diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.4502724420716605,
+  "epoch": 0.5503329847542517,
   "eval_steps": 500,
-  "global_step": 8181,
+  "global_step": 9999,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -57274,6 +57274,12732 @@
       "learning_rate": 8.822378335411856e-06,
       "loss": 0.765,
       "step": 8181
+    },
+    {
+      "epoch": 0.4503274808740162,
+      "grad_norm": 0.740364670753479,
+      "learning_rate": 8.822098885673346e-06,
+      "loss": 0.6354,
+      "step": 8182
+    },
+    {
+      "epoch": 0.45038251967637183,
+      "grad_norm": 0.8049225807189941,
+      "learning_rate": 8.821819407208963e-06,
+      "loss": 0.7023,
+      "step": 8183
+    },
+    {
+      "epoch": 0.4504375584787275,
+      "grad_norm": 0.7320911288261414,
+      "learning_rate": 8.821539900020808e-06,
+      "loss": 0.8429,
+      "step": 8184
+    },
+    {
+      "epoch": 0.45049259728108315,
+      "grad_norm": 0.7065376043319702,
+      "learning_rate": 8.821260364110984e-06,
+      "loss": 0.7283,
+      "step": 8185
+    },
+    {
+      "epoch": 0.45054763608343884,
+      "grad_norm": 0.7172972559928894,
+      "learning_rate": 8.820980799481588e-06,
+      "loss": 0.7673,
+      "step": 8186
+    },
+    {
+      "epoch": 0.4506026748857945,
+      "grad_norm": 0.712273895740509,
+      "learning_rate": 8.820701206134724e-06,
+      "loss": 0.7317,
+      "step": 8187
+    },
+    {
+      "epoch": 0.45065771368815016,
+      "grad_norm": 0.6954227685928345,
+      "learning_rate": 8.820421584072492e-06,
+      "loss": 0.7037,
+      "step": 8188
+    },
+    {
+      "epoch": 0.4507127524905058,
+      "grad_norm": 0.6790304780006409,
+      "learning_rate": 8.820141933296994e-06,
+      "loss": 0.7544,
+      "step": 8189
+    },
+    {
+      "epoch": 0.4507677912928615,
+      "grad_norm": 0.7483745813369751,
+      "learning_rate": 8.819862253810332e-06,
+      "loss": 0.7894,
+      "step": 8190
+    },
+    {
+      "epoch": 0.4508228300952171,
+      "grad_norm": 0.7926133871078491,
+      "learning_rate": 8.819582545614608e-06,
+      "loss": 0.8085,
+      "step": 8191
+    },
+    {
+      "epoch": 0.4508778688975728,
+      "grad_norm": 0.8442840576171875,
+      "learning_rate": 8.819302808711924e-06,
+      "loss": 0.8252,
+      "step": 8192
+    },
+    {
+      "epoch": 0.45093290769992844,
+      "grad_norm": 0.8359581232070923,
+      "learning_rate": 8.819023043104383e-06,
+      "loss": 0.8187,
+      "step": 8193
+    },
+    {
+      "epoch": 0.45098794650228413,
+      "grad_norm": 0.7793936133384705,
+      "learning_rate": 8.818743248794085e-06,
+      "loss": 0.8425,
+      "step": 8194
+    },
+    {
+      "epoch": 0.45104298530463977,
+      "grad_norm": 0.735509991645813,
+      "learning_rate": 8.818463425783136e-06,
+      "loss": 0.7781,
+      "step": 8195
+    },
+    {
+      "epoch": 0.45109802410699545,
+      "grad_norm": 0.6735361814498901,
+      "learning_rate": 8.818183574073639e-06,
+      "loss": 0.6987,
+      "step": 8196
+    },
+    {
+      "epoch": 0.4511530629093511,
+      "grad_norm": 0.7780157923698425,
+      "learning_rate": 8.817903693667695e-06,
+      "loss": 0.8474,
+      "step": 8197
+    },
+    {
+      "epoch": 0.4512081017117068,
+      "grad_norm": 0.6714445948600769,
+      "learning_rate": 8.817623784567411e-06,
+      "loss": 0.7216,
+      "step": 8198
+    },
+    {
+      "epoch": 0.4512631405140624,
+      "grad_norm": 0.6311395168304443,
+      "learning_rate": 8.817343846774886e-06,
+      "loss": 0.5724,
+      "step": 8199
+    },
+    {
+      "epoch": 0.4513181793164181,
+      "grad_norm": 0.7446333169937134,
+      "learning_rate": 8.817063880292227e-06,
+      "loss": 0.7867,
+      "step": 8200
+    },
+    {
+      "epoch": 0.45137321811877373,
+      "grad_norm": 0.7684246301651001,
+      "learning_rate": 8.816783885121539e-06,
+      "loss": 0.8141,
+      "step": 8201
+    },
+    {
+      "epoch": 0.4514282569211294,
+      "grad_norm": 0.754781186580658,
+      "learning_rate": 8.816503861264925e-06,
+      "loss": 0.8438,
+      "step": 8202
+    },
+    {
+      "epoch": 0.45148329572348506,
+      "grad_norm": 0.7705762982368469,
+      "learning_rate": 8.816223808724488e-06,
+      "loss": 0.8948,
+      "step": 8203
+    },
+    {
+      "epoch": 0.4515383345258407,
+      "grad_norm": 0.7731552720069885,
+      "learning_rate": 8.815943727502333e-06,
+      "loss": 0.7462,
+      "step": 8204
+    },
+    {
+      "epoch": 0.4515933733281964,
+      "grad_norm": 0.6615393757820129,
+      "learning_rate": 8.81566361760057e-06,
+      "loss": 0.7499,
+      "step": 8205
+    },
+    {
+      "epoch": 0.451648412130552,
+      "grad_norm": 0.724453866481781,
+      "learning_rate": 8.8153834790213e-06,
+      "loss": 0.7382,
+      "step": 8206
+    },
+    {
+      "epoch": 0.4517034509329077,
+      "grad_norm": 0.6369735598564148,
+      "learning_rate": 8.815103311766629e-06,
+      "loss": 0.7452,
+      "step": 8207
+    },
+    {
+      "epoch": 0.45175848973526334,
+      "grad_norm": 0.686000406742096,
+      "learning_rate": 8.814823115838659e-06,
+      "loss": 0.6971,
+      "step": 8208
+    },
+    {
+      "epoch": 0.451813528537619,
+      "grad_norm": 0.7372714281082153,
+      "learning_rate": 8.814542891239505e-06,
+      "loss": 0.8553,
+      "step": 8209
+    },
+    {
+      "epoch": 0.45186856733997466,
+      "grad_norm": 0.8348672986030579,
+      "learning_rate": 8.814262637971264e-06,
+      "loss": 0.7135,
+      "step": 8210
+    },
+    {
+      "epoch": 0.45192360614233035,
+      "grad_norm": 0.7829258441925049,
+      "learning_rate": 8.813982356036049e-06,
+      "loss": 0.7974,
+      "step": 8211
+    },
+    {
+      "epoch": 0.451978644944686,
+      "grad_norm": 0.7013983726501465,
+      "learning_rate": 8.81370204543596e-06,
+      "loss": 0.7531,
+      "step": 8212
+    },
+    {
+      "epoch": 0.45203368374704167,
+      "grad_norm": 0.8424196243286133,
+      "learning_rate": 8.81342170617311e-06,
+      "loss": 0.8217,
+      "step": 8213
+    },
+    {
+      "epoch": 0.4520887225493973,
+      "grad_norm": 0.7113365530967712,
+      "learning_rate": 8.813141338249603e-06,
+      "loss": 0.7728,
+      "step": 8214
+    },
+    {
+      "epoch": 0.452143761351753,
+      "grad_norm": 0.958642303943634,
+      "learning_rate": 8.812860941667545e-06,
+      "loss": 0.7234,
+      "step": 8215
+    },
+    {
+      "epoch": 0.4521988001541086,
+      "grad_norm": 0.6712706685066223,
+      "learning_rate": 8.812580516429045e-06,
+      "loss": 0.6998,
+      "step": 8216
+    },
+    {
+      "epoch": 0.4522538389564643,
+      "grad_norm": 0.7258469462394714,
+      "learning_rate": 8.812300062536212e-06,
+      "loss": 0.6758,
+      "step": 8217
+    },
+    {
+      "epoch": 0.45230887775881995,
+      "grad_norm": 0.735047459602356,
+      "learning_rate": 8.812019579991152e-06,
+      "loss": 0.7045,
+      "step": 8218
+    },
+    {
+      "epoch": 0.45236391656117564,
+      "grad_norm": 0.8339886665344238,
+      "learning_rate": 8.811739068795971e-06,
+      "loss": 0.8069,
+      "step": 8219
+    },
+    {
+      "epoch": 0.45241895536353127,
+      "grad_norm": 0.7170082926750183,
+      "learning_rate": 8.81145852895278e-06,
+      "loss": 0.6345,
+      "step": 8220
+    },
+    {
+      "epoch": 0.45247399416588696,
+      "grad_norm": 0.6892569661140442,
+      "learning_rate": 8.81117796046369e-06,
+      "loss": 0.712,
+      "step": 8221
+    },
+    {
+      "epoch": 0.4525290329682426,
+      "grad_norm": 0.6837140321731567,
+      "learning_rate": 8.810897363330804e-06,
+      "loss": 0.7184,
+      "step": 8222
+    },
+    {
+      "epoch": 0.4525840717705983,
+      "grad_norm": 0.7410069108009338,
+      "learning_rate": 8.810616737556235e-06,
+      "loss": 0.8265,
+      "step": 8223
+    },
+    {
+      "epoch": 0.4526391105729539,
+      "grad_norm": 0.6945875883102417,
+      "learning_rate": 8.810336083142089e-06,
+      "loss": 0.7163,
+      "step": 8224
+    },
+    {
+      "epoch": 0.4526941493753096,
+      "grad_norm": 0.6978884339332581,
+      "learning_rate": 8.810055400090477e-06,
+      "loss": 0.795,
+      "step": 8225
+    },
+    {
+      "epoch": 0.45274918817766524,
+      "grad_norm": 0.7209095358848572,
+      "learning_rate": 8.809774688403509e-06,
+      "loss": 0.7317,
+      "step": 8226
+    },
+    {
+      "epoch": 0.45280422698002093,
+      "grad_norm": 0.7279626727104187,
+      "learning_rate": 8.809493948083294e-06,
+      "loss": 0.7699,
+      "step": 8227
+    },
+    {
+      "epoch": 0.45285926578237656,
+      "grad_norm": 0.7642556428909302,
+      "learning_rate": 8.809213179131943e-06,
+      "loss": 0.8518,
+      "step": 8228
+    },
+    {
+      "epoch": 0.45291430458473225,
+      "grad_norm": 0.6868709325790405,
+      "learning_rate": 8.808932381551565e-06,
+      "loss": 0.737,
+      "step": 8229
+    },
+    {
+      "epoch": 0.4529693433870879,
+      "grad_norm": 0.7012789845466614,
+      "learning_rate": 8.80865155534427e-06,
+      "loss": 0.8146,
+      "step": 8230
+    },
+    {
+      "epoch": 0.4530243821894436,
+      "grad_norm": 0.678683340549469,
+      "learning_rate": 8.808370700512171e-06,
+      "loss": 0.7531,
+      "step": 8231
+    },
+    {
+      "epoch": 0.4530794209917992,
+      "grad_norm": 0.690559983253479,
+      "learning_rate": 8.808089817057377e-06,
+      "loss": 0.6779,
+      "step": 8232
+    },
+    {
+      "epoch": 0.4531344597941549,
+      "grad_norm": 0.7179763317108154,
+      "learning_rate": 8.807808904981997e-06,
+      "loss": 0.8815,
+      "step": 8233
+    },
+    {
+      "epoch": 0.45318949859651053,
+      "grad_norm": 0.7708277702331543,
+      "learning_rate": 8.807527964288147e-06,
+      "loss": 0.8084,
+      "step": 8234
+    },
+    {
+      "epoch": 0.4532445373988662,
+      "grad_norm": 0.6828494071960449,
+      "learning_rate": 8.807246994977936e-06,
+      "loss": 0.7587,
+      "step": 8235
+    },
+    {
+      "epoch": 0.45329957620122185,
+      "grad_norm": 0.7085250616073608,
+      "learning_rate": 8.806965997053475e-06,
+      "loss": 0.7894,
+      "step": 8236
+    },
+    {
+      "epoch": 0.45335461500357754,
+      "grad_norm": 0.7723467946052551,
+      "learning_rate": 8.806684970516876e-06,
+      "loss": 0.7408,
+      "step": 8237
+    },
+    {
+      "epoch": 0.4534096538059332,
+      "grad_norm": 0.8887566328048706,
+      "learning_rate": 8.806403915370253e-06,
+      "loss": 0.9022,
+      "step": 8238
+    },
+    {
+      "epoch": 0.45346469260828887,
+      "grad_norm": 0.7379833459854126,
+      "learning_rate": 8.806122831615718e-06,
+      "loss": 0.8264,
+      "step": 8239
+    },
+    {
+      "epoch": 0.4535197314106445,
+      "grad_norm": 0.903279721736908,
+      "learning_rate": 8.80584171925538e-06,
+      "loss": 0.7432,
+      "step": 8240
+    },
+    {
+      "epoch": 0.4535747702130002,
+      "grad_norm": 0.7671363353729248,
+      "learning_rate": 8.805560578291356e-06,
+      "loss": 0.8109,
+      "step": 8241
+    },
+    {
+      "epoch": 0.4536298090153558,
+      "grad_norm": 0.6047827005386353,
+      "learning_rate": 8.805279408725755e-06,
+      "loss": 0.6628,
+      "step": 8242
+    },
+    {
+      "epoch": 0.4536848478177115,
+      "grad_norm": 1.0570796728134155,
+      "learning_rate": 8.804998210560696e-06,
+      "loss": 0.7981,
+      "step": 8243
+    },
+    {
+      "epoch": 0.45373988662006715,
+      "grad_norm": 0.7116600871086121,
+      "learning_rate": 8.804716983798288e-06,
+      "loss": 0.7601,
+      "step": 8244
+    },
+    {
+      "epoch": 0.45379492542242283,
+      "grad_norm": 0.7162767648696899,
+      "learning_rate": 8.804435728440644e-06,
+      "loss": 0.8389,
+      "step": 8245
+    },
+    {
+      "epoch": 0.45384996422477847,
+      "grad_norm": 0.6715626120567322,
+      "learning_rate": 8.80415444448988e-06,
+      "loss": 0.6377,
+      "step": 8246
+    },
+    {
+      "epoch": 0.4539050030271341,
+      "grad_norm": 0.7168908715248108,
+      "learning_rate": 8.80387313194811e-06,
+      "loss": 0.7946,
+      "step": 8247
+    },
+    {
+      "epoch": 0.4539600418294898,
+      "grad_norm": 0.7497992515563965,
+      "learning_rate": 8.803591790817448e-06,
+      "loss": 0.8026,
+      "step": 8248
+    },
+    {
+      "epoch": 0.4540150806318454,
+      "grad_norm": 0.6665049195289612,
+      "learning_rate": 8.803310421100009e-06,
+      "loss": 0.779,
+      "step": 8249
+    },
+    {
+      "epoch": 0.4540701194342011,
+      "grad_norm": 0.766674280166626,
+      "learning_rate": 8.803029022797905e-06,
+      "loss": 0.7467,
+      "step": 8250
+    },
+    {
+      "epoch": 0.45412515823655675,
+      "grad_norm": 0.7306104302406311,
+      "learning_rate": 8.802747595913255e-06,
+      "loss": 0.8323,
+      "step": 8251
+    },
+    {
+      "epoch": 0.45418019703891244,
+      "grad_norm": 0.6425766944885254,
+      "learning_rate": 8.802466140448169e-06,
+      "loss": 0.7226,
+      "step": 8252
+    },
+    {
+      "epoch": 0.45423523584126807,
+      "grad_norm": 0.7992560267448425,
+      "learning_rate": 8.802184656404769e-06,
+      "loss": 0.7285,
+      "step": 8253
+    },
+    {
+      "epoch": 0.45429027464362376,
+      "grad_norm": 0.6935924887657166,
+      "learning_rate": 8.801903143785164e-06,
+      "loss": 0.5757,
+      "step": 8254
+    },
+    {
+      "epoch": 0.4543453134459794,
+      "grad_norm": 0.7091512084007263,
+      "learning_rate": 8.801621602591473e-06,
+      "loss": 0.7719,
+      "step": 8255
+    },
+    {
+      "epoch": 0.4544003522483351,
+      "grad_norm": 0.851231038570404,
+      "learning_rate": 8.801340032825814e-06,
+      "loss": 0.7804,
+      "step": 8256
+    },
+    {
+      "epoch": 0.4544553910506907,
+      "grad_norm": 0.7443445920944214,
+      "learning_rate": 8.801058434490298e-06,
+      "loss": 0.7172,
+      "step": 8257
+    },
+    {
+      "epoch": 0.4545104298530464,
+      "grad_norm": 0.7156546115875244,
+      "learning_rate": 8.800776807587046e-06,
+      "loss": 0.7756,
+      "step": 8258
+    },
+    {
+      "epoch": 0.45456546865540204,
+      "grad_norm": 0.8027580380439758,
+      "learning_rate": 8.800495152118172e-06,
+      "loss": 0.8035,
+      "step": 8259
+    },
+    {
+      "epoch": 0.4546205074577577,
+      "grad_norm": 0.6868240833282471,
+      "learning_rate": 8.800213468085794e-06,
+      "loss": 0.7159,
+      "step": 8260
+    },
+    {
+      "epoch": 0.45467554626011336,
+      "grad_norm": 0.9127504229545593,
+      "learning_rate": 8.79993175549203e-06,
+      "loss": 0.7705,
+      "step": 8261
+    },
+    {
+      "epoch": 0.45473058506246905,
+      "grad_norm": 0.7074575424194336,
+      "learning_rate": 8.799650014338994e-06,
+      "loss": 0.7841,
+      "step": 8262
+    },
+    {
+      "epoch": 0.4547856238648247,
+      "grad_norm": 0.7462378740310669,
+      "learning_rate": 8.799368244628807e-06,
+      "loss": 0.8125,
+      "step": 8263
+    },
+    {
+      "epoch": 0.4548406626671804,
+      "grad_norm": 0.7510300874710083,
+      "learning_rate": 8.799086446363585e-06,
+      "loss": 0.8354,
+      "step": 8264
+    },
+    {
+      "epoch": 0.454895701469536,
+      "grad_norm": 0.7134591937065125,
+      "learning_rate": 8.798804619545446e-06,
+      "loss": 0.7968,
+      "step": 8265
+    },
+    {
+      "epoch": 0.4549507402718917,
+      "grad_norm": 1.0424071550369263,
+      "learning_rate": 8.798522764176509e-06,
+      "loss": 0.8638,
+      "step": 8266
+    },
+    {
+      "epoch": 0.45500577907424733,
+      "grad_norm": 0.6805267930030823,
+      "learning_rate": 8.79824088025889e-06,
+      "loss": 0.757,
+      "step": 8267
+    },
+    {
+      "epoch": 0.455060817876603,
+      "grad_norm": 0.8145313262939453,
+      "learning_rate": 8.79795896779471e-06,
+      "loss": 0.7589,
+      "step": 8268
+    },
+    {
+      "epoch": 0.45511585667895865,
+      "grad_norm": 0.7611781358718872,
+      "learning_rate": 8.79767702678609e-06,
+      "loss": 0.8426,
+      "step": 8269
+    },
+    {
+      "epoch": 0.45517089548131434,
+      "grad_norm": 0.7639568448066711,
+      "learning_rate": 8.797395057235142e-06,
+      "loss": 0.6609,
+      "step": 8270
+    },
+    {
+      "epoch": 0.45522593428367,
+      "grad_norm": 0.8577544093132019,
+      "learning_rate": 8.79711305914399e-06,
+      "loss": 0.8085,
+      "step": 8271
+    },
+    {
+      "epoch": 0.45528097308602566,
+      "grad_norm": 0.7740383148193359,
+      "learning_rate": 8.796831032514754e-06,
+      "loss": 0.8689,
+      "step": 8272
+    },
+    {
+      "epoch": 0.4553360118883813,
+      "grad_norm": 0.7300885915756226,
+      "learning_rate": 8.796548977349553e-06,
+      "loss": 0.8303,
+      "step": 8273
+    },
+    {
+      "epoch": 0.455391050690737,
+      "grad_norm": 0.6677057147026062,
+      "learning_rate": 8.796266893650504e-06,
+      "loss": 0.7449,
+      "step": 8274
+    },
+    {
+      "epoch": 0.4554460894930926,
+      "grad_norm": 0.7269144058227539,
+      "learning_rate": 8.79598478141973e-06,
+      "loss": 0.8744,
+      "step": 8275
+    },
+    {
+      "epoch": 0.4555011282954483,
+      "grad_norm": 0.7458559274673462,
+      "learning_rate": 8.795702640659351e-06,
+      "loss": 0.8036,
+      "step": 8276
+    },
+    {
+      "epoch": 0.45555616709780394,
+      "grad_norm": 0.7693114280700684,
+      "learning_rate": 8.795420471371487e-06,
+      "loss": 0.7617,
+      "step": 8277
+    },
+    {
+      "epoch": 0.45561120590015963,
+      "grad_norm": 0.7594510316848755,
+      "learning_rate": 8.79513827355826e-06,
+      "loss": 0.7049,
+      "step": 8278
+    },
+    {
+      "epoch": 0.45566624470251527,
+      "grad_norm": 0.7481217980384827,
+      "learning_rate": 8.794856047221786e-06,
+      "loss": 0.804,
+      "step": 8279
+    },
+    {
+      "epoch": 0.45572128350487096,
+      "grad_norm": 0.726859986782074,
+      "learning_rate": 8.794573792364192e-06,
+      "loss": 0.7322,
+      "step": 8280
+    },
+    {
+      "epoch": 0.4557763223072266,
+      "grad_norm": 0.6800149083137512,
+      "learning_rate": 8.794291508987597e-06,
+      "loss": 0.8467,
+      "step": 8281
+    },
+    {
+      "epoch": 0.4558313611095823,
+      "grad_norm": 0.6264217495918274,
+      "learning_rate": 8.794009197094122e-06,
+      "loss": 0.6203,
+      "step": 8282
+    },
+    {
+      "epoch": 0.4558863999119379,
+      "grad_norm": 0.6973850131034851,
+      "learning_rate": 8.79372685668589e-06,
+      "loss": 0.8211,
+      "step": 8283
+    },
+    {
+      "epoch": 0.4559414387142936,
+      "grad_norm": 0.6992879509925842,
+      "learning_rate": 8.793444487765022e-06,
+      "loss": 0.7831,
+      "step": 8284
+    },
+    {
+      "epoch": 0.45599647751664923,
+      "grad_norm": 0.7641519904136658,
+      "learning_rate": 8.793162090333643e-06,
+      "loss": 0.7519,
+      "step": 8285
+    },
+    {
+      "epoch": 0.4560515163190049,
+      "grad_norm": 0.7296152710914612,
+      "learning_rate": 8.79287966439387e-06,
+      "loss": 0.8738,
+      "step": 8286
+    },
+    {
+      "epoch": 0.45610655512136056,
+      "grad_norm": 0.7549383044242859,
+      "learning_rate": 8.79259720994783e-06,
+      "loss": 0.7868,
+      "step": 8287
+    },
+    {
+      "epoch": 0.45616159392371625,
+      "grad_norm": 0.7932083606719971,
+      "learning_rate": 8.792314726997644e-06,
+      "loss": 0.8443,
+      "step": 8288
+    },
+    {
+      "epoch": 0.4562166327260719,
+      "grad_norm": 0.7999894022941589,
+      "learning_rate": 8.792032215545437e-06,
+      "loss": 0.852,
+      "step": 8289
+    },
+    {
+      "epoch": 0.4562716715284275,
+      "grad_norm": 0.8092383742332458,
+      "learning_rate": 8.79174967559333e-06,
+      "loss": 0.7922,
+      "step": 8290
+    },
+    {
+      "epoch": 0.4563267103307832,
+      "grad_norm": 0.7481340169906616,
+      "learning_rate": 8.791467107143447e-06,
+      "loss": 0.7086,
+      "step": 8291
+    },
+    {
+      "epoch": 0.45638174913313884,
+      "grad_norm": 0.8096129298210144,
+      "learning_rate": 8.791184510197912e-06,
+      "loss": 0.6645,
+      "step": 8292
+    },
+    {
+      "epoch": 0.4564367879354945,
+      "grad_norm": 0.7276492118835449,
+      "learning_rate": 8.79090188475885e-06,
+      "loss": 0.7174,
+      "step": 8293
+    },
+    {
+      "epoch": 0.45649182673785016,
+      "grad_norm": 0.815535843372345,
+      "learning_rate": 8.790619230828385e-06,
+      "loss": 0.8622,
+      "step": 8294
+    },
+    {
+      "epoch": 0.45654686554020585,
+      "grad_norm": 0.8191169500350952,
+      "learning_rate": 8.790336548408637e-06,
+      "loss": 0.8666,
+      "step": 8295
+    },
+    {
+      "epoch": 0.4566019043425615,
+      "grad_norm": 0.7449167966842651,
+      "learning_rate": 8.790053837501737e-06,
+      "loss": 0.7728,
+      "step": 8296
+    },
+    {
+      "epoch": 0.45665694314491717,
+      "grad_norm": 0.7311065196990967,
+      "learning_rate": 8.789771098109808e-06,
+      "loss": 0.8059,
+      "step": 8297
+    },
+    {
+      "epoch": 0.4567119819472728,
+      "grad_norm": 0.7381907105445862,
+      "learning_rate": 8.789488330234971e-06,
+      "loss": 0.7722,
+      "step": 8298
+    },
+    {
+      "epoch": 0.4567670207496285,
+      "grad_norm": 0.8180661201477051,
+      "learning_rate": 8.789205533879355e-06,
+      "loss": 0.9032,
+      "step": 8299
+    },
+    {
+      "epoch": 0.4568220595519841,
+      "grad_norm": 0.7993118762969971,
+      "learning_rate": 8.788922709045087e-06,
+      "loss": 0.8065,
+      "step": 8300
+    },
+    {
+      "epoch": 0.4568770983543398,
+      "grad_norm": 0.8449206948280334,
+      "learning_rate": 8.788639855734287e-06,
+      "loss": 0.7895,
+      "step": 8301
+    },
+    {
+      "epoch": 0.45693213715669545,
+      "grad_norm": 0.9224583506584167,
+      "learning_rate": 8.788356973949084e-06,
+      "loss": 0.78,
+      "step": 8302
+    },
+    {
+      "epoch": 0.45698717595905114,
+      "grad_norm": 0.7109915614128113,
+      "learning_rate": 8.788074063691604e-06,
+      "loss": 0.8029,
+      "step": 8303
+    },
+    {
+      "epoch": 0.4570422147614068,
+      "grad_norm": 0.7372310757637024,
+      "learning_rate": 8.787791124963976e-06,
+      "loss": 0.8118,
+      "step": 8304
+    },
+    {
+      "epoch": 0.45709725356376246,
+      "grad_norm": 0.8127168416976929,
+      "learning_rate": 8.787508157768323e-06,
+      "loss": 0.8665,
+      "step": 8305
+    },
+    {
+      "epoch": 0.4571522923661181,
+      "grad_norm": 0.7193050980567932,
+      "learning_rate": 8.787225162106771e-06,
+      "loss": 0.749,
+      "step": 8306
+    },
+    {
+      "epoch": 0.4572073311684738,
+      "grad_norm": 0.8825041651725769,
+      "learning_rate": 8.786942137981449e-06,
+      "loss": 0.9651,
+      "step": 8307
+    },
+    {
+      "epoch": 0.4572623699708294,
+      "grad_norm": 0.6854885816574097,
+      "learning_rate": 8.786659085394485e-06,
+      "loss": 0.8259,
+      "step": 8308
+    },
+    {
+      "epoch": 0.4573174087731851,
+      "grad_norm": 0.6698010563850403,
+      "learning_rate": 8.786376004348004e-06,
+      "loss": 0.7212,
+      "step": 8309
+    },
+    {
+      "epoch": 0.45737244757554074,
+      "grad_norm": 0.7706398963928223,
+      "learning_rate": 8.786092894844132e-06,
+      "loss": 0.719,
+      "step": 8310
+    },
+    {
+      "epoch": 0.45742748637789643,
+      "grad_norm": 0.8905620574951172,
+      "learning_rate": 8.785809756885002e-06,
+      "loss": 0.7518,
+      "step": 8311
+    },
+    {
+      "epoch": 0.45748252518025206,
+      "grad_norm": 0.7537117004394531,
+      "learning_rate": 8.78552659047274e-06,
+      "loss": 0.8267,
+      "step": 8312
+    },
+    {
+      "epoch": 0.45753756398260775,
+      "grad_norm": 0.7840754985809326,
+      "learning_rate": 8.78524339560947e-06,
+      "loss": 0.8417,
+      "step": 8313
+    },
+    {
+      "epoch": 0.4575926027849634,
+      "grad_norm": 0.7373713254928589,
+      "learning_rate": 8.784960172297327e-06,
+      "loss": 0.784,
+      "step": 8314
+    },
+    {
+      "epoch": 0.4576476415873191,
+      "grad_norm": 0.6648432016372681,
+      "learning_rate": 8.784676920538436e-06,
+      "loss": 0.7252,
+      "step": 8315
+    },
+    {
+      "epoch": 0.4577026803896747,
+      "grad_norm": 0.7904912829399109,
+      "learning_rate": 8.784393640334925e-06,
+      "loss": 0.7777,
+      "step": 8316
+    },
+    {
+      "epoch": 0.4577577191920304,
+      "grad_norm": 0.7691501379013062,
+      "learning_rate": 8.784110331688927e-06,
+      "loss": 0.733,
+      "step": 8317
+    },
+    {
+      "epoch": 0.45781275799438603,
+      "grad_norm": 0.6054617762565613,
+      "learning_rate": 8.783826994602566e-06,
+      "loss": 0.6367,
+      "step": 8318
+    },
+    {
+      "epoch": 0.4578677967967417,
+      "grad_norm": 0.7495457530021667,
+      "learning_rate": 8.783543629077976e-06,
+      "loss": 0.8672,
+      "step": 8319
+    },
+    {
+      "epoch": 0.45792283559909736,
+      "grad_norm": 0.6979867815971375,
+      "learning_rate": 8.783260235117283e-06,
+      "loss": 0.7338,
+      "step": 8320
+    },
+    {
+      "epoch": 0.45797787440145304,
+      "grad_norm": 0.6927759647369385,
+      "learning_rate": 8.78297681272262e-06,
+      "loss": 0.6925,
+      "step": 8321
+    },
+    {
+      "epoch": 0.4580329132038087,
+      "grad_norm": 0.9076687097549438,
+      "learning_rate": 8.782693361896115e-06,
+      "loss": 0.8225,
+      "step": 8322
+    },
+    {
+      "epoch": 0.45808795200616437,
+      "grad_norm": 0.7990893721580505,
+      "learning_rate": 8.782409882639902e-06,
+      "loss": 0.8144,
+      "step": 8323
+    },
+    {
+      "epoch": 0.45814299080852,
+      "grad_norm": 0.7958230376243591,
+      "learning_rate": 8.782126374956107e-06,
+      "loss": 0.7717,
+      "step": 8324
+    },
+    {
+      "epoch": 0.4581980296108757,
+      "grad_norm": 0.7694645524024963,
+      "learning_rate": 8.781842838846861e-06,
+      "loss": 0.8314,
+      "step": 8325
+    },
+    {
+      "epoch": 0.4582530684132313,
+      "grad_norm": 0.8653621077537537,
+      "learning_rate": 8.781559274314297e-06,
+      "loss": 0.7567,
+      "step": 8326
+    },
+    {
+      "epoch": 0.458308107215587,
+      "grad_norm": 0.7834668755531311,
+      "learning_rate": 8.781275681360548e-06,
+      "loss": 0.7431,
+      "step": 8327
+    },
+    {
+      "epoch": 0.45836314601794265,
+      "grad_norm": 0.6800104975700378,
+      "learning_rate": 8.780992059987742e-06,
+      "loss": 0.8266,
+      "step": 8328
+    },
+    {
+      "epoch": 0.45841818482029834,
+      "grad_norm": 0.7274910807609558,
+      "learning_rate": 8.780708410198011e-06,
+      "loss": 0.7358,
+      "step": 8329
+    },
+    {
+      "epoch": 0.45847322362265397,
+      "grad_norm": 0.8102344870567322,
+      "learning_rate": 8.780424731993488e-06,
+      "loss": 0.7397,
+      "step": 8330
+    },
+    {
+      "epoch": 0.45852826242500966,
+      "grad_norm": 0.7536956071853638,
+      "learning_rate": 8.780141025376305e-06,
+      "loss": 0.7053,
+      "step": 8331
+    },
+    {
+      "epoch": 0.4585833012273653,
+      "grad_norm": 0.678535521030426,
+      "learning_rate": 8.779857290348594e-06,
+      "loss": 0.792,
+      "step": 8332
+    },
+    {
+      "epoch": 0.4586383400297209,
+      "grad_norm": 0.8847216963768005,
+      "learning_rate": 8.779573526912487e-06,
+      "loss": 0.8117,
+      "step": 8333
+    },
+    {
+      "epoch": 0.4586933788320766,
+      "grad_norm": 0.6997288465499878,
+      "learning_rate": 8.779289735070117e-06,
+      "loss": 0.7797,
+      "step": 8334
+    },
+    {
+      "epoch": 0.45874841763443225,
+      "grad_norm": 0.7445441484451294,
+      "learning_rate": 8.779005914823617e-06,
+      "loss": 0.7505,
+      "step": 8335
+    },
+    {
+      "epoch": 0.45880345643678794,
+      "grad_norm": 0.618844211101532,
+      "learning_rate": 8.778722066175121e-06,
+      "loss": 0.661,
+      "step": 8336
+    },
+    {
+      "epoch": 0.45885849523914357,
+      "grad_norm": 0.6810492873191833,
+      "learning_rate": 8.778438189126761e-06,
+      "loss": 0.6819,
+      "step": 8337
+    },
+    {
+      "epoch": 0.45891353404149926,
+      "grad_norm": 0.6785591244697571,
+      "learning_rate": 8.778154283680671e-06,
+      "loss": 0.7808,
+      "step": 8338
+    },
+    {
+      "epoch": 0.4589685728438549,
+      "grad_norm": 0.7461212873458862,
+      "learning_rate": 8.777870349838984e-06,
+      "loss": 0.8566,
+      "step": 8339
+    },
+    {
+      "epoch": 0.4590236116462106,
+      "grad_norm": 0.6731496453285217,
+      "learning_rate": 8.777586387603836e-06,
+      "loss": 0.823,
+      "step": 8340
+    },
+    {
+      "epoch": 0.4590786504485662,
+      "grad_norm": 0.7295553684234619,
+      "learning_rate": 8.77730239697736e-06,
+      "loss": 0.9229,
+      "step": 8341
+    },
+    {
+      "epoch": 0.4591336892509219,
+      "grad_norm": 0.783275842666626,
+      "learning_rate": 8.77701837796169e-06,
+      "loss": 0.782,
+      "step": 8342
+    },
+    {
+      "epoch": 0.45918872805327754,
+      "grad_norm": 0.6952852606773376,
+      "learning_rate": 8.77673433055896e-06,
+      "loss": 0.7977,
+      "step": 8343
+    },
+    {
+      "epoch": 0.45924376685563323,
+      "grad_norm": 0.7381969094276428,
+      "learning_rate": 8.776450254771305e-06,
+      "loss": 0.768,
+      "step": 8344
+    },
+    {
+      "epoch": 0.45929880565798886,
+      "grad_norm": 0.7911093831062317,
+      "learning_rate": 8.776166150600862e-06,
+      "loss": 0.8284,
+      "step": 8345
+    },
+    {
+      "epoch": 0.45935384446034455,
+      "grad_norm": 0.7319246530532837,
+      "learning_rate": 8.775882018049765e-06,
+      "loss": 0.8135,
+      "step": 8346
+    },
+    {
+      "epoch": 0.4594088832627002,
+      "grad_norm": 0.7888429760932922,
+      "learning_rate": 8.77559785712015e-06,
+      "loss": 0.9001,
+      "step": 8347
+    },
+    {
+      "epoch": 0.4594639220650559,
+      "grad_norm": 0.6983326077461243,
+      "learning_rate": 8.775313667814151e-06,
+      "loss": 0.7537,
+      "step": 8348
+    },
+    {
+      "epoch": 0.4595189608674115,
+      "grad_norm": 0.7532416582107544,
+      "learning_rate": 8.775029450133905e-06,
+      "loss": 0.8307,
+      "step": 8349
+    },
+    {
+      "epoch": 0.4595739996697672,
+      "grad_norm": 0.7159993052482605,
+      "learning_rate": 8.774745204081549e-06,
+      "loss": 0.7874,
+      "step": 8350
+    },
+    {
+      "epoch": 0.45962903847212283,
+      "grad_norm": 0.6898767352104187,
+      "learning_rate": 8.774460929659218e-06,
+      "loss": 0.7453,
+      "step": 8351
+    },
+    {
+      "epoch": 0.4596840772744785,
+      "grad_norm": 0.6833236813545227,
+      "learning_rate": 8.774176626869051e-06,
+      "loss": 0.7281,
+      "step": 8352
+    },
+    {
+      "epoch": 0.45973911607683415,
+      "grad_norm": 0.7840244770050049,
+      "learning_rate": 8.77389229571318e-06,
+      "loss": 0.7194,
+      "step": 8353
+    },
+    {
+      "epoch": 0.45979415487918984,
+      "grad_norm": 0.7920441627502441,
+      "learning_rate": 8.773607936193747e-06,
+      "loss": 0.7135,
+      "step": 8354
+    },
+    {
+      "epoch": 0.4598491936815455,
+      "grad_norm": 0.7395668625831604,
+      "learning_rate": 8.773323548312884e-06,
+      "loss": 0.8162,
+      "step": 8355
+    },
+    {
+      "epoch": 0.45990423248390117,
+      "grad_norm": 0.7854128479957581,
+      "learning_rate": 8.773039132072734e-06,
+      "loss": 0.8252,
+      "step": 8356
+    },
+    {
+      "epoch": 0.4599592712862568,
+      "grad_norm": 0.694997251033783,
+      "learning_rate": 8.772754687475431e-06,
+      "loss": 0.6627,
+      "step": 8357
+    },
+    {
+      "epoch": 0.4600143100886125,
+      "grad_norm": 0.7698866724967957,
+      "learning_rate": 8.772470214523112e-06,
+      "loss": 0.8814,
+      "step": 8358
+    },
+    {
+      "epoch": 0.4600693488909681,
+      "grad_norm": 0.7323407530784607,
+      "learning_rate": 8.77218571321792e-06,
+      "loss": 0.7769,
+      "step": 8359
+    },
+    {
+      "epoch": 0.4601243876933238,
+      "grad_norm": 0.6637027263641357,
+      "learning_rate": 8.771901183561986e-06,
+      "loss": 0.6741,
+      "step": 8360
+    },
+    {
+      "epoch": 0.46017942649567944,
+      "grad_norm": 0.7423702478408813,
+      "learning_rate": 8.771616625557455e-06,
+      "loss": 0.7303,
+      "step": 8361
+    },
+    {
+      "epoch": 0.46023446529803513,
+      "grad_norm": 0.7599568367004395,
+      "learning_rate": 8.771332039206463e-06,
+      "loss": 0.8161,
+      "step": 8362
+    },
+    {
+      "epoch": 0.46028950410039077,
+      "grad_norm": 0.9063183069229126,
+      "learning_rate": 8.771047424511148e-06,
+      "loss": 0.8098,
+      "step": 8363
+    },
+    {
+      "epoch": 0.46034454290274646,
+      "grad_norm": 0.658210813999176,
+      "learning_rate": 8.770762781473651e-06,
+      "loss": 0.7097,
+      "step": 8364
+    },
+    {
+      "epoch": 0.4603995817051021,
+      "grad_norm": 0.8396975994110107,
+      "learning_rate": 8.770478110096111e-06,
+      "loss": 0.8731,
+      "step": 8365
+    },
+    {
+      "epoch": 0.4604546205074578,
+      "grad_norm": 0.7334815263748169,
+      "learning_rate": 8.770193410380663e-06,
+      "loss": 0.7689,
+      "step": 8366
+    },
+    {
+      "epoch": 0.4605096593098134,
+      "grad_norm": 0.8220386505126953,
+      "learning_rate": 8.769908682329453e-06,
+      "loss": 0.8139,
+      "step": 8367
+    },
+    {
+      "epoch": 0.4605646981121691,
+      "grad_norm": 0.8077995181083679,
+      "learning_rate": 8.76962392594462e-06,
+      "loss": 0.7379,
+      "step": 8368
+    },
+    {
+      "epoch": 0.46061973691452474,
+      "grad_norm": 0.8007730841636658,
+      "learning_rate": 8.7693391412283e-06,
+      "loss": 0.7835,
+      "step": 8369
+    },
+    {
+      "epoch": 0.4606747757168804,
+      "grad_norm": 0.7108187079429626,
+      "learning_rate": 8.769054328182637e-06,
+      "loss": 0.6787,
+      "step": 8370
+    },
+    {
+      "epoch": 0.46072981451923606,
+      "grad_norm": 0.7623056173324585,
+      "learning_rate": 8.768769486809772e-06,
+      "loss": 0.8056,
+      "step": 8371
+    },
+    {
+      "epoch": 0.46078485332159175,
+      "grad_norm": 0.6991614103317261,
+      "learning_rate": 8.768484617111843e-06,
+      "loss": 0.7404,
+      "step": 8372
+    },
+    {
+      "epoch": 0.4608398921239474,
+      "grad_norm": 0.7531471848487854,
+      "learning_rate": 8.768199719090991e-06,
+      "loss": 0.8104,
+      "step": 8373
+    },
+    {
+      "epoch": 0.46089493092630307,
+      "grad_norm": 1.0271111726760864,
+      "learning_rate": 8.76791479274936e-06,
+      "loss": 0.9028,
+      "step": 8374
+    },
+    {
+      "epoch": 0.4609499697286587,
+      "grad_norm": 0.7346897125244141,
+      "learning_rate": 8.76762983808909e-06,
+      "loss": 0.8179,
+      "step": 8375
+    },
+    {
+      "epoch": 0.46100500853101434,
+      "grad_norm": 0.6413559913635254,
+      "learning_rate": 8.767344855112324e-06,
+      "loss": 0.7995,
+      "step": 8376
+    },
+    {
+      "epoch": 0.46106004733337,
+      "grad_norm": 0.7187537550926208,
+      "learning_rate": 8.767059843821199e-06,
+      "loss": 0.7973,
+      "step": 8377
+    },
+    {
+      "epoch": 0.46111508613572566,
+      "grad_norm": 0.6819092035293579,
+      "learning_rate": 8.766774804217864e-06,
+      "loss": 0.8255,
+      "step": 8378
+    },
+    {
+      "epoch": 0.46117012493808135,
+      "grad_norm": 0.683318018913269,
+      "learning_rate": 8.766489736304457e-06,
+      "loss": 0.6794,
+      "step": 8379
+    },
+    {
+      "epoch": 0.461225163740437,
+      "grad_norm": 0.7345470786094666,
+      "learning_rate": 8.76620464008312e-06,
+      "loss": 0.8741,
+      "step": 8380
+    },
+    {
+      "epoch": 0.46128020254279267,
+      "grad_norm": 0.7369397282600403,
+      "learning_rate": 8.765919515556e-06,
+      "loss": 0.8301,
+      "step": 8381
+    },
+    {
+      "epoch": 0.4613352413451483,
+      "grad_norm": 0.7304979562759399,
+      "learning_rate": 8.765634362725233e-06,
+      "loss": 0.7507,
+      "step": 8382
+    },
+    {
+      "epoch": 0.461390280147504,
+      "grad_norm": 0.7968454957008362,
+      "learning_rate": 8.765349181592969e-06,
+      "loss": 0.7396,
+      "step": 8383
+    },
+    {
+      "epoch": 0.46144531894985963,
+      "grad_norm": 0.691439151763916,
+      "learning_rate": 8.765063972161347e-06,
+      "loss": 0.7199,
+      "step": 8384
+    },
+    {
+      "epoch": 0.4615003577522153,
+      "grad_norm": 0.8355879187583923,
+      "learning_rate": 8.764778734432513e-06,
+      "loss": 0.7369,
+      "step": 8385
+    },
+    {
+      "epoch": 0.46155539655457095,
+      "grad_norm": 0.908017098903656,
+      "learning_rate": 8.76449346840861e-06,
+      "loss": 0.8271,
+      "step": 8386
+    },
+    {
+      "epoch": 0.46161043535692664,
+      "grad_norm": 0.6426172852516174,
+      "learning_rate": 8.764208174091781e-06,
+      "loss": 0.6646,
+      "step": 8387
+    },
+    {
+      "epoch": 0.4616654741592823,
+      "grad_norm": 0.7003652453422546,
+      "learning_rate": 8.763922851484171e-06,
+      "loss": 0.7272,
+      "step": 8388
+    },
+    {
+      "epoch": 0.46172051296163796,
+      "grad_norm": 0.7470494508743286,
+      "learning_rate": 8.763637500587925e-06,
+      "loss": 0.8333,
+      "step": 8389
+    },
+    {
+      "epoch": 0.4617755517639936,
+      "grad_norm": 0.6974903345108032,
+      "learning_rate": 8.763352121405187e-06,
+      "loss": 0.834,
+      "step": 8390
+    },
+    {
+      "epoch": 0.4618305905663493,
+      "grad_norm": 0.8146659135818481,
+      "learning_rate": 8.7630667139381e-06,
+      "loss": 0.724,
+      "step": 8391
+    },
+    {
+      "epoch": 0.4618856293687049,
+      "grad_norm": 0.6614096164703369,
+      "learning_rate": 8.762781278188813e-06,
+      "loss": 0.6822,
+      "step": 8392
+    },
+    {
+      "epoch": 0.4619406681710606,
+      "grad_norm": 0.712944746017456,
+      "learning_rate": 8.762495814159469e-06,
+      "loss": 0.7864,
+      "step": 8393
+    },
+    {
+      "epoch": 0.46199570697341624,
+      "grad_norm": 0.7531552910804749,
+      "learning_rate": 8.762210321852213e-06,
+      "loss": 0.7494,
+      "step": 8394
+    },
+    {
+      "epoch": 0.46205074577577193,
+      "grad_norm": 0.8150199055671692,
+      "learning_rate": 8.761924801269191e-06,
+      "loss": 0.7869,
+      "step": 8395
+    },
+    {
+      "epoch": 0.46210578457812757,
+      "grad_norm": 0.8586462736129761,
+      "learning_rate": 8.76163925241255e-06,
+      "loss": 0.7647,
+      "step": 8396
+    },
+    {
+      "epoch": 0.46216082338048325,
+      "grad_norm": 0.7258061766624451,
+      "learning_rate": 8.761353675284434e-06,
+      "loss": 0.7672,
+      "step": 8397
+    },
+    {
+      "epoch": 0.4622158621828389,
+      "grad_norm": 0.6592851281166077,
+      "learning_rate": 8.761068069886992e-06,
+      "loss": 0.7488,
+      "step": 8398
+    },
+    {
+      "epoch": 0.4622709009851946,
+      "grad_norm": 0.7410836219787598,
+      "learning_rate": 8.760782436222368e-06,
+      "loss": 0.6669,
+      "step": 8399
+    },
+    {
+      "epoch": 0.4623259397875502,
+      "grad_norm": 0.7121642231941223,
+      "learning_rate": 8.76049677429271e-06,
+      "loss": 0.7005,
+      "step": 8400
+    },
+    {
+      "epoch": 0.4623809785899059,
+      "grad_norm": 0.7170663475990295,
+      "learning_rate": 8.760211084100166e-06,
+      "loss": 0.8154,
+      "step": 8401
+    },
+    {
+      "epoch": 0.46243601739226153,
+      "grad_norm": 0.6851769685745239,
+      "learning_rate": 8.759925365646882e-06,
+      "loss": 0.7948,
+      "step": 8402
+    },
+    {
+      "epoch": 0.4624910561946172,
+      "grad_norm": 0.7728533744812012,
+      "learning_rate": 8.759639618935006e-06,
+      "loss": 0.8263,
+      "step": 8403
+    },
+    {
+      "epoch": 0.46254609499697286,
+      "grad_norm": 0.7276784777641296,
+      "learning_rate": 8.759353843966682e-06,
+      "loss": 0.6992,
+      "step": 8404
+    },
+    {
+      "epoch": 0.46260113379932855,
+      "grad_norm": 0.7533649802207947,
+      "learning_rate": 8.759068040744063e-06,
+      "loss": 0.7744,
+      "step": 8405
+    },
+    {
+      "epoch": 0.4626561726016842,
+      "grad_norm": 0.6911979913711548,
+      "learning_rate": 8.758782209269294e-06,
+      "loss": 0.6977,
+      "step": 8406
+    },
+    {
+      "epoch": 0.46271121140403987,
+      "grad_norm": 0.6723766922950745,
