Upload 14 files

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:799aa3e96e60f5340d777ed0661bebe1ced4b6229995e9d6934bb2d9459c379e
+oid sha256:8362b3238b6d8ad257157ad1a100b5775db124e0f16793d6ae9bf9f0af7f4aab
 size 2384234968

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1b083c39fe6388e575235f5e9fc0de1f1f66b59a6209788eb53663fc005d70f1
+oid sha256:1dc6f2eb7585a3157fdebf182056facd58efd07c6d4a2f75bf05bb9c2bc6f807
 size 1064

tokenizer.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c760c4e715b8fa39254607191eca619d9b0612d5b29d3002ac512a5b6cad7d55
-size 11422934
+oid sha256:aeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4
+size 11422654

trainer_state.json

@@ -4,535 +4,325 @@
   "best_model_checkpoint": null,
   "epoch": 1.0,
   "eval_steps": 500,
-  "global_step": 3735,
+  "global_step": 4467,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.013386880856760375,
-      "grad_norm": 10.444840431213379,
-      "learning_rate": 9.995753909720682e-06,
-      "loss": 0.6863,
-      "step": 50
-    },
-    {
-      "epoch": 0.02677376171352075,
-      "grad_norm": 6.0297532081604,
-      "learning_rate": 9.98267481169144e-06,
-      "loss": 0.5107,
+      "epoch": 0.02238638907544213,
+      "grad_norm": 8.243054389953613,
+      "learning_rate": 4.993942787771599e-06,
+      "loss": 1.1762,
       "step": 100
     },
     {
-      "epoch": 0.040160642570281124,
-      "grad_norm": 10.061092376708984,
-      "learning_rate": 9.960784067018357e-06,
-      "loss": 0.461,
-      "step": 150
-    },
-    {
-      "epoch": 0.0535475234270415,
-      "grad_norm": 6.345846176147461,
-      "learning_rate": 9.930120388544258e-06,
-      "loss": 0.464,
+      "epoch": 0.04477277815088426,
+      "grad_norm": 13.294663429260254,
+      "learning_rate": 4.975555846433033e-06,
+      "loss": 1.1404,
       "step": 200
     },
     {
-      "epoch": 0.06693440428380187,
-      "grad_norm": 11.498032569885254,
-      "learning_rate": 9.890738003669029e-06,
-      "loss": 0.4628,
-      "step": 250
-    },
-    {
-      "epoch": 0.08032128514056225,
-      "grad_norm": 7.820000648498535,
-      "learning_rate": 9.842706558450787e-06,
-      "loss": 0.4477,
+      "epoch": 0.0671591672263264,
+      "grad_norm": 8.65041446685791,
+      "learning_rate": 4.9449294649220665e-06,
+      "loss": 1.0161,
       "step": 300
     },
     {
-      "epoch": 0.09370816599732262,
-      "grad_norm": 11.710145950317383,
-      "learning_rate": 9.786110994439813e-06,
-      "loss": 0.4411,
-      "step": 350
-    },
-    {
-      "epoch": 0.107095046854083,
-      "grad_norm": 7.274962902069092,
-      "learning_rate": 9.721051398463065e-06,
-      "loss": 0.4735,
+      "epoch": 0.08954555630176853,
+      "grad_norm": 10.313958168029785,
+      "learning_rate": 4.902215063682208e-06,
+      "loss": 0.9855,
       "step": 400
     },
     {
-      "epoch": 0.12048192771084337,
-      "grad_norm": 8.92360782623291,
-      "learning_rate": 9.647642825624925e-06,
-      "loss": 0.4501,
-      "step": 450
-    },
-    {
-      "epoch": 0.13386880856760375,
-      "grad_norm": 8.818862915039062,
-      "learning_rate": 9.56601509583717e-06,
-      "loss": 0.457,
+      "epoch": 0.11193194537721066,
+      "grad_norm": 10.495174407958984,
+      "learning_rate": 4.847623827752661e-06,
+      "loss": 0.9682,
       "step": 500
     },
     {
-      "epoch": 0.14725568942436412,
-      "grad_norm": 12.081962585449219,
-      "learning_rate": 9.476312564238035e-06,
-      "loss": 0.4308,
-      "step": 550
-    },
-    {
