Upload trainer_state.json with huggingface_hub

trainer_state.json

@@ -2,474 +2,784 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.3822651045256147,
+  "epoch": 1.0,
   "eval_steps": 500,
-  "global_step": 650,
+  "global_step": 1089,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.036756260050539856,
-      "grad_norm": 1.136057734489441,
-      "learning_rate": 1.0975609756097562e-05,
-      "loss": 1.1345,
+      "epoch": 0.00919064851513585,
+      "grad_norm": 0.8316559195518494,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 0.499,
       "step": 10
     },
     {
-      "epoch": 0.07351252010107971,
-      "grad_norm": 0.5754098296165466,
-      "learning_rate": 2.3170731707317075e-05,
-      "loss": 0.9743,
+      "epoch": 0.0183812970302717,
+      "grad_norm": 0.8773086667060852,
+      "learning_rate": 1.7272727272727274e-05,
+      "loss": 0.5179,
       "step": 20
     },
     {
-      "epoch": 0.11026878015161957,
-      "grad_norm": 0.5011259317398071,
-      "learning_rate": 3.5365853658536584e-05,
-      "loss": 0.9405,
+      "epoch": 0.02757194554540755,
+      "grad_norm": 0.9753805994987488,
+      "learning_rate": 2.636363636363636e-05,
+      "loss": 0.5032,
       "step": 30
     },
     {
-      "epoch": 0.14702504020215942,
-      "grad_norm": 0.453047513961792,
-      "learning_rate": 4.75609756097561e-05,
-      "loss": 0.9082,
+      "epoch": 0.0367625940605434,
+      "grad_norm": 0.8634973764419556,
+      "learning_rate": 3.545454545454546e-05,
+      "loss": 0.5424,
       "step": 40
     },
     {
-      "epoch": 0.1837813002526993,
-      "grad_norm": 0.3960385024547577,
-      "learning_rate": 4.998685537302135e-05,
-      "loss": 0.8931,
+      "epoch": 0.045953242575679246,
+      "grad_norm": 0.8948274254798889,
+      "learning_rate": 4.454545454545455e-05,
+      "loss": 0.5347,
       "step": 50
     },
     {
-      "epoch": 0.22053756030323915,
-      "grad_norm": 0.41752490401268005,
-      "learning_rate": 4.9933479014963055e-05,
-      "loss": 0.873,
+      "epoch": 0.0551438910908151,
+      "grad_norm": 0.9634304642677307,
+      "learning_rate": 4.999815378056639e-05,
+      "loss": 0.5524,
       "step": 60
     },
     {
-      "epoch": 0.257293820353779,
-      "grad_norm": 0.4332248270511627,
-      "learning_rate": 4.9839137016353147e-05,
-      "loss": 0.8851,
+      "epoch": 0.06433453960595094,
+      "grad_norm": 0.8109133243560791,
+      "learning_rate": 4.997738694339449e-05,
+      "loss": 0.5651,
       "step": 70
     },
     {
-      "epoch": 0.29405008040431885,
-      "grad_norm": 0.41378816962242126,
-      "learning_rate": 4.970398438084758e-05,
-      "loss": 0.9094,
+      "epoch": 0.0735251881210868,
+      "grad_norm": 0.8288258910179138,
+      "learning_rate": 4.993356472749955e-05,
+      "loss": 0.6061,
       "step": 80
     },
     {
-      "epoch": 0.33080634045485874,
-      "grad_norm": 0.40073126554489136,
-      "learning_rate": 4.952824316387163e-05,
-      "loss": 0.8963,
+      "epoch": 0.08271583663622264,
+      "grad_norm": 0.7533223628997803,
+      "learning_rate": 4.986672758298095e-05,
+      "loss": 0.6184,
       "step": 90
     },
     {
-      "epoch": 0.3675626005053986,
-      "grad_norm": 0.49767372012138367,
-      "learning_rate": 4.931220210778332e-05,
-      "loss": 0.8639,
+      "epoch": 0.09190648515135849,
+      "grad_norm": 0.8546445369720459,
+      "learning_rate": 4.97769372038695e-05,
+      "loss": 0.5879,
       "step": 100
     },
     {
-      "epoch": 0.4043188605559384,
-      "grad_norm": 0.3762946128845215,
-      "learning_rate": 4.905621616747054e-05,
-      "loss": 0.858,
+      "epoch": 0.10109713366649434,
+      "grad_norm": 0.8864842057228088,