+      "learning_rate": 8.758496349544526e-06,
+      "loss": 0.7286,
+      "step": 8407
+    },
+    {
+      "epoch": 0.4627662502063955,
+      "grad_norm": 0.7327921390533447,
+      "learning_rate": 8.758210461571903e-06,
+      "loss": 0.7708,
+      "step": 8408
+    },
+    {
+      "epoch": 0.4628212890087512,
+      "grad_norm": 0.7498626708984375,
+      "learning_rate": 8.757924545353578e-06,
+      "loss": 0.7476,
+      "step": 8409
+    },
+    {
+      "epoch": 0.4628763278111068,
+      "grad_norm": 0.8944914937019348,
+      "learning_rate": 8.757638600891696e-06,
+      "loss": 0.7814,
+      "step": 8410
+    },
+    {
+      "epoch": 0.4629313666134625,
+      "grad_norm": 0.7242841124534607,
+      "learning_rate": 8.757352628188411e-06,
+      "loss": 0.7564,
+      "step": 8411
+    },
+    {
+      "epoch": 0.46298640541581815,
+      "grad_norm": 0.6706324815750122,
+      "learning_rate": 8.757066627245866e-06,
+      "loss": 0.7792,
+      "step": 8412
+    },
+    {
+      "epoch": 0.46304144421817384,
+      "grad_norm": 0.8044155836105347,
+      "learning_rate": 8.756780598066218e-06,
+      "loss": 0.7873,
+      "step": 8413
+    },
+    {
+      "epoch": 0.46309648302052947,
+      "grad_norm": 0.9265295267105103,
+      "learning_rate": 8.75649454065161e-06,
+      "loss": 0.878,
+      "step": 8414
+    },
+    {
+      "epoch": 0.46315152182288516,
+      "grad_norm": 0.8162378668785095,
+      "learning_rate": 8.756208455004194e-06,
+      "loss": 0.8758,
+      "step": 8415
+    },
+    {
+      "epoch": 0.4632065606252408,
+      "grad_norm": 0.7081401348114014,
+      "learning_rate": 8.755922341126121e-06,
+      "loss": 0.8053,
+      "step": 8416
+    },
+    {
+      "epoch": 0.4632615994275965,
+      "grad_norm": 0.663885235786438,
+      "learning_rate": 8.755636199019544e-06,
+      "loss": 0.7456,
+      "step": 8417
+    },
+    {
+      "epoch": 0.4633166382299521,
+      "grad_norm": 0.6934974193572998,
+      "learning_rate": 8.755350028686608e-06,
+      "loss": 0.7316,
+      "step": 8418
+    },
+    {
+      "epoch": 0.46337167703230775,
+      "grad_norm": 0.7162168025970459,
+      "learning_rate": 8.755063830129467e-06,
+      "loss": 0.8566,
+      "step": 8419
+    },
+    {
+      "epoch": 0.46342671583466344,
+      "grad_norm": 0.7507640719413757,
+      "learning_rate": 8.75477760335027e-06,
+      "loss": 0.8141,
+      "step": 8420
+    },
+    {
+      "epoch": 0.46348175463701907,
+      "grad_norm": 0.6853382587432861,
+      "learning_rate": 8.754491348351172e-06,
+      "loss": 0.6995,
+      "step": 8421
+    },
+    {
+      "epoch": 0.46353679343937476,
+      "grad_norm": 0.6421381831169128,
+      "learning_rate": 8.75420506513432e-06,
+      "loss": 0.6344,
+      "step": 8422
+    },
+    {
+      "epoch": 0.4635918322417304,
+      "grad_norm": 0.8042624592781067,
+      "learning_rate": 8.753918753701868e-06,
+      "loss": 0.7506,
+      "step": 8423
+    },
+    {
+      "epoch": 0.4636468710440861,
+      "grad_norm": 0.7184088230133057,
+      "learning_rate": 8.753632414055969e-06,
+      "loss": 0.7997,
+      "step": 8424
+    },
+    {
+      "epoch": 0.4637019098464417,
+      "grad_norm": 0.749919593334198,
+      "learning_rate": 8.753346046198773e-06,
+      "loss": 0.8168,
+      "step": 8425
+    },
+    {
+      "epoch": 0.4637569486487974,
+      "grad_norm": 0.6583670973777771,
+      "learning_rate": 8.753059650132433e-06,
+      "loss": 0.6615,
+      "step": 8426
+    },
+    {
+      "epoch": 0.46381198745115304,
+      "grad_norm": 0.7560496926307678,
+      "learning_rate": 8.7527732258591e-06,
+      "loss": 0.7221,
+      "step": 8427
+    },
+    {
+      "epoch": 0.46386702625350873,
+      "grad_norm": 0.7031972408294678,
+      "learning_rate": 8.752486773380928e-06,
+      "loss": 0.8124,
+      "step": 8428
+    },
+    {
+      "epoch": 0.46392206505586436,
+      "grad_norm": 0.684124767780304,
+      "learning_rate": 8.752200292700072e-06,
+      "loss": 0.6862,
+      "step": 8429
+    },
+    {
+      "epoch": 0.46397710385822005,
+      "grad_norm": 0.8015589118003845,
+      "learning_rate": 8.751913783818682e-06,
+      "loss": 0.7863,
+      "step": 8430
+    },
+    {
+      "epoch": 0.4640321426605757,
+      "grad_norm": 0.6815705299377441,
+      "learning_rate": 8.751627246738912e-06,
+      "loss": 0.8116,
+      "step": 8431
+    },
+    {
+      "epoch": 0.4640871814629314,
+      "grad_norm": 0.7402058839797974,
+      "learning_rate": 8.751340681462914e-06,
+      "loss": 0.7341,
+      "step": 8432
+    },
+    {
+      "epoch": 0.464142220265287,
+      "grad_norm": 0.7484470009803772,
+      "learning_rate": 8.751054087992848e-06,
+      "loss": 0.8103,
+      "step": 8433
+    },
+    {
+      "epoch": 0.4641972590676427,
+      "grad_norm": 0.8148707151412964,
+      "learning_rate": 8.75076746633086e-06,
+      "loss": 0.8995,
+      "step": 8434
+    },
+    {
+      "epoch": 0.46425229786999833,
+      "grad_norm": 0.6403086185455322,
+      "learning_rate": 8.750480816479107e-06,
+      "loss": 0.6705,
+      "step": 8435
+    },
+    {
+      "epoch": 0.464307336672354,
+      "grad_norm": 0.7787690758705139,
+      "learning_rate": 8.750194138439748e-06,
+      "loss": 0.854,
+      "step": 8436
+    },
+    {
+      "epoch": 0.46436237547470965,
+      "grad_norm": 0.6975393891334534,
+      "learning_rate": 8.749907432214931e-06,
+      "loss": 0.7588,
+      "step": 8437
+    },
+    {
+      "epoch": 0.46441741427706534,
+      "grad_norm": 0.8002430200576782,
+      "learning_rate": 8.749620697806812e-06,
+      "loss": 0.8244,
+      "step": 8438
+    },
+    {
+      "epoch": 0.464472453079421,
+      "grad_norm": 0.8049100637435913,
+      "learning_rate": 8.74933393521755e-06,
+      "loss": 0.7686,
+      "step": 8439
+    },
+    {
+      "epoch": 0.46452749188177667,
+      "grad_norm": 0.6716971397399902,
+      "learning_rate": 8.749047144449298e-06,
+      "loss": 0.7823,
+      "step": 8440
+    },
+    {
+      "epoch": 0.4645825306841323,
+      "grad_norm": 0.7292011380195618,
+      "learning_rate": 8.748760325504212e-06,
+      "loss": 0.7643,
+      "step": 8441
+    },
+    {
+      "epoch": 0.464637569486488,
+      "grad_norm": 0.6823335886001587,
+      "learning_rate": 8.748473478384444e-06,
+      "loss": 0.7539,
+      "step": 8442
+    },
+    {
+      "epoch": 0.4646926082888436,
+      "grad_norm": 0.761730968952179,
+      "learning_rate": 8.748186603092155e-06,
+      "loss": 0.7279,
+      "step": 8443
+    },
+    {
+      "epoch": 0.4647476470911993,
+      "grad_norm": 0.694007933139801,
+      "learning_rate": 8.747899699629498e-06,
+      "loss": 0.7907,
+      "step": 8444
+    },
+    {
+      "epoch": 0.46480268589355495,
+      "grad_norm": 0.7638683319091797,
+      "learning_rate": 8.74761276799863e-06,
+      "loss": 0.7278,
+      "step": 8445
+    },
+    {
+      "epoch": 0.46485772469591063,
+      "grad_norm": 0.6281229853630066,
+      "learning_rate": 8.747325808201708e-06,
+      "loss": 0.6609,
+      "step": 8446
+    },
+    {
+      "epoch": 0.46491276349826627,
+      "grad_norm": 0.7273259162902832,
+      "learning_rate": 8.747038820240887e-06,
+      "loss": 0.7553,
+      "step": 8447
+    },
+    {
+      "epoch": 0.46496780230062196,
+      "grad_norm": 0.807482898235321,
+      "learning_rate": 8.746751804118326e-06,
+      "loss": 0.7783,
+      "step": 8448
+    },
+    {
+      "epoch": 0.4650228411029776,
+      "grad_norm": 0.7088230848312378,
+      "learning_rate": 8.746464759836182e-06,
+      "loss": 0.762,
+      "step": 8449
+    },
+    {
+      "epoch": 0.4650778799053333,
+      "grad_norm": 0.7039850354194641,
+      "learning_rate": 8.746177687396612e-06,
+      "loss": 0.7811,
+      "step": 8450
+    },
+    {
+      "epoch": 0.4651329187076889,
+      "grad_norm": 0.7154161334037781,
+      "learning_rate": 8.745890586801773e-06,
+      "loss": 0.76,
+      "step": 8451
+    },
+    {
+      "epoch": 0.4651879575100446,
+      "grad_norm": 0.6738846302032471,
+      "learning_rate": 8.745603458053822e-06,
+      "loss": 0.7119,
+      "step": 8452
+    },
+    {
+      "epoch": 0.46524299631240024,
+      "grad_norm": 0.6615753173828125,
+      "learning_rate": 8.745316301154919e-06,
+      "loss": 0.8061,
+      "step": 8453
+    },
+    {
+      "epoch": 0.4652980351147559,
+      "grad_norm": 0.7285076379776001,
+      "learning_rate": 8.74502911610722e-06,
+      "loss": 0.7522,
+      "step": 8454
+    },
+    {
+      "epoch": 0.46535307391711156,
+      "grad_norm": 0.7100732922554016,
+      "learning_rate": 8.744741902912886e-06,
+      "loss": 0.7665,
+      "step": 8455
+    },
+    {
+      "epoch": 0.46540811271946725,
+      "grad_norm": 0.6564487814903259,
+      "learning_rate": 8.744454661574074e-06,
+      "loss": 0.7352,
+      "step": 8456
+    },
+    {
+      "epoch": 0.4654631515218229,
+      "grad_norm": 0.689549446105957,
+      "learning_rate": 8.744167392092944e-06,
+      "loss": 0.7011,
+      "step": 8457
+    },
+    {
+      "epoch": 0.46551819032417857,
+      "grad_norm": 0.6660958528518677,
+      "learning_rate": 8.743880094471651e-06,
+      "loss": 0.7074,
+      "step": 8458
+    },
+    {
+      "epoch": 0.4655732291265342,
+      "grad_norm": 0.7470804452896118,
+      "learning_rate": 8.743592768712361e-06,
+      "loss": 0.6684,
+      "step": 8459
+    },
+    {
+      "epoch": 0.4656282679288899,
+      "grad_norm": 0.8058002591133118,
+      "learning_rate": 8.743305414817227e-06,
+      "loss": 0.7945,
+      "step": 8460
+    },
+    {
+      "epoch": 0.4656833067312455,
+      "grad_norm": 0.7756261825561523,
+      "learning_rate": 8.743018032788413e-06,
+      "loss": 0.8442,
+      "step": 8461
+    },
+    {
+      "epoch": 0.46573834553360116,
+      "grad_norm": 0.9267478585243225,
+      "learning_rate": 8.742730622628077e-06,
+      "loss": 0.8721,
+      "step": 8462
+    },
+    {
+      "epoch": 0.46579338433595685,
+      "grad_norm": 0.8684219121932983,
+      "learning_rate": 8.74244318433838e-06,
+      "loss": 0.7833,
+      "step": 8463
+    },
+    {
+      "epoch": 0.4658484231383125,
+      "grad_norm": 0.7060475945472717,
+      "learning_rate": 8.742155717921481e-06,
+      "loss": 0.7724,
+      "step": 8464
+    },
+    {
+      "epoch": 0.4659034619406682,
+      "grad_norm": 0.7316318154335022,
+      "learning_rate": 8.741868223379543e-06,
+      "loss": 0.7489,
+      "step": 8465
+    },
+    {
+      "epoch": 0.4659585007430238,
+      "grad_norm": 0.8131282925605774,
+      "learning_rate": 8.741580700714724e-06,
+      "loss": 0.7453,
+      "step": 8466
+    },
+    {
+      "epoch": 0.4660135395453795,
+      "grad_norm": 0.6985850930213928,
+      "learning_rate": 8.741293149929187e-06,
+      "loss": 0.7083,
+      "step": 8467
+    },
+    {
+      "epoch": 0.46606857834773513,
+      "grad_norm": 0.7512301206588745,
+      "learning_rate": 8.74100557102509e-06,
+      "loss": 0.7343,
+      "step": 8468
+    },
+    {
+      "epoch": 0.4661236171500908,
+      "grad_norm": 0.7547290921211243,
+      "learning_rate": 8.740717964004596e-06,
+      "loss": 0.8358,
+      "step": 8469
+    },
+    {
+      "epoch": 0.46617865595244645,
+      "grad_norm": 0.9091271758079529,
+      "learning_rate": 8.740430328869868e-06,
+      "loss": 0.762,
+      "step": 8470
+    },
+    {
+      "epoch": 0.46623369475480214,
+      "grad_norm": 0.6960130333900452,
+      "learning_rate": 8.740142665623069e-06,
+      "loss": 0.7317,
+      "step": 8471
+    },
+    {
+      "epoch": 0.4662887335571578,
+      "grad_norm": 0.684309184551239,
+      "learning_rate": 8.739854974266357e-06,
+      "loss": 0.7653,
+      "step": 8472
+    },
+    {
+      "epoch": 0.46634377235951346,
+      "grad_norm": 0.7669411301612854,
+      "learning_rate": 8.739567254801898e-06,
+      "loss": 0.7152,
+      "step": 8473
+    },
+    {
+      "epoch": 0.4663988111618691,
+      "grad_norm": 0.7072784900665283,
+      "learning_rate": 8.73927950723185e-06,
+      "loss": 0.7508,
+      "step": 8474
+    },
+    {
+      "epoch": 0.4664538499642248,
+      "grad_norm": 0.7249277234077454,
+      "learning_rate": 8.73899173155838e-06,
+      "loss": 0.7469,
+      "step": 8475
+    },
+    {
+      "epoch": 0.4665088887665804,
+      "grad_norm": 0.7664750218391418,
+      "learning_rate": 8.738703927783647e-06,
+      "loss": 0.8692,
+      "step": 8476
+    },
+    {
+      "epoch": 0.4665639275689361,
+      "grad_norm": 0.7579765319824219,
+      "learning_rate": 8.738416095909818e-06,
+      "loss": 0.8283,
+      "step": 8477
+    },
+    {
+      "epoch": 0.46661896637129174,
+      "grad_norm": 0.7066456079483032,
+      "learning_rate": 8.738128235939054e-06,
+      "loss": 0.7125,
+      "step": 8478
+    },
+    {
+      "epoch": 0.46667400517364743,
+      "grad_norm": 0.766106367111206,
+      "learning_rate": 8.737840347873518e-06,
+      "loss": 0.7683,
+      "step": 8479
+    },
+    {
+      "epoch": 0.46672904397600307,
+      "grad_norm": 0.7599226236343384,
+      "learning_rate": 8.737552431715374e-06,
+      "loss": 0.8375,
+      "step": 8480
+    },
+    {
+      "epoch": 0.46678408277835876,
+      "grad_norm": 0.6955341100692749,
+      "learning_rate": 8.737264487466789e-06,
+      "loss": 0.7012,
+      "step": 8481
+    },
+    {
+      "epoch": 0.4668391215807144,
+      "grad_norm": 0.6096246242523193,
+      "learning_rate": 8.736976515129923e-06,
+      "loss": 0.6126,
+      "step": 8482
+    },
+    {
+      "epoch": 0.4668941603830701,
+      "grad_norm": 0.7469536066055298,
+      "learning_rate": 8.73668851470694e-06,
+      "loss": 0.7675,
+      "step": 8483
+    },
+    {
+      "epoch": 0.4669491991854257,
+      "grad_norm": 0.8018775582313538,
+      "learning_rate": 8.73640048620001e-06,
+      "loss": 0.7372,
+      "step": 8484
+    },
+    {
+      "epoch": 0.4670042379877814,
+      "grad_norm": 0.7446827292442322,
+      "learning_rate": 8.736112429611293e-06,
+      "loss": 0.7277,
+      "step": 8485
+    },
+    {
+      "epoch": 0.46705927679013703,
+      "grad_norm": 0.6292026042938232,
+      "learning_rate": 8.735824344942954e-06,
+      "loss": 0.6172,
+      "step": 8486
+    },
+    {
+      "epoch": 0.4671143155924927,
+      "grad_norm": 0.7207980751991272,
+      "learning_rate": 8.735536232197159e-06,
+      "loss": 0.8363,
+      "step": 8487
+    },
+    {
+      "epoch": 0.46716935439484836,
+      "grad_norm": 0.8585891127586365,
+      "learning_rate": 8.735248091376073e-06,
+      "loss": 0.8006,
+      "step": 8488
+    },
+    {
+      "epoch": 0.46722439319720405,
+      "grad_norm": 0.8149702548980713,
+      "learning_rate": 8.734959922481863e-06,
+      "loss": 0.7869,
+      "step": 8489
+    },
+    {
+      "epoch": 0.4672794319995597,
+      "grad_norm": 0.7113268971443176,
+      "learning_rate": 8.734671725516695e-06,
+      "loss": 0.7774,
+      "step": 8490
+    },
+    {
+      "epoch": 0.46733447080191537,
+      "grad_norm": 0.6940683722496033,
+      "learning_rate": 8.734383500482733e-06,
+      "loss": 0.7157,
+      "step": 8491
+    },
+    {
+      "epoch": 0.467389509604271,
+      "grad_norm": 0.7823536396026611,
+      "learning_rate": 8.734095247382145e-06,
+      "loss": 0.8161,
+      "step": 8492
+    },
+    {
+      "epoch": 0.4674445484066267,
+      "grad_norm": 0.7094922065734863,
+      "learning_rate": 8.733806966217096e-06,
+      "loss": 0.7593,
+      "step": 8493
+    },
+    {
+      "epoch": 0.4674995872089823,
+      "grad_norm": 0.656432569026947,
+      "learning_rate": 8.733518656989753e-06,
+      "loss": 0.7853,
+      "step": 8494
+    },
+    {
+      "epoch": 0.467554626011338,
+      "grad_norm": 0.6715715527534485,
+      "learning_rate": 8.733230319702284e-06,
+      "loss": 0.839,
+      "step": 8495
+    },
+    {
+      "epoch": 0.46760966481369365,
+      "grad_norm": 0.7496705055236816,
+      "learning_rate": 8.732941954356854e-06,
+      "loss": 0.8231,
+      "step": 8496
+    },
+    {
+      "epoch": 0.46766470361604934,
+      "grad_norm": 0.7728047370910645,
+      "learning_rate": 8.732653560955635e-06,
+      "loss": 0.7852,
+      "step": 8497
+    },
+    {
+      "epoch": 0.46771974241840497,
+      "grad_norm": 1.5637458562850952,
+      "learning_rate": 8.732365139500787e-06,
+      "loss": 0.7749,
+      "step": 8498
+    },
+    {
+      "epoch": 0.46777478122076066,
+      "grad_norm": 0.6603190898895264,
+      "learning_rate": 8.732076689994484e-06,
+      "loss": 0.6628,
+      "step": 8499
+    },
+    {
+      "epoch": 0.4678298200231163,
+      "grad_norm": 0.7170974612236023,
+      "learning_rate": 8.73178821243889e-06,
+      "loss": 0.7855,
+      "step": 8500
+    },
+    {
+      "epoch": 0.467884858825472,
+      "grad_norm": 0.7220103740692139,
+      "learning_rate": 8.731499706836175e-06,
+      "loss": 0.7035,
+      "step": 8501
+    },
+    {
+      "epoch": 0.4679398976278276,
+      "grad_norm": 0.6940942406654358,
+      "learning_rate": 8.731211173188507e-06,
+      "loss": 0.7857,
+      "step": 8502
+    },
+    {
+      "epoch": 0.4679949364301833,
+      "grad_norm": 2.441596508026123,
+      "learning_rate": 8.730922611498057e-06,
+      "loss": 0.695,
+      "step": 8503
+    },
+    {
+      "epoch": 0.46804997523253894,
+      "grad_norm": 0.7654910087585449,
+      "learning_rate": 8.730634021766989e-06,
+      "loss": 0.788,
+      "step": 8504
+    },
+    {
+      "epoch": 0.4681050140348946,
+      "grad_norm": 0.791824996471405,
+      "learning_rate": 8.730345403997475e-06,
+      "loss": 0.7899,
+      "step": 8505
+    },
+    {
+      "epoch": 0.46816005283725026,
+      "grad_norm": 0.6863934993743896,
+      "learning_rate": 8.730056758191682e-06,
+      "loss": 0.7402,
+      "step": 8506
+    },
+    {
+      "epoch": 0.4682150916396059,
+      "grad_norm": 0.7920359373092651,
+      "learning_rate": 8.729768084351783e-06,
+      "loss": 0.7835,
+      "step": 8507
+    },
+    {
+      "epoch": 0.4682701304419616,
+      "grad_norm": 0.7077129483222961,
+      "learning_rate": 8.729479382479944e-06,
+      "loss": 0.7761,
+      "step": 8508
+    },
+    {
+      "epoch": 0.4683251692443172,
+      "grad_norm": 0.6870049238204956,
+      "learning_rate": 8.729190652578337e-06,
+      "loss": 0.8169,
+      "step": 8509
+    },
+    {
+      "epoch": 0.4683802080466729,
+      "grad_norm": 0.6802713871002197,
+      "learning_rate": 8.728901894649131e-06,
+      "loss": 0.7914,
+      "step": 8510
+    },
+    {
+      "epoch": 0.46843524684902854,
+      "grad_norm": 0.6645112633705139,
+      "learning_rate": 8.728613108694497e-06,
+      "loss": 0.7543,
+      "step": 8511
+    },
+    {
+      "epoch": 0.46849028565138423,
+      "grad_norm": 0.708292543888092,
+      "learning_rate": 8.728324294716604e-06,
+      "loss": 0.7015,
+      "step": 8512
+    },
+    {
+      "epoch": 0.46854532445373986,
+      "grad_norm": 0.7444465160369873,
+      "learning_rate": 8.728035452717625e-06,
+      "loss": 0.7999,
+      "step": 8513
+    },
+    {
+      "epoch": 0.46860036325609555,
+      "grad_norm": 0.7028616666793823,
+      "learning_rate": 8.727746582699728e-06,
+      "loss": 0.8094,
+      "step": 8514
+    },
+    {
+      "epoch": 0.4686554020584512,
+      "grad_norm": 0.7063208222389221,
+      "learning_rate": 8.727457684665088e-06,
+      "loss": 0.8028,
+      "step": 8515
+    },
+    {
+      "epoch": 0.4687104408608069,
+      "grad_norm": 0.8455138802528381,
+      "learning_rate": 8.727168758615871e-06,
+      "loss": 0.7691,
+      "step": 8516
+    },
+    {
+      "epoch": 0.4687654796631625,
+      "grad_norm": 1.0325778722763062,
+      "learning_rate": 8.726879804554252e-06,
+      "loss": 0.7042,
+      "step": 8517
+    },
+    {
+      "epoch": 0.4688205184655182,
+      "grad_norm": 0.7352754473686218,
+      "learning_rate": 8.726590822482402e-06,
+      "loss": 0.8467,
+      "step": 8518
+    },
+    {
+      "epoch": 0.46887555726787383,
+      "grad_norm": 0.7247193455696106,
+      "learning_rate": 8.726301812402494e-06,
+      "loss": 0.8034,
+      "step": 8519
+    },
+    {
+      "epoch": 0.4689305960702295,
+      "grad_norm": 0.6876820921897888,
+      "learning_rate": 8.726012774316699e-06,
+      "loss": 0.7308,
+      "step": 8520
+    },
+    {
+      "epoch": 0.46898563487258516,
+      "grad_norm": 0.6987231969833374,
+      "learning_rate": 8.725723708227188e-06,
+      "loss": 0.7655,
+      "step": 8521
+    },
+    {
+      "epoch": 0.46904067367494084,
+      "grad_norm": 0.7471843361854553,
+      "learning_rate": 8.725434614136135e-06,
+      "loss": 0.7271,
+      "step": 8522
+    },
+    {
+      "epoch": 0.4690957124772965,
+      "grad_norm": 0.7564642429351807,
+      "learning_rate": 8.725145492045715e-06,
+      "loss": 0.7335,
+      "step": 8523
+    },
+    {
+      "epoch": 0.46915075127965217,
+      "grad_norm": 0.7488992214202881,
+      "learning_rate": 8.724856341958095e-06,
+      "loss": 0.8815,
+      "step": 8524
+    },
+    {
+      "epoch": 0.4692057900820078,
+      "grad_norm": 0.6776759028434753,
+      "learning_rate": 8.724567163875455e-06,
+      "loss": 0.7452,
+      "step": 8525
+    },
+    {
+      "epoch": 0.4692608288843635,
+      "grad_norm": 0.6905981302261353,
+      "learning_rate": 8.724277957799963e-06,
+      "loss": 0.6815,
+      "step": 8526
+    },
+    {
+      "epoch": 0.4693158676867191,
+      "grad_norm": 0.7392297983169556,
+      "learning_rate": 8.723988723733795e-06,
+      "loss": 0.7546,
+      "step": 8527
+    },
+    {
+      "epoch": 0.4693709064890748,
+      "grad_norm": 0.7479110360145569,
+      "learning_rate": 8.723699461679128e-06,
+      "loss": 0.7455,
+      "step": 8528
+    },
+    {
+      "epoch": 0.46942594529143045,
+      "grad_norm": 0.7231360673904419,
+      "learning_rate": 8.723410171638129e-06,
+      "loss": 0.7611,
+      "step": 8529
+    },
+    {
+      "epoch": 0.46948098409378614,
+      "grad_norm": 0.7493714690208435,
+      "learning_rate": 8.723120853612976e-06,
+      "loss": 0.6997,
+      "step": 8530
+    },
+    {
+      "epoch": 0.46953602289614177,
+      "grad_norm": 0.8056793808937073,
+      "learning_rate": 8.722831507605844e-06,
+      "loss": 0.7431,
+      "step": 8531
+    },
+    {
+      "epoch": 0.46959106169849746,
+      "grad_norm": 0.7528547048568726,
+      "learning_rate": 8.722542133618907e-06,
+      "loss": 0.8798,
+      "step": 8532
+    },
+    {
+      "epoch": 0.4696461005008531,
+      "grad_norm": 0.6964863538742065,
+      "learning_rate": 8.72225273165434e-06,
+      "loss": 0.8462,
+      "step": 8533
+    },
+    {
+      "epoch": 0.4697011393032088,
+      "grad_norm": 0.7354302406311035,
+      "learning_rate": 8.721963301714318e-06,
+      "loss": 0.7882,
+      "step": 8534
+    },
+    {
+      "epoch": 0.4697561781055644,
+      "grad_norm": 0.7365205883979797,
+      "learning_rate": 8.721673843801014e-06,
+      "loss": 0.7483,
+      "step": 8535
+    },
+    {
+      "epoch": 0.4698112169079201,
+      "grad_norm": 0.7485378384590149,
+      "learning_rate": 8.72138435791661e-06,
+      "loss": 0.8539,
+      "step": 8536
+    },
+    {
+      "epoch": 0.46986625571027574,
+      "grad_norm": 0.7674353718757629,
+      "learning_rate": 8.721094844063274e-06,
+      "loss": 0.834,
+      "step": 8537
+    },
+    {
+      "epoch": 0.4699212945126314,
+      "grad_norm": 0.7054184079170227,
+      "learning_rate": 8.720805302243185e-06,
+      "loss": 0.7938,
+      "step": 8538
+    },
+    {
+      "epoch": 0.46997633331498706,
+      "grad_norm": 0.7414574027061462,
+      "learning_rate": 8.72051573245852e-06,
+      "loss": 0.7932,
+      "step": 8539
+    },
+    {
+      "epoch": 0.47003137211734275,
+      "grad_norm": 0.6734428405761719,
+      "learning_rate": 8.720226134711455e-06,
+      "loss": 0.8775,
+      "step": 8540
+    },
+    {
+      "epoch": 0.4700864109196984,
+      "grad_norm": 0.6588559150695801,
+      "learning_rate": 8.719936509004166e-06,
+      "loss": 0.6985,
+      "step": 8541
+    },
+    {
+      "epoch": 0.4701414497220541,
+      "grad_norm": 0.6557223200798035,
+      "learning_rate": 8.71964685533883e-06,
+      "loss": 0.7243,
+      "step": 8542
+    },
+    {
+      "epoch": 0.4701964885244097,
+      "grad_norm": 0.7876269221305847,
+      "learning_rate": 8.719357173717624e-06,
+      "loss": 0.8075,
+      "step": 8543
+    },
+    {
+      "epoch": 0.4702515273267654,
+      "grad_norm": 0.8346554040908813,
+      "learning_rate": 8.719067464142726e-06,
+      "loss": 0.8427,
+      "step": 8544
+    },
+    {
+      "epoch": 0.47030656612912103,
+      "grad_norm": 0.7190483808517456,
+      "learning_rate": 8.718777726616311e-06,
+      "loss": 0.7689,
+      "step": 8545
+    },
+    {
+      "epoch": 0.4703616049314767,
+      "grad_norm": 1.303118109703064,
+      "learning_rate": 8.718487961140558e-06,
+      "loss": 0.7537,
+      "step": 8546
+    },
+    {
+      "epoch": 0.47041664373383235,
+      "grad_norm": 0.7733024954795837,
+      "learning_rate": 8.718198167717647e-06,
+      "loss": 0.747,
+      "step": 8547
+    },
+    {
+      "epoch": 0.470471682536188,
+      "grad_norm": 0.6692484617233276,
+      "learning_rate": 8.717908346349751e-06,
+      "loss": 0.725,
+      "step": 8548
+    },
+    {
+      "epoch": 0.4705267213385437,
+      "grad_norm": 0.9639461636543274,
+      "learning_rate": 8.717618497039054e-06,
+      "loss": 0.8642,
+      "step": 8549
+    },
+    {
+      "epoch": 0.4705817601408993,
+      "grad_norm": 0.7584646344184875,
+      "learning_rate": 8.717328619787728e-06,
+      "loss": 0.8174,
+      "step": 8550
+    },
+    {
+      "epoch": 0.470636798943255,
+      "grad_norm": 0.7051709890365601,
+      "learning_rate": 8.717038714597957e-06,
+      "loss": 0.7962,
+      "step": 8551
+    },
+    {
+      "epoch": 0.47069183774561063,
+      "grad_norm": 0.738913893699646,
+      "learning_rate": 8.716748781471918e-06,
+      "loss": 0.7367,
+      "step": 8552
+    },
+    {
+      "epoch": 0.4707468765479663,
+      "grad_norm": 0.7027214169502258,
+      "learning_rate": 8.716458820411791e-06,
+      "loss": 0.7613,
+      "step": 8553
+    },
+    {
+      "epoch": 0.47080191535032195,
+      "grad_norm": 0.6701993346214294,
+      "learning_rate": 8.716168831419754e-06,
+      "loss": 0.638,
+      "step": 8554
+    },
+    {
+      "epoch": 0.47085695415267764,
+      "grad_norm": 0.7422072887420654,
+      "learning_rate": 8.715878814497984e-06,
+      "loss": 0.8338,
+      "step": 8555
+    },
+    {
+      "epoch": 0.4709119929550333,
+      "grad_norm": 0.985992968082428,
+      "learning_rate": 8.715588769648667e-06,
+      "loss": 0.7765,
+      "step": 8556
+    },
+    {
+      "epoch": 0.47096703175738897,
+      "grad_norm": 0.6937553882598877,
+      "learning_rate": 8.715298696873978e-06,
+      "loss": 0.7306,
+      "step": 8557
+    },
+    {
+      "epoch": 0.4710220705597446,
+      "grad_norm": 1.1683214902877808,
+      "learning_rate": 8.715008596176099e-06,
+      "loss": 0.7782,
+      "step": 8558
+    },
+    {
+      "epoch": 0.4710771093621003,
+      "grad_norm": 0.7493681907653809,
+      "learning_rate": 8.714718467557209e-06,
+      "loss": 0.9166,
+      "step": 8559
+    },
+    {
+      "epoch": 0.4711321481644559,
+      "grad_norm": 0.7562084794044495,
+      "learning_rate": 8.71442831101949e-06,
+      "loss": 0.7999,
+      "step": 8560
+    },
+    {
+      "epoch": 0.4711871869668116,
+      "grad_norm": 0.7950266003608704,
+      "learning_rate": 8.71413812656512e-06,
+      "loss": 0.8094,
+      "step": 8561
+    },
+    {
+      "epoch": 0.47124222576916724,
+      "grad_norm": 1.1411044597625732,
+      "learning_rate": 8.713847914196287e-06,
+      "loss": 0.7631,
+      "step": 8562
+    },
+    {
+      "epoch": 0.47129726457152293,
+      "grad_norm": 0.7270122170448303,
+      "learning_rate": 8.713557673915162e-06,
+      "loss": 0.7529,
+      "step": 8563
+    },
+    {
+      "epoch": 0.47135230337387857,
+      "grad_norm": 0.8138573169708252,
+      "learning_rate": 8.713267405723935e-06,
+      "loss": 0.8215,
+      "step": 8564
+    },
+    {
+      "epoch": 0.47140734217623426,
+      "grad_norm": 0.732982873916626,
+      "learning_rate": 8.712977109624783e-06,
+      "loss": 0.7099,
+      "step": 8565
+    },
+    {
+      "epoch": 0.4714623809785899,
+      "grad_norm": 0.7307591438293457,
+      "learning_rate": 8.712686785619888e-06,
+      "loss": 0.7035,
+      "step": 8566
+    },
+    {
+      "epoch": 0.4715174197809456,
+      "grad_norm": 0.8684857487678528,
+      "learning_rate": 8.712396433711434e-06,
+      "loss": 0.8605,
+      "step": 8567
+    },
+    {
+      "epoch": 0.4715724585833012,
+      "grad_norm": 0.7490718364715576,
+      "learning_rate": 8.712106053901603e-06,
+      "loss": 0.7439,
+      "step": 8568
+    },
+    {
+      "epoch": 0.4716274973856569,
+      "grad_norm": 0.8572973012924194,
+      "learning_rate": 8.711815646192575e-06,
+      "loss": 0.8187,
+      "step": 8569
+    },
+    {
+      "epoch": 0.47168253618801254,
+      "grad_norm": 0.785270094871521,
+      "learning_rate": 8.711525210586536e-06,
+      "loss": 0.7812,
+      "step": 8570
+    },
+    {
+      "epoch": 0.4717375749903682,
+      "grad_norm": 0.683651864528656,
+      "learning_rate": 8.711234747085663e-06,
+      "loss": 0.7682,
+      "step": 8571
+    },
+    {
+      "epoch": 0.47179261379272386,
+      "grad_norm": 0.7990714907646179,
+      "learning_rate": 8.710944255692147e-06,
+      "loss": 0.8114,
+      "step": 8572
+    },
+    {
+      "epoch": 0.47184765259507955,
+      "grad_norm": 0.9354856610298157,
+      "learning_rate": 8.710653736408165e-06,
+      "loss": 0.7353,
+      "step": 8573
+    },
+    {
+      "epoch": 0.4719026913974352,
+      "grad_norm": 0.8309356570243835,
+      "learning_rate": 8.710363189235904e-06,
+      "loss": 0.8635,
+      "step": 8574
+    },
+    {
+      "epoch": 0.47195773019979087,
+      "grad_norm": 0.7018463015556335,
+      "learning_rate": 8.710072614177547e-06,
+      "loss": 0.6372,
+      "step": 8575
+    },
+    {
+      "epoch": 0.4720127690021465,
+      "grad_norm": 0.7626469135284424,
+      "learning_rate": 8.709782011235277e-06,
+      "loss": 0.7684,
+      "step": 8576
+    },
+    {
+      "epoch": 0.4720678078045022,
+      "grad_norm": 0.6995826959609985,
+      "learning_rate": 8.70949138041128e-06,
+      "loss": 0.7301,
+      "step": 8577
+    },
+    {
+      "epoch": 0.4721228466068578,
+      "grad_norm": 0.719307541847229,
+      "learning_rate": 8.709200721707736e-06,
+      "loss": 0.7437,
+      "step": 8578
+    },
+    {
+      "epoch": 0.4721778854092135,
+      "grad_norm": 0.7355539202690125,
+      "learning_rate": 8.708910035126832e-06,
+      "loss": 0.7926,
+      "step": 8579
+    },
+    {
+      "epoch": 0.47223292421156915,
+      "grad_norm": 0.7262680530548096,
+      "learning_rate": 8.708619320670755e-06,
+      "loss": 0.7641,
+      "step": 8580
+    },
+    {
+      "epoch": 0.47228796301392484,
+      "grad_norm": 0.844745934009552,
+      "learning_rate": 8.708328578341687e-06,
+      "loss": 0.7228,
+      "step": 8581
+    },
+    {
+      "epoch": 0.47234300181628047,
+      "grad_norm": 0.8169287443161011,
+      "learning_rate": 8.708037808141814e-06,
+      "loss": 0.7076,
+      "step": 8582
+    },
+    {
+      "epoch": 0.47239804061863616,
+      "grad_norm": 0.7342209219932556,
+      "learning_rate": 8.707747010073322e-06,
+      "loss": 0.7997,
+      "step": 8583
+    },
+    {
+      "epoch": 0.4724530794209918,
+      "grad_norm": 0.7138200402259827,
+      "learning_rate": 8.707456184138394e-06,
+      "loss": 0.7796,
+      "step": 8584
+    },
+    {
+      "epoch": 0.4725081182233475,
+      "grad_norm": 0.7168061137199402,
+      "learning_rate": 8.70716533033922e-06,
+      "loss": 0.6876,
+      "step": 8585
+    },
+    {
+      "epoch": 0.4725631570257031,
+      "grad_norm": 0.7256397604942322,
+      "learning_rate": 8.706874448677982e-06,
+      "loss": 0.8296,
+      "step": 8586
+    },
+    {
+      "epoch": 0.4726181958280588,
+      "grad_norm": 0.8232730627059937,
+      "learning_rate": 8.70658353915687e-06,
+      "loss": 0.8001,
+      "step": 8587
+    },
+    {
+      "epoch": 0.47267323463041444,
+      "grad_norm": 0.7110162973403931,
+      "learning_rate": 8.706292601778067e-06,
+      "loss": 0.7061,
+      "step": 8588
+    },
+    {
+      "epoch": 0.47272827343277013,
+      "grad_norm": 0.9466721415519714,
+      "learning_rate": 8.706001636543761e-06,
+      "loss": 0.8713,
+      "step": 8589
+    },
+    {
+      "epoch": 0.47278331223512576,
+      "grad_norm": 0.7017776370048523,
+      "learning_rate": 8.705710643456138e-06,
+      "loss": 0.759,
+      "step": 8590
+    },
+    {
+      "epoch": 0.4728383510374814,
+      "grad_norm": 0.7140772938728333,
+      "learning_rate": 8.705419622517386e-06,
+      "loss": 0.6962,
+      "step": 8591
+    },
+    {
+      "epoch": 0.4728933898398371,
+      "grad_norm": 1.1076452732086182,
+      "learning_rate": 8.705128573729694e-06,
+      "loss": 0.8264,
+      "step": 8592
+    },
+    {
+      "epoch": 0.4729484286421927,
+      "grad_norm": 0.7308200597763062,
+      "learning_rate": 8.704837497095247e-06,
+      "loss": 0.6243,
+      "step": 8593
+    },
+    {
+      "epoch": 0.4730034674445484,
+      "grad_norm": 0.9445781111717224,
+      "learning_rate": 8.704546392616231e-06,
+      "loss": 0.6676,
+      "step": 8594
+    },
+    {
+      "epoch": 0.47305850624690404,
+      "grad_norm": 0.6527873277664185,
+      "learning_rate": 8.704255260294837e-06,
+      "loss": 0.6979,
+      "step": 8595
+    },
+    {
+      "epoch": 0.47311354504925973,
+      "grad_norm": 0.6732963919639587,
+      "learning_rate": 8.703964100133252e-06,
+      "loss": 0.7724,
+      "step": 8596
+    },
+    {
+      "epoch": 0.47316858385161537,
+      "grad_norm": 0.7661726474761963,
+      "learning_rate": 8.703672912133665e-06,
+      "loss": 0.7988,
+      "step": 8597
+    },
+    {
+      "epoch": 0.47322362265397105,
+      "grad_norm": 0.7006877660751343,
+      "learning_rate": 8.703381696298262e-06,
+      "loss": 0.6765,
+      "step": 8598
+    },
+    {
+      "epoch": 0.4732786614563267,
+      "grad_norm": 0.7195086479187012,
+      "learning_rate": 8.703090452629236e-06,
+      "loss": 0.6676,
+      "step": 8599
+    },
+    {
+      "epoch": 0.4733337002586824,
+      "grad_norm": 0.6692042350769043,
+      "learning_rate": 8.702799181128771e-06,
+      "loss": 0.7882,
+      "step": 8600
+    },
+    {
+      "epoch": 0.473388739061038,
+      "grad_norm": 0.7736524343490601,
+      "learning_rate": 8.70250788179906e-06,
+      "loss": 0.7977,
+      "step": 8601
+    },
+    {
+      "epoch": 0.4734437778633937,
+      "grad_norm": 0.8821607828140259,
+      "learning_rate": 8.70221655464229e-06,
+      "loss": 0.7465,
+      "step": 8602
+    },
+    {
+      "epoch": 0.47349881666574933,
+      "grad_norm": 0.7565156817436218,
+      "learning_rate": 8.701925199660652e-06,
+      "loss": 0.831,
+      "step": 8603
+    },
+    {
+      "epoch": 0.473553855468105,
+      "grad_norm": 0.8542304039001465,
+      "learning_rate": 8.701633816856335e-06,
+      "loss": 0.7538,
+      "step": 8604
+    },
+    {
+      "epoch": 0.47360889427046066,
+      "grad_norm": 0.6891050338745117,
+      "learning_rate": 8.701342406231529e-06,
+      "loss": 0.7687,
+      "step": 8605
+    },
+    {
+      "epoch": 0.47366393307281635,
+      "grad_norm": 0.8570719361305237,
+      "learning_rate": 8.701050967788424e-06,
+      "loss": 0.7236,
+      "step": 8606
+    },
+    {
+      "epoch": 0.473718971875172,
+      "grad_norm": 0.7921456098556519,
+      "learning_rate": 8.700759501529212e-06,
+      "loss": 0.8214,
+      "step": 8607
+    },
+    {
+      "epoch": 0.47377401067752767,
+      "grad_norm": 0.7584527730941772,
+      "learning_rate": 8.70046800745608e-06,
+      "loss": 0.8204,
+      "step": 8608
+    },
+    {
+      "epoch": 0.4738290494798833,
+      "grad_norm": 0.8033978343009949,
+      "learning_rate": 8.700176485571222e-06,
+      "loss": 0.8278,
+      "step": 8609
+    },
+    {
+      "epoch": 0.473884088282239,
+      "grad_norm": 0.9950750470161438,
+      "learning_rate": 8.699884935876828e-06,
+      "loss": 0.8181,
+      "step": 8610
+    },
+    {
+      "epoch": 0.4739391270845946,
+      "grad_norm": 0.7213684916496277,
+      "learning_rate": 8.69959335837509e-06,
+      "loss": 0.7099,
+      "step": 8611
+    },
+    {
+      "epoch": 0.4739941658869503,
+      "grad_norm": 0.7847200632095337,
+      "learning_rate": 8.699301753068199e-06,
+      "loss": 0.8272,
+      "step": 8612
+    },
+    {
+      "epoch": 0.47404920468930595,
+      "grad_norm": 0.7075058221817017,
+      "learning_rate": 8.699010119958344e-06,
+      "loss": 0.7127,
+      "step": 8613
+    },
+    {
+      "epoch": 0.47410424349166164,
+      "grad_norm": 0.682741641998291,
+      "learning_rate": 8.69871845904772e-06,
+      "loss": 0.8446,
+      "step": 8614
+    },
+    {
+      "epoch": 0.47415928229401727,
+      "grad_norm": 0.7120605111122131,
+      "learning_rate": 8.69842677033852e-06,
+      "loss": 0.7776,
+      "step": 8615
+    },
+    {
+      "epoch": 0.47421432109637296,
+      "grad_norm": 0.822405219078064,
+      "learning_rate": 8.698135053832933e-06,
+      "loss": 0.8018,
+      "step": 8616
+    },
+    {
+      "epoch": 0.4742693598987286,
+      "grad_norm": 0.6815186738967896,
+      "learning_rate": 8.697843309533152e-06,
+      "loss": 0.7413,
+      "step": 8617
+    },
+    {
+      "epoch": 0.4743243987010843,
+      "grad_norm": 0.7587849497795105,
+      "learning_rate": 8.69755153744137e-06,
+      "loss": 0.7809,
+      "step": 8618
+    },
+    {
+      "epoch": 0.4743794375034399,
+      "grad_norm": 0.7092488408088684,
+      "learning_rate": 8.697259737559782e-06,
+      "loss": 0.7921,
+      "step": 8619
+    },
+    {
+      "epoch": 0.4744344763057956,
+      "grad_norm": 0.7396836280822754,
+      "learning_rate": 8.69696790989058e-06,
+      "loss": 0.7946,
+      "step": 8620
+    },
+    {
+      "epoch": 0.47448951510815124,
+      "grad_norm": 0.6760729551315308,
+      "learning_rate": 8.696676054435955e-06,
+      "loss": 0.7389,
+      "step": 8621
+    },
+    {
+      "epoch": 0.4745445539105069,
+      "grad_norm": 1.1640692949295044,
+      "learning_rate": 8.696384171198105e-06,
+      "loss": 0.8291,
+      "step": 8622
+    },
+    {
+      "epoch": 0.47459959271286256,
+      "grad_norm": 0.7415158152580261,
+      "learning_rate": 8.696092260179219e-06,
+      "loss": 0.7534,
+      "step": 8623
+    },
+    {
+      "epoch": 0.47465463151521825,
+      "grad_norm": 0.7730052471160889,
+      "learning_rate": 8.695800321381492e-06,
+      "loss": 0.8447,
+      "step": 8624
+    },
+    {
+      "epoch": 0.4747096703175739,
+      "grad_norm": 0.811522364616394,
+      "learning_rate": 8.695508354807121e-06,
+      "loss": 0.7466,
+      "step": 8625
+    },
+    {
+      "epoch": 0.4747647091199296,
+      "grad_norm": 0.7908332347869873,
+      "learning_rate": 8.695216360458298e-06,
+      "loss": 0.7769,
+      "step": 8626
+    },
+    {
+      "epoch": 0.4748197479222852,
+      "grad_norm": 0.744971752166748,
+      "learning_rate": 8.694924338337217e-06,
+      "loss": 0.7651,
+      "step": 8627
+    },
+    {
+      "epoch": 0.4748747867246409,
+      "grad_norm": 0.705565869808197,
+      "learning_rate": 8.694632288446075e-06,
+      "loss": 0.8258,
+      "step": 8628
+    },
+    {
+      "epoch": 0.47492982552699653,
+      "grad_norm": 0.8199328780174255,
+      "learning_rate": 8.694340210787065e-06,
+      "loss": 0.733,
+      "step": 8629
+    },
+    {
+      "epoch": 0.4749848643293522,
+      "grad_norm": 0.6965511441230774,
+      "learning_rate": 8.694048105362382e-06,
+      "loss": 0.7548,
+      "step": 8630
+    },
+    {
+      "epoch": 0.47503990313170785,
+      "grad_norm": 0.7943055629730225,
+      "learning_rate": 8.693755972174225e-06,
+      "loss": 0.7518,
+      "step": 8631
+    },
+    {
+      "epoch": 0.47509494193406354,
+      "grad_norm": 0.6277437806129456,
+      "learning_rate": 8.693463811224785e-06,
+      "loss": 0.6941,
+      "step": 8632
+    },
+    {
+      "epoch": 0.4751499807364192,
+      "grad_norm": 1.0745574235916138,
+      "learning_rate": 8.693171622516259e-06,
+      "loss": 0.8056,
+      "step": 8633
+    },
+    {
+      "epoch": 0.4752050195387748,
+      "grad_norm": 0.7005153894424438,
+      "learning_rate": 8.692879406050844e-06,
