End of training

README.md

@@ -4,6 +4,7 @@ license: llama3.1
 base_model: meta-llama/Llama-3.1-8B
 tags:
 - llama-factory
+- full
 - generated_from_trainer
 model-index:
 - name: OH_DCFT_V3_wo_camel_ai_math
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # OH_DCFT_V3_wo_camel_ai_math
 
-This model is a fine-tuned version of [meta-llama/Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) on an unknown dataset.
+This model is a fine-tuned version of [meta-llama/Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) on the mlfoundations-dev/OH_DCFT_V3_wo_camel_ai_math dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.6600
 

all_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 2.9980781550288276,
-    "eval_loss": 0.6631707549095154,
-    "eval_runtime": 210.5047,
-    "eval_samples_per_second": 49.951,
+    "eval_loss": 0.6599797010421753,
+    "eval_runtime": 210.4663,
+    "eval_samples_per_second": 49.96,
     "eval_steps_per_second": 0.394,
     "total_flos": 1959374817853440.0,
-    "train_loss": 0.6376129688360752,
-    "train_runtime": 34970.0124,
-    "train_samples_per_second": 17.138,
+    "train_loss": 0.6400190381922274,
+    "train_runtime": 35097.8295,
+    "train_samples_per_second": 17.076,
     "train_steps_per_second": 0.033
 }

eval_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 2.9980781550288276,
-    "eval_loss": 0.6631707549095154,
-    "eval_runtime": 210.5047,
-    "eval_samples_per_second": 49.951,
+    "eval_loss": 0.6599797010421753,
+    "eval_runtime": 210.4663,
+    "eval_samples_per_second": 49.96,
     "eval_steps_per_second": 0.394
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 2.9980781550288276,
     "total_flos": 1959374817853440.0,
-    "train_loss": 0.6376129688360752,
-    "train_runtime": 34970.0124,
-    "train_samples_per_second": 17.138,
+    "train_loss": 0.6400190381922274,
+    "train_runtime": 35097.8295,
+    "train_samples_per_second": 17.076,
     "train_steps_per_second": 0.033
 }

trainer_state.json

@@ -10,854 +10,854 @@
   "log_history": [
     {
       "epoch": 0.025624599615631006,
-      "grad_norm": 3.2885472038218984,
+      "grad_norm": 23.87626578586006,
       "learning_rate": 5e-06,
-      "loss": 0.9386,
+      "loss": 0.9021,
       "step": 10
     },
     {
       "epoch": 0.05124919923126201,
-      "grad_norm": 4.456300061852855,
+      "grad_norm": 12.894025041742566,
       "learning_rate": 5e-06,
-      "loss": 0.8437,
+      "loss": 0.8278,
       "step": 20
     },
     {
       "epoch": 0.07687379884689302,
-      "grad_norm": 0.995400664964247,
+      "grad_norm": 1.9039308459782107,
       "learning_rate": 5e-06,
-      "loss": 0.803,
+      "loss": 0.7942,
       "step": 30
     },
     {
       "epoch": 0.10249839846252402,
-      "grad_norm": 0.9865743145936546,
+      "grad_norm": 0.8246425522140604,
       "learning_rate": 5e-06,
-      "loss": 0.7755,
+      "loss": 0.7682,
       "step": 40
     },
     {
       "epoch": 0.12812299807815503,
-      "grad_norm": 1.1445876439209104,
+      "grad_norm": 0.7903898281300727,
       "learning_rate": 5e-06,
-      "loss": 0.7491,
+      "loss": 0.7443,
       "step": 50
     },
     {
       "epoch": 0.15374759769378604,
-      "grad_norm": 0.7021402356716058,
+      "grad_norm": 0.7618274610091537,
       "learning_rate": 5e-06,
-      "loss": 0.7363,
+      "loss": 0.733,
       "step": 60
     },
     {
       "epoch": 0.17937219730941703,
-      "grad_norm": 0.6748305295476841,
+      "grad_norm": 0.6631690211716542,
       "learning_rate": 5e-06,
-      "loss": 0.7274,
+      "loss": 0.7251,
       "step": 70
     },
     {
       "epoch": 0.20499679692504805,
-      "grad_norm": 0.8880154082614105,
+      "grad_norm": 0.7752376118116371,
