Model save

all_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 1.0,
-    "train_loss": 0.4228534948309896,
-    "train_runtime": 5290.3981,
+    "train_loss": 0.4350618153878388,
+    "train_runtime": 5287.1692,
     "train_samples": 56236,
-    "train_samples_per_second": 10.63,
+    "train_samples_per_second": 10.636,
     "train_steps_per_second": 0.083
 }

model-00001-of-00003.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:75d6368d1fdee8107ded739a6b116ec4475e494955ea8e987c3fbe72511666ef
+oid sha256:1aad48cb8aa7b2903f93b8cfd7cb4737c66ec3d00b09189ab8e7eb0e36e7f63d
 size 4943162336

model-00002-of-00003.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9c5b6326f4bbfe2dc14bbb7c67e7fd495feaea85c802948941fd853302b36ac8
+oid sha256:df0b2b647126bc8b703c31af35bd8e2148134aa0e7a73730fa9484c71e1715ba
 size 4999819336

model-00003-of-00003.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6defafbfd4c91d1663c1f8e99d13d070ad43df11a0d33c2b6e84d278c3d22e01
+oid sha256:2c4087aac41dc68561e56069eabd2d1e6f707422d96ba377ee692b1d30a1ec69
 size 4540516344

runs/Jul03_00-12-44_n136-129-074/events.out.tfevents.1719936835.n136-129-074.2571438.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:54a34ee72ccfc24cb05608d7611fdc197f74a06d521c8a496ca268781f6087cd
-size 25561
+oid sha256:ad7ad8b7d03380518cec989fc964e28f0b31a5c1f9d4ae57780dc84d425cb203
+size 27979

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 1.0,
-    "train_loss": 0.4228534948309896,
-    "train_runtime": 5290.3981,
+    "train_loss": 0.4350618153878388,
+    "train_runtime": 5287.1692,
     "train_samples": 56236,
-    "train_samples_per_second": 10.63,
+    "train_samples_per_second": 10.636,
     "train_steps_per_second": 0.083
 }

trainer_state.json

@@ -10,7 +10,7 @@
   "log_history": [
     {
       "epoch": 0.0,
-      "grad_norm": 56.936522397845614,
+      "grad_norm": 22.786040256127208,
       "learning_rate": 2.2727272727272727e-09,
       "logits/chosen": -1.6768856048583984,
       "logits/rejected": -1.7259055376052856,
@@ -25,656 +25,656 @@
     },
     {
       "epoch": 0.02,
-      "grad_norm": 64.38180400614688,
+      "grad_norm": 25.6836964844408,
       "learning_rate": 2.2727272727272725e-08,
-      "logits/chosen": -1.7024906873703003,
-      "logits/rejected": -1.6682149171829224,
-      "logps/chosen": -1.21290922164917,
-      "logps/rejected": -1.2202513217926025,
-      "loss": 0.6934,
-      "rewards/accuracies": 0.4652777910232544,
-      "rewards/chosen": 0.00198164745233953,
-      "rewards/margins": -0.0004947604611515999,
-      "rewards/rejected": 0.00247640791349113,
+      "logits/chosen": -1.7025811672210693,
+      "logits/rejected": -1.6681201457977295,
+      "logps/chosen": -1.2131770849227905,
+      "logps/rejected": -1.2202790975570679,
+      "loss": 0.6931,
+      "rewards/accuracies": 0.4166666567325592,
+      "rewards/chosen": 0.00025717043899931014,
+      "rewards/margins": -0.0006778165698051453,
+      "rewards/rejected": 0.0009349870379082859,
       "step": 10
     },
     {
       "epoch": 0.05,
-      "grad_norm": 64.62406710950098,
+      "grad_norm": 26.405806920127038,
       "learning_rate": 4.545454545454545e-08,
-      "logits/chosen": -1.7796684503555298,
-      "logits/rejected": -1.7348577976226807,
-      "logps/chosen": -1.1446455717086792,
-      "logps/rejected": -1.185205340385437,
-      "loss": 0.6913,
-      "rewards/accuracies": 0.518750011920929,
-      "rewards/chosen": -0.007116294465959072,
-      "rewards/margins": 0.0016124974936246872,
-      "rewards/rejected": -0.008728791028261185,
+      "logits/chosen": -1.7794504165649414,
+      "logits/rejected": -1.734635591506958,