+      "learning_rate": 4.966427647118077e-05,
+      "loss": 0.6273,
       "step": 110
     },
     {
-      "epoch": 0.4410751206064783,
-      "grad_norm": 0.37841302156448364,
-      "learning_rate": 4.876070592716105e-05,
-      "loss": 0.8514,
+      "epoch": 0.1102877821816302,
+      "grad_norm": 0.7892295122146606,
+      "learning_rate": 4.952884937641176e-05,
+      "loss": 0.579,
       "step": 120
     },
     {
-      "epoch": 0.47783138065701813,
-      "grad_norm": 0.39247384667396545,
-      "learning_rate": 4.842615690940373e-05,
-      "loss": 0.848,
+      "epoch": 0.11947843069676604,
+      "grad_norm": 0.7917618751525879,
+      "learning_rate": 4.937078092555153e-05,
+      "loss": 0.5732,
       "step": 130
     },
     {
-      "epoch": 0.514587640707558,
-      "grad_norm": 0.36395716667175293,
-      "learning_rate": 4.80531187773563e-05,
-      "loss": 0.8735,
+      "epoch": 0.12866907921190188,
+      "grad_norm": 0.727471113204956,
+      "learning_rate": 4.919021702369445e-05,
+      "loss": 0.6135,
       "step": 140
     },
     {
-      "epoch": 0.5513439007580979,
-      "grad_norm": 0.369722455739975,
-      "learning_rate": 4.7642204431690206e-05,
-      "loss": 0.8525,
+      "epoch": 0.13785972772703775,
+      "grad_norm": 0.7850656509399414,
+      "learning_rate": 4.898732434036244e-05,
+      "loss": 0.5946,
       "step": 150
     },
     {
-      "epoch": 0.5881001608086377,
-      "grad_norm": 0.3555547297000885,
-      "learning_rate": 4.71940890035964e-05,
-      "loss": 0.8892,
+      "epoch": 0.1470503762421736,
+      "grad_norm": 0.6990341544151306,
+      "learning_rate": 4.876229015566071e-05,
+      "loss": 0.6023,
       "step": 160
     },
     {
-      "epoch": 0.6248564208591776,
-      "grad_norm": 0.36071062088012695,
-      "learning_rate": 4.6709508745546575e-05,
-      "loss": 0.835,
+      "epoch": 0.15624102475730944,
+      "grad_norm": 0.7127705216407776,
+      "learning_rate": 4.85153221874088e-05,
+      "loss": 0.608,
       "step": 170
     },
     {
-      "epoch": 0.6616126809097175,
-      "grad_norm": 0.33451011776924133,
-      "learning_rate": 4.618925982163232e-05,
-      "loss": 0.8515,
+      "epoch": 0.1654316732724453,
+      "grad_norm": 0.778217077255249,
+      "learning_rate": 4.824664839940667e-05,
+      "loss": 0.5726,
       "step": 180
     },
     {
-      "epoch": 0.6983689409602573,
-      "grad_norm": 0.3431645631790161,
-      "learning_rate": 4.563419699946956e-05,
-      "loss": 0.862,
+      "epoch": 0.17462232178758114,
+      "grad_norm": 0.7732360363006592,
+      "learning_rate": 4.795651679101267e-05,
+      "loss": 0.5802,
       "step": 190
     },
     {
-      "epoch": 0.7351252010107971,
-      "grad_norm": 0.3229110538959503,
-      "learning_rate": 4.504523224581762e-05,
-      "loss": 0.8708,
+      "epoch": 0.18381297030271698,
+      "grad_norm": 0.7755158543586731,
+      "learning_rate": 4.764519516822774e-05,
+      "loss": 0.6277,
       "step": 200
     },
     {
-      "epoch": 0.771881461061337,
-      "grad_norm": 0.3718363642692566,
-      "learning_rate": 4.442333322822028e-05,
-      "loss": 0.8445,
+      "epoch": 0.19300361881785283,
+      "grad_norm": 0.7841328382492065,
+      "learning_rate": 4.731297089649703e-05,
+      "loss": 0.6151,
       "step": 210
     },
     {
-      "epoch": 0.8086377211118768,
-      "grad_norm": 0.389436811208725,
-      "learning_rate": 4.376952172513046e-05,
-      "loss": 0.8711,
+      "epoch": 0.20219426733298868,
+      "grad_norm": 0.7142501473426819,
+      "learning_rate": 4.696015063545726e-05,
+      "loss": 0.5707,
       "step": 220
     },
     {
-      "epoch": 0.8453939811624167,
-      "grad_norm": 0.38827431201934814,