       "learning_rate": 5e-06,
-      "loss": 0.7169,
+      "loss": 0.7144,
       "step": 80
     },
     {
       "epoch": 0.23062139654067906,
-      "grad_norm": 0.9734749908597081,
+      "grad_norm": 0.751359072795012,
       "learning_rate": 5e-06,
-      "loss": 0.7133,
+      "loss": 0.7106,
       "step": 90
     },
     {
       "epoch": 0.25624599615631005,
-      "grad_norm": 0.586318081087361,
+      "grad_norm": 0.5794322814005122,
       "learning_rate": 5e-06,
-      "loss": 0.709,
+      "loss": 0.7063,
       "step": 100
     },
     {
       "epoch": 0.28187059577194107,
-      "grad_norm": 0.49195087780563773,
+      "grad_norm": 0.8734788690919498,
       "learning_rate": 5e-06,
-      "loss": 0.7021,
+      "loss": 0.6994,
       "step": 110
     },
     {
       "epoch": 0.3074951953875721,
-      "grad_norm": 0.5411392841704284,
+      "grad_norm": 0.5062452804420097,
       "learning_rate": 5e-06,
-      "loss": 0.708,
+      "loss": 0.7057,
       "step": 120
     },
     {
       "epoch": 0.3331197950032031,
-      "grad_norm": 0.6121381534822533,
+      "grad_norm": 0.6542917427814133,
       "learning_rate": 5e-06,
-      "loss": 0.702,
+      "loss": 0.6997,
       "step": 130
     },
     {
       "epoch": 0.35874439461883406,
-      "grad_norm": 0.5779295696071732,
+      "grad_norm": 0.592400656203172,
       "learning_rate": 5e-06,
-      "loss": 0.708,
+      "loss": 0.7055,
       "step": 140
     },
     {
       "epoch": 0.3843689942344651,
-      "grad_norm": 0.5431053824058845,
+      "grad_norm": 0.5861404729061817,
       "learning_rate": 5e-06,
-      "loss": 0.7024,
+      "loss": 0.7002,
       "step": 150
     },
     {
       "epoch": 0.4099935938500961,
-      "grad_norm": 0.6102291649380537,
+      "grad_norm": 0.5129403767949405,
       "learning_rate": 5e-06,
-      "loss": 0.6998,
+      "loss": 0.6977,
       "step": 160
     },
     {
       "epoch": 0.4356181934657271,
-      "grad_norm": 0.562363970013211,
+      "grad_norm": 0.5869968660626531,
       "learning_rate": 5e-06,
-      "loss": 0.6984,
+      "loss": 0.6961,
       "step": 170
     },
     {
       "epoch": 0.4612427930813581,
-      "grad_norm": 0.5583907505215044,
+      "grad_norm": 0.6257829680932973,
       "learning_rate": 5e-06,
-      "loss": 0.6945,
+      "loss": 0.6928,
       "step": 180
     },
     {
       "epoch": 0.4868673926969891,
-      "grad_norm": 0.6011083145558899,
+      "grad_norm": 0.5886221575021535,
       "learning_rate": 5e-06,
-      "loss": 0.6894,
+      "loss": 0.6876,
       "step": 190
     },
     {
       "epoch": 0.5124919923126201,
-      "grad_norm": 0.7575418941889398,
+      "grad_norm": 0.6006186676094072,
       "learning_rate": 5e-06,
-      "loss": 0.7,
+      "loss": 0.6983,
       "step": 200
     },
     {
       "epoch": 0.5381165919282511,
-      "grad_norm": 0.5081456852667208,
+      "grad_norm": 0.5015888297145041,
       "learning_rate": 5e-06,
-      "loss": 0.6948,
+      "loss": 0.6931,
       "step": 210
     },
     {
       "epoch": 0.5637411915438821,
-      "grad_norm": 0.5741517105893602,
+      "grad_norm": 0.5307259707540836,
       "learning_rate": 5e-06,
-      "loss": 0.6897,
+      "loss": 0.6881,
       "step": 220
     },
     {
       "epoch": 0.5893657911595132,
-      "grad_norm": 0.6030704306992282,
+      "grad_norm": 0.6093913911692712,
       "learning_rate": 5e-06,
-      "loss": 0.6813,
+      "loss": 0.6797,
       "step": 230
     },
     {
       "epoch": 0.6149903907751442,
-      "grad_norm": 0.6775227505967703,
+      "grad_norm": 0.6093826734084604,
       "learning_rate": 5e-06,
-      "loss": 0.6852,
+      "loss": 0.6838,
       "step": 240
     },
     {