+      "logps/chosen": -1.144683599472046,
+      "logps/rejected": -1.1852140426635742,
+      "loss": 0.6925,
+      "rewards/accuracies": 0.5375000238418579,
+      "rewards/chosen": -0.0029223989695310593,
+      "rewards/margins": 0.0005865652346983552,
+      "rewards/rejected": -0.0035089640878140926,
       "step": 20
     },
     {
       "epoch": 0.07,
-      "grad_norm": 65.81836343668654,
+      "grad_norm": 28.068883928310647,
       "learning_rate": 6.818181818181817e-08,
-      "logits/chosen": -1.7444534301757812,
-      "logits/rejected": -1.675607442855835,
-      "logps/chosen": -1.1950287818908691,
-      "logps/rejected": -1.2475531101226807,
-      "loss": 0.6826,
-      "rewards/accuracies": 0.6187499761581421,
-      "rewards/chosen": -0.06063701584935188,
-      "rewards/margins": 0.0264725498855114,
-      "rewards/rejected": -0.08710956573486328,
+      "logits/chosen": -1.7446922063827515,
+      "logits/rejected": -1.676011085510254,
+      "logps/chosen": -1.1949479579925537,
+      "logps/rejected": -1.2484194040298462,
+      "loss": 0.6882,
+      "rewards/accuracies": 0.6625000238418579,
+      "rewards/chosen": -0.024093259125947952,
+      "rewards/margins": 0.012483139522373676,
+      "rewards/rejected": -0.036576397716999054,
       "step": 30
     },
     {
       "epoch": 0.09,
-      "grad_norm": 48.424632790373714,
+      "grad_norm": 19.593533440172358,
       "learning_rate": 9.09090909090909e-08,
-      "logits/chosen": -1.7310537099838257,
-      "logits/rejected": -1.6650664806365967,
-      "logps/chosen": -1.2446069717407227,
-      "logps/rejected": -1.3240689039230347,
-      "loss": 0.6683,
-      "rewards/accuracies": 0.6312500238418579,
-      "rewards/chosen": -0.18829187750816345,
-      "rewards/margins": 0.1356922686100006,
-      "rewards/rejected": -0.3239841163158417,
+      "logits/chosen": -1.7312889099121094,
+      "logits/rejected": -1.6649024486541748,
+      "logps/chosen": -1.2515684366226196,
+      "logps/rejected": -1.3352997303009033,
+      "loss": 0.6797,
+      "rewards/accuracies": 0.6625000238418579,
+      "rewards/chosen": -0.08923971652984619,
+      "rewards/margins": 0.06281541287899017,
+      "rewards/rejected": -0.15205512940883636,
       "step": 40
     },
     {
       "epoch": 0.11,
-      "grad_norm": 50.59306703821365,
+      "grad_norm": 24.263100694250465,
       "learning_rate": 9.994307990108962e-08,
-      "logits/chosen": -1.695184350013733,
-      "logits/rejected": -1.6308221817016602,
-      "logps/chosen": -1.2791718244552612,
-      "logps/rejected": -1.3329250812530518,
-      "loss": 0.6419,
-      "rewards/accuracies": 0.6812499761581421,
-      "rewards/chosen": -0.35542386770248413,
-      "rewards/margins": 0.1683032065629959,
-      "rewards/rejected": -0.5237270593643188,
+      "logits/chosen": -1.6921930313110352,
+      "logits/rejected": -1.6271032094955444,
+      "logps/chosen": -1.298782467842102,
+      "logps/rejected": -1.3565475940704346,
+      "loss": 0.6637,
+      "rewards/accuracies": 0.7250000238418579,
+      "rewards/chosen": -0.1813907027244568,
+      "rewards/margins": 0.07534531503915787,
+      "rewards/rejected": -0.25673604011535645,
       "step": 50
     },
     {
       "epoch": 0.14,
-      "grad_norm": 70.52284728653416,
+      "grad_norm": 28.402662229181594,
       "learning_rate": 9.959570405988094e-08,
-      "logits/chosen": -1.7311099767684937,
-      "logits/rejected": -1.651442527770996,
-      "logps/chosen": -1.2228403091430664,
-      "logps/rejected": -1.3055191040039062,
-      "loss": 0.6443,
-      "rewards/accuracies": 0.6937500238418579,
-      "rewards/chosen": -0.5687106847763062,