+      "loss": 0.757,
+      "step": 8634
+    },
+    {
+      "epoch": 0.4752600583411305,
+      "grad_norm": 0.6971127986907959,
+      "learning_rate": 8.692587161830737e-06,
+      "loss": 0.7509,
+      "step": 8635
+    },
+    {
+      "epoch": 0.47531509714348613,
+      "grad_norm": 0.7583497762680054,
+      "learning_rate": 8.692294889858133e-06,
+      "loss": 0.7895,
+      "step": 8636
+    },
+    {
+      "epoch": 0.4753701359458418,
+      "grad_norm": 0.719932496547699,
+      "learning_rate": 8.692002590135228e-06,
+      "loss": 0.762,
+      "step": 8637
+    },
+    {
+      "epoch": 0.47542517474819745,
+      "grad_norm": 0.7041804790496826,
+      "learning_rate": 8.691710262664222e-06,
+      "loss": 0.7101,
+      "step": 8638
+    },
+    {
+      "epoch": 0.47548021355055314,
+      "grad_norm": 0.7395016551017761,
+      "learning_rate": 8.691417907447309e-06,
+      "loss": 0.723,
+      "step": 8639
+    },
+    {
+      "epoch": 0.4755352523529088,
+      "grad_norm": 0.6605637073516846,
+      "learning_rate": 8.691125524486686e-06,
+      "loss": 0.644,
+      "step": 8640
+    },
+    {
+      "epoch": 0.47559029115526447,
+      "grad_norm": 0.694732129573822,
+      "learning_rate": 8.690833113784552e-06,
+      "loss": 0.7162,
+      "step": 8641
+    },
+    {
+      "epoch": 0.4756453299576201,
+      "grad_norm": 0.7622451186180115,
+      "learning_rate": 8.690540675343105e-06,
+      "loss": 0.6995,
+      "step": 8642
+    },
+    {
+      "epoch": 0.4757003687599758,
+      "grad_norm": 0.6961628794670105,
+      "learning_rate": 8.69024820916454e-06,
+      "loss": 0.7955,
+      "step": 8643
+    },
+    {
+      "epoch": 0.4757554075623314,
+      "grad_norm": 0.706266462802887,
+      "learning_rate": 8.68995571525106e-06,
+      "loss": 0.7237,
+      "step": 8644
+    },
+    {
+      "epoch": 0.4758104463646871,
+      "grad_norm": 0.7727495431900024,
+      "learning_rate": 8.689663193604858e-06,
+      "loss": 0.7215,
+      "step": 8645
+    },
+    {
+      "epoch": 0.47586548516704275,
+      "grad_norm": 0.7320648431777954,
+      "learning_rate": 8.689370644228136e-06,
+      "loss": 0.7592,
+      "step": 8646
+    },
+    {
+      "epoch": 0.47592052396939843,
+      "grad_norm": 0.8149487376213074,
+      "learning_rate": 8.689078067123093e-06,
+      "loss": 0.7666,
+      "step": 8647
+    },
+    {
+      "epoch": 0.47597556277175407,
+      "grad_norm": 0.6584552526473999,
+      "learning_rate": 8.688785462291927e-06,
+      "loss": 0.7497,
+      "step": 8648
+    },
+    {
+      "epoch": 0.47603060157410976,
+      "grad_norm": 0.7197825312614441,
+      "learning_rate": 8.688492829736836e-06,
+      "loss": 0.7559,
+      "step": 8649
+    },
+    {
+      "epoch": 0.4760856403764654,
+      "grad_norm": 0.8116913437843323,
+      "learning_rate": 8.68820016946002e-06,
+      "loss": 0.7029,
+      "step": 8650
+    },
+    {
+      "epoch": 0.4761406791788211,
+      "grad_norm": 0.6733378171920776,
+      "learning_rate": 8.68790748146368e-06,
+      "loss": 0.7242,
+      "step": 8651
+    },
+    {
+      "epoch": 0.4761957179811767,
+      "grad_norm": 0.690464437007904,
+      "learning_rate": 8.687614765750012e-06,
+      "loss": 0.6668,
+      "step": 8652
+    },
+    {
+      "epoch": 0.4762507567835324,
+      "grad_norm": 0.7901185154914856,
+      "learning_rate": 8.687322022321221e-06,
+      "loss": 0.7436,
+      "step": 8653
+    },
+    {
+      "epoch": 0.47630579558588804,
+      "grad_norm": 0.7608267068862915,
+      "learning_rate": 8.687029251179504e-06,
+      "loss": 0.8292,
+      "step": 8654
+    },
+    {
+      "epoch": 0.4763608343882437,
+      "grad_norm": 0.6851119995117188,
+      "learning_rate": 8.686736452327062e-06,
+      "loss": 0.7974,
+      "step": 8655
+    },
+    {
+      "epoch": 0.47641587319059936,
+      "grad_norm": 0.6946395635604858,
+      "learning_rate": 8.686443625766094e-06,
+      "loss": 0.6745,
+      "step": 8656
+    },
+    {
+      "epoch": 0.47647091199295505,
+      "grad_norm": 0.7403521537780762,
+      "learning_rate": 8.686150771498804e-06,
+      "loss": 0.7759,
+      "step": 8657
+    },
+    {
+      "epoch": 0.4765259507953107,
+      "grad_norm": 0.8415689468383789,
+      "learning_rate": 8.685857889527393e-06,
+      "loss": 0.7911,
+      "step": 8658
+    },
+    {
+      "epoch": 0.47658098959766637,
+      "grad_norm": 0.6947778463363647,
+      "learning_rate": 8.68556497985406e-06,
+      "loss": 0.8026,
+      "step": 8659
+    },
+    {
+      "epoch": 0.476636028400022,
+      "grad_norm": 0.6807059645652771,
+      "learning_rate": 8.685272042481006e-06,
+      "loss": 0.7194,
+      "step": 8660
+    },
+    {
+      "epoch": 0.4766910672023777,
+      "grad_norm": 0.8948639631271362,
+      "learning_rate": 8.684979077410434e-06,
+      "loss": 0.8017,
+      "step": 8661
+    },
+    {
+      "epoch": 0.4767461060047333,
+      "grad_norm": 0.6697849035263062,
+      "learning_rate": 8.684686084644546e-06,
+      "loss": 0.7653,
+      "step": 8662
+    },
+    {
+      "epoch": 0.476801144807089,
+      "grad_norm": 0.7303311228752136,
+      "learning_rate": 8.684393064185543e-06,
+      "loss": 0.8287,
+      "step": 8663
+    },
+    {
+      "epoch": 0.47685618360944465,
+      "grad_norm": 0.6545100808143616,
+      "learning_rate": 8.68410001603563e-06,
+      "loss": 0.7438,
+      "step": 8664
+    },
+    {
+      "epoch": 0.47691122241180034,
+      "grad_norm": 0.8757766485214233,
+      "learning_rate": 8.683806940197006e-06,
+      "loss": 0.8343,
+      "step": 8665
+    },
+    {
+      "epoch": 0.476966261214156,
+      "grad_norm": 0.6414330005645752,
+      "learning_rate": 8.683513836671876e-06,
+      "loss": 0.7201,
+      "step": 8666
+    },
+    {
+      "epoch": 0.47702130001651166,
+      "grad_norm": 0.6736441850662231,
+      "learning_rate": 8.68322070546244e-06,
+      "loss": 0.7365,
+      "step": 8667
+    },
+    {
+      "epoch": 0.4770763388188673,
+      "grad_norm": 0.780491054058075,
+      "learning_rate": 8.682927546570905e-06,
+      "loss": 0.924,
+      "step": 8668
+    },
+    {
+      "epoch": 0.477131377621223,
+      "grad_norm": 0.6913807988166809,
+      "learning_rate": 8.68263435999947e-06,
+      "loss": 0.8269,
+      "step": 8669
+    },
+    {
+      "epoch": 0.4771864164235786,
+      "grad_norm": 0.7264360189437866,
+      "learning_rate": 8.682341145750344e-06,
+      "loss": 0.788,
+      "step": 8670
+    },
+    {
+      "epoch": 0.4772414552259343,
+      "grad_norm": 0.7777243852615356,
+      "learning_rate": 8.682047903825725e-06,
+      "loss": 0.8691,
+      "step": 8671
+    },
+    {
+      "epoch": 0.47729649402828994,
+      "grad_norm": 0.7590457797050476,
+      "learning_rate": 8.681754634227821e-06,
+      "loss": 0.8249,
+      "step": 8672
+    },
+    {
+      "epoch": 0.47735153283064563,
+      "grad_norm": 0.7672324776649475,
+      "learning_rate": 8.681461336958836e-06,
+      "loss": 0.8334,
+      "step": 8673
+    },
+    {
+      "epoch": 0.47740657163300126,
+      "grad_norm": 0.7181395888328552,
+      "learning_rate": 8.681168012020971e-06,
+      "loss": 0.8089,
+      "step": 8674
+    },
+    {
+      "epoch": 0.47746161043535695,
+      "grad_norm": 0.7671428918838501,
+      "learning_rate": 8.680874659416433e-06,
+      "loss": 0.7634,
+      "step": 8675
+    },
+    {
+      "epoch": 0.4775166492377126,
+      "grad_norm": 0.73219895362854,
+      "learning_rate": 8.680581279147427e-06,
+      "loss": 0.7013,
+      "step": 8676
+    },
+    {
+      "epoch": 0.4775716880400682,
+      "grad_norm": 0.8050867319107056,
+      "learning_rate": 8.680287871216158e-06,
+      "loss": 0.7524,
+      "step": 8677
+    },
+    {
+      "epoch": 0.4776267268424239,
+      "grad_norm": 0.7154340744018555,
+      "learning_rate": 8.679994435624828e-06,
+      "loss": 0.802,
+      "step": 8678
+    },
+    {
+      "epoch": 0.47768176564477954,
+      "grad_norm": 0.7005884051322937,
+      "learning_rate": 8.679700972375647e-06,
+      "loss": 0.7633,
+      "step": 8679
+    },
+    {
+      "epoch": 0.47773680444713523,
+      "grad_norm": 0.8203871846199036,
+      "learning_rate": 8.679407481470818e-06,
+      "loss": 0.7782,
+      "step": 8680
+    },
+    {
+      "epoch": 0.47779184324949087,
+      "grad_norm": 0.6582844853401184,
+      "learning_rate": 8.679113962912547e-06,
+      "loss": 0.6799,
+      "step": 8681
+    },
+    {
+      "epoch": 0.47784688205184656,
+      "grad_norm": 0.7052889466285706,
+      "learning_rate": 8.67882041670304e-06,
+      "loss": 0.7814,
+      "step": 8682
+    },
+    {
+      "epoch": 0.4779019208542022,
+      "grad_norm": 0.7533165812492371,
+      "learning_rate": 8.678526842844504e-06,
+      "loss": 0.7983,
+      "step": 8683
+    },
+    {
+      "epoch": 0.4779569596565579,
+      "grad_norm": 0.7335212230682373,
+      "learning_rate": 8.678233241339144e-06,
+      "loss": 0.8023,
+      "step": 8684
+    },
+    {
+      "epoch": 0.4780119984589135,
+      "grad_norm": 0.7824274897575378,
+      "learning_rate": 8.67793961218917e-06,
+      "loss": 0.8219,
+      "step": 8685
+    },
+    {
+      "epoch": 0.4780670372612692,
+      "grad_norm": 0.6547996401786804,
+      "learning_rate": 8.677645955396784e-06,
+      "loss": 0.715,
+      "step": 8686
+    },
+    {
+      "epoch": 0.47812207606362483,
+      "grad_norm": 0.7507368326187134,
+      "learning_rate": 8.677352270964196e-06,
+      "loss": 0.9379,
+      "step": 8687
+    },
+    {
+      "epoch": 0.4781771148659805,
+      "grad_norm": 0.6403020620346069,
+      "learning_rate": 8.677058558893613e-06,
+      "loss": 0.659,
+      "step": 8688
+    },
+    {
+      "epoch": 0.47823215366833616,
+      "grad_norm": 0.7075803279876709,
+      "learning_rate": 8.676764819187242e-06,
+      "loss": 0.7515,
+      "step": 8689
+    },
+    {
+      "epoch": 0.47828719247069185,
+      "grad_norm": 0.6899601817131042,
+      "learning_rate": 8.676471051847291e-06,
+      "loss": 0.8398,
+      "step": 8690
+    },
+    {
+      "epoch": 0.4783422312730475,
+      "grad_norm": 0.7145645618438721,
+      "learning_rate": 8.676177256875969e-06,
+      "loss": 0.7711,
+      "step": 8691
+    },
+    {
+      "epoch": 0.47839727007540317,
+      "grad_norm": 0.7139655351638794,
+      "learning_rate": 8.675883434275479e-06,
+      "loss": 0.8664,
+      "step": 8692
+    },
+    {
+      "epoch": 0.4784523088777588,
+      "grad_norm": 0.7100433111190796,
+      "learning_rate": 8.675589584048037e-06,
+      "loss": 0.7812,
+      "step": 8693
+    },
+    {
+      "epoch": 0.4785073476801145,
+      "grad_norm": 0.6103882789611816,
+      "learning_rate": 8.675295706195845e-06,
+      "loss": 0.6565,
+      "step": 8694
+    },
+    {
+      "epoch": 0.4785623864824701,
+      "grad_norm": 0.7236714959144592,
+      "learning_rate": 8.675001800721114e-06,
+      "loss": 0.6849,
+      "step": 8695
+    },
+    {
+      "epoch": 0.4786174252848258,
+      "grad_norm": 0.7567160129547119,
+      "learning_rate": 8.674707867626056e-06,
+      "loss": 0.8289,
+      "step": 8696
+    },
+    {
+      "epoch": 0.47867246408718145,
+      "grad_norm": 0.7004136443138123,
+      "learning_rate": 8.674413906912876e-06,
+      "loss": 0.7466,
+      "step": 8697
+    },
+    {
+      "epoch": 0.47872750288953714,
+      "grad_norm": 0.713835597038269,
+      "learning_rate": 8.674119918583783e-06,
+      "loss": 0.7875,
+      "step": 8698
+    },
+    {
+      "epoch": 0.47878254169189277,
+      "grad_norm": 0.8476874232292175,
+      "learning_rate": 8.67382590264099e-06,
+      "loss": 0.8028,
+      "step": 8699
+    },
+    {
+      "epoch": 0.47883758049424846,
+      "grad_norm": 0.720273494720459,
+      "learning_rate": 8.673531859086706e-06,
+      "loss": 0.7829,
+      "step": 8700
+    },
+    {
+      "epoch": 0.4788926192966041,
+      "grad_norm": 0.8042417168617249,
+      "learning_rate": 8.673237787923137e-06,
+      "loss": 0.7914,
+      "step": 8701
+    },
+    {
+      "epoch": 0.4789476580989598,
+      "grad_norm": 0.7779260277748108,
+      "learning_rate": 8.672943689152498e-06,
+      "loss": 0.6921,
+      "step": 8702
+    },
+    {
+      "epoch": 0.4790026969013154,
+      "grad_norm": 0.7957637906074524,
+      "learning_rate": 8.672649562776997e-06,
+      "loss": 0.8761,
+      "step": 8703
+    },
+    {
+      "epoch": 0.4790577357036711,
+      "grad_norm": 0.7467649579048157,
+      "learning_rate": 8.672355408798845e-06,
+      "loss": 0.7984,
+      "step": 8704
+    },
+    {
+      "epoch": 0.47911277450602674,
+      "grad_norm": 0.6746538877487183,
+      "learning_rate": 8.672061227220252e-06,
+      "loss": 0.7392,
+      "step": 8705
+    },
+    {
+      "epoch": 0.47916781330838243,
+      "grad_norm": 0.7331795692443848,
+      "learning_rate": 8.671767018043432e-06,
+      "loss": 0.7171,
+      "step": 8706
+    },
+    {
+      "epoch": 0.47922285211073806,
+      "grad_norm": 0.7879608273506165,
+      "learning_rate": 8.671472781270592e-06,
+      "loss": 0.8497,
+      "step": 8707
+    },
+    {
+      "epoch": 0.47927789091309375,
+      "grad_norm": 0.8659428358078003,
+      "learning_rate": 8.671178516903946e-06,
+      "loss": 0.8102,
+      "step": 8708
+    },
+    {
+      "epoch": 0.4793329297154494,
+      "grad_norm": 0.6489408612251282,
+      "learning_rate": 8.670884224945704e-06,
+      "loss": 0.6752,
+      "step": 8709
+    },
+    {
+      "epoch": 0.4793879685178051,
+      "grad_norm": 0.8182825446128845,
+      "learning_rate": 8.670589905398079e-06,
+      "loss": 0.7972,
+      "step": 8710
+    },
+    {
+      "epoch": 0.4794430073201607,
+      "grad_norm": 0.7759343981742859,
+      "learning_rate": 8.670295558263285e-06,
+      "loss": 0.7856,
+      "step": 8711
+    },
+    {
+      "epoch": 0.4794980461225164,
+      "grad_norm": 0.7421835064888,
+      "learning_rate": 8.670001183543528e-06,
+      "loss": 0.8165,
+      "step": 8712
+    },
+    {
+      "epoch": 0.47955308492487203,
+      "grad_norm": 0.6498512625694275,
+      "learning_rate": 8.669706781241028e-06,
+      "loss": 0.7212,
+      "step": 8713
+    },
+    {
+      "epoch": 0.4796081237272277,
+      "grad_norm": 0.8493219614028931,
+      "learning_rate": 8.669412351357993e-06,
+      "loss": 0.8036,
+      "step": 8714
+    },
+    {
+      "epoch": 0.47966316252958335,
+      "grad_norm": 0.6834331750869751,
+      "learning_rate": 8.669117893896637e-06,
+      "loss": 0.8127,
+      "step": 8715
+    },
+    {
+      "epoch": 0.47971820133193904,
+      "grad_norm": 0.7793670296669006,
+      "learning_rate": 8.668823408859172e-06,
+      "loss": 0.7276,
+      "step": 8716
+    },
+    {
+      "epoch": 0.4797732401342947,
+      "grad_norm": 0.7108075022697449,
+      "learning_rate": 8.668528896247815e-06,
+      "loss": 0.8328,
+      "step": 8717
+    },
+    {
+      "epoch": 0.47982827893665037,
+      "grad_norm": 0.6662433743476868,
+      "learning_rate": 8.668234356064774e-06,
+      "loss": 0.6751,
+      "step": 8718
+    },
+    {
+      "epoch": 0.479883317739006,
+      "grad_norm": 0.6595591902732849,
+      "learning_rate": 8.667939788312267e-06,
+      "loss": 0.707,
+      "step": 8719
+    },
+    {
+      "epoch": 0.47993835654136163,
+      "grad_norm": 0.7435836791992188,
+      "learning_rate": 8.667645192992506e-06,
+      "loss": 0.7885,
+      "step": 8720
+    },
+    {
+      "epoch": 0.4799933953437173,
+      "grad_norm": 0.6999356746673584,
+      "learning_rate": 8.667350570107706e-06,
+      "loss": 0.7538,
+      "step": 8721
+    },
+    {
+      "epoch": 0.48004843414607296,
+      "grad_norm": 0.7111191749572754,
+      "learning_rate": 8.66705591966008e-06,
+      "loss": 0.6814,
+      "step": 8722
+    },
+    {
+      "epoch": 0.48010347294842864,
+      "grad_norm": 0.6752734780311584,
+      "learning_rate": 8.666761241651844e-06,
+      "loss": 0.7221,
+      "step": 8723
+    },
+    {
+      "epoch": 0.4801585117507843,
+      "grad_norm": 0.7432951331138611,
+      "learning_rate": 8.666466536085212e-06,
+      "loss": 0.7689,
+      "step": 8724
+    },
+    {
+      "epoch": 0.48021355055313997,
+      "grad_norm": 0.7384392023086548,
+      "learning_rate": 8.666171802962398e-06,
+      "loss": 0.7862,
+      "step": 8725
+    },
+    {
+      "epoch": 0.4802685893554956,
+      "grad_norm": 0.6878762245178223,
+      "learning_rate": 8.66587704228562e-06,
+      "loss": 0.7246,
+      "step": 8726
+    },
+    {
+      "epoch": 0.4803236281578513,
+      "grad_norm": 0.6640586853027344,
+      "learning_rate": 8.66558225405709e-06,
+      "loss": 0.7181,
+      "step": 8727
+    },
+    {
+      "epoch": 0.4803786669602069,
+      "grad_norm": 0.6808595061302185,
+      "learning_rate": 8.665287438279024e-06,
+      "loss": 0.7866,
+      "step": 8728
+    },
+    {
+      "epoch": 0.4804337057625626,
+      "grad_norm": 0.5966268181800842,
+      "learning_rate": 8.66499259495364e-06,
+      "loss": 0.6755,
+      "step": 8729
+    },
+    {
+      "epoch": 0.48048874456491825,
+      "grad_norm": 0.742016077041626,
+      "learning_rate": 8.664697724083152e-06,
+      "loss": 0.8682,
+      "step": 8730
+    },
+    {
+      "epoch": 0.48054378336727394,
+      "grad_norm": 0.6621154546737671,
+      "learning_rate": 8.66440282566978e-06,
+      "loss": 0.7525,
+      "step": 8731
+    },
+    {
+      "epoch": 0.48059882216962957,
+      "grad_norm": 0.7347434759140015,
+      "learning_rate": 8.664107899715733e-06,
+      "loss": 0.7919,
+      "step": 8732
+    },
+    {
+      "epoch": 0.48065386097198526,
+      "grad_norm": 0.7564681172370911,
+      "learning_rate": 8.663812946223234e-06,
+      "loss": 0.9172,
+      "step": 8733
+    },
+    {
+      "epoch": 0.4807088997743409,
+      "grad_norm": 0.7193084359169006,
+      "learning_rate": 8.663517965194497e-06,
+      "loss": 0.7931,
+      "step": 8734
+    },
+    {
+      "epoch": 0.4807639385766966,
+      "grad_norm": 0.6882064938545227,
+      "learning_rate": 8.66322295663174e-06,
+      "loss": 0.7678,
+      "step": 8735
+    },
+    {
+      "epoch": 0.4808189773790522,
+      "grad_norm": 0.7954713106155396,
+      "learning_rate": 8.662927920537179e-06,
+      "loss": 0.6357,
+      "step": 8736
+    },
+    {
+      "epoch": 0.4808740161814079,
+      "grad_norm": 0.7123041749000549,
+      "learning_rate": 8.662632856913034e-06,
+      "loss": 0.7234,
+      "step": 8737
+    },
+    {
+      "epoch": 0.48092905498376354,
+      "grad_norm": 0.745145320892334,
+      "learning_rate": 8.66233776576152e-06,
+      "loss": 0.7516,
+      "step": 8738
+    },
+    {
+      "epoch": 0.4809840937861192,
+      "grad_norm": 0.6904219388961792,
+      "learning_rate": 8.662042647084856e-06,
+      "loss": 0.7995,
+      "step": 8739
+    },
+    {
+      "epoch": 0.48103913258847486,
+      "grad_norm": 0.71831214427948,
+      "learning_rate": 8.661747500885258e-06,
+      "loss": 0.7965,
+      "step": 8740
+    },
+    {
+      "epoch": 0.48109417139083055,
+      "grad_norm": 0.8514378666877747,
+      "learning_rate": 8.661452327164948e-06,
+      "loss": 0.8023,
+      "step": 8741
+    },
+    {
+      "epoch": 0.4811492101931862,
+      "grad_norm": 0.7411143779754639,
+      "learning_rate": 8.66115712592614e-06,
+      "loss": 0.797,
+      "step": 8742
+    },
+    {
+      "epoch": 0.4812042489955419,
+      "grad_norm": 0.737178385257721,
+      "learning_rate": 8.660861897171057e-06,
+      "loss": 0.7286,
+      "step": 8743
+    },
+    {
+      "epoch": 0.4812592877978975,
+      "grad_norm": 0.6823513507843018,
+      "learning_rate": 8.660566640901918e-06,
+      "loss": 0.7482,
+      "step": 8744
+    },
+    {
+      "epoch": 0.4813143266002532,
+      "grad_norm": 0.7205879092216492,
+      "learning_rate": 8.660271357120937e-06,
+      "loss": 0.8294,
+      "step": 8745
+    },
+    {
+      "epoch": 0.48136936540260883,
+      "grad_norm": 0.6887338757514954,
+      "learning_rate": 8.659976045830337e-06,
+      "loss": 0.7711,
+      "step": 8746
+    },
+    {
+      "epoch": 0.4814244042049645,
+      "grad_norm": 0.7498533129692078,
+      "learning_rate": 8.659680707032336e-06,
+      "loss": 0.7296,
+      "step": 8747
+    },
+    {
+      "epoch": 0.48147944300732015,
+      "grad_norm": 0.8041636943817139,
+      "learning_rate": 8.659385340729155e-06,
+      "loss": 0.9213,
+      "step": 8748
+    },
+    {
+      "epoch": 0.48153448180967584,
+      "grad_norm": 0.8623721599578857,
+      "learning_rate": 8.659089946923014e-06,
+      "loss": 0.8024,
+      "step": 8749
+    },
+    {
+      "epoch": 0.4815895206120315,
+      "grad_norm": 0.7212050557136536,
+      "learning_rate": 8.658794525616132e-06,
+      "loss": 0.732,
+      "step": 8750
+    },
+    {
+      "epoch": 0.48164455941438716,
+      "grad_norm": 0.7141492366790771,
+      "learning_rate": 8.658499076810729e-06,
+      "loss": 0.8062,
+      "step": 8751
+    },
+    {
+      "epoch": 0.4816995982167428,
+      "grad_norm": 0.7191516160964966,
+      "learning_rate": 8.658203600509027e-06,
+      "loss": 0.805,
+      "step": 8752
+    },
+    {
+      "epoch": 0.4817546370190985,
+      "grad_norm": 0.71059650182724,
+      "learning_rate": 8.657908096713245e-06,
+      "loss": 0.6755,
+      "step": 8753
+    },
+    {
+      "epoch": 0.4818096758214541,
+      "grad_norm": 0.6715459823608398,
+      "learning_rate": 8.657612565425607e-06,
+      "loss": 0.8093,
+      "step": 8754
+    },
+    {
+      "epoch": 0.4818647146238098,
+      "grad_norm": 0.7438814640045166,
+      "learning_rate": 8.65731700664833e-06,
+      "loss": 0.8059,
+      "step": 8755
+    },
+    {
+      "epoch": 0.48191975342616544,
+      "grad_norm": 0.7295387387275696,
+      "learning_rate": 8.657021420383637e-06,
+      "loss": 0.8437,
+      "step": 8756
+    },
+    {
+      "epoch": 0.48197479222852113,
+      "grad_norm": 0.7053797245025635,
+      "learning_rate": 8.656725806633753e-06,
+      "loss": 0.8424,
+      "step": 8757
+    },
+    {
+      "epoch": 0.48202983103087677,
+      "grad_norm": 0.6902007460594177,
+      "learning_rate": 8.656430165400894e-06,
+      "loss": 0.6967,
+      "step": 8758
+    },
+    {
+      "epoch": 0.48208486983323245,
+      "grad_norm": 0.66749507188797,
+      "learning_rate": 8.656134496687286e-06,
+      "loss": 0.7858,
+      "step": 8759
+    },
+    {
+      "epoch": 0.4821399086355881,
+      "grad_norm": 0.6755428314208984,
+      "learning_rate": 8.65583880049515e-06,
+      "loss": 0.6669,
+      "step": 8760
+    },
+    {
+      "epoch": 0.4821949474379438,
+      "grad_norm": 0.921096920967102,
+      "learning_rate": 8.655543076826706e-06,
+      "loss": 0.8545,
+      "step": 8761
+    },
+    {
+      "epoch": 0.4822499862402994,
+      "grad_norm": 0.7931553721427917,
+      "learning_rate": 8.65524732568418e-06,
+      "loss": 0.8708,
+      "step": 8762
+    },
+    {
+      "epoch": 0.48230502504265504,
+      "grad_norm": 0.7891780734062195,
+      "learning_rate": 8.654951547069794e-06,
+      "loss": 0.687,
+      "step": 8763
+    },
+    {
+      "epoch": 0.48236006384501073,
+      "grad_norm": 0.747662365436554,
+      "learning_rate": 8.65465574098577e-06,
+      "loss": 0.8153,
+      "step": 8764
+    },
+    {
+      "epoch": 0.48241510264736637,
+      "grad_norm": 0.7758497595787048,
+      "learning_rate": 8.65435990743433e-06,
+      "loss": 0.8018,
+      "step": 8765
+    },
+    {
+      "epoch": 0.48247014144972206,
+      "grad_norm": 0.6997805237770081,
+      "learning_rate": 8.654064046417703e-06,
+      "loss": 0.7845,
+      "step": 8766
+    },
+    {
+      "epoch": 0.4825251802520777,
+      "grad_norm": 0.7188366651535034,
+      "learning_rate": 8.653768157938106e-06,
+      "loss": 0.7528,
+      "step": 8767
+    },
+    {
+      "epoch": 0.4825802190544334,
+      "grad_norm": 0.6848055124282837,
+      "learning_rate": 8.653472241997767e-06,
+      "loss": 0.7658,
+      "step": 8768
+    },
+    {
+      "epoch": 0.482635257856789,
+      "grad_norm": 1.0603824853897095,
+      "learning_rate": 8.653176298598907e-06,
+      "loss": 0.7692,
+      "step": 8769
+    },
+    {
+      "epoch": 0.4826902966591447,
+      "grad_norm": 0.8191514611244202,
+      "learning_rate": 8.652880327743753e-06,
+      "loss": 0.7706,
+      "step": 8770
+    },
+    {
+      "epoch": 0.48274533546150034,
+      "grad_norm": 0.6318503618240356,
+      "learning_rate": 8.652584329434527e-06,
+      "loss": 0.6635,
+      "step": 8771
+    },
+    {
+      "epoch": 0.482800374263856,
+      "grad_norm": 0.6860769391059875,
+      "learning_rate": 8.652288303673457e-06,
+      "loss": 0.739,
+      "step": 8772
+    },
+    {
+      "epoch": 0.48285541306621166,
+      "grad_norm": 0.7414761185646057,
+      "learning_rate": 8.651992250462765e-06,
+      "loss": 0.7949,
+      "step": 8773
+    },
+    {
+      "epoch": 0.48291045186856735,
+      "grad_norm": 0.7255183458328247,
+      "learning_rate": 8.651696169804676e-06,
+      "loss": 0.8569,
+      "step": 8774
+    },
+    {
+      "epoch": 0.482965490670923,
+      "grad_norm": 0.7034135460853577,
+      "learning_rate": 8.651400061701417e-06,
+      "loss": 0.7562,
+      "step": 8775
+    },
+    {
+      "epoch": 0.48302052947327867,
+      "grad_norm": 0.7041038274765015,
+      "learning_rate": 8.651103926155212e-06,
+      "loss": 0.7194,
+      "step": 8776
+    },
+    {
+      "epoch": 0.4830755682756343,
+      "grad_norm": 1.0965619087219238,
+      "learning_rate": 8.650807763168287e-06,
+      "loss": 0.9033,
+      "step": 8777
+    },
+    {
+      "epoch": 0.48313060707799,
+      "grad_norm": 0.7400044798851013,
+      "learning_rate": 8.650511572742869e-06,
+      "loss": 0.7626,
+      "step": 8778
+    },
+    {
+      "epoch": 0.4831856458803456,
+      "grad_norm": 0.6957885026931763,
+      "learning_rate": 8.650215354881182e-06,
+      "loss": 0.7283,
+      "step": 8779
+    },
+    {
+      "epoch": 0.4832406846827013,
+      "grad_norm": 0.7992473840713501,
+      "learning_rate": 8.649919109585454e-06,
+      "loss": 0.8376,
+      "step": 8780
+    },
+    {
+      "epoch": 0.48329572348505695,
+      "grad_norm": 0.8556981086730957,
+      "learning_rate": 8.649622836857911e-06,
+      "loss": 0.7737,
+      "step": 8781
+    },
+    {
+      "epoch": 0.48335076228741264,
+      "grad_norm": 0.8476192355155945,
+      "learning_rate": 8.64932653670078e-06,
+      "loss": 0.8926,
+      "step": 8782
+    },
+    {
+      "epoch": 0.48340580108976827,
+      "grad_norm": 0.6461093425750732,
+      "learning_rate": 8.649030209116289e-06,
+      "loss": 0.7452,
+      "step": 8783
+    },
+    {
+      "epoch": 0.48346083989212396,
+      "grad_norm": 0.6997528076171875,
+      "learning_rate": 8.648733854106661e-06,
+      "loss": 0.7962,
+      "step": 8784
+    },
+    {
+      "epoch": 0.4835158786944796,
+      "grad_norm": 0.7606356739997864,
+      "learning_rate": 8.648437471674128e-06,
+      "loss": 0.6517,
+      "step": 8785
+    },
+    {
+      "epoch": 0.4835709174968353,
+      "grad_norm": 0.8118630051612854,
+      "learning_rate": 8.648141061820913e-06,
+      "loss": 0.7539,
+      "step": 8786
+    },
+    {
+      "epoch": 0.4836259562991909,
+      "grad_norm": 0.8778805136680603,
+      "learning_rate": 8.64784462454925e-06,
+      "loss": 0.763,
+      "step": 8787
+    },
+    {
+      "epoch": 0.4836809951015466,
+      "grad_norm": 0.7741022706031799,
+      "learning_rate": 8.647548159861361e-06,
+      "loss": 0.7749,
+      "step": 8788
+    },
+    {
+      "epoch": 0.48373603390390224,
+      "grad_norm": 0.76578688621521,
+      "learning_rate": 8.647251667759478e-06,
+      "loss": 0.6968,
+      "step": 8789
+    },
+    {
+      "epoch": 0.48379107270625793,
+      "grad_norm": 0.8477250933647156,
+      "learning_rate": 8.646955148245827e-06,
+      "loss": 0.8364,
+      "step": 8790
+    },
+    {
+      "epoch": 0.48384611150861356,
+      "grad_norm": 0.9105041027069092,
+      "learning_rate": 8.646658601322635e-06,
+      "loss": 0.823,
+      "step": 8791
+    },
+    {
+      "epoch": 0.48390115031096925,
+      "grad_norm": 0.7642726898193359,
+      "learning_rate": 8.646362026992135e-06,
+      "loss": 0.721,
+      "step": 8792
+    },
+    {
+      "epoch": 0.4839561891133249,
+      "grad_norm": 0.7567259669303894,
+      "learning_rate": 8.646065425256555e-06,
+      "loss": 0.7876,
+      "step": 8793
+    },
+    {
+      "epoch": 0.4840112279156806,
+      "grad_norm": 0.7691231966018677,
+      "learning_rate": 8.64576879611812e-06,
+      "loss": 0.8308,
+      "step": 8794
+    },
+    {
+      "epoch": 0.4840662667180362,
+      "grad_norm": 1.0769426822662354,
+      "learning_rate": 8.645472139579067e-06,
+      "loss": 0.892,
+      "step": 8795
+    },
+    {
+      "epoch": 0.4841213055203919,
+      "grad_norm": 0.6987955570220947,
+      "learning_rate": 8.64517545564162e-06,
+      "loss": 0.8254,
+      "step": 8796
+    },
+    {
+      "epoch": 0.48417634432274753,
+      "grad_norm": 0.7736005783081055,
+      "learning_rate": 8.644878744308007e-06,
+      "loss": 0.7666,
+      "step": 8797
+    },
+    {
+      "epoch": 0.4842313831251032,
+      "grad_norm": 0.6233380436897278,
+      "learning_rate": 8.644582005580464e-06,
+      "loss": 0.6443,
+      "step": 8798
+    },
+    {
+      "epoch": 0.48428642192745885,
+      "grad_norm": 0.7343530654907227,
+      "learning_rate": 8.644285239461217e-06,
+      "loss": 0.724,
+      "step": 8799
+    },
+    {
+      "epoch": 0.48434146072981454,
+      "grad_norm": 0.725321352481842,
+      "learning_rate": 8.643988445952499e-06,
+      "loss": 0.7249,
+      "step": 8800
+    },
+    {
+      "epoch": 0.4843964995321702,
+      "grad_norm": 0.7256256341934204,
+      "learning_rate": 8.643691625056539e-06,
+      "loss": 0.8656,
+      "step": 8801
+    },
+    {
+      "epoch": 0.48445153833452587,
+      "grad_norm": 0.8559528589248657,
+      "learning_rate": 8.643394776775567e-06,
+      "loss": 0.9186,
+      "step": 8802
+    },
+    {
+      "epoch": 0.4845065771368815,
+      "grad_norm": 0.6735692024230957,
+      "learning_rate": 8.643097901111815e-06,
+      "loss": 0.7007,
+      "step": 8803
+    },
+    {
+      "epoch": 0.4845616159392372,
+      "grad_norm": 0.8373280167579651,
+      "learning_rate": 8.642800998067515e-06,
+      "loss": 0.8774,
+      "step": 8804
+    },
+    {
+      "epoch": 0.4846166547415928,
+      "grad_norm": 0.731311023235321,
+      "learning_rate": 8.642504067644898e-06,
+      "loss": 0.7102,
+      "step": 8805
+    },
+    {
+      "epoch": 0.48467169354394846,
+      "grad_norm": 0.7259742617607117,
+      "learning_rate": 8.642207109846195e-06,
+      "loss": 0.7174,
+      "step": 8806
+    },
+    {
+      "epoch": 0.48472673234630415,
+      "grad_norm": 0.6454386115074158,
+      "learning_rate": 8.641910124673638e-06,
+      "loss": 0.7656,
+      "step": 8807
+    },
+    {
+      "epoch": 0.4847817711486598,
+      "grad_norm": 0.7701624631881714,
+      "learning_rate": 8.641613112129462e-06,
+      "loss": 0.7926,
+      "step": 8808
+    },
+    {
+      "epoch": 0.48483680995101547,
+      "grad_norm": 0.6812854409217834,
+      "learning_rate": 8.641316072215893e-06,
+      "loss": 0.7072,
+      "step": 8809
+    },
+    {
+      "epoch": 0.4848918487533711,
+      "grad_norm": 0.8180119395256042,
+      "learning_rate": 8.641019004935169e-06,
+      "loss": 0.8621,
+      "step": 8810
+    },
+    {
+      "epoch": 0.4849468875557268,
+      "grad_norm": 0.6346331834793091,
+      "learning_rate": 8.64072191028952e-06,
+      "loss": 0.6907,
+      "step": 8811
+    },
+    {
+      "epoch": 0.4850019263580824,
+      "grad_norm": 0.6819741129875183,
+      "learning_rate": 8.64042478828118e-06,
+      "loss": 0.77,
+      "step": 8812
+    },
+    {
+      "epoch": 0.4850569651604381,
+      "grad_norm": 0.9074214100837708,
+      "learning_rate": 8.640127638912383e-06,
+      "loss": 0.7799,
+      "step": 8813
+    },
+    {
+      "epoch": 0.48511200396279375,
+      "grad_norm": 0.8065158724784851,
+      "learning_rate": 8.63983046218536e-06,
+      "loss": 0.8033,
+      "step": 8814
+    },
+    {
+      "epoch": 0.48516704276514944,
+      "grad_norm": 0.6241241097450256,
+      "learning_rate": 8.639533258102345e-06,
+      "loss": 0.6936,
+      "step": 8815
+    },
+    {
+      "epoch": 0.48522208156750507,
+      "grad_norm": 0.6928265690803528,
+      "learning_rate": 8.639236026665573e-06,
+      "loss": 0.7526,
+      "step": 8816
+    },
+    {
+      "epoch": 0.48527712036986076,
+      "grad_norm": 0.8171425461769104,
+      "learning_rate": 8.638938767877276e-06,
+      "loss": 0.8227,
+      "step": 8817
+    },
+    {
+      "epoch": 0.4853321591722164,
+      "grad_norm": 0.7007083296775818,
+      "learning_rate": 8.638641481739692e-06,
+      "loss": 0.7439,
+      "step": 8818
+    },
+    {
+      "epoch": 0.4853871979745721,
+      "grad_norm": 0.8905115127563477,
+      "learning_rate": 8.63834416825505e-06,
+      "loss": 0.6873,
+      "step": 8819
+    },
+    {
+      "epoch": 0.4854422367769277,
+      "grad_norm": 0.702198326587677,
+      "learning_rate": 8.638046827425588e-06,
+      "loss": 0.7999,
+      "step": 8820
+    },
+    {
+      "epoch": 0.4854972755792834,
+      "grad_norm": 0.7280104160308838,
+      "learning_rate": 8.63774945925354e-06,
+      "loss": 0.8562,
+      "step": 8821
+    },
+    {
+      "epoch": 0.48555231438163904,
+      "grad_norm": 0.9803630113601685,
+      "learning_rate": 8.63745206374114e-06,
+      "loss": 0.8347,
+      "step": 8822
+    },
+    {
+      "epoch": 0.4856073531839947,
+      "grad_norm": 0.6781168580055237,
+      "learning_rate": 8.637154640890625e-06,
+      "loss": 0.8124,
+      "step": 8823
+    },
+    {
+      "epoch": 0.48566239198635036,
+      "grad_norm": 0.7219669222831726,
+      "learning_rate": 8.63685719070423e-06,
+      "loss": 0.8053,
+      "step": 8824
+    },
+    {
+      "epoch": 0.48571743078870605,
+      "grad_norm": 0.7077241539955139,
+      "learning_rate": 8.636559713184187e-06,
+      "loss": 0.7534,
+      "step": 8825
+    },
+    {
+      "epoch": 0.4857724695910617,
+      "grad_norm": 0.70063316822052,
+      "learning_rate": 8.636262208332737e-06,
+      "loss": 0.7509,
+      "step": 8826
+    },
+    {
+      "epoch": 0.4858275083934174,
+      "grad_norm": 0.7292184233665466,
+      "learning_rate": 8.635964676152114e-06,
+      "loss": 0.7485,
+      "step": 8827
+    },
+    {
+      "epoch": 0.485882547195773,
+      "grad_norm": 0.7970258593559265,
+      "learning_rate": 8.635667116644552e-06,
+      "loss": 0.8874,
+      "step": 8828
+    },
+    {
+      "epoch": 0.4859375859981287,
+      "grad_norm": 0.7090024352073669,
+      "learning_rate": 8.63536952981229e-06,
+      "loss": 0.7665,
+      "step": 8829
+    },
+    {
+      "epoch": 0.48599262480048433,
+      "grad_norm": 0.761409342288971,
+      "learning_rate": 8.635071915657565e-06,
+      "loss": 0.7977,
+      "step": 8830
+    },
+    {
+      "epoch": 0.48604766360284,
+      "grad_norm": 0.724896252155304,
+      "learning_rate": 8.634774274182611e-06,
+      "loss": 0.8591,
+      "step": 8831
+    },
+    {
+      "epoch": 0.48610270240519565,
+      "grad_norm": 0.737424910068512,
+      "learning_rate": 8.634476605389666e-06,
+      "loss": 0.8256,
+      "step": 8832
+    },
+    {
+      "epoch": 0.48615774120755134,
+      "grad_norm": 0.8261227607727051,
+      "learning_rate": 8.63417890928097e-06,
+      "loss": 0.8089,
+      "step": 8833
+    },
+    {
+      "epoch": 0.486212780009907,
+      "grad_norm": 0.6744595766067505,
+      "learning_rate": 8.633881185858756e-06,
+      "loss": 0.7821,
+      "step": 8834
+    },
+    {
+      "epoch": 0.48626781881226266,
+      "grad_norm": 0.6717672944068909,
+      "learning_rate": 8.633583435125263e-06,
+      "loss": 0.7823,
+      "step": 8835
+    },
+    {
+      "epoch": 0.4863228576146183,
+      "grad_norm": 0.753616213798523,
+      "learning_rate": 8.633285657082732e-06,
+      "loss": 0.8044,
+      "step": 8836
+    },
+    {
+      "epoch": 0.486377896416974,
+      "grad_norm": 0.6910914182662964,
+      "learning_rate": 8.632987851733397e-06,
+      "loss": 0.8244,
+      "step": 8837
+    },
+    {
+      "epoch": 0.4864329352193296,
+      "grad_norm": 0.9127064347267151,
+      "learning_rate": 8.632690019079499e-06,
+      "loss": 0.7918,
+      "step": 8838
+    },
+    {
+      "epoch": 0.4864879740216853,
+      "grad_norm": 0.715918779373169,
+      "learning_rate": 8.632392159123274e-06,
+      "loss": 0.744,
+      "step": 8839
+    },
+    {
+      "epoch": 0.48654301282404094,
+      "grad_norm": 0.8206684589385986,
+      "learning_rate": 8.632094271866963e-06,
+      "loss": 0.7852,
+      "step": 8840
+    },
+    {
+      "epoch": 0.48659805162639663,
+      "grad_norm": 0.6502171158790588,
+      "learning_rate": 8.631796357312802e-06,
+      "loss": 0.7653,
+      "step": 8841
+    },
+    {
+      "epoch": 0.48665309042875227,
+      "grad_norm": 0.6987786889076233,
+      "learning_rate": 8.631498415463033e-06,
+      "loss": 0.7669,
+      "step": 8842
+    },
+    {
+      "epoch": 0.48670812923110796,
+      "grad_norm": 0.7902390360832214,
+      "learning_rate": 8.631200446319894e-06,
+      "loss": 0.8438,
+      "step": 8843
+    },
+    {
+      "epoch": 0.4867631680334636,
+      "grad_norm": 0.7464659810066223,
+      "learning_rate": 8.630902449885625e-06,
+      "loss": 0.8276,
+      "step": 8844
+    },
+    {
+      "epoch": 0.4868182068358193,
+      "grad_norm": 0.7375630736351013,
+      "learning_rate": 8.630604426162465e-06,
+      "loss": 0.7921,
+      "step": 8845
+    },
+    {
+      "epoch": 0.4868732456381749,
+      "grad_norm": 0.7206295728683472,
+      "learning_rate": 8.630306375152653e-06,
+      "loss": 0.8424,
+      "step": 8846
+    },
+    {
+      "epoch": 0.4869282844405306,
+      "grad_norm": 0.7384368181228638,
+      "learning_rate": 8.63000829685843e-06,
+      "loss": 0.8702,
+      "step": 8847
+    },
+    {
+      "epoch": 0.48698332324288623,
+      "grad_norm": 0.7839015126228333,
+      "learning_rate": 8.629710191282037e-06,
+      "loss": 0.7064,
+      "step": 8848
+    },
+    {
+      "epoch": 0.48703836204524187,
+      "grad_norm": 0.6909724473953247,
+      "learning_rate": 8.629412058425712e-06,
+      "loss": 0.6924,
+      "step": 8849
+    },
+    {
+      "epoch": 0.48709340084759756,
+      "grad_norm": 0.6553036570549011,
+      "learning_rate": 8.6291138982917e-06,
+      "loss": 0.6526,
+      "step": 8850
+    },
+    {
+      "epoch": 0.4871484396499532,
+      "grad_norm": 0.7202072143554688,
+      "learning_rate": 8.628815710882239e-06,
+      "loss": 0.7272,
+      "step": 8851
+    },
+    {
+      "epoch": 0.4872034784523089,
+      "grad_norm": 0.6898619532585144,
+      "learning_rate": 8.62851749619957e-06,
+      "loss": 0.7687,
+      "step": 8852
+    },
+    {
+      "epoch": 0.4872585172546645,
+      "grad_norm": 0.7888908386230469,
+      "learning_rate": 8.628219254245935e-06,
+      "loss": 0.7654,
+      "step": 8853
+    },
+    {
+      "epoch": 0.4873135560570202,
+      "grad_norm": 0.7312424778938293,
+      "learning_rate": 8.627920985023575e-06,
+      "loss": 0.8053,
+      "step": 8854
+    },
+    {
+      "epoch": 0.48736859485937584,
+      "grad_norm": 0.6588439345359802,
+      "learning_rate": 8.627622688534731e-06,
+      "loss": 0.7229,
+      "step": 8855
+    },
+    {
+      "epoch": 0.4874236336617315,
+      "grad_norm": 0.8292293548583984,
+      "learning_rate": 8.627324364781647e-06,
+      "loss": 0.8482,
+      "step": 8856
+    },
+    {
+      "epoch": 0.48747867246408716,
+      "grad_norm": 0.7573973536491394,
+      "learning_rate": 8.627026013766564e-06,
+      "loss": 0.7282,
+      "step": 8857
+    },
+    {
+      "epoch": 0.48753371126644285,
+      "grad_norm": 1.2215768098831177,
+      "learning_rate": 8.626727635491726e-06,
+      "loss": 0.7771,
+      "step": 8858
+    },
+    {
+      "epoch": 0.4875887500687985,
+      "grad_norm": 0.7324759364128113,
+      "learning_rate": 8.626429229959369e-06,
+      "loss": 0.781,
+      "step": 8859
+    },
+    {
+      "epoch": 0.48764378887115417,
+      "grad_norm": 0.6995676159858704,
+      "learning_rate": 8.626130797171745e-06,
+      "loss": 0.6907,
+      "step": 8860
+    },
+    {
+      "epoch": 0.4876988276735098,
+      "grad_norm": 0.7400509119033813,
+      "learning_rate": 8.625832337131092e-06,
+      "loss": 0.6572,
+      "step": 8861
+    },