-      "epoch": 0.1606425702811245,
-      "grad_norm": 10.395655632019043,
-      "learning_rate": 9.378693865906347e-06,
-      "loss": 0.4408,
+      "epoch": 0.1343183344526528,
+      "grad_norm": 12.217084884643555,
+      "learning_rate": 4.781425662644569e-06,
+      "loss": 0.988,
       "step": 600
     },
     {
-      "epoch": 0.17402945113788487,
-      "grad_norm": 8.816801071166992,
-      "learning_rate": 9.273331635322185e-06,
-      "loss": 0.4497,
-      "step": 650
-    },
-    {
-      "epoch": 0.18741633199464525,
-      "grad_norm": 5.368481636047363,
-      "learning_rate": 9.160412201070217e-06,
-      "loss": 0.4447,
+      "epoch": 0.15670472352809492,
+      "grad_norm": 10.601492881774902,
+      "learning_rate": 4.703947859896326e-06,
+      "loss": 1.0177,
       "step": 700
     },
     {
-      "epoch": 0.20080321285140562,
-      "grad_norm": 6.392587184906006,
-      "learning_rate": 9.040135256325584e-06,
-      "loss": 0.4407,
-      "step": 750
-    },
-    {
-      "epoch": 0.214190093708166,
-      "grad_norm": 7.844484806060791,
-      "learning_rate": 8.912713505705093e-06,
-      "loss": 0.4528,
+      "epoch": 0.17909111260353705,
+      "grad_norm": 7.260547637939453,
+      "learning_rate": 4.615573478905602e-06,
+      "loss": 0.9601,
       "step": 800
     },
     {
-      "epoch": 0.22757697456492637,
-      "grad_norm": 8.644186019897461,
-      "learning_rate": 8.778372289108224e-06,
-      "loss": 0.4218,
-      "step": 850
-    },
-    {
-      "epoch": 0.24096385542168675,
-      "grad_norm": 10.462806701660156,
-      "learning_rate": 8.637349183213186e-06,
-      "loss": 0.4194,
+      "epoch": 0.20147750167897918,
+      "grad_norm": 10.645676612854004,
+      "learning_rate": 4.5167394530384775e-06,
+      "loss": 0.9422,
       "step": 900
     },
     {
-      "epoch": 0.2543507362784471,
-      "grad_norm": 8.724151611328125,
-      "learning_rate": 8.489893581332753e-06,
-      "loss": 0.4529,
-      "step": 950
-    },
-    {
-      "epoch": 0.2677376171352075,
-      "grad_norm": 8.977086067199707,
-      "learning_rate": 8.33626625237289e-06,
-      "loss": 0.4354,
+      "epoch": 0.2238638907544213,
+      "grad_norm": 10.681941986083984,
+      "learning_rate": 4.407934429379341e-06,
+      "loss": 1.0196,
       "step": 1000
     },
     {
-      "epoch": 0.28112449799196787,
-      "grad_norm": 7.824547290802002,
-      "learning_rate": 8.176738879674129e-06,
-      "loss": 0.4314,
-      "step": 1050
-    },
-    {
-      "epoch": 0.29451137884872824,
-      "grad_norm": 5.790531635284424,
-      "learning_rate": 8.01159358055124e-06,
-      "loss": 0.4286,
+      "epoch": 0.24625027982986344,
+      "grad_norm": 9.875200271606445,
+      "learning_rate": 4.28969635280205e-06,
+      "loss": 1.0191,
       "step": 1100
     },
     {
-      "epoch": 0.3078982597054886,
-      "grad_norm": 9.854269027709961,
-      "learning_rate": 7.84112240738086e-06,
-      "loss": 0.4201,
-      "step": 1150
-    },
-    {
-      "epoch": 0.321285140562249,
-      "grad_norm": 10.543339729309082,
-      "learning_rate": 7.665626831119396e-06,
-      "loss": 0.4405,
+      "epoch": 0.2686366689053056,
+      "grad_norm": 10.363883018493652,
+      "learning_rate": 4.162609806307003e-06,
+      "loss": 0.9449,
       "step": 1200
     },
     {
-      "epoch": 0.33467202141900937,
-      "grad_norm": 4.100381851196289,
-      "learning_rate": 7.485417208164567e-06,
-      "loss": 0.4256,
-      "step": 1250
-    },
-    {
-      "epoch": 0.34805890227576974,
-      "grad_norm": 6.328100681304932,
