate": 2.8169035935568144e-06,
+      "loss": 1.5198,
+      "step": 74080
+    },
+    {
+      "epoch": 2.15551095209006,
+      "grad_norm": 13.5,
+      "learning_rate": 2.8149643118841995e-06,
+      "loss": 1.4427,
+      "step": 74100
+    },
+    {
+      "epoch": 2.1560927364226083,
+      "grad_norm": 13.625,
+      "learning_rate": 2.8130250302115845e-06,
+      "loss": 1.4765,
+      "step": 74120
+    },
+    {
+      "epoch": 2.156674520755156,
+      "grad_norm": 11.125,
+      "learning_rate": 2.8110857485389696e-06,
+      "loss": 1.4595,
+      "step": 74140
+    },
+    {
+      "epoch": 2.157256305087704,
+      "grad_norm": 13.5625,
+      "learning_rate": 2.8091464668663547e-06,
+      "loss": 1.5147,
+      "step": 74160
+    },
+    {
+      "epoch": 2.157838089420252,
+      "grad_norm": 12.8125,
+      "learning_rate": 2.80720718519374e-06,
+      "loss": 1.5748,
+      "step": 74180
+    },
+    {
+      "epoch": 2.1584198737527998,
+      "grad_norm": 13.4375,
+      "learning_rate": 2.805267903521125e-06,
+      "loss": 1.5331,
+      "step": 74200
+    },
+    {
+      "epoch": 2.159001658085348,
+      "grad_norm": 14.4375,
+      "learning_rate": 2.80332862184851e-06,
+      "loss": 1.4877,
+      "step": 74220
+    },
+    {
+      "epoch": 2.1595834424178957,
+      "grad_norm": 13.25,
+      "learning_rate": 2.801389340175895e-06,
+      "loss": 1.4491,
+      "step": 74240
+    },
+    {
+      "epoch": 2.1601652267504434,
+      "grad_norm": 14.125,
+      "learning_rate": 2.7994500585032802e-06,
+      "loss": 1.4192,
+      "step": 74260
+    },
+    {
+      "epoch": 2.1607470110829916,
+      "grad_norm": 15.125,
+      "learning_rate": 2.7975107768306653e-06,
+      "loss": 1.5257,
+      "step": 74280
+    },
+    {
+      "epoch": 2.1613287954155393,
+      "grad_norm": 13.75,
+      "learning_rate": 2.7955714951580504e-06,
+      "loss": 1.4681,
+      "step": 74300
+    },
+    {
+      "epoch": 2.1619105797480875,
+      "grad_norm": 17.0,
+      "learning_rate": 2.7936322134854355e-06,
+      "loss": 1.5076,
+      "step": 74320
+    },
+    {
+      "epoch": 2.1624923640806353,
+      "grad_norm": 12.625,
+      "learning_rate": 2.7916929318128206e-06,
+      "loss": 1.5422,
+      "step": 74340
+    },
+    {
+      "epoch": 2.163074148413183,
+      "grad_norm": 13.9375,
+      "learning_rate": 2.7897536501402057e-06,
+      "loss": 1.5043,
+      "step": 74360
+    },
+    {
+      "epoch": 2.163655932745731,
+      "grad_norm": 13.5,
+      "learning_rate": 2.787814368467591e-06,
+      "loss": 1.5271,
+      "step": 74380
+    },
+    {
+      "epoch": 2.164237717078279,
+      "grad_norm": 13.875,
+      "learning_rate": 2.785875086794976e-06,
+      "loss": 1.3912,
+      "step": 74400
+    },
+    {
+      "epoch": 2.164819501410827,
+      "grad_norm": 20.125,
+      "learning_rate": 2.783935805122361e-06,
+      "loss": 1.5108,
+      "step": 74420
+    },
+    {
+      "epoch": 2.165401285743375,
+      "grad_norm": 15.6875,
+      "learning_rate": 2.781996523449746e-06,
+      "loss": 1.5373,
+      "step": 74440
+    },
+    {
+      "epoch": 2.165983070075923,
+      "grad_norm": 12.25,
+      "learning_rate": 2.7800572417771308e-06,
+      "loss": 1.5474,
+      "step": 74460
+    },
+    {
+      "epoch": 2.166564854408471,
+      "grad_norm": 13.6875,
+      "learning_rate": 2.778117960104516e-06,
+      "loss": 1.4762,
+      "step": 74480
+    },
+    {
+      "epoch": 2.1671466387410185,
+      "grad_norm": 13.1875,
+      "learning_rate": 2.776178678431901e-06,
+      "loss": 1.485,
+      "step": 74500
+    },
+    {
+      "epoch": 2.1677284230735667,
+      "grad_norm": 13.375,
+      "learning_rate": 2.774239396759286e-06,
+      "loss": 1.4827,
+      "step": 74520
+    },
+    {
+      "epoch": 2.1683102074061145,
+      "grad_norm": 15.0,
+      "learning_rate": 2.772300115086671e-06,
+      "loss": 1.4605,
+      "step": 74540
+    },
+    {
+      "epoch": 2.1688919917386627,
+      "grad_norm": 14.1875,
+      "learning_rate": 2.7703608334140563e-06,
+      "loss": 1.4337,
+      "step": 74560
+    },
+    {
+      "epoch": 2.1694737760712104,
+      "grad_norm": 13.625,
+      "learning_rate": 2.7684215517414414e-06,
+      "loss": 1.4334,
+      "step": 74580
+    },
+    {
+      "epoch": 2.170055560403758,
+      "grad_norm": 13.1875,
+      "learning_rate": 2.7664822700688265e-06,
+      "loss": 1.5381,
+      "step": 74600
+    },
+    {
+      "epoch": 2.1706373447363063,
+      "grad_norm": 14.0625,
+      "learning_rate": 2.7645429883962116e-06,
+      "loss": 1.4843,
+      "step": 74620
+    },
+    {
+      "epoch": 2.171219129068854,
+      "grad_norm": 14.5,
+      "learning_rate": 2.7626037067235967e-06,
+      "loss": 1.5029,
+      "step": 74640
+    },
+    {
+      "epoch": 2.1718009134014022,
+      "grad_norm": 11.8125,
+      "learning_rate": 2.7606644250509818e-06,
+      "loss": 1.4978,
+      "step": 74660
+    },
+    {
+      "epoch": 2.17238269773395,
+      "grad_norm": 13.625,
+      "learning_rate": 2.758725143378367e-06,
+      "loss": 1.537,
+      "step": 74680
+    },
+    {
+      "epoch": 2.172964482066498,
+      "grad_norm": 14.5,
+      "learning_rate": 2.7567858617057515e-06,
+      "loss": 1.5458,
+      "step": 74700
+    },
+    {
+      "epoch": 2.173546266399046,
+      "grad_norm": 15.875,
+      "learning_rate": 2.7548465800331366e-06,
+      "loss": 1.544,
+      "step": 74720
+    },
+    {
+      "epoch": 2.1741280507315937,
+      "grad_norm": 12.625,
+      "learning_rate": 2.7529072983605213e-06,
+      "loss": 1.4829,
+      "step": 74740
+    },
+    {
+      "epoch": 2.174709835064142,
+      "grad_norm": 15.5,
+      "learning_rate": 2.7509680166879064e-06,
+      "loss": 1.5048,
+      "step": 74760
+    },
+    {
+      "epoch": 2.1752916193966896,
+      "grad_norm": 13.3125,
+      "learning_rate": 2.7490287350152915e-06,
+      "loss": 1.5169,
+      "step": 74780
+    },
+    {
+      "epoch": 2.1758734037292378,
+      "grad_norm": 12.1875,
+      "learning_rate": 2.7470894533426766e-06,
+      "loss": 1.5291,
+      "step": 74800
+    },
+    {
+      "epoch": 2.1764551880617855,
+      "grad_norm": 16.125,
+      "learning_rate": 2.7451501716700617e-06,
+      "loss": 1.5344,
+      "step": 74820
+    },
+    {
+      "epoch": 2.1770369723943332,
+      "grad_norm": 11.375,
+      "learning_rate": 2.743210889997447e-06,
+      "loss": 1.558,
+      "step": 74840
+    },
+    {
+      "epoch": 2.1776187567268814,
+      "grad_norm": 10.875,
+      "learning_rate": 2.741271608324832e-06,
+      "loss": 1.4821,
+      "step": 74860
+    },
+    {
+      "epoch": 2.178200541059429,
+      "grad_norm": 13.4375,
+      "learning_rate": 2.739332326652217e-06,
+      "loss": 1.4888,
+      "step": 74880
+    },
+    {
+      "epoch": 2.1787823253919774,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.737393044979602e-06,
+      "loss": 1.5181,
+      "step": 74900
+    },
+    {
+      "epoch": 2.179364109724525,
+      "grad_norm": 13.625,
+      "learning_rate": 2.735453763306987e-06,
+      "loss": 1.5534,
+      "step": 74920
+    },
+    {
+      "epoch": 2.179945894057073,
+      "grad_norm": 15.8125,
+      "learning_rate": 2.7335144816343723e-06,
+      "loss": 1.4914,
+      "step": 74940
+    },
+    {
+      "epoch": 2.180527678389621,
+      "grad_norm": 12.5,
+      "learning_rate": 2.7315751999617574e-06,
+      "loss": 1.4862,
+      "step": 74960
+    },
+    {
+      "epoch": 2.1811094627221688,
+      "grad_norm": 22.5,
+      "learning_rate": 2.7296359182891425e-06,
+      "loss": 1.5262,
+      "step": 74980
+    },
+    {
+      "epoch": 2.181691247054717,
+      "grad_norm": 13.8125,
+      "learning_rate": 2.7276966366165276e-06,
+      "loss": 1.4733,
+      "step": 75000
+    },
+    {
+      "epoch": 2.1822730313872647,
+      "grad_norm": 13.1875,
+      "learning_rate": 2.7257573549439127e-06,
+      "loss": 1.3873,
+      "step": 75020
+    },
+    {
+      "epoch": 2.1828548157198124,
+      "grad_norm": 11.375,
+      "learning_rate": 2.7238180732712978e-06,
+      "loss": 1.5585,
+      "step": 75040
+    },
+    {
+      "epoch": 2.1834366000523606,
+      "grad_norm": 18.25,
+      "learning_rate": 2.721878791598683e-06,
+      "loss": 1.5439,
+      "step": 75060
+    },
+    {
+      "epoch": 2.1840183843849084,
+      "grad_norm": 12.5,
+      "learning_rate": 2.719939509926068e-06,
+      "loss": 1.4403,
+      "step": 75080
+    },
+    {
+      "epoch": 2.1846001687174565,
+      "grad_norm": 12.625,
+      "learning_rate": 2.718000228253453e-06,
+      "loss": 1.4441,
+      "step": 75100
+    },
+    {
+      "epoch": 2.1851819530500043,
+      "grad_norm": 14.0625,
+      "learning_rate": 2.716060946580838e-06,
+      "loss": 1.6107,
+      "step": 75120
+    },
+    {
+      "epoch": 2.1857637373825525,
+      "grad_norm": 13.0,
+      "learning_rate": 2.7141216649082233e-06,
+      "loss": 1.4784,
+      "step": 75140
+    },
+    {
+      "epoch": 2.1863455217151,
+      "grad_norm": 15.6875,
+      "learning_rate": 2.712182383235608e-06,
+      "loss": 1.5381,
+      "step": 75160
+    },
+    {
+      "epoch": 2.186927306047648,
+      "grad_norm": 13.1875,
+      "learning_rate": 2.710243101562993e-06,
+      "loss": 1.4521,
+      "step": 75180
+    },
+    {
+      "epoch": 2.187509090380196,
+      "grad_norm": 11.9375,
+      "learning_rate": 2.708303819890378e-06,
+      "loss": 1.4395,
+      "step": 75200
+    },
+    {
+      "epoch": 2.188090874712744,
+      "grad_norm": 12.75,
+      "learning_rate": 2.7063645382177632e-06,
+      "loss": 1.5815,
+      "step": 75220
+    },
+    {
+      "epoch": 2.188672659045292,
+      "grad_norm": 10.625,
+      "learning_rate": 2.7044252565451483e-06,
+      "loss": 1.4308,
+      "step": 75240
+    },
+    {
+      "epoch": 2.18925444337784,
+      "grad_norm": 14.75,
+      "learning_rate": 2.7024859748725334e-06,
+      "loss": 1.5653,
+      "step": 75260
+    },
+    {
+      "epoch": 2.189836227710388,
+      "grad_norm": 12.25,
+      "learning_rate": 2.7005466931999185e-06,
+      "loss": 1.5253,
+      "step": 75280
+    },
+    {
+      "epoch": 2.1904180120429357,
+      "grad_norm": 14.25,
+      "learning_rate": 2.6986074115273036e-06,
+      "loss": 1.4762,
+      "step": 75300
+    },
+    {
+      "epoch": 2.1909997963754835,
+      "grad_norm": 11.1875,
+      "learning_rate": 2.6966681298546887e-06,
+      "loss": 1.5231,
+      "step": 75320
+    },
+    {
+      "epoch": 2.1915815807080317,
+      "grad_norm": 12.4375,
+      "learning_rate": 2.694728848182074e-06,
+      "loss": 1.566,
+      "step": 75340
+    },
+    {
+      "epoch": 2.1921633650405794,
+      "grad_norm": 11.75,
+      "learning_rate": 2.692789566509459e-06,
+      "loss": 1.5473,
+      "step": 75360
+    },
+    {
+      "epoch": 2.1927451493731276,
+      "grad_norm": 13.625,
+      "learning_rate": 2.690850284836844e-06,
+      "loss": 1.4593,
+      "step": 75380
+    },
+    {
+      "epoch": 2.1933269337056753,
+      "grad_norm": 14.3125,
+      "learning_rate": 2.688911003164229e-06,
+      "loss": 1.4661,
+      "step": 75400
+    },
+    {
+      "epoch": 2.193908718038223,
+      "grad_norm": 13.3125,
+      "learning_rate": 2.686971721491614e-06,
+      "loss": 1.4729,
+      "step": 75420
+    },
+    {
+      "epoch": 2.1944905023707713,
+      "grad_norm": 14.4375,
+      "learning_rate": 2.6850324398189993e-06,
+      "loss": 1.4975,
+      "step": 75440
+    },
+    {
+      "epoch": 2.195072286703319,
+      "grad_norm": 11.375,
+      "learning_rate": 2.6830931581463844e-06,
+      "loss": 1.4666,
+      "step": 75460
+    },
+    {
+      "epoch": 2.195654071035867,
+      "grad_norm": 14.4375,
+      "learning_rate": 2.6811538764737695e-06,
+      "loss": 1.4428,
+      "step": 75480
+    },
+    {
+      "epoch": 2.196235855368415,
+      "grad_norm": 14.0625,
+      "learning_rate": 2.6792145948011546e-06,
+      "loss": 1.5145,
+      "step": 75500
+    },
+    {
+      "epoch": 2.1968176397009627,
+      "grad_norm": 15.25,
+      "learning_rate": 2.6772753131285397e-06,
+      "loss": 1.5086,
+      "step": 75520
+    },
+    {
+      "epoch": 2.197399424033511,
+      "grad_norm": 13.5,
+      "learning_rate": 2.6753360314559248e-06,
+      "loss": 1.4997,
+      "step": 75540
+    },
+    {
+      "epoch": 2.1979812083660586,
+      "grad_norm": 14.3125,
+      "learning_rate": 2.6733967497833095e-06,
+      "loss": 1.4753,
+      "step": 75560
+    },
+    {
+      "epoch": 2.1985629926986068,
+      "grad_norm": 12.75,
+      "learning_rate": 2.6714574681106945e-06,
+      "loss": 1.5529,
+      "step": 75580
+    },
+    {
+      "epoch": 2.1991447770311545,
+      "grad_norm": 13.4375,
+      "learning_rate": 2.6695181864380796e-06,
+      "loss": 1.5082,
+      "step": 75600
+    },
+    {
+      "epoch": 2.1997265613637023,
+      "grad_norm": 12.0625,
+      "learning_rate": 2.6675789047654647e-06,
+      "loss": 1.6554,
+      "step": 75620
+    },
+    {
+      "epoch": 2.2003083456962504,
+      "grad_norm": 14.5,
+      "learning_rate": 2.66563962309285e-06,
+      "loss": 1.5155,
+      "step": 75640
+    },
+    {
+      "epoch": 2.200890130028798,
+      "grad_norm": 10.625,
+      "learning_rate": 2.663700341420235e-06,
+      "loss": 1.4859,
+      "step": 75660
+    },
+    {
+      "epoch": 2.2014719143613464,
+      "grad_norm": 13.0625,
+      "learning_rate": 2.66176105974762e-06,
+      "loss": 1.4973,
+      "step": 75680
+    },
+    {
+      "epoch": 2.202053698693894,
+      "grad_norm": 11.9375,
+      "learning_rate": 2.659821778075005e-06,
+      "loss": 1.4262,
+      "step": 75700
+    },
+    {
+      "epoch": 2.2026354830264423,
+      "grad_norm": 14.0625,
+      "learning_rate": 2.6578824964023902e-06,
+      "loss": 1.4614,
+      "step": 75720
+    },
+    {
+      "epoch": 2.20321726735899,
+      "grad_norm": 12.0,
+      "learning_rate": 2.6559432147297753e-06,
+      "loss": 1.5203,
+      "step": 75740
+    },
+    {
+      "epoch": 2.203799051691538,
+      "grad_norm": 13.0625,
+      "learning_rate": 2.6540039330571604e-06,
+      "loss": 1.5362,
+      "step": 75760
+    },
+    {
+      "epoch": 2.204380836024086,
+      "grad_norm": 12.8125,
+      "learning_rate": 2.6520646513845455e-06,
+      "loss": 1.5148,
+      "step": 75780
+    },
+    {
+      "epoch": 2.2049626203566337,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.6501253697119306e-06,
+      "loss": 1.4596,
+      "step": 75800
+    },
+    {
+      "epoch": 2.205544404689182,
+      "grad_norm": 14.0,
+      "learning_rate": 2.6481860880393157e-06,
+      "loss": 1.5355,
+      "step": 75820
+    },
+    {
+      "epoch": 2.2061261890217296,
+      "grad_norm": 14.75,
+      "learning_rate": 2.646246806366701e-06,
+      "loss": 1.5136,
+      "step": 75840
+    },
+    {
+      "epoch": 2.2067079733542774,
+      "grad_norm": 11.4375,
+      "learning_rate": 2.644307524694086e-06,
+      "loss": 1.4803,
+      "step": 75860
+    },
+    {
+      "epoch": 2.2072897576868256,
+      "grad_norm": 12.9375,
+      "learning_rate": 2.642368243021471e-06,
+      "loss": 1.5108,
+      "step": 75880
+    },
+    {
+      "epoch": 2.2078715420193733,
+      "grad_norm": 13.1875,
+      "learning_rate": 2.640428961348856e-06,
+      "loss": 1.4315,
+      "step": 75900
+    },
+    {
+      "epoch": 2.2084533263519215,
+      "grad_norm": 12.3125,
+      "learning_rate": 2.638489679676241e-06,
+      "loss": 1.487,
+      "step": 75920
+    },
+    {
+      "epoch": 2.2090351106844692,
+      "grad_norm": 11.9375,
+      "learning_rate": 2.6365503980036263e-06,
+      "loss": 1.486,
+      "step": 75940
+    },
+    {
+      "epoch": 2.2096168950170174,
+      "grad_norm": 11.9375,
+      "learning_rate": 2.634611116331011e-06,
+      "loss": 1.574,
+      "step": 75960
+    },
+    {
+      "epoch": 2.210198679349565,
+      "grad_norm": 13.8125,
+      "learning_rate": 2.632671834658396e-06,
+      "loss": 1.4623,
+      "step": 75980
+    },
+    {
+      "epoch": 2.210780463682113,
+      "grad_norm": 13.125,
+      "learning_rate": 2.630732552985781e-06,
+      "loss": 1.4845,
+      "step": 76000
+    },
+    {
+      "epoch": 2.211362248014661,
+      "grad_norm": 16.125,
+      "learning_rate": 2.6287932713131663e-06,
+      "loss": 1.4634,
+      "step": 76020
+    },
+    {
+      "epoch": 2.211944032347209,
+      "grad_norm": 15.5625,
+      "learning_rate": 2.6268539896405514e-06,
+      "loss": 1.4881,
+      "step": 76040
+    },
+    {
+      "epoch": 2.212525816679757,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.6249147079679365e-06,
+      "loss": 1.5763,
+      "step": 76060
+    },
+    {
+      "epoch": 2.2131076010123047,
+      "grad_norm": 10.8125,
+      "learning_rate": 2.6229754262953216e-06,
+      "loss": 1.3982,
+      "step": 76080
+    },
+    {
+      "epoch": 2.2136893853448525,
+      "grad_norm": 11.9375,
+      "learning_rate": 2.6210361446227067e-06,
+      "loss": 1.5016,
+      "step": 76100
+    },
+    {
+      "epoch": 2.2142711696774007,
+      "grad_norm": 13.5625,
+      "learning_rate": 2.6190968629500918e-06,
+      "loss": 1.5301,
+      "step": 76120
+    },
+    {
+      "epoch": 2.2148529540099484,
+      "grad_norm": 14.25,
+      "learning_rate": 2.617157581277477e-06,
+      "loss": 1.5179,
+      "step": 76140
+    },
+    {
+      "epoch": 2.2154347383424966,
+      "grad_norm": 10.9375,
+      "learning_rate": 2.615218299604862e-06,
+      "loss": 1.4687,
+      "step": 76160
+    },
+    {
+      "epoch": 2.2160165226750443,
+      "grad_norm": 12.75,
+      "learning_rate": 2.613279017932247e-06,
+      "loss": 1.4994,
+      "step": 76180
+    },
+    {
+      "epoch": 2.216598307007592,
+      "grad_norm": 12.4375,
+      "learning_rate": 2.611339736259632e-06,
+      "loss": 1.4692,
+      "step": 76200
+    },
+    {
+      "epoch": 2.2171800913401403,
+      "grad_norm": 13.875,
+      "learning_rate": 2.6094004545870172e-06,
+      "loss": 1.5165,
+      "step": 76220
+    },
+    {
+      "epoch": 2.217761875672688,
+      "grad_norm": 13.1875,
+      "learning_rate": 2.6074611729144023e-06,
+      "loss": 1.4748,
+      "step": 76240
+    },
+    {
+      "epoch": 2.218343660005236,
+      "grad_norm": 14.625,
+      "learning_rate": 2.6055218912417874e-06,
+      "loss": 1.4563,
+      "step": 76260
+    },
+    {
+      "epoch": 2.218925444337784,
+      "grad_norm": 13.0625,
+      "learning_rate": 2.6035826095691725e-06,
+      "loss": 1.5219,
+      "step": 76280
+    },
+    {
+      "epoch": 2.2195072286703317,
+      "grad_norm": 14.0,
+      "learning_rate": 2.6016433278965576e-06,
+      "loss": 1.5895,
+      "step": 76300
+    },
+    {
+      "epoch": 2.22008901300288,
+      "grad_norm": 14.5625,
+      "learning_rate": 2.5997040462239427e-06,
+      "loss": 1.534,
+      "step": 76320
+    },
+    {
+      "epoch": 2.2206707973354276,
+      "grad_norm": 13.3125,
+      "learning_rate": 2.597764764551328e-06,
+      "loss": 1.4324,
+      "step": 76340
+    },
+    {
+      "epoch": 2.221252581667976,
+      "grad_norm": 13.5,
+      "learning_rate": 2.595825482878713e-06,
+      "loss": 1.5725,
+      "step": 76360
+    },
+    {
+      "epoch": 2.2218343660005235,
+      "grad_norm": 15.125,
+      "learning_rate": 2.5938862012060976e-06,
+      "loss": 1.4665,
+      "step": 76380
+    },
+    {
+      "epoch": 2.2224161503330717,
+      "grad_norm": 15.875,
+      "learning_rate": 2.5919469195334827e-06,
+      "loss": 1.5412,
+      "step": 76400
+    },
+    {
+      "epoch": 2.2229979346656195,
+      "grad_norm": 12.875,
+      "learning_rate": 2.590007637860868e-06,