+    {
+      "epoch": 0.4877538664758655,
+      "grad_norm": 0.6634842753410339,
+      "learning_rate": 8.625533849839653e-06,
+      "loss": 0.7229,
+      "step": 8862
+    },
+    {
+      "epoch": 0.4878089052782211,
+      "grad_norm": 0.7357299327850342,
+      "learning_rate": 8.625235335299673e-06,
+      "loss": 0.6418,
+      "step": 8863
+    },
+    {
+      "epoch": 0.4878639440805768,
+      "grad_norm": 0.6473466157913208,
+      "learning_rate": 8.624936793513394e-06,
+      "loss": 0.6796,
+      "step": 8864
+    },
+    {
+      "epoch": 0.48791898288293245,
+      "grad_norm": 0.9110734462738037,
+      "learning_rate": 8.62463822448306e-06,
+      "loss": 0.8143,
+      "step": 8865
+    },
+    {
+      "epoch": 0.48797402168528814,
+      "grad_norm": 0.7932308316230774,
+      "learning_rate": 8.624339628210916e-06,
+      "loss": 0.9103,
+      "step": 8866
+    },
+    {
+      "epoch": 0.4880290604876438,
+      "grad_norm": 0.6677752137184143,
+      "learning_rate": 8.624041004699205e-06,
+      "loss": 0.8073,
+      "step": 8867
+    },
+    {
+      "epoch": 0.48808409928999946,
+      "grad_norm": 0.7379121780395508,
+      "learning_rate": 8.623742353950171e-06,
+      "loss": 0.8643,
+      "step": 8868
+    },
+    {
+      "epoch": 0.4881391380923551,
+      "grad_norm": 0.7479479312896729,
+      "learning_rate": 8.623443675966062e-06,
+      "loss": 0.6117,
+      "step": 8869
+    },
+    {
+      "epoch": 0.4881941768947108,
+      "grad_norm": 0.7822794914245605,
+      "learning_rate": 8.623144970749118e-06,
+      "loss": 0.8629,
+      "step": 8870
+    },
+    {
+      "epoch": 0.4882492156970664,
+      "grad_norm": 0.7040950655937195,
+      "learning_rate": 8.622846238301587e-06,
+      "loss": 0.7519,
+      "step": 8871
+    },
+    {
+      "epoch": 0.4883042544994221,
+      "grad_norm": 0.747368574142456,
+      "learning_rate": 8.622547478625714e-06,
+      "loss": 0.7459,
+      "step": 8872
+    },
+    {
+      "epoch": 0.48835929330177774,
+      "grad_norm": 0.6755948066711426,
+      "learning_rate": 8.622248691723742e-06,
+      "loss": 0.7515,
+      "step": 8873
+    },
+    {
+      "epoch": 0.48841433210413343,
+      "grad_norm": 0.7265586256980896,
+      "learning_rate": 8.62194987759792e-06,
+      "loss": 0.7691,
+      "step": 8874
+    },
+    {
+      "epoch": 0.48846937090648906,
+      "grad_norm": 0.6696380972862244,
+      "learning_rate": 8.621651036250493e-06,
+      "loss": 0.778,
+      "step": 8875
+    },
+    {
+      "epoch": 0.48852440970884475,
+      "grad_norm": 0.7666454911231995,
+      "learning_rate": 8.621352167683705e-06,
+      "loss": 0.7396,
+      "step": 8876
+    },
+    {
+      "epoch": 0.4885794485112004,
+      "grad_norm": 0.7079235315322876,
+      "learning_rate": 8.621053271899803e-06,
+      "loss": 0.7917,
+      "step": 8877
+    },
+    {
+      "epoch": 0.4886344873135561,
+      "grad_norm": 0.6888919472694397,
+      "learning_rate": 8.620754348901034e-06,
+      "loss": 0.605,
+      "step": 8878
+    },
+    {
+      "epoch": 0.4886895261159117,
+      "grad_norm": 0.7177572250366211,
+      "learning_rate": 8.620455398689645e-06,
+      "loss": 0.7534,
+      "step": 8879
+    },
+    {
+      "epoch": 0.4887445649182674,
+      "grad_norm": 0.7268772721290588,
+      "learning_rate": 8.620156421267883e-06,
+      "loss": 0.7748,
+      "step": 8880
+    },
+    {
+      "epoch": 0.48879960372062303,
+      "grad_norm": 0.8015080690383911,
+      "learning_rate": 8.619857416637993e-06,
+      "loss": 0.6716,
+      "step": 8881
+    },
+    {
+      "epoch": 0.4888546425229787,
+      "grad_norm": 0.7464118599891663,
+      "learning_rate": 8.619558384802226e-06,
+      "loss": 0.796,
+      "step": 8882
+    },
+    {
+      "epoch": 0.48890968132533436,
+      "grad_norm": 0.6829718351364136,
+      "learning_rate": 8.619259325762826e-06,
+      "loss": 0.788,
+      "step": 8883
+    },
+    {
+      "epoch": 0.48896472012769004,
+      "grad_norm": 0.6553084850311279,
+      "learning_rate": 8.618960239522041e-06,
+      "loss": 0.7215,
+      "step": 8884
+    },
+    {
+      "epoch": 0.4890197589300457,
+      "grad_norm": 0.8056252598762512,
+      "learning_rate": 8.618661126082119e-06,
+      "loss": 0.8588,
+      "step": 8885
+    },
+    {
+      "epoch": 0.48907479773240137,
+      "grad_norm": 0.8145674467086792,
+      "learning_rate": 8.618361985445309e-06,
+      "loss": 0.8095,
+      "step": 8886
+    },
+    {
+      "epoch": 0.489129836534757,
+      "grad_norm": 0.740031898021698,
+      "learning_rate": 8.61806281761386e-06,
+      "loss": 0.7029,
+      "step": 8887
+    },
+    {
+      "epoch": 0.4891848753371127,
+      "grad_norm": 0.7442640662193298,
+      "learning_rate": 8.617763622590019e-06,
+      "loss": 0.782,
+      "step": 8888
+    },
+    {
+      "epoch": 0.4892399141394683,
+      "grad_norm": 0.6992725133895874,
+      "learning_rate": 8.617464400376035e-06,
+      "loss": 0.7877,
+      "step": 8889
+    },
+    {
+      "epoch": 0.489294952941824,
+      "grad_norm": 1.19756281375885,
+      "learning_rate": 8.617165150974157e-06,
+      "loss": 0.6985,
+      "step": 8890
+    },
+    {
+      "epoch": 0.48934999174417965,
+      "grad_norm": 0.6418262720108032,
+      "learning_rate": 8.616865874386633e-06,
+      "loss": 0.7385,
+      "step": 8891
+    },
+    {
+      "epoch": 0.4894050305465353,
+      "grad_norm": 0.787406325340271,
+      "learning_rate": 8.616566570615714e-06,
+      "loss": 0.8686,
+      "step": 8892
+    },
+    {
+      "epoch": 0.48946006934889097,
+      "grad_norm": 0.6990430951118469,
+      "learning_rate": 8.616267239663648e-06,
+      "loss": 0.7683,
+      "step": 8893
+    },
+    {
+      "epoch": 0.4895151081512466,
+      "grad_norm": 0.7180235981941223,
+      "learning_rate": 8.615967881532687e-06,
+      "loss": 0.8337,
+      "step": 8894
+    },
+    {
+      "epoch": 0.4895701469536023,
+      "grad_norm": 0.7647475600242615,
+      "learning_rate": 8.615668496225077e-06,
+      "loss": 0.8668,
+      "step": 8895
+    },
+    {
+      "epoch": 0.4896251857559579,
+      "grad_norm": 0.843063473701477,
+      "learning_rate": 8.615369083743072e-06,
+      "loss": 0.7968,
+      "step": 8896
+    },
+    {
+      "epoch": 0.4896802245583136,
+      "grad_norm": 0.9526075124740601,
+      "learning_rate": 8.61506964408892e-06,
+      "loss": 0.8766,
+      "step": 8897
+    },
+    {
+      "epoch": 0.48973526336066925,
+      "grad_norm": 0.7850056290626526,
+      "learning_rate": 8.614770177264874e-06,
+      "loss": 0.8033,
+      "step": 8898
+    },
+    {
+      "epoch": 0.48979030216302494,
+      "grad_norm": 0.8658629655838013,
+      "learning_rate": 8.614470683273182e-06,
+      "loss": 0.8206,
+      "step": 8899
+    },
+    {
+      "epoch": 0.48984534096538057,
+      "grad_norm": 0.8060176968574524,
+      "learning_rate": 8.614171162116096e-06,
+      "loss": 0.7602,
+      "step": 8900
+    },
+    {
+      "epoch": 0.48990037976773626,
+      "grad_norm": 0.7398280501365662,
+      "learning_rate": 8.613871613795865e-06,
+      "loss": 0.8067,
+      "step": 8901
+    },
+    {
+      "epoch": 0.4899554185700919,
+      "grad_norm": 0.7341256141662598,
+      "learning_rate": 8.613572038314744e-06,
+      "loss": 0.7305,
+      "step": 8902
+    },
+    {
+      "epoch": 0.4900104573724476,
+      "grad_norm": 0.7832887172698975,
+      "learning_rate": 8.613272435674984e-06,
+      "loss": 0.7012,
+      "step": 8903
+    },
+    {
+      "epoch": 0.4900654961748032,
+      "grad_norm": 0.6536995768547058,
+      "learning_rate": 8.612972805878834e-06,
+      "loss": 0.745,
+      "step": 8904
+    },
+    {
+      "epoch": 0.4901205349771589,
+      "grad_norm": 0.7511856555938721,
+      "learning_rate": 8.612673148928547e-06,
+      "loss": 0.7741,
+      "step": 8905
+    },
+    {
+      "epoch": 0.49017557377951454,
+      "grad_norm": 0.6117261648178101,
+      "learning_rate": 8.612373464826377e-06,
+      "loss": 0.5813,
+      "step": 8906
+    },
+    {
+      "epoch": 0.49023061258187023,
+      "grad_norm": 0.7832254767417908,
+      "learning_rate": 8.612073753574574e-06,
+      "loss": 0.7426,
+      "step": 8907
+    },
+    {
+      "epoch": 0.49028565138422586,
+      "grad_norm": 0.7516622543334961,
+      "learning_rate": 8.611774015175393e-06,
+      "loss": 0.8205,
+      "step": 8908
+    },
+    {
+      "epoch": 0.49034069018658155,
+      "grad_norm": 0.7776936888694763,
+      "learning_rate": 8.611474249631085e-06,
+      "loss": 0.8457,
+      "step": 8909
+    },
+    {
+      "epoch": 0.4903957289889372,
+      "grad_norm": 0.9364853501319885,
+      "learning_rate": 8.6111744569439e-06,
+      "loss": 0.9114,
+      "step": 8910
+    },
+    {
+      "epoch": 0.4904507677912929,
+      "grad_norm": 0.7584181427955627,
+      "learning_rate": 8.610874637116099e-06,
+      "loss": 0.6852,
+      "step": 8911
+    },
+    {
+      "epoch": 0.4905058065936485,
+      "grad_norm": 0.7326254844665527,
+      "learning_rate": 8.610574790149929e-06,
+      "loss": 0.7843,
+      "step": 8912
+    },
+    {
+      "epoch": 0.4905608453960042,
+      "grad_norm": 0.918258547782898,
+      "learning_rate": 8.610274916047645e-06,
+      "loss": 0.766,
+      "step": 8913
+    },
+    {
+      "epoch": 0.49061588419835983,
+      "grad_norm": 1.0083420276641846,
+      "learning_rate": 8.609975014811502e-06,
+      "loss": 0.7436,
+      "step": 8914
+    },
+    {
+      "epoch": 0.4906709230007155,
+      "grad_norm": 0.712664783000946,
+      "learning_rate": 8.609675086443752e-06,
+      "loss": 0.7891,
+      "step": 8915
+    },
+    {
+      "epoch": 0.49072596180307115,
+      "grad_norm": 0.7635206580162048,
+      "learning_rate": 8.609375130946651e-06,
+      "loss": 0.7842,
+      "step": 8916
+    },
+    {
+      "epoch": 0.49078100060542684,
+      "grad_norm": 0.7567723989486694,
+      "learning_rate": 8.609075148322452e-06,
+      "loss": 0.8435,
+      "step": 8917
+    },
+    {
+      "epoch": 0.4908360394077825,
+      "grad_norm": 0.8918718099594116,
+      "learning_rate": 8.60877513857341e-06,
+      "loss": 0.8015,
+      "step": 8918
+    },
+    {
+      "epoch": 0.49089107821013817,
+      "grad_norm": 0.8701914548873901,
+      "learning_rate": 8.608475101701781e-06,
+      "loss": 0.7806,
+      "step": 8919
+    },
+    {
+      "epoch": 0.4909461170124938,
+      "grad_norm": 0.7528215646743774,
+      "learning_rate": 8.608175037709819e-06,
+      "loss": 0.7958,
+      "step": 8920
+    },
+    {
+      "epoch": 0.4910011558148495,
+      "grad_norm": 0.7277387380599976,
+      "learning_rate": 8.60787494659978e-06,
+      "loss": 0.7878,
+      "step": 8921
+    },
+    {
+      "epoch": 0.4910561946172051,
+      "grad_norm": 0.6739892959594727,
+      "learning_rate": 8.607574828373917e-06,
+      "loss": 0.7212,
+      "step": 8922
+    },
+    {
+      "epoch": 0.4911112334195608,
+      "grad_norm": 0.712480366230011,
+      "learning_rate": 8.607274683034487e-06,
+      "loss": 0.7966,
+      "step": 8923
+    },
+    {
+      "epoch": 0.49116627222191644,
+      "grad_norm": 0.7192126512527466,
+      "learning_rate": 8.606974510583747e-06,
+      "loss": 0.7032,
+      "step": 8924
+    },
+    {
+      "epoch": 0.49122131102427213,
+      "grad_norm": 0.7502614855766296,
+      "learning_rate": 8.606674311023953e-06,
+      "loss": 0.7465,
+      "step": 8925
+    },
+    {
+      "epoch": 0.49127634982662777,
+      "grad_norm": 0.8475236892700195,
+      "learning_rate": 8.606374084357361e-06,
+      "loss": 0.8083,
+      "step": 8926
+    },
+    {
+      "epoch": 0.49133138862898346,
+      "grad_norm": 0.6972761750221252,
+      "learning_rate": 8.606073830586224e-06,
+      "loss": 0.7206,
+      "step": 8927
+    },
+    {
+      "epoch": 0.4913864274313391,
+      "grad_norm": 0.6209561824798584,
+      "learning_rate": 8.605773549712803e-06,
+      "loss": 0.6664,
+      "step": 8928
+    },
+    {
+      "epoch": 0.4914414662336948,
+      "grad_norm": 0.7905771732330322,
+      "learning_rate": 8.605473241739353e-06,
+      "loss": 0.7243,
+      "step": 8929
+    },
+    {
+      "epoch": 0.4914965050360504,
+      "grad_norm": 0.762959897518158,
+      "learning_rate": 8.605172906668131e-06,
+      "loss": 0.7747,
+      "step": 8930
+    },
+    {
+      "epoch": 0.4915515438384061,
+      "grad_norm": 0.7297530174255371,
+      "learning_rate": 8.604872544501394e-06,
+      "loss": 0.7441,
+      "step": 8931
+    },
+    {
+      "epoch": 0.49160658264076174,
+      "grad_norm": 0.6732318997383118,
+      "learning_rate": 8.6045721552414e-06,
+      "loss": 0.7621,
+      "step": 8932
+    },
+    {
+      "epoch": 0.4916616214431174,
+      "grad_norm": 0.7010045647621155,
+      "learning_rate": 8.604271738890407e-06,
+      "loss": 0.7971,
+      "step": 8933
+    },
+    {
+      "epoch": 0.49171666024547306,
+      "grad_norm": 0.6996648907661438,
+      "learning_rate": 8.603971295450672e-06,
+      "loss": 0.8119,
+      "step": 8934
+    },
+    {
+      "epoch": 0.4917716990478287,
+      "grad_norm": 0.7679941058158875,
+      "learning_rate": 8.603670824924456e-06,
+      "loss": 0.8035,
+      "step": 8935
+    },
+    {
+      "epoch": 0.4918267378501844,
+      "grad_norm": 0.8009630441665649,
+      "learning_rate": 8.603370327314011e-06,
+      "loss": 0.7817,
+      "step": 8936
+    },
+    {
+      "epoch": 0.49188177665254,
+      "grad_norm": 0.7167709469795227,
+      "learning_rate": 8.603069802621601e-06,
+      "loss": 0.7621,
+      "step": 8937
+    },
+    {
+      "epoch": 0.4919368154548957,
+      "grad_norm": 0.7447960376739502,
+      "learning_rate": 8.602769250849483e-06,
+      "loss": 0.7664,
+      "step": 8938
+    },
+    {
+      "epoch": 0.49199185425725134,
+      "grad_norm": 0.653131365776062,
+      "learning_rate": 8.602468671999915e-06,
+      "loss": 0.6927,
+      "step": 8939
+    },
+    {
+      "epoch": 0.492046893059607,
+      "grad_norm": 0.6758691072463989,
+      "learning_rate": 8.602168066075158e-06,
+      "loss": 0.7519,
+      "step": 8940
+    },
+    {
+      "epoch": 0.49210193186196266,
+      "grad_norm": 0.9186220765113831,
+      "learning_rate": 8.60186743307747e-06,
+      "loss": 0.7265,
+      "step": 8941
+    },
+    {
+      "epoch": 0.49215697066431835,
+      "grad_norm": 0.6781855225563049,
+      "learning_rate": 8.60156677300911e-06,
+      "loss": 0.6719,
+      "step": 8942
+    },
+    {
+      "epoch": 0.492212009466674,
+      "grad_norm": 0.7262865304946899,
+      "learning_rate": 8.601266085872336e-06,
+      "loss": 0.6449,
+      "step": 8943
+    },
+    {
+      "epoch": 0.4922670482690297,
+      "grad_norm": 0.6877585053443909,
+      "learning_rate": 8.600965371669411e-06,
+      "loss": 0.6999,
+      "step": 8944
+    },
+    {
+      "epoch": 0.4923220870713853,
+      "grad_norm": 1.1133443117141724,
+      "learning_rate": 8.600664630402596e-06,
+      "loss": 0.7842,
+      "step": 8945
+    },
+    {
+      "epoch": 0.492377125873741,
+      "grad_norm": 0.643478274345398,
+      "learning_rate": 8.600363862074149e-06,
+      "loss": 0.7009,
+      "step": 8946
+    },
+    {
+      "epoch": 0.49243216467609663,
+      "grad_norm": 0.7692574262619019,
+      "learning_rate": 8.600063066686331e-06,
+      "loss": 0.7777,
+      "step": 8947
+    },
+    {
+      "epoch": 0.4924872034784523,
+      "grad_norm": 0.884963870048523,
+      "learning_rate": 8.599762244241403e-06,
+      "loss": 0.7789,
+      "step": 8948
+    },
+    {
+      "epoch": 0.49254224228080795,
+      "grad_norm": 0.6918813586235046,
+      "learning_rate": 8.599461394741624e-06,
+      "loss": 0.7769,
+      "step": 8949
+    },
+    {
+      "epoch": 0.49259728108316364,
+      "grad_norm": 0.7432044148445129,
+      "learning_rate": 8.599160518189258e-06,
+      "loss": 0.7972,
+      "step": 8950
+    },
+    {
+      "epoch": 0.4926523198855193,
+      "grad_norm": 0.7530491948127747,
+      "learning_rate": 8.598859614586564e-06,
+      "loss": 0.8812,
+      "step": 8951
+    },
+    {
+      "epoch": 0.49270735868787496,
+      "grad_norm": 0.8738592267036438,
+      "learning_rate": 8.598558683935806e-06,
+      "loss": 0.6967,
+      "step": 8952
+    },
+    {
+      "epoch": 0.4927623974902306,
+      "grad_norm": 1.032084584236145,
+      "learning_rate": 8.598257726239242e-06,
+      "loss": 0.8513,
+      "step": 8953
+    },
+    {
+      "epoch": 0.4928174362925863,
+      "grad_norm": 0.8717961311340332,
+      "learning_rate": 8.597956741499136e-06,
+      "loss": 0.7703,
+      "step": 8954
+    },
+    {
+      "epoch": 0.4928724750949419,
+      "grad_norm": 0.6788356900215149,
+      "learning_rate": 8.597655729717753e-06,
+      "loss": 0.7649,
+      "step": 8955
+    },
+    {
+      "epoch": 0.4929275138972976,
+      "grad_norm": 1.0595613718032837,
+      "learning_rate": 8.59735469089735e-06,
+      "loss": 0.6967,
+      "step": 8956
+    },
+    {
+      "epoch": 0.49298255269965324,
+      "grad_norm": 0.7583820819854736,
+      "learning_rate": 8.597053625040193e-06,
+      "loss": 0.8384,
+      "step": 8957
+    },
+    {
+      "epoch": 0.49303759150200893,
+      "grad_norm": 0.7232168912887573,
+      "learning_rate": 8.596752532148545e-06,
+      "loss": 0.7643,
+      "step": 8958
+    },
+    {
+      "epoch": 0.49309263030436457,
+      "grad_norm": 0.727190375328064,
+      "learning_rate": 8.596451412224666e-06,
+      "loss": 0.845,
+      "step": 8959
+    },
+    {
+      "epoch": 0.49314766910672025,
+      "grad_norm": 0.6844252347946167,
+      "learning_rate": 8.596150265270821e-06,
+      "loss": 0.7099,
+      "step": 8960
+    },
+    {
+      "epoch": 0.4932027079090759,
+      "grad_norm": 0.7379910945892334,
+      "learning_rate": 8.595849091289275e-06,
+      "loss": 0.8168,
+      "step": 8961
+    },
+    {
+      "epoch": 0.4932577467114316,
+      "grad_norm": 0.77718186378479,
+      "learning_rate": 8.595547890282288e-06,
+      "loss": 0.8457,
+      "step": 8962
+    },
+    {
+      "epoch": 0.4933127855137872,
+      "grad_norm": 0.686126172542572,
+      "learning_rate": 8.595246662252127e-06,
+      "loss": 0.7918,
+      "step": 8963
+    },
+    {
+      "epoch": 0.4933678243161429,
+      "grad_norm": 0.7406145930290222,
+      "learning_rate": 8.594945407201051e-06,
+      "loss": 0.6866,
+      "step": 8964
+    },
+    {
+      "epoch": 0.49342286311849853,
+      "grad_norm": 0.9543277025222778,
+      "learning_rate": 8.594644125131331e-06,
+      "loss": 0.8444,
+      "step": 8965
+    },
+    {
+      "epoch": 0.4934779019208542,
+      "grad_norm": 0.8659517765045166,
+      "learning_rate": 8.594342816045228e-06,
+      "loss": 0.7661,
+      "step": 8966
+    },
+    {
+      "epoch": 0.49353294072320986,
+      "grad_norm": 0.7289552092552185,
+      "learning_rate": 8.594041479945005e-06,
+      "loss": 0.7734,
+      "step": 8967
+    },
+    {
+      "epoch": 0.49358797952556555,
+      "grad_norm": 0.7232840657234192,
+      "learning_rate": 8.59374011683293e-06,
+      "loss": 0.8557,
+      "step": 8968
+    },
+    {
+      "epoch": 0.4936430183279212,
+      "grad_norm": 0.738684356212616,
+      "learning_rate": 8.593438726711265e-06,
+      "loss": 0.7779,
+      "step": 8969
+    },
+    {
+      "epoch": 0.49369805713027687,
+      "grad_norm": 0.7486668229103088,
+      "learning_rate": 8.593137309582276e-06,
+      "loss": 0.7326,
+      "step": 8970
+    },
+    {
+      "epoch": 0.4937530959326325,
+      "grad_norm": 0.6564297080039978,
+      "learning_rate": 8.59283586544823e-06,
+      "loss": 0.6927,
+      "step": 8971
+    },
+    {
+      "epoch": 0.4938081347349882,
+      "grad_norm": 0.722540557384491,
+      "learning_rate": 8.592534394311392e-06,
+      "loss": 0.7254,
+      "step": 8972
+    },
+    {
+      "epoch": 0.4938631735373438,
+      "grad_norm": 0.7466141581535339,
+      "learning_rate": 8.592232896174026e-06,
+      "loss": 0.8551,
+      "step": 8973
+    },
+    {
+      "epoch": 0.4939182123396995,
+      "grad_norm": 0.7819109559059143,
+      "learning_rate": 8.591931371038398e-06,
+      "loss": 0.7271,
+      "step": 8974
+    },
+    {
+      "epoch": 0.49397325114205515,
+      "grad_norm": 0.7847672700881958,
+      "learning_rate": 8.591629818906776e-06,
+      "loss": 0.8404,
+      "step": 8975
+    },
+    {
+      "epoch": 0.49402828994441084,
+      "grad_norm": 0.8167426586151123,
+      "learning_rate": 8.591328239781428e-06,
+      "loss": 0.7375,
+      "step": 8976
+    },
+    {
+      "epoch": 0.49408332874676647,
+      "grad_norm": 0.7894755005836487,
+      "learning_rate": 8.591026633664615e-06,
+      "loss": 0.7872,
+      "step": 8977
+    },
+    {
+      "epoch": 0.4941383675491221,
+      "grad_norm": 0.726204514503479,
+      "learning_rate": 8.590725000558609e-06,
+      "loss": 0.7289,
+      "step": 8978
+    },
+    {
+      "epoch": 0.4941934063514778,
+      "grad_norm": 0.7116577625274658,
+      "learning_rate": 8.590423340465675e-06,
+      "loss": 0.7379,
+      "step": 8979
+    },
+    {
+      "epoch": 0.4942484451538334,
+      "grad_norm": 0.7302193641662598,
+      "learning_rate": 8.59012165338808e-06,
+      "loss": 0.7951,
+      "step": 8980
+    },
+    {
+      "epoch": 0.4943034839561891,
+      "grad_norm": 0.680555522441864,
+      "learning_rate": 8.58981993932809e-06,
+      "loss": 0.7609,
+      "step": 8981
+    },
+    {
+      "epoch": 0.49435852275854475,
+      "grad_norm": 0.874546229839325,
+      "learning_rate": 8.589518198287976e-06,
+      "loss": 0.8025,
+      "step": 8982
+    },
+    {
+      "epoch": 0.49441356156090044,
+      "grad_norm": 0.7164583206176758,
+      "learning_rate": 8.589216430270004e-06,
+      "loss": 0.7466,
+      "step": 8983
+    },
+    {
+      "epoch": 0.49446860036325607,
+      "grad_norm": 0.9155141115188599,
+      "learning_rate": 8.588914635276442e-06,
+      "loss": 0.7896,
+      "step": 8984
+    },
+    {
+      "epoch": 0.49452363916561176,
+      "grad_norm": 0.6777059435844421,
+      "learning_rate": 8.588612813309558e-06,
+      "loss": 0.7468,
+      "step": 8985
+    },
+    {
+      "epoch": 0.4945786779679674,
+      "grad_norm": 0.7100371718406677,
+      "learning_rate": 8.58831096437162e-06,
+      "loss": 0.7216,
+      "step": 8986
+    },
+    {
+      "epoch": 0.4946337167703231,
+      "grad_norm": 0.6842584609985352,
+      "learning_rate": 8.5880090884649e-06,
+      "loss": 0.7103,
+      "step": 8987
+    },
+    {
+      "epoch": 0.4946887555726787,
+      "grad_norm": 0.6347573399543762,
+      "learning_rate": 8.587707185591661e-06,
+      "loss": 0.7103,
+      "step": 8988
+    },
+    {
+      "epoch": 0.4947437943750344,
+      "grad_norm": 0.7175829410552979,
+      "learning_rate": 8.587405255754177e-06,
+      "loss": 0.8375,
+      "step": 8989
+    },
+    {
+      "epoch": 0.49479883317739004,
+      "grad_norm": 0.8402735590934753,
+      "learning_rate": 8.587103298954715e-06,
+      "loss": 0.6841,
+      "step": 8990
+    },
+    {
+      "epoch": 0.49485387197974573,
+      "grad_norm": 0.6988743543624878,
+      "learning_rate": 8.586801315195545e-06,
+      "loss": 0.7637,
+      "step": 8991
+    },
+    {
+      "epoch": 0.49490891078210136,
+      "grad_norm": 0.6672561168670654,
+      "learning_rate": 8.586499304478934e-06,
+      "loss": 0.7103,
+      "step": 8992
+    },
+    {
+      "epoch": 0.49496394958445705,
+      "grad_norm": 0.6821330189704895,
+      "learning_rate": 8.586197266807158e-06,
+      "loss": 0.6881,
+      "step": 8993
+    },
+    {
+      "epoch": 0.4950189883868127,
+      "grad_norm": 0.7886170744895935,
+      "learning_rate": 8.585895202182482e-06,
+      "loss": 0.7892,
+      "step": 8994
+    },
+    {
+      "epoch": 0.4950740271891684,
+      "grad_norm": 0.7348074913024902,
+      "learning_rate": 8.585593110607177e-06,
+      "loss": 0.7835,
+      "step": 8995
+    },
+    {
+      "epoch": 0.495129065991524,
+      "grad_norm": 0.9375506639480591,
+      "learning_rate": 8.585290992083514e-06,
+      "loss": 0.8017,
+      "step": 8996
+    },
+    {
+      "epoch": 0.4951841047938797,
+      "grad_norm": 0.7442331910133362,
+      "learning_rate": 8.584988846613765e-06,
+      "loss": 0.72,
+      "step": 8997
+    },
+    {
+      "epoch": 0.49523914359623533,
+      "grad_norm": 0.7347918748855591,
+      "learning_rate": 8.584686674200197e-06,
+      "loss": 0.8229,
+      "step": 8998
+    },
+    {
+      "epoch": 0.495294182398591,
+      "grad_norm": 0.7168740630149841,
+      "learning_rate": 8.584384474845084e-06,
+      "loss": 0.7288,
+      "step": 8999
+    },
+    {
+      "epoch": 0.49534922120094665,
+      "grad_norm": 0.7834853529930115,
+      "learning_rate": 8.584082248550697e-06,
+      "loss": 0.8521,
+      "step": 9000
+    },
+    {
+      "epoch": 0.49540426000330234,
+      "grad_norm": 0.6499035358428955,
+      "learning_rate": 8.58377999531931e-06,
+      "loss": 0.6887,
+      "step": 9001
+    },
+    {
+      "epoch": 0.495459298805658,
+      "grad_norm": 0.8000181913375854,
+      "learning_rate": 8.583477715153189e-06,
+      "loss": 0.8688,
+      "step": 9002
+    },
+    {
+      "epoch": 0.49551433760801367,
+      "grad_norm": 0.7539342045783997,
+      "learning_rate": 8.58317540805461e-06,
+      "loss": 0.6151,
+      "step": 9003
+    },
+    {
+      "epoch": 0.4955693764103693,
+      "grad_norm": 0.7677812576293945,
+      "learning_rate": 8.582873074025841e-06,
+      "loss": 0.8168,
+      "step": 9004
+    },
+    {
+      "epoch": 0.495624415212725,
+      "grad_norm": 0.7679157853126526,
+      "learning_rate": 8.58257071306916e-06,
+      "loss": 0.7719,
+      "step": 9005
+    },
+    {
+      "epoch": 0.4956794540150806,
+      "grad_norm": 0.9745703935623169,
+      "learning_rate": 8.582268325186836e-06,
+      "loss": 0.8272,
+      "step": 9006
+    },
+    {
+      "epoch": 0.4957344928174363,
+      "grad_norm": 0.66932612657547,
+      "learning_rate": 8.581965910381143e-06,
+      "loss": 0.7256,
+      "step": 9007
+    },
+    {
+      "epoch": 0.49578953161979195,
+      "grad_norm": 0.7630981206893921,
+      "learning_rate": 8.581663468654351e-06,
+      "loss": 0.7594,
+      "step": 9008
+    },
+    {
+      "epoch": 0.49584457042214763,
+      "grad_norm": 0.7420778870582581,
+      "learning_rate": 8.581361000008737e-06,
+      "loss": 0.7834,
+      "step": 9009
+    },
+    {
+      "epoch": 0.49589960922450327,
+      "grad_norm": 0.6775205731391907,
+      "learning_rate": 8.58105850444657e-06,
+      "loss": 0.7609,
+      "step": 9010
+    },
+    {
+      "epoch": 0.49595464802685896,
+      "grad_norm": 0.6588264107704163,
+      "learning_rate": 8.580755981970128e-06,
+      "loss": 0.805,
+      "step": 9011
+    },
+    {
+      "epoch": 0.4960096868292146,
+      "grad_norm": 0.7325689196586609,
+      "learning_rate": 8.580453432581681e-06,
+      "loss": 0.8817,
+      "step": 9012
+    },
+    {
+      "epoch": 0.4960647256315703,
+      "grad_norm": 0.7319273948669434,
+      "learning_rate": 8.580150856283505e-06,
+      "loss": 0.8001,
+      "step": 9013
+    },
+    {
+      "epoch": 0.4961197644339259,
+      "grad_norm": 0.7841789126396179,
+      "learning_rate": 8.579848253077875e-06,
+      "loss": 0.8415,
+      "step": 9014
+    },
+    {
+      "epoch": 0.4961748032362816,
+      "grad_norm": 0.7593979239463806,
+      "learning_rate": 8.579545622967062e-06,
+      "loss": 0.8238,
+      "step": 9015
+    },
+    {
+      "epoch": 0.49622984203863724,
+      "grad_norm": 0.6938808560371399,
+      "learning_rate": 8.579242965953343e-06,
+      "loss": 0.7325,
+      "step": 9016
+    },
+    {
+      "epoch": 0.4962848808409929,
+      "grad_norm": 0.7907594442367554,
+      "learning_rate": 8.578940282038993e-06,
+      "loss": 0.6947,
+      "step": 9017
+    },
+    {
+      "epoch": 0.49633991964334856,
+      "grad_norm": 0.708703875541687,
+      "learning_rate": 8.578637571226283e-06,
+      "loss": 0.6712,
+      "step": 9018
+    },
+    {
+      "epoch": 0.49639495844570425,
+      "grad_norm": 0.6820377707481384,
+      "learning_rate": 8.578334833517492e-06,
+      "loss": 0.7269,
+      "step": 9019
+    },
+    {
+      "epoch": 0.4964499972480599,
+      "grad_norm": 0.6858653426170349,
+      "learning_rate": 8.578032068914896e-06,
+      "loss": 0.7325,
+      "step": 9020
+    },
+    {
+      "epoch": 0.4965050360504155,
+      "grad_norm": 0.8758736848831177,
+      "learning_rate": 8.577729277420768e-06,
+      "loss": 0.6652,
+      "step": 9021
+    },
+    {
+      "epoch": 0.4965600748527712,
+      "grad_norm": 0.731316328048706,
+      "learning_rate": 8.577426459037383e-06,
+      "loss": 0.7835,
+      "step": 9022
+    },
+    {
+      "epoch": 0.49661511365512684,
+      "grad_norm": 0.813778817653656,
+      "learning_rate": 8.57712361376702e-06,
+      "loss": 0.8025,
+      "step": 9023
+    },
+    {
+      "epoch": 0.4966701524574825,
+      "grad_norm": 0.7167351841926575,
+      "learning_rate": 8.576820741611952e-06,
+      "loss": 0.7483,
+      "step": 9024
+    },
+    {
+      "epoch": 0.49672519125983816,
+      "grad_norm": 0.7243192791938782,
+      "learning_rate": 8.576517842574457e-06,
+      "loss": 0.8411,
+      "step": 9025
+    },
+    {
+      "epoch": 0.49678023006219385,
+      "grad_norm": 0.5869036316871643,
+      "learning_rate": 8.576214916656814e-06,
+      "loss": 0.6661,
+      "step": 9026
+    },
+    {
+      "epoch": 0.4968352688645495,
+      "grad_norm": 0.7502203583717346,
+      "learning_rate": 8.575911963861293e-06,
+      "loss": 0.8838,
+      "step": 9027
+    },
+    {
+      "epoch": 0.4968903076669052,
+      "grad_norm": 0.687562108039856,
+      "learning_rate": 8.575608984190177e-06,
+      "loss": 0.7446,
+      "step": 9028
+    },
+    {
+      "epoch": 0.4969453464692608,
+      "grad_norm": 0.7735342383384705,
+      "learning_rate": 8.57530597764574e-06,
+      "loss": 0.8464,
+      "step": 9029
+    },
+    {
+      "epoch": 0.4970003852716165,
+      "grad_norm": 0.7828487753868103,
+      "learning_rate": 8.575002944230261e-06,
+      "loss": 0.7504,
+      "step": 9030
+    },
+    {
+      "epoch": 0.49705542407397213,
+      "grad_norm": 0.6359286904335022,
+      "learning_rate": 8.574699883946018e-06,
+      "loss": 0.6805,
+      "step": 9031
+    },
+    {
+      "epoch": 0.4971104628763278,
+      "grad_norm": 0.7462830543518066,
+      "learning_rate": 8.574396796795285e-06,
+      "loss": 0.8317,
+      "step": 9032
+    },
+    {
+      "epoch": 0.49716550167868345,
+      "grad_norm": 0.705115795135498,
+      "learning_rate": 8.574093682780344e-06,
+      "loss": 0.7401,
+      "step": 9033
+    },
+    {
+      "epoch": 0.49722054048103914,
+      "grad_norm": 0.6466538310050964,
+      "learning_rate": 8.573790541903472e-06,
+      "loss": 0.7761,
+      "step": 9034
+    },
+    {
+      "epoch": 0.4972755792833948,
+      "grad_norm": 0.7479867339134216,
+      "learning_rate": 8.573487374166946e-06,
+      "loss": 0.8394,
+      "step": 9035
+    },
+    {
+      "epoch": 0.49733061808575046,
+      "grad_norm": 0.7378019094467163,
+      "learning_rate": 8.573184179573046e-06,
+      "loss": 0.8215,
+      "step": 9036
+    },
+    {
+      "epoch": 0.4973856568881061,
+      "grad_norm": 0.6526094675064087,
+      "learning_rate": 8.57288095812405e-06,
+      "loss": 0.8055,
+      "step": 9037
+    },
+    {
+      "epoch": 0.4974406956904618,
+      "grad_norm": 0.679595947265625,
+      "learning_rate": 8.572577709822238e-06,
+      "loss": 0.8241,
+      "step": 9038
+    },
+    {
+      "epoch": 0.4974957344928174,
+      "grad_norm": 0.753466010093689,
+      "learning_rate": 8.572274434669886e-06,
+      "loss": 0.896,
+      "step": 9039
+    },
+    {
+      "epoch": 0.4975507732951731,
+      "grad_norm": 0.7068368792533875,
+      "learning_rate": 8.571971132669277e-06,
+      "loss": 0.778,
+      "step": 9040
+    },
+    {
+      "epoch": 0.49760581209752874,
+      "grad_norm": 0.7397973537445068,
+      "learning_rate": 8.571667803822689e-06,
+      "loss": 0.782,
+      "step": 9041
+    },
+    {
+      "epoch": 0.49766085089988443,
+      "grad_norm": 0.7837033271789551,
+      "learning_rate": 8.571364448132402e-06,
+      "loss": 0.7509,
+      "step": 9042
+    },
+    {
+      "epoch": 0.49771588970224007,
+      "grad_norm": 0.6808765530586243,
+      "learning_rate": 8.571061065600696e-06,
+      "loss": 0.672,
+      "step": 9043
+    },
+    {
+      "epoch": 0.49777092850459576,
+      "grad_norm": 0.6574100255966187,
+      "learning_rate": 8.570757656229852e-06,
+      "loss": 0.751,
+      "step": 9044
+    },
+    {
+      "epoch": 0.4978259673069514,
+      "grad_norm": 0.7357671856880188,
+      "learning_rate": 8.570454220022146e-06,
+      "loss": 0.7977,
+      "step": 9045
+    },
+    {
+      "epoch": 0.4978810061093071,
+      "grad_norm": 0.7937216758728027,
+      "learning_rate": 8.570150756979865e-06,
+      "loss": 0.8151,
+      "step": 9046
+    },
+    {
+      "epoch": 0.4979360449116627,
+      "grad_norm": 0.7050907611846924,
+      "learning_rate": 8.569847267105285e-06,
+      "loss": 0.7667,
+      "step": 9047
+    },
+    {
+      "epoch": 0.4979910837140184,
+      "grad_norm": 0.7105300426483154,
+      "learning_rate": 8.569543750400688e-06,
+      "loss": 0.7031,
+      "step": 9048
+    },
+    {
+      "epoch": 0.49804612251637403,
+      "grad_norm": 0.7174646854400635,
+      "learning_rate": 8.569240206868358e-06,
+      "loss": 0.7692,
+      "step": 9049
+    },
+    {
+      "epoch": 0.4981011613187297,
+      "grad_norm": 0.7525906562805176,
+      "learning_rate": 8.568936636510573e-06,
+      "loss": 0.7584,
+      "step": 9050
+    },
+    {
+      "epoch": 0.49815620012108536,
+      "grad_norm": 1.5518100261688232,
+      "learning_rate": 8.568633039329615e-06,
+      "loss": 0.7932,
+      "step": 9051
+    },
+    {
+      "epoch": 0.49821123892344105,
+      "grad_norm": 0.7037720084190369,
+      "learning_rate": 8.568329415327766e-06,
+      "loss": 0.8345,
+      "step": 9052
+    },
+    {
+      "epoch": 0.4982662777257967,
+      "grad_norm": 0.6422694325447083,
+      "learning_rate": 8.568025764507308e-06,
+      "loss": 0.7396,
+      "step": 9053
+    },
+    {
+      "epoch": 0.49832131652815237,
+      "grad_norm": 0.777306854724884,
+      "learning_rate": 8.567722086870525e-06,
+      "loss": 0.8605,
+      "step": 9054
+    },
+    {
+      "epoch": 0.498376355330508,
+      "grad_norm": 0.6619865298271179,
+      "learning_rate": 8.567418382419697e-06,
+      "loss": 0.7395,
+      "step": 9055
+    },
+    {
+      "epoch": 0.4984313941328637,
+      "grad_norm": 0.7214456796646118,
+      "learning_rate": 8.567114651157106e-06,
+      "loss": 0.7932,
+      "step": 9056
+    },
+    {
+      "epoch": 0.4984864329352193,
+      "grad_norm": 0.75806725025177,
+      "learning_rate": 8.566810893085037e-06,
+      "loss": 0.7998,
+      "step": 9057
+    },
+    {
+      "epoch": 0.498541471737575,
+      "grad_norm": 0.8089895844459534,
+      "learning_rate": 8.566507108205773e-06,
+      "loss": 0.7849,
+      "step": 9058
+    },
+    {
+      "epoch": 0.49859651053993065,
+      "grad_norm": 0.817814290523529,
+      "learning_rate": 8.566203296521597e-06,
+      "loss": 0.7261,
+      "step": 9059
+    },
+    {
+      "epoch": 0.49865154934228634,
+      "grad_norm": 0.7417539954185486,
+      "learning_rate": 8.56589945803479e-06,
+      "loss": 0.7087,
+      "step": 9060
+    },
+    {
+      "epoch": 0.49870658814464197,
+      "grad_norm": 0.7518000602722168,
+      "learning_rate": 8.565595592747639e-06,
+      "loss": 0.7245,
+      "step": 9061
+    },
+    {
+      "epoch": 0.49876162694699766,
+      "grad_norm": 0.9537304043769836,
+      "learning_rate": 8.565291700662423e-06,
+      "loss": 0.901,
+      "step": 9062
+    },
+    {
+      "epoch": 0.4988166657493533,
+      "grad_norm": 0.784545361995697,
+      "learning_rate": 8.56498778178143e-06,
+      "loss": 0.7813,
+      "step": 9063
+    },
+    {
+      "epoch": 0.4988717045517089,
+      "grad_norm": 0.9218429923057556,
+      "learning_rate": 8.564683836106945e-06,
+      "loss": 0.8452,
+      "step": 9064
+    },
+    {
+      "epoch": 0.4989267433540646,
+      "grad_norm": 0.6902065277099609,
+      "learning_rate": 8.56437986364125e-06,
+      "loss": 0.7527,
+      "step": 9065
+    },
+    {
+      "epoch": 0.49898178215642025,
+      "grad_norm": 0.7388677000999451,
+      "learning_rate": 8.56407586438663e-06,
+      "loss": 0.82,
+      "step": 9066
+    },
+    {
+      "epoch": 0.49903682095877594,
+      "grad_norm": 0.6959313154220581,
+      "learning_rate": 8.563771838345369e-06,
+      "loss": 0.7274,
+      "step": 9067
+    },
+    {
+      "epoch": 0.4990918597611316,
+      "grad_norm": 0.6582610607147217,
+      "learning_rate": 8.563467785519753e-06,
+      "loss": 0.6518,
+      "step": 9068
+    },
+    {
+      "epoch": 0.49914689856348726,
+      "grad_norm": 0.6525924801826477,
+      "learning_rate": 8.563163705912066e-06,
+      "loss": 0.7006,
+      "step": 9069
+    },
+    {
+      "epoch": 0.4992019373658429,
+      "grad_norm": 0.8092843890190125,
+      "learning_rate": 8.562859599524596e-06,
+      "loss": 0.6915,
+      "step": 9070
+    },
+    {
+      "epoch": 0.4992569761681986,
+      "grad_norm": 0.6540575623512268,
+      "learning_rate": 8.562555466359626e-06,
+      "loss": 0.6729,
+      "step": 9071
+    },
+    {
+      "epoch": 0.4993120149705542,
+      "grad_norm": 0.8220445513725281,
+      "learning_rate": 8.562251306419443e-06,
+      "loss": 0.8172,
+      "step": 9072
+    },
+    {
+      "epoch": 0.4993670537729099,
+      "grad_norm": 0.7461502552032471,
+      "learning_rate": 8.561947119706334e-06,
+      "loss": 0.6902,
+      "step": 9073
+    },
+    {
+      "epoch": 0.49942209257526554,
+      "grad_norm": 0.8166316151618958,
+      "learning_rate": 8.56164290622258e-06,
+      "loss": 0.8238,
+      "step": 9074
+    },
+    {
+      "epoch": 0.49947713137762123,
+      "grad_norm": 0.8453896641731262,
+      "learning_rate": 8.561338665970476e-06,
+      "loss": 0.7697,
+      "step": 9075
+    },
+    {
+      "epoch": 0.49953217017997686,
+      "grad_norm": 0.7606340050697327,
+      "learning_rate": 8.5610343989523e-06,
+      "loss": 0.6951,
+      "step": 9076
+    },
+    {
+      "epoch": 0.49958720898233255,
+      "grad_norm": 0.7408013343811035,
+      "learning_rate": 8.560730105170345e-06,
+      "loss": 0.8298,
+      "step": 9077
+    },
+    {
+      "epoch": 0.4996422477846882,
+      "grad_norm": 0.7625541090965271,
+      "learning_rate": 8.560425784626896e-06,
+      "loss": 0.6738,
+      "step": 9078
+    },
+    {
+      "epoch": 0.4996972865870439,
+      "grad_norm": 0.6940996646881104,
+      "learning_rate": 8.560121437324238e-06,
+      "loss": 0.78,
+      "step": 9079
+    },
+    {
+      "epoch": 0.4997523253893995,
+      "grad_norm": 0.8087461590766907,
+      "learning_rate": 8.559817063264661e-06,
+      "loss": 0.7831,
+      "step": 9080
+    },
+    {
+      "epoch": 0.4998073641917552,
+      "grad_norm": 0.7418510317802429,
+      "learning_rate": 8.559512662450452e-06,
+      "loss": 0.801,
+      "step": 9081
+    },
+    {
+      "epoch": 0.49986240299411083,
+      "grad_norm": 0.6793946027755737,
+      "learning_rate": 8.5592082348839e-06,
+      "loss": 0.7329,
+      "step": 9082
+    },
+    {
+      "epoch": 0.4999174417964665,
+      "grad_norm": 0.8197429180145264,
+      "learning_rate": 8.55890378056729e-06,
+      "loss": 0.804,
+      "step": 9083
+    },
+    {
+      "epoch": 0.49997248059882216,
+      "grad_norm": 0.7526460886001587,
+      "learning_rate": 8.558599299502912e-06,
+      "loss": 0.8378,
+      "step": 9084
+    },
+    {
+      "epoch": 0.5000275194011778,
+      "grad_norm": 0.8169133067131042,
+      "learning_rate": 8.558294791693055e-06,
+      "loss": 0.828,
+      "step": 9085
+    },
+    {
+      "epoch": 0.5000825582035335,
+      "grad_norm": 0.8386932015419006,
+      "learning_rate": 8.557990257140007e-06,
+      "loss": 0.7961,
+      "step": 9086
+    },
+    {
+      "epoch": 0.5001375970058891,
+      "grad_norm": 0.7183443903923035,
+      "learning_rate": 8.557685695846057e-06,
+      "loss": 0.6964,
+      "step": 9087
+    },
+    {
+      "epoch": 0.5001926358082448,
+      "grad_norm": 0.77079176902771,
+      "learning_rate": 8.557381107813491e-06,
+      "loss": 0.8222,
+      "step": 9088
+    },
+    {
+      "epoch": 0.5002476746106005,
+      "grad_norm": 0.6519342660903931,