-      "learning_rate": 4.308487194713097e-05,
-      "loss": 0.884,
+      "epoch": 0.21138491584812452,
+      "grad_norm": 0.7081753015518188,
+      "learning_rate": 4.658706005587443e-05,
+      "loss": 0.605,
       "step": 230
     },
     {
-      "epoch": 0.8821502412129566,
-      "grad_norm": 0.34896695613861084,
-      "learning_rate": 4.2370508772009334e-05,
-      "loss": 0.8524,
+      "epoch": 0.2205755643632604,
+      "grad_norm": 0.7052861452102661,
+      "learning_rate": 4.619404353903353e-05,
+      "loss": 0.5783,
       "step": 240
     },
     {
-      "epoch": 0.9189065012634965,
-      "grad_norm": 0.35946622490882874,
-      "learning_rate": 4.162760589658649e-05,
-      "loss": 0.847,
+      "epoch": 0.22976621287839624,
+      "grad_norm": 0.7916448712348938,
+      "learning_rate": 4.578146385885723e-05,
+      "loss": 0.6093,
       "step": 250
     },
     {
-      "epoch": 0.9556627613140363,
-      "grad_norm": 0.4428149461746216,
-      "learning_rate": 4.0857383908336076e-05,
-      "loss": 0.8585,
+      "epoch": 0.2389568613935321,
+      "grad_norm": 0.8103572130203247,
+      "learning_rate": 4.5349701847047496e-05,
+      "loss": 0.6432,
       "step": 260
     },
     {
-      "epoch": 0.9924190213645762,
-      "grad_norm": 0.3346478044986725,
-      "learning_rate": 4.0061108279962336e-05,
-      "loss": 0.8696,
+      "epoch": 0.24814750990866793,
+      "grad_norm": 0.715683102607727,
+      "learning_rate": 4.489915604155879e-05,
+      "loss": 0.5833,
       "step": 270
     },
     {
-      "epoch": 1.025729382035378,
-      "grad_norm": 0.47293058037757874,
-      "learning_rate": 3.924008729023185e-05,
-      "loss": 0.7543,
+      "epoch": 0.25733815842380375,
+      "grad_norm": 0.7637775540351868,
+      "learning_rate": 4.4430242318727624e-05,
+      "loss": 0.6096,
       "step": 280
     },
     {
-      "epoch": 1.0624856420859177,
-      "grad_norm": 0.35213029384613037,
-      "learning_rate": 3.8395669874474915e-05,
-      "loss": 0.7643,
+      "epoch": 0.26652880693893966,
+      "grad_norm": 0.7922810912132263,
+      "learning_rate": 4.394339350939787e-05,
+      "loss": 0.6341,
       "step": 290
     },
     {
-      "epoch": 1.0992419021364577,
-      "grad_norm": 0.3495759069919586,
-      "learning_rate": 3.752924340828837e-05,
-      "loss": 0.7147,
+      "epoch": 0.2757194554540755,
+      "grad_norm": 0.6947622895240784,
+      "learning_rate": 4.34390589993962e-05,
+      "loss": 0.6911,
       "step": 300
     },
     {
-      "epoch": 1.1359981621869975,
-      "grad_norm": 0.33258387446403503,
-      "learning_rate": 3.66422314280811e-05,
-      "loss": 0.7164,
+      "epoch": 0.28491010396921135,
+      "grad_norm": 0.74177485704422,
+      "learning_rate": 4.2917704314726505e-05,
+      "loss": 0.5604,
       "step": 310
     },
     {
-      "epoch": 1.1727544222375372,
-      "grad_norm": 0.3696906566619873,
-      "learning_rate": 3.573609129220748e-05,
-      "loss": 0.6989,
+      "epoch": 0.2941007524843472,
+      "grad_norm": 0.8070971369743347,
+      "learning_rate": 4.2379810691866064e-05,
+      "loss": 0.6613,
       "step": 320
     },
     {
-      "epoch": 1.2095106822880772,
-      "grad_norm": 0.3357951045036316,
-      "learning_rate": 3.4812311786531426e-05,
-      "loss": 0.7079,
+      "epoch": 0.30329140099948304,
+      "grad_norm": 0.7629824280738831,
+      "learning_rate": 4.182587463356016e-05,
+      "loss": 0.6023,
       "step": 330
     },
     {
-      "epoch": 1.246266942338617,
-      "grad_norm": 0.3242957592010498,
-      "learning_rate": 3.387241067835513e-05,
-      "loss": 0.7193,
+      "epoch": 0.3124820495146189,
+      "grad_norm": 0.8028804063796997,