+      "loss": 1.452,
+      "step": 76420
+    },
+    {
+      "epoch": 2.223579718998167,
+      "grad_norm": 12.9375,
+      "learning_rate": 2.588068356188253e-06,
+      "loss": 1.5128,
+      "step": 76440
+    },
+    {
+      "epoch": 2.2241615033307154,
+      "grad_norm": 13.0625,
+      "learning_rate": 2.586129074515638e-06,
+      "loss": 1.5441,
+      "step": 76460
+    },
+    {
+      "epoch": 2.224743287663263,
+      "grad_norm": 11.625,
+      "learning_rate": 2.584189792843023e-06,
+      "loss": 1.5348,
+      "step": 76480
+    },
+    {
+      "epoch": 2.2253250719958113,
+      "grad_norm": 12.625,
+      "learning_rate": 2.582250511170408e-06,
+      "loss": 1.4825,
+      "step": 76500
+    },
+    {
+      "epoch": 2.225906856328359,
+      "grad_norm": 12.3125,
+      "learning_rate": 2.5803112294977933e-06,
+      "loss": 1.5192,
+      "step": 76520
+    },
+    {
+      "epoch": 2.2264886406609072,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.5783719478251784e-06,
+      "loss": 1.5104,
+      "step": 76540
+    },
+    {
+      "epoch": 2.227070424993455,
+      "grad_norm": 12.4375,
+      "learning_rate": 2.5764326661525635e-06,
+      "loss": 1.4994,
+      "step": 76560
+    },
+    {
+      "epoch": 2.2276522093260027,
+      "grad_norm": 14.5,
+      "learning_rate": 2.5744933844799486e-06,
+      "loss": 1.5483,
+      "step": 76580
+    },
+    {
+      "epoch": 2.228233993658551,
+      "grad_norm": 16.75,
+      "learning_rate": 2.5725541028073337e-06,
+      "loss": 1.4975,
+      "step": 76600
+    },
+    {
+      "epoch": 2.2288157779910986,
+      "grad_norm": 13.0,
+      "learning_rate": 2.5706148211347188e-06,
+      "loss": 1.4466,
+      "step": 76620
+    },
+    {
+      "epoch": 2.229397562323647,
+      "grad_norm": 12.0,
+      "learning_rate": 2.568675539462104e-06,
+      "loss": 1.444,
+      "step": 76640
+    },
+    {
+      "epoch": 2.2299793466561946,
+      "grad_norm": 15.0,
+      "learning_rate": 2.566736257789489e-06,
+      "loss": 1.4896,
+      "step": 76660
+    },
+    {
+      "epoch": 2.2305611309887423,
+      "grad_norm": 13.5625,
+      "learning_rate": 2.564796976116874e-06,
+      "loss": 1.4873,
+      "step": 76680
+    },
+    {
+      "epoch": 2.2311429153212905,
+      "grad_norm": 15.0625,
+      "learning_rate": 2.562857694444259e-06,
+      "loss": 1.4281,
+      "step": 76700
+    },
+    {
+      "epoch": 2.2317246996538382,
+      "grad_norm": 13.8125,
+      "learning_rate": 2.5609184127716443e-06,
+      "loss": 1.5257,
+      "step": 76720
+    },
+    {
+      "epoch": 2.2323064839863864,
+      "grad_norm": 14.6875,
+      "learning_rate": 2.5589791310990294e-06,
+      "loss": 1.4407,
+      "step": 76740
+    },
+    {
+      "epoch": 2.232888268318934,
+      "grad_norm": 12.8125,
+      "learning_rate": 2.5570398494264145e-06,
+      "loss": 1.4499,
+      "step": 76760
+    },
+    {
+      "epoch": 2.233470052651482,
+      "grad_norm": 16.25,
+      "learning_rate": 2.555100567753799e-06,
+      "loss": 1.4619,
+      "step": 76780
+    },
+    {
+      "epoch": 2.23405183698403,
+      "grad_norm": 13.5625,
+      "learning_rate": 2.5531612860811842e-06,
+      "loss": 1.5441,
+      "step": 76800
+    },
+    {
+      "epoch": 2.234633621316578,
+      "grad_norm": 11.125,
+      "learning_rate": 2.551222004408569e-06,
+      "loss": 1.5077,
+      "step": 76820
+    },
+    {
+      "epoch": 2.235215405649126,
+      "grad_norm": 12.125,
+      "learning_rate": 2.549282722735954e-06,
+      "loss": 1.5183,
+      "step": 76840
+    },
+    {
+      "epoch": 2.2357971899816738,
+      "grad_norm": 14.3125,
+      "learning_rate": 2.547343441063339e-06,
+      "loss": 1.4919,
+      "step": 76860
+    },
+    {
+      "epoch": 2.2363789743142215,
+      "grad_norm": 16.0,
+      "learning_rate": 2.545404159390724e-06,
+      "loss": 1.544,
+      "step": 76880
+    },
+    {
+      "epoch": 2.2369607586467697,
+      "grad_norm": 13.4375,
+      "learning_rate": 2.5434648777181093e-06,
+      "loss": 1.4385,
+      "step": 76900
+    },
+    {
+      "epoch": 2.2375425429793174,
+      "grad_norm": 12.375,
+      "learning_rate": 2.5415255960454944e-06,
+      "loss": 1.5499,
+      "step": 76920
+    },
+    {
+      "epoch": 2.2381243273118656,
+      "grad_norm": 11.75,
+      "learning_rate": 2.5395863143728795e-06,
+      "loss": 1.5124,
+      "step": 76940
+    },
+    {
+      "epoch": 2.2387061116444134,
+      "grad_norm": 13.875,
+      "learning_rate": 2.5376470327002646e-06,
+      "loss": 1.5421,
+      "step": 76960
+    },
+    {
+      "epoch": 2.2392878959769615,
+      "grad_norm": 14.1875,
+      "learning_rate": 2.5357077510276497e-06,
+      "loss": 1.5072,
+      "step": 76980
+    },
+    {
+      "epoch": 2.2398696803095093,
+      "grad_norm": 13.0,
+      "learning_rate": 2.5337684693550348e-06,
+      "loss": 1.5009,
+      "step": 77000
+    },
+    {
+      "epoch": 2.240451464642057,
+      "grad_norm": 13.1875,
+      "learning_rate": 2.53182918768242e-06,
+      "loss": 1.4685,
+      "step": 77020
+    },
+    {
+      "epoch": 2.241033248974605,
+      "grad_norm": 13.0,
+      "learning_rate": 2.529889906009805e-06,
+      "loss": 1.5346,
+      "step": 77040
+    },
+    {
+      "epoch": 2.241615033307153,
+      "grad_norm": 11.875,
+      "learning_rate": 2.5279506243371896e-06,
+      "loss": 1.4772,
+      "step": 77060
+    },
+    {
+      "epoch": 2.242196817639701,
+      "grad_norm": 14.5625,
+      "learning_rate": 2.5260113426645747e-06,
+      "loss": 1.5247,
+      "step": 77080
+    },
+    {
+      "epoch": 2.242778601972249,
+      "grad_norm": 14.5,
+      "learning_rate": 2.52407206099196e-06,
+      "loss": 1.5067,
+      "step": 77100
+    },
+    {
+      "epoch": 2.2433603863047966,
+      "grad_norm": 13.25,
+      "learning_rate": 2.522132779319345e-06,
+      "loss": 1.5404,
+      "step": 77120
+    },
+    {
+      "epoch": 2.243942170637345,
+      "grad_norm": 11.0625,
+      "learning_rate": 2.52019349764673e-06,
+      "loss": 1.4202,
+      "step": 77140
+    },
+    {
+      "epoch": 2.2445239549698925,
+      "grad_norm": 12.5625,
+      "learning_rate": 2.518254215974115e-06,
+      "loss": 1.512,
+      "step": 77160
+    },
+    {
+      "epoch": 2.2451057393024407,
+      "grad_norm": 12.25,
+      "learning_rate": 2.5163149343015002e-06,
+      "loss": 1.4833,
+      "step": 77180
+    },
+    {
+      "epoch": 2.2456875236349885,
+      "grad_norm": 11.3125,
+      "learning_rate": 2.5143756526288853e-06,
+      "loss": 1.4444,
+      "step": 77200
+    },
+    {
+      "epoch": 2.2462693079675367,
+      "grad_norm": 14.9375,
+      "learning_rate": 2.5124363709562704e-06,
+      "loss": 1.4819,
+      "step": 77220
+    },
+    {
+      "epoch": 2.2468510923000844,
+      "grad_norm": 14.9375,
+      "learning_rate": 2.5104970892836555e-06,
+      "loss": 1.4552,
+      "step": 77240
+    },
+    {
+      "epoch": 2.247432876632632,
+      "grad_norm": 12.625,
+      "learning_rate": 2.5085578076110406e-06,
+      "loss": 1.6139,
+      "step": 77260
+    },
+    {
+      "epoch": 2.2480146609651803,
+      "grad_norm": 15.8125,
+      "learning_rate": 2.5066185259384257e-06,
+      "loss": 1.5939,
+      "step": 77280
+    },
+    {
+      "epoch": 2.248596445297728,
+      "grad_norm": 10.625,
+      "learning_rate": 2.504679244265811e-06,
+      "loss": 1.4841,
+      "step": 77300
+    },
+    {
+      "epoch": 2.2491782296302762,
+      "grad_norm": 13.125,
+      "learning_rate": 2.502739962593196e-06,
+      "loss": 1.5859,
+      "step": 77320
+    },
+    {
+      "epoch": 2.249760013962824,
+      "grad_norm": 13.5,
+      "learning_rate": 2.500800680920581e-06,
+      "loss": 1.4813,
+      "step": 77340
+    },
+    {
+      "epoch": 2.2503417982953717,
+      "grad_norm": 14.125,
+      "learning_rate": 2.498861399247966e-06,
+      "loss": 1.4882,
+      "step": 77360
+    },
+    {
+      "epoch": 2.25092358262792,
+      "grad_norm": 14.1875,
+      "learning_rate": 2.496922117575351e-06,
+      "loss": 1.445,
+      "step": 77380
+    },
+    {
+      "epoch": 2.2515053669604677,
+      "grad_norm": 12.625,
+      "learning_rate": 2.4949828359027363e-06,
+      "loss": 1.5974,
+      "step": 77400
+    },
+    {
+      "epoch": 2.252087151293016,
+      "grad_norm": 12.8125,
+      "learning_rate": 2.4930435542301214e-06,
+      "loss": 1.562,
+      "step": 77420
+    },
+    {
+      "epoch": 2.2526689356255636,
+      "grad_norm": 14.4375,
+      "learning_rate": 2.4911042725575065e-06,
+      "loss": 1.5872,
+      "step": 77440
+    },
+    {
+      "epoch": 2.2532507199581113,
+      "grad_norm": 14.4375,
+      "learning_rate": 2.4891649908848916e-06,
+      "loss": 1.5831,
+      "step": 77460
+    },
+    {
+      "epoch": 2.2538325042906595,
+      "grad_norm": 11.75,
+      "learning_rate": 2.4872257092122763e-06,
+      "loss": 1.5468,
+      "step": 77480
+    },
+    {
+      "epoch": 2.2544142886232073,
+      "grad_norm": 12.875,
+      "learning_rate": 2.4852864275396614e-06,
+      "loss": 1.4803,
+      "step": 77500
+    },
+    {
+      "epoch": 2.2549960729557554,
+      "grad_norm": 12.375,
+      "learning_rate": 2.4833471458670465e-06,
+      "loss": 1.4564,
+      "step": 77520
+    },
+    {
+      "epoch": 2.255577857288303,
+      "grad_norm": 13.0625,
+      "learning_rate": 2.4814078641944316e-06,
+      "loss": 1.5319,
+      "step": 77540
+    },
+    {
+      "epoch": 2.256159641620851,
+      "grad_norm": 16.375,
+      "learning_rate": 2.4794685825218167e-06,
+      "loss": 1.5106,
+      "step": 77560
+    },
+    {
+      "epoch": 2.256741425953399,
+      "grad_norm": 13.8125,
+      "learning_rate": 2.4775293008492018e-06,
+      "loss": 1.5285,
+      "step": 77580
+    },
+    {
+      "epoch": 2.257323210285947,
+      "grad_norm": 12.375,
+      "learning_rate": 2.475590019176587e-06,
+      "loss": 1.47,
+      "step": 77600
+    },
+    {
+      "epoch": 2.257904994618495,
+      "grad_norm": 13.6875,
+      "learning_rate": 2.473650737503972e-06,
+      "loss": 1.4906,
+      "step": 77620
+    },
+    {
+      "epoch": 2.2584867789510428,
+      "grad_norm": 15.375,
+      "learning_rate": 2.471711455831357e-06,
+      "loss": 1.5675,
+      "step": 77640
+    },
+    {
+      "epoch": 2.259068563283591,
+      "grad_norm": 13.5,
+      "learning_rate": 2.469772174158742e-06,
+      "loss": 1.3821,
+      "step": 77660
+    },
+    {
+      "epoch": 2.2596503476161387,
+      "grad_norm": 13.625,
+      "learning_rate": 2.4678328924861272e-06,
+      "loss": 1.5581,
+      "step": 77680
+    },
+    {
+      "epoch": 2.2602321319486864,
+      "grad_norm": 12.0,
+      "learning_rate": 2.4658936108135123e-06,
+      "loss": 1.4272,
+      "step": 77700
+    },
+    {
+      "epoch": 2.2608139162812346,
+      "grad_norm": 13.0,
+      "learning_rate": 2.4639543291408974e-06,
+      "loss": 1.4901,
+      "step": 77720
+    },
+    {
+      "epoch": 2.2613957006137824,
+      "grad_norm": 10.625,
+      "learning_rate": 2.4620150474682825e-06,
+      "loss": 1.4473,
+      "step": 77740
+    },
+    {
+      "epoch": 2.2619774849463306,
+      "grad_norm": 12.875,
+      "learning_rate": 2.4600757657956676e-06,
+      "loss": 1.5642,
+      "step": 77760
+    },
+    {
+      "epoch": 2.2625592692788783,
+      "grad_norm": 13.9375,
+      "learning_rate": 2.4581364841230527e-06,
+      "loss": 1.4057,
+      "step": 77780
+    },
+    {
+      "epoch": 2.2631410536114265,
+      "grad_norm": 11.6875,
+      "learning_rate": 2.456197202450438e-06,
+      "loss": 1.5791,
+      "step": 77800
+    },
+    {
+      "epoch": 2.263722837943974,
+      "grad_norm": 12.3125,
+      "learning_rate": 2.454257920777823e-06,
+      "loss": 1.4757,
+      "step": 77820
+    },
+    {
+      "epoch": 2.264304622276522,
+      "grad_norm": 13.5,
+      "learning_rate": 2.452318639105208e-06,
+      "loss": 1.4728,
+      "step": 77840
+    },
+    {
+      "epoch": 2.26488640660907,
+      "grad_norm": 12.625,
+      "learning_rate": 2.450379357432593e-06,
+      "loss": 1.5004,
+      "step": 77860
+    },
+    {
+      "epoch": 2.265468190941618,
+      "grad_norm": 13.0,
+      "learning_rate": 2.448440075759978e-06,
+      "loss": 1.4871,
+      "step": 77880
+    },
+    {
+      "epoch": 2.266049975274166,
+      "grad_norm": 15.1875,
+      "learning_rate": 2.446500794087363e-06,
+      "loss": 1.5657,
+      "step": 77900
+    },
+    {
+      "epoch": 2.266631759606714,
+      "grad_norm": 11.9375,
+      "learning_rate": 2.444561512414748e-06,
+      "loss": 1.553,
+      "step": 77920
+    },
+    {
+      "epoch": 2.2672135439392616,
+      "grad_norm": 12.4375,
+      "learning_rate": 2.442622230742133e-06,
+      "loss": 1.4405,
+      "step": 77940
+    },
+    {
+      "epoch": 2.2677953282718097,
+      "grad_norm": 14.3125,
+      "learning_rate": 2.440682949069518e-06,
+      "loss": 1.5006,
+      "step": 77960
+    },
+    {
+      "epoch": 2.2683771126043575,
+      "grad_norm": 12.5,
+      "learning_rate": 2.4387436673969033e-06,
+      "loss": 1.4933,
+      "step": 77980
+    },
+    {
+      "epoch": 2.2689588969369057,
+      "grad_norm": 13.75,
+      "learning_rate": 2.4368043857242884e-06,
+      "loss": 1.452,
+      "step": 78000
+    },
+    {
+      "epoch": 2.2695406812694534,
+      "grad_norm": 14.75,
+      "learning_rate": 2.4348651040516735e-06,
+      "loss": 1.6161,
+      "step": 78020
+    },
+    {
+      "epoch": 2.270122465602001,
+      "grad_norm": 11.8125,
+      "learning_rate": 2.4329258223790586e-06,
+      "loss": 1.5615,
+      "step": 78040
+    },
+    {
+      "epoch": 2.2707042499345493,
+      "grad_norm": 13.0,
+      "learning_rate": 2.4309865407064437e-06,
+      "loss": 1.468,
+      "step": 78060
+    },
+    {
+      "epoch": 2.271286034267097,
+      "grad_norm": 10.1875,
+      "learning_rate": 2.4290472590338288e-06,
+      "loss": 1.5075,
+      "step": 78080
+    },
+    {
+      "epoch": 2.2718678185996453,
+      "grad_norm": 17.5,
+      "learning_rate": 2.427107977361214e-06,
+      "loss": 1.5169,
+      "step": 78100
+    },
+    {
+      "epoch": 2.272449602932193,
+      "grad_norm": 15.125,
+      "learning_rate": 2.425168695688599e-06,
+      "loss": 1.5831,
+      "step": 78120
+    },
+    {
+      "epoch": 2.2730313872647407,
+      "grad_norm": 13.6875,
+      "learning_rate": 2.423229414015984e-06,
+      "loss": 1.4668,
+      "step": 78140
+    },
+    {
+      "epoch": 2.273613171597289,
+      "grad_norm": 13.0,
+      "learning_rate": 2.421290132343369e-06,
+      "loss": 1.5313,
+      "step": 78160
+    },
+    {
+      "epoch": 2.2741949559298367,
+      "grad_norm": 12.4375,
+      "learning_rate": 2.4193508506707543e-06,
+      "loss": 1.5282,
+      "step": 78180
+    },
+    {
+      "epoch": 2.274776740262385,
+      "grad_norm": 13.6875,
+      "learning_rate": 2.4174115689981394e-06,
+      "loss": 1.5503,
+      "step": 78200
+    },
+    {
+      "epoch": 2.2753585245949326,
+      "grad_norm": 16.0,
+      "learning_rate": 2.4154722873255245e-06,
+      "loss": 1.5384,
+      "step": 78220
+    },
+    {
+      "epoch": 2.2759403089274803,
+      "grad_norm": 15.75,
+      "learning_rate": 2.4135330056529095e-06,
+      "loss": 1.4713,
+      "step": 78240
+    },
+    {
+      "epoch": 2.2765220932600285,
+      "grad_norm": 13.3125,
+      "learning_rate": 2.4115937239802946e-06,
+      "loss": 1.4791,
+      "step": 78260
+    },
+    {
+      "epoch": 2.2771038775925763,
+      "grad_norm": 14.75,
+      "learning_rate": 2.4096544423076797e-06,
+      "loss": 1.5024,
+      "step": 78280
+    },
+    {
+      "epoch": 2.2776856619251244,
+      "grad_norm": 14.75,
+      "learning_rate": 2.4077151606350644e-06,
+      "loss": 1.4897,
+      "step": 78300
+    },
+    {
+      "epoch": 2.278267446257672,
+      "grad_norm": 14.9375,
+      "learning_rate": 2.4057758789624495e-06,
+      "loss": 1.4642,
+      "step": 78320
+    },
+    {
+      "epoch": 2.2788492305902204,
+      "grad_norm": 13.5,
+      "learning_rate": 2.4038365972898346e-06,
+      "loss": 1.4546,
+      "step": 78340
+    },
+    {
+      "epoch": 2.279431014922768,
+      "grad_norm": 15.75,
+      "learning_rate": 2.4018973156172197e-06,
+      "loss": 1.544,
+      "step": 78360
+    },
+    {
+      "epoch": 2.2800127992553163,
+      "grad_norm": 16.0,
+      "learning_rate": 2.399958033944605e-06,
+      "loss": 1.477,
+      "step": 78380
+    },
+    {
+      "epoch": 2.280594583587864,
+      "grad_norm": 14.6875,
+      "learning_rate": 2.39801875227199e-06,
+      "loss": 1.5448,
+      "step": 78400
+    },
+    {
+      "epoch": 2.281176367920412,
+      "grad_norm": 14.5,
+      "learning_rate": 2.396079470599375e-06,
+      "loss": 1.4392,
+      "step": 78420
+    },
+    {
+      "epoch": 2.28175815225296,
+      "grad_norm": 15.125,
+      "learning_rate": 2.3941401889267597e-06,
+      "loss": 1.5462,
+      "step": 78440
+    },
+    {
+      "epoch": 2.2823399365855077,
+      "grad_norm": 16.375,
+      "learning_rate": 2.3922009072541448e-06,
+      "loss": 1.5173,
+      "step": 78460
+    },
+    {
+      "epoch": 2.282921720918056,
+      "grad_norm": 13.5,
+      "learning_rate": 2.39026162558153e-06,
+      "loss": 1.4916,
+      "step": 78480
+    },
+    {
+      "epoch": 2.2835035052506036,
+      "grad_norm": 12.875,
+      "learning_rate": 2.388322343908915e-06,
+      "loss": 1.4347,
+      "step": 78500
+    },
+    {
+      "epoch": 2.2840852895831514,
+      "grad_norm": 11.3125,
+      "learning_rate": 2.3863830622363e-06,
+      "loss": 1.5101,
+      "step": 78520
+    },
+    {
+      "epoch": 2.2846670739156996,
+      "grad_norm": 15.0625,
+      "learning_rate": 2.384443780563685e-06,
+      "loss": 1.4632,
+      "step": 78540
+    },
+    {
+      "epoch": 2.2852488582482473,
+      "grad_norm": 14.875,
+      "learning_rate": 2.3825044988910703e-06,
+      "loss": 1.5455,
+      "step": 78560
+    },
+    {
+      "epoch": 2.2858306425807955,
+      "grad_norm": 14.125,
+      "learning_rate": 2.3805652172184554e-06,
+      "loss": 1.4849,
+      "step": 78580
+    },
+    {
+      "epoch": 2.2864124269133432,
+      "grad_norm": 12.5625,
+      "learning_rate": 2.3786259355458405e-06,
+      "loss": 1.5309,
+      "step": 78600
+    },
+    {
+      "epoch": 2.286994211245891,
+      "grad_norm": 11.75,
+      "learning_rate": 2.3766866538732256e-06,
+      "loss": 1.5735,
+      "step": 78620
+    },
+    {
+      "epoch": 2.287575995578439,
+      "grad_norm": 14.5,
+      "learning_rate": 2.3747473722006107e-06,
+      "loss": 1.5037,
+      "step": 78640
+    },
+    {
+      "epoch": 2.288157779910987,
+      "grad_norm": 13.0,
+      "learning_rate": 2.3728080905279957e-06,
+      "loss": 1.4424,
+      "step": 78660
+    },
+    {
+      "epoch": 2.288739564243535,
+      "grad_norm": 14.25,
+      "learning_rate": 2.370868808855381e-06,
+      "loss": 1.4635,
+      "step": 78680
+    },
+    {
+      "epoch": 2.289321348576083,
+      "grad_norm": 13.125,
+      "learning_rate": 2.368929527182766e-06,
+      "loss": 1.4897,
+      "step": 78700
+    },
+    {
+      "epoch": 2.2899031329086306,
+      "grad_norm": 14.3125,
+      "learning_rate": 2.366990245510151e-06,
+      "loss": 1.4852,
+      "step": 78720
+    },
+    {
+      "epoch": 2.2904849172411788,
+      "grad_norm": 13.875,
+      "learning_rate": 2.365050963837536e-06,
+      "loss": 1.5879,
+      "step": 78740
+    },
+    {
+      "epoch": 2.2910667015737265,
+      "grad_norm": 12.375,
+      "learning_rate": 2.3631116821649212e-06,
+      "loss": 1.5611,