+      "learning_rate": 8.557076493044603e-06,
+      "loss": 0.772,
+      "step": 9089
+    },
+    {
+      "epoch": 0.5003027134129562,
+      "grad_norm": 0.7039975523948669,
+      "learning_rate": 8.556771851541678e-06,
+      "loss": 0.7491,
+      "step": 9090
+    },
+    {
+      "epoch": 0.5003577522153118,
+      "grad_norm": 0.6459039449691772,
+      "learning_rate": 8.556467183307012e-06,
+      "loss": 0.7104,
+      "step": 9091
+    },
+    {
+      "epoch": 0.5004127910176674,
+      "grad_norm": 0.7359183430671692,
+      "learning_rate": 8.556162488342887e-06,
+      "loss": 0.829,
+      "step": 9092
+    },
+    {
+      "epoch": 0.5004678298200231,
+      "grad_norm": 0.7029602527618408,
+      "learning_rate": 8.555857766651599e-06,
+      "loss": 0.8163,
+      "step": 9093
+    },
+    {
+      "epoch": 0.5005228686223788,
+      "grad_norm": 0.6687049865722656,
+      "learning_rate": 8.555553018235435e-06,
+      "loss": 0.7589,
+      "step": 9094
+    },
+    {
+      "epoch": 0.5005779074247344,
+      "grad_norm": 0.7277147173881531,
+      "learning_rate": 8.555248243096686e-06,
+      "loss": 0.8334,
+      "step": 9095
+    },
+    {
+      "epoch": 0.5006329462270901,
+      "grad_norm": 0.6512065529823303,
+      "learning_rate": 8.554943441237642e-06,
+      "loss": 0.7174,
+      "step": 9096
+    },
+    {
+      "epoch": 0.5006879850294458,
+      "grad_norm": 0.725351095199585,
+      "learning_rate": 8.554638612660594e-06,
+      "loss": 0.6514,
+      "step": 9097
+    },
+    {
+      "epoch": 0.5007430238318015,
+      "grad_norm": 0.7983208894729614,
+      "learning_rate": 8.554333757367836e-06,
+      "loss": 0.8385,
+      "step": 9098
+    },
+    {
+      "epoch": 0.500798062634157,
+      "grad_norm": 0.6631388068199158,
+      "learning_rate": 8.554028875361657e-06,
+      "loss": 0.7103,
+      "step": 9099
+    },
+    {
+      "epoch": 0.5008531014365127,
+      "grad_norm": 0.730421245098114,
+      "learning_rate": 8.553723966644347e-06,
+      "loss": 0.8005,
+      "step": 9100
+    },
+    {
+      "epoch": 0.5009081402388684,
+      "grad_norm": 0.7385838627815247,
+      "learning_rate": 8.5534190312182e-06,
+      "loss": 0.7586,
+      "step": 9101
+    },
+    {
+      "epoch": 0.5009631790412241,
+      "grad_norm": 0.712458610534668,
+      "learning_rate": 8.553114069085506e-06,
+      "loss": 0.7587,
+      "step": 9102
+    },
+    {
+      "epoch": 0.5010182178435797,
+      "grad_norm": 0.7393542528152466,
+      "learning_rate": 8.552809080248559e-06,
+      "loss": 0.746,
+      "step": 9103
+    },
+    {
+      "epoch": 0.5010732566459354,
+      "grad_norm": 0.6596370935440063,
+      "learning_rate": 8.552504064709649e-06,
+      "loss": 0.6968,
+      "step": 9104
+    },
+    {
+      "epoch": 0.5011282954482911,
+      "grad_norm": 0.7340545654296875,
+      "learning_rate": 8.552199022471069e-06,
+      "loss": 0.8326,
+      "step": 9105
+    },
+    {
+      "epoch": 0.5011833342506467,
+      "grad_norm": 0.6586140990257263,
+      "learning_rate": 8.55189395353511e-06,
+      "loss": 0.7144,
+      "step": 9106
+    },
+    {
+      "epoch": 0.5012383730530023,
+      "grad_norm": 0.6875959038734436,
+      "learning_rate": 8.551588857904071e-06,
+      "loss": 0.721,
+      "step": 9107
+    },
+    {
+      "epoch": 0.501293411855358,
+      "grad_norm": 0.6754499077796936,
+      "learning_rate": 8.551283735580238e-06,
+      "loss": 0.6771,
+      "step": 9108
+    },
+    {
+      "epoch": 0.5013484506577137,
+      "grad_norm": 0.8027325868606567,
+      "learning_rate": 8.55097858656591e-06,
+      "loss": 0.8196,
+      "step": 9109
+    },
+    {
+      "epoch": 0.5014034894600693,
+      "grad_norm": 0.6992260217666626,
+      "learning_rate": 8.550673410863376e-06,
+      "loss": 0.7923,
+      "step": 9110
+    },
+    {
+      "epoch": 0.501458528262425,
+      "grad_norm": 0.741205632686615,
+      "learning_rate": 8.550368208474928e-06,
+      "loss": 0.7036,
+      "step": 9111
+    },
+    {
+      "epoch": 0.5015135670647807,
+      "grad_norm": 0.6485981345176697,
+      "learning_rate": 8.550062979402866e-06,
+      "loss": 0.6351,
+      "step": 9112
+    },
+    {
+      "epoch": 0.5015686058671364,
+      "grad_norm": 0.6984226703643799,
+      "learning_rate": 8.549757723649481e-06,
+      "loss": 0.7714,
+      "step": 9113
+    },
+    {
+      "epoch": 0.5016236446694919,
+      "grad_norm": 0.7773998975753784,
+      "learning_rate": 8.549452441217067e-06,
+      "loss": 0.8901,
+      "step": 9114
+    },
+    {
+      "epoch": 0.5016786834718476,
+      "grad_norm": 0.6912227272987366,
+      "learning_rate": 8.549147132107918e-06,
+      "loss": 0.7702,
+      "step": 9115
+    },
+    {
+      "epoch": 0.5017337222742033,
+      "grad_norm": 0.6742583513259888,
+      "learning_rate": 8.54884179632433e-06,
+      "loss": 0.7789,
+      "step": 9116
+    },
+    {
+      "epoch": 0.501788761076559,
+      "grad_norm": 0.7896195650100708,
+      "learning_rate": 8.548536433868595e-06,
+      "loss": 0.7358,
+      "step": 9117
+    },
+    {
+      "epoch": 0.5018437998789146,
+      "grad_norm": 0.7112523913383484,
+      "learning_rate": 8.548231044743011e-06,
+      "loss": 0.7286,
+      "step": 9118
+    },
+    {
+      "epoch": 0.5018988386812703,
+      "grad_norm": 0.9162774085998535,
+      "learning_rate": 8.547925628949873e-06,
+      "loss": 0.935,
+      "step": 9119
+    },
+    {
+      "epoch": 0.501953877483626,
+      "grad_norm": 0.6319599747657776,
+      "learning_rate": 8.547620186491477e-06,
+      "loss": 0.625,
+      "step": 9120
+    },
+    {
+      "epoch": 0.5020089162859817,
+      "grad_norm": 0.7074719667434692,
+      "learning_rate": 8.547314717370115e-06,
+      "loss": 0.6614,
+      "step": 9121
+    },
+    {
+      "epoch": 0.5020639550883372,
+      "grad_norm": 0.7417262196540833,
+      "learning_rate": 8.547009221588086e-06,
+      "loss": 0.8476,
+      "step": 9122
+    },
+    {
+      "epoch": 0.5021189938906929,
+      "grad_norm": 0.7057339549064636,
+      "learning_rate": 8.546703699147685e-06,
+      "loss": 0.805,
+      "step": 9123
+    },
+    {
+      "epoch": 0.5021740326930486,
+      "grad_norm": 0.7420887351036072,
+      "learning_rate": 8.546398150051207e-06,
+      "loss": 0.7331,
+      "step": 9124
+    },
+    {
+      "epoch": 0.5022290714954043,
+      "grad_norm": 0.9526195526123047,
+      "learning_rate": 8.546092574300953e-06,
+      "loss": 0.7803,
+      "step": 9125
+    },
+    {
+      "epoch": 0.5022841102977599,
+      "grad_norm": 0.748130202293396,
+      "learning_rate": 8.545786971899214e-06,
+      "loss": 0.7998,
+      "step": 9126
+    },
+    {
+      "epoch": 0.5023391491001156,
+      "grad_norm": 0.7266026139259338,
+      "learning_rate": 8.545481342848289e-06,
+      "loss": 0.8377,
+      "step": 9127
+    },
+    {
+      "epoch": 0.5023941879024713,
+      "grad_norm": 0.6762456893920898,
+      "learning_rate": 8.545175687150478e-06,
+      "loss": 0.7312,
+      "step": 9128
+    },
+    {
+      "epoch": 0.502449226704827,
+      "grad_norm": 0.7011429667472839,
+      "learning_rate": 8.544870004808072e-06,
+      "loss": 0.7666,
+      "step": 9129
+    },
+    {
+      "epoch": 0.5025042655071825,
+      "grad_norm": 0.6652229428291321,
+      "learning_rate": 8.544564295823375e-06,
+      "loss": 0.6904,
+      "step": 9130
+    },
+    {
+      "epoch": 0.5025593043095382,
+      "grad_norm": 0.8333765268325806,
+      "learning_rate": 8.54425856019868e-06,
+      "loss": 0.7318,
+      "step": 9131
+    },
+    {
+      "epoch": 0.5026143431118939,
+      "grad_norm": 0.6827245950698853,
+      "learning_rate": 8.543952797936285e-06,
+      "loss": 0.7692,
+      "step": 9132
+    },
+    {
+      "epoch": 0.5026693819142496,
+      "grad_norm": 0.8744323253631592,
+      "learning_rate": 8.543647009038491e-06,
+      "loss": 0.7316,
+      "step": 9133
+    },
+    {
+      "epoch": 0.5027244207166052,
+      "grad_norm": 0.7024276852607727,
+      "learning_rate": 8.543341193507594e-06,
+      "loss": 0.7008,
+      "step": 9134
+    },
+    {
+      "epoch": 0.5027794595189609,
+      "grad_norm": 0.8786055445671082,
+      "learning_rate": 8.543035351345895e-06,
+      "loss": 0.7054,
+      "step": 9135
+    },
+    {
+      "epoch": 0.5028344983213165,
+      "grad_norm": 0.727924108505249,
+      "learning_rate": 8.54272948255569e-06,
+      "loss": 0.8049,
+      "step": 9136
+    },
+    {
+      "epoch": 0.5028895371236722,
+      "grad_norm": 0.8366256356239319,
+      "learning_rate": 8.542423587139277e-06,
+      "loss": 0.7926,
+      "step": 9137
+    },
+    {
+      "epoch": 0.5029445759260278,
+      "grad_norm": 0.7657913565635681,
+      "learning_rate": 8.542117665098958e-06,
+      "loss": 0.8152,
+      "step": 9138
+    },
+    {
+      "epoch": 0.5029996147283835,
+      "grad_norm": 0.7543498277664185,
+      "learning_rate": 8.54181171643703e-06,
+      "loss": 0.7566,
+      "step": 9139
+    },
+    {
+      "epoch": 0.5030546535307392,
+      "grad_norm": 0.7771349549293518,
+      "learning_rate": 8.541505741155794e-06,
+      "loss": 0.7907,
+      "step": 9140
+    },
+    {
+      "epoch": 0.5031096923330949,
+      "grad_norm": 0.6661877632141113,
+      "learning_rate": 8.541199739257548e-06,
+      "loss": 0.7481,
+      "step": 9141
+    },
+    {
+      "epoch": 0.5031647311354505,
+      "grad_norm": 0.7700417637825012,
+      "learning_rate": 8.540893710744593e-06,
+      "loss": 0.7544,
+      "step": 9142
+    },
+    {
+      "epoch": 0.5032197699378061,
+      "grad_norm": 0.6476640105247498,
+      "learning_rate": 8.54058765561923e-06,
+      "loss": 0.7221,
+      "step": 9143
+    },
+    {
+      "epoch": 0.5032748087401618,
+      "grad_norm": 0.7098944187164307,
+      "learning_rate": 8.540281573883755e-06,
+      "loss": 0.8083,
+      "step": 9144
+    },
+    {
+      "epoch": 0.5033298475425175,
+      "grad_norm": 0.9733545184135437,
+      "learning_rate": 8.539975465540473e-06,
+      "loss": 0.7381,
+      "step": 9145
+    },
+    {
+      "epoch": 0.5033848863448731,
+      "grad_norm": 0.641211986541748,
+      "learning_rate": 8.539669330591685e-06,
+      "loss": 0.7511,
+      "step": 9146
+    },
+    {
+      "epoch": 0.5034399251472288,
+      "grad_norm": 0.626027524471283,
+      "learning_rate": 8.539363169039687e-06,
+      "loss": 0.7321,
+      "step": 9147
+    },
+    {
+      "epoch": 0.5034949639495845,
+      "grad_norm": 0.7627241611480713,
+      "learning_rate": 8.539056980886785e-06,
+      "loss": 0.7269,
+      "step": 9148
+    },
+    {
+      "epoch": 0.5035500027519401,
+      "grad_norm": 0.6711145639419556,
+      "learning_rate": 8.538750766135275e-06,
+      "loss": 0.8179,
+      "step": 9149
+    },
+    {
+      "epoch": 0.5036050415542958,
+      "grad_norm": 0.6981950998306274,
+      "learning_rate": 8.538444524787463e-06,
+      "loss": 0.8095,
+      "step": 9150
+    },
+    {
+      "epoch": 0.5036600803566514,
+      "grad_norm": 0.8869871497154236,
+      "learning_rate": 8.53813825684565e-06,
+      "loss": 0.8549,
+      "step": 9151
+    },
+    {
+      "epoch": 0.5037151191590071,
+      "grad_norm": 0.6461544036865234,
+      "learning_rate": 8.537831962312137e-06,
+      "loss": 0.7388,
+      "step": 9152
+    },
+    {
+      "epoch": 0.5037701579613627,
+      "grad_norm": 0.8279222249984741,
+      "learning_rate": 8.537525641189224e-06,
+      "loss": 0.8609,
+      "step": 9153
+    },
+    {
+      "epoch": 0.5038251967637184,
+      "grad_norm": 0.7117578387260437,
+      "learning_rate": 8.537219293479217e-06,
+      "loss": 0.802,
+      "step": 9154
+    },
+    {
+      "epoch": 0.5038802355660741,
+      "grad_norm": 0.6831860542297363,
+      "learning_rate": 8.536912919184416e-06,
+      "loss": 0.7821,
+      "step": 9155
+    },
+    {
+      "epoch": 0.5039352743684298,
+      "grad_norm": 1.1528539657592773,
+      "learning_rate": 8.536606518307125e-06,
+      "loss": 0.8578,
+      "step": 9156
+    },
+    {
+      "epoch": 0.5039903131707854,
+      "grad_norm": 0.6545060873031616,
+      "learning_rate": 8.536300090849645e-06,
+      "loss": 0.7744,
+      "step": 9157
+    },
+    {
+      "epoch": 0.504045351973141,
+      "grad_norm": 0.7176601886749268,
+      "learning_rate": 8.535993636814281e-06,
+      "loss": 0.8104,
+      "step": 9158
+    },
+    {
+      "epoch": 0.5041003907754967,
+      "grad_norm": 0.8458410501480103,
+      "learning_rate": 8.535687156203334e-06,
+      "loss": 0.8653,
+      "step": 9159
+    },
+    {
+      "epoch": 0.5041554295778524,
+      "grad_norm": 0.7500274777412415,
+      "learning_rate": 8.53538064901911e-06,
+      "loss": 0.8043,
+      "step": 9160
+    },
+    {
+      "epoch": 0.504210468380208,
+      "grad_norm": 0.6982965469360352,
+      "learning_rate": 8.535074115263911e-06,
+      "loss": 0.7564,
+      "step": 9161
+    },
+    {
+      "epoch": 0.5042655071825637,
+      "grad_norm": 0.8344218134880066,
+      "learning_rate": 8.534767554940042e-06,
+      "loss": 0.7575,
+      "step": 9162
+    },
+    {
+      "epoch": 0.5043205459849194,
+      "grad_norm": 0.7527137398719788,
+      "learning_rate": 8.534460968049806e-06,
+      "loss": 0.7757,
+      "step": 9163
+    },
+    {
+      "epoch": 0.5043755847872751,
+      "grad_norm": 0.7136969566345215,
+      "learning_rate": 8.534154354595508e-06,
+      "loss": 0.826,
+      "step": 9164
+    },
+    {
+      "epoch": 0.5044306235896306,
+      "grad_norm": 0.8102819919586182,
+      "learning_rate": 8.533847714579449e-06,
+      "loss": 0.7247,
+      "step": 9165
+    },
+    {
+      "epoch": 0.5044856623919863,
+      "grad_norm": 0.7568309903144836,
+      "learning_rate": 8.53354104800394e-06,
+      "loss": 0.8509,
+      "step": 9166
+    },
+    {
+      "epoch": 0.504540701194342,
+      "grad_norm": 0.7719592452049255,
+      "learning_rate": 8.53323435487128e-06,
+      "loss": 0.8039,
+      "step": 9167
+    },
+    {
+      "epoch": 0.5045957399966977,
+      "grad_norm": 0.7514411807060242,
+      "learning_rate": 8.532927635183778e-06,
+      "loss": 0.8759,
+      "step": 9168
+    },
+    {
+      "epoch": 0.5046507787990533,
+      "grad_norm": 0.9781903028488159,
+      "learning_rate": 8.532620888943736e-06,
+      "loss": 0.8022,
+      "step": 9169
+    },
+    {
+      "epoch": 0.504705817601409,
+      "grad_norm": 0.7713304758071899,
+      "learning_rate": 8.532314116153462e-06,
+      "loss": 0.8372,
+      "step": 9170
+    },
+    {
+      "epoch": 0.5047608564037647,
+      "grad_norm": 0.7519709467887878,
+      "learning_rate": 8.53200731681526e-06,
+      "loss": 0.7374,
+      "step": 9171
+    },
+    {
+      "epoch": 0.5048158952061204,
+      "grad_norm": 0.6923980712890625,
+      "learning_rate": 8.531700490931438e-06,
+      "loss": 0.7511,
+      "step": 9172
+    },
+    {
+      "epoch": 0.5048709340084759,
+      "grad_norm": 0.682357907295227,
+      "learning_rate": 8.5313936385043e-06,
+      "loss": 0.7647,
+      "step": 9173
+    },
+    {
+      "epoch": 0.5049259728108316,
+      "grad_norm": 0.8255659341812134,
+      "learning_rate": 8.531086759536152e-06,
+      "loss": 0.7533,
+      "step": 9174
+    },
+    {
+      "epoch": 0.5049810116131873,
+      "grad_norm": 0.6774975061416626,
+      "learning_rate": 8.530779854029301e-06,
+      "loss": 0.7019,
+      "step": 9175
+    },
+    {
+      "epoch": 0.505036050415543,
+      "grad_norm": 0.7973241209983826,
+      "learning_rate": 8.530472921986053e-06,
+      "loss": 0.7824,
+      "step": 9176
+    },
+    {
+      "epoch": 0.5050910892178986,
+      "grad_norm": 0.8216109275817871,
+      "learning_rate": 8.530165963408716e-06,
+      "loss": 0.8063,
+      "step": 9177
+    },
+    {
+      "epoch": 0.5051461280202543,
+      "grad_norm": 0.7277935743331909,
+      "learning_rate": 8.5298589782996e-06,
+      "loss": 0.7631,
+      "step": 9178
+    },
+    {
+      "epoch": 0.50520116682261,
+      "grad_norm": 0.6647855043411255,
+      "learning_rate": 8.529551966661004e-06,
+      "loss": 0.7462,
+      "step": 9179
+    },
+    {
+      "epoch": 0.5052562056249656,
+      "grad_norm": 0.766272783279419,
+      "learning_rate": 8.529244928495241e-06,
+      "loss": 0.8075,
+      "step": 9180
+    },
+    {
+      "epoch": 0.5053112444273212,
+      "grad_norm": 0.7276293635368347,
+      "learning_rate": 8.52893786380462e-06,
+      "loss": 0.7908,
+      "step": 9181
+    },
+    {
+      "epoch": 0.5053662832296769,
+      "grad_norm": 0.7864169478416443,
+      "learning_rate": 8.528630772591447e-06,
+      "loss": 0.8082,
+      "step": 9182
+    },
+    {
+      "epoch": 0.5054213220320326,
+      "grad_norm": 0.9106804132461548,
+      "learning_rate": 8.528323654858028e-06,
+      "loss": 0.8989,
+      "step": 9183
+    },
+    {
+      "epoch": 0.5054763608343883,
+      "grad_norm": 0.7288523316383362,
+      "learning_rate": 8.52801651060667e-06,
+      "loss": 0.7972,
+      "step": 9184
+    },
+    {
+      "epoch": 0.5055313996367439,
+      "grad_norm": 0.7149643301963806,
+      "learning_rate": 8.527709339839689e-06,
+      "loss": 0.8191,
+      "step": 9185
+    },
+    {
+      "epoch": 0.5055864384390996,
+      "grad_norm": 0.6661714911460876,
+      "learning_rate": 8.527402142559388e-06,
+      "loss": 0.6596,
+      "step": 9186
+    },
+    {
+      "epoch": 0.5056414772414553,
+      "grad_norm": 0.7071447372436523,
+      "learning_rate": 8.527094918768076e-06,
+      "loss": 0.7633,
+      "step": 9187
+    },
+    {
+      "epoch": 0.5056965160438109,
+      "grad_norm": 0.7314093112945557,
+      "learning_rate": 8.526787668468064e-06,
+      "loss": 0.7815,
+      "step": 9188
+    },
+    {
+      "epoch": 0.5057515548461665,
+      "grad_norm": 0.8200539946556091,
+      "learning_rate": 8.526480391661657e-06,
+      "loss": 0.8376,
+      "step": 9189
+    },
+    {
+      "epoch": 0.5058065936485222,
+      "grad_norm": 0.7422435283660889,
+      "learning_rate": 8.52617308835117e-06,
+      "loss": 0.8783,
+      "step": 9190
+    },
+    {
+      "epoch": 0.5058616324508779,
+      "grad_norm": 0.7845084071159363,
+      "learning_rate": 8.525865758538909e-06,
+      "loss": 0.8005,
+      "step": 9191
+    },
+    {
+      "epoch": 0.5059166712532335,
+      "grad_norm": 0.6854296922683716,
+      "learning_rate": 8.525558402227185e-06,
+      "loss": 0.8118,
+      "step": 9192
+    },
+    {
+      "epoch": 0.5059717100555892,
+      "grad_norm": 0.6805297136306763,
+      "learning_rate": 8.525251019418309e-06,
+      "loss": 0.6765,
+      "step": 9193
+    },
+    {
+      "epoch": 0.5060267488579449,
+      "grad_norm": 0.7194867134094238,
+      "learning_rate": 8.524943610114587e-06,
+      "loss": 0.6752,
+      "step": 9194
+    },
+    {
+      "epoch": 0.5060817876603005,
+      "grad_norm": 0.6935137510299683,
+      "learning_rate": 8.524636174318335e-06,
+      "loss": 0.7122,
+      "step": 9195
+    },
+    {
+      "epoch": 0.5061368264626561,
+      "grad_norm": 0.8652825951576233,
+      "learning_rate": 8.52432871203186e-06,
+      "loss": 0.7725,
+      "step": 9196
+    },
+    {
+      "epoch": 0.5061918652650118,
+      "grad_norm": 0.9104461669921875,
+      "learning_rate": 8.524021223257472e-06,
+      "loss": 0.8589,
+      "step": 9197
+    },
+    {
+      "epoch": 0.5062469040673675,
+      "grad_norm": 0.7680580019950867,
+      "learning_rate": 8.523713707997486e-06,
+      "loss": 0.842,
+      "step": 9198
+    },
+    {
+      "epoch": 0.5063019428697232,
+      "grad_norm": 0.7324872612953186,
+      "learning_rate": 8.52340616625421e-06,
+      "loss": 0.802,
+      "step": 9199
+    },
+    {
+      "epoch": 0.5063569816720788,
+      "grad_norm": 0.8812359571456909,
+      "learning_rate": 8.523098598029958e-06,
+      "loss": 0.8286,
+      "step": 9200
+    },
+    {
+      "epoch": 0.5064120204744345,
+      "grad_norm": 0.6992496848106384,
+      "learning_rate": 8.522791003327038e-06,
+      "loss": 0.811,
+      "step": 9201
+    },
+    {
+      "epoch": 0.5064670592767901,
+      "grad_norm": 0.8191942572593689,
+      "learning_rate": 8.522483382147766e-06,
+      "loss": 0.7192,
+      "step": 9202
+    },
+    {
+      "epoch": 0.5065220980791458,
+      "grad_norm": 0.9354501366615295,
+      "learning_rate": 8.522175734494452e-06,
+      "loss": 0.7424,
+      "step": 9203
+    },
+    {
+      "epoch": 0.5065771368815014,
+      "grad_norm": 0.6481999754905701,
+      "learning_rate": 8.521868060369405e-06,
+      "loss": 0.6385,
+      "step": 9204
+    },
+    {
+      "epoch": 0.5066321756838571,
+      "grad_norm": 0.7158499360084534,
+      "learning_rate": 8.521560359774943e-06,
+      "loss": 0.6116,
+      "step": 9205
+    },
+    {
+      "epoch": 0.5066872144862128,
+      "grad_norm": 0.8738408088684082,
+      "learning_rate": 8.521252632713376e-06,
+      "loss": 0.894,
+      "step": 9206
+    },
+    {
+      "epoch": 0.5067422532885685,
+      "grad_norm": 0.7037062644958496,
+      "learning_rate": 8.520944879187015e-06,
+      "loss": 0.6958,
+      "step": 9207
+    },
+    {
+      "epoch": 0.5067972920909241,
+      "grad_norm": 0.7205594778060913,
+      "learning_rate": 8.520637099198175e-06,
+      "loss": 0.7188,
+      "step": 9208
+    },
+    {
+      "epoch": 0.5068523308932797,
+      "grad_norm": 0.6761966347694397,
+      "learning_rate": 8.520329292749169e-06,
+      "loss": 0.7669,
+      "step": 9209
+    },
+    {
+      "epoch": 0.5069073696956354,
+      "grad_norm": 0.682556688785553,
+      "learning_rate": 8.520021459842312e-06,
+      "loss": 0.7745,
+      "step": 9210
+    },
+    {
+      "epoch": 0.5069624084979911,
+      "grad_norm": 0.6687794923782349,
+      "learning_rate": 8.519713600479913e-06,
+      "loss": 0.7814,
+      "step": 9211
+    },
+    {
+      "epoch": 0.5070174473003467,
+      "grad_norm": 0.6391967535018921,
+      "learning_rate": 8.51940571466429e-06,
+      "loss": 0.7331,
+      "step": 9212
+    },
+    {
+      "epoch": 0.5070724861027024,
+      "grad_norm": 0.8420151472091675,
+      "learning_rate": 8.519097802397758e-06,
+      "loss": 0.8257,
+      "step": 9213
+    },
+    {
+      "epoch": 0.5071275249050581,
+      "grad_norm": 0.692787230014801,
+      "learning_rate": 8.518789863682625e-06,
+      "loss": 0.7179,
+      "step": 9214
+    },
+    {
+      "epoch": 0.5071825637074138,
+      "grad_norm": 0.6874318718910217,
+      "learning_rate": 8.518481898521213e-06,
+      "loss": 0.6847,
+      "step": 9215
+    },
+    {
+      "epoch": 0.5072376025097693,
+      "grad_norm": 0.8107750415802002,
+      "learning_rate": 8.518173906915832e-06,
+      "loss": 0.8459,
+      "step": 9216
+    },
+    {
+      "epoch": 0.507292641312125,
+      "grad_norm": 0.7952812910079956,
+      "learning_rate": 8.517865888868797e-06,
+      "loss": 0.8503,
+      "step": 9217
+    },
+    {
+      "epoch": 0.5073476801144807,
+      "grad_norm": 0.6926921606063843,
+      "learning_rate": 8.517557844382424e-06,
+      "loss": 0.6713,
+      "step": 9218
+    },
+    {
+      "epoch": 0.5074027189168364,
+      "grad_norm": 0.8203585147857666,
+      "learning_rate": 8.517249773459026e-06,
+      "loss": 0.8483,
+      "step": 9219
+    },
+    {
+      "epoch": 0.507457757719192,
+      "grad_norm": 0.6788125038146973,
+      "learning_rate": 8.516941676100923e-06,
+      "loss": 0.7521,
+      "step": 9220
+    },
+    {
+      "epoch": 0.5075127965215477,
+      "grad_norm": 0.6439838409423828,
+      "learning_rate": 8.516633552310426e-06,
+      "loss": 0.7359,
+      "step": 9221
+    },
+    {
+      "epoch": 0.5075678353239034,
+      "grad_norm": 0.6872217655181885,
+      "learning_rate": 8.516325402089854e-06,
+      "loss": 0.73,
+      "step": 9222
+    },
+    {
+      "epoch": 0.5076228741262591,
+      "grad_norm": 0.6695985794067383,
+      "learning_rate": 8.51601722544152e-06,
+      "loss": 0.7519,
+      "step": 9223
+    },
+    {
+      "epoch": 0.5076779129286146,
+      "grad_norm": 0.7779402136802673,
+      "learning_rate": 8.515709022367741e-06,
+      "loss": 0.7325,
+      "step": 9224
+    },
+    {
+      "epoch": 0.5077329517309703,
+      "grad_norm": 0.9289746284484863,
+      "learning_rate": 8.515400792870836e-06,
+      "loss": 0.7839,
+      "step": 9225
+    },
+    {
+      "epoch": 0.507787990533326,
+      "grad_norm": 0.6949248313903809,
+      "learning_rate": 8.51509253695312e-06,
+      "loss": 0.7363,
+      "step": 9226
+    },
+    {
+      "epoch": 0.5078430293356817,
+      "grad_norm": 0.6463130116462708,
+      "learning_rate": 8.514784254616908e-06,
+      "loss": 0.7607,
+      "step": 9227
+    },
+    {
+      "epoch": 0.5078980681380373,
+      "grad_norm": 0.7332046031951904,
+      "learning_rate": 8.514475945864519e-06,
+      "loss": 0.6833,
+      "step": 9228
+    },
+    {
+      "epoch": 0.507953106940393,
+      "grad_norm": 0.8674100637435913,
+      "learning_rate": 8.51416761069827e-06,
+      "loss": 0.669,
+      "step": 9229
+    },
+    {
+      "epoch": 0.5080081457427487,
+      "grad_norm": 0.8073185682296753,
+      "learning_rate": 8.513859249120477e-06,
+      "loss": 0.7215,
+      "step": 9230
+    },
+    {
+      "epoch": 0.5080631845451044,
+      "grad_norm": 0.674117386341095,
+      "learning_rate": 8.51355086113346e-06,
+      "loss": 0.7813,
+      "step": 9231
+    },
+    {
+      "epoch": 0.5081182233474599,
+      "grad_norm": 0.8564596176147461,
+      "learning_rate": 8.513242446739534e-06,
+      "loss": 0.7393,
+      "step": 9232
+    },
+    {
+      "epoch": 0.5081732621498156,
+      "grad_norm": 0.684637188911438,
+      "learning_rate": 8.512934005941015e-06,
+      "loss": 0.781,
+      "step": 9233
+    },
+    {
+      "epoch": 0.5082283009521713,
+      "grad_norm": 0.816123902797699,
+      "learning_rate": 8.51262553874023e-06,
+      "loss": 0.8597,
+      "step": 9234
+    },
+    {
+      "epoch": 0.5082833397545269,
+      "grad_norm": 0.6582320332527161,
+      "learning_rate": 8.512317045139488e-06,
+      "loss": 0.6654,
+      "step": 9235
+    },
+    {
+      "epoch": 0.5083383785568826,
+      "grad_norm": 1.0153518915176392,
+      "learning_rate": 8.512008525141113e-06,
+      "loss": 0.7946,
+      "step": 9236
+    },
+    {
+      "epoch": 0.5083934173592383,
+      "grad_norm": 0.7455416917800903,
+      "learning_rate": 8.511699978747422e-06,
+      "loss": 0.8365,
+      "step": 9237
+    },
+    {
+      "epoch": 0.508448456161594,
+      "grad_norm": 0.6498221755027771,
+      "learning_rate": 8.511391405960733e-06,
+      "loss": 0.7252,
+      "step": 9238
+    },
+    {
+      "epoch": 0.5085034949639495,
+      "grad_norm": 0.6856792569160461,
+      "learning_rate": 8.511082806783368e-06,
+      "loss": 0.7282,
+      "step": 9239
+    },
+    {
+      "epoch": 0.5085585337663052,
+      "grad_norm": 0.6930065751075745,
+      "learning_rate": 8.510774181217643e-06,
+      "loss": 0.7404,
+      "step": 9240
+    },
+    {
+      "epoch": 0.5086135725686609,
+      "grad_norm": 0.6953150033950806,
+      "learning_rate": 8.51046552926588e-06,
+      "loss": 0.7684,
+      "step": 9241
+    },
+    {
+      "epoch": 0.5086686113710166,
+      "grad_norm": 0.7307711839675903,
+      "learning_rate": 8.510156850930395e-06,
+      "loss": 0.7557,
+      "step": 9242
+    },
+    {
+      "epoch": 0.5087236501733722,
+      "grad_norm": 0.7296478152275085,
+      "learning_rate": 8.509848146213513e-06,
+      "loss": 0.7469,
+      "step": 9243
+    },
+    {
+      "epoch": 0.5087786889757279,
+      "grad_norm": 0.7035672664642334,
+      "learning_rate": 8.509539415117553e-06,
+      "loss": 0.7151,
+      "step": 9244
+    },
+    {
+      "epoch": 0.5088337277780836,
+      "grad_norm": 0.7818698883056641,
+      "learning_rate": 8.509230657644832e-06,
+      "loss": 0.7134,
+      "step": 9245
+    },
+    {
+      "epoch": 0.5088887665804392,
+      "grad_norm": 0.7503119111061096,
+      "learning_rate": 8.508921873797674e-06,
+      "loss": 0.7028,
+      "step": 9246
+    },
+    {
+      "epoch": 0.5089438053827948,
+      "grad_norm": 0.7733498215675354,
+      "learning_rate": 8.508613063578397e-06,
+      "loss": 0.8159,
+      "step": 9247
+    },
+    {
+      "epoch": 0.5089988441851505,
+      "grad_norm": 0.9236353635787964,
+      "learning_rate": 8.508304226989326e-06,
+      "loss": 0.8013,
+      "step": 9248
+    },
+    {
+      "epoch": 0.5090538829875062,
+      "grad_norm": 0.6567198634147644,
+      "learning_rate": 8.507995364032777e-06,
+      "loss": 0.8285,
+      "step": 9249
+    },
+    {
+      "epoch": 0.5091089217898619,
+      "grad_norm": 0.6555445790290833,
+      "learning_rate": 8.507686474711074e-06,
+      "loss": 0.6917,
+      "step": 9250
+    },
+    {
+      "epoch": 0.5091639605922175,
+      "grad_norm": 0.8505375385284424,
+      "learning_rate": 8.507377559026539e-06,
+      "loss": 0.824,
+      "step": 9251
+    },
+    {
+      "epoch": 0.5092189993945732,
+      "grad_norm": 0.703413188457489,
+      "learning_rate": 8.507068616981493e-06,
+      "loss": 0.7162,
+      "step": 9252
+    },
+    {
+      "epoch": 0.5092740381969288,
+      "grad_norm": 0.7257823944091797,
+      "learning_rate": 8.50675964857826e-06,
+      "loss": 0.8031,
+      "step": 9253
+    },
+    {
+      "epoch": 0.5093290769992845,
+      "grad_norm": 0.6861198544502258,
+      "learning_rate": 8.506450653819159e-06,
+      "loss": 0.7724,
+      "step": 9254
+    },
+    {
+      "epoch": 0.5093841158016401,
+      "grad_norm": 0.7733107209205627,
+      "learning_rate": 8.506141632706512e-06,
+      "loss": 0.7834,
+      "step": 9255
+    },
+    {
+      "epoch": 0.5094391546039958,
+      "grad_norm": 0.7472217082977295,
+      "learning_rate": 8.505832585242644e-06,
+      "loss": 0.7594,
+      "step": 9256
+    },
+    {
+      "epoch": 0.5094941934063515,
+      "grad_norm": 0.6273325085639954,
+      "learning_rate": 8.505523511429876e-06,
+      "loss": 0.6798,
+      "step": 9257
+    },
+    {
+      "epoch": 0.5095492322087072,
+      "grad_norm": 0.7366517186164856,
+      "learning_rate": 8.505214411270533e-06,
+      "loss": 0.7916,
+      "step": 9258
+    },
+    {
+      "epoch": 0.5096042710110628,
+      "grad_norm": 0.6654453873634338,
+      "learning_rate": 8.504905284766936e-06,
+      "loss": 0.7228,
+      "step": 9259
+    },
+    {
+      "epoch": 0.5096593098134184,
+      "grad_norm": 0.7926275134086609,
+      "learning_rate": 8.50459613192141e-06,
+      "loss": 0.8303,
+      "step": 9260
+    },
+    {
+      "epoch": 0.5097143486157741,
+      "grad_norm": 0.7256377935409546,
+      "learning_rate": 8.504286952736277e-06,
+      "loss": 0.7977,
+      "step": 9261
+    },
+    {
+      "epoch": 0.5097693874181298,
+      "grad_norm": 0.7333946824073792,
+      "learning_rate": 8.50397774721386e-06,
+      "loss": 0.7978,
+      "step": 9262
+    },
+    {
+      "epoch": 0.5098244262204854,
+      "grad_norm": 0.6102882623672485,
+      "learning_rate": 8.503668515356485e-06,
+      "loss": 0.6386,
+      "step": 9263
+    },
+    {
+      "epoch": 0.5098794650228411,
+      "grad_norm": 0.7939823865890503,
+      "learning_rate": 8.503359257166477e-06,
+      "loss": 0.7328,
+      "step": 9264
+    },
+    {
+      "epoch": 0.5099345038251968,
+      "grad_norm": 0.7245013117790222,
+      "learning_rate": 8.503049972646157e-06,
+      "loss": 0.795,
+      "step": 9265
+    },
+    {
+      "epoch": 0.5099895426275525,
+      "grad_norm": 0.6722108125686646,
+      "learning_rate": 8.502740661797852e-06,
+      "loss": 0.7062,
+      "step": 9266
+    },
+    {
+      "epoch": 0.510044581429908,
+      "grad_norm": 0.6759012341499329,
+      "learning_rate": 8.502431324623884e-06,
+      "loss": 0.7427,
+      "step": 9267
+    },
+    {
+      "epoch": 0.5100996202322637,
+      "grad_norm": 0.6448835730552673,
+      "learning_rate": 8.502121961126581e-06,
+      "loss": 0.7381,
+      "step": 9268
+    },
+    {
+      "epoch": 0.5101546590346194,
+      "grad_norm": 0.6437426209449768,
+      "learning_rate": 8.501812571308266e-06,
+      "loss": 0.6733,
+      "step": 9269
+    },
+    {
+      "epoch": 0.5102096978369751,
+      "grad_norm": 0.6879013776779175,
+      "learning_rate": 8.501503155171267e-06,
+      "loss": 0.7227,
+      "step": 9270
+    },
+    {
+      "epoch": 0.5102647366393307,
+      "grad_norm": 0.6628512740135193,
+      "learning_rate": 8.501193712717906e-06,
+      "loss": 0.7151,
+      "step": 9271
+    },
+    {
+      "epoch": 0.5103197754416864,
+      "grad_norm": 0.7653747797012329,
+      "learning_rate": 8.500884243950511e-06,
+      "loss": 0.8189,
+      "step": 9272
+    },
+    {
+      "epoch": 0.5103748142440421,
+      "grad_norm": 0.7180060148239136,
+      "learning_rate": 8.500574748871407e-06,
+      "loss": 0.7633,
+      "step": 9273
+    },
+    {
+      "epoch": 0.5104298530463978,
+      "grad_norm": 0.7045086622238159,
+      "learning_rate": 8.50026522748292e-06,
+      "loss": 0.746,
+      "step": 9274
+    },
+    {
+      "epoch": 0.5104848918487533,
+      "grad_norm": 0.6224614381790161,
+      "learning_rate": 8.499955679787376e-06,
+      "loss": 0.7436,
+      "step": 9275
+    },
+    {
+      "epoch": 0.510539930651109,
+      "grad_norm": 0.6716495156288147,
+      "learning_rate": 8.499646105787103e-06,
+      "loss": 0.8006,
+      "step": 9276
+    },
+    {
+      "epoch": 0.5105949694534647,
+      "grad_norm": 0.83705735206604,
+      "learning_rate": 8.499336505484426e-06,
+      "loss": 0.886,
+      "step": 9277
+    },
+    {
+      "epoch": 0.5106500082558203,
+      "grad_norm": 0.7942199110984802,
+      "learning_rate": 8.499026878881673e-06,
+      "loss": 0.7709,
+      "step": 9278
+    },
+    {
+      "epoch": 0.510705047058176,
+      "grad_norm": 0.7500330209732056,
+      "learning_rate": 8.49871722598117e-06,
+      "loss": 0.7737,
+      "step": 9279
+    },
+    {
+      "epoch": 0.5107600858605317,
+      "grad_norm": 0.7283433675765991,
+      "learning_rate": 8.498407546785245e-06,
+      "loss": 0.8345,
+      "step": 9280
+    },
+    {
+      "epoch": 0.5108151246628874,
+      "grad_norm": 0.6970989108085632,
+      "learning_rate": 8.498097841296224e-06,
+      "loss": 0.7451,
+      "step": 9281
+    },
+    {
+      "epoch": 0.5108701634652429,
+      "grad_norm": 0.8338573575019836,
+      "learning_rate": 8.497788109516438e-06,
+      "loss": 0.8198,
+      "step": 9282
+    },
+    {
+      "epoch": 0.5109252022675986,
+      "grad_norm": 0.6544861197471619,
+      "learning_rate": 8.497478351448213e-06,
+      "loss": 0.7549,
+      "step": 9283
+    },
+    {
+      "epoch": 0.5109802410699543,
+      "grad_norm": 0.6627360582351685,
+      "learning_rate": 8.497168567093876e-06,
+      "loss": 0.7136,
+      "step": 9284
+    },
+    {
+      "epoch": 0.51103527987231,
+      "grad_norm": 0.7176669239997864,
+      "learning_rate": 8.496858756455755e-06,
+      "loss": 0.766,
+      "step": 9285
+    },
+    {
+      "epoch": 0.5110903186746656,
+      "grad_norm": 0.8260897397994995,
+      "learning_rate": 8.496548919536183e-06,
+      "loss": 0.8167,
+      "step": 9286
+    },
+    {
+      "epoch": 0.5111453574770213,
+      "grad_norm": 0.7077773809432983,
+      "learning_rate": 8.496239056337483e-06,
+      "loss": 0.776,
+      "step": 9287
+    },
+    {
+      "epoch": 0.511200396279377,
+      "grad_norm": 0.7609447836875916,
+      "learning_rate": 8.495929166861988e-06,
+      "loss": 0.7339,
+      "step": 9288
+    },
+    {
+      "epoch": 0.5112554350817327,
+      "grad_norm": 0.6896487474441528,
+      "learning_rate": 8.495619251112022e-06,
+      "loss": 0.7639,
+      "step": 9289
+    },
+    {
+      "epoch": 0.5113104738840882,
+      "grad_norm": 0.6946871280670166,
+      "learning_rate": 8.495309309089918e-06,
+      "loss": 0.8242,
+      "step": 9290
+    },
+    {
+      "epoch": 0.5113655126864439,
+      "grad_norm": 0.79847252368927,
+      "learning_rate": 8.494999340798007e-06,
+      "loss": 0.8226,
+      "step": 9291
+    },
+    {
+      "epoch": 0.5114205514887996,
+      "grad_norm": 0.7845447063446045,
+      "learning_rate": 8.494689346238615e-06,
+      "loss": 0.8593,
+      "step": 9292
+    },
+    {
+      "epoch": 0.5114755902911553,
+      "grad_norm": 1.1577119827270508,
+      "learning_rate": 8.494379325414074e-06,
+      "loss": 0.746,
+      "step": 9293
+    },
+    {
+      "epoch": 0.5115306290935109,
+      "grad_norm": 0.6720938682556152,
+      "learning_rate": 8.494069278326713e-06,
+      "loss": 0.6768,
+      "step": 9294
+    },
+    {
+      "epoch": 0.5115856678958666,
+      "grad_norm": 0.7389395833015442,
+      "learning_rate": 8.493759204978862e-06,
+      "loss": 0.8126,
+      "step": 9295
+    },
+    {
+      "epoch": 0.5116407066982223,
+      "grad_norm": 0.7629536986351013,
+      "learning_rate": 8.493449105372853e-06,
+      "loss": 0.7107,
+      "step": 9296
+    },
+    {
+      "epoch": 0.511695745500578,
+      "grad_norm": 0.7339474558830261,
+      "learning_rate": 8.493138979511015e-06,
+      "loss": 0.8144,
+      "step": 9297
+    },
+    {
+      "epoch": 0.5117507843029335,
+      "grad_norm": 0.7222825288772583,
+      "learning_rate": 8.49282882739568e-06,
+      "loss": 0.7512,
+      "step": 9298
+    },
+    {
+      "epoch": 0.5118058231052892,
+      "grad_norm": 0.676659107208252,
+      "learning_rate": 8.49251864902918e-06,
+      "loss": 0.6515,
+      "step": 9299
+    },
+    {
+      "epoch": 0.5118608619076449,
+      "grad_norm": 0.6336323618888855,
+      "learning_rate": 8.492208444413844e-06,
+      "loss": 0.719,
+      "step": 9300
+    },
+    {
+      "epoch": 0.5119159007100006,
+      "grad_norm": 0.701543927192688,
+      "learning_rate": 8.491898213552e-06,
+      "loss": 0.728,
+      "step": 9301
+    },
+    {
+      "epoch": 0.5119709395123562,
+      "grad_norm": 0.6809069514274597,
+      "learning_rate": 8.491587956445988e-06,
+      "loss": 0.8844,
+      "step": 9302
+    },
+    {
+      "epoch": 0.5120259783147119,
+      "grad_norm": 0.8046489357948303,
+      "learning_rate": 8.491277673098135e-06,
+      "loss": 0.817,
+      "step": 9303
+    },
+    {
+      "epoch": 0.5120810171170675,
+      "grad_norm": 0.8630616068840027,
+      "learning_rate": 8.490967363510774e-06,
+      "loss": 0.7745,
+      "step": 9304
+    },
+    {
+      "epoch": 0.5121360559194232,
+      "grad_norm": 0.7457678914070129,
+      "learning_rate": 8.490657027686235e-06,
+      "loss": 0.7956,
+      "step": 9305
+    },
+    {
+      "epoch": 0.5121910947217788,
+      "grad_norm": 0.6383466124534607,
+      "learning_rate": 8.490346665626854e-06,
+      "loss": 0.8046,
+      "step": 9306
+    },
+    {
+      "epoch": 0.5122461335241345,
+      "grad_norm": 0.7658202052116394,
+      "learning_rate": 8.49003627733496e-06,
+      "loss": 0.7905,
+      "step": 9307
+    },
+    {
+      "epoch": 0.5123011723264902,
+      "grad_norm": 0.6793283224105835,
+      "learning_rate": 8.48972586281289e-06,
+      "loss": 0.6646,
+      "step": 9308
+    },
+    {
+      "epoch": 0.5123562111288459,
+      "grad_norm": 0.7345246076583862,
+      "learning_rate": 8.489415422062972e-06,
+      "loss": 0.788,
+      "step": 9309
+    },
+    {
+      "epoch": 0.5124112499312015,
+      "grad_norm": 0.6665463447570801,
+      "learning_rate": 8.489104955087542e-06,
+      "loss": 0.706,
+      "step": 9310
+    },
+    {
+      "epoch": 0.5124662887335572,
+      "grad_norm": 0.7895458936691284,
+      "learning_rate": 8.488794461888934e-06,
+      "loss": 0.7464,
+      "step": 9311
+    },
+    {
+      "epoch": 0.5125213275359128,
+      "grad_norm": 0.7375221252441406,
+      "learning_rate": 8.488483942469481e-06,
+      "loss": 0.8029,
+      "step": 9312
+    },
+    {
+      "epoch": 0.5125763663382685,
+      "grad_norm": 0.792348325252533,
+      "learning_rate": 8.488173396831514e-06,
+      "loss": 0.7324,
+      "step": 9313
+    },
+    {
+      "epoch": 0.5126314051406241,
+      "grad_norm": 0.6500192880630493,
+      "learning_rate": 8.487862824977373e-06,
+      "loss": 0.7331,
+      "step": 9314
+    },
+    {
+      "epoch": 0.5126864439429798,
+      "grad_norm": 0.6607314348220825,
+      "learning_rate": 8.487552226909386e-06,
+      "loss": 0.7782,
+      "step": 9315
+    },
+    {
+      "epoch": 0.5127414827453355,
+      "grad_norm": 0.8261791467666626,
+      "learning_rate": 8.487241602629892e-06,
+      "loss": 0.8036,
+      "step": 9316
+    },
+    {
+      "epoch": 0.5127965215476912,