+      "learning_rate": 4.125640745052522e-05,
+      "loss": 0.6769,
       "step": 340
     },
     {
-      "epoch": 1.283023202389157,
-      "grad_norm": 0.34572961926460266,
-      "learning_rate": 3.2917932222731325e-05,
-      "loss": 0.7268,
+      "epoch": 0.32167269802975473,
+      "grad_norm": 0.7006043791770935,
+      "learning_rate": 4.067193478948344e-05,
+      "loss": 0.6306,
       "step": 350
     },
     {
-      "epoch": 1.3197794624396968,
-      "grad_norm": 0.33331671357154846,
-      "learning_rate": 3.195044462525636e-05,
-      "loss": 0.7029,
+      "epoch": 0.3308633465448906,
+      "grad_norm": 0.6782275438308716,
+      "learning_rate": 4.007299614796447e-05,
+      "loss": 0.6056,
       "step": 360
     },
     {
-      "epoch": 1.3565357224902366,
-      "grad_norm": 0.3374245762825012,
-      "learning_rate": 3.097153746551248e-05,
-      "loss": 0.7378,
+      "epoch": 0.3400539950600264,
+      "grad_norm": 0.746510922908783,
+      "learning_rate": 3.9460144376322256e-05,
+      "loss": 0.5728,
       "step": 370
     },
     {
-      "epoch": 1.3932919825407764,
-      "grad_norm": 0.32257580757141113,
-      "learning_rate": 2.998281908539272e-05,
-      "loss": 0.7119,
+      "epoch": 0.3492446435751623,
+      "grad_norm": 0.7944929599761963,
+      "learning_rate": 3.8833945167426434e-05,
+      "loss": 0.59,
       "step": 380
     },
     {
-      "epoch": 1.4300482425913164,
-      "grad_norm": 0.31712806224823,
-      "learning_rate": 2.898591394659948e-05,
-      "loss": 0.7254,
+      "epoch": 0.3584352920902981,
+      "grad_norm": 0.7733877897262573,
+      "learning_rate": 3.819497653449954e-05,
+      "loss": 0.6037,
       "step": 390
     },
     {
-      "epoch": 1.4668045026418561,
-      "grad_norm": 0.32865336537361145,
-      "learning_rate": 2.798245996165813e-05,
-      "loss": 0.7233,
+      "epoch": 0.36762594060543397,
+      "grad_norm": 0.6790263652801514,
+      "learning_rate": 3.754382827758179e-05,
+      "loss": 0.6148,
       "step": 400
     },
     {
-      "epoch": 1.5035607626923961,
-      "grad_norm": 0.33550089597702026,
-      "learning_rate": 2.697410580283107e-05,
-      "loss": 0.7412,
+      "epoch": 0.3768165891205698,
+      "grad_norm": 0.6839562058448792,
+      "learning_rate": 3.688110143911618e-05,
+      "loss": 0.5848,
       "step": 410
     },
     {
-      "epoch": 1.540317022742936,
-      "grad_norm": 0.3236851394176483,
-      "learning_rate": 2.5962508193353542e-05,
-      "loss": 0.7113,
+      "epoch": 0.38600723763570566,
+      "grad_norm": 0.8358755707740784,
+      "learning_rate": 3.620740774915622e-05,
+      "loss": 0.584,
       "step": 420
     },
     {
-      "epoch": 1.5770732827934757,
-      "grad_norm": 0.32259735465049744,
-      "learning_rate": 2.4949329185441666e-05,
-      "loss": 0.7152,
+      "epoch": 0.3951978861508415,
+      "grad_norm": 0.6800047755241394,
+      "learning_rate": 3.552336906070838e-05,
+      "loss": 0.5998,
       "step": 430
     },
     {
-      "epoch": 1.6138295428440155,
-      "grad_norm": 0.3176753520965576,
-      "learning_rate": 2.3936233429545054e-05,
-      "loss": 0.7301,
+      "epoch": 0.40438853466597735,
+      "grad_norm": 0.7366592884063721,
+      "learning_rate": 3.482961677573074e-05,
+      "loss": 0.5783,
       "step": 440
     },
     {
-      "epoch": 1.6505858028945555,
-      "grad_norm": 0.3171616792678833,
-      "learning_rate": 2.2924885439330477e-05,
-      "loss": 0.71,
+      "epoch": 0.4135791831811132,
+      "grad_norm": 0.7072996497154236,
+      "learning_rate": 3.41267912623173e-05,
+      "loss": 0.561,
       "step": 450
     },
     {
-      "epoch": 1.6873420629450955,
-      "grad_norm": 0.34114307165145874,