       "epoch": 0.6406149903907752,
-      "grad_norm": 0.6338125454546812,
+      "grad_norm": 0.607941001127991,
       "learning_rate": 5e-06,
-      "loss": 0.6804,
+      "loss": 0.6789,
       "step": 250
     },
     {
       "epoch": 0.6662395900064062,
-      "grad_norm": 0.5669879790281377,
+      "grad_norm": 0.4897426114233287,
       "learning_rate": 5e-06,
-      "loss": 0.6855,
+      "loss": 0.684,
       "step": 260
     },
     {
       "epoch": 0.6918641896220371,
-      "grad_norm": 0.586776540269306,
+      "grad_norm": 0.4483386511091874,
       "learning_rate": 5e-06,
-      "loss": 0.6782,
+      "loss": 0.6769,
       "step": 270
     },
     {
       "epoch": 0.7174887892376681,
-      "grad_norm": 0.6203573082289723,
+      "grad_norm": 0.5533227188891904,
       "learning_rate": 5e-06,
-      "loss": 0.6796,
+      "loss": 0.6785,
       "step": 280
     },
     {
       "epoch": 0.7431133888532991,
-      "grad_norm": 0.5394149301983193,
+      "grad_norm": 0.5333807928895044,
       "learning_rate": 5e-06,
-      "loss": 0.6878,
+      "loss": 0.6866,
       "step": 290
     },
     {
       "epoch": 0.7687379884689302,
-      "grad_norm": 0.5647794597410413,
+      "grad_norm": 0.5209254483197653,
       "learning_rate": 5e-06,
-      "loss": 0.6774,
+      "loss": 0.6762,
       "step": 300
     },
     {
       "epoch": 0.7943625880845612,
-      "grad_norm": 0.5886707944345854,
+      "grad_norm": 0.45530525150524676,
       "learning_rate": 5e-06,
-      "loss": 0.6784,
+      "loss": 0.6773,
       "step": 310
     },
     {
       "epoch": 0.8199871877001922,
-      "grad_norm": 0.7892741242506496,
+      "grad_norm": 0.6616235203126,
       "learning_rate": 5e-06,
-      "loss": 0.6802,
+      "loss": 0.679,
       "step": 320
     },
     {
       "epoch": 0.8456117873158232,
-      "grad_norm": 0.6133781414876699,
+      "grad_norm": 0.5707652588164589,
       "learning_rate": 5e-06,
-      "loss": 0.6721,
+      "loss": 0.671,
       "step": 330
     },
     {
       "epoch": 0.8712363869314542,
-      "grad_norm": 0.6377632743431111,
+      "grad_norm": 0.6197974812556873,
       "learning_rate": 5e-06,
-      "loss": 0.6711,
+      "loss": 0.6701,
       "step": 340
     },
     {
       "epoch": 0.8968609865470852,
-      "grad_norm": 0.530582926146814,
+      "grad_norm": 0.5705968336277791,
       "learning_rate": 5e-06,
-      "loss": 0.6752,
+      "loss": 0.674,
       "step": 350
     },
     {
       "epoch": 0.9224855861627163,
-      "grad_norm": 0.520444529526131,
+      "grad_norm": 0.46125889577293033,
       "learning_rate": 5e-06,
-      "loss": 0.6685,
+      "loss": 0.6675,
       "step": 360
     },
     {
       "epoch": 0.9481101857783472,
-      "grad_norm": 0.575279318340331,
+      "grad_norm": 0.5263500841671853,
       "learning_rate": 5e-06,
-      "loss": 0.6824,
+      "loss": 0.6816,
       "step": 370
     },
     {
       "epoch": 0.9737347853939782,
-      "grad_norm": 0.8704274805859955,
+      "grad_norm": 0.6701180427430501,
       "learning_rate": 5e-06,
-      "loss": 0.6763,
+      "loss": 0.6753,
       "step": 380
     },
     {
       "epoch": 0.9993593850096092,
-      "grad_norm": 0.5818270555924433,
+      "grad_norm": 0.6408384613157992,
       "learning_rate": 5e-06,
-      "loss": 0.6796,
+      "loss": 0.6786,
       "step": 390
     },
     {
       "epoch": 0.9993593850096092,
-      "eval_loss": 0.6705073118209839,
-      "eval_runtime": 210.2321,
-      "eval_samples_per_second": 50.016,
+      "eval_loss": 0.669628381729126,
+      "eval_runtime": 210.2529,
+      "eval_samples_per_second": 50.011,
       "eval_steps_per_second": 0.395,
       "step": 390
     },
     {
       "epoch": 1.0249839846252402,