+      "step": 78760
+    },
+    {
+      "epoch": 2.2916484859062747,
+      "grad_norm": 13.75,
+      "learning_rate": 2.3611724004923063e-06,
+      "loss": 1.5068,
+      "step": 78780
+    },
+    {
+      "epoch": 2.2922302702388224,
+      "grad_norm": 15.0,
+      "learning_rate": 2.3592331188196914e-06,
+      "loss": 1.4878,
+      "step": 78800
+    },
+    {
+      "epoch": 2.29281205457137,
+      "grad_norm": 17.5,
+      "learning_rate": 2.3572938371470765e-06,
+      "loss": 1.3869,
+      "step": 78820
+    },
+    {
+      "epoch": 2.2933938389039183,
+      "grad_norm": 15.25,
+      "learning_rate": 2.355354555474461e-06,
+      "loss": 1.5822,
+      "step": 78840
+    },
+    {
+      "epoch": 2.293975623236466,
+      "grad_norm": 12.25,
+      "learning_rate": 2.3534152738018463e-06,
+      "loss": 1.5676,
+      "step": 78860
+    },
+    {
+      "epoch": 2.2945574075690143,
+      "grad_norm": 10.375,
+      "learning_rate": 2.3514759921292314e-06,
+      "loss": 1.4993,
+      "step": 78880
+    },
+    {
+      "epoch": 2.295139191901562,
+      "grad_norm": 13.25,
+      "learning_rate": 2.3495367104566165e-06,
+      "loss": 1.4298,
+      "step": 78900
+    },
+    {
+      "epoch": 2.29572097623411,
+      "grad_norm": 12.4375,
+      "learning_rate": 2.3475974287840016e-06,
+      "loss": 1.5453,
+      "step": 78920
+    },
+    {
+      "epoch": 2.296302760566658,
+      "grad_norm": 15.3125,
+      "learning_rate": 2.3456581471113867e-06,
+      "loss": 1.4572,
+      "step": 78940
+    },
+    {
+      "epoch": 2.2968845448992057,
+      "grad_norm": 16.5,
+      "learning_rate": 2.3437188654387718e-06,
+      "loss": 1.4612,
+      "step": 78960
+    },
+    {
+      "epoch": 2.297466329231754,
+      "grad_norm": 13.375,
+      "learning_rate": 2.341779583766157e-06,
+      "loss": 1.584,
+      "step": 78980
+    },
+    {
+      "epoch": 2.2980481135643016,
+      "grad_norm": 14.125,
+      "learning_rate": 2.339840302093542e-06,
+      "loss": 1.4853,
+      "step": 79000
+    },
+    {
+      "epoch": 2.29862989789685,
+      "grad_norm": 9.25,
+      "learning_rate": 2.337901020420927e-06,
+      "loss": 1.4815,
+      "step": 79020
+    },
+    {
+      "epoch": 2.2992116822293975,
+      "grad_norm": 12.25,
+      "learning_rate": 2.335961738748312e-06,
+      "loss": 1.4859,
+      "step": 79040
+    },
+    {
+      "epoch": 2.2997934665619457,
+      "grad_norm": 14.0625,
+      "learning_rate": 2.3340224570756973e-06,
+      "loss": 1.5659,
+      "step": 79060
+    },
+    {
+      "epoch": 2.3003752508944935,
+      "grad_norm": 13.6875,
+      "learning_rate": 2.3320831754030824e-06,
+      "loss": 1.498,
+      "step": 79080
+    },
+    {
+      "epoch": 2.300957035227041,
+      "grad_norm": 12.8125,
+      "learning_rate": 2.3301438937304675e-06,
+      "loss": 1.4827,
+      "step": 79100
+    },
+    {
+      "epoch": 2.3015388195595894,
+      "grad_norm": 12.0625,
+      "learning_rate": 2.3282046120578526e-06,
+      "loss": 1.4264,
+      "step": 79120
+    },
+    {
+      "epoch": 2.302120603892137,
+      "grad_norm": 15.3125,
+      "learning_rate": 2.3262653303852377e-06,
+      "loss": 1.4727,
+      "step": 79140
+    },
+    {
+      "epoch": 2.3027023882246853,
+      "grad_norm": 14.25,
+      "learning_rate": 2.3243260487126228e-06,
+      "loss": 1.4976,
+      "step": 79160
+    },
+    {
+      "epoch": 2.303284172557233,
+      "grad_norm": 14.125,
+      "learning_rate": 2.322386767040008e-06,
+      "loss": 1.4303,
+      "step": 79180
+    },
+    {
+      "epoch": 2.303865956889781,
+      "grad_norm": 17.875,
+      "learning_rate": 2.320447485367393e-06,
+      "loss": 1.4618,
+      "step": 79200
+    },
+    {
+      "epoch": 2.304447741222329,
+      "grad_norm": 14.0625,
+      "learning_rate": 2.318508203694778e-06,
+      "loss": 1.463,
+      "step": 79220
+    },
+    {
+      "epoch": 2.3050295255548767,
+      "grad_norm": 12.6875,
+      "learning_rate": 2.3165689220221627e-06,
+      "loss": 1.4862,
+      "step": 79240
+    },
+    {
+      "epoch": 2.305611309887425,
+      "grad_norm": 14.5625,
+      "learning_rate": 2.314629640349548e-06,
+      "loss": 1.4419,
+      "step": 79260
+    },
+    {
+      "epoch": 2.3061930942199727,
+      "grad_norm": 11.75,
+      "learning_rate": 2.312690358676933e-06,
+      "loss": 1.4422,
+      "step": 79280
+    },
+    {
+      "epoch": 2.3067748785525204,
+      "grad_norm": 14.5,
+      "learning_rate": 2.310751077004318e-06,
+      "loss": 1.5241,
+      "step": 79300
+    },
+    {
+      "epoch": 2.3073566628850686,
+      "grad_norm": 13.25,
+      "learning_rate": 2.308811795331703e-06,
+      "loss": 1.4881,
+      "step": 79320
+    },
+    {
+      "epoch": 2.3079384472176163,
+      "grad_norm": 13.25,
+      "learning_rate": 2.3068725136590882e-06,
+      "loss": 1.5045,
+      "step": 79340
+    },
+    {
+      "epoch": 2.3085202315501645,
+      "grad_norm": 13.8125,
+      "learning_rate": 2.3049332319864733e-06,
+      "loss": 1.4625,
+      "step": 79360
+    },
+    {
+      "epoch": 2.3091020158827122,
+      "grad_norm": 13.75,
+      "learning_rate": 2.3029939503138584e-06,
+      "loss": 1.5605,
+      "step": 79380
+    },
+    {
+      "epoch": 2.30968380021526,
+      "grad_norm": 13.625,
+      "learning_rate": 2.3010546686412435e-06,
+      "loss": 1.4955,
+      "step": 79400
+    },
+    {
+      "epoch": 2.310265584547808,
+      "grad_norm": 13.4375,
+      "learning_rate": 2.2991153869686286e-06,
+      "loss": 1.5451,
+      "step": 79420
+    },
+    {
+      "epoch": 2.310847368880356,
+      "grad_norm": 15.5,
+      "learning_rate": 2.2971761052960137e-06,
+      "loss": 1.4791,
+      "step": 79440
+    },
+    {
+      "epoch": 2.311429153212904,
+      "grad_norm": 13.1875,
+      "learning_rate": 2.295236823623399e-06,
+      "loss": 1.4711,
+      "step": 79460
+    },
+    {
+      "epoch": 2.312010937545452,
+      "grad_norm": 14.375,
+      "learning_rate": 2.2932975419507835e-06,
+      "loss": 1.5716,
+      "step": 79480
+    },
+    {
+      "epoch": 2.3125927218779996,
+      "grad_norm": 13.5,
+      "learning_rate": 2.2913582602781686e-06,
+      "loss": 1.5021,
+      "step": 79500
+    },
+    {
+      "epoch": 2.3131745062105478,
+      "grad_norm": 15.0625,
+      "learning_rate": 2.2894189786055537e-06,
+      "loss": 1.4842,
+      "step": 79520
+    },
+    {
+      "epoch": 2.3137562905430955,
+      "grad_norm": 14.25,
+      "learning_rate": 2.2874796969329388e-06,
+      "loss": 1.4506,
+      "step": 79540
+    },
+    {
+      "epoch": 2.3143380748756437,
+      "grad_norm": 11.875,
+      "learning_rate": 2.285540415260324e-06,
+      "loss": 1.5117,
+      "step": 79560
+    },
+    {
+      "epoch": 2.3149198592081914,
+      "grad_norm": 11.8125,
+      "learning_rate": 2.283601133587709e-06,
+      "loss": 1.5294,
+      "step": 79580
+    },
+    {
+      "epoch": 2.3155016435407396,
+      "grad_norm": 12.9375,
+      "learning_rate": 2.281661851915094e-06,
+      "loss": 1.4868,
+      "step": 79600
+    },
+    {
+      "epoch": 2.3160834278732874,
+      "grad_norm": 12.0625,
+      "learning_rate": 2.279722570242479e-06,
+      "loss": 1.5197,
+      "step": 79620
+    },
+    {
+      "epoch": 2.3166652122058355,
+      "grad_norm": 15.0,
+      "learning_rate": 2.2777832885698643e-06,
+      "loss": 1.5152,
+      "step": 79640
+    },
+    {
+      "epoch": 2.3172469965383833,
+      "grad_norm": 15.8125,
+      "learning_rate": 2.2758440068972494e-06,
+      "loss": 1.5211,
+      "step": 79660
+    },
+    {
+      "epoch": 2.317828780870931,
+      "grad_norm": 12.625,
+      "learning_rate": 2.2739047252246344e-06,
+      "loss": 1.4938,
+      "step": 79680
+    },
+    {
+      "epoch": 2.318410565203479,
+      "grad_norm": 13.0,
+      "learning_rate": 2.2719654435520195e-06,
+      "loss": 1.5331,
+      "step": 79700
+    },
+    {
+      "epoch": 2.318992349536027,
+      "grad_norm": 12.8125,
+      "learning_rate": 2.2700261618794046e-06,
+      "loss": 1.4789,
+      "step": 79720
+    },
+    {
+      "epoch": 2.319574133868575,
+      "grad_norm": 16.875,
+      "learning_rate": 2.2680868802067897e-06,
+      "loss": 1.4278,
+      "step": 79740
+    },
+    {
+      "epoch": 2.320155918201123,
+      "grad_norm": 14.5625,
+      "learning_rate": 2.266147598534175e-06,
+      "loss": 1.4458,
+      "step": 79760
+    },
+    {
+      "epoch": 2.3207377025336706,
+      "grad_norm": 13.8125,
+      "learning_rate": 2.26420831686156e-06,
+      "loss": 1.4504,
+      "step": 79780
+    },
+    {
+      "epoch": 2.321319486866219,
+      "grad_norm": 12.875,
+      "learning_rate": 2.2622690351889446e-06,
+      "loss": 1.4409,
+      "step": 79800
+    },
+    {
+      "epoch": 2.3219012711987665,
+      "grad_norm": 11.75,
+      "learning_rate": 2.2603297535163297e-06,
+      "loss": 1.51,
+      "step": 79820
+    },
+    {
+      "epoch": 2.3224830555313147,
+      "grad_norm": 15.625,
+      "learning_rate": 2.258390471843715e-06,
+      "loss": 1.4906,
+      "step": 79840
+    },
+    {
+      "epoch": 2.3230648398638625,
+      "grad_norm": 12.5625,
+      "learning_rate": 2.2564511901711e-06,
+      "loss": 1.526,
+      "step": 79860
+    },
+    {
+      "epoch": 2.32364662419641,
+      "grad_norm": 13.875,
+      "learning_rate": 2.254511908498485e-06,
+      "loss": 1.4035,
+      "step": 79880
+    },
+    {
+      "epoch": 2.3242284085289584,
+      "grad_norm": 15.6875,
+      "learning_rate": 2.25257262682587e-06,
+      "loss": 1.4974,
+      "step": 79900
+    },
+    {
+      "epoch": 2.324810192861506,
+      "grad_norm": 13.125,
+      "learning_rate": 2.250633345153255e-06,
+      "loss": 1.4432,
+      "step": 79920
+    },
+    {
+      "epoch": 2.3253919771940543,
+      "grad_norm": 11.875,
+      "learning_rate": 2.2486940634806403e-06,
+      "loss": 1.4812,
+      "step": 79940
+    },
+    {
+      "epoch": 2.325973761526602,
+      "grad_norm": 12.4375,
+      "learning_rate": 2.2467547818080254e-06,
+      "loss": 1.5583,
+      "step": 79960
+    },
+    {
+      "epoch": 2.32655554585915,
+      "grad_norm": 12.25,
+      "learning_rate": 2.2448155001354105e-06,
+      "loss": 1.4932,
+      "step": 79980
+    },
+    {
+      "epoch": 2.327137330191698,
+      "grad_norm": 13.4375,
+      "learning_rate": 2.2428762184627956e-06,
+      "loss": 1.5036,
+      "step": 80000
+    },
+    {
+      "epoch": 2.3277191145242457,
+      "grad_norm": 13.9375,
+      "learning_rate": 2.2409369367901807e-06,
+      "loss": 1.5853,
+      "step": 80020
+    },
+    {
+      "epoch": 2.328300898856794,
+      "grad_norm": 15.0,
+      "learning_rate": 2.2389976551175658e-06,
+      "loss": 1.492,
+      "step": 80040
+    },
+    {
+      "epoch": 2.3288826831893417,
+      "grad_norm": 18.0,
+      "learning_rate": 2.237058373444951e-06,
+      "loss": 1.4796,
+      "step": 80060
+    },
+    {
+      "epoch": 2.3294644675218894,
+      "grad_norm": 12.25,
+      "learning_rate": 2.235119091772336e-06,
+      "loss": 1.5104,
+      "step": 80080
+    },
+    {
+      "epoch": 2.3300462518544376,
+      "grad_norm": 13.5,
+      "learning_rate": 2.233179810099721e-06,
+      "loss": 1.5557,
+      "step": 80100
+    },
+    {
+      "epoch": 2.3306280361869853,
+      "grad_norm": 16.375,
+      "learning_rate": 2.231240528427106e-06,
+      "loss": 1.5181,
+      "step": 80120
+    },
+    {
+      "epoch": 2.3312098205195335,
+      "grad_norm": 14.875,
+      "learning_rate": 2.2293012467544913e-06,
+      "loss": 1.5075,
+      "step": 80140
+    },
+    {
+      "epoch": 2.3317916048520813,
+      "grad_norm": 11.0,
+      "learning_rate": 2.2273619650818764e-06,
+      "loss": 1.5286,
+      "step": 80160
+    },
+    {
+      "epoch": 2.3323733891846294,
+      "grad_norm": 12.1875,
+      "learning_rate": 2.2254226834092615e-06,
+      "loss": 1.4977,
+      "step": 80180
+    },
+    {
+      "epoch": 2.332955173517177,
+      "grad_norm": 12.0,
+      "learning_rate": 2.223483401736646e-06,
+      "loss": 1.5509,
+      "step": 80200
+    },
+    {
+      "epoch": 2.333536957849725,
+      "grad_norm": 14.6875,
+      "learning_rate": 2.2215441200640312e-06,
+      "loss": 1.5433,
+      "step": 80220
+    },
+    {
+      "epoch": 2.334118742182273,
+      "grad_norm": 13.0625,
+      "learning_rate": 2.2196048383914163e-06,
+      "loss": 1.4002,
+      "step": 80240
+    },
+    {
+      "epoch": 2.334700526514821,
+      "grad_norm": 13.0,
+      "learning_rate": 2.2176655567188014e-06,
+      "loss": 1.4283,
+      "step": 80260
+    },
+    {
+      "epoch": 2.335282310847369,
+      "grad_norm": 12.5625,
+      "learning_rate": 2.2157262750461865e-06,
+      "loss": 1.619,
+      "step": 80280
+    },
+    {
+      "epoch": 2.3358640951799168,
+      "grad_norm": 14.3125,
+      "learning_rate": 2.2137869933735716e-06,
+      "loss": 1.5174,
+      "step": 80300
+    },
+    {
+      "epoch": 2.336445879512465,
+      "grad_norm": 14.625,
+      "learning_rate": 2.2118477117009567e-06,
+      "loss": 1.4691,
+      "step": 80320
+    },
+    {
+      "epoch": 2.3370276638450127,
+      "grad_norm": 14.3125,
+      "learning_rate": 2.209908430028342e-06,
+      "loss": 1.5607,
+      "step": 80340
+    },
+    {
+      "epoch": 2.3376094481775604,
+      "grad_norm": 13.5,
+      "learning_rate": 2.207969148355727e-06,
+      "loss": 1.4788,
+      "step": 80360
+    },
+    {
+      "epoch": 2.3381912325101086,
+      "grad_norm": 12.875,
+      "learning_rate": 2.206029866683112e-06,
+      "loss": 1.5194,
+      "step": 80380
+    },
+    {
+      "epoch": 2.3387730168426564,
+      "grad_norm": 9.4375,
+      "learning_rate": 2.204090585010497e-06,
+      "loss": 1.5032,
+      "step": 80400
+    },
+    {
+      "epoch": 2.3393548011752046,
+      "grad_norm": 14.9375,
+      "learning_rate": 2.202151303337882e-06,
+      "loss": 1.4447,
+      "step": 80420
+    },
+    {
+      "epoch": 2.3399365855077523,
+      "grad_norm": 17.5,
+      "learning_rate": 2.2002120216652673e-06,
+      "loss": 1.4967,
+      "step": 80440
+    },
+    {
+      "epoch": 2.3405183698403,
+      "grad_norm": 12.5,
+      "learning_rate": 2.1982727399926524e-06,
+      "loss": 1.4616,
+      "step": 80460
+    },
+    {
+      "epoch": 2.3411001541728482,
+      "grad_norm": 11.6875,
+      "learning_rate": 2.1963334583200375e-06,
+      "loss": 1.4305,
+      "step": 80480
+    },
+    {
+      "epoch": 2.341681938505396,
+      "grad_norm": 14.6875,
+      "learning_rate": 2.1943941766474226e-06,
+      "loss": 1.5134,
+      "step": 80500
+    },
+    {
+      "epoch": 2.342263722837944,
+      "grad_norm": 12.125,
+      "learning_rate": 2.1924548949748077e-06,
+      "loss": 1.4399,
+      "step": 80520
+    },
+    {
+      "epoch": 2.342845507170492,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.1905156133021924e-06,
+      "loss": 1.4176,
+      "step": 80540
+    },
+    {
+      "epoch": 2.3434272915030396,
+      "grad_norm": 13.3125,
+      "learning_rate": 2.1885763316295775e-06,
+      "loss": 1.5658,
+      "step": 80560
+    },
+    {
+      "epoch": 2.344009075835588,
+      "grad_norm": 11.3125,
+      "learning_rate": 2.1866370499569626e-06,
+      "loss": 1.6313,
+      "step": 80580
+    },
+    {
+      "epoch": 2.3445908601681356,
+      "grad_norm": 15.1875,
+      "learning_rate": 2.1846977682843477e-06,
+      "loss": 1.5179,
+      "step": 80600
+    },
+    {
+      "epoch": 2.3451726445006837,
+      "grad_norm": 13.0625,
+      "learning_rate": 2.1827584866117328e-06,
+      "loss": 1.501,
+      "step": 80620
+    },
+    {
+      "epoch": 2.3457544288332315,
+      "grad_norm": 13.0,
+      "learning_rate": 2.180819204939118e-06,
+      "loss": 1.5677,
+      "step": 80640
+    },
+    {
+      "epoch": 2.3463362131657792,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.178879923266503e-06,
+      "loss": 1.4799,
+      "step": 80660
+    },
+    {
+      "epoch": 2.3469179974983274,
+      "grad_norm": 13.6875,
+      "learning_rate": 2.176940641593888e-06,
+      "loss": 1.437,
+      "step": 80680
+    },
+    {
+      "epoch": 2.347499781830875,
+      "grad_norm": 13.5,
+      "learning_rate": 2.175001359921273e-06,
+      "loss": 1.5217,
+      "step": 80700
+    },
+    {
+      "epoch": 2.3480815661634233,
+      "grad_norm": 13.125,
+      "learning_rate": 2.1730620782486582e-06,
+      "loss": 1.5573,
+      "step": 80720
+    },
+    {
+      "epoch": 2.348663350495971,
+      "grad_norm": 14.6875,
+      "learning_rate": 2.1711227965760433e-06,
+      "loss": 1.4689,
+      "step": 80740
+    },
+    {
+      "epoch": 2.349245134828519,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.169183514903428e-06,
+      "loss": 1.5098,
+      "step": 80760
+    },
+    {
+      "epoch": 2.349826919161067,
+      "grad_norm": 15.0,
+      "learning_rate": 2.167244233230813e-06,
+      "loss": 1.5224,
+      "step": 80780
+    },
+    {
+      "epoch": 2.3504087034936147,
+      "grad_norm": 13.25,
+      "learning_rate": 2.1653049515581982e-06,
+      "loss": 1.5475,
+      "step": 80800
+    },
+    {
+      "epoch": 2.350990487826163,
+      "grad_norm": 14.5,
+      "learning_rate": 2.1633656698855833e-06,
+      "loss": 1.5486,
+      "step": 80820
+    },
+    {
+      "epoch": 2.3515722721587107,
+      "grad_norm": 15.6875,
+      "learning_rate": 2.1614263882129684e-06,
+      "loss": 1.4639,
+      "step": 80840
+    },
+    {
+      "epoch": 2.352154056491259,
+      "grad_norm": 13.3125,
+      "learning_rate": 2.1594871065403535e-06,
+      "loss": 1.4772,
+      "step": 80860
+    },
+    {
+      "epoch": 2.3527358408238066,
+      "grad_norm": 11.875,
+      "learning_rate": 2.1575478248677386e-06,
+      "loss": 1.5233,
+      "step": 80880
+    },
+    {
+      "epoch": 2.353317625156355,
+      "grad_norm": 14.0625,
+      "learning_rate": 2.1556085431951237e-06,
+      "loss": 1.4569,
+      "step": 80900
+    },
+    {
+      "epoch": 2.3538994094889025,
+      "grad_norm": 15.125,
+      "learning_rate": 2.153669261522509e-06,
+      "loss": 1.5201,
+      "step": 80920
+    },
+    {
+      "epoch": 2.3544811938214503,
+      "grad_norm": 13.9375,
+      "learning_rate": 2.151729979849894e-06,
+      "loss": 1.5271,
+      "step": 80940
+    },
+    {
+      "epoch": 2.3550629781539985,
+      "grad_norm": 14.3125,
+      "learning_rate": 2.149790698177279e-06,
+      "loss": 1.525,
+      "step": 80960
+    },
+    {
+      "epoch": 2.355644762486546,
+      "grad_norm": 13.5625,
+      "learning_rate": 2.147851416504664e-06,
+      "loss": 1.5975,
+      "step": 80980
+    },
+    {
+      "epoch": 2.3562265468190944,
+      "grad_norm": 14.1875,
+      "learning_rate": 2.145912134832049e-06,
+      "loss": 1.4742,
+      "step": 81000
+    },
+    {
+      "epoch": 2.356808331151642,
+      "grad_norm": 12.8125,
+      "learning_rate": 2.1439728531594343e-06,
+      "loss": 1.4587,
+      "step": 81020
+    },
+    {
+      "epoch": 2.35739011548419,
+      "grad_norm": 11.5,
+      "learning_rate": 2.1420335714868194e-06,
+      "loss": 1.47,
+      "step": 81040
+    },
+    {
+      "epoch": 2.357971899816738,
+      "grad_norm": 13.75,
+      "learning_rate": 2.1400942898142045e-06,
+      "loss": 1.4474,
+      "step": 81060
+    },
+    {
+      "epoch": 2.358553684149286,
+      "grad_norm": 14.9375,
+      "learning_rate": 2.1381550081415896e-06,
+      "loss": 1.5069,
+      "step": 81080
+    },
+    {
+      "epoch": 2.359135468481834,
+      "grad_norm": 14.0625,
+      "learning_rate": 2.1362157264689747e-06,
+      "loss": 1.5437,
+      "step": 81100
+    },
+    {