-      "learning_rate": 7.300812231503435e-06,
-      "loss": 0.4333,
+      "epoch": 0.2910230579807477,
+      "grad_norm": 8.477913856506348,
+      "learning_rate": 4.027303120773824e-06,
+      "loss": 0.9039,
       "step": 1300
     },
     {
-      "epoch": 0.3614457831325301,
-      "grad_norm": 5.430217742919922,
-      "learning_rate": 7.112138367117529e-06,
-      "loss": 0.3958,
-      "step": 1350
-    },
-    {
-      "epoch": 0.3748326639892905,
-      "grad_norm": 10.457396507263184,
-      "learning_rate": 6.91972927664176e-06,
-      "loss": 0.3899,
+      "epoch": 0.31340944705618984,
+      "grad_norm": 9.964315414428711,
+      "learning_rate": 3.884445268419355e-06,
+      "loss": 1.0146,
       "step": 1400
     },
     {
-      "epoch": 0.38821954484605087,
-      "grad_norm": 6.345710754394531,
-      "learning_rate": 6.723925227298133e-06,
-      "loss": 0.3759,
-      "step": 1450
-    },
-    {
-      "epoch": 0.40160642570281124,
-      "grad_norm": 9.008039474487305,
-      "learning_rate": 6.525072490147766e-06,
-      "loss": 0.3912,
+      "epoch": 0.33579583613163194,
+      "grad_norm": 10.20909309387207,
+      "learning_rate": 3.734742555320098e-06,
+      "loss": 1.0387,
       "step": 1500
     },
     {
-      "epoch": 0.4149933065595716,
-      "grad_norm": 7.655404090881348,
-      "learning_rate": 6.323522727725371e-06,
-      "loss": 0.3992,
-      "step": 1550
-    },
-    {
-      "epoch": 0.428380187416332,
-      "grad_norm": 7.3834967613220215,
-      "learning_rate": 6.119632372139152e-06,
-      "loss": 0.3655,
+      "epoch": 0.3581822252070741,
+      "grad_norm": 12.675057411193848,
+      "learning_rate": 3.578935129351634e-06,
+      "loss": 1.0002,
       "step": 1600
     },
     {
-      "epoch": 0.44176706827309237,
-      "grad_norm": 8.772950172424316,
-      "learning_rate": 5.913761994735908e-06,
-      "loss": 0.4217,
-      "step": 1650
-    },
-    {
-      "epoch": 0.45515394912985274,
-      "grad_norm": 5.111259937286377,
-      "learning_rate": 5.706275668446074e-06,
-      "loss": 0.4281,
+      "epoch": 0.3805686142825162,
+      "grad_norm": 11.391951560974121,
+      "learning_rate": 3.4177933208102103e-06,
+      "loss": 0.9976,
       "step": 1700
     },
     {
-      "epoch": 0.4685408299866131,
-      "grad_norm": 5.894604682922363,
-      "learning_rate": 5.497540323936371e-06,
-      "loss": 0.4276,
-      "step": 1750
-    },
-    {
-      "epoch": 0.4819277108433735,
-      "grad_norm": 9.776211738586426,
-      "learning_rate": 5.2879251007086555e-06,
-      "loss": 0.4073,
+      "epoch": 0.40295500335795836,
+      "grad_norm": 10.763954162597656,
+      "learning_rate": 3.2521138338088676e-06,
+      "loss": 0.9535,
       "step": 1800
     },
     {
-      "epoch": 0.49531459170013387,
-      "grad_norm": 8.807998657226562,
-      "learning_rate": 5.077800694292546e-06,
-      "loss": 0.4339,
-      "step": 1850
-    },
-    {
-      "epoch": 0.5087014725568942,
-      "grad_norm": 8.668907165527344,
-      "learning_rate": 4.867538700686292e-06,
-      "loss": 0.432,
+      "epoch": 0.42534139243340047,
+      "grad_norm": 9.04262638092041,
+      "learning_rate": 3.0827158072783113e-06,
+      "loss": 1.0119,
       "step": 1900
     },
     {
-      "epoch": 0.5220883534136547,
-      "grad_norm": 7.573647499084473,
-      "learning_rate": 4.657510959205182e-06,
-      "loss": 0.402,
-      "step": 1950
-    },
-    {
-      "epoch": 0.535475234270415,
-      "grad_norm": 7.700588703155518,
-      "learning_rate": 4.448088894899669e-06,