-      "rewards/margins": 0.15375135838985443,
-      "rewards/rejected": -0.7224621176719666,
+      "logits/chosen": -1.7207958698272705,
+      "logits/rejected": -1.6402089595794678,
+      "logps/chosen": -1.2883288860321045,
+      "logps/rejected": -1.378244161605835,
+      "loss": 0.6599,
+      "rewards/accuracies": 0.6499999761581421,
+      "rewards/chosen": -0.35846152901649475,
+      "rewards/margins": 0.07597381621599197,
+      "rewards/rejected": -0.4344353675842285,
       "step": 60
     },
     {
       "epoch": 0.16,
-      "grad_norm": 45.679190031541204,
+      "grad_norm": 21.201003636828116,
       "learning_rate": 9.893476820924666e-08,
-      "logits/chosen": -1.8201316595077515,
-      "logits/rejected": -1.7361564636230469,
-      "logps/chosen": -1.3600715398788452,
-      "logps/rejected": -1.4583489894866943,
-      "loss": 0.6162,
-      "rewards/accuracies": 0.6625000238418579,
-      "rewards/chosen": -0.7361480593681335,
-      "rewards/margins": 0.2580623924732208,
-      "rewards/rejected": -0.994210422039032,
+      "logits/chosen": -1.801526665687561,
+      "logits/rejected": -1.7130231857299805,
+      "logps/chosen": -1.4759318828582764,
+      "logps/rejected": -1.5954515933990479,
+      "loss": 0.6498,
+      "rewards/accuracies": 0.699999988079071,
+      "rewards/chosen": -0.5261799097061157,
+      "rewards/margins": 0.14570936560630798,
+      "rewards/rejected": -0.6718892455101013,
       "step": 70
     },
     {
       "epoch": 0.18,
-      "grad_norm": 54.41920848920781,
+      "grad_norm": 26.86998789465821,
       "learning_rate": 9.796445099843647e-08,
-      "logits/chosen": -1.8112220764160156,
-      "logits/rejected": -1.7294734716415405,
-      "logps/chosen": -1.3612664937973022,
-      "logps/rejected": -1.476828694343567,
-      "loss": 0.6324,
-      "rewards/accuracies": 0.6625000238418579,
-      "rewards/chosen": -0.7714906334877014,
-      "rewards/margins": 0.2751607894897461,
-      "rewards/rejected": -1.0466516017913818,
+      "logits/chosen": -1.7849477529525757,
+      "logits/rejected": -1.6990400552749634,
+      "logps/chosen": -1.5336757898330688,
+      "logps/rejected": -1.6886417865753174,
+      "loss": 0.654,
+      "rewards/accuracies": 0.668749988079071,
+      "rewards/chosen": -0.6534146070480347,
+      "rewards/margins": 0.18887227773666382,
+      "rewards/rejected": -0.8422868847846985,
       "step": 80
     },
     {
       "epoch": 0.2,
-      "grad_norm": 55.97475150728081,
+      "grad_norm": 25.145260129960253,
       "learning_rate": 9.669088708527066e-08,
-      "logits/chosen": -1.7606143951416016,
-      "logits/rejected": -1.6983873844146729,
-      "logps/chosen": -1.4084397554397583,
-      "logps/rejected": -1.47458815574646,
-      "loss": 0.6039,
-      "rewards/accuracies": 0.6812499761581421,
-      "rewards/chosen": -0.8637123107910156,
-      "rewards/margins": 0.29551851749420166,
-      "rewards/rejected": -1.1592309474945068,
+      "logits/chosen": -1.7331278324127197,
+      "logits/rejected": -1.6643707752227783,
+      "logps/chosen": -1.604683518409729,
+      "logps/rejected": -1.6745525598526,
+      "loss": 0.6397,
+      "rewards/accuracies": 0.643750011920929,
+      "rewards/chosen": -0.7379721999168396,
+      "rewards/margins": 0.12564878165721893,
+      "rewards/rejected": -0.8636210560798645,
       "step": 90
     },
     {
       "epoch": 0.23,
-      "grad_norm": 59.68871990537901,
+      "grad_norm": 28.208412906348087,
       "learning_rate": 9.512212835085849e-08,
-      "logits/chosen": -1.7991968393325806,
-      "logits/rejected": -1.7159475088119507,