+      "epoch": 2.3597172528143817,
+      "grad_norm": 15.875,
+      "learning_rate": 2.1342764447963598e-06,
+      "loss": 1.4676,
+      "step": 81120
+    },
+    {
+      "epoch": 2.3602990371469295,
+      "grad_norm": 13.4375,
+      "learning_rate": 2.132337163123745e-06,
+      "loss": 1.5713,
+      "step": 81140
+    },
+    {
+      "epoch": 2.3608808214794776,
+      "grad_norm": 13.3125,
+      "learning_rate": 2.1303978814511295e-06,
+      "loss": 1.513,
+      "step": 81160
+    },
+    {
+      "epoch": 2.3614626058120254,
+      "grad_norm": 12.5,
+      "learning_rate": 2.1284585997785146e-06,
+      "loss": 1.589,
+      "step": 81180
+    },
+    {
+      "epoch": 2.3620443901445736,
+      "grad_norm": 10.25,
+      "learning_rate": 2.1265193181058997e-06,
+      "loss": 1.5326,
+      "step": 81200
+    },
+    {
+      "epoch": 2.3626261744771213,
+      "grad_norm": 16.25,
+      "learning_rate": 2.124580036433285e-06,
+      "loss": 1.5033,
+      "step": 81220
+    },
+    {
+      "epoch": 2.363207958809669,
+      "grad_norm": 11.75,
+      "learning_rate": 2.12264075476067e-06,
+      "loss": 1.5411,
+      "step": 81240
+    },
+    {
+      "epoch": 2.3637897431422172,
+      "grad_norm": 13.1875,
+      "learning_rate": 2.120701473088055e-06,
+      "loss": 1.5313,
+      "step": 81260
+    },
+    {
+      "epoch": 2.364371527474765,
+      "grad_norm": 12.75,
+      "learning_rate": 2.11876219141544e-06,
+      "loss": 1.4944,
+      "step": 81280
+    },
+    {
+      "epoch": 2.364953311807313,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.1168229097428252e-06,
+      "loss": 1.4613,
+      "step": 81300
+    },
+    {
+      "epoch": 2.365535096139861,
+      "grad_norm": 13.5,
+      "learning_rate": 2.1148836280702103e-06,
+      "loss": 1.5501,
+      "step": 81320
+    },
+    {
+      "epoch": 2.3661168804724086,
+      "grad_norm": 12.75,
+      "learning_rate": 2.1129443463975954e-06,
+      "loss": 1.546,
+      "step": 81340
+    },
+    {
+      "epoch": 2.366698664804957,
+      "grad_norm": 12.875,
+      "learning_rate": 2.1110050647249805e-06,
+      "loss": 1.4815,
+      "step": 81360
+    },
+    {
+      "epoch": 2.3672804491375046,
+      "grad_norm": 12.6875,
+      "learning_rate": 2.1090657830523656e-06,
+      "loss": 1.4589,
+      "step": 81380
+    },
+    {
+      "epoch": 2.3678622334700528,
+      "grad_norm": 11.4375,
+      "learning_rate": 2.1071265013797507e-06,
+      "loss": 1.5355,
+      "step": 81400
+    },
+    {
+      "epoch": 2.3684440178026005,
+      "grad_norm": 13.125,
+      "learning_rate": 2.105187219707136e-06,
+      "loss": 1.5244,
+      "step": 81420
+    },
+    {
+      "epoch": 2.3690258021351487,
+      "grad_norm": 12.1875,
+      "learning_rate": 2.103247938034521e-06,
+      "loss": 1.4817,
+      "step": 81440
+    },
+    {
+      "epoch": 2.3696075864676964,
+      "grad_norm": 13.375,
+      "learning_rate": 2.101308656361906e-06,
+      "loss": 1.4578,
+      "step": 81460
+    },
+    {
+      "epoch": 2.370189370800244,
+      "grad_norm": 12.8125,
+      "learning_rate": 2.099369374689291e-06,
+      "loss": 1.4558,
+      "step": 81480
+    },
+    {
+      "epoch": 2.3707711551327924,
+      "grad_norm": 13.0625,
+      "learning_rate": 2.097430093016676e-06,
+      "loss": 1.5164,
+      "step": 81500
+    },
+    {
+      "epoch": 2.37135293946534,
+      "grad_norm": 13.0625,
+      "learning_rate": 2.0954908113440613e-06,
+      "loss": 1.5401,
+      "step": 81520
+    },
+    {
+      "epoch": 2.3719347237978883,
+      "grad_norm": 14.25,
+      "learning_rate": 2.0935515296714464e-06,
+      "loss": 1.5447,
+      "step": 81540
+    },
+    {
+      "epoch": 2.372516508130436,
+      "grad_norm": 13.6875,
+      "learning_rate": 2.091612247998831e-06,
+      "loss": 1.524,
+      "step": 81560
+    },
+    {
+      "epoch": 2.373098292462984,
+      "grad_norm": 12.125,
+      "learning_rate": 2.089672966326216e-06,
+      "loss": 1.5211,
+      "step": 81580
+    },
+    {
+      "epoch": 2.373680076795532,
+      "grad_norm": 16.625,
+      "learning_rate": 2.0877336846536013e-06,
+      "loss": 1.5303,
+      "step": 81600
+    },
+    {
+      "epoch": 2.3742618611280797,
+      "grad_norm": 14.625,
+      "learning_rate": 2.0857944029809864e-06,
+      "loss": 1.5686,
+      "step": 81620
+    },
+    {
+      "epoch": 2.374843645460628,
+      "grad_norm": 13.4375,
+      "learning_rate": 2.0838551213083715e-06,
+      "loss": 1.5455,
+      "step": 81640
+    },
+    {
+      "epoch": 2.3754254297931756,
+      "grad_norm": 13.0625,
+      "learning_rate": 2.0819158396357566e-06,
+      "loss": 1.4859,
+      "step": 81660
+    },
+    {
+      "epoch": 2.376007214125724,
+      "grad_norm": 11.625,
+      "learning_rate": 2.0799765579631417e-06,
+      "loss": 1.4473,
+      "step": 81680
+    },
+    {
+      "epoch": 2.3765889984582715,
+      "grad_norm": 14.9375,
+      "learning_rate": 2.0780372762905263e-06,
+      "loss": 1.5468,
+      "step": 81700
+    },
+    {
+      "epoch": 2.3771707827908193,
+      "grad_norm": 13.25,
+      "learning_rate": 2.0760979946179114e-06,
+      "loss": 1.4947,
+      "step": 81720
+    },
+    {
+      "epoch": 2.3777525671233675,
+      "grad_norm": 14.25,
+      "learning_rate": 2.0741587129452965e-06,
+      "loss": 1.4489,
+      "step": 81740
+    },
+    {
+      "epoch": 2.378334351455915,
+      "grad_norm": 14.375,
+      "learning_rate": 2.0722194312726816e-06,
+      "loss": 1.4634,
+      "step": 81760
+    },
+    {
+      "epoch": 2.3789161357884634,
+      "grad_norm": 12.875,
+      "learning_rate": 2.0702801496000667e-06,
+      "loss": 1.5208,
+      "step": 81780
+    },
+    {
+      "epoch": 2.379497920121011,
+      "grad_norm": 14.5,
+      "learning_rate": 2.068340867927452e-06,
+      "loss": 1.4866,
+      "step": 81800
+    },
+    {
+      "epoch": 2.380079704453559,
+      "grad_norm": 14.5,
+      "learning_rate": 2.066401586254837e-06,
+      "loss": 1.4926,
+      "step": 81820
+    },
+    {
+      "epoch": 2.380661488786107,
+      "grad_norm": 12.6875,
+      "learning_rate": 2.064462304582222e-06,
+      "loss": 1.5062,
+      "step": 81840
+    },
+    {
+      "epoch": 2.381243273118655,
+      "grad_norm": 15.625,
+      "learning_rate": 2.062523022909607e-06,
+      "loss": 1.4599,
+      "step": 81860
+    },
+    {
+      "epoch": 2.381825057451203,
+      "grad_norm": 12.0,
+      "learning_rate": 2.060583741236992e-06,
+      "loss": 1.516,
+      "step": 81880
+    },
+    {
+      "epoch": 2.3824068417837507,
+      "grad_norm": 13.25,
+      "learning_rate": 2.0586444595643773e-06,
+      "loss": 1.4665,
+      "step": 81900
+    },
+    {
+      "epoch": 2.3829886261162985,
+      "grad_norm": 17.375,
+      "learning_rate": 2.0567051778917624e-06,
+      "loss": 1.5056,
+      "step": 81920
+    },
+    {
+      "epoch": 2.3835704104488467,
+      "grad_norm": 14.125,
+      "learning_rate": 2.0547658962191475e-06,
+      "loss": 1.4422,
+      "step": 81940
+    },
+    {
+      "epoch": 2.3841521947813944,
+      "grad_norm": 13.875,
+      "learning_rate": 2.0528266145465326e-06,
+      "loss": 1.5094,
+      "step": 81960
+    },
+    {
+      "epoch": 2.3847339791139426,
+      "grad_norm": 15.5,
+      "learning_rate": 2.0508873328739177e-06,
+      "loss": 1.6147,
+      "step": 81980
+    },
+    {
+      "epoch": 2.3853157634464903,
+      "grad_norm": 12.6875,
+      "learning_rate": 2.048948051201303e-06,
+      "loss": 1.4948,
+      "step": 82000
+    },
+    {
+      "epoch": 2.385897547779038,
+      "grad_norm": 13.625,
+      "learning_rate": 2.047008769528688e-06,
+      "loss": 1.4844,
+      "step": 82020
+    },
+    {
+      "epoch": 2.3864793321115862,
+      "grad_norm": 13.625,
+      "learning_rate": 2.045069487856073e-06,
+      "loss": 1.5263,
+      "step": 82040
+    },
+    {
+      "epoch": 2.387061116444134,
+      "grad_norm": 10.125,
+      "learning_rate": 2.043130206183458e-06,
+      "loss": 1.4915,
+      "step": 82060
+    },
+    {
+      "epoch": 2.387642900776682,
+      "grad_norm": 12.875,
+      "learning_rate": 2.041190924510843e-06,
+      "loss": 1.4541,
+      "step": 82080
+    },
+    {
+      "epoch": 2.38822468510923,
+      "grad_norm": 13.625,
+      "learning_rate": 2.0392516428382283e-06,
+      "loss": 1.4651,
+      "step": 82100
+    },
+    {
+      "epoch": 2.388806469441778,
+      "grad_norm": 14.0,
+      "learning_rate": 2.037312361165613e-06,
+      "loss": 1.4914,
+      "step": 82120
+    },
+    {
+      "epoch": 2.389388253774326,
+      "grad_norm": 13.5625,
+      "learning_rate": 2.035373079492998e-06,
+      "loss": 1.498,
+      "step": 82140
+    },
+    {
+      "epoch": 2.389970038106874,
+      "grad_norm": 13.5,
+      "learning_rate": 2.033433797820383e-06,
+      "loss": 1.4332,
+      "step": 82160
+    },
+    {
+      "epoch": 2.3905518224394218,
+      "grad_norm": 11.3125,
+      "learning_rate": 2.0314945161477682e-06,
+      "loss": 1.4921,
+      "step": 82180
+    },
+    {
+      "epoch": 2.3911336067719695,
+      "grad_norm": 10.875,
+      "learning_rate": 2.0295552344751533e-06,
+      "loss": 1.4908,
+      "step": 82200
+    },
+    {
+      "epoch": 2.3917153911045177,
+      "grad_norm": 15.625,
+      "learning_rate": 2.0276159528025384e-06,
+      "loss": 1.4405,
+      "step": 82220
+    },
+    {
+      "epoch": 2.3922971754370654,
+      "grad_norm": 12.875,
+      "learning_rate": 2.0256766711299235e-06,
+      "loss": 1.5628,
+      "step": 82240
+    },
+    {
+      "epoch": 2.3928789597696136,
+      "grad_norm": 13.6875,
+      "learning_rate": 2.0237373894573086e-06,
+      "loss": 1.4273,
+      "step": 82260
+    },
+    {
+      "epoch": 2.3934607441021614,
+      "grad_norm": 13.375,
+      "learning_rate": 2.0217981077846937e-06,
+      "loss": 1.5593,
+      "step": 82280
+    },
+    {
+      "epoch": 2.394042528434709,
+      "grad_norm": 14.375,
+      "learning_rate": 2.019858826112079e-06,
+      "loss": 1.5318,
+      "step": 82300
+    },
+    {
+      "epoch": 2.3946243127672573,
+      "grad_norm": 13.5,
+      "learning_rate": 2.017919544439464e-06,
+      "loss": 1.4808,
+      "step": 82320
+    },
+    {
+      "epoch": 2.395206097099805,
+      "grad_norm": 14.5625,
+      "learning_rate": 2.015980262766849e-06,
+      "loss": 1.5295,
+      "step": 82340
+    },
+    {
+      "epoch": 2.395787881432353,
+      "grad_norm": 12.0,
+      "learning_rate": 2.014040981094234e-06,
+      "loss": 1.6113,
+      "step": 82360
+    },
+    {
+      "epoch": 2.396369665764901,
+      "grad_norm": 14.0,
+      "learning_rate": 2.0121016994216192e-06,
+      "loss": 1.5515,
+      "step": 82380
+    },
+    {
+      "epoch": 2.3969514500974487,
+      "grad_norm": 13.25,
+      "learning_rate": 2.0101624177490043e-06,
+      "loss": 1.4164,
+      "step": 82400
+    },
+    {
+      "epoch": 2.397533234429997,
+      "grad_norm": 14.4375,
+      "learning_rate": 2.0082231360763894e-06,
+      "loss": 1.5213,
+      "step": 82420
+    },
+    {
+      "epoch": 2.3981150187625446,
+      "grad_norm": 11.9375,
+      "learning_rate": 2.0062838544037745e-06,
+      "loss": 1.4736,
+      "step": 82440
+    },
+    {
+      "epoch": 2.398696803095093,
+      "grad_norm": 15.0,
+      "learning_rate": 2.0043445727311596e-06,
+      "loss": 1.5123,
+      "step": 82460
+    },
+    {
+      "epoch": 2.3992785874276406,
+      "grad_norm": 14.1875,
+      "learning_rate": 2.0024052910585447e-06,
+      "loss": 1.399,
+      "step": 82480
+    },
+    {
+      "epoch": 2.3998603717601883,
+      "grad_norm": 17.625,
+      "learning_rate": 2.00046600938593e-06,
+      "loss": 1.4898,
+      "step": 82500
+    },
+    {
+      "epoch": 2.4004421560927365,
+      "grad_norm": 13.5,
+      "learning_rate": 1.9985267277133145e-06,
+      "loss": 1.4799,
+      "step": 82520
+    },
+    {
+      "epoch": 2.401023940425284,
+      "grad_norm": 13.5,
+      "learning_rate": 1.9965874460406996e-06,
+      "loss": 1.5474,
+      "step": 82540
+    },
+    {
+      "epoch": 2.4016057247578324,
+      "grad_norm": 14.625,
+      "learning_rate": 1.9946481643680847e-06,
+      "loss": 1.4957,
+      "step": 82560
+    },
+    {
+      "epoch": 2.40218750909038,
+      "grad_norm": 12.375,
+      "learning_rate": 1.9927088826954698e-06,
+      "loss": 1.4922,
+      "step": 82580
+    },
+    {
+      "epoch": 2.402769293422928,
+      "grad_norm": 16.875,
+      "learning_rate": 1.990769601022855e-06,
+      "loss": 1.4538,
+      "step": 82600
+    },
+    {
+      "epoch": 2.403351077755476,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.98883031935024e-06,
+      "loss": 1.474,
+      "step": 82620
+    },
+    {
+      "epoch": 2.403932862088024,
+      "grad_norm": 13.875,
+      "learning_rate": 1.986891037677625e-06,
+      "loss": 1.531,
+      "step": 82640
+    },
+    {
+      "epoch": 2.404514646420572,
+      "grad_norm": 13.5,
+      "learning_rate": 1.9849517560050097e-06,
+      "loss": 1.5763,
+      "step": 82660
+    },
+    {
+      "epoch": 2.4050964307531197,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.983012474332395e-06,
+      "loss": 1.5443,
+      "step": 82680
+    },
+    {
+      "epoch": 2.405678215085668,
+      "grad_norm": 11.125,
+      "learning_rate": 1.98107319265978e-06,
+      "loss": 1.5399,
+      "step": 82700
+    },
+    {
+      "epoch": 2.4062599994182157,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.979133910987165e-06,
+      "loss": 1.5399,
+      "step": 82720
+    },
+    {
+      "epoch": 2.4068417837507634,
+      "grad_norm": 15.625,
+      "learning_rate": 1.97719462931455e-06,
+      "loss": 1.4963,
+      "step": 82740
+    },
+    {
+      "epoch": 2.4074235680833116,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.9752553476419352e-06,
+      "loss": 1.6571,
+      "step": 82760
+    },
+    {
+      "epoch": 2.4080053524158593,
+      "grad_norm": 13.5,
+      "learning_rate": 1.9733160659693203e-06,
+      "loss": 1.4761,
+      "step": 82780
+    },
+    {
+      "epoch": 2.4085871367484075,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.9713767842967054e-06,
+      "loss": 1.5335,
+      "step": 82800
+    },
+    {
+      "epoch": 2.4091689210809553,
+      "grad_norm": 17.625,
+      "learning_rate": 1.9694375026240905e-06,
+      "loss": 1.4916,
+      "step": 82820
+    },
+    {
+      "epoch": 2.4097507054135034,
+      "grad_norm": 16.75,
+      "learning_rate": 1.9674982209514756e-06,
+      "loss": 1.5589,
+      "step": 82840
+    },
+    {
+      "epoch": 2.410332489746051,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.9655589392788607e-06,
+      "loss": 1.4952,
+      "step": 82860
+    },
+    {
+      "epoch": 2.410914274078599,
+      "grad_norm": 12.875,
+      "learning_rate": 1.963619657606246e-06,
+      "loss": 1.4848,
+      "step": 82880
+    },
+    {
+      "epoch": 2.411496058411147,
+      "grad_norm": 13.625,
+      "learning_rate": 1.961680375933631e-06,
+      "loss": 1.4803,
+      "step": 82900
+    },
+    {
+      "epoch": 2.412077842743695,
+      "grad_norm": 15.0,
+      "learning_rate": 1.959741094261016e-06,
+      "loss": 1.5074,
+      "step": 82920
+    },
+    {
+      "epoch": 2.412659627076243,
+      "grad_norm": 15.25,
+      "learning_rate": 1.957801812588401e-06,
+      "loss": 1.4036,
+      "step": 82940
+    },
+    {
+      "epoch": 2.413241411408791,
+      "grad_norm": 11.4375,
+      "learning_rate": 1.955862530915786e-06,
+      "loss": 1.4796,
+      "step": 82960
+    },
+    {
+      "epoch": 2.4138231957413385,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.9539232492431713e-06,
+      "loss": 1.449,
+      "step": 82980
+    },
+    {
+      "epoch": 2.4144049800738867,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.9519839675705564e-06,
+      "loss": 1.4359,
+      "step": 83000
+    },
+    {
+      "epoch": 2.4149867644064344,
+      "grad_norm": 15.0,
+      "learning_rate": 1.9500446858979415e-06,
+      "loss": 1.4583,
+      "step": 83020
+    },
+    {
+      "epoch": 2.4155685487389826,
+      "grad_norm": 16.0,
+      "learning_rate": 1.9481054042253266e-06,
+      "loss": 1.5212,
+      "step": 83040
+    },
+    {
+      "epoch": 2.4161503330715304,
+      "grad_norm": 18.125,
+      "learning_rate": 1.9461661225527117e-06,
+      "loss": 1.5268,
+      "step": 83060
+    },
+    {
+      "epoch": 2.416732117404078,
+      "grad_norm": 12.375,
+      "learning_rate": 1.9442268408800964e-06,
+      "loss": 1.5542,
+      "step": 83080
+    },
+    {
+      "epoch": 2.4173139017366263,
+      "grad_norm": 16.5,
+      "learning_rate": 1.9422875592074815e-06,
+      "loss": 1.3943,
+      "step": 83100
+    },
+    {
+      "epoch": 2.417895686069174,
+      "grad_norm": 12.125,
+      "learning_rate": 1.9403482775348666e-06,
+      "loss": 1.4659,
+      "step": 83120
+    },
+    {
+      "epoch": 2.4184774704017222,
+      "grad_norm": 15.0,
+      "learning_rate": 1.9384089958622517e-06,
+      "loss": 1.4778,
+      "step": 83140
+    },
+    {
+      "epoch": 2.41905925473427,
+      "grad_norm": 14.125,
+      "learning_rate": 1.9364697141896368e-06,
+      "loss": 1.4649,
+      "step": 83160
+    },
+    {
+      "epoch": 2.4196410390668177,
+      "grad_norm": 14.375,
+      "learning_rate": 1.934530432517022e-06,
+      "loss": 1.5323,
+      "step": 83180
+    },
+    {
+      "epoch": 2.420222823399366,
+      "grad_norm": 18.625,
+      "learning_rate": 1.932591150844407e-06,
+      "loss": 1.5303,
+      "step": 83200
+    },
+    {
+      "epoch": 2.4208046077319136,
+      "grad_norm": 11.4375,
+      "learning_rate": 1.930651869171792e-06,
+      "loss": 1.4254,
+      "step": 83220
+    },
+    {
+      "epoch": 2.421386392064462,
+      "grad_norm": 14.5,
+      "learning_rate": 1.928712587499177e-06,
+      "loss": 1.4953,
+      "step": 83240
+    },
+    {
+      "epoch": 2.4219681763970096,
+      "grad_norm": 14.375,
+      "learning_rate": 1.9267733058265622e-06,
+      "loss": 1.4438,
+      "step": 83260
+    },
+    {
+      "epoch": 2.4225499607295573,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.9248340241539473e-06,
+      "loss": 1.4613,
+      "step": 83280
+    },
+    {
+      "epoch": 2.4231317450621055,
+      "grad_norm": 12.0,
+      "learning_rate": 1.9228947424813324e-06,
+      "loss": 1.5023,
+      "step": 83300
+    },
+    {
+      "epoch": 2.4237135293946532,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.9209554608087175e-06,
+      "loss": 1.5418,
+      "step": 83320
+    },
+    {
+      "epoch": 2.4242953137272014,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.9190161791361026e-06,
+      "loss": 1.497,
+      "step": 83340
+    },
+    {
+      "epoch": 2.424877098059749,
+      "grad_norm": 10.0,
+      "learning_rate": 1.9170768974634877e-06,
+      "loss": 1.559,
+      "step": 83360
+    },
+    {
+      "epoch": 2.4254588823922973,
+      "grad_norm": 13.125,
+      "learning_rate": 1.915137615790873e-06,
+      "loss": 1.5427,
+      "step": 83380
+    },
+    {
+      "epoch": 2.426040666724845,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.913198334118258e-06,
+      "loss": 1.4278,
+      "step": 83400
+    },
+    {
+      "epoch": 2.4266224510573933,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.911259052445643e-06,
+      "loss": 1.4889,
+      "step": 83420
+    },
+    {
+      "epoch": 2.427204235389941,
+      "grad_norm": 12.625,
+      "learning_rate": 1.909319770773028e-06,
+      "loss": 1.4222,
+      "step": 83440
+    },
+    {
+      "epoch": 2.4277860197224888,
+      "grad_norm": 15.0,