-      "grad_norm": 0.7758678598730366,
+      "grad_norm": 0.6401466301705097,
       "learning_rate": 5e-06,
-      "loss": 0.6304,
+      "loss": 0.6342,
       "step": 400
     },
     {
       "epoch": 1.0506085842408712,
-      "grad_norm": 0.6249211500646218,
+      "grad_norm": 0.561321939868088,
       "learning_rate": 5e-06,
-      "loss": 0.6246,
+      "loss": 0.6286,
       "step": 410
     },
     {
       "epoch": 1.0762331838565022,
-      "grad_norm": 0.6047928898745287,
+      "grad_norm": 0.5873132600959697,
       "learning_rate": 5e-06,
-      "loss": 0.6304,
+      "loss": 0.634,
       "step": 420
     },
     {
       "epoch": 1.1018577834721333,
-      "grad_norm": 0.5812394406224998,
+      "grad_norm": 0.4982492625003527,
       "learning_rate": 5e-06,
-      "loss": 0.6259,
+      "loss": 0.6297,
       "step": 430
     },
     {
       "epoch": 1.1274823830877643,
-      "grad_norm": 0.522119175741154,
+      "grad_norm": 0.5451894371727711,
       "learning_rate": 5e-06,
-      "loss": 0.6269,
+      "loss": 0.6307,
       "step": 440
     },
     {
       "epoch": 1.1531069827033953,
-      "grad_norm": 0.5901119686651206,
+      "grad_norm": 0.505854015646168,
       "learning_rate": 5e-06,
-      "loss": 0.6245,
+      "loss": 0.6283,
       "step": 450
     },
     {
       "epoch": 1.1787315823190263,
-      "grad_norm": 0.5228301365724772,
+      "grad_norm": 0.47382835639154014,
       "learning_rate": 5e-06,
-      "loss": 0.6258,
+      "loss": 0.6296,
       "step": 460
     },
     {
       "epoch": 1.2043561819346573,
-      "grad_norm": 0.5223355205495337,
+      "grad_norm": 0.49475730559885767,
       "learning_rate": 5e-06,
-      "loss": 0.6308,
+      "loss": 0.6347,
       "step": 470
     },
     {
       "epoch": 1.2299807815502883,
-      "grad_norm": 0.5682544518265831,
+      "grad_norm": 0.4985567194890561,
       "learning_rate": 5e-06,
-      "loss": 0.6277,
+      "loss": 0.6315,
       "step": 480
     },
     {
       "epoch": 1.2556053811659194,
-      "grad_norm": 0.5767233826653153,
+      "grad_norm": 0.5560313955644473,
       "learning_rate": 5e-06,
-      "loss": 0.624,
+      "loss": 0.6275,
       "step": 490
     },
     {
       "epoch": 1.2812299807815504,
-      "grad_norm": 0.5410081056861039,
+      "grad_norm": 0.49255815849421947,
       "learning_rate": 5e-06,
-      "loss": 0.6099,
+      "loss": 0.6134,
       "step": 500
     },
     {
       "epoch": 1.3068545803971814,
-      "grad_norm": 0.5399164809119454,
+      "grad_norm": 0.5136446413395007,
       "learning_rate": 5e-06,
-      "loss": 0.6259,
+      "loss": 0.6294,
       "step": 510
     },
     {
       "epoch": 1.3324791800128124,
-      "grad_norm": 0.6426051464181596,
+      "grad_norm": 0.715579880532086,
       "learning_rate": 5e-06,
-      "loss": 0.6302,
+      "loss": 0.634,
       "step": 520
     },
     {
       "epoch": 1.3581037796284434,
-      "grad_norm": 0.5537518586318295,
+      "grad_norm": 0.539683933602384,
       "learning_rate": 5e-06,
-      "loss": 0.6241,
+      "loss": 0.6273,
       "step": 530
     },
     {
       "epoch": 1.3837283792440744,
-      "grad_norm": 0.5136102716991056,
+      "grad_norm": 0.47743309373776915,
       "learning_rate": 5e-06,
-      "loss": 0.6279,
+      "loss": 0.6312,
       "step": 540
     },
     {
       "epoch": 1.4093529788597055,
-      "grad_norm": 0.5691815192061714,
+      "grad_norm": 0.5510090178389563,
       "learning_rate": 5e-06,
-      "loss": 0.6257,
+      "loss": 0.6291,
       "step": 550
     },
     {
       "epoch": 1.4349775784753362,
-      "grad_norm": 0.5956632063571426,
+      "grad_norm": 0.4885531066853449,
       "learning_rate": 5e-06,
-      "loss": 0.6265,