-      "logps/chosen": -1.3749934434890747,
-      "logps/rejected": -1.490278959274292,
-      "loss": 0.5966,
+      "logits/chosen": -1.770500898361206,
+      "logits/rejected": -1.6813795566558838,
+      "logps/chosen": -1.557369351387024,
+      "logps/rejected": -1.7003123760223389,
+      "loss": 0.623,
       "rewards/accuracies": 0.6937500238418579,
-      "rewards/chosen": -1.0326907634735107,
-      "rewards/margins": 0.36287835240364075,
-      "rewards/rejected": -1.3955689668655396,
+      "rewards/chosen": -0.7778279781341553,
+      "rewards/margins": 0.200466588139534,
+      "rewards/rejected": -0.9782945513725281,
       "step": 100
     },
     {
       "epoch": 0.25,
-      "grad_norm": 51.2753686121636,
+      "grad_norm": 25.857658698377822,
       "learning_rate": 9.326809299301306e-08,
-      "logits/chosen": -1.796499252319336,
-      "logits/rejected": -1.6970577239990234,
-      "logps/chosen": -1.4165159463882446,
-      "logps/rejected": -1.5609779357910156,
-      "loss": 0.5906,
-      "rewards/accuracies": 0.699999988079071,
-      "rewards/chosen": -1.0793695449829102,
-      "rewards/margins": 0.46429458260536194,
-      "rewards/rejected": -1.5436642169952393,
+      "logits/chosen": -1.7706959247589111,
+      "logits/rejected": -1.666680932044983,
+      "logps/chosen": -1.6019655466079712,
+      "logps/rejected": -1.7736393213272095,
+      "loss": 0.6212,
+      "rewards/accuracies": 0.7124999761581421,
+      "rewards/chosen": -0.8026469945907593,
+      "rewards/margins": 0.240141361951828,
+      "rewards/rejected": -1.0427882671356201,
       "step": 110
     },
     {
       "epoch": 0.27,
-      "grad_norm": 55.503246670500936,
+      "grad_norm": 27.113925141570313,
       "learning_rate": 9.114050282021158e-08,
-      "logits/chosen": -1.7899627685546875,
-      "logits/rejected": -1.7314037084579468,
-      "logps/chosen": -1.3753345012664795,
-      "logps/rejected": -1.5070170164108276,
-      "loss": 0.5881,
-      "rewards/accuracies": 0.7749999761581421,
-      "rewards/chosen": -1.1388596296310425,
-      "rewards/margins": 0.4214509427547455,
-      "rewards/rejected": -1.5603104829788208,
+      "logits/chosen": -1.7661209106445312,
+      "logits/rejected": -1.7044270038604736,
+      "logps/chosen": -1.5668163299560547,
+      "logps/rejected": -1.7313486337661743,
+      "loss": 0.6144,
+      "rewards/accuracies": 0.731249988079071,
+      "rewards/chosen": -0.8385075330734253,
+      "rewards/margins": 0.2342797964811325,
+      "rewards/rejected": -1.0727872848510742,
       "step": 120
     },
     {
       "epoch": 0.3,
-      "grad_norm": 53.36720259691284,
+      "grad_norm": 25.192318558632277,
       "learning_rate": 8.875280914254802e-08,
-      "logits/chosen": -1.7889665365219116,
-      "logits/rejected": -1.699845552444458,
-      "logps/chosen": -1.4648027420043945,
-      "logps/rejected": -1.600429892539978,
-      "loss": 0.5731,
-      "rewards/accuracies": 0.7250000238418579,
-      "rewards/chosen": -1.3598629236221313,
-      "rewards/margins": 0.5360117554664612,
-      "rewards/rejected": -1.8958743810653687,
+      "logits/chosen": -1.7651408910751343,
+      "logits/rejected": -1.6746151447296143,
+      "logps/chosen": -1.7618343830108643,
+      "logps/rejected": -1.9648230075836182,
+      "loss": 0.5995,
+      "rewards/accuracies": 0.668749988079071,
+      "rewards/chosen": -1.1380090713500977,
+      "rewards/margins": 0.34912705421447754,
+      "rewards/rejected": -1.4871362447738647,
       "step": 130
     },
     {
       "epoch": 0.32,
-      "grad_norm": 43.48973975700312,
+      "grad_norm": 24.555074276571457,
       "learning_rate": 8.612010772821971e-08,