+      "learning_rate": 1.9073804891004132e-06,
+      "loss": 1.5064,
+      "step": 83460
+    },
+    {
+      "epoch": 2.428367804055037,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.905441207427798e-06,
+      "loss": 1.5303,
+      "step": 83480
+    },
+    {
+      "epoch": 2.4289495883875847,
+      "grad_norm": 14.875,
+      "learning_rate": 1.9035019257551832e-06,
+      "loss": 1.4656,
+      "step": 83500
+    },
+    {
+      "epoch": 2.429531372720133,
+      "grad_norm": 15.125,
+      "learning_rate": 1.9015626440825683e-06,
+      "loss": 1.5907,
+      "step": 83520
+    },
+    {
+      "epoch": 2.4301131570526806,
+      "grad_norm": 12.8125,
+      "learning_rate": 1.8996233624099534e-06,
+      "loss": 1.4374,
+      "step": 83540
+    },
+    {
+      "epoch": 2.4306949413852283,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.8976840807373385e-06,
+      "loss": 1.5289,
+      "step": 83560
+    },
+    {
+      "epoch": 2.4312767257177765,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.8957447990647234e-06,
+      "loss": 1.5548,
+      "step": 83580
+    },
+    {
+      "epoch": 2.4318585100503243,
+      "grad_norm": 13.5,
+      "learning_rate": 1.8938055173921085e-06,
+      "loss": 1.4922,
+      "step": 83600
+    },
+    {
+      "epoch": 2.4324402943828725,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.8918662357194936e-06,
+      "loss": 1.4578,
+      "step": 83620
+    },
+    {
+      "epoch": 2.43302207871542,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.8899269540468787e-06,
+      "loss": 1.4666,
+      "step": 83640
+    },
+    {
+      "epoch": 2.433603863047968,
+      "grad_norm": 16.25,
+      "learning_rate": 1.8879876723742638e-06,
+      "loss": 1.5009,
+      "step": 83660
+    },
+    {
+      "epoch": 2.434185647380516,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.8860483907016489e-06,
+      "loss": 1.4532,
+      "step": 83680
+    },
+    {
+      "epoch": 2.434767431713064,
+      "grad_norm": 18.5,
+      "learning_rate": 1.884109109029034e-06,
+      "loss": 1.4642,
+      "step": 83700
+    },
+    {
+      "epoch": 2.435349216045612,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.8821698273564186e-06,
+      "loss": 1.569,
+      "step": 83720
+    },
+    {
+      "epoch": 2.43593100037816,
+      "grad_norm": 14.25,
+      "learning_rate": 1.8802305456838037e-06,
+      "loss": 1.5864,
+      "step": 83740
+    },
+    {
+      "epoch": 2.4365127847107075,
+      "grad_norm": 11.0625,
+      "learning_rate": 1.8782912640111888e-06,
+      "loss": 1.5732,
+      "step": 83760
+    },
+    {
+      "epoch": 2.4370945690432557,
+      "grad_norm": 14.375,
+      "learning_rate": 1.876351982338574e-06,
+      "loss": 1.4666,
+      "step": 83780
+    },
+    {
+      "epoch": 2.4376763533758035,
+      "grad_norm": 12.375,
+      "learning_rate": 1.874412700665959e-06,
+      "loss": 1.4964,
+      "step": 83800
+    },
+    {
+      "epoch": 2.4382581377083516,
+      "grad_norm": 12.375,
+      "learning_rate": 1.8724734189933441e-06,
+      "loss": 1.5458,
+      "step": 83820
+    },
+    {
+      "epoch": 2.4388399220408994,
+      "grad_norm": 13.125,
+      "learning_rate": 1.8705341373207292e-06,
+      "loss": 1.522,
+      "step": 83840
+    },
+    {
+      "epoch": 2.439421706373447,
+      "grad_norm": 14.875,
+      "learning_rate": 1.8685948556481143e-06,
+      "loss": 1.5432,
+      "step": 83860
+    },
+    {
+      "epoch": 2.4400034907059953,
+      "grad_norm": 14.0,
+      "learning_rate": 1.8666555739754994e-06,
+      "loss": 1.4725,
+      "step": 83880
+    },
+    {
+      "epoch": 2.440585275038543,
+      "grad_norm": 14.0,
+      "learning_rate": 1.8647162923028845e-06,
+      "loss": 1.5032,
+      "step": 83900
+    },
+    {
+      "epoch": 2.4411670593710912,
+      "grad_norm": 15.375,
+      "learning_rate": 1.8627770106302694e-06,
+      "loss": 1.4386,
+      "step": 83920
+    },
+    {
+      "epoch": 2.441748843703639,
+      "grad_norm": 10.5625,
+      "learning_rate": 1.8608377289576545e-06,
+      "loss": 1.5889,
+      "step": 83940
+    },
+    {
+      "epoch": 2.442330628036187,
+      "grad_norm": 13.75,
+      "learning_rate": 1.8588984472850396e-06,
+      "loss": 1.5089,
+      "step": 83960
+    },
+    {
+      "epoch": 2.442912412368735,
+      "grad_norm": 16.0,
+      "learning_rate": 1.8569591656124247e-06,
+      "loss": 1.5052,
+      "step": 83980
+    },
+    {
+      "epoch": 2.4434941967012827,
+      "grad_norm": 14.125,
+      "learning_rate": 1.8550198839398098e-06,
+      "loss": 1.4531,
+      "step": 84000
+    },
+    {
+      "epoch": 2.444075981033831,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.8530806022671949e-06,
+      "loss": 1.5164,
+      "step": 84020
+    },
+    {
+      "epoch": 2.4446577653663786,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.85114132059458e-06,
+      "loss": 1.5387,
+      "step": 84040
+    },
+    {
+      "epoch": 2.4452395496989268,
+      "grad_norm": 16.25,
+      "learning_rate": 1.849202038921965e-06,
+      "loss": 1.4059,
+      "step": 84060
+    },
+    {
+      "epoch": 2.4458213340314745,
+      "grad_norm": 12.4375,
+      "learning_rate": 1.8472627572493502e-06,
+      "loss": 1.4368,
+      "step": 84080
+    },
+    {
+      "epoch": 2.4464031183640227,
+      "grad_norm": 12.125,
+      "learning_rate": 1.8453234755767353e-06,
+      "loss": 1.461,
+      "step": 84100
+    },
+    {
+      "epoch": 2.4469849026965704,
+      "grad_norm": 13.25,
+      "learning_rate": 1.8433841939041204e-06,
+      "loss": 1.5223,
+      "step": 84120
+    },
+    {
+      "epoch": 2.447566687029118,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.8414449122315053e-06,
+      "loss": 1.5211,
+      "step": 84140
+    },
+    {
+      "epoch": 2.4481484713616664,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.8395056305588904e-06,
+      "loss": 1.4521,
+      "step": 84160
+    },
+    {
+      "epoch": 2.448730255694214,
+      "grad_norm": 14.25,
+      "learning_rate": 1.8375663488862755e-06,
+      "loss": 1.5413,
+      "step": 84180
+    },
+    {
+      "epoch": 2.4493120400267623,
+      "grad_norm": 11.0,
+      "learning_rate": 1.8356270672136605e-06,
+      "loss": 1.5836,
+      "step": 84200
+    },
+    {
+      "epoch": 2.44989382435931,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.8336877855410456e-06,
+      "loss": 1.5283,
+      "step": 84220
+    },
+    {
+      "epoch": 2.4504756086918578,
+      "grad_norm": 12.4375,
+      "learning_rate": 1.8317485038684307e-06,
+      "loss": 1.3703,
+      "step": 84240
+    },
+    {
+      "epoch": 2.451057393024406,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.8298092221958158e-06,
+      "loss": 1.5435,
+      "step": 84260
+    },
+    {
+      "epoch": 2.4516391773569537,
+      "grad_norm": 13.875,
+      "learning_rate": 1.827869940523201e-06,
+      "loss": 1.4617,
+      "step": 84280
+    },
+    {
+      "epoch": 2.452220961689502,
+      "grad_norm": 13.875,
+      "learning_rate": 1.825930658850586e-06,
+      "loss": 1.4998,
+      "step": 84300
+    },
+    {
+      "epoch": 2.4528027460220496,
+      "grad_norm": 14.625,
+      "learning_rate": 1.8239913771779711e-06,
+      "loss": 1.5317,
+      "step": 84320
+    },
+    {
+      "epoch": 2.4533845303545974,
+      "grad_norm": 13.5,
+      "learning_rate": 1.822052095505356e-06,
+      "loss": 1.4806,
+      "step": 84340
+    },
+    {
+      "epoch": 2.4539663146871455,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.8201128138327411e-06,
+      "loss": 1.5489,
+      "step": 84360
+    },
+    {
+      "epoch": 2.4545480990196933,
+      "grad_norm": 11.6875,
+      "learning_rate": 1.8181735321601262e-06,
+      "loss": 1.466,
+      "step": 84380
+    },
+    {
+      "epoch": 2.4551298833522415,
+      "grad_norm": 16.75,
+      "learning_rate": 1.8162342504875113e-06,
+      "loss": 1.4696,
+      "step": 84400
+    },
+    {
+      "epoch": 2.455711667684789,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.8142949688148964e-06,
+      "loss": 1.5239,
+      "step": 84420
+    },
+    {
+      "epoch": 2.456293452017337,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.8123556871422815e-06,
+      "loss": 1.5328,
+      "step": 84440
+    },
+    {
+      "epoch": 2.456875236349885,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.8104164054696666e-06,
+      "loss": 1.449,
+      "step": 84460
+    },
+    {
+      "epoch": 2.457457020682433,
+      "grad_norm": 12.25,
+      "learning_rate": 1.8084771237970517e-06,
+      "loss": 1.527,
+      "step": 84480
+    },
+    {
+      "epoch": 2.458038805014981,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.8065378421244368e-06,
+      "loss": 1.5881,
+      "step": 84500
+    },
+    {
+      "epoch": 2.458620589347529,
+      "grad_norm": 16.5,
+      "learning_rate": 1.804598560451822e-06,
+      "loss": 1.4317,
+      "step": 84520
+    },
+    {
+      "epoch": 2.4592023736800765,
+      "grad_norm": 15.25,
+      "learning_rate": 1.8026592787792068e-06,
+      "loss": 1.4889,
+      "step": 84540
+    },
+    {
+      "epoch": 2.4597841580126247,
+      "grad_norm": 14.25,
+      "learning_rate": 1.8007199971065919e-06,
+      "loss": 1.5569,
+      "step": 84560
+    },
+    {
+      "epoch": 2.4603659423451725,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.798780715433977e-06,
+      "loss": 1.4411,
+      "step": 84580
+    },
+    {
+      "epoch": 2.4609477266777207,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.796841433761362e-06,
+      "loss": 1.4549,
+      "step": 84600
+    },
+    {
+      "epoch": 2.4615295110102684,
+      "grad_norm": 12.75,
+      "learning_rate": 1.7949021520887472e-06,
+      "loss": 1.5386,
+      "step": 84620
+    },
+    {
+      "epoch": 2.4621112953428166,
+      "grad_norm": 14.0,
+      "learning_rate": 1.7929628704161323e-06,
+      "loss": 1.6189,
+      "step": 84640
+    },
+    {
+      "epoch": 2.4626930796753643,
+      "grad_norm": 16.75,
+      "learning_rate": 1.7910235887435174e-06,
+      "loss": 1.5666,
+      "step": 84660
+    },
+    {
+      "epoch": 2.4632748640079125,
+      "grad_norm": 12.125,
+      "learning_rate": 1.7890843070709025e-06,
+      "loss": 1.4781,
+      "step": 84680
+    },
+    {
+      "epoch": 2.4638566483404603,
+      "grad_norm": 13.0,
+      "learning_rate": 1.7871450253982876e-06,
+      "loss": 1.4691,
+      "step": 84700
+    },
+    {
+      "epoch": 2.464438432673008,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.7852057437256727e-06,
+      "loss": 1.5353,
+      "step": 84720
+    },
+    {
+      "epoch": 2.465020217005556,
+      "grad_norm": 12.5,
+      "learning_rate": 1.7832664620530575e-06,
+      "loss": 1.3974,
+      "step": 84740
+    },
+    {
+      "epoch": 2.465602001338104,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.7813271803804426e-06,
+      "loss": 1.4819,
+      "step": 84760
+    },
+    {
+      "epoch": 2.466183785670652,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.7793878987078275e-06,
+      "loss": 1.5848,
+      "step": 84780
+    },
+    {
+      "epoch": 2.4667655700032,
+      "grad_norm": 11.625,
+      "learning_rate": 1.7774486170352126e-06,
+      "loss": 1.4865,
+      "step": 84800
+    },
+    {
+      "epoch": 2.4673473543357476,
+      "grad_norm": 14.375,
+      "learning_rate": 1.7755093353625977e-06,
+      "loss": 1.4644,
+      "step": 84820
+    },
+    {
+      "epoch": 2.4679291386682958,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.7735700536899828e-06,
+      "loss": 1.4834,
+      "step": 84840
+    },
+    {
+      "epoch": 2.4685109230008435,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.771630772017368e-06,
+      "loss": 1.5563,
+      "step": 84860
+    },
+    {
+      "epoch": 2.4690927073333917,
+      "grad_norm": 12.375,
+      "learning_rate": 1.7696914903447528e-06,
+      "loss": 1.5087,
+      "step": 84880
+    },
+    {
+      "epoch": 2.4696744916659394,
+      "grad_norm": 14.0,
+      "learning_rate": 1.767752208672138e-06,
+      "loss": 1.4945,
+      "step": 84900
+    },
+    {
+      "epoch": 2.470256275998487,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.765812926999523e-06,
+      "loss": 1.4434,
+      "step": 84920
+    },
+    {
+      "epoch": 2.4708380603310354,
+      "grad_norm": 15.75,
+      "learning_rate": 1.763873645326908e-06,
+      "loss": 1.5025,
+      "step": 84940
+    },
+    {
+      "epoch": 2.471419844663583,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.7619343636542932e-06,
+      "loss": 1.5365,
+      "step": 84960
+    },
+    {
+      "epoch": 2.4720016289961313,
+      "grad_norm": 14.625,
+      "learning_rate": 1.7599950819816783e-06,
+      "loss": 1.4666,
+      "step": 84980
+    },
+    {
+      "epoch": 2.472583413328679,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.7580558003090634e-06,
+      "loss": 1.5651,
+      "step": 85000
+    },
+    {
+      "epoch": 2.4731651976612268,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.7561165186364485e-06,
+      "loss": 1.4765,
+      "step": 85020
+    },
+    {
+      "epoch": 2.473746981993775,
+      "grad_norm": 12.3125,
+      "learning_rate": 1.7541772369638336e-06,
+      "loss": 1.4852,
+      "step": 85040
+    },
+    {
+      "epoch": 2.4743287663263227,
+      "grad_norm": 12.1875,
+      "learning_rate": 1.7522379552912187e-06,
+      "loss": 1.5172,
+      "step": 85060
+    },
+    {
+      "epoch": 2.474910550658871,
+      "grad_norm": 16.125,
+      "learning_rate": 1.7502986736186036e-06,
+      "loss": 1.5443,
+      "step": 85080
+    },
+    {
+      "epoch": 2.4754923349914186,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.7483593919459887e-06,
+      "loss": 1.4828,
+      "step": 85100
+    },
+    {
+      "epoch": 2.4760741193239664,
+      "grad_norm": 15.0,
+      "learning_rate": 1.7464201102733738e-06,
+      "loss": 1.5358,
+      "step": 85120
+    },
+    {
+      "epoch": 2.4766559036565146,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.7444808286007589e-06,
+      "loss": 1.4983,
+      "step": 85140
+    },
+    {
+      "epoch": 2.4772376879890623,
+      "grad_norm": 16.375,
+      "learning_rate": 1.742541546928144e-06,
+      "loss": 1.4259,
+      "step": 85160
+    },
+    {
+      "epoch": 2.4778194723216105,
+      "grad_norm": 10.5,
+      "learning_rate": 1.740602265255529e-06,
+      "loss": 1.473,
+      "step": 85180
+    },
+    {
+      "epoch": 2.4784012566541582,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.7386629835829142e-06,
+      "loss": 1.5687,
+      "step": 85200
+    },
+    {
+      "epoch": 2.4789830409867064,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.7367237019102992e-06,
+      "loss": 1.3671,
+      "step": 85220
+    },
+    {
+      "epoch": 2.479564825319254,
+      "grad_norm": 14.625,
+      "learning_rate": 1.7347844202376843e-06,
+      "loss": 1.4198,
+      "step": 85240
+    },
+    {
+      "epoch": 2.480146609651802,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.7328451385650694e-06,
+      "loss": 1.543,
+      "step": 85260
+    },
+    {
+      "epoch": 2.48072839398435,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.7309058568924545e-06,
+      "loss": 1.5451,
+      "step": 85280
+    },
+    {
+      "epoch": 2.481310178316898,
+      "grad_norm": 16.25,
+      "learning_rate": 1.7289665752198394e-06,
+      "loss": 1.5445,
+      "step": 85300
+    },
+    {
+      "epoch": 2.481891962649446,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.7270272935472245e-06,
+      "loss": 1.5054,
+      "step": 85320
+    },
+    {
+      "epoch": 2.4824737469819937,
+      "grad_norm": 12.125,
+      "learning_rate": 1.7250880118746096e-06,
+      "loss": 1.563,
+      "step": 85340
+    },
+    {
+      "epoch": 2.483055531314542,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.7231487302019947e-06,
+      "loss": 1.4879,
+      "step": 85360
+    },
+    {
+      "epoch": 2.4836373156470897,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.7212094485293798e-06,
+      "loss": 1.4319,
+      "step": 85380
+    },
+    {
+      "epoch": 2.4842190999796374,
+      "grad_norm": 13.125,
+      "learning_rate": 1.719270166856765e-06,
+      "loss": 1.4411,
+      "step": 85400
+    },
+    {
+      "epoch": 2.4848008843121856,
+      "grad_norm": 14.625,
+      "learning_rate": 1.71733088518415e-06,
+      "loss": 1.5097,
+      "step": 85420
+    },
+    {
+      "epoch": 2.4853826686447333,
+      "grad_norm": 16.75,
+      "learning_rate": 1.7153916035115351e-06,
+      "loss": 1.524,
+      "step": 85440
+    },
+    {
+      "epoch": 2.4859644529772815,
+      "grad_norm": 16.25,
+      "learning_rate": 1.7134523218389202e-06,
+      "loss": 1.5037,
+      "step": 85460
+    },
+    {
+      "epoch": 2.4865462373098293,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.7115130401663053e-06,
+      "loss": 1.5351,
+      "step": 85480
+    },
+    {
+      "epoch": 2.487128021642377,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.7095737584936902e-06,
+      "loss": 1.454,
+      "step": 85500
+    },
+    {
+      "epoch": 2.487709805974925,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.7076344768210753e-06,
+      "loss": 1.484,
+      "step": 85520
+    },
+    {
+      "epoch": 2.488291590307473,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.7056951951484604e-06,
+      "loss": 1.5233,
+      "step": 85540
+    },
+    {
+      "epoch": 2.488873374640021,
+      "grad_norm": 11.625,
+      "learning_rate": 1.7037559134758455e-06,
+      "loss": 1.4799,
+      "step": 85560
+    },
+    {
+      "epoch": 2.489455158972569,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.7018166318032306e-06,
+      "loss": 1.4214,
+      "step": 85580
+    },
+    {
+      "epoch": 2.4900369433051166,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.6998773501306157e-06,
+      "loss": 1.3949,
+      "step": 85600
+    },
+    {
+      "epoch": 2.490618727637665,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.6979380684580008e-06,
+      "loss": 1.5463,
+      "step": 85620
+    },
+    {
+      "epoch": 2.4912005119702125,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.6959987867853859e-06,
+      "loss": 1.455,
+      "step": 85640
+    },
+    {
+      "epoch": 2.4917822963027607,
+      "grad_norm": 16.25,
+      "learning_rate": 1.694059505112771e-06,
+      "loss": 1.4947,
+      "step": 85660
+    },
+    {
+      "epoch": 2.4923640806353085,
+      "grad_norm": 10.6875,
+      "learning_rate": 1.692120223440156e-06,
+      "loss": 1.5354,
+      "step": 85680
+    },
+    {
+      "epoch": 2.492945864967856,
+      "grad_norm": 13.5,
+      "learning_rate": 1.690180941767541e-06,
+      "loss": 1.5464,
+      "step": 85700
+    },
+    {
+      "epoch": 2.4935276493004044,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.688241660094926e-06,
+      "loss": 1.5747,
+      "step": 85720
+    },
+    {
+      "epoch": 2.494109433632952,
+      "grad_norm": 13.375,
+      "learning_rate": 1.6863023784223111e-06,
+      "loss": 1.3954,
+      "step": 85740
+    },
+    {
+      "epoch": 2.4946912179655003,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.6843630967496962e-06,
+      "loss": 1.5419,
+      "step": 85760
+    },
+    {
+      "epoch": 2.495273002298048,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.6824238150770813e-06,
+      "loss": 1.4634,
+      "step": 85780
+    },
+    {
+      "epoch": 2.495854786630596,
+      "grad_norm": 13.375,
+      "learning_rate": 1.6804845334044664e-06,
+      "loss": 1.5401,
+      "step": 85800