-      "learning_rate": 2.1916946856890293e-05,
-      "loss": 0.7058,
+      "epoch": 0.42276983169624904,
+      "grad_norm": 0.7341826558113098,
+      "learning_rate": 3.341554126360619e-05,
+      "loss": 0.5687,
       "step": 460
     },
     {
-      "epoch": 1.7240983229956353,
-      "grad_norm": 0.3279208838939667,
-      "learning_rate": 2.0914073722668788e-05,
-      "loss": 0.6893,
+      "epoch": 0.4319604802113849,
+      "grad_norm": 0.9071962237358093,
+      "learning_rate": 3.269652329895728e-05,
+      "loss": 0.6271,
       "step": 470
     },
     {
-      "epoch": 1.760854583046175,
-      "grad_norm": 0.34568503499031067,
-      "learning_rate": 1.9917913754592183e-05,
-      "loss": 0.7079,
+      "epoch": 0.4411511287265208,
+      "grad_norm": 0.664733350276947,
+      "learning_rate": 3.1970401057951905e-05,
+      "loss": 0.5826,
       "step": 480
     },
     {
-      "epoch": 1.7976108430967148,
-      "grad_norm": 0.3167240023612976,
-      "learning_rate": 1.8930103640872313e-05,
-      "loss": 0.7249,
+      "epoch": 0.45034177724165664,
+      "grad_norm": 0.7106197476387024,
+      "learning_rate": 3.12378447877742e-05,
+      "loss": 0.581,
       "step": 490
     },
     {
-      "epoch": 1.8343671031472548,
-      "grad_norm": 0.31115061044692993,
-      "learning_rate": 1.795226635093226e-05,
-      "loss": 0.686,
+      "epoch": 0.4595324257567925,
+      "grad_norm": 0.7017560005187988,
+      "learning_rate": 3.049953067453934e-05,
+      "loss": 0.5701,
       "step": 500
     },
     {
-      "epoch": 1.8711233631977946,
-      "grad_norm": 0.31580448150634766,
-      "learning_rate": 1.6986008468871783e-05,
-      "loss": 0.6973,
+      "epoch": 0.46872307427192833,
+      "grad_norm": 0.7051500678062439,
+      "learning_rate": 2.9756140219139957e-05,
+      "loss": 0.5772,
       "step": 510
     },
     {
-      "epoch": 1.9078796232483346,
-      "grad_norm": 0.3217413127422333,
-      "learning_rate": 1.6032917553853936e-05,
-      "loss": 0.6929,
+      "epoch": 0.4779137227870642,
+      "grad_norm": 0.7792097330093384,
+      "learning_rate": 2.900835960818664e-05,
+      "loss": 0.6002,
       "step": 520
     },
     {
-      "epoch": 1.9446358832988744,
-      "grad_norm": 0.31961649656295776,
-      "learning_rate": 1.509455953174948e-05,
-      "loss": 0.6886,
+      "epoch": 0.4871043713022,
+      "grad_norm": 0.7199306488037109,
+      "learning_rate": 2.8256879080623394e-05,
+      "loss": 0.5871,
       "step": 530
     },
     {
-      "epoch": 1.9813921433494142,
-      "grad_norm": 0.32721585035324097,
-      "learning_rate": 1.4172476122324806e-05,
-      "loss": 0.6806,
+      "epoch": 0.49629501981733587,
+      "grad_norm": 0.7342507839202881,
+      "learning_rate": 2.7502392290602463e-05,
+      "loss": 0.6032,
       "step": 540
     },
     {
-      "epoch": 2.014702504020216,
-      "grad_norm": 0.513198733329773,
-      "learning_rate": 1.3268182306200405e-05,
-      "loss": 0.6179,
+      "epoch": 0.5054856683324717,
+      "grad_norm": 0.7658795118331909,
+      "learning_rate": 2.6745595667206873e-05,
+      "loss": 0.595,
       "step": 550
     },
     {
-      "epoch": 2.051458764070756,
-      "grad_norm": 0.3842564523220062,
-      "learning_rate": 1.2383163835741692e-05,
-      "loss": 0.5208,
+      "epoch": 0.5146763168476075,
+      "grad_norm": 0.7520851492881775,
+      "learning_rate": 2.598718777161152e-05,
+      "loss": 0.6061,
       "step": 560
     },
     {
-      "epoch": 2.0882150241212956,
-      "grad_norm": 0.3991941809654236,
-      "learning_rate": 1.1518874793971754e-05,
-      "loss": 0.5122,
+      "epoch": 0.5238669653627434,
+      "grad_norm": 0.7485829591751099,