+      "grad_norm": 0.8301663994789124,
+      "learning_rate": 8.486930952141222e-06,
+      "loss": 0.7928,
+      "step": 9317
+    },
+    {
+      "epoch": 0.5128515603500468,
+      "grad_norm": 0.6957940459251404,
+      "learning_rate": 8.486620275445713e-06,
+      "loss": 0.7359,
+      "step": 9318
+    },
+    {
+      "epoch": 0.5129065991524024,
+      "grad_norm": 0.7562606334686279,
+      "learning_rate": 8.4863095725457e-06,
+      "loss": 0.7546,
+      "step": 9319
+    },
+    {
+      "epoch": 0.5129616379547581,
+      "grad_norm": 0.795886218547821,
+      "learning_rate": 8.485998843443517e-06,
+      "loss": 0.7558,
+      "step": 9320
+    },
+    {
+      "epoch": 0.5130166767571137,
+      "grad_norm": 0.6558147072792053,
+      "learning_rate": 8.4856880881415e-06,
+      "loss": 0.6832,
+      "step": 9321
+    },
+    {
+      "epoch": 0.5130717155594694,
+      "grad_norm": 0.7300151586532593,
+      "learning_rate": 8.485377306641984e-06,
+      "loss": 0.8018,
+      "step": 9322
+    },
+    {
+      "epoch": 0.5131267543618251,
+      "grad_norm": 0.7114105224609375,
+      "learning_rate": 8.485066498947305e-06,
+      "loss": 0.7374,
+      "step": 9323
+    },
+    {
+      "epoch": 0.5131817931641808,
+      "grad_norm": 0.7061085104942322,
+      "learning_rate": 8.484755665059798e-06,
+      "loss": 0.7905,
+      "step": 9324
+    },
+    {
+      "epoch": 0.5132368319665364,
+      "grad_norm": 0.8481647968292236,
+      "learning_rate": 8.484444804981802e-06,
+      "loss": 0.8518,
+      "step": 9325
+    },
+    {
+      "epoch": 0.513291870768892,
+      "grad_norm": 0.7583557367324829,
+      "learning_rate": 8.48413391871565e-06,
+      "loss": 0.8328,
+      "step": 9326
+    },
+    {
+      "epoch": 0.5133469095712477,
+      "grad_norm": 0.7381925582885742,
+      "learning_rate": 8.483823006263683e-06,
+      "loss": 0.76,
+      "step": 9327
+    },
+    {
+      "epoch": 0.5134019483736034,
+      "grad_norm": 0.8037852644920349,
+      "learning_rate": 8.483512067628232e-06,
+      "loss": 0.711,
+      "step": 9328
+    },
+    {
+      "epoch": 0.513456987175959,
+      "grad_norm": 0.6682618260383606,
+      "learning_rate": 8.483201102811637e-06,
+      "loss": 0.7479,
+      "step": 9329
+    },
+    {
+      "epoch": 0.5135120259783147,
+      "grad_norm": 0.662234365940094,
+      "learning_rate": 8.482890111816237e-06,
+      "loss": 0.7701,
+      "step": 9330
+    },
+    {
+      "epoch": 0.5135670647806704,
+      "grad_norm": 0.7081482410430908,
+      "learning_rate": 8.482579094644365e-06,
+      "loss": 0.8255,
+      "step": 9331
+    },
+    {
+      "epoch": 0.5136221035830261,
+      "grad_norm": 0.9659954905509949,
+      "learning_rate": 8.482268051298364e-06,
+      "loss": 0.8742,
+      "step": 9332
+    },
+    {
+      "epoch": 0.5136771423853816,
+      "grad_norm": 0.7837772369384766,
+      "learning_rate": 8.481956981780564e-06,
+      "loss": 0.7692,
+      "step": 9333
+    },
+    {
+      "epoch": 0.5137321811877373,
+      "grad_norm": 0.681918203830719,
+      "learning_rate": 8.481645886093311e-06,
+      "loss": 0.6952,
+      "step": 9334
+    },
+    {
+      "epoch": 0.513787219990093,
+      "grad_norm": 0.7253187894821167,
+      "learning_rate": 8.481334764238937e-06,
+      "loss": 0.7074,
+      "step": 9335
+    },
+    {
+      "epoch": 0.5138422587924487,
+      "grad_norm": 0.8845877051353455,
+      "learning_rate": 8.481023616219783e-06,
+      "loss": 0.675,
+      "step": 9336
+    },
+    {
+      "epoch": 0.5138972975948043,
+      "grad_norm": 0.6569344401359558,
+      "learning_rate": 8.480712442038188e-06,
+      "loss": 0.7181,
+      "step": 9337
+    },
+    {
+      "epoch": 0.51395233639716,
+      "grad_norm": 0.7372813820838928,
+      "learning_rate": 8.480401241696491e-06,
+      "loss": 0.8137,
+      "step": 9338
+    },
+    {
+      "epoch": 0.5140073751995157,
+      "grad_norm": 0.843099057674408,
+      "learning_rate": 8.48009001519703e-06,
+      "loss": 0.7648,
+      "step": 9339
+    },
+    {
+      "epoch": 0.5140624140018714,
+      "grad_norm": 0.7762032747268677,
+      "learning_rate": 8.479778762542142e-06,
+      "loss": 0.7805,
+      "step": 9340
+    },
+    {
+      "epoch": 0.5141174528042269,
+      "grad_norm": 0.739086925983429,
+      "learning_rate": 8.479467483734169e-06,
+      "loss": 0.7125,
+      "step": 9341
+    },
+    {
+      "epoch": 0.5141724916065826,
+      "grad_norm": 0.7351683974266052,
+      "learning_rate": 8.479156178775451e-06,
+      "loss": 0.7855,
+      "step": 9342
+    },
+    {
+      "epoch": 0.5142275304089383,
+      "grad_norm": 0.7601314187049866,
+      "learning_rate": 8.478844847668325e-06,
+      "loss": 0.8349,
+      "step": 9343
+    },
+    {
+      "epoch": 0.514282569211294,
+      "grad_norm": 0.6841638684272766,
+      "learning_rate": 8.478533490415133e-06,
+      "loss": 0.7986,
+      "step": 9344
+    },
+    {
+      "epoch": 0.5143376080136496,
+      "grad_norm": 0.6734872460365295,
+      "learning_rate": 8.478222107018213e-06,
+      "loss": 0.6941,
+      "step": 9345
+    },
+    {
+      "epoch": 0.5143926468160053,
+      "grad_norm": 0.801930844783783,
+      "learning_rate": 8.47791069747991e-06,
+      "loss": 0.8537,
+      "step": 9346
+    },
+    {
+      "epoch": 0.514447685618361,
+      "grad_norm": 0.6960629224777222,
+      "learning_rate": 8.477599261802558e-06,
+      "loss": 0.6629,
+      "step": 9347
+    },
+    {
+      "epoch": 0.5145027244207167,
+      "grad_norm": 0.7791358232498169,
+      "learning_rate": 8.477287799988502e-06,
+      "loss": 0.8777,
+      "step": 9348
+    },
+    {
+      "epoch": 0.5145577632230722,
+      "grad_norm": 0.7022722959518433,
+      "learning_rate": 8.476976312040082e-06,
+      "loss": 0.7116,
+      "step": 9349
+    },
+    {
+      "epoch": 0.5146128020254279,
+      "grad_norm": 0.7791306376457214,
+      "learning_rate": 8.476664797959639e-06,
+      "loss": 0.7262,
+      "step": 9350
+    },
+    {
+      "epoch": 0.5146678408277836,
+      "grad_norm": 0.7391177415847778,
+      "learning_rate": 8.476353257749514e-06,
+      "loss": 0.7308,
+      "step": 9351
+    },
+    {
+      "epoch": 0.5147228796301393,
+      "grad_norm": 0.6989552974700928,
+      "learning_rate": 8.476041691412046e-06,
+      "loss": 0.7754,
+      "step": 9352
+    },
+    {
+      "epoch": 0.5147779184324949,
+      "grad_norm": 0.7639930844306946,
+      "learning_rate": 8.475730098949582e-06,
+      "loss": 0.8385,
+      "step": 9353
+    },
+    {
+      "epoch": 0.5148329572348506,
+      "grad_norm": 0.7687931060791016,
+      "learning_rate": 8.47541848036446e-06,
+      "loss": 0.8118,
+      "step": 9354
+    },
+    {
+      "epoch": 0.5148879960372063,
+      "grad_norm": 0.8831589221954346,
+      "learning_rate": 8.475106835659024e-06,
+      "loss": 0.7705,
+      "step": 9355
+    },
+    {
+      "epoch": 0.5149430348395619,
+      "grad_norm": 0.7585502862930298,
+      "learning_rate": 8.474795164835614e-06,
+      "loss": 0.8167,
+      "step": 9356
+    },
+    {
+      "epoch": 0.5149980736419175,
+      "grad_norm": 0.7078690528869629,
+      "learning_rate": 8.474483467896572e-06,
+      "loss": 0.7412,
+      "step": 9357
+    },
+    {
+      "epoch": 0.5150531124442732,
+      "grad_norm": 0.8950889706611633,
+      "learning_rate": 8.474171744844246e-06,
+      "loss": 0.8132,
+      "step": 9358
+    },
+    {
+      "epoch": 0.5151081512466289,
+      "grad_norm": 0.7196077704429626,
+      "learning_rate": 8.473859995680973e-06,
+      "loss": 0.8041,
+      "step": 9359
+    },
+    {
+      "epoch": 0.5151631900489846,
+      "grad_norm": 0.7705141305923462,
+      "learning_rate": 8.473548220409099e-06,
+      "loss": 0.8437,
+      "step": 9360
+    },
+    {
+      "epoch": 0.5152182288513402,
+      "grad_norm": 0.6507467031478882,
+      "learning_rate": 8.473236419030966e-06,
+      "loss": 0.7713,
+      "step": 9361
+    },
+    {
+      "epoch": 0.5152732676536959,
+      "grad_norm": 0.7120817303657532,
+      "learning_rate": 8.472924591548917e-06,
+      "loss": 0.7688,
+      "step": 9362
+    },
+    {
+      "epoch": 0.5153283064560515,
+      "grad_norm": 0.7830487489700317,
+      "learning_rate": 8.472612737965297e-06,
+      "loss": 0.8875,
+      "step": 9363
+    },
+    {
+      "epoch": 0.5153833452584071,
+      "grad_norm": 0.8790529370307922,
+      "learning_rate": 8.47230085828245e-06,
+      "loss": 0.7648,
+      "step": 9364
+    },
+    {
+      "epoch": 0.5154383840607628,
+      "grad_norm": 0.8956806659698486,
+      "learning_rate": 8.471988952502718e-06,
+      "loss": 0.7891,
+      "step": 9365
+    },
+    {
+      "epoch": 0.5154934228631185,
+      "grad_norm": 0.7370011210441589,
+      "learning_rate": 8.471677020628448e-06,
+      "loss": 0.7609,
+      "step": 9366
+    },
+    {
+      "epoch": 0.5155484616654742,
+      "grad_norm": 0.6794238090515137,
+      "learning_rate": 8.471365062661982e-06,
+      "loss": 0.6679,
+      "step": 9367
+    },
+    {
+      "epoch": 0.5156035004678298,
+      "grad_norm": 0.7330273985862732,
+      "learning_rate": 8.471053078605664e-06,
+      "loss": 0.7276,
+      "step": 9368
+    },
+    {
+      "epoch": 0.5156585392701855,
+      "grad_norm": 0.7796601057052612,
+      "learning_rate": 8.470741068461843e-06,
+      "loss": 0.7897,
+      "step": 9369
+    },
+    {
+      "epoch": 0.5157135780725411,
+      "grad_norm": 0.6834099888801575,
+      "learning_rate": 8.470429032232858e-06,
+      "loss": 0.7924,
+      "step": 9370
+    },
+    {
+      "epoch": 0.5157686168748968,
+      "grad_norm": 0.6991616487503052,
+      "learning_rate": 8.47011696992106e-06,
+      "loss": 0.7901,
+      "step": 9371
+    },
+    {
+      "epoch": 0.5158236556772524,
+      "grad_norm": 0.7321401834487915,
+      "learning_rate": 8.469804881528792e-06,
+      "loss": 0.6718,
+      "step": 9372
+    },
+    {
+      "epoch": 0.5158786944796081,
+      "grad_norm": 0.7091043591499329,
+      "learning_rate": 8.469492767058398e-06,
+      "loss": 0.8204,
+      "step": 9373
+    },
+    {
+      "epoch": 0.5159337332819638,
+      "grad_norm": 0.8777012825012207,
+      "learning_rate": 8.469180626512223e-06,
+      "loss": 0.8045,
+      "step": 9374
+    },
+    {
+      "epoch": 0.5159887720843195,
+      "grad_norm": 0.6652738451957703,
+      "learning_rate": 8.468868459892619e-06,
+      "loss": 0.7248,
+      "step": 9375
+    },
+    {
+      "epoch": 0.5160438108866751,
+      "grad_norm": 0.7209659218788147,
+      "learning_rate": 8.468556267201925e-06,
+      "loss": 0.7508,
+      "step": 9376
+    },
+    {
+      "epoch": 0.5160988496890307,
+      "grad_norm": 0.7685441970825195,
+      "learning_rate": 8.468244048442494e-06,
+      "loss": 0.7501,
+      "step": 9377
+    },
+    {
+      "epoch": 0.5161538884913864,
+      "grad_norm": 0.6773725152015686,
+      "learning_rate": 8.467931803616665e-06,
+      "loss": 0.8036,
+      "step": 9378
+    },
+    {
+      "epoch": 0.5162089272937421,
+      "grad_norm": 0.7167890071868896,
+      "learning_rate": 8.467619532726792e-06,
+      "loss": 0.7229,
+      "step": 9379
+    },
+    {
+      "epoch": 0.5162639660960977,
+      "grad_norm": 0.7066929340362549,
+      "learning_rate": 8.467307235775218e-06,
+      "loss": 0.7433,
+      "step": 9380
+    },
+    {
+      "epoch": 0.5163190048984534,
+      "grad_norm": 0.7261828780174255,
+      "learning_rate": 8.46699491276429e-06,
+      "loss": 0.7873,
+      "step": 9381
+    },
+    {
+      "epoch": 0.5163740437008091,
+      "grad_norm": 0.7442463636398315,
+      "learning_rate": 8.466682563696356e-06,
+      "loss": 0.7953,
+      "step": 9382
+    },
+    {
+      "epoch": 0.5164290825031648,
+      "grad_norm": 0.5668768286705017,
+      "learning_rate": 8.466370188573765e-06,
+      "loss": 0.5602,
+      "step": 9383
+    },
+    {
+      "epoch": 0.5164841213055203,
+      "grad_norm": 0.7364997267723083,
+      "learning_rate": 8.466057787398864e-06,
+      "loss": 0.8274,
+      "step": 9384
+    },
+    {
+      "epoch": 0.516539160107876,
+      "grad_norm": 0.7793132066726685,
+      "learning_rate": 8.465745360174e-06,
+      "loss": 0.7832,
+      "step": 9385
+    },
+    {
+      "epoch": 0.5165941989102317,
+      "grad_norm": 0.6818128824234009,
+      "learning_rate": 8.46543290690152e-06,
+      "loss": 0.8314,
+      "step": 9386
+    },
+    {
+      "epoch": 0.5166492377125874,
+      "grad_norm": 0.7392195463180542,
+      "learning_rate": 8.465120427583778e-06,
+      "loss": 0.8124,
+      "step": 9387
+    },
+    {
+      "epoch": 0.516704276514943,
+      "grad_norm": 0.8582521677017212,
+      "learning_rate": 8.464807922223115e-06,
+      "loss": 0.7417,
+      "step": 9388
+    },
+    {
+      "epoch": 0.5167593153172987,
+      "grad_norm": 0.7322097420692444,
+      "learning_rate": 8.464495390821882e-06,
+      "loss": 0.7408,
+      "step": 9389
+    },
+    {
+      "epoch": 0.5168143541196544,
+      "grad_norm": 0.8177433013916016,
+      "learning_rate": 8.464182833382432e-06,
+      "loss": 0.87,
+      "step": 9390
+    },
+    {
+      "epoch": 0.5168693929220101,
+      "grad_norm": 0.7088115215301514,
+      "learning_rate": 8.46387024990711e-06,
+      "loss": 0.7748,
+      "step": 9391
+    },
+    {
+      "epoch": 0.5169244317243656,
+      "grad_norm": 0.6648650169372559,
+      "learning_rate": 8.463557640398268e-06,
+      "loss": 0.6302,
+      "step": 9392
+    },
+    {
+      "epoch": 0.5169794705267213,
+      "grad_norm": 0.6688859462738037,
+      "learning_rate": 8.463245004858251e-06,
+      "loss": 0.7252,
+      "step": 9393
+    },
+    {
+      "epoch": 0.517034509329077,
+      "grad_norm": 0.7231030464172363,
+      "learning_rate": 8.462932343289412e-06,
+      "loss": 0.8497,
+      "step": 9394
+    },
+    {
+      "epoch": 0.5170895481314327,
+      "grad_norm": 0.7142065763473511,
+      "learning_rate": 8.462619655694103e-06,
+      "loss": 0.7041,
+      "step": 9395
+    },
+    {
+      "epoch": 0.5171445869337883,
+      "grad_norm": 0.7197136878967285,
+      "learning_rate": 8.462306942074669e-06,
+      "loss": 0.7022,
+      "step": 9396
+    },
+    {
+      "epoch": 0.517199625736144,
+      "grad_norm": 0.7620192766189575,
+      "learning_rate": 8.461994202433463e-06,
+      "loss": 0.8243,
+      "step": 9397
+    },
+    {
+      "epoch": 0.5172546645384997,
+      "grad_norm": 0.7697533965110779,
+      "learning_rate": 8.461681436772836e-06,
+      "loss": 0.7861,
+      "step": 9398
+    },
+    {
+      "epoch": 0.5173097033408554,
+      "grad_norm": 0.7224711179733276,
+      "learning_rate": 8.461368645095138e-06,
+      "loss": 0.7588,
+      "step": 9399
+    },
+    {
+      "epoch": 0.5173647421432109,
+      "grad_norm": 0.9285979270935059,
+      "learning_rate": 8.46105582740272e-06,
+      "loss": 0.8113,
+      "step": 9400
+    },
+    {
+      "epoch": 0.5174197809455666,
+      "grad_norm": 0.7297842502593994,
+      "learning_rate": 8.460742983697934e-06,
+      "loss": 0.7115,
+      "step": 9401
+    },
+    {
+      "epoch": 0.5174748197479223,
+      "grad_norm": 0.6712872982025146,
+      "learning_rate": 8.460430113983126e-06,
+      "loss": 0.751,
+      "step": 9402
+    },
+    {
+      "epoch": 0.517529858550278,
+      "grad_norm": 0.7807186245918274,
+      "learning_rate": 8.460117218260657e-06,
+      "loss": 0.8375,
+      "step": 9403
+    },
+    {
+      "epoch": 0.5175848973526336,
+      "grad_norm": 0.621530294418335,
+      "learning_rate": 8.45980429653287e-06,
+      "loss": 0.638,
+      "step": 9404
+    },
+    {
+      "epoch": 0.5176399361549893,
+      "grad_norm": 0.7086256146430969,
+      "learning_rate": 8.45949134880212e-06,
+      "loss": 0.8304,
+      "step": 9405
+    },
+    {
+      "epoch": 0.517694974957345,
+      "grad_norm": 0.62705397605896,
+      "learning_rate": 8.45917837507076e-06,
+      "loss": 0.7008,
+      "step": 9406
+    },
+    {
+      "epoch": 0.5177500137597005,
+      "grad_norm": 0.9109121561050415,
+      "learning_rate": 8.458865375341142e-06,
+      "loss": 0.7529,
+      "step": 9407
+    },
+    {
+      "epoch": 0.5178050525620562,
+      "grad_norm": 0.6909900903701782,
+      "learning_rate": 8.458552349615615e-06,
+      "loss": 0.8453,
+      "step": 9408
+    },
+    {
+      "epoch": 0.5178600913644119,
+      "grad_norm": 0.7548434138298035,
+      "learning_rate": 8.458239297896536e-06,
+      "loss": 0.7516,
+      "step": 9409
+    },
+    {
+      "epoch": 0.5179151301667676,
+      "grad_norm": 0.7595730423927307,
+      "learning_rate": 8.457926220186257e-06,
+      "loss": 0.7599,
+      "step": 9410
+    },
+    {
+      "epoch": 0.5179701689691232,
+      "grad_norm": 0.7449337840080261,
+      "learning_rate": 8.45761311648713e-06,
+      "loss": 0.8236,
+      "step": 9411
+    },
+    {
+      "epoch": 0.5180252077714789,
+      "grad_norm": 0.7529160976409912,
+      "learning_rate": 8.457299986801507e-06,
+      "loss": 0.8655,
+      "step": 9412
+    },
+    {
+      "epoch": 0.5180802465738346,
+      "grad_norm": 0.6777701377868652,
+      "learning_rate": 8.456986831131742e-06,
+      "loss": 0.7737,
+      "step": 9413
+    },
+    {
+      "epoch": 0.5181352853761902,
+      "grad_norm": 0.9363510012626648,
+      "learning_rate": 8.456673649480191e-06,
+      "loss": 0.8227,
+      "step": 9414
+    },
+    {
+      "epoch": 0.5181903241785458,
+      "grad_norm": 0.798001229763031,
+      "learning_rate": 8.456360441849206e-06,
+      "loss": 0.8881,
+      "step": 9415
+    },
+    {
+      "epoch": 0.5182453629809015,
+      "grad_norm": 0.7212072610855103,
+      "learning_rate": 8.456047208241141e-06,
+      "loss": 0.8165,
+      "step": 9416
+    },
+    {
+      "epoch": 0.5183004017832572,
+      "grad_norm": 0.6918027997016907,
+      "learning_rate": 8.45573394865835e-06,
+      "loss": 0.8048,
+      "step": 9417
+    },
+    {
+      "epoch": 0.5183554405856129,
+      "grad_norm": 0.6474916338920593,
+      "learning_rate": 8.455420663103187e-06,
+      "loss": 0.6502,
+      "step": 9418
+    },
+    {
+      "epoch": 0.5184104793879685,
+      "grad_norm": 0.6592364311218262,
+      "learning_rate": 8.455107351578008e-06,
+      "loss": 0.7509,
+      "step": 9419
+    },
+    {
+      "epoch": 0.5184655181903242,
+      "grad_norm": 0.7658745646476746,
+      "learning_rate": 8.454794014085168e-06,
+      "loss": 0.8444,
+      "step": 9420
+    },
+    {
+      "epoch": 0.5185205569926798,
+      "grad_norm": 0.6814215183258057,
+      "learning_rate": 8.45448065062702e-06,
+      "loss": 0.7367,
+      "step": 9421
+    },
+    {
+      "epoch": 0.5185755957950355,
+      "grad_norm": 0.644740104675293,
+      "learning_rate": 8.45416726120592e-06,
+      "loss": 0.7456,
+      "step": 9422
+    },
+    {
+      "epoch": 0.5186306345973911,
+      "grad_norm": 0.8578751087188721,
+      "learning_rate": 8.453853845824225e-06,
+      "loss": 0.8481,
+      "step": 9423
+    },
+    {
+      "epoch": 0.5186856733997468,
+      "grad_norm": 0.6630389094352722,
+      "learning_rate": 8.453540404484288e-06,
+      "loss": 0.7487,
+      "step": 9424
+    },
+    {
+      "epoch": 0.5187407122021025,
+      "grad_norm": 0.7756431698799133,
+      "learning_rate": 8.453226937188466e-06,
+      "loss": 0.798,
+      "step": 9425
+    },
+    {
+      "epoch": 0.5187957510044582,
+      "grad_norm": 0.7856318354606628,
+      "learning_rate": 8.452913443939113e-06,
+      "loss": 0.785,
+      "step": 9426
+    },
+    {
+      "epoch": 0.5188507898068138,
+      "grad_norm": 0.7563977837562561,
+      "learning_rate": 8.45259992473859e-06,
+      "loss": 0.8182,
+      "step": 9427
+    },
+    {
+      "epoch": 0.5189058286091695,
+      "grad_norm": 0.6945043802261353,
+      "learning_rate": 8.452286379589247e-06,
+      "loss": 0.7262,
+      "step": 9428
+    },
+    {
+      "epoch": 0.5189608674115251,
+      "grad_norm": 0.6607717275619507,
+      "learning_rate": 8.451972808493444e-06,
+      "loss": 0.7257,
+      "step": 9429
+    },
+    {
+      "epoch": 0.5190159062138808,
+      "grad_norm": 0.6682843565940857,
+      "learning_rate": 8.451659211453539e-06,
+      "loss": 0.6775,
+      "step": 9430
+    },
+    {
+      "epoch": 0.5190709450162364,
+      "grad_norm": 0.7175559401512146,
+      "learning_rate": 8.451345588471886e-06,
+      "loss": 0.7154,
+      "step": 9431
+    },
+    {
+      "epoch": 0.5191259838185921,
+      "grad_norm": 0.7499119639396667,
+      "learning_rate": 8.451031939550845e-06,
+      "loss": 0.7537,
+      "step": 9432
+    },
+    {
+      "epoch": 0.5191810226209478,
+      "grad_norm": 0.65048748254776,
+      "learning_rate": 8.450718264692771e-06,
+      "loss": 0.7253,
+      "step": 9433
+    },
+    {
+      "epoch": 0.5192360614233035,
+      "grad_norm": 0.7067640423774719,
+      "learning_rate": 8.450404563900022e-06,
+      "loss": 0.7245,
+      "step": 9434
+    },
+    {
+      "epoch": 0.519291100225659,
+      "grad_norm": 0.7079932689666748,
+      "learning_rate": 8.450090837174956e-06,
+      "loss": 0.7776,
+      "step": 9435
+    },
+    {
+      "epoch": 0.5193461390280147,
+      "grad_norm": 0.8260107636451721,
+      "learning_rate": 8.44977708451993e-06,
+      "loss": 0.8529,
+      "step": 9436
+    },
+    {
+      "epoch": 0.5194011778303704,
+      "grad_norm": 0.6412167549133301,
+      "learning_rate": 8.449463305937304e-06,
+      "loss": 0.7371,
+      "step": 9437
+    },
+    {
+      "epoch": 0.5194562166327261,
+      "grad_norm": 0.7067576050758362,
+      "learning_rate": 8.449149501429435e-06,
+      "loss": 0.7161,
+      "step": 9438
+    },
+    {
+      "epoch": 0.5195112554350817,
+      "grad_norm": 0.6966904997825623,
+      "learning_rate": 8.448835670998681e-06,
+      "loss": 0.7285,
+      "step": 9439
+    },
+    {
+      "epoch": 0.5195662942374374,
+      "grad_norm": 0.8066132664680481,
+      "learning_rate": 8.448521814647401e-06,
+      "loss": 0.8265,
+      "step": 9440
+    },
+    {
+      "epoch": 0.5196213330397931,
+      "grad_norm": 0.7597149610519409,
+      "learning_rate": 8.448207932377957e-06,
+      "loss": 0.7721,
+      "step": 9441
+    },
+    {
+      "epoch": 0.5196763718421488,
+      "grad_norm": 0.6965302228927612,
+      "learning_rate": 8.447894024192702e-06,
+      "loss": 0.749,
+      "step": 9442
+    },
+    {
+      "epoch": 0.5197314106445043,
+      "grad_norm": 0.7032600045204163,
+      "learning_rate": 8.447580090094e-06,
+      "loss": 0.7923,
+      "step": 9443
+    },
+    {
+      "epoch": 0.51978644944686,
+      "grad_norm": 0.7255309820175171,
+      "learning_rate": 8.447266130084208e-06,
+      "loss": 0.6739,
+      "step": 9444
+    },
+    {
+      "epoch": 0.5198414882492157,
+      "grad_norm": 0.6602993011474609,
+      "learning_rate": 8.446952144165686e-06,
+      "loss": 0.7886,
+      "step": 9445
+    },
+    {
+      "epoch": 0.5198965270515714,
+      "grad_norm": 0.7017884850502014,
+      "learning_rate": 8.446638132340796e-06,
+      "loss": 0.7554,
+      "step": 9446
+    },
+    {
+      "epoch": 0.519951565853927,
+      "grad_norm": 0.7234843969345093,
+      "learning_rate": 8.446324094611894e-06,
+      "loss": 0.8294,
+      "step": 9447
+    },
+    {
+      "epoch": 0.5200066046562827,
+      "grad_norm": 0.6859332919120789,
+      "learning_rate": 8.446010030981347e-06,
+      "loss": 0.7563,
+      "step": 9448
+    },
+    {
+      "epoch": 0.5200616434586384,
+      "grad_norm": 0.7759458422660828,
+      "learning_rate": 8.445695941451507e-06,
+      "loss": 0.7577,
+      "step": 9449
+    },
+    {
+      "epoch": 0.520116682260994,
+      "grad_norm": 0.7852263450622559,
+      "learning_rate": 8.44538182602474e-06,
+      "loss": 0.7446,
+      "step": 9450
+    },
+    {
+      "epoch": 0.5201717210633496,
+      "grad_norm": 0.8143053650856018,
+      "learning_rate": 8.445067684703406e-06,
+      "loss": 0.7995,
+      "step": 9451
+    },
+    {
+      "epoch": 0.5202267598657053,
+      "grad_norm": 0.692738950252533,
+      "learning_rate": 8.444753517489865e-06,
+      "loss": 0.7185,
+      "step": 9452
+    },
+    {
+      "epoch": 0.520281798668061,
+      "grad_norm": 0.6615390181541443,
+      "learning_rate": 8.444439324386478e-06,
+      "loss": 0.7128,
+      "step": 9453
+    },
+    {
+      "epoch": 0.5203368374704166,
+      "grad_norm": 0.7360419034957886,
+      "learning_rate": 8.444125105395608e-06,
+      "loss": 0.6565,
+      "step": 9454
+    },
+    {
+      "epoch": 0.5203918762727723,
+      "grad_norm": 0.7280182838439941,
+      "learning_rate": 8.443810860519615e-06,
+      "loss": 0.7295,
+      "step": 9455
+    },
+    {
+      "epoch": 0.520446915075128,
+      "grad_norm": 0.787367582321167,
+      "learning_rate": 8.44349658976086e-06,
+      "loss": 0.7342,
+      "step": 9456
+    },
+    {
+      "epoch": 0.5205019538774837,
+      "grad_norm": 0.7496024966239929,
+      "learning_rate": 8.44318229312171e-06,
+      "loss": 0.7499,
+      "step": 9457
+    },
+    {
+      "epoch": 0.5205569926798392,
+      "grad_norm": 0.9167383909225464,
+      "learning_rate": 8.44286797060452e-06,
+      "loss": 0.7797,
+      "step": 9458
+    },
+    {
+      "epoch": 0.5206120314821949,
+      "grad_norm": 0.7032341957092285,
+      "learning_rate": 8.442553622211659e-06,
+      "loss": 0.7627,
+      "step": 9459
+    },
+    {
+      "epoch": 0.5206670702845506,
+      "grad_norm": 1.2905993461608887,
+      "learning_rate": 8.442239247945485e-06,
+      "loss": 0.7841,
+      "step": 9460
+    },
+    {
+      "epoch": 0.5207221090869063,
+      "grad_norm": 0.6909230351448059,
+      "learning_rate": 8.441924847808362e-06,
+      "loss": 0.7234,
+      "step": 9461
+    },
+    {
+      "epoch": 0.5207771478892619,
+      "grad_norm": 0.6632175445556641,
+      "learning_rate": 8.441610421802653e-06,
+      "loss": 0.6733,
+      "step": 9462
+    },
+    {
+      "epoch": 0.5208321866916176,
+      "grad_norm": 0.7838154435157776,
+      "learning_rate": 8.441295969930722e-06,
+      "loss": 0.7583,
+      "step": 9463
+    },
+    {
+      "epoch": 0.5208872254939733,
+      "grad_norm": 0.6380481123924255,
+      "learning_rate": 8.440981492194932e-06,
+      "loss": 0.7109,
+      "step": 9464
+    },
+    {
+      "epoch": 0.520942264296329,
+      "grad_norm": 0.6859052181243896,
+      "learning_rate": 8.440666988597646e-06,
+      "loss": 0.7387,
+      "step": 9465
+    },
+    {
+      "epoch": 0.5209973030986845,
+      "grad_norm": 0.7411379814147949,
+      "learning_rate": 8.440352459141226e-06,
+      "loss": 0.7852,
+      "step": 9466
+    },
+    {
+      "epoch": 0.5210523419010402,
+      "grad_norm": 0.6925216913223267,
+      "learning_rate": 8.44003790382804e-06,
+      "loss": 0.8228,
+      "step": 9467
+    },
+    {
+      "epoch": 0.5211073807033959,
+      "grad_norm": 0.7136396169662476,
+      "learning_rate": 8.43972332266045e-06,
+      "loss": 0.8168,
+      "step": 9468
+    },
+    {
+      "epoch": 0.5211624195057516,
+      "grad_norm": 0.719639003276825,
+      "learning_rate": 8.43940871564082e-06,
+      "loss": 0.6728,
+      "step": 9469
+    },
+    {
+      "epoch": 0.5212174583081072,
+      "grad_norm": 0.647861897945404,
+      "learning_rate": 8.439094082771513e-06,
+      "loss": 0.6986,
+      "step": 9470
+    },
+    {
+      "epoch": 0.5212724971104629,
+      "grad_norm": 0.6644579172134399,
+      "learning_rate": 8.438779424054897e-06,
+      "loss": 0.6263,
+      "step": 9471
+    },
+    {
+      "epoch": 0.5213275359128186,
+      "grad_norm": 0.7157352566719055,
+      "learning_rate": 8.438464739493335e-06,
+      "loss": 0.827,
+      "step": 9472
+    },
+    {
+      "epoch": 0.5213825747151742,
+      "grad_norm": 0.793765127658844,
+      "learning_rate": 8.438150029089193e-06,
+      "loss": 0.741,
+      "step": 9473
+    },
+    {
+      "epoch": 0.5214376135175298,
+      "grad_norm": 0.7078518867492676,
+      "learning_rate": 8.437835292844836e-06,
+      "loss": 0.7618,
+      "step": 9474
+    },
+    {
+      "epoch": 0.5214926523198855,
+      "grad_norm": 0.7492140531539917,
+      "learning_rate": 8.437520530762628e-06,
+      "loss": 0.7894,
+      "step": 9475
+    },
+    {
+      "epoch": 0.5215476911222412,
+      "grad_norm": 0.6534473299980164,
+      "learning_rate": 8.437205742844937e-06,
+      "loss": 0.7567,
+      "step": 9476
+    },
+    {
+      "epoch": 0.5216027299245969,
+      "grad_norm": 0.8745388984680176,
+      "learning_rate": 8.436890929094126e-06,
+      "loss": 0.8758,
+      "step": 9477
+    },
+    {
+      "epoch": 0.5216577687269525,
+      "grad_norm": 0.6804752349853516,
+      "learning_rate": 8.436576089512564e-06,
+      "loss": 0.7841,
+      "step": 9478
+    },
+    {
+      "epoch": 0.5217128075293082,
+      "grad_norm": 0.712065577507019,
+      "learning_rate": 8.436261224102615e-06,
+      "loss": 0.8079,
+      "step": 9479
+    },
+    {
+      "epoch": 0.5217678463316638,
+      "grad_norm": 0.8733783960342407,
+      "learning_rate": 8.435946332866648e-06,
+      "loss": 0.8295,
+      "step": 9480
+    },
+    {
+      "epoch": 0.5218228851340195,
+      "grad_norm": 0.6871289610862732,
+      "learning_rate": 8.435631415807028e-06,
+      "loss": 0.7087,
+      "step": 9481
+    },
+    {
+      "epoch": 0.5218779239363751,
+      "grad_norm": 0.8363185524940491,
+      "learning_rate": 8.43531647292612e-06,
+      "loss": 0.7329,
+      "step": 9482
+    },
+    {
+      "epoch": 0.5219329627387308,
+      "grad_norm": 0.6845195293426514,
+      "learning_rate": 8.435001504226295e-06,
+      "loss": 0.7651,
+      "step": 9483
+    },
+    {
+      "epoch": 0.5219880015410865,
+      "grad_norm": 0.7527645826339722,
+      "learning_rate": 8.434686509709917e-06,
+      "loss": 0.6856,
+      "step": 9484
+    },
+    {
+      "epoch": 0.5220430403434422,
+      "grad_norm": 0.6945710778236389,
+      "learning_rate": 8.434371489379356e-06,
+      "loss": 0.6875,
+      "step": 9485
+    },
+    {
+      "epoch": 0.5220980791457978,
+      "grad_norm": 0.7668873071670532,
+      "learning_rate": 8.434056443236977e-06,
+      "loss": 0.7662,
+      "step": 9486
+    },
+    {
+      "epoch": 0.5221531179481534,
+      "grad_norm": 0.9873473048210144,
+      "learning_rate": 8.433741371285148e-06,
+      "loss": 0.7662,
+      "step": 9487
+    },
+    {
+      "epoch": 0.5222081567505091,
+      "grad_norm": 0.8635447025299072,
+      "learning_rate": 8.43342627352624e-06,
+      "loss": 0.645,
+      "step": 9488
+    },
+    {
+      "epoch": 0.5222631955528648,
+      "grad_norm": 0.7836978435516357,
+      "learning_rate": 8.43311114996262e-06,
+      "loss": 0.7647,
+      "step": 9489
+    },
+    {
+      "epoch": 0.5223182343552204,
+      "grad_norm": 0.8370835185050964,
+      "learning_rate": 8.432796000596652e-06,
+      "loss": 0.8402,
+      "step": 9490
+    },
+    {
+      "epoch": 0.5223732731575761,
+      "grad_norm": 0.9627843499183655,
+      "learning_rate": 8.432480825430712e-06,
+      "loss": 0.6985,
+      "step": 9491
+    },
+    {
+      "epoch": 0.5224283119599318,
+      "grad_norm": 0.6774263978004456,
+      "learning_rate": 8.432165624467163e-06,
+      "loss": 0.7051,
+      "step": 9492
+    },
+    {
+      "epoch": 0.5224833507622874,
+      "grad_norm": 0.6590597033500671,
+      "learning_rate": 8.431850397708375e-06,
+      "loss": 0.7147,
+      "step": 9493
+    },
+    {
+      "epoch": 0.522538389564643,
+      "grad_norm": 0.8153522610664368,
+      "learning_rate": 8.43153514515672e-06,
+      "loss": 0.6759,
+      "step": 9494
+    },
+    {
+      "epoch": 0.5225934283669987,
+      "grad_norm": 0.7457708716392517,
+      "learning_rate": 8.431219866814563e-06,
+      "loss": 0.7168,
+      "step": 9495
+    },
+    {
+      "epoch": 0.5226484671693544,
+      "grad_norm": 0.6994161009788513,
+      "learning_rate": 8.430904562684278e-06,
+      "loss": 0.8393,
+      "step": 9496
+    },
+    {
+      "epoch": 0.52270350597171,
+      "grad_norm": 0.780337393283844,
+      "learning_rate": 8.430589232768232e-06,
+      "loss": 0.6528,
+      "step": 9497
+    },
+    {
+      "epoch": 0.5227585447740657,
+      "grad_norm": 0.6833232641220093,
+      "learning_rate": 8.430273877068796e-06,
+      "loss": 0.7545,
+      "step": 9498
+    },
+    {
+      "epoch": 0.5228135835764214,
+      "grad_norm": 0.7330057621002197,
+      "learning_rate": 8.42995849558834e-06,
+      "loss": 0.7932,
+      "step": 9499
+    },
+    {
+      "epoch": 0.5228686223787771,
+      "grad_norm": 0.8131541609764099,
+      "learning_rate": 8.429643088329233e-06,
+      "loss": 0.7546,
+      "step": 9500
+    },
+    {
+      "epoch": 0.5229236611811326,
+      "grad_norm": 0.7353833317756653,
+      "learning_rate": 8.42932765529385e-06,
+      "loss": 0.7508,
+      "step": 9501
+    },
+    {
+      "epoch": 0.5229786999834883,
+      "grad_norm": 0.7166246771812439,
+      "learning_rate": 8.429012196484554e-06,
+      "loss": 0.728,
+      "step": 9502
+    },
+    {
+      "epoch": 0.523033738785844,
+      "grad_norm": 0.732064962387085,
+      "learning_rate": 8.428696711903721e-06,
+      "loss": 0.8306,
+      "step": 9503
+    },
+    {
+      "epoch": 0.5230887775881997,
+      "grad_norm": 0.6858934164047241,
+      "learning_rate": 8.428381201553721e-06,
+      "loss": 0.7801,
+      "step": 9504
+    },
+    {
+      "epoch": 0.5231438163905553,
+      "grad_norm": 0.7046478986740112,
+      "learning_rate": 8.428065665436928e-06,
+      "loss": 0.7365,
+      "step": 9505
+    },
+    {
+      "epoch": 0.523198855192911,
+      "grad_norm": 0.6669325828552246,
+      "learning_rate": 8.42775010355571e-06,
+      "loss": 0.7764,
+      "step": 9506
+    },
+    {
+      "epoch": 0.5232538939952667,
+      "grad_norm": 0.655619740486145,
+      "learning_rate": 8.427434515912438e-06,
+      "loss": 0.7919,
+      "step": 9507
+    },
+    {
+      "epoch": 0.5233089327976224,
+      "grad_norm": 0.6236690878868103,
+      "learning_rate": 8.427118902509487e-06,
+      "loss": 0.6653,
+      "step": 9508
+    },
+    {
+      "epoch": 0.5233639715999779,
+      "grad_norm": 0.8233165740966797,
+      "learning_rate": 8.426803263349228e-06,
+      "loss": 0.8012,
+      "step": 9509
+    },
+    {
+      "epoch": 0.5234190104023336,
+      "grad_norm": 0.6626759171485901,
+      "learning_rate": 8.426487598434035e-06,
+      "loss": 0.7728,
+      "step": 9510
+    },
+    {
+      "epoch": 0.5234740492046893,
+      "grad_norm": 0.9209974408149719,
+      "learning_rate": 8.426171907766275e-06,
+      "loss": 0.769,
+      "step": 9511
+    },
+    {
+      "epoch": 0.523529088007045,
+      "grad_norm": 0.6297587156295776,
+      "learning_rate": 8.425856191348325e-06,
+      "loss": 0.7333,
+      "step": 9512
+    },
+    {
+      "epoch": 0.5235841268094006,
+      "grad_norm": 0.6995256543159485,
+      "learning_rate": 8.425540449182558e-06,
+      "loss": 0.7486,
+      "step": 9513
+    },
+    {
+      "epoch": 0.5236391656117563,
+      "grad_norm": 0.8076607584953308,
+      "learning_rate": 8.425224681271345e-06,
+      "loss": 0.8533,
+      "step": 9514
+    },
+    {
+      "epoch": 0.523694204414112,
+      "grad_norm": 1.2198601961135864,
+      "learning_rate": 8.42490888761706e-06,
+      "loss": 0.7291,
+      "step": 9515
+    },
+    {
+      "epoch": 0.5237492432164677,
+      "grad_norm": 0.7047159671783447,
+      "learning_rate": 8.424593068222076e-06,
+      "loss": 0.713,
+      "step": 9516
+    },
+    {
+      "epoch": 0.5238042820188232,
+      "grad_norm": 0.7652333378791809,
+      "learning_rate": 8.424277223088768e-06,
+      "loss": 0.8149,
+      "step": 9517
+    },
+    {
+      "epoch": 0.5238593208211789,
+      "grad_norm": 1.1311010122299194,
+      "learning_rate": 8.42396135221951e-06,
+      "loss": 0.8195,
+      "step": 9518
+    },
+    {
+      "epoch": 0.5239143596235346,
+      "grad_norm": 0.7855533957481384,
+      "learning_rate": 8.423645455616674e-06,
+      "loss": 0.7901,
+      "step": 9519
+    },
+    {
+      "epoch": 0.5239693984258903,
+      "grad_norm": 0.7028971314430237,
+      "learning_rate": 8.423329533282635e-06,
+      "loss": 0.8006,
+      "step": 9520
+    },
+    {
+      "epoch": 0.5240244372282459,
+      "grad_norm": 0.703809916973114,
+      "learning_rate": 8.423013585219769e-06,
+      "loss": 0.7581,
+      "step": 9521
+    },
+    {
+      "epoch": 0.5240794760306016,
+      "grad_norm": 0.94233238697052,
+      "learning_rate": 8.422697611430448e-06,
+      "loss": 0.7689,
+      "step": 9522
+    },
+    {
+      "epoch": 0.5241345148329573,
+      "grad_norm": 0.8164071440696716,
+      "learning_rate": 8.422381611917047e-06,
+      "loss": 0.8761,
+      "step": 9523
+    },
+    {
+      "epoch": 0.5241895536353129,
+      "grad_norm": 0.6242091059684753,
+      "learning_rate": 8.422065586681944e-06,
+      "loss": 0.6975,
+      "step": 9524
+    },
+    {
+      "epoch": 0.5242445924376685,
+      "grad_norm": 0.6607261300086975,
+      "learning_rate": 8.42174953572751e-06,
+      "loss": 0.6847,
+      "step": 9525
+    },
+    {
+      "epoch": 0.5242996312400242,
+      "grad_norm": 0.7174261212348938,
+      "learning_rate": 8.421433459056123e-06,
+      "loss": 0.7905,
+      "step": 9526
+    },
+    {
+      "epoch": 0.5243546700423799,
+      "grad_norm": 0.7414089441299438,
+      "learning_rate": 8.42111735667016e-06,
+      "loss": 0.7788,
+      "step": 9527
+    },
+    {
+      "epoch": 0.5244097088447356,
+      "grad_norm": 0.7347442507743835,
+      "learning_rate": 8.420801228571992e-06,
+      "loss": 0.7691,
+      "step": 9528
+    },
+    {
+      "epoch": 0.5244647476470912,
+      "grad_norm": 0.6947832107543945,
+      "learning_rate": 8.420485074763999e-06,
+      "loss": 0.6702,
+      "step": 9529
+    },
+    {
+      "epoch": 0.5245197864494469,
+      "grad_norm": 0.6865423321723938,
+      "learning_rate": 8.420168895248557e-06,
+      "loss": 0.7577,
+      "step": 9530
+    },
+    {
+      "epoch": 0.5245748252518025,
+      "grad_norm": 0.7023190855979919,
+      "learning_rate": 8.419852690028039e-06,
+      "loss": 0.7711,
+      "step": 9531
+    },
+    {
+      "epoch": 0.5246298640541582,
+      "grad_norm": 0.8312145471572876,
+      "learning_rate": 8.419536459104824e-06,
+      "loss": 0.7999,
+      "step": 9532
+    },
+    {
+      "epoch": 0.5246849028565138,
+      "grad_norm": 0.6700688600540161,
+      "learning_rate": 8.419220202481288e-06,
+      "loss": 0.7163,
+      "step": 9533
+    },
+    {
+      "epoch": 0.5247399416588695,
+      "grad_norm": 0.767062246799469,
+      "learning_rate": 8.418903920159809e-06,
+      "loss": 0.7451,
+      "step": 9534
+    },
+    {
+      "epoch": 0.5247949804612252,
+      "grad_norm": 0.6814010143280029,
+      "learning_rate": 8.418587612142763e-06,
+      "loss": 0.771,
+      "step": 9535
+    },
+    {
+      "epoch": 0.5248500192635808,
+      "grad_norm": 0.6728426218032837,
+      "learning_rate": 8.418271278432528e-06,
+      "loss": 0.8336,
+      "step": 9536
+    },
+    {
+      "epoch": 0.5249050580659365,
+      "grad_norm": 0.7112382650375366,
+      "learning_rate": 8.417954919031482e-06,
+      "loss": 0.7392,
+      "step": 9537
+    },
+    {
+      "epoch": 0.5249600968682921,
+      "grad_norm": 0.7371365427970886,
+      "learning_rate": 8.417638533942e-06,
+      "loss": 0.8233,
+      "step": 9538
+    },
+    {
+      "epoch": 0.5250151356706478,
+      "grad_norm": 0.6593502163887024,
+      "learning_rate": 8.41732212316646e-06,
+      "loss": 0.7455,
+      "step": 9539
+    },
+    {
+      "epoch": 0.5250701744730034,
+      "grad_norm": 0.685553252696991,
+      "learning_rate": 8.417005686707245e-06,
+      "loss": 0.7783,
+      "step": 9540
+    },
+    {
+      "epoch": 0.5251252132753591,
+      "grad_norm": 0.7003353238105774,
+      "learning_rate": 8.41668922456673e-06,
+      "loss": 0.7733,
+      "step": 9541
+    },
+    {
+      "epoch": 0.5251802520777148,
+      "grad_norm": 0.7602891325950623,
+      "learning_rate": 8.416372736747292e-06,
+      "loss": 0.7236,
+      "step": 9542
+    },
+    {
+      "epoch": 0.5252352908800705,
+      "grad_norm": 0.647531270980835,
+      "learning_rate": 8.41605622325131e-06,
+      "loss": 0.7388,
+      "step": 9543
+    },
+    {
+      "epoch": 0.5252903296824261,
+      "grad_norm": 0.7309756875038147,
+      "learning_rate": 8.415739684081165e-06,
+      "loss": 0.7178,
+      "step": 9544
+    },
+    {
+      "epoch": 0.5253453684847817,