-      "loss": 0.3972,
+      "epoch": 0.4477277815088426,
+      "grad_norm": 7.757719993591309,
+      "learning_rate": 2.9104367650473923e-06,
+      "loss": 0.9353,
       "step": 2000
     },
     {
-      "epoch": 0.5488621151271754,
-      "grad_norm": 7.340878009796143,
-      "learning_rate": 4.23964286170611e-06,
-      "loss": 0.3288,
-      "step": 2050
-    },
-    {
-      "epoch": 0.5622489959839357,
-      "grad_norm": 7.881528377532959,
-      "learning_rate": 4.032541487491709e-06,
-      "loss": 0.4066,
+      "epoch": 0.47011417058428473,
+      "grad_norm": 8.288910865783691,
+      "learning_rate": 2.7361284750264927e-06,
+      "loss": 1.0068,
       "step": 2100
     },
     {
-      "epoch": 0.5756358768406962,
-      "grad_norm": 7.409366607666016,
-      "learning_rate": 3.827151022151955e-06,
-      "loss": 0.3516,
-      "step": 2150
-    },
-    {
-      "epoch": 0.5890227576974565,
-      "grad_norm": 7.922168731689453,
-      "learning_rate": 3.623834689913387e-06,
-      "loss": 0.3729,
+      "epoch": 0.4925005596597269,
+      "grad_norm": 10.480790138244629,
+      "learning_rate": 2.5606527379664746e-06,
+      "loss": 0.9621,
       "step": 2200
     },
     {
-      "epoch": 0.6024096385542169,
-      "grad_norm": 8.52475643157959,
-      "learning_rate": 3.4229520469871445e-06,
-      "loss": 0.4058,
-      "step": 2250
-    },
-    {
-      "epoch": 0.6157965194109772,
-      "grad_norm": 8.120729446411133,
-      "learning_rate": 3.2248583457092163e-06,
-      "loss": 0.3451,
+      "epoch": 0.514886948735169,
+      "grad_norm": 12.541658401489258,
+      "learning_rate": 2.384877126614103e-06,
+      "loss": 0.9756,
       "step": 2300
     },
     {
-      "epoch": 0.6291834002677377,
-      "grad_norm": 8.336807250976562,
-      "learning_rate": 3.0299039062919417e-06,
-      "loss": 0.3691,
-      "step": 2350
-    },
-    {
-      "epoch": 0.642570281124498,
-      "grad_norm": 8.232279777526855,
-      "learning_rate": 2.8384334972977275e-06,
-      "loss": 0.412,
+      "epoch": 0.5372733378106112,
+      "grad_norm": 13.887962341308594,
+      "learning_rate": 2.20967069633002e-06,
+      "loss": 0.9665,
       "step": 2400
     },
     {
-      "epoch": 0.6559571619812584,
-      "grad_norm": 5.542270183563232,
-      "learning_rate": 2.650785725930657e-06,
-      "loss": 0.3768,
-      "step": 2450
-    },
-    {
-      "epoch": 0.6693440428380187,
-      "grad_norm": 5.975722312927246,
-      "learning_rate": 2.4672924392241493e-06,
-      "loss": 0.3842,
+      "epoch": 0.5596597268860533,
+      "grad_norm": 11.765827178955078,
+      "learning_rate": 2.035899688376515e-06,
+      "loss": 0.9511,
       "step": 2500
     },
     {
-      "epoch": 0.6827309236947792,
-      "grad_norm": 6.6057868003845215,
-      "learning_rate": 2.288278137183748e-06,
-      "loss": 0.3278,
-      "step": 2550
-    },
-    {
-      "epoch": 0.6961178045515395,
-      "grad_norm": 7.060661792755127,
-      "learning_rate": 2.11405939892275e-06,
-      "loss": 0.3673,
+      "epoch": 0.5820461159614954,
+      "grad_norm": 11.959718704223633,
+      "learning_rate": 1.8644232471185239e-06,
+      "loss": 0.9713,
       "step": 2600
     },
     {
-      "epoch": 0.7095046854082999,
-      "grad_norm": 5.489977836608887,
-      "learning_rate": 1.9449443228056565e-06,
-      "loss": 0.3522,
-      "step": 2650
-    },
-    {
-      "epoch": 0.7228915662650602,
-      "grad_norm": 7.412909507751465,
-      "learning_rate": 1.7812319815894096e-06,
-      "loss": 0.366,