-      "logits/chosen": -1.8176887035369873,
-      "logits/rejected": -1.7743492126464844,
-      "logps/chosen": -1.4533226490020752,
-      "logps/rejected": -1.5335619449615479,
-      "loss": 0.5749,
-      "rewards/accuracies": 0.75,
-      "rewards/chosen": -1.1972180604934692,
-      "rewards/margins": 0.5257461667060852,
-      "rewards/rejected": -1.7229640483856201,
+      "logits/chosen": -1.7977707386016846,
+      "logits/rejected": -1.7540991306304932,
+      "logps/chosen": -1.8272440433502197,
+      "logps/rejected": -1.9743133783340454,
+      "loss": 0.5993,
+      "rewards/accuracies": 0.737500011920929,
+      "rewards/chosen": -1.2267296314239502,
+      "rewards/margins": 0.34395888447761536,
+      "rewards/rejected": -1.5706886053085327,
       "step": 140
     },
     {
       "epoch": 0.34,
-      "grad_norm": 42.61041473166327,
+      "grad_norm": 29.798327176387467,
       "learning_rate": 8.325904336322055e-08,
-      "logits/chosen": -1.7937091588974,
-      "logits/rejected": -1.7397197484970093,
-      "logps/chosen": -1.4411404132843018,
-      "logps/rejected": -1.5950322151184082,
-      "loss": 0.5797,
-      "rewards/accuracies": 0.71875,
-      "rewards/chosen": -1.3688783645629883,
-      "rewards/margins": 0.5335904359817505,
-      "rewards/rejected": -1.9024689197540283,
+      "logits/chosen": -1.7762537002563477,
+      "logits/rejected": -1.722198247909546,
+      "logps/chosen": -1.9330543279647827,
+      "logps/rejected": -2.1491074562072754,
+      "loss": 0.6088,
+      "rewards/accuracies": 0.6812499761581421,
+      "rewards/chosen": -1.5313791036605835,
+      "rewards/margins": 0.3377588093280792,
+      "rewards/rejected": -1.8691380023956299,
       "step": 150
     },
     {
       "epoch": 0.36,
-      "grad_norm": 43.447112350155855,
+      "grad_norm": 22.379243172505653,
       "learning_rate": 8.01877046176447e-08,
-      "logits/chosen": -1.7396624088287354,
-      "logits/rejected": -1.6674416065216064,
-      "logps/chosen": -1.4853546619415283,
-      "logps/rejected": -1.6183408498764038,
-      "loss": 0.5754,
-      "rewards/accuracies": 0.6625000238418579,
-      "rewards/chosen": -1.5352730751037598,
-      "rewards/margins": 0.40800127387046814,
-      "rewards/rejected": -1.9432742595672607,
+      "logits/chosen": -1.7237749099731445,
+      "logits/rejected": -1.6540390253067017,
+      "logps/chosen": -1.9900233745574951,
+      "logps/rejected": -2.1875462532043457,
+      "loss": 0.5903,
+      "rewards/accuracies": 0.675000011920929,
+      "rewards/chosen": -1.6234468221664429,
+      "rewards/margins": 0.29227370023727417,
+      "rewards/rejected": -1.9157205820083618,
       "step": 160
     },
     {
       "epoch": 0.39,
-      "grad_norm": 42.81233538750994,
+      "grad_norm": 19.935363442551864,
       "learning_rate": 7.692550948392249e-08,
-      "logits/chosen": -1.7920860052108765,
-      "logits/rejected": -1.7322765588760376,
-      "logps/chosen": -1.4762592315673828,
-      "logps/rejected": -1.5927846431732178,
-      "loss": 0.5744,
-      "rewards/accuracies": 0.731249988079071,
-      "rewards/chosen": -1.330018162727356,
-      "rewards/margins": 0.5304046869277954,
-      "rewards/rejected": -1.8604227304458618,
+      "logits/chosen": -1.7721624374389648,
+      "logits/rejected": -1.7196706533432007,
+      "logps/chosen": -1.960934042930603,
+      "logps/rejected": -2.164457321166992,
+      "loss": 0.5921,
+      "rewards/accuracies": 0.706250011920929,
+      "rewards/chosen": -1.501357078552246,
+      "rewards/margins": 0.38615694642066956,
+      "rewards/rejected": -1.887514352798462,