+    },
+    {
+      "epoch": 2.496436570963144,
+      "grad_norm": 12.1875,
+      "learning_rate": 1.6785452517318515e-06,
+      "loss": 1.4788,
+      "step": 85820
+    },
+    {
+      "epoch": 2.4970183552956917,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.6766059700592362e-06,
+      "loss": 1.4634,
+      "step": 85840
+    },
+    {
+      "epoch": 2.49760013962824,
+      "grad_norm": 15.875,
+      "learning_rate": 1.6746666883866213e-06,
+      "loss": 1.5184,
+      "step": 85860
+    },
+    {
+      "epoch": 2.4981819239607876,
+      "grad_norm": 13.5,
+      "learning_rate": 1.6727274067140064e-06,
+      "loss": 1.5197,
+      "step": 85880
+    },
+    {
+      "epoch": 2.498763708293336,
+      "grad_norm": 11.5625,
+      "learning_rate": 1.6707881250413915e-06,
+      "loss": 1.5003,
+      "step": 85900
+    },
+    {
+      "epoch": 2.4993454926258836,
+      "grad_norm": 12.875,
+      "learning_rate": 1.6688488433687766e-06,
+      "loss": 1.4576,
+      "step": 85920
+    },
+    {
+      "epoch": 2.4999272769584318,
+      "grad_norm": 12.875,
+      "learning_rate": 1.6669095616961617e-06,
+      "loss": 1.5049,
+      "step": 85940
+    },
+    {
+      "epoch": 2.5005090612909795,
+      "grad_norm": 13.75,
+      "learning_rate": 1.6649702800235468e-06,
+      "loss": 1.5107,
+      "step": 85960
+    },
+    {
+      "epoch": 2.5010908456235272,
+      "grad_norm": 12.125,
+      "learning_rate": 1.663030998350932e-06,
+      "loss": 1.4735,
+      "step": 85980
+    },
+    {
+      "epoch": 2.5016726299560754,
+      "grad_norm": 16.5,
+      "learning_rate": 1.661091716678317e-06,
+      "loss": 1.4646,
+      "step": 86000
+    },
+    {
+      "epoch": 2.502254414288623,
+      "grad_norm": 12.75,
+      "learning_rate": 1.659152435005702e-06,
+      "loss": 1.4567,
+      "step": 86020
+    },
+    {
+      "epoch": 2.5028361986211713,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.657213153333087e-06,
+      "loss": 1.4665,
+      "step": 86040
+    },
+    {
+      "epoch": 2.503417982953719,
+      "grad_norm": 15.25,
+      "learning_rate": 1.655273871660472e-06,
+      "loss": 1.5428,
+      "step": 86060
+    },
+    {
+      "epoch": 2.503999767286267,
+      "grad_norm": 11.8125,
+      "learning_rate": 1.6533345899878572e-06,
+      "loss": 1.5118,
+      "step": 86080
+    },
+    {
+      "epoch": 2.504581551618815,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.6513953083152423e-06,
+      "loss": 1.5431,
+      "step": 86100
+    },
+    {
+      "epoch": 2.5051633359513628,
+      "grad_norm": 14.375,
+      "learning_rate": 1.6494560266426274e-06,
+      "loss": 1.5274,
+      "step": 86120
+    },
+    {
+      "epoch": 2.505745120283911,
+      "grad_norm": 13.75,
+      "learning_rate": 1.6475167449700125e-06,
+      "loss": 1.5037,
+      "step": 86140
+    },
+    {
+      "epoch": 2.5063269046164587,
+      "grad_norm": 15.5,
+      "learning_rate": 1.6455774632973976e-06,
+      "loss": 1.5036,
+      "step": 86160
+    },
+    {
+      "epoch": 2.5069086889490064,
+      "grad_norm": 19.5,
+      "learning_rate": 1.6436381816247827e-06,
+      "loss": 1.4302,
+      "step": 86180
+    },
+    {
+      "epoch": 2.5074904732815546,
+      "grad_norm": 11.75,
+      "learning_rate": 1.6416988999521678e-06,
+      "loss": 1.5004,
+      "step": 86200
+    },
+    {
+      "epoch": 2.5080722576141024,
+      "grad_norm": 12.0,
+      "learning_rate": 1.6397596182795529e-06,
+      "loss": 1.4269,
+      "step": 86220
+    },
+    {
+      "epoch": 2.5086540419466505,
+      "grad_norm": 12.4375,
+      "learning_rate": 1.6378203366069377e-06,
+      "loss": 1.4466,
+      "step": 86240
+    },
+    {
+      "epoch": 2.5092358262791983,
+      "grad_norm": 13.875,
+      "learning_rate": 1.6358810549343228e-06,
+      "loss": 1.5068,
+      "step": 86260
+    },
+    {
+      "epoch": 2.509817610611746,
+      "grad_norm": 14.75,
+      "learning_rate": 1.633941773261708e-06,
+      "loss": 1.4814,
+      "step": 86280
+    },
+    {
+      "epoch": 2.510399394944294,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.632002491589093e-06,
+      "loss": 1.4891,
+      "step": 86300
+    },
+    {
+      "epoch": 2.510981179276842,
+      "grad_norm": 11.375,
+      "learning_rate": 1.6300632099164781e-06,
+      "loss": 1.5594,
+      "step": 86320
+    },
+    {
+      "epoch": 2.51156296360939,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.6281239282438632e-06,
+      "loss": 1.5167,
+      "step": 86340
+    },
+    {
+      "epoch": 2.512144747941938,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.6261846465712483e-06,
+      "loss": 1.5381,
+      "step": 86360
+    },
+    {
+      "epoch": 2.5127265322744856,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.6242453648986334e-06,
+      "loss": 1.445,
+      "step": 86380
+    },
+    {
+      "epoch": 2.513308316607034,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.6223060832260185e-06,
+      "loss": 1.5091,
+      "step": 86400
+    },
+    {
+      "epoch": 2.5138901009395815,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.6203668015534036e-06,
+      "loss": 1.5194,
+      "step": 86420
+    },
+    {
+      "epoch": 2.5144718852721297,
+      "grad_norm": 16.375,
+      "learning_rate": 1.6184275198807887e-06,
+      "loss": 1.4995,
+      "step": 86440
+    },
+    {
+      "epoch": 2.5150536696046775,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.6164882382081736e-06,
+      "loss": 1.5071,
+      "step": 86460
+    },
+    {
+      "epoch": 2.515635453937225,
+      "grad_norm": 10.1875,
+      "learning_rate": 1.6145489565355587e-06,
+      "loss": 1.4894,
+      "step": 86480
+    },
+    {
+      "epoch": 2.5162172382697734,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.6126096748629438e-06,
+      "loss": 1.5848,
+      "step": 86500
+    },
+    {
+      "epoch": 2.5167990226023216,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.6106703931903289e-06,
+      "loss": 1.5928,
+      "step": 86520
+    },
+    {
+      "epoch": 2.5173808069348693,
+      "grad_norm": 12.4375,
+      "learning_rate": 1.608731111517714e-06,
+      "loss": 1.4975,
+      "step": 86540
+    },
+    {
+      "epoch": 2.517962591267417,
+      "grad_norm": 14.0,
+      "learning_rate": 1.606791829845099e-06,
+      "loss": 1.5314,
+      "step": 86560
+    },
+    {
+      "epoch": 2.5185443755999652,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.6048525481724842e-06,
+      "loss": 1.4563,
+      "step": 86580
+    },
+    {
+      "epoch": 2.519126159932513,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.6029132664998693e-06,
+      "loss": 1.4916,
+      "step": 86600
+    },
+    {
+      "epoch": 2.519707944265061,
+      "grad_norm": 16.375,
+      "learning_rate": 1.6009739848272544e-06,
+      "loss": 1.5425,
+      "step": 86620
+    },
+    {
+      "epoch": 2.520289728597609,
+      "grad_norm": 13.5,
+      "learning_rate": 1.5990347031546395e-06,
+      "loss": 1.4995,
+      "step": 86640
+    },
+    {
+      "epoch": 2.5208715129301567,
+      "grad_norm": 14.75,
+      "learning_rate": 1.5970954214820244e-06,
+      "loss": 1.4136,
+      "step": 86660
+    },
+    {
+      "epoch": 2.521453297262705,
+      "grad_norm": 11.25,
+      "learning_rate": 1.5951561398094095e-06,
+      "loss": 1.5536,
+      "step": 86680
+    },
+    {
+      "epoch": 2.5220350815952526,
+      "grad_norm": 11.0,
+      "learning_rate": 1.5932168581367946e-06,
+      "loss": 1.5162,
+      "step": 86700
+    },
+    {
+      "epoch": 2.5226168659278008,
+      "grad_norm": 16.125,
+      "learning_rate": 1.5912775764641797e-06,
+      "loss": 1.5166,
+      "step": 86720
+    },
+    {
+      "epoch": 2.5231986502603485,
+      "grad_norm": 10.4375,
+      "learning_rate": 1.5893382947915648e-06,
+      "loss": 1.4429,
+      "step": 86740
+    },
+    {
+      "epoch": 2.5237804345928962,
+      "grad_norm": 17.5,
+      "learning_rate": 1.5873990131189498e-06,
+      "loss": 1.4458,
+      "step": 86760
+    },
+    {
+      "epoch": 2.5243622189254444,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.585459731446335e-06,
+      "loss": 1.4595,
+      "step": 86780
+    },
+    {
+      "epoch": 2.524944003257992,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.58352044977372e-06,
+      "loss": 1.5339,
+      "step": 86800
+    },
+    {
+      "epoch": 2.5255257875905404,
+      "grad_norm": 14.5,
+      "learning_rate": 1.5815811681011051e-06,
+      "loss": 1.5444,
+      "step": 86820
+    },
+    {
+      "epoch": 2.526107571923088,
+      "grad_norm": 16.5,
+      "learning_rate": 1.5796418864284902e-06,
+      "loss": 1.4465,
+      "step": 86840
+    },
+    {
+      "epoch": 2.526689356255636,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.5777026047558751e-06,
+      "loss": 1.5121,
+      "step": 86860
+    },
+    {
+      "epoch": 2.527271140588184,
+      "grad_norm": 11.5625,
+      "learning_rate": 1.5757633230832602e-06,
+      "loss": 1.4736,
+      "step": 86880
+    },
+    {
+      "epoch": 2.5278529249207318,
+      "grad_norm": 16.375,
+      "learning_rate": 1.5738240414106451e-06,
+      "loss": 1.546,
+      "step": 86900
+    },
+    {
+      "epoch": 2.52843470925328,
+      "grad_norm": 13.5,
+      "learning_rate": 1.5718847597380302e-06,
+      "loss": 1.4419,
+      "step": 86920
+    },
+    {
+      "epoch": 2.5290164935858277,
+      "grad_norm": 12.1875,
+      "learning_rate": 1.5699454780654153e-06,
+      "loss": 1.4853,
+      "step": 86940
+    },
+    {
+      "epoch": 2.5295982779183754,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.5680061963928004e-06,
+      "loss": 1.492,
+      "step": 86960
+    },
+    {
+      "epoch": 2.5301800622509236,
+      "grad_norm": 14.0,
+      "learning_rate": 1.5660669147201855e-06,
+      "loss": 1.4927,
+      "step": 86980
+    },
+    {
+      "epoch": 2.5307618465834714,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.5641276330475704e-06,
+      "loss": 1.504,
+      "step": 87000
+    },
+    {
+      "epoch": 2.5313436309160195,
+      "grad_norm": 14.0,
+      "learning_rate": 1.5621883513749555e-06,
+      "loss": 1.4816,
+      "step": 87020
+    },
+    {
+      "epoch": 2.5319254152485673,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.5602490697023406e-06,
+      "loss": 1.5125,
+      "step": 87040
+    },
+    {
+      "epoch": 2.532507199581115,
+      "grad_norm": 11.4375,
+      "learning_rate": 1.5583097880297257e-06,
+      "loss": 1.5393,
+      "step": 87060
+    },
+    {
+      "epoch": 2.533088983913663,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.5563705063571108e-06,
+      "loss": 1.4948,
+      "step": 87080
+    },
+    {
+      "epoch": 2.5336707682462114,
+      "grad_norm": 17.375,
+      "learning_rate": 1.5544312246844959e-06,
+      "loss": 1.5009,
+      "step": 87100
+    },
+    {
+      "epoch": 2.534252552578759,
+      "grad_norm": 13.25,
+      "learning_rate": 1.552491943011881e-06,
+      "loss": 1.491,
+      "step": 87120
+    },
+    {
+      "epoch": 2.534834336911307,
+      "grad_norm": 12.25,
+      "learning_rate": 1.550552661339266e-06,
+      "loss": 1.4954,
+      "step": 87140
+    },
+    {
+      "epoch": 2.5354161212438546,
+      "grad_norm": 14.625,
+      "learning_rate": 1.5486133796666512e-06,
+      "loss": 1.4881,
+      "step": 87160
+    },
+    {
+      "epoch": 2.535997905576403,
+      "grad_norm": 11.375,
+      "learning_rate": 1.5466740979940363e-06,
+      "loss": 1.4248,
+      "step": 87180
+    },
+    {
+      "epoch": 2.536579689908951,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.5447348163214211e-06,
+      "loss": 1.4026,
+      "step": 87200
+    },
+    {
+      "epoch": 2.5371614742414987,
+      "grad_norm": 15.375,
+      "learning_rate": 1.5427955346488062e-06,
+      "loss": 1.5348,
+      "step": 87220
+    },
+    {
+      "epoch": 2.5377432585740465,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.5408562529761913e-06,
+      "loss": 1.4726,
+      "step": 87240
+    },
+    {
+      "epoch": 2.5383250429065947,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.5389169713035764e-06,
+      "loss": 1.4464,
+      "step": 87260
+    },
+    {
+      "epoch": 2.5389068272391424,
+      "grad_norm": 11.625,
+      "learning_rate": 1.5369776896309615e-06,
+      "loss": 1.5215,
+      "step": 87280
+    },
+    {
+      "epoch": 2.5394886115716906,
+      "grad_norm": 12.25,
+      "learning_rate": 1.5350384079583466e-06,
+      "loss": 1.5135,
+      "step": 87300
+    },
+    {
+      "epoch": 2.5400703959042383,
+      "grad_norm": 14.25,
+      "learning_rate": 1.5330991262857317e-06,
+      "loss": 1.5343,
+      "step": 87320
+    },
+    {
+      "epoch": 2.540652180236786,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.5311598446131168e-06,
+      "loss": 1.4614,
+      "step": 87340
+    },
+    {
+      "epoch": 2.5412339645693343,
+      "grad_norm": 12.5,
+      "learning_rate": 1.529220562940502e-06,
+      "loss": 1.3836,
+      "step": 87360
+    },
+    {
+      "epoch": 2.541815748901882,
+      "grad_norm": 15.625,
+      "learning_rate": 1.527281281267887e-06,
+      "loss": 1.4345,
+      "step": 87380
+    },
+    {
+      "epoch": 2.54239753323443,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.5253419995952721e-06,
+      "loss": 1.5226,
+      "step": 87400
+    },
+    {
+      "epoch": 2.542979317566978,
+      "grad_norm": 11.8125,
+      "learning_rate": 1.523402717922657e-06,
+      "loss": 1.4521,
+      "step": 87420
+    },
+    {
+      "epoch": 2.5435611018995257,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.521463436250042e-06,
+      "loss": 1.5146,
+      "step": 87440
+    },
+    {
+      "epoch": 2.544142886232074,
+      "grad_norm": 11.6875,
+      "learning_rate": 1.5195241545774272e-06,
+      "loss": 1.6255,
+      "step": 87460
+    },
+    {
+      "epoch": 2.5447246705646216,
+      "grad_norm": 14.125,
+      "learning_rate": 1.5175848729048123e-06,
+      "loss": 1.43,
+      "step": 87480
+    },
+    {
+      "epoch": 2.54530645489717,
+      "grad_norm": 14.375,
+      "learning_rate": 1.5156455912321974e-06,
+      "loss": 1.4608,
+      "step": 87500
+    },
+    {
+      "epoch": 2.5458882392297175,
+      "grad_norm": 12.0,
+      "learning_rate": 1.5137063095595825e-06,
+      "loss": 1.5352,
+      "step": 87520
+    },
+    {
+      "epoch": 2.5464700235622653,
+      "grad_norm": 13.625,
+      "learning_rate": 1.5117670278869676e-06,
+      "loss": 1.575,
+      "step": 87540
+    },
+    {
+      "epoch": 2.5470518078948134,
+      "grad_norm": 12.8125,
+      "learning_rate": 1.5098277462143527e-06,
+      "loss": 1.4455,
+      "step": 87560
+    },
+    {
+      "epoch": 2.547633592227361,
+      "grad_norm": 12.5,
+      "learning_rate": 1.5078884645417378e-06,
+      "loss": 1.4836,
+      "step": 87580
+    },
+    {
+      "epoch": 2.5482153765599094,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.5059491828691229e-06,
+      "loss": 1.3722,
+      "step": 87600
+    },
+    {
+      "epoch": 2.548797160892457,
+      "grad_norm": 13.125,
+      "learning_rate": 1.5040099011965078e-06,
+      "loss": 1.4748,
+      "step": 87620
+    },
+    {
+      "epoch": 2.549378945225005,
+      "grad_norm": 11.875,
+      "learning_rate": 1.5020706195238929e-06,
+      "loss": 1.5758,
+      "step": 87640
+    },
+    {
+      "epoch": 2.549960729557553,
+      "grad_norm": 16.875,
+      "learning_rate": 1.500131337851278e-06,
+      "loss": 1.5294,
+      "step": 87660
+    },
+    {
+      "epoch": 2.550542513890101,
+      "grad_norm": 13.625,
+      "learning_rate": 1.498192056178663e-06,
+      "loss": 1.5163,
+      "step": 87680
+    },
+    {
+      "epoch": 2.551124298222649,
+      "grad_norm": 15.125,
+      "learning_rate": 1.4962527745060482e-06,
+      "loss": 1.5919,
+      "step": 87700
+    },
+    {
+      "epoch": 2.5517060825551967,
+      "grad_norm": 12.25,
+      "learning_rate": 1.4943134928334333e-06,
+      "loss": 1.5041,
+      "step": 87720
+    },
+    {
+      "epoch": 2.5522878668877444,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.4923742111608184e-06,
+      "loss": 1.4404,
+      "step": 87740
+    },
+    {
+      "epoch": 2.5528696512202926,
+      "grad_norm": 11.75,
+      "learning_rate": 1.4904349294882035e-06,
+      "loss": 1.5303,
+      "step": 87760
+    },
+    {
+      "epoch": 2.553451435552841,
+      "grad_norm": 12.8125,
+      "learning_rate": 1.4884956478155885e-06,
+      "loss": 1.5042,
+      "step": 87780
+    },
+    {
+      "epoch": 2.5540332198853886,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.4865563661429736e-06,
+      "loss": 1.4936,
+      "step": 87800
+    },
+    {
+      "epoch": 2.5546150042179363,
+      "grad_norm": 16.25,
+      "learning_rate": 1.4846170844703585e-06,
+      "loss": 1.5642,
+      "step": 87820
+    },
+    {
+      "epoch": 2.5551967885504845,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.4826778027977436e-06,
+      "loss": 1.4549,
+      "step": 87840
+    },
+    {
+      "epoch": 2.5557785728830322,
+      "grad_norm": 13.75,
+      "learning_rate": 1.4807385211251287e-06,
+      "loss": 1.514,
+      "step": 87860
+    },
+    {
+      "epoch": 2.5563603572155804,
+      "grad_norm": 13.5,
+      "learning_rate": 1.4787992394525138e-06,
+      "loss": 1.4641,
+      "step": 87880
+    },
+    {
+      "epoch": 2.556942141548128,
+      "grad_norm": 12.1875,
+      "learning_rate": 1.476859957779899e-06,
+      "loss": 1.5321,
+      "step": 87900
+    },
+    {
+      "epoch": 2.557523925880676,
+      "grad_norm": 17.25,
+      "learning_rate": 1.474920676107284e-06,
+      "loss": 1.462,
+      "step": 87920
+    },
+    {
+      "epoch": 2.558105710213224,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.4729813944346691e-06,
+      "loss": 1.5143,
+      "step": 87940
+    },
+    {
+      "epoch": 2.558687494545772,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.4710421127620538e-06,
+      "loss": 1.4539,
+      "step": 87960
+    },
+    {
+      "epoch": 2.55926927887832,
+      "grad_norm": 12.125,
+      "learning_rate": 1.4691028310894389e-06,
+      "loss": 1.451,
+      "step": 87980
+    },
+    {
+      "epoch": 2.5598510632108677,
+      "grad_norm": 13.5,
+      "learning_rate": 1.467163549416824e-06,
+      "loss": 1.5119,
+      "step": 88000
+    },
+    {
+      "epoch": 2.5604328475434155,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.465224267744209e-06,
+      "loss": 1.4436,
+      "step": 88020
+    },
+    {
+      "epoch": 2.5610146318759637,
+      "grad_norm": 11.5,
+      "learning_rate": 1.4632849860715942e-06,
+      "loss": 1.436,
+      "step": 88040
+    },
+    {
+      "epoch": 2.5615964162085114,
+      "grad_norm": 12.1875,
+      "learning_rate": 1.4613457043989793e-06,
+      "loss": 1.4874,
+      "step": 88060
+    },
+    {
+      "epoch": 2.5621782005410596,
+      "grad_norm": 16.25,
+      "learning_rate": 1.4594064227263644e-06,
+      "loss": 1.4848,
+      "step": 88080
+    },
+    {
+      "epoch": 2.5627599848736073,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.4574671410537495e-06,
+      "loss": 1.4135,
+      "step": 88100
+    },
+    {
+      "epoch": 2.563341769206155,
+      "grad_norm": 13.625,
+      "learning_rate": 1.4555278593811346e-06,
+      "loss": 1.4105,
+      "step": 88120
+    },
+    {
+      "epoch": 2.5639235535387033,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.4535885777085197e-06,
+      "loss": 1.4694,
+      "step": 88140
+    },
+    {