+      "learning_rate": 2.5227868652276205e-05,
+      "loss": 0.5787,
       "step": 570
     },
     {
-      "epoch": 2.1249712841718353,
-      "grad_norm": 0.3792611062526703,
-      "learning_rate": 1.0676735205516788e-05,
-      "loss": 0.4966,
+      "epoch": 0.5330576138778793,
+      "grad_norm": 0.7010724544525146,
+      "learning_rate": 2.446833919876584e-05,
+      "loss": 0.6055,
       "step": 580
     },
     {
-      "epoch": 2.1617275442223756,
-      "grad_norm": 0.41136300563812256,
-      "learning_rate": 9.858128703509348e-06,
-      "loss": 0.5039,
+      "epoch": 0.5422482623930152,
+      "grad_norm": 0.7274252772331238,
+      "learning_rate": 2.3709300494794347e-05,
+      "loss": 0.5696,
       "step": 590
     },
     {
-      "epoch": 2.1984838042729153,
-      "grad_norm": 0.40203046798706055,
-      "learning_rate": 9.064400256282757e-06,
-      "loss": 0.5089,
+      "epoch": 0.551438910908151,
+      "grad_norm": 0.6527493000030518,
+      "learning_rate": 2.295145317108924e-05,
+      "loss": 0.5651,
       "step": 600
     },
     {
-      "epoch": 2.235240064323455,
-      "grad_norm": 0.4289482533931732,
-      "learning_rate": 8.296853957591638e-06,
-      "loss": 0.4895,
+      "epoch": 0.5606295594232868,
+      "grad_norm": 0.752870500087738,
+      "learning_rate": 2.219549675867439e-05,
+      "loss": 0.6001,
       "step": 610
     },
     {
-      "epoch": 2.271996324373995,
-      "grad_norm": 0.4751777946949005,
-      "learning_rate": 7.556750883989324e-06,
-      "loss": 0.4958,
+      "epoch": 0.5698202079384227,
+      "grad_norm": 0.8178837895393372,
+      "learning_rate": 2.1442129043167874e-05,
+      "loss": 0.6227,
       "step": 620
     },
     {
-      "epoch": 2.3087525844245347,
-      "grad_norm": 0.39896517992019653,
-      "learning_rate": 6.845307022882405e-06,
-      "loss": 0.4994,
+      "epoch": 0.5790108564535585,
+      "grad_norm": 0.7618655562400818,
+      "learning_rate": 2.0692045420690925e-05,
+      "loss": 0.5446,
       "step": 630
     },
     {
-      "epoch": 2.3455088444750745,
-      "grad_norm": 0.40157681703567505,
-      "learning_rate": 6.1636912746665695e-06,
-      "loss": 0.5074,
+      "epoch": 0.5882015049686944,
+      "grad_norm": 0.7821996808052063,
+      "learning_rate": 1.9945938255982466e-05,
+      "loss": 0.6276,
       "step": 640
     },
     {
-      "epoch": 2.3822651045256147,
-      "grad_norm": 0.3953118622303009,
-      "learning_rate": 5.5130235322263634e-06,
-      "loss": 0.4776,
+      "epoch": 0.5973921534838302,
+      "grad_norm": 0.7382956147193909,
+      "learning_rate": 1.920449624331179e-05,
+      "loss": 0.5368,
       "step": 650
+    },
+    {
+      "epoch": 0.6065828019989661,
+      "grad_norm": 0.7067505717277527,
+      "learning_rate": 1.8468403770779337e-05,
+      "loss": 0.5906,
+      "step": 660
+    },
+    {
+      "epoch": 0.6157734505141019,
+      "grad_norm": 0.7150029540061951,
+      "learning_rate": 1.7738340288592103e-05,
+      "loss": 0.5398,
+      "step": 670
+    },
+    {
+      "epoch": 0.6249640990292378,
+      "grad_norm": 0.6986609697341919,
+      "learning_rate": 1.7014979681897164e-05,
+      "loss": 0.5463,
+      "step": 680
+    },
+    {
+      "epoch": 0.6341547475443736,
+      "grad_norm": 0.8039021492004395,
+      "learning_rate": 1.629898964875189e-05,
+      "loss": 0.5595,
+      "step": 690
+    },
+    {
+      "epoch": 0.6433453960595095,
+      "grad_norm": 0.772060751914978,
+      "learning_rate": 1.5591031083805265e-05,
+      "loss": 0.5648,
+      "step": 700
+    },
+    {
+      "epoch": 0.6525360445746453,
+      "grad_norm": 0.7191370129585266,
+      "learning_rate": 1.4891757468258983e-05,
+      "loss": 0.5658,