       "step": 170
     },
     {
       "epoch": 0.41,
-      "grad_norm": 54.867303914998864,
+      "grad_norm": 27.99675949527233,
       "learning_rate": 7.349308261002021e-08,
-      "logits/chosen": -1.7482887506484985,
-      "logits/rejected": -1.6929025650024414,
-      "logps/chosen": -1.4677311182022095,
-      "logps/rejected": -1.6114797592163086,
-      "loss": 0.5673,
-      "rewards/accuracies": 0.7437499761581421,
-      "rewards/chosen": -1.3532251119613647,
-      "rewards/margins": 0.47049036622047424,
-      "rewards/rejected": -1.8237155675888062,
+      "logits/chosen": -1.7341270446777344,
+      "logits/rejected": -1.6832844018936157,
+      "logps/chosen": -1.9921613931655884,
+      "logps/rejected": -2.2205045223236084,
+      "loss": 0.5854,
+      "rewards/accuracies": 0.699999988079071,
+      "rewards/chosen": -1.5901505947113037,
+      "rewards/margins": 0.35738545656204224,
+      "rewards/rejected": -1.9475358724594116,
       "step": 180
     },
     {
       "epoch": 0.43,
-      "grad_norm": 47.004742188004336,
+      "grad_norm": 24.529175323127856,
       "learning_rate": 6.991212490377531e-08,
-      "logits/chosen": -1.8027311563491821,
-      "logits/rejected": -1.751875638961792,
-      "logps/chosen": -1.5003658533096313,
-      "logps/rejected": -1.639650583267212,
-      "loss": 0.5419,
-      "rewards/accuracies": 0.731249988079071,
-      "rewards/chosen": -1.3513133525848389,
-      "rewards/margins": 0.588715136051178,
-      "rewards/rejected": -1.940028429031372,
+      "logits/chosen": -1.7890892028808594,
+      "logits/rejected": -1.746469259262085,
+      "logps/chosen": -2.087700843811035,
+      "logps/rejected": -2.3541736602783203,
+      "loss": 0.5608,
+      "rewards/accuracies": 0.75,
+      "rewards/chosen": -1.7151954174041748,
+      "rewards/margins": 0.4898623526096344,
+      "rewards/rejected": -2.2050578594207764,
       "step": 190
     },
     {
       "epoch": 0.46,
-      "grad_norm": 60.076327584157355,
+      "grad_norm": 30.651962115004057,
       "learning_rate": 6.620527633276978e-08,
-      "logits/chosen": -1.742418646812439,
-      "logits/rejected": -1.66952383518219,
-      "logps/chosen": -1.4815881252288818,
-      "logps/rejected": -1.7071993350982666,
-      "loss": 0.5591,
-      "rewards/accuracies": 0.768750011920929,
-      "rewards/chosen": -1.4356744289398193,
-      "rewards/margins": 0.6941120028495789,
-      "rewards/rejected": -2.129786491394043,
+      "logits/chosen": -1.7287298440933228,
+      "logits/rejected": -1.6635551452636719,
+      "logps/chosen": -2.149373769760132,
+      "logps/rejected": -2.5112643241882324,
+      "loss": 0.5803,
+      "rewards/accuracies": 0.71875,
+      "rewards/chosen": -1.909841537475586,
+      "rewards/margins": 0.5502029657363892,
+      "rewards/rejected": -2.4600441455841064,
       "step": 200
     },
     {
       "epoch": 0.48,
-      "grad_norm": 47.83824182965551,
+      "grad_norm": 27.408811645376396,
       "learning_rate": 6.239597278716581e-08,
-      "logits/chosen": -1.8077137470245361,
-      "logits/rejected": -1.7449321746826172,
-      "logps/chosen": -1.5601786375045776,
-      "logps/rejected": -1.6718772649765015,
-      "loss": 0.5475,
-      "rewards/accuracies": 0.75,
-      "rewards/chosen": -1.5742429494857788,
-      "rewards/margins": 0.6618110537528992,
-      "rewards/rejected": -2.236053943634033,
+      "logits/chosen": -1.7830358743667603,
+      "logits/rejected": -1.7272449731826782,
+      "logps/chosen": -2.303990125656128,
+      "logps/rejected": -2.549621105194092,
+      "loss": 0.5627,
+      "rewards/accuracies": 0.7250000238418579,