+      "epoch": 0.6044325050369376,
+      "grad_norm": 8.351862907409668,
+      "learning_rate": 1.6960891723125235e-06,
+      "loss": 1.0633,
       "step": 2700
     },
     {
-      "epoch": 0.7362784471218207,
-      "grad_norm": 5.627942085266113,
-      "learning_rate": 1.6232118935260628e-06,
-      "loss": 0.3906,
-      "step": 2750
-    },
-    {
-      "epoch": 0.749665327978581,
-      "grad_norm": 6.9760212898254395,
-      "learning_rate": 1.4711635103621718e-06,
-      "loss": 0.3912,
+      "epoch": 0.6268188941123797,
+      "grad_norm": 7.508358478546143,
+      "learning_rate": 1.5317297274845156e-06,
+      "loss": 0.895,
       "step": 2800
     },
     {
-      "epoch": 0.7630522088353414,
-      "grad_norm": 7.291072368621826,
-      "learning_rate": 1.325355723140367e-06,
-      "loss": 0.3753,
-      "step": 2850
-    },
-    {
-      "epoch": 0.7764390896921017,
-      "grad_norm": 6.7273478507995605,
-      "learning_rate": 1.1860463866770826e-06,
-      "loss": 0.3694,
+      "epoch": 0.6492052831878218,
+      "grad_norm": 12.4747314453125,
+      "learning_rate": 1.372157525120959e-06,
+      "loss": 0.9394,
       "step": 2900
     },
     {
-      "epoch": 0.7898259705488622,
-      "grad_norm": 7.016156196594238,
-      "learning_rate": 1.0534818635573751e-06,
-      "loss": 0.3636,
-      "step": 2950
-    },
-    {
-      "epoch": 0.8032128514056225,
-      "grad_norm": 7.053563117980957,
-      "learning_rate": 9.278965884532598e-07,
-      "loss": 0.4107,
+      "epoch": 0.6715916722632639,
+      "grad_norm": 12.205965042114258,
+      "learning_rate": 1.2181615090167711e-06,
+      "loss": 0.9485,
       "step": 3000
     },
     {
-      "epoch": 0.8165997322623829,
-      "grad_norm": 6.060437202453613,
-      "learning_rate": 8.09512653536052e-07,
-      "loss": 0.3908,
-      "step": 3050
+      "epoch": 0.6939780613387061,
+      "grad_norm": 6.783346652984619,
+      "learning_rate": 1.0705030536441147e-06,
+      "loss": 0.8932,
+      "step": 3100
     },
     {
-      "epoch": 0.8299866131191432,
-      "grad_norm": 7.361605644226074,
-      "learning_rate": 6.985394157158893e-07,
-      "loss": 0.3662,
-      "step": 3100
+      "epoch": 0.7163644504141482,
+      "grad_norm": 9.786669731140137,
+      "learning_rate": 9.299121998271918e-07,
+      "loss": 0.9629,
+      "step": 3200
     },
     {
-      "epoch": 0.8433734939759037,
-      "grad_norm": 5.547077655792236,
-      "learning_rate": 5.951731264030202e-07,
-      "loss": 0.3513,
-      "step": 3150
+      "epoch": 0.7387508394895903,
+      "grad_norm": 8.05592155456543,
+      "learning_rate": 7.970840453342679e-07,
+      "loss": 0.9799,
+      "step": 3300
     },
     {
-      "epoch": 0.856760374832664,
-      "grad_norm": 9.102234840393066,
-      "learning_rate": 4.995965844455969e-07,
-      "loss": 0.409,
-      "step": 3200
+      "epoch": 0.7611372285650324,
+      "grad_norm": 11.015290260314941,
+      "learning_rate": 6.726753082323087e-07,
+      "loss": 0.8752,
+      "step": 3400
     },
     {
-      "epoch": 0.8701472556894244,
-      "grad_norm": 6.737813472747803,
-      "learning_rate": 4.1197881285776675e-07,
-      "loss": 0.3838,
-      "step": 3250
+      "epoch": 0.7835236176404746,
+      "grad_norm": 8.78658390045166,
+      "learning_rate": 5.573010799953652e-07,
+      "loss": 0.8905,
+      "step": 3500
     },
     {
-      "epoch": 0.8835341365461847,
-      "grad_norm": 6.354872226715088,
-      "learning_rate": 3.324747599097078e-07,
-      "loss": 0.4013,
-      "step": 3300