+      "loss": 0.63,
       "step": 560
     },
     {
       "epoch": 1.4606021780909673,
-      "grad_norm": 0.581023423899474,
+      "grad_norm": 0.4550390667985221,
       "learning_rate": 5e-06,
-      "loss": 0.627,
+      "loss": 0.63,
       "step": 570
     },
     {
       "epoch": 1.4862267777065983,
-      "grad_norm": 0.5178152762952138,
+      "grad_norm": 0.5094454871437174,
       "learning_rate": 5e-06,
-      "loss": 0.6268,
+      "loss": 0.6301,
       "step": 580
     },
     {
       "epoch": 1.5118513773222293,
-      "grad_norm": 0.624224902789754,
+      "grad_norm": 0.5378797481591068,
       "learning_rate": 5e-06,
-      "loss": 0.6329,
+      "loss": 0.6362,
       "step": 590
     },
     {
       "epoch": 1.5374759769378603,
-      "grad_norm": 0.6214150588168393,
+      "grad_norm": 0.4964570729916681,
       "learning_rate": 5e-06,
-      "loss": 0.6244,
+      "loss": 0.6275,
       "step": 600
     },
     {
       "epoch": 1.5631005765534913,
-      "grad_norm": 0.510471747049578,
+      "grad_norm": 0.5150915164078523,
       "learning_rate": 5e-06,
-      "loss": 0.6304,
+      "loss": 0.6336,
       "step": 610
     },
     {
       "epoch": 1.5887251761691223,
-      "grad_norm": 0.511918307352263,
+      "grad_norm": 0.45827307808132584,
       "learning_rate": 5e-06,
-      "loss": 0.6273,
+      "loss": 0.6305,
       "step": 620
     },
     {
       "epoch": 1.6143497757847534,
-      "grad_norm": 0.7352297634775503,
+      "grad_norm": 0.670050138542801,
       "learning_rate": 5e-06,
-      "loss": 0.626,
+      "loss": 0.629,
       "step": 630
     },
     {
       "epoch": 1.6399743754003844,
-      "grad_norm": 0.6567654313011416,
+      "grad_norm": 0.5390177749556742,
       "learning_rate": 5e-06,
-      "loss": 0.6159,
+      "loss": 0.6189,
       "step": 640
     },
     {
       "epoch": 1.6655989750160154,
-      "grad_norm": 0.549652992706915,
+      "grad_norm": 0.47514950562866276,
       "learning_rate": 5e-06,
-      "loss": 0.6305,
+      "loss": 0.6335,
       "step": 650
     },
     {
       "epoch": 1.6912235746316464,
-      "grad_norm": 0.5634051713923773,
+      "grad_norm": 0.45626516699726205,
       "learning_rate": 5e-06,
-      "loss": 0.6269,
+      "loss": 0.63,
       "step": 660
     },
     {
       "epoch": 1.7168481742472774,
-      "grad_norm": 0.529903884524912,
+      "grad_norm": 0.5259666631069414,
       "learning_rate": 5e-06,
-      "loss": 0.6259,
+      "loss": 0.629,
       "step": 670
     },
     {
       "epoch": 1.7424727738629084,
-      "grad_norm": 0.506816776536443,
+      "grad_norm": 0.45958090467852913,
       "learning_rate": 5e-06,
-      "loss": 0.6273,
+      "loss": 0.6304,
       "step": 680
     },
     {
       "epoch": 1.7680973734785392,
-      "grad_norm": 0.6450672440371147,
+      "grad_norm": 0.537603671552474,
       "learning_rate": 5e-06,
-      "loss": 0.617,
+      "loss": 0.6201,
       "step": 690
     },
     {
       "epoch": 1.7937219730941703,
-      "grad_norm": 0.6192873375131733,
+      "grad_norm": 0.4944510355408107,
       "learning_rate": 5e-06,
-      "loss": 0.6196,
+      "loss": 0.6227,
       "step": 700
     },
     {
       "epoch": 1.8193465727098013,
-      "grad_norm": 0.5352789663117377,
+      "grad_norm": 0.4459500383872668,
       "learning_rate": 5e-06,
-      "loss": 0.6262,
+      "loss": 0.6293,
       "step": 710
     },
     {
       "epoch": 1.8449711723254323,
-      "grad_norm": 0.4668080446143915,
+      "grad_norm": 0.41107024989867014,
       "learning_rate": 5e-06,
-      "loss": 0.6294,
+      "loss": 0.6325,
       "step": 720
     },
     {
       "epoch": 1.8705957719410633,
-      "grad_norm": 0.6648243829055217,