+      "rewards/chosen": -2.1173198223114014,
+      "rewards/margins": 0.5325894355773926,
+      "rewards/rejected": -2.649909496307373,
       "step": 210
     },
     {
       "epoch": 0.5,
-      "grad_norm": 42.258080587167875,
+      "grad_norm": 25.141892376540234,
       "learning_rate": 5.8508297910462456e-08,
-      "logits/chosen": -1.7566890716552734,
-      "logits/rejected": -1.6750905513763428,
-      "logps/chosen": -1.4759745597839355,
-      "logps/rejected": -1.6640676259994507,
-      "loss": 0.5499,
-      "rewards/accuracies": 0.7437499761581421,
-      "rewards/chosen": -1.562295913696289,
-      "rewards/margins": 0.6515985727310181,
-      "rewards/rejected": -2.2138943672180176,
+      "logits/chosen": -1.729365587234497,
+      "logits/rejected": -1.6552003622055054,
+      "logps/chosen": -2.227386713027954,
+      "logps/rejected": -2.558889627456665,
+      "loss": 0.5666,
+      "rewards/accuracies": 0.6937500238418579,
+      "rewards/chosen": -2.1277427673339844,
+      "rewards/margins": 0.5474593043327332,
+      "rewards/rejected": -2.675201892852783,
       "step": 220
     },
     {
       "epoch": 0.52,
-      "grad_norm": 52.34767465528489,
+      "grad_norm": 25.203206141016988,
       "learning_rate": 5.456683083494731e-08,
-      "logits/chosen": -1.7469978332519531,
-      "logits/rejected": -1.708335518836975,
-      "logps/chosen": -1.4775021076202393,
-      "logps/rejected": -1.5822871923446655,
-      "loss": 0.5699,
-      "rewards/accuracies": 0.625,
-      "rewards/chosen": -1.4184671640396118,
-      "rewards/margins": 0.40521565079689026,
-      "rewards/rejected": -1.8236827850341797,
+      "logits/chosen": -1.7191989421844482,
+      "logits/rejected": -1.6832917928695679,
+      "logps/chosen": -2.193159580230713,
+      "logps/rejected": -2.4023752212524414,
+      "loss": 0.5847,
+      "rewards/accuracies": 0.65625,
+      "rewards/chosen": -1.9987010955810547,
+      "rewards/margins": 0.3709481358528137,
+      "rewards/rejected": -2.3696494102478027,
       "step": 230
     },
     {
       "epoch": 0.55,
-      "grad_norm": 61.39373208172248,
+      "grad_norm": 31.30257665339357,
       "learning_rate": 5.059649078450834e-08,
-      "logits/chosen": -1.73995840549469,
-      "logits/rejected": -1.692831039428711,
-      "logps/chosen": -1.4603643417358398,
-      "logps/rejected": -1.6224209070205688,
-      "loss": 0.5397,
-      "rewards/accuracies": 0.6875,
-      "rewards/chosen": -1.4600539207458496,
-      "rewards/margins": 0.5000897645950317,
-      "rewards/rejected": -1.9601436853408813,
+      "logits/chosen": -1.7015504837036133,
+      "logits/rejected": -1.662376046180725,
+      "logps/chosen": -2.2371976375579834,
+      "logps/rejected": -2.5139238834381104,
+      "loss": 0.5533,
+      "rewards/accuracies": 0.6625000238418579,
+      "rewards/chosen": -2.137688398361206,
+      "rewards/margins": 0.4293746054172516,
+      "rewards/rejected": -2.567063570022583,
       "step": 240
     },
     {
       "epoch": 0.57,
-      "grad_norm": 46.81475246774513,
+      "grad_norm": 24.832978144124045,
       "learning_rate": 4.6622379527277186e-08,
-      "logits/chosen": -1.7453975677490234,
-      "logits/rejected": -1.6935020685195923,
-      "logps/chosen": -1.460304617881775,
-      "logps/rejected": -1.592510461807251,
-      "loss": 0.5459,
-      "rewards/accuracies": 0.6937500238418579,
-      "rewards/chosen": -1.6070802211761475,
-      "rewards/margins": 0.49562931060791016,
-      "rewards/rejected": -2.1027095317840576,
+      "logits/chosen": -1.7127622365951538,
+      "logits/rejected": -1.6652097702026367,
+      "logps/chosen": -2.328359603881836,