+      "epoch": 0.8059100067159167,
+      "grad_norm": 11.667632102966309,
+      "learning_rate": 4.515317844197653e-07,
+      "loss": 0.9166,
+      "step": 3600
     },
     {
-      "epoch": 0.8969210174029452,
-      "grad_norm": 7.315886497497559,
-      "learning_rate": 2.61225025108276e-07,
-      "loss": 0.4003,
-      "step": 3350
+      "epoch": 0.8282963957913588,
+      "grad_norm": 8.403421401977539,
+      "learning_rate": 3.5589035738156305e-07,
+      "loss": 0.9322,
+      "step": 3700
     },
     {
-      "epoch": 0.9103078982597055,
-      "grad_norm": 7.8258562088012695,
-      "learning_rate": 1.9835561055279728e-07,
-      "loss": 0.3364,
-      "step": 3400
+      "epoch": 0.8506827848668009,
+      "grad_norm": 11.51596450805664,
+      "learning_rate": 2.708496613798717e-07,
+      "loss": 0.909,
+      "step": 3800
     },
     {
-      "epoch": 0.9236947791164659,
-      "grad_norm": 3.8167738914489746,
-      "learning_rate": 1.4397769810576668e-07,
-      "loss": 0.3611,
-      "step": 3450
+      "epoch": 0.8730691739422431,
+      "grad_norm": 6.4879045486450195,
+      "learning_rate": 1.9683014764887682e-07,
+      "loss": 0.9552,
+      "step": 3900
     },
     {
-      "epoch": 0.9370816599732262,
-      "grad_norm": 9.796895980834961,
-      "learning_rate": 9.818745277249076e-08,
-      "loss": 0.363,
-      "step": 3500
+      "epoch": 0.8954555630176853,
+      "grad_norm": 11.149494171142578,
+      "learning_rate": 1.3419777739733408e-07,
+      "loss": 0.9242,
+      "step": 4000
     },
     {
-      "epoch": 0.9504685408299867,
-      "grad_norm": 9.241850852966309,
-      "learning_rate": 6.106585263740528e-08,
-      "loss": 0.4065,
-      "step": 3550
+      "epoch": 0.9178419520931274,
+      "grad_norm": 8.041584014892578,
+      "learning_rate": 8.326221245317373e-08,
+      "loss": 0.9278,
+      "step": 4100
     },
     {
-      "epoch": 0.963855421686747,
-      "grad_norm": 6.8581624031066895,
-      "learning_rate": 3.267854565780326e-08,
-      "loss": 0.361,
-      "step": 3600
+      "epoch": 0.9402283411685695,
+      "grad_norm": 9.19898796081543,
+      "learning_rate": 4.427528425888977e-08,
+      "loss": 0.8939,
+      "step": 4200
     },
     {
-      "epoch": 0.9772423025435074,
-      "grad_norm": 6.801017761230469,
-      "learning_rate": 1.3075733568240212e-08,
-      "loss": 0.3531,
-      "step": 3650
+      "epoch": 0.9626147302440117,
+      "grad_norm": 9.239889144897461,
+      "learning_rate": 1.7429748787176626e-08,
+      "loss": 0.9563,
+      "step": 4300
     },
     {
-      "epoch": 0.9906291834002677,
-      "grad_norm": 7.733842372894287,
-      "learning_rate": 2.2920831009209944e-09,
-      "loss": 0.4049,
-      "step": 3700
+      "epoch": 0.9850011193194538,
+      "grad_norm": 6.8510565757751465,
+      "learning_rate": 2.8583335326598516e-09,
+      "loss": 0.961,
+      "step": 4400
     }
   ],
-  "logging_steps": 50,
-  "max_steps": 3735,
+  "logging_steps": 100,
+  "max_steps": 4467,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 1,
-  "save_steps": 50000,
+  "save_steps": 100000,
   "stateful_callbacks": {
     "TrainerControl": {
       "args": {
@@ -545,8 +335,8 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.010777670549504e+16,
-  "train_batch_size": 2,
+  "total_flos": 2.417341778362368e+16,
+  "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cd3bc88ff5b25f40cd4e4bf31742ef38e7a91fb605a571a3ce68503effacfe00
+oid sha256:18779069295d4fb496b6ef1c071438cc814f890abd5a4efec426d22b7fe46a44
 size 5304