+      "epoch": 2.564505337871251,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.4516492960359046e-06,
+      "loss": 1.4228,
+      "step": 88160
+    },
+    {
+      "epoch": 2.565087122203799,
+      "grad_norm": 12.1875,
+      "learning_rate": 1.4497100143632897e-06,
+      "loss": 1.4681,
+      "step": 88180
+    },
+    {
+      "epoch": 2.565668906536347,
+      "grad_norm": 13.375,
+      "learning_rate": 1.4477707326906747e-06,
+      "loss": 1.552,
+      "step": 88200
+    },
+    {
+      "epoch": 2.5662506908688947,
+      "grad_norm": 10.75,
+      "learning_rate": 1.4458314510180598e-06,
+      "loss": 1.5275,
+      "step": 88220
+    },
+    {
+      "epoch": 2.566832475201443,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.443892169345445e-06,
+      "loss": 1.4954,
+      "step": 88240
+    },
+    {
+      "epoch": 2.5674142595339906,
+      "grad_norm": 12.625,
+      "learning_rate": 1.44195288767283e-06,
+      "loss": 1.5454,
+      "step": 88260
+    },
+    {
+      "epoch": 2.567996043866539,
+      "grad_norm": 10.6875,
+      "learning_rate": 1.4400136060002151e-06,
+      "loss": 1.4329,
+      "step": 88280
+    },
+    {
+      "epoch": 2.5685778281990865,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.4380743243276002e-06,
+      "loss": 1.4786,
+      "step": 88300
+    },
+    {
+      "epoch": 2.5691596125316343,
+      "grad_norm": 15.0,
+      "learning_rate": 1.4361350426549853e-06,
+      "loss": 1.4678,
+      "step": 88320
+    },
+    {
+      "epoch": 2.5697413968641825,
+      "grad_norm": 11.25,
+      "learning_rate": 1.4341957609823704e-06,
+      "loss": 1.5508,
+      "step": 88340
+    },
+    {
+      "epoch": 2.5703231811967306,
+      "grad_norm": 11.0625,
+      "learning_rate": 1.4322564793097553e-06,
+      "loss": 1.4506,
+      "step": 88360
+    },
+    {
+      "epoch": 2.5709049655292784,
+      "grad_norm": 12.0,
+      "learning_rate": 1.4303171976371404e-06,
+      "loss": 1.4891,
+      "step": 88380
+    },
+    {
+      "epoch": 2.571486749861826,
+      "grad_norm": 15.125,
+      "learning_rate": 1.4283779159645255e-06,
+      "loss": 1.5108,
+      "step": 88400
+    },
+    {
+      "epoch": 2.5720685341943743,
+      "grad_norm": 12.875,
+      "learning_rate": 1.4264386342919106e-06,
+      "loss": 1.4628,
+      "step": 88420
+    },
+    {
+      "epoch": 2.572650318526922,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.4244993526192957e-06,
+      "loss": 1.4989,
+      "step": 88440
+    },
+    {
+      "epoch": 2.5732321028594702,
+      "grad_norm": 14.0,
+      "learning_rate": 1.4225600709466808e-06,
+      "loss": 1.5064,
+      "step": 88460
+    },
+    {
+      "epoch": 2.573813887192018,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.420620789274066e-06,
+      "loss": 1.4556,
+      "step": 88480
+    },
+    {
+      "epoch": 2.5743956715245657,
+      "grad_norm": 16.25,
+      "learning_rate": 1.418681507601451e-06,
+      "loss": 1.3963,
+      "step": 88500
+    },
+    {
+      "epoch": 2.574977455857114,
+      "grad_norm": 13.75,
+      "learning_rate": 1.416742225928836e-06,
+      "loss": 1.5654,
+      "step": 88520
+    },
+    {
+      "epoch": 2.5755592401896616,
+      "grad_norm": 16.5,
+      "learning_rate": 1.4148029442562212e-06,
+      "loss": 1.5069,
+      "step": 88540
+    },
+    {
+      "epoch": 2.57614102452221,
+      "grad_norm": 13.875,
+      "learning_rate": 1.4128636625836063e-06,
+      "loss": 1.5128,
+      "step": 88560
+    },
+    {
+      "epoch": 2.5767228088547576,
+      "grad_norm": 13.5,
+      "learning_rate": 1.4109243809109912e-06,
+      "loss": 1.4906,
+      "step": 88580
+    },
+    {
+      "epoch": 2.5773045931873053,
+      "grad_norm": 15.375,
+      "learning_rate": 1.4089850992383763e-06,
+      "loss": 1.5292,
+      "step": 88600
+    },
+    {
+      "epoch": 2.5778863775198535,
+      "grad_norm": 15.75,
+      "learning_rate": 1.4070458175657614e-06,
+      "loss": 1.4152,
+      "step": 88620
+    },
+    {
+      "epoch": 2.5784681618524012,
+      "grad_norm": 13.375,
+      "learning_rate": 1.4051065358931465e-06,
+      "loss": 1.4743,
+      "step": 88640
+    },
+    {
+      "epoch": 2.5790499461849494,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.4031672542205316e-06,
+      "loss": 1.5106,
+      "step": 88660
+    },
+    {
+      "epoch": 2.579631730517497,
+      "grad_norm": 14.125,
+      "learning_rate": 1.4012279725479167e-06,
+      "loss": 1.4018,
+      "step": 88680
+    },
+    {
+      "epoch": 2.580213514850045,
+      "grad_norm": 14.0,
+      "learning_rate": 1.3992886908753018e-06,
+      "loss": 1.5591,
+      "step": 88700
+    },
+    {
+      "epoch": 2.580795299182593,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.3973494092026869e-06,
+      "loss": 1.4992,
+      "step": 88720
+    },
+    {
+      "epoch": 2.581377083515141,
+      "grad_norm": 12.125,
+      "learning_rate": 1.395410127530072e-06,
+      "loss": 1.478,
+      "step": 88740
+    },
+    {
+      "epoch": 2.581958867847689,
+      "grad_norm": 12.25,
+      "learning_rate": 1.393470845857457e-06,
+      "loss": 1.4129,
+      "step": 88760
+    },
+    {
+      "epoch": 2.5825406521802368,
+      "grad_norm": 17.75,
+      "learning_rate": 1.391531564184842e-06,
+      "loss": 1.4793,
+      "step": 88780
+    },
+    {
+      "epoch": 2.5831224365127845,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.389592282512227e-06,
+      "loss": 1.5708,
+      "step": 88800
+    },
+    {
+      "epoch": 2.5837042208453327,
+      "grad_norm": 12.75,
+      "learning_rate": 1.3876530008396121e-06,
+      "loss": 1.5207,
+      "step": 88820
+    },
+    {
+      "epoch": 2.5842860051778804,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.3857137191669972e-06,
+      "loss": 1.5063,
+      "step": 88840
+    },
+    {
+      "epoch": 2.5848677895104286,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.3837744374943823e-06,
+      "loss": 1.5515,
+      "step": 88860
+    },
+    {
+      "epoch": 2.5854495738429764,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.3818351558217674e-06,
+      "loss": 1.5853,
+      "step": 88880
+    },
+    {
+      "epoch": 2.586031358175524,
+      "grad_norm": 13.625,
+      "learning_rate": 1.3798958741491525e-06,
+      "loss": 1.569,
+      "step": 88900
+    },
+    {
+      "epoch": 2.5866131425080723,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.3779565924765376e-06,
+      "loss": 1.5617,
+      "step": 88920
+    },
+    {
+      "epoch": 2.58719492684062,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.3760173108039227e-06,
+      "loss": 1.4424,
+      "step": 88940
+    },
+    {
+      "epoch": 2.587776711173168,
+      "grad_norm": 12.75,
+      "learning_rate": 1.3740780291313078e-06,
+      "loss": 1.5559,
+      "step": 88960
+    },
+    {
+      "epoch": 2.588358495505716,
+      "grad_norm": 11.375,
+      "learning_rate": 1.3721387474586927e-06,
+      "loss": 1.5516,
+      "step": 88980
+    },
+    {
+      "epoch": 2.5889402798382637,
+      "grad_norm": 12.625,
+      "learning_rate": 1.3701994657860778e-06,
+      "loss": 1.5243,
+      "step": 89000
+    },
+    {
+      "epoch": 2.589522064170812,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.3682601841134627e-06,
+      "loss": 1.4705,
+      "step": 89020
+    },
+    {
+      "epoch": 2.59010384850336,
+      "grad_norm": 14.375,
+      "learning_rate": 1.3663209024408478e-06,
+      "loss": 1.4796,
+      "step": 89040
+    },
+    {
+      "epoch": 2.590685632835908,
+      "grad_norm": 12.3125,
+      "learning_rate": 1.3643816207682329e-06,
+      "loss": 1.5771,
+      "step": 89060
+    },
+    {
+      "epoch": 2.5912674171684555,
+      "grad_norm": 16.25,
+      "learning_rate": 1.362442339095618e-06,
+      "loss": 1.5066,
+      "step": 89080
+    },
+    {
+      "epoch": 2.5918492015010037,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.360503057423003e-06,
+      "loss": 1.5614,
+      "step": 89100
+    },
+    {
+      "epoch": 2.5924309858335515,
+      "grad_norm": 15.5,
+      "learning_rate": 1.358563775750388e-06,
+      "loss": 1.5666,
+      "step": 89120
+    },
+    {
+      "epoch": 2.5930127701660997,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.356624494077773e-06,
+      "loss": 1.5544,
+      "step": 89140
+    },
+    {
+      "epoch": 2.5935945544986474,
+      "grad_norm": 14.5,
+      "learning_rate": 1.3546852124051582e-06,
+      "loss": 1.4609,
+      "step": 89160
+    },
+    {
+      "epoch": 2.594176338831195,
+      "grad_norm": 14.375,
+      "learning_rate": 1.3527459307325433e-06,
+      "loss": 1.4836,
+      "step": 89180
+    },
+    {
+      "epoch": 2.5947581231637433,
+      "grad_norm": 8.625,
+      "learning_rate": 1.3508066490599284e-06,
+      "loss": 1.4095,
+      "step": 89200
+    },
+    {
+      "epoch": 2.595339907496291,
+      "grad_norm": 17.875,
+      "learning_rate": 1.3488673673873135e-06,
+      "loss": 1.4704,
+      "step": 89220
+    },
+    {
+      "epoch": 2.5959216918288393,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.3469280857146985e-06,
+      "loss": 1.5322,
+      "step": 89240
+    },
+    {
+      "epoch": 2.596503476161387,
+      "grad_norm": 13.125,
+      "learning_rate": 1.3449888040420836e-06,
+      "loss": 1.5451,
+      "step": 89260
+    },
+    {
+      "epoch": 2.5970852604939347,
+      "grad_norm": 13.125,
+      "learning_rate": 1.3430495223694687e-06,
+      "loss": 1.5784,
+      "step": 89280
+    },
+    {
+      "epoch": 2.597667044826483,
+      "grad_norm": 13.125,
+      "learning_rate": 1.3411102406968538e-06,
+      "loss": 1.5883,
+      "step": 89300
+    },
+    {
+      "epoch": 2.5982488291590307,
+      "grad_norm": 15.375,
+      "learning_rate": 1.3391709590242387e-06,
+      "loss": 1.5509,
+      "step": 89320
+    },
+    {
+      "epoch": 2.598830613491579,
+      "grad_norm": 10.75,
+      "learning_rate": 1.3372316773516238e-06,
+      "loss": 1.5002,
+      "step": 89340
+    },
+    {
+      "epoch": 2.5994123978241266,
+      "grad_norm": 16.25,
+      "learning_rate": 1.335292395679009e-06,
+      "loss": 1.5081,
+      "step": 89360
+    },
+    {
+      "epoch": 2.5999941821566743,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.333353114006394e-06,
+      "loss": 1.4374,
+      "step": 89380
+    },
+    {
+      "epoch": 2.6005759664892225,
+      "grad_norm": 13.5,
+      "learning_rate": 1.3314138323337791e-06,
+      "loss": 1.5208,
+      "step": 89400
+    },
+    {
+      "epoch": 2.6011577508217703,
+      "grad_norm": 15.25,
+      "learning_rate": 1.3294745506611642e-06,
+      "loss": 1.5339,
+      "step": 89420
+    },
+    {
+      "epoch": 2.6017395351543184,
+      "grad_norm": 12.3125,
+      "learning_rate": 1.3275352689885493e-06,
+      "loss": 1.4808,
+      "step": 89440
+    },
+    {
+      "epoch": 2.602321319486866,
+      "grad_norm": 11.8125,
+      "learning_rate": 1.3255959873159344e-06,
+      "loss": 1.4434,
+      "step": 89460
+    },
+    {
+      "epoch": 2.602903103819414,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.3236567056433195e-06,
+      "loss": 1.4781,
+      "step": 89480
+    },
+    {
+      "epoch": 2.603484888151962,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.3217174239707046e-06,
+      "loss": 1.4987,
+      "step": 89500
+    },
+    {
+      "epoch": 2.60406667248451,
+      "grad_norm": 16.625,
+      "learning_rate": 1.3197781422980895e-06,
+      "loss": 1.4983,
+      "step": 89520
+    },
+    {
+      "epoch": 2.604648456817058,
+      "grad_norm": 10.9375,
+      "learning_rate": 1.3178388606254746e-06,
+      "loss": 1.5454,
+      "step": 89540
+    },
+    {
+      "epoch": 2.6052302411496058,
+      "grad_norm": 12.75,
+      "learning_rate": 1.3158995789528597e-06,
+      "loss": 1.569,
+      "step": 89560
+    },
+    {
+      "epoch": 2.6058120254821535,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.3139602972802448e-06,
+      "loss": 1.4457,
+      "step": 89580
+    },
+    {
+      "epoch": 2.6063938098147017,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.3120210156076299e-06,
+      "loss": 1.5113,
+      "step": 89600
+    },
+    {
+      "epoch": 2.60697559414725,
+      "grad_norm": 16.125,
+      "learning_rate": 1.310081733935015e-06,
+      "loss": 1.4794,
+      "step": 89620
+    },
+    {
+      "epoch": 2.6075573784797976,
+      "grad_norm": 15.125,
+      "learning_rate": 1.3081424522624e-06,
+      "loss": 1.4272,
+      "step": 89640
+    },
+    {
+      "epoch": 2.6081391628123454,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.3062031705897852e-06,
+      "loss": 1.5364,
+      "step": 89660
+    },
+    {
+      "epoch": 2.6087209471448936,
+      "grad_norm": 11.6875,
+      "learning_rate": 1.3042638889171703e-06,
+      "loss": 1.4851,
+      "step": 89680
+    },
+    {
+      "epoch": 2.6093027314774413,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.3023246072445554e-06,
+      "loss": 1.4141,
+      "step": 89700
+    },
+    {
+      "epoch": 2.6098845158099895,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.3003853255719405e-06,
+      "loss": 1.5233,
+      "step": 89720
+    },
+    {
+      "epoch": 2.610466300142537,
+      "grad_norm": 12.4375,
+      "learning_rate": 1.2984460438993253e-06,
+      "loss": 1.4895,
+      "step": 89740
+    },
+    {
+      "epoch": 2.611048084475085,
+      "grad_norm": 16.0,
+      "learning_rate": 1.2965067622267104e-06,
+      "loss": 1.4961,
+      "step": 89760
+    },
+    {
+      "epoch": 2.611629868807633,
+      "grad_norm": 10.625,
+      "learning_rate": 1.2945674805540955e-06,
+      "loss": 1.5121,
+      "step": 89780
+    },
+    {
+      "epoch": 2.612211653140181,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.2926281988814806e-06,
+      "loss": 1.5176,
+      "step": 89800
+    },
+    {
+      "epoch": 2.612793437472729,
+      "grad_norm": 14.625,
+      "learning_rate": 1.2906889172088657e-06,
+      "loss": 1.488,
+      "step": 89820
+    },
+    {
+      "epoch": 2.613375221805277,
+      "grad_norm": 15.0,
+      "learning_rate": 1.2887496355362508e-06,
+      "loss": 1.5184,
+      "step": 89840
+    },
+    {
+      "epoch": 2.6139570061378246,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.286810353863636e-06,
+      "loss": 1.525,
+      "step": 89860
+    },
+    {
+      "epoch": 2.6145387904703727,
+      "grad_norm": 12.375,
+      "learning_rate": 1.284871072191021e-06,
+      "loss": 1.4853,
+      "step": 89880
+    },
+    {
+      "epoch": 2.6151205748029205,
+      "grad_norm": 12.1875,
+      "learning_rate": 1.2829317905184061e-06,
+      "loss": 1.5604,
+      "step": 89900
+    },
+    {
+      "epoch": 2.6157023591354687,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.2809925088457912e-06,
+      "loss": 1.5214,
+      "step": 89920
+    },
+    {
+      "epoch": 2.6162841434680164,
+      "grad_norm": 15.25,
+      "learning_rate": 1.2790532271731761e-06,
+      "loss": 1.4771,
+      "step": 89940
+    },
+    {
+      "epoch": 2.616865927800564,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.2771139455005612e-06,
+      "loss": 1.414,
+      "step": 89960
+    },
+    {
+      "epoch": 2.6174477121331123,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.2751746638279463e-06,
+      "loss": 1.43,
+      "step": 89980
+    },
+    {
+      "epoch": 2.61802949646566,
+      "grad_norm": 13.75,
+      "learning_rate": 1.2732353821553314e-06,
+      "loss": 1.3977,
+      "step": 90000
+    },
+    {
+      "epoch": 2.6186112807982083,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.2712961004827165e-06,
+      "loss": 1.491,
+      "step": 90020
+    },
+    {
+      "epoch": 2.619193065130756,
+      "grad_norm": 14.25,
+      "learning_rate": 1.2693568188101016e-06,
+      "loss": 1.541,
+      "step": 90040
+    },
+    {
+      "epoch": 2.6197748494633037,
+      "grad_norm": 13.875,
+      "learning_rate": 1.2674175371374867e-06,
+      "loss": 1.4521,
+      "step": 90060
+    },
+    {
+      "epoch": 2.620356633795852,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.2654782554648714e-06,
+      "loss": 1.5054,
+      "step": 90080
+    },
+    {
+      "epoch": 2.6209384181283997,
+      "grad_norm": 11.5,
+      "learning_rate": 1.2635389737922565e-06,
+      "loss": 1.5068,
+      "step": 90100
+    },
+    {
+      "epoch": 2.621520202460948,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.2615996921196416e-06,
+      "loss": 1.4902,
+      "step": 90120
+    },
+    {
+      "epoch": 2.6221019867934956,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.2596604104470267e-06,
+      "loss": 1.3949,
+      "step": 90140
+    },
+    {
+      "epoch": 2.6226837711260433,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.2577211287744118e-06,
+      "loss": 1.4714,
+      "step": 90160
+    },
+    {
+      "epoch": 2.6232655554585915,
+      "grad_norm": 14.5,
+      "learning_rate": 1.2557818471017969e-06,
+      "loss": 1.4378,
+      "step": 90180
+    },
+    {
+      "epoch": 2.6238473397911393,
+      "grad_norm": 14.0,
+      "learning_rate": 1.253842565429182e-06,
+      "loss": 1.5399,
+      "step": 90200
+    },
+    {
+      "epoch": 2.6244291241236875,
+      "grad_norm": 12.375,
+      "learning_rate": 1.251903283756567e-06,
+      "loss": 1.5688,
+      "step": 90220
+    },
+    {
+      "epoch": 2.625010908456235,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.2499640020839522e-06,
+      "loss": 1.4745,
+      "step": 90240
+    },
+    {
+      "epoch": 2.625592692788783,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.2480247204113372e-06,
+      "loss": 1.5646,
+      "step": 90260
+    },
+    {
+      "epoch": 2.626174477121331,
+      "grad_norm": 12.625,
+      "learning_rate": 1.2460854387387223e-06,
+      "loss": 1.435,
+      "step": 90280
+    },
+    {
+      "epoch": 2.6267562614538793,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.2441461570661074e-06,
+      "loss": 1.558,
+      "step": 90300
+    },
+    {
+      "epoch": 2.627338045786427,
+      "grad_norm": 12.875,
+      "learning_rate": 1.2422068753934925e-06,
+      "loss": 1.4167,
+      "step": 90320
+    },
+    {
+      "epoch": 2.627919830118975,
+      "grad_norm": 15.25,
+      "learning_rate": 1.2402675937208776e-06,
+      "loss": 1.4858,
+      "step": 90340
+    },
+    {
+      "epoch": 2.628501614451523,
+      "grad_norm": 12.8125,
+      "learning_rate": 1.2383283120482625e-06,
+      "loss": 1.4881,
+      "step": 90360
+    },
+    {
+      "epoch": 2.6290833987840707,
+      "grad_norm": 15.75,
+      "learning_rate": 1.2363890303756476e-06,
+      "loss": 1.5406,
+      "step": 90380
+    },
+    {
+      "epoch": 2.629665183116619,
+      "grad_norm": 13.0,
+      "learning_rate": 1.2344497487030327e-06,
+      "loss": 1.5149,
+      "step": 90400
+    },
+    {
+      "epoch": 2.6302469674491666,
+      "grad_norm": 13.875,
+      "learning_rate": 1.2325104670304178e-06,
+      "loss": 1.5103,
+      "step": 90420
+    },
+    {
+      "epoch": 2.6308287517817144,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.230571185357803e-06,
+      "loss": 1.459,
+      "step": 90440
+    },
+    {
+      "epoch": 2.6314105361142626,
+      "grad_norm": 16.0,
+      "learning_rate": 1.228631903685188e-06,
+      "loss": 1.4727,
+      "step": 90460
+    },
+    {
+      "epoch": 2.6319923204468103,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.226692622012573e-06,
+      "loss": 1.4068,
+      "step": 90480
+    },
+    {
+      "epoch": 2.6325741047793585,
+      "grad_norm": 15.75,
+      "learning_rate": 1.224753340339958e-06,
+      "loss": 1.532,