+      "step": 710
+    },
+    {
+      "epoch": 0.6617266930897812,
+      "grad_norm": 0.692693293094635,
+      "learning_rate": 1.4201814266671603e-05,
+      "loss": 0.5544,
+      "step": 720
+    },
+    {
+      "epoch": 0.670917341604917,
+      "grad_norm": 0.8597797155380249,
+      "learning_rate": 1.3521838331162439e-05,
+      "loss": 0.5671,
+      "step": 730
+    },
+    {
+      "epoch": 0.6801079901200529,
+      "grad_norm": 0.7617805600166321,
+      "learning_rate": 1.2852457313565187e-05,
+      "loss": 0.5965,
+      "step": 740
+    },
+    {
+      "epoch": 0.6892986386351887,
+      "grad_norm": 0.7422612905502319,
+      "learning_rate": 1.2194289086073837e-05,
+      "loss": 0.5756,
+      "step": 750
+    },
+    {
+      "epoch": 0.6984892871503245,
+      "grad_norm": 0.6952013969421387,
+      "learning_rate": 1.1547941170915686e-05,
+      "loss": 0.5227,
+      "step": 760
+    },
+    {
+      "epoch": 0.7076799356654604,
+      "grad_norm": 0.7390100359916687,
+      "learning_rate": 1.0914010179577952e-05,
+      "loss": 0.5638,
+      "step": 770
+    },
+    {
+      "epoch": 0.7168705841805962,
+      "grad_norm": 0.8268740177154541,
+      "learning_rate": 1.0293081262105448e-05,
+      "loss": 0.5406,
+      "step": 780
+    },
+    {
+      "epoch": 0.7260612326957321,
+      "grad_norm": 0.8330510854721069,
+      "learning_rate": 9.685727566977837e-06,
+      "loss": 0.5548,
+      "step": 790
+    },
+    {
+      "epoch": 0.7352518812108679,
+      "grad_norm": 0.7260822057723999,
+      "learning_rate": 9.092509712064875e-06,
+      "loss": 0.6003,
+      "step": 800
+    },
+    {
+      "epoch": 0.7444425297260038,
+      "grad_norm": 0.818632960319519,
+      "learning_rate": 8.513975267148013e-06,
+      "loss": 0.5384,
+      "step": 810
+    },
+    {
+      "epoch": 0.7536331782411396,
+      "grad_norm": 0.7371208667755127,
+      "learning_rate": 7.950658248486131e-06,
+      "loss": 0.5694,
+      "step": 820
+    },
+    {
+      "epoch": 0.7628238267562755,
+      "grad_norm": 0.7147781848907471,
+      "learning_rate": 7.403078625891757e-06,
+      "loss": 0.5314,
+      "step": 830
+    },
+    {
+      "epoch": 0.7720144752714113,
+      "grad_norm": 0.7499004006385803,
+      "learning_rate": 6.8717418427729054e-06,
+      "loss": 0.5395,
+      "step": 840
+    },
+    {
+      "epoch": 0.7812051237865472,
+      "grad_norm": 0.8400412797927856,
+      "learning_rate": 6.357138349583497e-06,
+      "loss": 0.5305,
+      "step": 850
+    },
+    {
+      "epoch": 0.790395772301683,
+      "grad_norm": 0.783181369304657,
+      "learning_rate": 5.859743151113059e-06,
+      "loss": 0.5798,
+      "step": 860
+    },
+    {
+      "epoch": 0.7995864208168189,
+      "grad_norm": 0.7410508394241333,
+      "learning_rate": 5.380015368033476e-06,
+      "loss": 0.5668,
+      "step": 870
+    },
+    {
+      "epoch": 0.8087770693319547,
+      "grad_norm": 0.8269330263137817,
+      "learning_rate": 4.918397813107678e-06,
+      "loss": 0.5836,
+      "step": 880
+    },
+    {
+      "epoch": 0.8179677178470905,
+      "grad_norm": 0.8294151425361633,
+      "learning_rate": 4.4753165824512335e-06,
+      "loss": 0.562,
+      "step": 890
+    },
+    {
+      "epoch": 0.8271583663622264,
+      "grad_norm": 0.798109233379364,
+      "learning_rate": 4.051180662224358e-06,
+      "loss": 0.5628,
+      "step": 900
+    },
+    {
+      "epoch": 0.8363490148773622,
+      "grad_norm": 0.7758229970932007,
+      "learning_rate": 3.6463815511171646e-06,
+      "loss": 0.5716,
+      "step": 910
+    },
+    {
+      "epoch": 0.8455396633924981,
+      "grad_norm": 0.8326278328895569,
+      "learning_rate": 3.2612928989767894e-06,