+      "grad_norm": 0.5205263432281845,
       "learning_rate": 5e-06,
-      "loss": 0.6231,
+      "loss": 0.626,
       "step": 730
     },
     {
       "epoch": 1.8962203715566943,
-      "grad_norm": 0.5272415641474717,
+      "grad_norm": 0.5178259439056772,
       "learning_rate": 5e-06,
-      "loss": 0.6224,
+      "loss": 0.6253,
       "step": 740
     },
     {
       "epoch": 1.9218449711723253,
-      "grad_norm": 0.5810543386827824,
+      "grad_norm": 0.5186310082942625,
       "learning_rate": 5e-06,
-      "loss": 0.6225,
+      "loss": 0.6254,
       "step": 750
     },
     {
       "epoch": 1.9474695707879563,
-      "grad_norm": 0.6085709674004274,
+      "grad_norm": 0.48818671836894234,
       "learning_rate": 5e-06,
-      "loss": 0.6217,
+      "loss": 0.6244,
       "step": 760
     },
     {
       "epoch": 1.9730941704035874,
-      "grad_norm": 0.5375279101850413,
+      "grad_norm": 0.4816352801983277,
       "learning_rate": 5e-06,
-      "loss": 0.6272,
+      "loss": 0.6301,
       "step": 770
     },
     {
       "epoch": 1.9987187700192184,
-      "grad_norm": 0.6762226804445376,
+      "grad_norm": 0.5573576817650256,
       "learning_rate": 5e-06,
-      "loss": 0.6256,
+      "loss": 0.6284,
       "step": 780
     },
     {
       "epoch": 1.9987187700192184,
-      "eval_loss": 0.6599456071853638,
-      "eval_runtime": 209.9848,
-      "eval_samples_per_second": 50.075,
-      "eval_steps_per_second": 0.395,
+      "eval_loss": 0.6584250926971436,
+      "eval_runtime": 210.7445,
+      "eval_samples_per_second": 49.895,
+      "eval_steps_per_second": 0.394,
       "step": 780
     },
     {
       "epoch": 2.0243433696348494,
-      "grad_norm": 0.6490424719585334,
+      "grad_norm": 0.6596303880050641,
       "learning_rate": 5e-06,
-      "loss": 0.574,
+      "loss": 0.582,
       "step": 790
     },
     {
       "epoch": 2.0499679692504804,
-      "grad_norm": 0.5740794707367172,
+      "grad_norm": 0.5374575961558432,
       "learning_rate": 5e-06,
-      "loss": 0.5772,
+      "loss": 0.5856,
       "step": 800
     },
     {
       "epoch": 2.0755925688661114,
-      "grad_norm": 0.6139160456025744,
+      "grad_norm": 0.565805764328584,
       "learning_rate": 5e-06,
-      "loss": 0.5716,
+      "loss": 0.5799,
       "step": 810
     },
     {
       "epoch": 2.1012171684817424,
-      "grad_norm": 0.5042331647002983,
+      "grad_norm": 0.45938077865068583,
       "learning_rate": 5e-06,
-      "loss": 0.5658,
+      "loss": 0.574,
       "step": 820
     },
     {
       "epoch": 2.1268417680973735,
-      "grad_norm": 0.525240715956572,
+      "grad_norm": 0.5086155425168429,
       "learning_rate": 5e-06,
-      "loss": 0.5766,
+      "loss": 0.5846,
       "step": 830
     },
     {
       "epoch": 2.1524663677130045,
-      "grad_norm": 0.6301792211373436,
+      "grad_norm": 0.5951424838463266,
       "learning_rate": 5e-06,
-      "loss": 0.5727,
+      "loss": 0.5804,
       "step": 840
     },
     {
       "epoch": 2.1780909673286355,
-      "grad_norm": 0.5695524017278328,
+      "grad_norm": 0.5389501938241722,
       "learning_rate": 5e-06,
-      "loss": 0.5653,
+      "loss": 0.5733,
       "step": 850
     },
     {
       "epoch": 2.2037155669442665,
-      "grad_norm": 0.5036859440783635,
+      "grad_norm": 0.49715965924754885,
       "learning_rate": 5e-06,
-      "loss": 0.582,
+      "loss": 0.5901,
       "step": 860
     },
     {
       "epoch": 2.2293401665598975,
-      "grad_norm": 0.6707419650010094,
+      "grad_norm": 0.6416154288767167,
       "learning_rate": 5e-06,
-      "loss": 0.5738,
+      "loss": 0.5814,
       "step": 870
     },
     {
       "epoch": 2.2549647661755285,
-      "grad_norm": 0.643330757681528,