+      "grad_norm": 0.6991532444953918,
+      "learning_rate": 8.415423119239236e-06,
+      "loss": 0.8078,
+      "step": 9545
+    },
+    {
+      "epoch": 0.5254004072871374,
+      "grad_norm": 0.7392330765724182,
+      "learning_rate": 8.4151065287279e-06,
+      "loss": 0.8452,
+      "step": 9546
+    },
+    {
+      "epoch": 0.5254554460894931,
+      "grad_norm": 0.7617329955101013,
+      "learning_rate": 8.414789912549537e-06,
+      "loss": 0.7885,
+      "step": 9547
+    },
+    {
+      "epoch": 0.5255104848918487,
+      "grad_norm": 1.160125732421875,
+      "learning_rate": 8.414473270706527e-06,
+      "loss": 0.9628,
+      "step": 9548
+    },
+    {
+      "epoch": 0.5255655236942044,
+      "grad_norm": 0.7578685879707336,
+      "learning_rate": 8.414156603201252e-06,
+      "loss": 0.7745,
+      "step": 9549
+    },
+    {
+      "epoch": 0.5256205624965601,
+      "grad_norm": 0.6963017582893372,
+      "learning_rate": 8.413839910036089e-06,
+      "loss": 0.7693,
+      "step": 9550
+    },
+    {
+      "epoch": 0.5256756012989158,
+      "grad_norm": 0.6631398797035217,
+      "learning_rate": 8.413523191213415e-06,
+      "loss": 0.6606,
+      "step": 9551
+    },
+    {
+      "epoch": 0.5257306401012714,
+      "grad_norm": 0.707343339920044,
+      "learning_rate": 8.41320644673562e-06,
+      "loss": 0.7161,
+      "step": 9552
+    },
+    {
+      "epoch": 0.525785678903627,
+      "grad_norm": 0.833448588848114,
+      "learning_rate": 8.412889676605075e-06,
+      "loss": 0.7509,
+      "step": 9553
+    },
+    {
+      "epoch": 0.5258407177059827,
+      "grad_norm": 0.6214264631271362,
+      "learning_rate": 8.412572880824168e-06,
+      "loss": 0.7436,
+      "step": 9554
+    },
+    {
+      "epoch": 0.5258957565083384,
+      "grad_norm": 0.6479233503341675,
+      "learning_rate": 8.412256059395274e-06,
+      "loss": 0.7359,
+      "step": 9555
+    },
+    {
+      "epoch": 0.525950795310694,
+      "grad_norm": 0.7596501111984253,
+      "learning_rate": 8.411939212320778e-06,
+      "loss": 0.7422,
+      "step": 9556
+    },
+    {
+      "epoch": 0.5260058341130497,
+      "grad_norm": 0.8040934205055237,
+      "learning_rate": 8.41162233960306e-06,
+      "loss": 0.7721,
+      "step": 9557
+    },
+    {
+      "epoch": 0.5260608729154054,
+      "grad_norm": 0.7190027832984924,
+      "learning_rate": 8.411305441244505e-06,
+      "loss": 0.8794,
+      "step": 9558
+    },
+    {
+      "epoch": 0.5261159117177611,
+      "grad_norm": 0.8002649545669556,
+      "learning_rate": 8.410988517247486e-06,
+      "loss": 0.7958,
+      "step": 9559
+    },
+    {
+      "epoch": 0.5261709505201166,
+      "grad_norm": 0.7151750326156616,
+      "learning_rate": 8.410671567614394e-06,
+      "loss": 0.7597,
+      "step": 9560
+    },
+    {
+      "epoch": 0.5262259893224723,
+      "grad_norm": 0.9718102812767029,
+      "learning_rate": 8.410354592347607e-06,
+      "loss": 0.8272,
+      "step": 9561
+    },
+    {
+      "epoch": 0.526281028124828,
+      "grad_norm": 0.701932966709137,
+      "learning_rate": 8.410037591449506e-06,
+      "loss": 0.808,
+      "step": 9562
+    },
+    {
+      "epoch": 0.5263360669271837,
+      "grad_norm": 0.8247585296630859,
+      "learning_rate": 8.409720564922476e-06,
+      "loss": 0.7598,
+      "step": 9563
+    },
+    {
+      "epoch": 0.5263911057295393,
+      "grad_norm": 0.7305104732513428,
+      "learning_rate": 8.409403512768899e-06,
+      "loss": 0.8161,
+      "step": 9564
+    },
+    {
+      "epoch": 0.526446144531895,
+      "grad_norm": 0.8726410865783691,
+      "learning_rate": 8.409086434991158e-06,
+      "loss": 0.8598,
+      "step": 9565
+    },
+    {
+      "epoch": 0.5265011833342507,
+      "grad_norm": 0.7329155802726746,
+      "learning_rate": 8.408769331591637e-06,
+      "loss": 0.7355,
+      "step": 9566
+    },
+    {
+      "epoch": 0.5265562221366064,
+      "grad_norm": 0.8227902054786682,
+      "learning_rate": 8.408452202572716e-06,
+      "loss": 0.7888,
+      "step": 9567
+    },
+    {
+      "epoch": 0.5266112609389619,
+      "grad_norm": 0.7190666794776917,
+      "learning_rate": 8.408135047936783e-06,
+      "loss": 0.669,
+      "step": 9568
+    },
+    {
+      "epoch": 0.5266662997413176,
+      "grad_norm": 0.6529938578605652,
+      "learning_rate": 8.407817867686217e-06,
+      "loss": 0.7345,
+      "step": 9569
+    },
+    {
+      "epoch": 0.5267213385436733,
+      "grad_norm": 0.6985379457473755,
+      "learning_rate": 8.407500661823407e-06,
+      "loss": 0.852,
+      "step": 9570
+    },
+    {
+      "epoch": 0.526776377346029,
+      "grad_norm": 0.7480047345161438,
+      "learning_rate": 8.407183430350732e-06,
+      "loss": 0.7422,
+      "step": 9571
+    },
+    {
+      "epoch": 0.5268314161483846,
+      "grad_norm": 0.7599420547485352,
+      "learning_rate": 8.406866173270579e-06,
+      "loss": 0.7499,
+      "step": 9572
+    },
+    {
+      "epoch": 0.5268864549507403,
+      "grad_norm": 0.813448965549469,
+      "learning_rate": 8.406548890585331e-06,
+      "loss": 0.7979,
+      "step": 9573
+    },
+    {
+      "epoch": 0.526941493753096,
+      "grad_norm": 0.6029278039932251,
+      "learning_rate": 8.406231582297374e-06,
+      "loss": 0.7289,
+      "step": 9574
+    },
+    {
+      "epoch": 0.5269965325554516,
+      "grad_norm": 0.656829297542572,
+      "learning_rate": 8.40591424840909e-06,
+      "loss": 0.6778,
+      "step": 9575
+    },
+    {
+      "epoch": 0.5270515713578072,
+      "grad_norm": 0.7147198915481567,
+      "learning_rate": 8.405596888922869e-06,
+      "loss": 0.7212,
+      "step": 9576
+    },
+    {
+      "epoch": 0.5271066101601629,
+      "grad_norm": 0.7722035050392151,
+      "learning_rate": 8.405279503841094e-06,
+      "loss": 0.8008,
+      "step": 9577
+    },
+    {
+      "epoch": 0.5271616489625186,
+      "grad_norm": 0.6828493475914001,
+      "learning_rate": 8.40496209316615e-06,
+      "loss": 0.787,
+      "step": 9578
+    },
+    {
+      "epoch": 0.5272166877648742,
+      "grad_norm": 0.6965187788009644,
+      "learning_rate": 8.40464465690042e-06,
+      "loss": 0.6803,
+      "step": 9579
+    },
+    {
+      "epoch": 0.5272717265672299,
+      "grad_norm": 0.7300547957420349,
+      "learning_rate": 8.404327195046293e-06,
+      "loss": 0.8165,
+      "step": 9580
+    },
+    {
+      "epoch": 0.5273267653695856,
+      "grad_norm": 0.7367526292800903,
+      "learning_rate": 8.404009707606153e-06,
+      "loss": 0.7709,
+      "step": 9581
+    },
+    {
+      "epoch": 0.5273818041719412,
+      "grad_norm": 0.6694689989089966,
+      "learning_rate": 8.40369219458239e-06,
+      "loss": 0.7971,
+      "step": 9582
+    },
+    {
+      "epoch": 0.5274368429742968,
+      "grad_norm": 0.6723141074180603,
+      "learning_rate": 8.403374655977384e-06,
+      "loss": 0.695,
+      "step": 9583
+    },
+    {
+      "epoch": 0.5274918817766525,
+      "grad_norm": 0.7737089395523071,
+      "learning_rate": 8.403057091793528e-06,
+      "loss": 0.7765,
+      "step": 9584
+    },
+    {
+      "epoch": 0.5275469205790082,
+      "grad_norm": 0.8378487825393677,
+      "learning_rate": 8.402739502033204e-06,
+      "loss": 0.7984,
+      "step": 9585
+    },
+    {
+      "epoch": 0.5276019593813639,
+      "grad_norm": 0.7496509552001953,
+      "learning_rate": 8.402421886698802e-06,
+      "loss": 0.7846,
+      "step": 9586
+    },
+    {
+      "epoch": 0.5276569981837195,
+      "grad_norm": 0.7020435929298401,
+      "learning_rate": 8.402104245792706e-06,
+      "loss": 0.8102,
+      "step": 9587
+    },
+    {
+      "epoch": 0.5277120369860752,
+      "grad_norm": 0.8877277374267578,
+      "learning_rate": 8.401786579317308e-06,
+      "loss": 0.6995,
+      "step": 9588
+    },
+    {
+      "epoch": 0.5277670757884309,
+      "grad_norm": 0.6975196599960327,
+      "learning_rate": 8.401468887274991e-06,
+      "loss": 0.7475,
+      "step": 9589
+    },
+    {
+      "epoch": 0.5278221145907865,
+      "grad_norm": 0.8267357349395752,
+      "learning_rate": 8.401151169668144e-06,
+      "loss": 0.7091,
+      "step": 9590
+    },
+    {
+      "epoch": 0.5278771533931421,
+      "grad_norm": 0.6778179407119751,
+      "learning_rate": 8.400833426499156e-06,
+      "loss": 0.8198,
+      "step": 9591
+    },
+    {
+      "epoch": 0.5279321921954978,
+      "grad_norm": 0.7343330979347229,
+      "learning_rate": 8.400515657770414e-06,
+      "loss": 0.7565,
+      "step": 9592
+    },
+    {
+      "epoch": 0.5279872309978535,
+      "grad_norm": 0.7745271325111389,
+      "learning_rate": 8.400197863484307e-06,
+      "loss": 0.7991,
+      "step": 9593
+    },
+    {
+      "epoch": 0.5280422698002092,
+      "grad_norm": 0.7652345895767212,
+      "learning_rate": 8.399880043643224e-06,
+      "loss": 0.7752,
+      "step": 9594
+    },
+    {
+      "epoch": 0.5280973086025648,
+      "grad_norm": 0.9764432311058044,
+      "learning_rate": 8.399562198249551e-06,
+      "loss": 0.784,
+      "step": 9595
+    },
+    {
+      "epoch": 0.5281523474049205,
+      "grad_norm": 0.6763052940368652,
+      "learning_rate": 8.399244327305678e-06,
+      "loss": 0.7695,
+      "step": 9596
+    },
+    {
+      "epoch": 0.5282073862072761,
+      "grad_norm": 0.7788934111595154,
+      "learning_rate": 8.398926430813996e-06,
+      "loss": 0.8152,
+      "step": 9597
+    },
+    {
+      "epoch": 0.5282624250096318,
+      "grad_norm": 0.8088317513465881,
+      "learning_rate": 8.398608508776894e-06,
+      "loss": 0.7751,
+      "step": 9598
+    },
+    {
+      "epoch": 0.5283174638119874,
+      "grad_norm": 0.6735319495201111,
+      "learning_rate": 8.398290561196756e-06,
+      "loss": 0.7305,
+      "step": 9599
+    },
+    {
+      "epoch": 0.5283725026143431,
+      "grad_norm": 0.7279297113418579,
+      "learning_rate": 8.39797258807598e-06,
+      "loss": 0.7381,
+      "step": 9600
+    },
+    {
+      "epoch": 0.5284275414166988,
+      "grad_norm": 0.74604332447052,
+      "learning_rate": 8.39765458941695e-06,
+      "loss": 0.8138,
+      "step": 9601
+    },
+    {
+      "epoch": 0.5284825802190545,
+      "grad_norm": 0.7735850214958191,
+      "learning_rate": 8.397336565222057e-06,
+      "loss": 0.7364,
+      "step": 9602
+    },
+    {
+      "epoch": 0.52853761902141,
+      "grad_norm": 0.7890003323554993,
+      "learning_rate": 8.397018515493693e-06,
+      "loss": 0.8301,
+      "step": 9603
+    },
+    {
+      "epoch": 0.5285926578237657,
+      "grad_norm": 0.739054262638092,
+      "learning_rate": 8.396700440234245e-06,
+      "loss": 0.7503,
+      "step": 9604
+    },
+    {
+      "epoch": 0.5286476966261214,
+      "grad_norm": 0.7611023783683777,
+      "learning_rate": 8.396382339446108e-06,
+      "loss": 0.7225,
+      "step": 9605
+    },
+    {
+      "epoch": 0.5287027354284771,
+      "grad_norm": 0.770602285861969,
+      "learning_rate": 8.39606421313167e-06,
+      "loss": 0.71,
+      "step": 9606
+    },
+    {
+      "epoch": 0.5287577742308327,
+      "grad_norm": 0.7495261430740356,
+      "learning_rate": 8.395746061293322e-06,
+      "loss": 0.7729,
+      "step": 9607
+    },
+    {
+      "epoch": 0.5288128130331884,
+      "grad_norm": 0.7159668207168579,
+      "learning_rate": 8.395427883933456e-06,
+      "loss": 0.8457,
+      "step": 9608
+    },
+    {
+      "epoch": 0.5288678518355441,
+      "grad_norm": 0.7663426399230957,
+      "learning_rate": 8.395109681054463e-06,
+      "loss": 0.784,
+      "step": 9609
+    },
+    {
+      "epoch": 0.5289228906378998,
+      "grad_norm": 0.7271933555603027,
+      "learning_rate": 8.394791452658732e-06,
+      "loss": 0.7981,
+      "step": 9610
+    },
+    {
+      "epoch": 0.5289779294402553,
+      "grad_norm": 0.7782096266746521,
+      "learning_rate": 8.394473198748661e-06,
+      "loss": 0.7953,
+      "step": 9611
+    },
+    {
+      "epoch": 0.529032968242611,
+      "grad_norm": 0.8318955302238464,
+      "learning_rate": 8.394154919326636e-06,
+      "loss": 0.6875,
+      "step": 9612
+    },
+    {
+      "epoch": 0.5290880070449667,
+      "grad_norm": 0.7402167916297913,
+      "learning_rate": 8.393836614395051e-06,
+      "loss": 0.7805,
+      "step": 9613
+    },
+    {
+      "epoch": 0.5291430458473224,
+      "grad_norm": 0.6314370632171631,
+      "learning_rate": 8.393518283956299e-06,
+      "loss": 0.6841,
+      "step": 9614
+    },
+    {
+      "epoch": 0.529198084649678,
+      "grad_norm": 0.8387365937232971,
+      "learning_rate": 8.393199928012772e-06,
+      "loss": 0.8503,
+      "step": 9615
+    },
+    {
+      "epoch": 0.5292531234520337,
+      "grad_norm": 0.7066243886947632,
+      "learning_rate": 8.392881546566863e-06,
+      "loss": 0.8494,
+      "step": 9616
+    },
+    {
+      "epoch": 0.5293081622543894,
+      "grad_norm": 0.7034226059913635,
+      "learning_rate": 8.392563139620964e-06,
+      "loss": 0.7335,
+      "step": 9617
+    },
+    {
+      "epoch": 0.5293632010567451,
+      "grad_norm": 0.6969622373580933,
+      "learning_rate": 8.392244707177468e-06,
+      "loss": 0.7203,
+      "step": 9618
+    },
+    {
+      "epoch": 0.5294182398591006,
+      "grad_norm": 0.7694050073623657,
+      "learning_rate": 8.391926249238768e-06,
+      "loss": 0.7864,
+      "step": 9619
+    },
+    {
+      "epoch": 0.5294732786614563,
+      "grad_norm": 0.7284281253814697,
+      "learning_rate": 8.391607765807262e-06,
+      "loss": 0.6704,
+      "step": 9620
+    },
+    {
+      "epoch": 0.529528317463812,
+      "grad_norm": 1.0466688871383667,
+      "learning_rate": 8.391289256885337e-06,
+      "loss": 0.7807,
+      "step": 9621
+    },
+    {
+      "epoch": 0.5295833562661676,
+      "grad_norm": 0.7118388414382935,
+      "learning_rate": 8.39097072247539e-06,
+      "loss": 0.738,
+      "step": 9622
+    },
+    {
+      "epoch": 0.5296383950685233,
+      "grad_norm": 0.794377863407135,
+      "learning_rate": 8.390652162579815e-06,
+      "loss": 0.6831,
+      "step": 9623
+    },
+    {
+      "epoch": 0.529693433870879,
+      "grad_norm": 0.6042492389678955,
+      "learning_rate": 8.390333577201007e-06,
+      "loss": 0.6773,
+      "step": 9624
+    },
+    {
+      "epoch": 0.5297484726732347,
+      "grad_norm": 0.6452521681785583,
+      "learning_rate": 8.390014966341357e-06,
+      "loss": 0.7168,
+      "step": 9625
+    },
+    {
+      "epoch": 0.5298035114755902,
+      "grad_norm": 0.7113651633262634,
+      "learning_rate": 8.389696330003265e-06,
+      "loss": 0.709,
+      "step": 9626
+    },
+    {
+      "epoch": 0.5298585502779459,
+      "grad_norm": 0.6469250917434692,
+      "learning_rate": 8.38937766818912e-06,
+      "loss": 0.6804,
+      "step": 9627
+    },
+    {
+      "epoch": 0.5299135890803016,
+      "grad_norm": 0.7529417872428894,
+      "learning_rate": 8.389058980901322e-06,
+      "loss": 0.8537,
+      "step": 9628
+    },
+    {
+      "epoch": 0.5299686278826573,
+      "grad_norm": 0.7681186199188232,
+      "learning_rate": 8.388740268142262e-06,
+      "loss": 0.7383,
+      "step": 9629
+    },
+    {
+      "epoch": 0.5300236666850129,
+      "grad_norm": 0.6585648655891418,
+      "learning_rate": 8.388421529914337e-06,
+      "loss": 0.7535,
+      "step": 9630
+    },
+    {
+      "epoch": 0.5300787054873686,
+      "grad_norm": 0.7432085871696472,
+      "learning_rate": 8.388102766219943e-06,
+      "loss": 0.7391,
+      "step": 9631
+    },
+    {
+      "epoch": 0.5301337442897243,
+      "grad_norm": 0.6672815084457397,
+      "learning_rate": 8.387783977061476e-06,
+      "loss": 0.8056,
+      "step": 9632
+    },
+    {
+      "epoch": 0.53018878309208,
+      "grad_norm": 0.7566675543785095,
+      "learning_rate": 8.387465162441332e-06,
+      "loss": 0.7858,
+      "step": 9633
+    },
+    {
+      "epoch": 0.5302438218944355,
+      "grad_norm": 0.6522077322006226,
+      "learning_rate": 8.387146322361907e-06,
+      "loss": 0.759,
+      "step": 9634
+    },
+    {
+      "epoch": 0.5302988606967912,
+      "grad_norm": 0.7246397137641907,
+      "learning_rate": 8.386827456825597e-06,
+      "loss": 0.8158,
+      "step": 9635
+    },
+    {
+      "epoch": 0.5303538994991469,
+      "grad_norm": 0.7577807307243347,
+      "learning_rate": 8.386508565834797e-06,
+      "loss": 0.7495,
+      "step": 9636
+    },
+    {
+      "epoch": 0.5304089383015026,
+      "grad_norm": 0.7080703973770142,
+      "learning_rate": 8.386189649391906e-06,
+      "loss": 0.8086,
+      "step": 9637
+    },
+    {
+      "epoch": 0.5304639771038582,
+      "grad_norm": 0.7505277395248413,
+      "learning_rate": 8.385870707499321e-06,
+      "loss": 0.7206,
+      "step": 9638
+    },
+    {
+      "epoch": 0.5305190159062139,
+      "grad_norm": 0.7044165134429932,
+      "learning_rate": 8.385551740159437e-06,
+      "loss": 0.7838,
+      "step": 9639
+    },
+    {
+      "epoch": 0.5305740547085696,
+      "grad_norm": 0.7921645641326904,
+      "learning_rate": 8.385232747374652e-06,
+      "loss": 0.7604,
+      "step": 9640
+    },
+    {
+      "epoch": 0.5306290935109252,
+      "grad_norm": 0.9930111169815063,
+      "learning_rate": 8.384913729147364e-06,
+      "loss": 0.7839,
+      "step": 9641
+    },
+    {
+      "epoch": 0.5306841323132808,
+      "grad_norm": 0.7333244681358337,
+      "learning_rate": 8.38459468547997e-06,
+      "loss": 0.7941,
+      "step": 9642
+    },
+    {
+      "epoch": 0.5307391711156365,
+      "grad_norm": 0.7857590913772583,
+      "learning_rate": 8.384275616374868e-06,
+      "loss": 0.8535,
+      "step": 9643
+    },
+    {
+      "epoch": 0.5307942099179922,
+      "grad_norm": 0.8568746447563171,
+      "learning_rate": 8.383956521834459e-06,
+      "loss": 0.6586,
+      "step": 9644
+    },
+    {
+      "epoch": 0.5308492487203479,
+      "grad_norm": 0.7061276435852051,
+      "learning_rate": 8.383637401861136e-06,
+      "loss": 0.7288,
+      "step": 9645
+    },
+    {
+      "epoch": 0.5309042875227035,
+      "grad_norm": 0.7348940968513489,
+      "learning_rate": 8.383318256457303e-06,
+      "loss": 0.8099,
+      "step": 9646
+    },
+    {
+      "epoch": 0.5309593263250592,
+      "grad_norm": 0.6526725888252258,
+      "learning_rate": 8.382999085625353e-06,
+      "loss": 0.6702,
+      "step": 9647
+    },
+    {
+      "epoch": 0.5310143651274148,
+      "grad_norm": 0.8122747540473938,
+      "learning_rate": 8.382679889367687e-06,
+      "loss": 0.67,
+      "step": 9648
+    },
+    {
+      "epoch": 0.5310694039297705,
+      "grad_norm": 0.9145376682281494,
+      "learning_rate": 8.382360667686706e-06,
+      "loss": 0.7719,
+      "step": 9649
+    },
+    {
+      "epoch": 0.5311244427321261,
+      "grad_norm": 0.6659818887710571,
+      "learning_rate": 8.382041420584807e-06,
+      "loss": 0.806,
+      "step": 9650
+    },
+    {
+      "epoch": 0.5311794815344818,
+      "grad_norm": 0.7088539004325867,
+      "learning_rate": 8.381722148064391e-06,
+      "loss": 0.7046,
+      "step": 9651
+    },
+    {
+      "epoch": 0.5312345203368375,
+      "grad_norm": 0.8610590696334839,
+      "learning_rate": 8.381402850127854e-06,
+      "loss": 0.6998,
+      "step": 9652
+    },
+    {
+      "epoch": 0.5312895591391932,
+      "grad_norm": 0.775830864906311,
+      "learning_rate": 8.3810835267776e-06,
+      "loss": 0.8874,
+      "step": 9653
+    },
+    {
+      "epoch": 0.5313445979415488,
+      "grad_norm": 0.6871606707572937,
+      "learning_rate": 8.380764178016028e-06,
+      "loss": 0.7903,
+      "step": 9654
+    },
+    {
+      "epoch": 0.5313996367439044,
+      "grad_norm": 0.7005272507667542,
+      "learning_rate": 8.380444803845537e-06,
+      "loss": 0.6685,
+      "step": 9655
+    },
+    {
+      "epoch": 0.5314546755462601,
+      "grad_norm": 0.8922042846679688,
+      "learning_rate": 8.380125404268527e-06,
+      "loss": 0.7797,
+      "step": 9656
+    },
+    {
+      "epoch": 0.5315097143486158,
+      "grad_norm": 0.7242267727851868,
+      "learning_rate": 8.3798059792874e-06,
+      "loss": 0.863,
+      "step": 9657
+    },
+    {
+      "epoch": 0.5315647531509714,
+      "grad_norm": 0.6625328660011292,
+      "learning_rate": 8.379486528904555e-06,
+      "loss": 0.7,
+      "step": 9658
+    },
+    {
+      "epoch": 0.5316197919533271,
+      "grad_norm": 0.9882226586341858,
+      "learning_rate": 8.379167053122394e-06,
+      "loss": 0.7534,
+      "step": 9659
+    },
+    {
+      "epoch": 0.5316748307556828,
+      "grad_norm": 0.6894702911376953,
+      "learning_rate": 8.378847551943318e-06,
+      "loss": 0.7503,
+      "step": 9660
+    },
+    {
+      "epoch": 0.5317298695580385,
+      "grad_norm": 0.6820259690284729,
+      "learning_rate": 8.37852802536973e-06,
+      "loss": 0.7713,
+      "step": 9661
+    },
+    {
+      "epoch": 0.531784908360394,
+      "grad_norm": 0.667918860912323,
+      "learning_rate": 8.378208473404028e-06,
+      "loss": 0.7524,
+      "step": 9662
+    },
+    {
+      "epoch": 0.5318399471627497,
+      "grad_norm": 0.7789241075515747,
+      "learning_rate": 8.377888896048617e-06,
+      "loss": 0.6906,
+      "step": 9663
+    },
+    {
+      "epoch": 0.5318949859651054,
+      "grad_norm": 0.7264542579650879,
+      "learning_rate": 8.377569293305894e-06,
+      "loss": 0.7836,
+      "step": 9664
+    },
+    {
+      "epoch": 0.531950024767461,
+      "grad_norm": 0.6979835629463196,
+      "learning_rate": 8.377249665178267e-06,
+      "loss": 0.7739,
+      "step": 9665
+    },
+    {
+      "epoch": 0.5320050635698167,
+      "grad_norm": 0.8008072376251221,
+      "learning_rate": 8.376930011668136e-06,
+      "loss": 0.7853,
+      "step": 9666
+    },
+    {
+      "epoch": 0.5320601023721724,
+      "grad_norm": 0.7185621857643127,
+      "learning_rate": 8.376610332777901e-06,
+      "loss": 0.7311,
+      "step": 9667
+    },
+    {
+      "epoch": 0.5321151411745281,
+      "grad_norm": 0.7644047141075134,
+      "learning_rate": 8.376290628509969e-06,
+      "loss": 0.6919,
+      "step": 9668
+    },
+    {
+      "epoch": 0.5321701799768837,
+      "grad_norm": 0.7387600541114807,
+      "learning_rate": 8.37597089886674e-06,
+      "loss": 0.7285,
+      "step": 9669
+    },
+    {
+      "epoch": 0.5322252187792393,
+      "grad_norm": 0.7344895005226135,
+      "learning_rate": 8.375651143850614e-06,
+      "loss": 0.7514,
+      "step": 9670
+    },
+    {
+      "epoch": 0.532280257581595,
+      "grad_norm": 0.6930707097053528,
+      "learning_rate": 8.375331363464002e-06,
+      "loss": 0.8318,
+      "step": 9671
+    },
+    {
+      "epoch": 0.5323352963839507,
+      "grad_norm": 0.678162693977356,
+      "learning_rate": 8.3750115577093e-06,
+      "loss": 0.7123,
+      "step": 9672
+    },
+    {
+      "epoch": 0.5323903351863063,
+      "grad_norm": 0.7780481576919556,
+      "learning_rate": 8.374691726588914e-06,
+      "loss": 0.7672,
+      "step": 9673
+    },
+    {
+      "epoch": 0.532445373988662,
+      "grad_norm": 0.6664674282073975,
+      "learning_rate": 8.374371870105252e-06,
+      "loss": 0.6994,
+      "step": 9674
+    },
+    {
+      "epoch": 0.5325004127910177,
+      "grad_norm": 0.6952562928199768,
+      "learning_rate": 8.374051988260712e-06,
+      "loss": 0.8638,
+      "step": 9675
+    },
+    {
+      "epoch": 0.5325554515933734,
+      "grad_norm": 0.764005184173584,
+      "learning_rate": 8.373732081057699e-06,
+      "loss": 0.756,
+      "step": 9676
+    },
+    {
+      "epoch": 0.5326104903957289,
+      "grad_norm": 0.9434393048286438,
+      "learning_rate": 8.373412148498621e-06,
+      "loss": 0.8668,
+      "step": 9677
+    },
+    {
+      "epoch": 0.5326655291980846,
+      "grad_norm": 0.752609133720398,
+      "learning_rate": 8.373092190585878e-06,
+      "loss": 0.8078,
+      "step": 9678
+    },
+    {
+      "epoch": 0.5327205680004403,
+      "grad_norm": 0.671940803527832,
+      "learning_rate": 8.37277220732188e-06,
+      "loss": 0.7726,
+      "step": 9679
+    },
+    {
+      "epoch": 0.532775606802796,
+      "grad_norm": 0.7824863791465759,
+      "learning_rate": 8.372452198709027e-06,
+      "loss": 0.8246,
+      "step": 9680
+    },
+    {
+      "epoch": 0.5328306456051516,
+      "grad_norm": 0.7300587892532349,
+      "learning_rate": 8.372132164749726e-06,
+      "loss": 0.7953,
+      "step": 9681
+    },
+    {
+      "epoch": 0.5328856844075073,
+      "grad_norm": 0.7146018743515015,
+      "learning_rate": 8.371812105446384e-06,
+      "loss": 0.7409,
+      "step": 9682
+    },
+    {
+      "epoch": 0.532940723209863,
+      "grad_norm": 0.73857581615448,
+      "learning_rate": 8.371492020801404e-06,
+      "loss": 0.8067,
+      "step": 9683
+    },
+    {
+      "epoch": 0.5329957620122187,
+      "grad_norm": 0.6760877966880798,
+      "learning_rate": 8.37117191081719e-06,
+      "loss": 0.7363,
+      "step": 9684
+    },
+    {
+      "epoch": 0.5330508008145742,
+      "grad_norm": 0.766482412815094,
+      "learning_rate": 8.370851775496154e-06,
+      "loss": 0.7358,
+      "step": 9685
+    },
+    {
+      "epoch": 0.5331058396169299,
+      "grad_norm": 0.7230576276779175,
+      "learning_rate": 8.370531614840697e-06,
+      "loss": 0.8154,
+      "step": 9686
+    },
+    {
+      "epoch": 0.5331608784192856,
+      "grad_norm": 0.7357933521270752,
+      "learning_rate": 8.370211428853225e-06,
+      "loss": 0.7187,
+      "step": 9687
+    },
+    {
+      "epoch": 0.5332159172216413,
+      "grad_norm": 0.8208534121513367,
+      "learning_rate": 8.369891217536148e-06,
+      "loss": 0.8037,
+      "step": 9688
+    },
+    {
+      "epoch": 0.5332709560239969,
+      "grad_norm": 0.6771863698959351,
+      "learning_rate": 8.36957098089187e-06,
+      "loss": 0.733,
+      "step": 9689
+    },
+    {
+      "epoch": 0.5333259948263526,
+      "grad_norm": 0.6382480263710022,
+      "learning_rate": 8.369250718922798e-06,
+      "loss": 0.7391,
+      "step": 9690
+    },
+    {
+      "epoch": 0.5333810336287083,
+      "grad_norm": 0.6638994812965393,
+      "learning_rate": 8.368930431631342e-06,
+      "loss": 0.7176,
+      "step": 9691
+    },
+    {
+      "epoch": 0.533436072431064,
+      "grad_norm": 0.7599604725837708,
+      "learning_rate": 8.368610119019903e-06,
+      "loss": 0.8814,
+      "step": 9692
+    },
+    {
+      "epoch": 0.5334911112334195,
+      "grad_norm": 0.6896547079086304,
+      "learning_rate": 8.368289781090894e-06,
+      "loss": 0.7618,
+      "step": 9693
+    },
+    {
+      "epoch": 0.5335461500357752,
+      "grad_norm": 0.7081224918365479,
+      "learning_rate": 8.36796941784672e-06,
+      "loss": 0.656,
+      "step": 9694
+    },
+    {
+      "epoch": 0.5336011888381309,
+      "grad_norm": 0.8819646835327148,
+      "learning_rate": 8.367649029289791e-06,
+      "loss": 0.8946,
+      "step": 9695
+    },
+    {
+      "epoch": 0.5336562276404866,
+      "grad_norm": 0.6597925424575806,
+      "learning_rate": 8.367328615422512e-06,
+      "loss": 0.6891,
+      "step": 9696
+    },
+    {
+      "epoch": 0.5337112664428422,
+      "grad_norm": 0.6855770945549011,
+      "learning_rate": 8.367008176247294e-06,
+      "loss": 0.7158,
+      "step": 9697
+    },
+    {
+      "epoch": 0.5337663052451979,
+      "grad_norm": 0.6874905228614807,
+      "learning_rate": 8.366687711766541e-06,
+      "loss": 0.7445,
+      "step": 9698
+    },
+    {
+      "epoch": 0.5338213440475535,
+      "grad_norm": 0.6990895867347717,
+      "learning_rate": 8.366367221982666e-06,
+      "loss": 0.6189,
+      "step": 9699
+    },
+    {
+      "epoch": 0.5338763828499092,
+      "grad_norm": 0.7235365509986877,
+      "learning_rate": 8.366046706898075e-06,
+      "loss": 0.6406,
+      "step": 9700
+    },
+    {
+      "epoch": 0.5339314216522648,
+      "grad_norm": 0.7563154697418213,
+      "learning_rate": 8.36572616651518e-06,
+      "loss": 0.7798,
+      "step": 9701
+    },
+    {
+      "epoch": 0.5339864604546205,
+      "grad_norm": 0.6845980286598206,
+      "learning_rate": 8.365405600836387e-06,
+      "loss": 0.7665,
+      "step": 9702
+    },
+    {
+      "epoch": 0.5340414992569762,
+      "grad_norm": 0.6374378204345703,
+      "learning_rate": 8.365085009864106e-06,
+      "loss": 0.6935,
+      "step": 9703
+    },
+    {
+      "epoch": 0.5340965380593319,
+      "grad_norm": 0.726672887802124,
+      "learning_rate": 8.364764393600747e-06,
+      "loss": 0.7821,
+      "step": 9704
+    },
+    {
+      "epoch": 0.5341515768616875,
+      "grad_norm": 0.6784456372261047,
+      "learning_rate": 8.364443752048719e-06,
+      "loss": 0.7722,
+      "step": 9705
+    },
+    {
+      "epoch": 0.5342066156640431,
+      "grad_norm": 0.6344080567359924,
+      "learning_rate": 8.364123085210433e-06,
+      "loss": 0.7256,
+      "step": 9706
+    },
+    {
+      "epoch": 0.5342616544663988,
+      "grad_norm": 0.7913152575492859,
+      "learning_rate": 8.363802393088299e-06,
+      "loss": 0.7892,
+      "step": 9707
+    },
+    {
+      "epoch": 0.5343166932687544,
+      "grad_norm": 0.6792107820510864,
+      "learning_rate": 8.363481675684726e-06,
+      "loss": 0.7374,
+      "step": 9708
+    },
+    {
+      "epoch": 0.5343717320711101,
+      "grad_norm": 1.0153685808181763,
+      "learning_rate": 8.363160933002126e-06,
+      "loss": 0.7396,
+      "step": 9709
+    },
+    {
+      "epoch": 0.5344267708734658,
+      "grad_norm": 0.7655258774757385,
+      "learning_rate": 8.362840165042906e-06,
+      "loss": 0.7746,
+      "step": 9710
+    },
+    {
+      "epoch": 0.5344818096758215,
+      "grad_norm": 0.7830179929733276,
+      "learning_rate": 8.362519371809483e-06,
+      "loss": 0.7082,
+      "step": 9711
+    },
+    {
+      "epoch": 0.5345368484781771,
+      "grad_norm": 0.7410556674003601,
+      "learning_rate": 8.362198553304261e-06,
+      "loss": 0.7055,
+      "step": 9712
+    },
+    {
+      "epoch": 0.5345918872805328,
+      "grad_norm": 0.6542297005653381,
+      "learning_rate": 8.361877709529658e-06,
+      "loss": 0.7153,
+      "step": 9713
+    },
+    {
+      "epoch": 0.5346469260828884,
+      "grad_norm": 0.6752653121948242,
+      "learning_rate": 8.36155684048808e-06,
+      "loss": 0.6901,
+      "step": 9714
+    },
+    {
+      "epoch": 0.5347019648852441,
+      "grad_norm": 0.7158684134483337,
+      "learning_rate": 8.361235946181943e-06,
+      "loss": 0.7775,
+      "step": 9715
+    },
+    {
+      "epoch": 0.5347570036875997,
+      "grad_norm": 0.6174392700195312,
+      "learning_rate": 8.360915026613652e-06,
+      "loss": 0.6501,
+      "step": 9716
+    },
+    {
+      "epoch": 0.5348120424899554,
+      "grad_norm": 0.7110500931739807,
+      "learning_rate": 8.360594081785627e-06,
+      "loss": 0.742,
+      "step": 9717
+    },
+    {
+      "epoch": 0.5348670812923111,
+      "grad_norm": 0.8456488251686096,
+      "learning_rate": 8.360273111700276e-06,
+      "loss": 0.8237,
+      "step": 9718
+    },
+    {
+      "epoch": 0.5349221200946668,
+      "grad_norm": 0.6660711169242859,
+      "learning_rate": 8.359952116360011e-06,
+      "loss": 0.7856,
+      "step": 9719
+    },
+    {
+      "epoch": 0.5349771588970224,
+      "grad_norm": 0.7661204934120178,
+      "learning_rate": 8.359631095767244e-06,
+      "loss": 0.8336,
+      "step": 9720
+    },
+    {
+      "epoch": 0.535032197699378,
+      "grad_norm": 0.7747855186462402,
+      "learning_rate": 8.359310049924392e-06,
+      "loss": 0.7302,
+      "step": 9721
+    },
+    {
+      "epoch": 0.5350872365017337,
+      "grad_norm": 0.8156001567840576,
+      "learning_rate": 8.358988978833864e-06,
+      "loss": 0.7878,
+      "step": 9722
+    },
+    {
+      "epoch": 0.5351422753040894,
+      "grad_norm": 0.7371010780334473,
+      "learning_rate": 8.358667882498073e-06,
+      "loss": 0.803,
+      "step": 9723
+    },
+    {
+      "epoch": 0.535197314106445,
+      "grad_norm": 0.7141744494438171,
+      "learning_rate": 8.358346760919431e-06,
+      "loss": 0.687,
+      "step": 9724
+    },
+    {
+      "epoch": 0.5352523529088007,
+      "grad_norm": 0.6395956873893738,
+      "learning_rate": 8.358025614100358e-06,
+      "loss": 0.7052,
+      "step": 9725
+    },
+    {
+      "epoch": 0.5353073917111564,
+      "grad_norm": 0.7135289311408997,
+      "learning_rate": 8.35770444204326e-06,
+      "loss": 0.7882,
+      "step": 9726
+    },
+    {
+      "epoch": 0.5353624305135121,
+      "grad_norm": 0.702408492565155,
+      "learning_rate": 8.357383244750557e-06,
+      "loss": 0.6965,
+      "step": 9727
+    },
+    {
+      "epoch": 0.5354174693158676,
+      "grad_norm": 0.731193482875824,
+      "learning_rate": 8.357062022224658e-06,
+      "loss": 0.7525,
+      "step": 9728
+    },
+    {
+      "epoch": 0.5354725081182233,
+      "grad_norm": 0.8115057945251465,
+      "learning_rate": 8.356740774467982e-06,
+      "loss": 0.7466,
+      "step": 9729
+    },
+    {
+      "epoch": 0.535527546920579,
+      "grad_norm": 0.8644380569458008,
+      "learning_rate": 8.356419501482938e-06,
+      "loss": 0.7989,
+      "step": 9730
+    },
+    {
+      "epoch": 0.5355825857229347,
+      "grad_norm": 1.414620041847229,
+      "learning_rate": 8.356098203271945e-06,
+      "loss": 0.7782,
+      "step": 9731
+    },
+    {
+      "epoch": 0.5356376245252903,
+      "grad_norm": 0.7355421185493469,
+      "learning_rate": 8.355776879837417e-06,
+      "loss": 0.7163,
+      "step": 9732
+    },
+    {
+      "epoch": 0.535692663327646,
+      "grad_norm": 0.6556879281997681,
+      "learning_rate": 8.355455531181766e-06,
+      "loss": 0.7543,
+      "step": 9733
+    },
+    {
+      "epoch": 0.5357477021300017,
+      "grad_norm": 0.6632516980171204,
+      "learning_rate": 8.355134157307412e-06,
+      "loss": 0.7382,
+      "step": 9734
+    },
+    {
+      "epoch": 0.5358027409323574,
+      "grad_norm": 0.7096145153045654,
+      "learning_rate": 8.354812758216767e-06,
+      "loss": 0.7797,
+      "step": 9735
+    },
+    {
+      "epoch": 0.5358577797347129,
+      "grad_norm": 0.6404649019241333,
+      "learning_rate": 8.354491333912244e-06,
+      "loss": 0.6637,
+      "step": 9736
+    },
+    {
+      "epoch": 0.5359128185370686,
+      "grad_norm": 0.6987022757530212,
+      "learning_rate": 8.354169884396266e-06,
+      "loss": 0.7682,
+      "step": 9737
+    },
+    {
+      "epoch": 0.5359678573394243,
+      "grad_norm": 0.6593581438064575,
+      "learning_rate": 8.353848409671245e-06,
+      "loss": 0.6747,
+      "step": 9738
+    },
+    {
+      "epoch": 0.53602289614178,
+      "grad_norm": 0.6999880075454712,
+      "learning_rate": 8.353526909739596e-06,
+      "loss": 0.6659,
+      "step": 9739
+    },
+    {
+      "epoch": 0.5360779349441356,
+      "grad_norm": 0.6448989510536194,
+      "learning_rate": 8.353205384603735e-06,
+      "loss": 0.7297,
+      "step": 9740
+    },
+    {
+      "epoch": 0.5361329737464913,
+      "grad_norm": 0.6666765213012695,
+      "learning_rate": 8.352883834266082e-06,
+      "loss": 0.6459,
+      "step": 9741
+    },
+    {
+      "epoch": 0.536188012548847,
+      "grad_norm": 0.8020225763320923,
+      "learning_rate": 8.352562258729051e-06,
+      "loss": 0.8122,
+      "step": 9742
+    },
+    {
+      "epoch": 0.5362430513512026,
+      "grad_norm": 0.6883382201194763,
+      "learning_rate": 8.35224065799506e-06,
+      "loss": 0.7084,
+      "step": 9743
+    },
+    {
+      "epoch": 0.5362980901535582,
+      "grad_norm": 0.7366660237312317,
+      "learning_rate": 8.351919032066525e-06,
+      "loss": 0.848,
+      "step": 9744
+    },
+    {
+      "epoch": 0.5363531289559139,
+      "grad_norm": 0.7408311367034912,
+      "learning_rate": 8.351597380945863e-06,
+      "loss": 0.798,
+      "step": 9745
+    },
+    {
+      "epoch": 0.5364081677582696,
+      "grad_norm": 0.6841676235198975,
+      "learning_rate": 8.351275704635495e-06,
+      "loss": 0.7372,
+      "step": 9746
+    },
+    {
+      "epoch": 0.5364632065606253,
+      "grad_norm": 0.6903505325317383,
+      "learning_rate": 8.350954003137833e-06,
+      "loss": 0.7371,
+      "step": 9747
+    },
+    {
+      "epoch": 0.5365182453629809,
+      "grad_norm": 0.6444700956344604,
+      "learning_rate": 8.350632276455298e-06,
+      "loss": 0.6685,
+      "step": 9748
+    },
+    {
+      "epoch": 0.5365732841653366,
+      "grad_norm": 0.6821029186248779,
+      "learning_rate": 8.350310524590307e-06,
+      "loss": 0.8796,
+      "step": 9749
+    },
+    {
+      "epoch": 0.5366283229676923,
+      "grad_norm": 0.6733999848365784,
+      "learning_rate": 8.349988747545282e-06,
+      "loss": 0.6833,
+      "step": 9750
+    },
+    {
+      "epoch": 0.5366833617700478,
+      "grad_norm": 0.8097321391105652,
+      "learning_rate": 8.349666945322636e-06,
+      "loss": 0.834,
+      "step": 9751
+    },
+    {
+      "epoch": 0.5367384005724035,
+      "grad_norm": 0.7692395448684692,
+      "learning_rate": 8.34934511792479e-06,
+      "loss": 0.7866,
+      "step": 9752
+    },
+    {
+      "epoch": 0.5367934393747592,
+      "grad_norm": 0.7551112174987793,
+      "learning_rate": 8.349023265354164e-06,
+      "loss": 0.8378,
+      "step": 9753
+    },
+    {
+      "epoch": 0.5368484781771149,
+      "grad_norm": 0.5796393156051636,
+      "learning_rate": 8.348701387613176e-06,
+      "loss": 0.5995,
+      "step": 9754
+    },
+    {
+      "epoch": 0.5369035169794705,
+      "grad_norm": 0.6839799284934998,
+      "learning_rate": 8.348379484704244e-06,
+      "loss": 0.8262,
+      "step": 9755
+    },
+    {
+      "epoch": 0.5369585557818262,
+      "grad_norm": 0.7710869908332825,
+      "learning_rate": 8.348057556629786e-06,
+      "loss": 0.7796,
+      "step": 9756
+    },
+    {
+      "epoch": 0.5370135945841819,
+      "grad_norm": 0.733096718788147,
+      "learning_rate": 8.347735603392225e-06,
+      "loss": 0.8233,
+      "step": 9757
+    },
+    {
+      "epoch": 0.5370686333865375,
+      "grad_norm": 0.6438466906547546,
+      "learning_rate": 8.347413624993982e-06,
+      "loss": 0.7582,
+      "step": 9758
+    },
+    {
+      "epoch": 0.5371236721888931,
+      "grad_norm": 0.6877560615539551,
+      "learning_rate": 8.34709162143747e-06,
+      "loss": 0.7428,
+      "step": 9759
+    },
+    {
+      "epoch": 0.5371787109912488,
+      "grad_norm": 1.060831069946289,
+      "learning_rate": 8.346769592725115e-06,
+      "loss": 0.8636,
+      "step": 9760
+    },
+    {
+      "epoch": 0.5372337497936045,
+      "grad_norm": 0.6828434467315674,
+      "learning_rate": 8.346447538859334e-06,
+      "loss": 0.7801,
+      "step": 9761
+    },
+    {
+      "epoch": 0.5372887885959602,
+      "grad_norm": 0.6784753203392029,
+      "learning_rate": 8.346125459842552e-06,
+      "loss": 0.7356,
+      "step": 9762
+    },
+    {
+      "epoch": 0.5373438273983158,
+      "grad_norm": 0.6493560075759888,
+      "learning_rate": 8.345803355677185e-06,
+      "loss": 0.749,
+      "step": 9763
+    },
+    {
+      "epoch": 0.5373988662006715,
+      "grad_norm": 0.7109258770942688,
+      "learning_rate": 8.345481226365657e-06,
+      "loss": 0.7599,
+      "step": 9764
+    },
+    {
+      "epoch": 0.5374539050030271,
+      "grad_norm": 0.8526985049247742,
+      "learning_rate": 8.345159071910387e-06,
+      "loss": 0.6605,
+      "step": 9765
+    },
+    {
+      "epoch": 0.5375089438053828,
+      "grad_norm": 0.9194039702415466,
+      "learning_rate": 8.344836892313797e-06,
+      "loss": 0.794,
+      "step": 9766
+    },
+    {
+      "epoch": 0.5375639826077384,
+      "grad_norm": 0.7258954048156738,
+      "learning_rate": 8.344514687578307e-06,
+      "loss": 0.871,
+      "step": 9767
+    },
+    {
+      "epoch": 0.5376190214100941,
+      "grad_norm": 0.7099377512931824,
+      "learning_rate": 8.34419245770634e-06,
+      "loss": 0.8098,
+      "step": 9768
+    },
+    {
+      "epoch": 0.5376740602124498,
+      "grad_norm": 0.7883020639419556,
+      "learning_rate": 8.34387020270032e-06,
+      "loss": 0.8383,
+      "step": 9769
+    },
+    {
+      "epoch": 0.5377290990148055,
+      "grad_norm": 0.7009730339050293,
+      "learning_rate": 8.343547922562664e-06,
+      "loss": 0.7794,
+      "step": 9770
+    },
+    {
+      "epoch": 0.5377841378171611,
+      "grad_norm": 0.6569581031799316,
+      "learning_rate": 8.343225617295798e-06,
+      "loss": 0.7574,
+      "step": 9771
+    },
+    {
+      "epoch": 0.5378391766195167,
+      "grad_norm": 0.6159278154373169,
+      "learning_rate": 8.342903286902142e-06,
+      "loss": 0.7136,
+      "step": 9772
+    },
+    {
+      "epoch": 0.5378942154218724,