+      "step": 90500
+    },
+    {
+      "epoch": 2.6331558891119062,
+      "grad_norm": 14.0,
+      "learning_rate": 1.222814058667343e-06,
+      "loss": 1.5277,
+      "step": 90520
+    },
+    {
+      "epoch": 2.633737673444454,
+      "grad_norm": 16.25,
+      "learning_rate": 1.2208747769947282e-06,
+      "loss": 1.5638,
+      "step": 90540
+    },
+    {
+      "epoch": 2.634319457777002,
+      "grad_norm": 14.625,
+      "learning_rate": 1.2189354953221133e-06,
+      "loss": 1.4908,
+      "step": 90560
+    },
+    {
+      "epoch": 2.63490124210955,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.2169962136494984e-06,
+      "loss": 1.541,
+      "step": 90580
+    },
+    {
+      "epoch": 2.635483026442098,
+      "grad_norm": 14.125,
+      "learning_rate": 1.2150569319768835e-06,
+      "loss": 1.5557,
+      "step": 90600
+    },
+    {
+      "epoch": 2.636064810774646,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.2131176503042686e-06,
+      "loss": 1.4827,
+      "step": 90620
+    },
+    {
+      "epoch": 2.6366465951071936,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.2111783686316537e-06,
+      "loss": 1.4529,
+      "step": 90640
+    },
+    {
+      "epoch": 2.6372283794397418,
+      "grad_norm": 12.875,
+      "learning_rate": 1.2092390869590388e-06,
+      "loss": 1.5072,
+      "step": 90660
+    },
+    {
+      "epoch": 2.6378101637722895,
+      "grad_norm": 12.4375,
+      "learning_rate": 1.2072998052864237e-06,
+      "loss": 1.4667,
+      "step": 90680
+    },
+    {
+      "epoch": 2.6383919481048377,
+      "grad_norm": 15.625,
+      "learning_rate": 1.2053605236138088e-06,
+      "loss": 1.4551,
+      "step": 90700
+    },
+    {
+      "epoch": 2.6389737324373854,
+      "grad_norm": 12.0,
+      "learning_rate": 1.2034212419411939e-06,
+      "loss": 1.4715,
+      "step": 90720
+    },
+    {
+      "epoch": 2.639555516769933,
+      "grad_norm": 12.25,
+      "learning_rate": 1.201481960268579e-06,
+      "loss": 1.4934,
+      "step": 90740
+    },
+    {
+      "epoch": 2.6401373011024813,
+      "grad_norm": 16.125,
+      "learning_rate": 1.199542678595964e-06,
+      "loss": 1.4717,
+      "step": 90760
+    },
+    {
+      "epoch": 2.640719085435029,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.1976033969233491e-06,
+      "loss": 1.4353,
+      "step": 90780
+    },
+    {
+      "epoch": 2.6413008697675773,
+      "grad_norm": 10.0,
+      "learning_rate": 1.1956641152507342e-06,
+      "loss": 1.4902,
+      "step": 90800
+    },
+    {
+      "epoch": 2.641882654100125,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.1937248335781193e-06,
+      "loss": 1.4954,
+      "step": 90820
+    },
+    {
+      "epoch": 2.6424644384326728,
+      "grad_norm": 12.8125,
+      "learning_rate": 1.1917855519055044e-06,
+      "loss": 1.4846,
+      "step": 90840
+    },
+    {
+      "epoch": 2.643046222765221,
+      "grad_norm": 17.0,
+      "learning_rate": 1.1898462702328895e-06,
+      "loss": 1.5161,
+      "step": 90860
+    },
+    {
+      "epoch": 2.643628007097769,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.1879069885602744e-06,
+      "loss": 1.4884,
+      "step": 90880
+    },
+    {
+      "epoch": 2.644209791430317,
+      "grad_norm": 17.75,
+      "learning_rate": 1.1859677068876595e-06,
+      "loss": 1.4544,
+      "step": 90900
+    },
+    {
+      "epoch": 2.6447915757628646,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.1840284252150446e-06,
+      "loss": 1.5529,
+      "step": 90920
+    },
+    {
+      "epoch": 2.645373360095413,
+      "grad_norm": 13.625,
+      "learning_rate": 1.1820891435424297e-06,
+      "loss": 1.4614,
+      "step": 90940
+    },
+    {
+      "epoch": 2.6459551444279605,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.1801498618698146e-06,
+      "loss": 1.4789,
+      "step": 90960
+    },
+    {
+      "epoch": 2.6465369287605087,
+      "grad_norm": 15.0,
+      "learning_rate": 1.1782105801971997e-06,
+      "loss": 1.5151,
+      "step": 90980
+    },
+    {
+      "epoch": 2.6471187130930565,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.1762712985245848e-06,
+      "loss": 1.4421,
+      "step": 91000
+    },
+    {
+      "epoch": 2.647700497425604,
+      "grad_norm": 12.4375,
+      "learning_rate": 1.1743320168519699e-06,
+      "loss": 1.5315,
+      "step": 91020
+    },
+    {
+      "epoch": 2.6482822817581524,
+      "grad_norm": 14.75,
+      "learning_rate": 1.172392735179355e-06,
+      "loss": 1.4751,
+      "step": 91040
+    },
+    {
+      "epoch": 2.6488640660907,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.17045345350674e-06,
+      "loss": 1.5342,
+      "step": 91060
+    },
+    {
+      "epoch": 2.6494458504232483,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.1685141718341252e-06,
+      "loss": 1.5897,
+      "step": 91080
+    },
+    {
+      "epoch": 2.650027634755796,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.1665748901615103e-06,
+      "loss": 1.4432,
+      "step": 91100
+    },
+    {
+      "epoch": 2.650609419088344,
+      "grad_norm": 12.3125,
+      "learning_rate": 1.1646356084888954e-06,
+      "loss": 1.471,
+      "step": 91120
+    },
+    {
+      "epoch": 2.651191203420892,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.1626963268162805e-06,
+      "loss": 1.5025,
+      "step": 91140
+    },
+    {
+      "epoch": 2.6517729877534397,
+      "grad_norm": 13.25,
+      "learning_rate": 1.1607570451436654e-06,
+      "loss": 1.5553,
+      "step": 91160
+    },
+    {
+      "epoch": 2.652354772085988,
+      "grad_norm": 13.125,
+      "learning_rate": 1.1588177634710505e-06,
+      "loss": 1.4099,
+      "step": 91180
+    },
+    {
+      "epoch": 2.6529365564185357,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.1568784817984356e-06,
+      "loss": 1.5497,
+      "step": 91200
+    },
+    {
+      "epoch": 2.6535183407510834,
+      "grad_norm": 15.5,
+      "learning_rate": 1.1549392001258207e-06,
+      "loss": 1.5543,
+      "step": 91220
+    },
+    {
+      "epoch": 2.6541001250836316,
+      "grad_norm": 8.625,
+      "learning_rate": 1.1529999184532058e-06,
+      "loss": 1.4677,
+      "step": 91240
+    },
+    {
+      "epoch": 2.6546819094161793,
+      "grad_norm": 15.125,
+      "learning_rate": 1.1510606367805909e-06,
+      "loss": 1.5021,
+      "step": 91260
+    },
+    {
+      "epoch": 2.6552636937487275,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.149121355107976e-06,
+      "loss": 1.5207,
+      "step": 91280
+    },
+    {
+      "epoch": 2.6558454780812752,
+      "grad_norm": 14.625,
+      "learning_rate": 1.147182073435361e-06,
+      "loss": 1.5079,
+      "step": 91300
+    },
+    {
+      "epoch": 2.656427262413823,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.1452427917627461e-06,
+      "loss": 1.3869,
+      "step": 91320
+    },
+    {
+      "epoch": 2.657009046746371,
+      "grad_norm": 10.875,
+      "learning_rate": 1.1433035100901312e-06,
+      "loss": 1.4704,
+      "step": 91340
+    },
+    {
+      "epoch": 2.657590831078919,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.1413642284175163e-06,
+      "loss": 1.5158,
+      "step": 91360
+    },
+    {
+      "epoch": 2.658172615411467,
+      "grad_norm": 13.625,
+      "learning_rate": 1.1394249467449012e-06,
+      "loss": 1.4783,
+      "step": 91380
+    },
+    {
+      "epoch": 2.658754399744015,
+      "grad_norm": 12.5,
+      "learning_rate": 1.1374856650722863e-06,
+      "loss": 1.4656,
+      "step": 91400
+    },
+    {
+      "epoch": 2.6593361840765626,
+      "grad_norm": 16.625,
+      "learning_rate": 1.1355463833996714e-06,
+      "loss": 1.5551,
+      "step": 91420
+    },
+    {
+      "epoch": 2.6599179684091108,
+      "grad_norm": 11.4375,
+      "learning_rate": 1.1336071017270563e-06,
+      "loss": 1.5305,
+      "step": 91440
+    },
+    {
+      "epoch": 2.6604997527416585,
+      "grad_norm": 19.0,
+      "learning_rate": 1.1316678200544414e-06,
+      "loss": 1.5697,
+      "step": 91460
+    },
+    {
+      "epoch": 2.6610815370742067,
+      "grad_norm": 10.625,
+      "learning_rate": 1.1297285383818265e-06,
+      "loss": 1.4515,
+      "step": 91480
+    },
+    {
+      "epoch": 2.6616633214067544,
+      "grad_norm": 13.25,
+      "learning_rate": 1.1277892567092116e-06,
+      "loss": 1.5167,
+      "step": 91500
+    },
+    {
+      "epoch": 2.662245105739302,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.1258499750365967e-06,
+      "loss": 1.4129,
+      "step": 91520
+    },
+    {
+      "epoch": 2.6628268900718504,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.1239106933639818e-06,
+      "loss": 1.5106,
+      "step": 91540
+    },
+    {
+      "epoch": 2.6634086744043985,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.1219714116913669e-06,
+      "loss": 1.5203,
+      "step": 91560
+    },
+    {
+      "epoch": 2.6639904587369463,
+      "grad_norm": 14.5,
+      "learning_rate": 1.120032130018752e-06,
+      "loss": 1.5161,
+      "step": 91580
+    },
+    {
+      "epoch": 2.664572243069494,
+      "grad_norm": 14.125,
+      "learning_rate": 1.118092848346137e-06,
+      "loss": 1.5112,
+      "step": 91600
+    },
+    {
+      "epoch": 2.665154027402042,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.1161535666735222e-06,
+      "loss": 1.417,
+      "step": 91620
+    },
+    {
+      "epoch": 2.66573581173459,
+      "grad_norm": 13.375,
+      "learning_rate": 1.114214285000907e-06,
+      "loss": 1.4804,
+      "step": 91640
+    },
+    {
+      "epoch": 2.666317596067138,
+      "grad_norm": 11.5625,
+      "learning_rate": 1.1122750033282922e-06,
+      "loss": 1.4938,
+      "step": 91660
+    },
+    {
+      "epoch": 2.666899380399686,
+      "grad_norm": 11.3125,
+      "learning_rate": 1.1103357216556773e-06,
+      "loss": 1.5055,
+      "step": 91680
+    },
+    {
+      "epoch": 2.6674811647322336,
+      "grad_norm": 9.25,
+      "learning_rate": 1.1083964399830624e-06,
+      "loss": 1.4847,
+      "step": 91700
+    },
+    {
+      "epoch": 2.668062949064782,
+      "grad_norm": 15.0,
+      "learning_rate": 1.1064571583104475e-06,
+      "loss": 1.438,
+      "step": 91720
+    },
+    {
+      "epoch": 2.6686447333973295,
+      "grad_norm": 11.5625,
+      "learning_rate": 1.1045178766378326e-06,
+      "loss": 1.5066,
+      "step": 91740
+    },
+    {
+      "epoch": 2.6692265177298777,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.1025785949652177e-06,
+      "loss": 1.4833,
+      "step": 91760
+    },
+    {
+      "epoch": 2.6698083020624255,
+      "grad_norm": 10.625,
+      "learning_rate": 1.1006393132926028e-06,
+      "loss": 1.509,
+      "step": 91780
+    },
+    {
+      "epoch": 2.670390086394973,
+      "grad_norm": 16.625,
+      "learning_rate": 1.0987000316199878e-06,
+      "loss": 1.584,
+      "step": 91800
+    },
+    {
+      "epoch": 2.6709718707275214,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.096760749947373e-06,
+      "loss": 1.4688,
+      "step": 91820
+    },
+    {
+      "epoch": 2.671553655060069,
+      "grad_norm": 15.625,
+      "learning_rate": 1.0948214682747578e-06,
+      "loss": 1.4821,
+      "step": 91840
+    },
+    {
+      "epoch": 2.6721354393926173,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.092882186602143e-06,
+      "loss": 1.4099,
+      "step": 91860
+    },
+    {
+      "epoch": 2.672717223725165,
+      "grad_norm": 16.375,
+      "learning_rate": 1.090942904929528e-06,
+      "loss": 1.4567,
+      "step": 91880
+    },
+    {
+      "epoch": 2.673299008057713,
+      "grad_norm": 14.125,
+      "learning_rate": 1.0890036232569131e-06,
+      "loss": 1.5113,
+      "step": 91900
+    },
+    {
+      "epoch": 2.673880792390261,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.0870643415842982e-06,
+      "loss": 1.4174,
+      "step": 91920
+    },
+    {
+      "epoch": 2.6744625767228087,
+      "grad_norm": 15.875,
+      "learning_rate": 1.0851250599116831e-06,
+      "loss": 1.5172,
+      "step": 91940
+    },
+    {
+      "epoch": 2.675044361055357,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.0831857782390682e-06,
+      "loss": 1.5045,
+      "step": 91960
+    },
+    {
+      "epoch": 2.6756261453879047,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.0812464965664533e-06,
+      "loss": 1.4731,
+      "step": 91980
+    },
+    {
+      "epoch": 2.6762079297204524,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.0793072148938384e-06,
+      "loss": 1.4835,
+      "step": 92000
+    },
+    {
+      "epoch": 2.6767897140530006,
+      "grad_norm": 12.3125,
+      "learning_rate": 1.0773679332212235e-06,
+      "loss": 1.4878,
+      "step": 92020
+    },
+    {
+      "epoch": 2.6773714983855483,
+      "grad_norm": 14.375,
+      "learning_rate": 1.0754286515486086e-06,
+      "loss": 1.4873,
+      "step": 92040
+    },
+    {
+      "epoch": 2.6779532827180965,
+      "grad_norm": 12.625,
+      "learning_rate": 1.0734893698759937e-06,
+      "loss": 1.5304,
+      "step": 92060
+    },
+    {
+      "epoch": 2.6785350670506443,
+      "grad_norm": 11.8125,
+      "learning_rate": 1.0715500882033788e-06,
+      "loss": 1.4758,
+      "step": 92080
+    },
+    {
+      "epoch": 2.679116851383192,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.0696108065307639e-06,
+      "loss": 1.4777,
+      "step": 92100
+    },
+    {
+      "epoch": 2.67969863571574,
+      "grad_norm": 11.75,
+      "learning_rate": 1.0676715248581488e-06,
+      "loss": 1.4517,
+      "step": 92120
+    },
+    {
+      "epoch": 2.6802804200482884,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.0657322431855339e-06,
+      "loss": 1.5198,
+      "step": 92140
+    },
+    {
+      "epoch": 2.680862204380836,
+      "grad_norm": 14.0,
+      "learning_rate": 1.063792961512919e-06,
+      "loss": 1.4252,
+      "step": 92160
+    },
+    {
+      "epoch": 2.681443988713384,
+      "grad_norm": 12.75,
+      "learning_rate": 1.061853679840304e-06,
+      "loss": 1.5404,
+      "step": 92180
+    },
+    {
+      "epoch": 2.682025773045932,
+      "grad_norm": 12.3125,
+      "learning_rate": 1.0599143981676892e-06,
+      "loss": 1.4618,
+      "step": 92200
+    },
+    {
+      "epoch": 2.68260755737848,
+      "grad_norm": 13.875,
+      "learning_rate": 1.0579751164950743e-06,
+      "loss": 1.6344,
+      "step": 92220
+    },
+    {
+      "epoch": 2.683189341711028,
+      "grad_norm": 14.125,
+      "learning_rate": 1.0560358348224594e-06,
+      "loss": 1.3988,
+      "step": 92240
+    },
+    {
+      "epoch": 2.6837711260435757,
+      "grad_norm": 12.375,
+      "learning_rate": 1.0540965531498445e-06,
+      "loss": 1.4478,
+      "step": 92260
+    },
+    {
+      "epoch": 2.6843529103761234,
+      "grad_norm": 11.875,
+      "learning_rate": 1.0521572714772296e-06,
+      "loss": 1.4711,
+      "step": 92280
+    },
+    {
+      "epoch": 2.6849346947086716,
+      "grad_norm": 16.75,
+      "learning_rate": 1.0502179898046146e-06,
+      "loss": 1.48,
+      "step": 92300
+    },
+    {
+      "epoch": 2.6855164790412194,
+      "grad_norm": 11.25,
+      "learning_rate": 1.0482787081319995e-06,
+      "loss": 1.4759,
+      "step": 92320
+    },
+    {
+      "epoch": 2.6860982633737676,
+      "grad_norm": 13.75,
+      "learning_rate": 1.0463394264593846e-06,
+      "loss": 1.4739,
+      "step": 92340
+    },
+    {
+      "epoch": 2.6866800477063153,
+      "grad_norm": 13.375,
+      "learning_rate": 1.0444001447867697e-06,
+      "loss": 1.4706,
+      "step": 92360
+    },
+    {
+      "epoch": 2.687261832038863,
+      "grad_norm": 12.5,
+      "learning_rate": 1.0424608631141548e-06,
+      "loss": 1.5112,
+      "step": 92380
+    },
+    {
+      "epoch": 2.6878436163714112,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.04052158144154e-06,
+      "loss": 1.5621,
+      "step": 92400
+    },
+    {
+      "epoch": 2.688425400703959,
+      "grad_norm": 12.4375,
+      "learning_rate": 1.038582299768925e-06,
+      "loss": 1.5636,
+      "step": 92420
+    },
+    {
+      "epoch": 2.689007185036507,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.0366430180963101e-06,
+      "loss": 1.4564,
+      "step": 92440
+    },
+    {
+      "epoch": 2.689588969369055,
+      "grad_norm": 12.125,
+      "learning_rate": 1.0347037364236952e-06,
+      "loss": 1.4961,
+      "step": 92460
+    },
+    {
+      "epoch": 2.6901707537016026,
+      "grad_norm": 14.125,
+      "learning_rate": 1.03276445475108e-06,
+      "loss": 1.4921,
+      "step": 92480
+    },
+    {
+      "epoch": 2.690752538034151,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.0308251730784652e-06,
+      "loss": 1.5171,
+      "step": 92500
+    },
+    {
+      "epoch": 2.6913343223666986,
+      "grad_norm": 12.25,
+      "learning_rate": 1.0288858914058503e-06,
+      "loss": 1.4865,
+      "step": 92520
+    },
+    {
+      "epoch": 2.6919161066992467,
+      "grad_norm": 10.375,
+      "learning_rate": 1.0269466097332354e-06,
+      "loss": 1.4231,
+      "step": 92540
+    },
+    {
+      "epoch": 2.6924978910317945,
+      "grad_norm": 17.0,
+      "learning_rate": 1.0250073280606205e-06,
+      "loss": 1.468,
+      "step": 92560
+    },
+    {
+      "epoch": 2.6930796753643422,
+      "grad_norm": 10.8125,
+      "learning_rate": 1.0230680463880056e-06,
+      "loss": 1.5362,
+      "step": 92580
+    },
+    {
+      "epoch": 2.6936614596968904,
+      "grad_norm": 11.625,
+      "learning_rate": 1.0211287647153905e-06,
+      "loss": 1.4689,
+      "step": 92600
+    },
+    {
+      "epoch": 2.694243244029438,
+      "grad_norm": 14.0,
+      "learning_rate": 1.0191894830427756e-06,
+      "loss": 1.4739,
+      "step": 92620
+    },
+    {
+      "epoch": 2.6948250283619863,
+      "grad_norm": 12.125,
+      "learning_rate": 1.0172502013701607e-06,
+      "loss": 1.4328,
+      "step": 92640
+    },
+    {
+      "epoch": 2.695406812694534,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.0153109196975458e-06,
+      "loss": 1.5491,
+      "step": 92660
+    },
+    {
+      "epoch": 2.695988597027082,
+      "grad_norm": 12.125,
+      "learning_rate": 1.0133716380249309e-06,
+      "loss": 1.5193,
+      "step": 92680
+    },
+    {
+      "epoch": 2.69657038135963,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.011432356352316e-06,
+      "loss": 1.539,
+      "step": 92700
+    },
+    {
+      "epoch": 2.6971521656921777,
+      "grad_norm": 13.125,
+      "learning_rate": 1.009493074679701e-06,
+      "loss": 1.4706,
+      "step": 92720
+    },
+    {
+      "epoch": 2.697733950024726,
+      "grad_norm": 15.0,
+      "learning_rate": 1.0075537930070862e-06,
+      "loss": 1.6127,
+      "step": 92740
+    },
+    {
+      "epoch": 2.6983157343572737,
+      "grad_norm": 12.1875,
+      "learning_rate": 1.0056145113344713e-06,
+      "loss": 1.6467,
+      "step": 92760
+    },
+    {
+      "epoch": 2.6988975186898214,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.0036752296618564e-06,
+      "loss": 1.5672,
+      "step": 92780
+    },
+    {
+      "epoch": 2.6994793030223696,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.0017359479892412e-06,
+      "loss": 1.4779,
+      "step": 92800
+    },
+    {
+      "epoch": 2.700061087354918,
+      "grad_norm": 12.375,
+      "learning_rate": 9.997966663166263e-07,
+      "loss": 1.4741,
+      "step": 92820
+    }
+  ],
+  "logging_steps": 20,
+  "max_steps": 103131,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 10314,
+  "total_flos": 6.503393522895938e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}