+      "grad_norm": 0.5840426113039779,
       "learning_rate": 5e-06,
-      "loss": 0.5757,
+      "loss": 0.5833,
       "step": 880
     },
     {
       "epoch": 2.2805893657911596,
-      "grad_norm": 0.5671913495072154,
+      "grad_norm": 0.5363577230603178,
       "learning_rate": 5e-06,
-      "loss": 0.5833,
+      "loss": 0.5913,
       "step": 890
     },
     {
       "epoch": 2.3062139654067906,
-      "grad_norm": 0.7335441489473075,
+      "grad_norm": 0.4999642649696085,
       "learning_rate": 5e-06,
-      "loss": 0.5779,
+      "loss": 0.5853,
       "step": 900
     },
     {
       "epoch": 2.3318385650224216,
-      "grad_norm": 0.5959572278401346,
+      "grad_norm": 0.6712704299303816,
       "learning_rate": 5e-06,
-      "loss": 0.5752,
+      "loss": 0.5824,
       "step": 910
     },
     {
       "epoch": 2.3574631646380526,
-      "grad_norm": 0.5395801639714329,
+      "grad_norm": 0.4993129343338316,
       "learning_rate": 5e-06,
-      "loss": 0.5806,
+      "loss": 0.5882,
       "step": 920
     },
     {
       "epoch": 2.3830877642536836,
-      "grad_norm": 0.7361927812478924,
+      "grad_norm": 0.6482429574217665,
       "learning_rate": 5e-06,
-      "loss": 0.5779,
+      "loss": 0.5852,
       "step": 930
     },
     {
       "epoch": 2.4087123638693146,
-      "grad_norm": 0.6995594552919696,
+      "grad_norm": 0.5858234169687565,
       "learning_rate": 5e-06,
-      "loss": 0.5815,
+      "loss": 0.5891,
       "step": 940
     },
     {
       "epoch": 2.4343369634849457,
-      "grad_norm": 0.5261563010125221,
+      "grad_norm": 0.4606498702171825,
       "learning_rate": 5e-06,
-      "loss": 0.5807,
+      "loss": 0.5882,
       "step": 950
     },
     {
       "epoch": 2.4599615631005767,
-      "grad_norm": 0.5407517395465042,
+      "grad_norm": 0.4665245302124569,
       "learning_rate": 5e-06,
-      "loss": 0.581,
+      "loss": 0.5882,
       "step": 960
     },
     {
       "epoch": 2.4855861627162077,
-      "grad_norm": 0.5289063213570133,
+      "grad_norm": 0.4787609348846855,
       "learning_rate": 5e-06,
-      "loss": 0.583,
+      "loss": 0.5904,
       "step": 970
     },
     {
       "epoch": 2.5112107623318387,
-      "grad_norm": 0.5796407083708335,
+      "grad_norm": 0.5112448288708001,
       "learning_rate": 5e-06,
-      "loss": 0.5839,
+      "loss": 0.5915,
       "step": 980
     },
     {
       "epoch": 2.5368353619474697,
-      "grad_norm": 0.598887219069754,
+      "grad_norm": 0.5584692413737392,
       "learning_rate": 5e-06,
-      "loss": 0.5783,
+      "loss": 0.5855,
       "step": 990
     },
     {
       "epoch": 2.5624599615631007,
-      "grad_norm": 0.5930341940197382,
+      "grad_norm": 0.5151021464825077,
       "learning_rate": 5e-06,
-      "loss": 0.5825,
+      "loss": 0.5895,
       "step": 1000
     },
     {
       "epoch": 2.5880845611787313,
-      "grad_norm": 0.5331038782110779,
+      "grad_norm": 0.47435391792664705,
       "learning_rate": 5e-06,
-      "loss": 0.5749,
+      "loss": 0.5818,
       "step": 1010
     },
     {
       "epoch": 2.6137091607943628,
-      "grad_norm": 0.543876112697049,
+      "grad_norm": 0.47062007111208676,
       "learning_rate": 5e-06,
-      "loss": 0.58,
+      "loss": 0.5869,
       "step": 1020
     },
     {
       "epoch": 2.6393337604099933,
-      "grad_norm": 0.6281847381865296,
+      "grad_norm": 0.5030724465847995,
       "learning_rate": 5e-06,
-      "loss": 0.5749,
+      "loss": 0.5817,
       "step": 1030
     },
     {
       "epoch": 2.664958360025625,
-      "grad_norm": 0.5842056088552238,
+      "grad_norm": 0.5696236917490695,
       "learning_rate": 5e-06,
-      "loss": 0.5776,
+      "loss": 0.5846,