+      "grad_norm": 0.6594879627227783,
+      "learning_rate": 8.342580931384121e-06,
+      "loss": 0.6906,
+      "step": 9773
+    },
+    {
+      "epoch": 0.5379492542242281,
+      "grad_norm": 0.7002933025360107,
+      "learning_rate": 8.342258550744156e-06,
+      "loss": 0.7272,
+      "step": 9774
+    },
+    {
+      "epoch": 0.5380042930265837,
+      "grad_norm": 0.8243216276168823,
+      "learning_rate": 8.341936144984672e-06,
+      "loss": 0.8105,
+      "step": 9775
+    },
+    {
+      "epoch": 0.5380593318289394,
+      "grad_norm": 0.8358921408653259,
+      "learning_rate": 8.34161371410809e-06,
+      "loss": 0.7118,
+      "step": 9776
+    },
+    {
+      "epoch": 0.5381143706312951,
+      "grad_norm": 0.6339066028594971,
+      "learning_rate": 8.34129125811683e-06,
+      "loss": 0.7035,
+      "step": 9777
+    },
+    {
+      "epoch": 0.5381694094336508,
+      "grad_norm": 0.7407625317573547,
+      "learning_rate": 8.340968777013324e-06,
+      "loss": 0.7447,
+      "step": 9778
+    },
+    {
+      "epoch": 0.5382244482360063,
+      "grad_norm": 0.6876600384712219,
+      "learning_rate": 8.340646270799991e-06,
+      "loss": 0.7298,
+      "step": 9779
+    },
+    {
+      "epoch": 0.538279487038362,
+      "grad_norm": 0.7021264433860779,
+      "learning_rate": 8.340323739479251e-06,
+      "loss": 0.7869,
+      "step": 9780
+    },
+    {
+      "epoch": 0.5383345258407177,
+      "grad_norm": 0.7341023087501526,
+      "learning_rate": 8.340001183053535e-06,
+      "loss": 0.7447,
+      "step": 9781
+    },
+    {
+      "epoch": 0.5383895646430734,
+      "grad_norm": 0.6829406023025513,
+      "learning_rate": 8.339678601525263e-06,
+      "loss": 0.7438,
+      "step": 9782
+    },
+    {
+      "epoch": 0.538444603445429,
+      "grad_norm": 0.7671583294868469,
+      "learning_rate": 8.33935599489686e-06,
+      "loss": 0.8678,
+      "step": 9783
+    },
+    {
+      "epoch": 0.5384996422477847,
+      "grad_norm": 0.701797366142273,
+      "learning_rate": 8.339033363170753e-06,
+      "loss": 0.8431,
+      "step": 9784
+    },
+    {
+      "epoch": 0.5385546810501404,
+      "grad_norm": 0.748235285282135,
+      "learning_rate": 8.338710706349363e-06,
+      "loss": 0.7905,
+      "step": 9785
+    },
+    {
+      "epoch": 0.5386097198524961,
+      "grad_norm": 0.8202430605888367,
+      "learning_rate": 8.338388024435119e-06,
+      "loss": 0.7734,
+      "step": 9786
+    },
+    {
+      "epoch": 0.5386647586548516,
+      "grad_norm": 0.8218014240264893,
+      "learning_rate": 8.338065317430442e-06,
+      "loss": 0.846,
+      "step": 9787
+    },
+    {
+      "epoch": 0.5387197974572073,
+      "grad_norm": 0.6773214936256409,
+      "learning_rate": 8.337742585337762e-06,
+      "loss": 0.7692,
+      "step": 9788
+    },
+    {
+      "epoch": 0.538774836259563,
+      "grad_norm": 0.7011464834213257,
+      "learning_rate": 8.337419828159501e-06,
+      "loss": 0.7534,
+      "step": 9789
+    },
+    {
+      "epoch": 0.5388298750619187,
+      "grad_norm": 0.8299004435539246,
+      "learning_rate": 8.337097045898087e-06,
+      "loss": 0.7997,
+      "step": 9790
+    },
+    {
+      "epoch": 0.5388849138642743,
+      "grad_norm": 0.8600753545761108,
+      "learning_rate": 8.336774238555942e-06,
+      "loss": 0.8307,
+      "step": 9791
+    },
+    {
+      "epoch": 0.53893995266663,
+      "grad_norm": 0.676490843296051,
+      "learning_rate": 8.336451406135498e-06,
+      "loss": 0.7748,
+      "step": 9792
+    },
+    {
+      "epoch": 0.5389949914689857,
+      "grad_norm": 0.7094627618789673,
+      "learning_rate": 8.336128548639177e-06,
+      "loss": 0.7524,
+      "step": 9793
+    },
+    {
+      "epoch": 0.5390500302713412,
+      "grad_norm": 0.6804066896438599,
+      "learning_rate": 8.335805666069407e-06,
+      "loss": 0.8299,
+      "step": 9794
+    },
+    {
+      "epoch": 0.5391050690736969,
+      "grad_norm": 0.6992025971412659,
+      "learning_rate": 8.335482758428614e-06,
+      "loss": 0.7548,
+      "step": 9795
+    },
+    {
+      "epoch": 0.5391601078760526,
+      "grad_norm": 0.6649640798568726,
+      "learning_rate": 8.335159825719227e-06,
+      "loss": 0.6595,
+      "step": 9796
+    },
+    {
+      "epoch": 0.5392151466784083,
+      "grad_norm": 0.7292002439498901,
+      "learning_rate": 8.33483686794367e-06,
+      "loss": 0.7944,
+      "step": 9797
+    },
+    {
+      "epoch": 0.5392701854807639,
+      "grad_norm": 0.9124587178230286,
+      "learning_rate": 8.334513885104375e-06,
+      "loss": 0.8586,
+      "step": 9798
+    },
+    {
+      "epoch": 0.5393252242831196,
+      "grad_norm": 0.7091020941734314,
+      "learning_rate": 8.334190877203761e-06,
+      "loss": 0.7019,
+      "step": 9799
+    },
+    {
+      "epoch": 0.5393802630854753,
+      "grad_norm": 0.7470952272415161,
+      "learning_rate": 8.333867844244265e-06,
+      "loss": 0.7866,
+      "step": 9800
+    },
+    {
+      "epoch": 0.539435301887831,
+      "grad_norm": 0.7368966341018677,
+      "learning_rate": 8.333544786228309e-06,
+      "loss": 0.8135,
+      "step": 9801
+    },
+    {
+      "epoch": 0.5394903406901865,
+      "grad_norm": 0.668305516242981,
+      "learning_rate": 8.333221703158322e-06,
+      "loss": 0.7549,
+      "step": 9802
+    },
+    {
+      "epoch": 0.5395453794925422,
+      "grad_norm": 0.6788874268531799,
+      "learning_rate": 8.332898595036735e-06,
+      "loss": 0.8077,
+      "step": 9803
+    },
+    {
+      "epoch": 0.5396004182948979,
+      "grad_norm": 0.654863715171814,
+      "learning_rate": 8.332575461865972e-06,
+      "loss": 0.7695,
+      "step": 9804
+    },
+    {
+      "epoch": 0.5396554570972536,
+      "grad_norm": 0.7460314631462097,
+      "learning_rate": 8.332252303648464e-06,
+      "loss": 0.7711,
+      "step": 9805
+    },
+    {
+      "epoch": 0.5397104958996092,
+      "grad_norm": 0.7923582792282104,
+      "learning_rate": 8.331929120386643e-06,
+      "loss": 0.7348,
+      "step": 9806
+    },
+    {
+      "epoch": 0.5397655347019649,
+      "grad_norm": 0.6570843458175659,
+      "learning_rate": 8.331605912082932e-06,
+      "loss": 0.7029,
+      "step": 9807
+    },
+    {
+      "epoch": 0.5398205735043206,
+      "grad_norm": 0.7728865742683411,
+      "learning_rate": 8.331282678739762e-06,
+      "loss": 0.8249,
+      "step": 9808
+    },
+    {
+      "epoch": 0.5398756123066762,
+      "grad_norm": 0.7121468186378479,
+      "learning_rate": 8.330959420359565e-06,
+      "loss": 0.8698,
+      "step": 9809
+    },
+    {
+      "epoch": 0.5399306511090318,
+      "grad_norm": 0.7779444456100464,
+      "learning_rate": 8.330636136944768e-06,
+      "loss": 0.7448,
+      "step": 9810
+    },
+    {
+      "epoch": 0.5399856899113875,
+      "grad_norm": 0.7770833373069763,
+      "learning_rate": 8.330312828497801e-06,
+      "loss": 0.8489,
+      "step": 9811
+    },
+    {
+      "epoch": 0.5400407287137432,
+      "grad_norm": 0.6705769896507263,
+      "learning_rate": 8.329989495021096e-06,
+      "loss": 0.7349,
+      "step": 9812
+    },
+    {
+      "epoch": 0.5400957675160989,
+      "grad_norm": 0.6775381565093994,
+      "learning_rate": 8.329666136517079e-06,
+      "loss": 0.8093,
+      "step": 9813
+    },
+    {
+      "epoch": 0.5401508063184545,
+      "grad_norm": 0.6621832251548767,
+      "learning_rate": 8.329342752988183e-06,
+      "loss": 0.7877,
+      "step": 9814
+    },
+    {
+      "epoch": 0.5402058451208102,
+      "grad_norm": 0.704339861869812,
+      "learning_rate": 8.329019344436839e-06,
+      "loss": 0.7708,
+      "step": 9815
+    },
+    {
+      "epoch": 0.5402608839231658,
+      "grad_norm": 0.789944052696228,
+      "learning_rate": 8.328695910865476e-06,
+      "loss": 0.7563,
+      "step": 9816
+    },
+    {
+      "epoch": 0.5403159227255215,
+      "grad_norm": 0.6997420191764832,
+      "learning_rate": 8.328372452276525e-06,
+      "loss": 0.7023,
+      "step": 9817
+    },
+    {
+      "epoch": 0.5403709615278771,
+      "grad_norm": 0.6453180313110352,
+      "learning_rate": 8.328048968672418e-06,
+      "loss": 0.7193,
+      "step": 9818
+    },
+    {
+      "epoch": 0.5404260003302328,
+      "grad_norm": 0.7059640884399414,
+      "learning_rate": 8.327725460055586e-06,
+      "loss": 0.7875,
+      "step": 9819
+    },
+    {
+      "epoch": 0.5404810391325885,
+      "grad_norm": 0.7725005745887756,
+      "learning_rate": 8.327401926428461e-06,
+      "loss": 0.7503,
+      "step": 9820
+    },
+    {
+      "epoch": 0.5405360779349442,
+      "grad_norm": 0.7710940837860107,
+      "learning_rate": 8.327078367793473e-06,
+      "loss": 0.8314,
+      "step": 9821
+    },
+    {
+      "epoch": 0.5405911167372998,
+      "grad_norm": 0.9090666770935059,
+      "learning_rate": 8.326754784153055e-06,
+      "loss": 0.8021,
+      "step": 9822
+    },
+    {
+      "epoch": 0.5406461555396554,
+      "grad_norm": 0.7135322690010071,
+      "learning_rate": 8.326431175509638e-06,
+      "loss": 0.8084,
+      "step": 9823
+    },
+    {
+      "epoch": 0.5407011943420111,
+      "grad_norm": 0.9126102328300476,
+      "learning_rate": 8.326107541865656e-06,
+      "loss": 0.75,
+      "step": 9824
+    },
+    {
+      "epoch": 0.5407562331443668,
+      "grad_norm": 0.7263361215591431,
+      "learning_rate": 8.325783883223539e-06,
+      "loss": 0.6808,
+      "step": 9825
+    },
+    {
+      "epoch": 0.5408112719467224,
+      "grad_norm": 0.7234700918197632,
+      "learning_rate": 8.32546019958572e-06,
+      "loss": 0.7582,
+      "step": 9826
+    },
+    {
+      "epoch": 0.5408663107490781,
+      "grad_norm": 0.7043294310569763,
+      "learning_rate": 8.325136490954633e-06,
+      "loss": 0.8421,
+      "step": 9827
+    },
+    {
+      "epoch": 0.5409213495514338,
+      "grad_norm": 0.7947664856910706,
+      "learning_rate": 8.32481275733271e-06,
+      "loss": 0.8672,
+      "step": 9828
+    },
+    {
+      "epoch": 0.5409763883537895,
+      "grad_norm": 0.704590916633606,
+      "learning_rate": 8.324488998722384e-06,
+      "loss": 0.7356,
+      "step": 9829
+    },
+    {
+      "epoch": 0.541031427156145,
+      "grad_norm": 0.7630662322044373,
+      "learning_rate": 8.32416521512609e-06,
+      "loss": 0.7082,
+      "step": 9830
+    },
+    {
+      "epoch": 0.5410864659585007,
+      "grad_norm": 0.728721022605896,
+      "learning_rate": 8.323841406546259e-06,
+      "loss": 0.7987,
+      "step": 9831
+    },
+    {
+      "epoch": 0.5411415047608564,
+      "grad_norm": 0.7164294719696045,
+      "learning_rate": 8.323517572985326e-06,
+      "loss": 0.721,
+      "step": 9832
+    },
+    {
+      "epoch": 0.5411965435632121,
+      "grad_norm": 0.7555723190307617,
+      "learning_rate": 8.323193714445722e-06,
+      "loss": 0.814,
+      "step": 9833
+    },
+    {
+      "epoch": 0.5412515823655677,
+      "grad_norm": 0.827485978603363,
+      "learning_rate": 8.322869830929887e-06,
+      "loss": 0.8817,
+      "step": 9834
+    },
+    {
+      "epoch": 0.5413066211679234,
+      "grad_norm": 0.718950092792511,
+      "learning_rate": 8.322545922440252e-06,
+      "loss": 0.8648,
+      "step": 9835
+    },
+    {
+      "epoch": 0.5413616599702791,
+      "grad_norm": 0.7361611723899841,
+      "learning_rate": 8.32222198897925e-06,
+      "loss": 0.7392,
+      "step": 9836
+    },
+    {
+      "epoch": 0.5414166987726347,
+      "grad_norm": 0.6712168455123901,
+      "learning_rate": 8.321898030549316e-06,
+      "loss": 0.7505,
+      "step": 9837
+    },
+    {
+      "epoch": 0.5414717375749903,
+      "grad_norm": 0.7475710511207581,
+      "learning_rate": 8.321574047152887e-06,
+      "loss": 0.7969,
+      "step": 9838
+    },
+    {
+      "epoch": 0.541526776377346,
+      "grad_norm": 0.9751361608505249,
+      "learning_rate": 8.321250038792397e-06,
+      "loss": 0.8534,
+      "step": 9839
+    },
+    {
+      "epoch": 0.5415818151797017,
+      "grad_norm": 0.6858723163604736,
+      "learning_rate": 8.32092600547028e-06,
+      "loss": 0.8277,
+      "step": 9840
+    },
+    {
+      "epoch": 0.5416368539820573,
+      "grad_norm": 0.8899725675582886,
+      "learning_rate": 8.320601947188971e-06,
+      "loss": 0.8599,
+      "step": 9841
+    },
+    {
+      "epoch": 0.541691892784413,
+      "grad_norm": 0.7140665650367737,
+      "learning_rate": 8.320277863950907e-06,
+      "loss": 0.7429,
+      "step": 9842
+    },
+    {
+      "epoch": 0.5417469315867687,
+      "grad_norm": 0.7467615604400635,
+      "learning_rate": 8.319953755758525e-06,
+      "loss": 0.7826,
+      "step": 9843
+    },
+    {
+      "epoch": 0.5418019703891244,
+      "grad_norm": 0.6578202843666077,
+      "learning_rate": 8.319629622614258e-06,
+      "loss": 0.6833,
+      "step": 9844
+    },
+    {
+      "epoch": 0.5418570091914799,
+      "grad_norm": 0.9430698156356812,
+      "learning_rate": 8.319305464520543e-06,
+      "loss": 0.8243,
+      "step": 9845
+    },
+    {
+      "epoch": 0.5419120479938356,
+      "grad_norm": 0.8632097840309143,
+      "learning_rate": 8.318981281479817e-06,
+      "loss": 0.7975,
+      "step": 9846
+    },
+    {
+      "epoch": 0.5419670867961913,
+      "grad_norm": 0.7241839170455933,
+      "learning_rate": 8.318657073494517e-06,
+      "loss": 0.7226,
+      "step": 9847
+    },
+    {
+      "epoch": 0.542022125598547,
+      "grad_norm": 0.6927164196968079,
+      "learning_rate": 8.318332840567078e-06,
+      "loss": 0.7125,
+      "step": 9848
+    },
+    {
+      "epoch": 0.5420771644009026,
+      "grad_norm": 0.6414939761161804,
+      "learning_rate": 8.318008582699937e-06,
+      "loss": 0.7366,
+      "step": 9849
+    },
+    {
+      "epoch": 0.5421322032032583,
+      "grad_norm": 0.7584436535835266,
+      "learning_rate": 8.317684299895533e-06,
+      "loss": 0.8601,
+      "step": 9850
+    },
+    {
+      "epoch": 0.542187242005614,
+      "grad_norm": 0.6045856475830078,
+      "learning_rate": 8.317359992156302e-06,
+      "loss": 0.6697,
+      "step": 9851
+    },
+    {
+      "epoch": 0.5422422808079697,
+      "grad_norm": 0.715048611164093,
+      "learning_rate": 8.31703565948468e-06,
+      "loss": 0.7535,
+      "step": 9852
+    },
+    {
+      "epoch": 0.5422973196103252,
+      "grad_norm": 0.6925113201141357,
+      "learning_rate": 8.316711301883106e-06,
+      "loss": 0.8122,
+      "step": 9853
+    },
+    {
+      "epoch": 0.5423523584126809,
+      "grad_norm": 0.6787780523300171,
+      "learning_rate": 8.316386919354018e-06,
+      "loss": 0.7428,
+      "step": 9854
+    },
+    {
+      "epoch": 0.5424073972150366,
+      "grad_norm": 0.6831366419792175,
+      "learning_rate": 8.316062511899855e-06,
+      "loss": 0.767,
+      "step": 9855
+    },
+    {
+      "epoch": 0.5424624360173923,
+      "grad_norm": 0.6865691542625427,
+      "learning_rate": 8.315738079523053e-06,
+      "loss": 0.6549,
+      "step": 9856
+    },
+    {
+      "epoch": 0.5425174748197479,
+      "grad_norm": 0.7149406671524048,
+      "learning_rate": 8.31541362222605e-06,
+      "loss": 0.8127,
+      "step": 9857
+    },
+    {
+      "epoch": 0.5425725136221036,
+      "grad_norm": 0.6826779842376709,
+      "learning_rate": 8.315089140011286e-06,
+      "loss": 0.706,
+      "step": 9858
+    },
+    {
+      "epoch": 0.5426275524244593,
+      "grad_norm": 0.688204288482666,
+      "learning_rate": 8.3147646328812e-06,
+      "loss": 0.8675,
+      "step": 9859
+    },
+    {
+      "epoch": 0.542682591226815,
+      "grad_norm": 0.6659492254257202,
+      "learning_rate": 8.31444010083823e-06,
+      "loss": 0.7851,
+      "step": 9860
+    },
+    {
+      "epoch": 0.5427376300291705,
+      "grad_norm": 0.8049291372299194,
+      "learning_rate": 8.314115543884816e-06,
+      "loss": 0.7442,
+      "step": 9861
+    },
+    {
+      "epoch": 0.5427926688315262,
+      "grad_norm": 0.7505989670753479,
+      "learning_rate": 8.313790962023397e-06,
+      "loss": 0.8391,
+      "step": 9862
+    },
+    {
+      "epoch": 0.5428477076338819,
+      "grad_norm": 0.6810199618339539,
+      "learning_rate": 8.31346635525641e-06,
+      "loss": 0.8131,
+      "step": 9863
+    },
+    {
+      "epoch": 0.5429027464362376,
+      "grad_norm": 0.6724215745925903,
+      "learning_rate": 8.313141723586298e-06,
+      "loss": 0.75,
+      "step": 9864
+    },
+    {
+      "epoch": 0.5429577852385932,
+      "grad_norm": 0.7804376482963562,
+      "learning_rate": 8.3128170670155e-06,
+      "loss": 0.704,
+      "step": 9865
+    },
+    {
+      "epoch": 0.5430128240409489,
+      "grad_norm": 0.9494230151176453,
+      "learning_rate": 8.312492385546455e-06,
+      "loss": 0.8578,
+      "step": 9866
+    },
+    {
+      "epoch": 0.5430678628433045,
+      "grad_norm": 0.6780333518981934,
+      "learning_rate": 8.312167679181606e-06,
+      "loss": 0.701,
+      "step": 9867
+    },
+    {
+      "epoch": 0.5431229016456602,
+      "grad_norm": 0.7407701015472412,
+      "learning_rate": 8.31184294792339e-06,
+      "loss": 0.8505,
+      "step": 9868
+    },
+    {
+      "epoch": 0.5431779404480158,
+      "grad_norm": 0.680903434753418,
+      "learning_rate": 8.311518191774249e-06,
+      "loss": 0.7645,
+      "step": 9869
+    },
+    {
+      "epoch": 0.5432329792503715,
+      "grad_norm": 0.6695752143859863,
+      "learning_rate": 8.311193410736622e-06,
+      "loss": 0.816,
+      "step": 9870
+    },
+    {
+      "epoch": 0.5432880180527272,
+      "grad_norm": 0.6725142598152161,
+      "learning_rate": 8.310868604812954e-06,
+      "loss": 0.7044,
+      "step": 9871
+    },
+    {
+      "epoch": 0.5433430568550829,
+      "grad_norm": 0.922627866268158,
+      "learning_rate": 8.310543774005684e-06,
+      "loss": 0.7589,
+      "step": 9872
+    },
+    {
+      "epoch": 0.5433980956574385,
+      "grad_norm": 1.0136839151382446,
+      "learning_rate": 8.310218918317251e-06,
+      "loss": 0.7573,
+      "step": 9873
+    },
+    {
+      "epoch": 0.5434531344597942,
+      "grad_norm": 0.9053532481193542,
+      "learning_rate": 8.309894037750099e-06,
+      "loss": 0.8269,
+      "step": 9874
+    },
+    {
+      "epoch": 0.5435081732621498,
+      "grad_norm": 0.6800149083137512,
+      "learning_rate": 8.309569132306671e-06,
+      "loss": 0.716,
+      "step": 9875
+    },
+    {
+      "epoch": 0.5435632120645055,
+      "grad_norm": 0.7157679796218872,
+      "learning_rate": 8.309244201989408e-06,
+      "loss": 0.7433,
+      "step": 9876
+    },
+    {
+      "epoch": 0.5436182508668611,
+      "grad_norm": 0.9316089749336243,
+      "learning_rate": 8.308919246800748e-06,
+      "loss": 0.7499,
+      "step": 9877
+    },
+    {
+      "epoch": 0.5436732896692168,
+      "grad_norm": 0.6682490110397339,
+      "learning_rate": 8.308594266743139e-06,
+      "loss": 0.7286,
+      "step": 9878
+    },
+    {
+      "epoch": 0.5437283284715725,
+      "grad_norm": 0.7241143584251404,
+      "learning_rate": 8.308269261819022e-06,
+      "loss": 0.7934,
+      "step": 9879
+    },
+    {
+      "epoch": 0.5437833672739281,
+      "grad_norm": 0.7402396202087402,
+      "learning_rate": 8.307944232030838e-06,
+      "loss": 0.7361,
+      "step": 9880
+    },
+    {
+      "epoch": 0.5438384060762838,
+      "grad_norm": 0.6839993596076965,
+      "learning_rate": 8.307619177381029e-06,
+      "loss": 0.749,
+      "step": 9881
+    },
+    {
+      "epoch": 0.5438934448786394,
+      "grad_norm": 0.6536363363265991,
+      "learning_rate": 8.307294097872041e-06,
+      "loss": 0.706,
+      "step": 9882
+    },
+    {
+      "epoch": 0.5439484836809951,
+      "grad_norm": 0.602644681930542,
+      "learning_rate": 8.306968993506317e-06,
+      "loss": 0.6857,
+      "step": 9883
+    },
+    {
+      "epoch": 0.5440035224833507,
+      "grad_norm": 0.6567881107330322,
+      "learning_rate": 8.306643864286297e-06,
+      "loss": 0.6989,
+      "step": 9884
+    },
+    {
+      "epoch": 0.5440585612857064,
+      "grad_norm": 1.0013506412506104,
+      "learning_rate": 8.306318710214427e-06,
+      "loss": 0.7251,
+      "step": 9885
+    },
+    {
+      "epoch": 0.5441136000880621,
+      "grad_norm": 0.7016813158988953,
+      "learning_rate": 8.305993531293153e-06,
+      "loss": 0.7535,
+      "step": 9886
+    },
+    {
+      "epoch": 0.5441686388904178,
+      "grad_norm": 0.7345741391181946,
+      "learning_rate": 8.305668327524915e-06,
+      "loss": 0.887,
+      "step": 9887
+    },
+    {
+      "epoch": 0.5442236776927734,
+      "grad_norm": 1.0925754308700562,
+      "learning_rate": 8.305343098912158e-06,
+      "loss": 0.7779,
+      "step": 9888
+    },
+    {
+      "epoch": 0.544278716495129,
+      "grad_norm": 0.79815274477005,
+      "learning_rate": 8.305017845457328e-06,
+      "loss": 0.7736,
+      "step": 9889
+    },
+    {
+      "epoch": 0.5443337552974847,
+      "grad_norm": 0.6324154138565063,
+      "learning_rate": 8.304692567162868e-06,
+      "loss": 0.6823,
+      "step": 9890
+    },
+    {
+      "epoch": 0.5443887940998404,
+      "grad_norm": 0.6990262866020203,
+      "learning_rate": 8.304367264031223e-06,
+      "loss": 0.7804,
+      "step": 9891
+    },
+    {
+      "epoch": 0.544443832902196,
+      "grad_norm": 1.4203195571899414,
+      "learning_rate": 8.304041936064839e-06,
+      "loss": 0.8702,
+      "step": 9892
+    },
+    {
+      "epoch": 0.5444988717045517,
+      "grad_norm": 0.6986544132232666,
+      "learning_rate": 8.303716583266161e-06,
+      "loss": 0.7666,
+      "step": 9893
+    },
+    {
+      "epoch": 0.5445539105069074,
+      "grad_norm": 0.7037138938903809,
+      "learning_rate": 8.303391205637632e-06,
+      "loss": 0.7995,
+      "step": 9894
+    },
+    {
+      "epoch": 0.5446089493092631,
+      "grad_norm": 0.7101728320121765,
+      "learning_rate": 8.3030658031817e-06,
+      "loss": 0.8185,
+      "step": 9895
+    },
+    {
+      "epoch": 0.5446639881116186,
+      "grad_norm": 0.6571425795555115,
+      "learning_rate": 8.302740375900808e-06,
+      "loss": 0.6152,
+      "step": 9896
+    },
+    {
+      "epoch": 0.5447190269139743,
+      "grad_norm": 0.7560263276100159,
+      "learning_rate": 8.302414923797406e-06,
+      "loss": 0.9037,
+      "step": 9897
+    },
+    {
+      "epoch": 0.54477406571633,
+      "grad_norm": 0.8692007064819336,
+      "learning_rate": 8.302089446873935e-06,
+      "loss": 0.7689,
+      "step": 9898
+    },
+    {
+      "epoch": 0.5448291045186857,
+      "grad_norm": 0.7533506751060486,
+      "learning_rate": 8.301763945132845e-06,
+      "loss": 0.7671,
+      "step": 9899
+    },
+    {
+      "epoch": 0.5448841433210413,
+      "grad_norm": 0.6992233991622925,
+      "learning_rate": 8.301438418576581e-06,
+      "loss": 0.723,
+      "step": 9900
+    },
+    {
+      "epoch": 0.544939182123397,
+      "grad_norm": 0.7966120839118958,
+      "learning_rate": 8.301112867207589e-06,
+      "loss": 0.7968,
+      "step": 9901
+    },
+    {
+      "epoch": 0.5449942209257527,
+      "grad_norm": 0.800558865070343,
+      "learning_rate": 8.300787291028316e-06,
+      "loss": 0.8583,
+      "step": 9902
+    },
+    {
+      "epoch": 0.5450492597281084,
+      "grad_norm": 0.7019909024238586,
+      "learning_rate": 8.30046169004121e-06,
+      "loss": 0.7045,
+      "step": 9903
+    },
+    {
+      "epoch": 0.5451042985304639,
+      "grad_norm": 0.7778449654579163,
+      "learning_rate": 8.300136064248717e-06,
+      "loss": 0.7964,
+      "step": 9904
+    },
+    {
+      "epoch": 0.5451593373328196,
+      "grad_norm": 0.6894309520721436,
+      "learning_rate": 8.299810413653284e-06,
+      "loss": 0.7382,
+      "step": 9905
+    },
+    {
+      "epoch": 0.5452143761351753,
+      "grad_norm": 0.6942182183265686,
+      "learning_rate": 8.299484738257361e-06,
+      "loss": 0.73,
+      "step": 9906
+    },
+    {
+      "epoch": 0.545269414937531,
+      "grad_norm": 0.6607787609100342,
+      "learning_rate": 8.299159038063394e-06,
+      "loss": 0.6987,
+      "step": 9907
+    },
+    {
+      "epoch": 0.5453244537398866,
+      "grad_norm": 0.7447709441184998,
+      "learning_rate": 8.29883331307383e-06,
+      "loss": 0.7787,
+      "step": 9908
+    },
+    {
+      "epoch": 0.5453794925422423,
+      "grad_norm": 0.6315301656723022,
+      "learning_rate": 8.298507563291116e-06,
+      "loss": 0.7047,
+      "step": 9909
+    },
+    {
+      "epoch": 0.545434531344598,
+      "grad_norm": 0.8095656633377075,
+      "learning_rate": 8.298181788717705e-06,
+      "loss": 0.691,
+      "step": 9910
+    },
+    {
+      "epoch": 0.5454895701469537,
+      "grad_norm": 0.6419453024864197,
+      "learning_rate": 8.29785598935604e-06,
+      "loss": 0.7333,
+      "step": 9911
+    },
+    {
+      "epoch": 0.5455446089493092,
+      "grad_norm": 0.7209222316741943,
+      "learning_rate": 8.297530165208574e-06,
+      "loss": 0.8174,
+      "step": 9912
+    },
+    {
+      "epoch": 0.5455996477516649,
+      "grad_norm": 0.6778598427772522,
+      "learning_rate": 8.297204316277754e-06,
+      "loss": 0.7696,
+      "step": 9913
+    },
+    {
+      "epoch": 0.5456546865540206,
+      "grad_norm": 0.6573307514190674,
+      "learning_rate": 8.296878442566028e-06,
+      "loss": 0.7843,
+      "step": 9914
+    },
+    {
+      "epoch": 0.5457097253563763,
+      "grad_norm": 0.6987473964691162,
+      "learning_rate": 8.296552544075847e-06,
+      "loss": 0.809,
+      "step": 9915
+    },
+    {
+      "epoch": 0.5457647641587319,
+      "grad_norm": 0.7149204015731812,
+      "learning_rate": 8.29622662080966e-06,
+      "loss": 0.848,
+      "step": 9916
+    },
+    {
+      "epoch": 0.5458198029610876,
+      "grad_norm": 0.6252632141113281,
+      "learning_rate": 8.295900672769913e-06,
+      "loss": 0.7029,
+      "step": 9917
+    },
+    {
+      "epoch": 0.5458748417634433,
+      "grad_norm": 0.713376522064209,
+      "learning_rate": 8.295574699959062e-06,
+      "loss": 0.726,
+      "step": 9918
+    },
+    {
+      "epoch": 0.5459298805657989,
+      "grad_norm": 0.6864717602729797,
+      "learning_rate": 8.295248702379552e-06,
+      "loss": 0.7428,
+      "step": 9919
+    },
+    {
+      "epoch": 0.5459849193681545,
+      "grad_norm": 0.8085678219795227,
+      "learning_rate": 8.294922680033837e-06,
+      "loss": 0.8697,
+      "step": 9920
+    },
+    {
+      "epoch": 0.5460399581705102,
+      "grad_norm": 0.7366700768470764,
+      "learning_rate": 8.294596632924363e-06,
+      "loss": 0.7714,
+      "step": 9921
+    },
+    {
+      "epoch": 0.5460949969728659,
+      "grad_norm": 0.670632541179657,
+      "learning_rate": 8.294270561053583e-06,
+      "loss": 0.7032,
+      "step": 9922
+    },
+    {
+      "epoch": 0.5461500357752215,
+      "grad_norm": 0.7867220640182495,
+      "learning_rate": 8.293944464423946e-06,
+      "loss": 0.8903,
+      "step": 9923
+    },
+    {
+      "epoch": 0.5462050745775772,
+      "grad_norm": 0.8441565632820129,
+      "learning_rate": 8.293618343037907e-06,
+      "loss": 0.8694,
+      "step": 9924
+    },
+    {
+      "epoch": 0.5462601133799329,
+      "grad_norm": 0.7048027515411377,
+      "learning_rate": 8.293292196897913e-06,
+      "loss": 0.8226,
+      "step": 9925
+    },
+    {
+      "epoch": 0.5463151521822885,
+      "grad_norm": 0.6344078779220581,
+      "learning_rate": 8.292966026006416e-06,
+      "loss": 0.7615,
+      "step": 9926
+    },
+    {
+      "epoch": 0.5463701909846441,
+      "grad_norm": 0.6744484901428223,
+      "learning_rate": 8.292639830365867e-06,
+      "loss": 0.6944,
+      "step": 9927
+    },
+    {
+      "epoch": 0.5464252297869998,
+      "grad_norm": 0.8113303780555725,
+      "learning_rate": 8.292313609978721e-06,
+      "loss": 0.7558,
+      "step": 9928
+    },
+    {
+      "epoch": 0.5464802685893555,
+      "grad_norm": 0.640190839767456,
+      "learning_rate": 8.291987364847425e-06,
+      "loss": 0.7167,
+      "step": 9929
+    },
+    {
+      "epoch": 0.5465353073917112,
+      "grad_norm": 0.7714816331863403,
+      "learning_rate": 8.291661094974434e-06,
+      "loss": 0.8662,
+      "step": 9930
+    },
+    {
+      "epoch": 0.5465903461940668,
+      "grad_norm": 0.6785402894020081,
+      "learning_rate": 8.291334800362199e-06,
+      "loss": 0.6835,
+      "step": 9931
+    },
+    {
+      "epoch": 0.5466453849964225,
+      "grad_norm": 0.704868495464325,
+      "learning_rate": 8.291008481013173e-06,
+      "loss": 0.7343,
+      "step": 9932
+    },
+    {
+      "epoch": 0.5467004237987781,
+      "grad_norm": 0.7587466239929199,
+      "learning_rate": 8.290682136929809e-06,
+      "loss": 0.7856,
+      "step": 9933
+    },
+    {
+      "epoch": 0.5467554626011338,
+      "grad_norm": 0.7460505962371826,
+      "learning_rate": 8.290355768114557e-06,
+      "loss": 0.7463,
+      "step": 9934
+    },
+    {
+      "epoch": 0.5468105014034894,
+      "grad_norm": 0.7185021042823792,
+      "learning_rate": 8.290029374569873e-06,
+      "loss": 0.8106,
+      "step": 9935
+    },
+    {
+      "epoch": 0.5468655402058451,
+      "grad_norm": 0.7023874521255493,
+      "learning_rate": 8.289702956298209e-06,
+      "loss": 0.6863,
+      "step": 9936
+    },
+    {
+      "epoch": 0.5469205790082008,
+      "grad_norm": 0.8688495755195618,
+      "learning_rate": 8.289376513302017e-06,
+      "loss": 0.8898,
+      "step": 9937
+    },
+    {
+      "epoch": 0.5469756178105565,
+      "grad_norm": 0.6405122876167297,
+      "learning_rate": 8.289050045583752e-06,
+      "loss": 0.6804,
+      "step": 9938
+    },
+    {
+      "epoch": 0.5470306566129121,
+      "grad_norm": 0.8364881277084351,
+      "learning_rate": 8.288723553145868e-06,
+      "loss": 0.8356,
+      "step": 9939
+    },
+    {
+      "epoch": 0.5470856954152677,
+      "grad_norm": 0.6621617078781128,
+      "learning_rate": 8.288397035990818e-06,
+      "loss": 0.7508,
+      "step": 9940
+    },
+    {
+      "epoch": 0.5471407342176234,
+      "grad_norm": 0.6822347640991211,
+      "learning_rate": 8.288070494121056e-06,
+      "loss": 0.7722,
+      "step": 9941
+    },
+    {
+      "epoch": 0.5471957730199791,
+      "grad_norm": 0.6727223992347717,
+      "learning_rate": 8.287743927539036e-06,
+      "loss": 0.743,
+      "step": 9942
+    },
+    {
+      "epoch": 0.5472508118223347,
+      "grad_norm": 0.7852441668510437,
+      "learning_rate": 8.287417336247214e-06,
+      "loss": 0.8321,
+      "step": 9943
+    },
+    {
+      "epoch": 0.5473058506246904,
+      "grad_norm": 0.6982126235961914,
+      "learning_rate": 8.287090720248041e-06,
+      "loss": 0.6669,
+      "step": 9944
+    },
+    {
+      "epoch": 0.5473608894270461,
+      "grad_norm": 0.7820166945457458,
+      "learning_rate": 8.286764079543976e-06,
+      "loss": 0.7592,
+      "step": 9945
+    },
+    {
+      "epoch": 0.5474159282294018,
+      "grad_norm": 0.6868422627449036,
+      "learning_rate": 8.28643741413747e-06,
+      "loss": 0.8308,
+      "step": 9946
+    },
+    {
+      "epoch": 0.5474709670317573,
+      "grad_norm": 0.8227942585945129,
+      "learning_rate": 8.286110724030982e-06,
+      "loss": 0.7982,
+      "step": 9947
+    },
+    {
+      "epoch": 0.547526005834113,
+      "grad_norm": 0.6838171482086182,
+      "learning_rate": 8.285784009226964e-06,
+      "loss": 0.7907,
+      "step": 9948
+    },
+    {
+      "epoch": 0.5475810446364687,
+      "grad_norm": 0.7200812697410583,
+      "learning_rate": 8.285457269727875e-06,
+      "loss": 0.88,
+      "step": 9949
+    },
+    {
+      "epoch": 0.5476360834388244,
+      "grad_norm": 0.7469412684440613,
+      "learning_rate": 8.285130505536168e-06,
+      "loss": 0.8167,
+      "step": 9950
+    },
+    {
+      "epoch": 0.54769112224118,
+      "grad_norm": 0.6660227179527283,
+      "learning_rate": 8.284803716654298e-06,
+      "loss": 0.7685,
+      "step": 9951
+    },
+    {
+      "epoch": 0.5477461610435357,
+      "grad_norm": 0.7116572260856628,
+      "learning_rate": 8.284476903084723e-06,
+      "loss": 0.7415,
+      "step": 9952
+    },
+    {
+      "epoch": 0.5478011998458914,
+      "grad_norm": 0.6540791988372803,
+      "learning_rate": 8.284150064829899e-06,
+      "loss": 0.6571,
+      "step": 9953
+    },
+    {
+      "epoch": 0.5478562386482471,
+      "grad_norm": 0.7527759075164795,
+      "learning_rate": 8.283823201892283e-06,
+      "loss": 0.8678,
+      "step": 9954
+    },
+    {
+      "epoch": 0.5479112774506026,
+      "grad_norm": 0.7795953750610352,
+      "learning_rate": 8.283496314274331e-06,
+      "loss": 0.8086,
+      "step": 9955
+    },
+    {
+      "epoch": 0.5479663162529583,
+      "grad_norm": 0.862503170967102,
+      "learning_rate": 8.283169401978498e-06,
+      "loss": 0.7442,
+      "step": 9956
+    },
+    {
+      "epoch": 0.548021355055314,
+      "grad_norm": 0.6552054286003113,
+      "learning_rate": 8.282842465007244e-06,
+      "loss": 0.6664,
+      "step": 9957
+    },
+    {
+      "epoch": 0.5480763938576697,
+      "grad_norm": 0.7242427468299866,
+      "learning_rate": 8.282515503363024e-06,
+      "loss": 0.8199,
+      "step": 9958
+    },
+    {
+      "epoch": 0.5481314326600253,
+      "grad_norm": 0.7529763579368591,
+      "learning_rate": 8.282188517048295e-06,
+      "loss": 0.761,
+      "step": 9959
+    },
+    {
+      "epoch": 0.548186471462381,
+      "grad_norm": 0.7909425497055054,
+      "learning_rate": 8.281861506065519e-06,
+      "loss": 0.7389,
+      "step": 9960
+    },
+    {
+      "epoch": 0.5482415102647367,
+      "grad_norm": 0.6594850420951843,
+      "learning_rate": 8.281534470417147e-06,
+      "loss": 0.7473,
+      "step": 9961
+    },
+    {
+      "epoch": 0.5482965490670924,
+      "grad_norm": 0.6900844573974609,
+      "learning_rate": 8.281207410105642e-06,
+      "loss": 0.7551,
+      "step": 9962
+    },
+    {
+      "epoch": 0.5483515878694479,
+      "grad_norm": 0.6922640204429626,
+      "learning_rate": 8.28088032513346e-06,
+      "loss": 0.7654,
+      "step": 9963
+    },
+    {
+      "epoch": 0.5484066266718036,
+      "grad_norm": 0.7758432626724243,
+      "learning_rate": 8.28055321550306e-06,
+      "loss": 0.8033,
+      "step": 9964
+    },
+    {
+      "epoch": 0.5484616654741593,
+      "grad_norm": 0.7074280977249146,
+      "learning_rate": 8.2802260812169e-06,
+      "loss": 0.7302,
+      "step": 9965
+    },
+    {
+      "epoch": 0.5485167042765149,
+      "grad_norm": 0.7724928259849548,
+      "learning_rate": 8.27989892227744e-06,
+      "loss": 0.7621,
+      "step": 9966
+    },
+    {
+      "epoch": 0.5485717430788706,
+      "grad_norm": 0.7364168167114258,
+      "learning_rate": 8.279571738687137e-06,
+      "loss": 0.7587,
+      "step": 9967
+    },
+    {
+      "epoch": 0.5486267818812263,
+      "grad_norm": 0.7298350930213928,
+      "learning_rate": 8.27924453044845e-06,
+      "loss": 0.7371,
+      "step": 9968
+    },
+    {
+      "epoch": 0.548681820683582,
+      "grad_norm": 0.8056737780570984,
+      "learning_rate": 8.27891729756384e-06,
+      "loss": 0.9871,
+      "step": 9969
+    },
+    {
+      "epoch": 0.5487368594859375,
+      "grad_norm": 0.7499688267707825,
+      "learning_rate": 8.278590040035763e-06,
+      "loss": 0.8574,
+      "step": 9970
+    },
+    {
+      "epoch": 0.5487918982882932,
+      "grad_norm": 0.7398175001144409,
+      "learning_rate": 8.278262757866683e-06,
+      "loss": 0.744,
+      "step": 9971
+    },
+    {
+      "epoch": 0.5488469370906489,
+      "grad_norm": 0.7099171876907349,
+      "learning_rate": 8.277935451059058e-06,
+      "loss": 0.7108,
+      "step": 9972
+    },
+    {
+      "epoch": 0.5489019758930046,
+      "grad_norm": 0.6720188856124878,
+      "learning_rate": 8.277608119615345e-06,
+      "loss": 0.8565,
+      "step": 9973
+    },
+    {
+      "epoch": 0.5489570146953602,
+      "grad_norm": 0.7870737910270691,
+      "learning_rate": 8.27728076353801e-06,
+      "loss": 0.7429,
+      "step": 9974
+    },
+    {
+      "epoch": 0.5490120534977159,
+      "grad_norm": 0.7358133792877197,
+      "learning_rate": 8.276953382829507e-06,
+      "loss": 0.7549,
+      "step": 9975
+    },
+    {
+      "epoch": 0.5490670923000716,
+      "grad_norm": 0.8968467116355896,
+      "learning_rate": 8.276625977492303e-06,
+      "loss": 0.6983,
+      "step": 9976
+    },
+    {
+      "epoch": 0.5491221311024272,
+      "grad_norm": 0.7346875071525574,
+      "learning_rate": 8.276298547528852e-06,
+      "loss": 0.8541,
+      "step": 9977
+    },
+    {
+      "epoch": 0.5491771699047828,
+      "grad_norm": 0.7297229170799255,
+      "learning_rate": 8.27597109294162e-06,
+      "loss": 0.8378,
+      "step": 9978
+    },
+    {
+      "epoch": 0.5492322087071385,
+      "grad_norm": 0.6907635927200317,
+      "learning_rate": 8.275643613733064e-06,
+      "loss": 0.7058,
+      "step": 9979
+    },
+    {
+      "epoch": 0.5492872475094942,
+      "grad_norm": 0.7612239718437195,
+      "learning_rate": 8.27531610990565e-06,
+      "loss": 0.6827,
+      "step": 9980
+    },
+    {
+      "epoch": 0.5493422863118499,
+      "grad_norm": 1.3160386085510254,
+      "learning_rate": 8.274988581461837e-06,
+      "loss": 0.7357,
+      "step": 9981
+    },
+    {
+      "epoch": 0.5493973251142055,
+      "grad_norm": 0.6370541453361511,
+      "learning_rate": 8.274661028404083e-06,
+      "loss": 0.7323,
+      "step": 9982
+    },
+    {
+      "epoch": 0.5494523639165612,
+      "grad_norm": 0.7051724195480347,
+      "learning_rate": 8.274333450734856e-06,
+      "loss": 0.7714,
+      "step": 9983
+    },
+    {
+      "epoch": 0.5495074027189168,
+      "grad_norm": 0.7452969551086426,
+      "learning_rate": 8.274005848456614e-06,
+      "loss": 0.7516,
+      "step": 9984
+    },
+    {
+      "epoch": 0.5495624415212725,
+      "grad_norm": 0.7132626175880432,
+      "learning_rate": 8.273678221571823e-06,
+      "loss": 0.6417,
+      "step": 9985
+    },
+    {
+      "epoch": 0.5496174803236281,
+      "grad_norm": 0.7873446345329285,
+      "learning_rate": 8.273350570082941e-06,
+      "loss": 0.8457,
+      "step": 9986
+    },
+    {
+      "epoch": 0.5496725191259838,
+      "grad_norm": 0.691470205783844,
+      "learning_rate": 8.273022893992432e-06,
+      "loss": 0.7871,
+      "step": 9987
+    },
+    {
+      "epoch": 0.5497275579283395,
+      "grad_norm": 0.6671431064605713,
+      "learning_rate": 8.27269519330276e-06,
+      "loss": 0.6919,
+      "step": 9988
+    },
+    {
+      "epoch": 0.5497825967306952,
+      "grad_norm": 0.8026914596557617,
+      "learning_rate": 8.272367468016387e-06,
+      "loss": 0.6885,
+      "step": 9989
+    },
+    {
+      "epoch": 0.5498376355330508,
+      "grad_norm": 0.9003152251243591,
+      "learning_rate": 8.272039718135774e-06,
+      "loss": 0.7671,
+      "step": 9990
+    },
+    {
+      "epoch": 0.5498926743354065,
+      "grad_norm": 0.6515254378318787,
+      "learning_rate": 8.271711943663388e-06,
+      "loss": 0.7589,
+      "step": 9991
+    },
+    {
+      "epoch": 0.5499477131377621,
+      "grad_norm": 0.6495782136917114,
+      "learning_rate": 8.27138414460169e-06,
+      "loss": 0.7277,
+      "step": 9992
+    },
+    {
+      "epoch": 0.5500027519401178,
+      "grad_norm": 0.7564565539360046,
+      "learning_rate": 8.271056320953146e-06,
+      "loss": 0.6977,
+      "step": 9993
+    },
+    {
+      "epoch": 0.5500577907424734,
+      "grad_norm": 0.8551548719406128,
+      "learning_rate": 8.270728472720218e-06,
+      "loss": 0.684,
+      "step": 9994
+    },
+    {
+      "epoch": 0.5501128295448291,
+      "grad_norm": 0.6614843010902405,
+      "learning_rate": 8.270400599905369e-06,
+      "loss": 0.6559,
+      "step": 9995
+    },
+    {
+      "epoch": 0.5501678683471848,
+      "grad_norm": 0.6920068264007568,
+      "learning_rate": 8.270072702511065e-06,
+      "loss": 0.7497,
+      "step": 9996
+    },
+    {
+      "epoch": 0.5502229071495405,
+      "grad_norm": 0.7426198124885559,
+      "learning_rate": 8.26974478053977e-06,
+      "loss": 0.7434,
+      "step": 9997
+    },
+    {
+      "epoch": 0.550277945951896,
+      "grad_norm": 1.2630934715270996,
+      "learning_rate": 8.269416833993949e-06,
+      "loss": 0.7306,
+      "step": 9998
+    },
+    {
+      "epoch": 0.5503329847542517,
+      "grad_norm": 0.7069457769393921,
+      "learning_rate": 8.269088862876066e-06,
+      "loss": 0.6735,
+      "step": 9999
     }
   ],
   "logging_steps": 1,
@@ -57293,7 +70019,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 2.414263242709795e+19,
+  "total_flos": 2.9507661855341937e+19,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null