+      "loss": 0.6003,
+      "step": 920
+    },
+    {
+      "epoch": 0.8547303119076339,
+      "grad_norm": 0.8162908554077148,
+      "learning_rate": 2.8962701619098754e-06,
+      "loss": 0.5603,
+      "step": 930
+    },
+    {
+      "epoch": 0.8639209604227698,
+      "grad_norm": 0.7787085771560669,
+      "learning_rate": 2.5516502741787963e-06,
+      "loss": 0.5119,
+      "step": 940
+    },
+    {
+      "epoch": 0.8731116089379056,
+      "grad_norm": 0.7192712426185608,
+      "learning_rate": 2.2277513371944365e-06,
+      "loss": 0.5594,
+      "step": 950
+    },
+    {
+      "epoch": 0.8823022574530416,
+      "grad_norm": 0.8983786702156067,
+      "learning_rate": 1.9248723258926904e-06,
+      "loss": 0.56,
+      "step": 960
+    },
+    {
+      "epoch": 0.8914929059681774,
+      "grad_norm": 0.7046878933906555,
+      "learning_rate": 1.6432928127656206e-06,
+      "loss": 0.5398,
+      "step": 970
+    },
+    {
+      "epoch": 0.9006835544833133,
+      "grad_norm": 0.8238187432289124,
+      "learning_rate": 1.3832727098020332e-06,
+      "loss": 0.5311,
+      "step": 980
+    },
+    {
+      "epoch": 0.9098742029984491,
+      "grad_norm": 0.8765665888786316,
+      "learning_rate": 1.1450520285757161e-06,
+      "loss": 0.5756,
+      "step": 990
+    },
+    {
+      "epoch": 0.919064851513585,
+      "grad_norm": 0.7299582362174988,
+      "learning_rate": 9.288506587027429e-07,
+      "loss": 0.5271,
+      "step": 1000
+    },
+    {
+      "epoch": 0.9282555000287208,
+      "grad_norm": 0.9001564383506775,
+      "learning_rate": 7.348681648723238e-07,
+      "loss": 0.5796,
+      "step": 1010
+    },
+    {
+      "epoch": 0.9374461485438567,
+      "grad_norm": 0.8588452935218811,
+      "learning_rate": 5.632836026386395e-07,
+      "loss": 0.5541,
+      "step": 1020
+    },
+    {
+      "epoch": 0.9466367970589925,
+      "grad_norm": 0.8667751550674438,
+      "learning_rate": 4.1425535314356667e-07,
+      "loss": 0.5385,
+      "step": 1030
+    },
+    {
+      "epoch": 0.9558274455741284,
+      "grad_norm": 0.7498716115951538,
+      "learning_rate": 2.8792097692297294e-07,
+      "loss": 0.5387,
+      "step": 1040
+    },
+    {
+      "epoch": 0.9650180940892642,
+      "grad_norm": 0.7537438273429871,
+      "learning_rate": 1.843970869314593e-07,
+      "loss": 0.582,
+      "step": 1050
+    },
+    {
+      "epoch": 0.9742087426044,
+      "grad_norm": 0.7668646574020386,
+      "learning_rate": 1.0377924090273771e-07,
+      "loss": 0.5493,
+      "step": 1060
+    },
+    {
+      "epoch": 0.9833993911195359,
+      "grad_norm": 0.7778952121734619,
+      "learning_rate": 4.614185314509856e-08,
+      "loss": 0.5798,
+      "step": 1070
+    },
+    {
+      "epoch": 0.9925900396346717,
+      "grad_norm": 0.8097951412200928,
+      "learning_rate": 1.1538125853238768e-08,
+      "loss": 0.5628,
+      "step": 1080
+    },
+    {
+      "epoch": 1.0,
+      "step": 1089,
+      "total_flos": 6818710278897664.0,
+      "train_loss": 0.5775129214900694,
+      "train_runtime": 138111.1092,
+      "train_samples_per_second": 1.008,
+      "train_steps_per_second": 0.008
     }
   ],
   "logging_steps": 10,
-  "max_steps": 816,
+  "max_steps": 1089,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 3,
-  "save_steps": 50,
+  "num_train_epochs": 1,
+  "save_steps": 200,
   "stateful_callbacks": {
     "TrainerControl": {
       "args": {
@@ -477,12 +787,12 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": false
+        "should_training_stop": true
       },
       "attributes": {}
     }
   },
-  "total_flos": 1.616348144730112e+16,
+  "total_flos": 6818710278897664.0,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null