       "step": 1040
     },
     {
       "epoch": 2.6905829596412554,
-      "grad_norm": 0.5715121557218589,
+      "grad_norm": 0.5097571703440519,
       "learning_rate": 5e-06,
-      "loss": 0.5769,
+      "loss": 0.5838,
       "step": 1050
     },
     {
       "epoch": 2.716207559256887,
-      "grad_norm": 0.4892217105691407,
+      "grad_norm": 0.46906709023993814,
       "learning_rate": 5e-06,
-      "loss": 0.5779,
+      "loss": 0.5849,
       "step": 1060
     },
     {
       "epoch": 2.7418321588725174,
-      "grad_norm": 0.6043402014925059,
+      "grad_norm": 0.6048108768327223,
       "learning_rate": 5e-06,
-      "loss": 0.5794,
+      "loss": 0.5861,
       "step": 1070
     },
     {
       "epoch": 2.767456758488149,
-      "grad_norm": 0.6212507294354088,
+      "grad_norm": 0.5763148469790699,
       "learning_rate": 5e-06,
-      "loss": 0.5831,
+      "loss": 0.59,
       "step": 1080
     },
     {
       "epoch": 2.7930813581037794,
-      "grad_norm": 0.5086022566714822,
+      "grad_norm": 0.5325120981655176,
       "learning_rate": 5e-06,
-      "loss": 0.5852,
+      "loss": 0.5922,
       "step": 1090
     },
     {
       "epoch": 2.818705957719411,
-      "grad_norm": 0.6072722346149179,
+      "grad_norm": 0.47425803790021404,
       "learning_rate": 5e-06,
-      "loss": 0.5851,
+      "loss": 0.5919,
       "step": 1100
     },
     {
       "epoch": 2.8443305573350415,
-      "grad_norm": 0.5800686633562157,
+      "grad_norm": 0.5025999987290302,
       "learning_rate": 5e-06,
-      "loss": 0.5784,
+      "loss": 0.5851,
       "step": 1110
     },
     {
       "epoch": 2.8699551569506725,
-      "grad_norm": 0.6470106482983882,
+      "grad_norm": 0.5898623437912127,
       "learning_rate": 5e-06,
-      "loss": 0.5775,
+      "loss": 0.584,
       "step": 1120
     },
     {
       "epoch": 2.8955797565663035,
-      "grad_norm": 0.5990632250208906,
+      "grad_norm": 0.49890831764955235,
       "learning_rate": 5e-06,
-      "loss": 0.5779,
+      "loss": 0.5847,
       "step": 1130
     },
     {
       "epoch": 2.9212043561819345,
-      "grad_norm": 0.5694303405669764,
+      "grad_norm": 0.4911458303208662,
       "learning_rate": 5e-06,
-      "loss": 0.5803,
+      "loss": 0.5868,
       "step": 1140
     },
     {
       "epoch": 2.9468289557975655,
-      "grad_norm": 0.5184100107485876,
+      "grad_norm": 0.4862045432861365,
       "learning_rate": 5e-06,
-      "loss": 0.5831,
+      "loss": 0.5896,
       "step": 1150
     },
     {
       "epoch": 2.9724535554131966,
-      "grad_norm": 0.6155448734761894,
+      "grad_norm": 0.574968732277202,
       "learning_rate": 5e-06,
-      "loss": 0.5848,
+      "loss": 0.5915,
       "step": 1160
     },
     {
       "epoch": 2.9980781550288276,
-      "grad_norm": 0.6461874509945199,
+      "grad_norm": 0.536851160723009,
       "learning_rate": 5e-06,
-      "loss": 0.5851,
+      "loss": 0.5918,
       "step": 1170
     },
     {
       "epoch": 2.9980781550288276,
-      "eval_loss": 0.6631707549095154,
-      "eval_runtime": 211.1021,
-      "eval_samples_per_second": 49.81,
-      "eval_steps_per_second": 0.393,
+      "eval_loss": 0.6599797010421753,
+      "eval_runtime": 210.167,
+      "eval_samples_per_second": 50.032,
+      "eval_steps_per_second": 0.395,
       "step": 1170
     },
     {
       "epoch": 2.9980781550288276,
       "step": 1170,
       "total_flos": 1959374817853440.0,
-      "train_loss": 0.6376129688360752,
-      "train_runtime": 34970.0124,
-      "train_samples_per_second": 17.138,
+      "train_loss": 0.6400190381922274,
+      "train_runtime": 35097.8295,
+      "train_samples_per_second": 17.076,
       "train_steps_per_second": 0.033
     }
   ],