+      "logps/rejected": -2.558023452758789,
+      "loss": 0.5645,
+      "rewards/accuracies": 0.6187499761581421,
+      "rewards/chosen": -2.378941535949707,
+      "rewards/margins": 0.3931688666343689,
+      "rewards/rejected": -2.7721104621887207,
       "step": 250
     },
     {
       "epoch": 0.59,
-      "grad_norm": 45.05028544056684,
+      "grad_norm": 28.702758237373775,
       "learning_rate": 4.26696226741691e-08,
-      "logits/chosen": -1.760898232460022,
-      "logits/rejected": -1.6884838342666626,
-      "logps/chosen": -1.5728627443313599,
-      "logps/rejected": -1.7121139764785767,
-      "loss": 0.5508,
-      "rewards/accuracies": 0.6625000238418579,
-      "rewards/chosen": -1.7001960277557373,
-      "rewards/margins": 0.5964212417602539,
-      "rewards/rejected": -2.2966175079345703,
+      "logits/chosen": -1.725927710533142,
+      "logits/rejected": -1.6632360219955444,
+      "logps/chosen": -2.4254117012023926,
+      "logps/rejected": -2.7066590785980225,
+      "loss": 0.5648,
+      "rewards/accuracies": 0.65625,
+      "rewards/chosen": -2.385176181793213,
+      "rewards/margins": 0.522560715675354,
+      "rewards/rejected": -2.9077370166778564,
       "step": 260
     },
     {
       "epoch": 0.61,
-      "grad_norm": 45.62774521822856,
+      "grad_norm": 24.57643832571065,
       "learning_rate": 3.876321082668098e-08,
-      "logits/chosen": -1.8219407796859741,
-      "logits/rejected": -1.7494175434112549,
-      "logps/chosen": -1.5225193500518799,
-      "logps/rejected": -1.6650638580322266,
-      "loss": 0.5391,
-      "rewards/accuracies": 0.75,
-      "rewards/chosen": -1.4201425313949585,
-      "rewards/margins": 0.6661779284477234,
-      "rewards/rejected": -2.086320638656616,
+      "logits/chosen": -1.7822297811508179,
+      "logits/rejected": -1.7207348346710205,
+      "logps/chosen": -2.3284859657287598,
+      "logps/rejected": -2.6171371936798096,
+      "loss": 0.5581,
+      "rewards/accuracies": 0.7250000238418579,
+      "rewards/chosen": -2.179990530014038,
+      "rewards/margins": 0.5586847066879272,
+      "rewards/rejected": -2.7386748790740967,
       "step": 270
     },
     {
       "epoch": 0.64,
-      "grad_norm": 49.38048288534598,
+      "grad_norm": 26.182300162855046,
       "learning_rate": 3.492784157826244e-08,
-      "logits/chosen": -1.7487083673477173,
-      "logits/rejected": -1.6498956680297852,
-      "logps/chosen": -1.5093038082122803,
-      "logps/rejected": -1.6828804016113281,
-      "loss": 0.5459,
-      "rewards/accuracies": 0.71875,
-      "rewards/chosen": -1.4321379661560059,
-      "rewards/margins": 0.6909336447715759,
-      "rewards/rejected": -2.1230716705322266,
+      "logits/chosen": -1.721687912940979,
+      "logits/rejected": -1.6329882144927979,
+      "logps/chosen": -2.33394193649292,
+      "logps/rejected": -2.6748275756835938,
+      "loss": 0.5564,
+      "rewards/accuracies": 0.7124999761581421,
+      "rewards/chosen": -2.2221310138702393,
+      "rewards/margins": 0.610991895198822,
+      "rewards/rejected": -2.833122730255127,
       "step": 280
     },
     {
       "epoch": 0.66,
-      "grad_norm": 39.80249587523834,
+      "grad_norm": 22.815010055443274,
       "learning_rate": 3.118776336817812e-08,
-      "logits/chosen": -1.7935447692871094,
-      "logits/rejected": -1.7284294366836548,
-      "logps/chosen": -1.4926804304122925,
-      "logps/rejected": -1.641869306564331,
-      "loss": 0.5366,
+      "logits/chosen": -1.755319595336914,
+      "logits/rejected": -1.7038185596466064,
+      "logps/chosen": -2.277468204498291,
+      "logps/rejected": -2.616783618927002,
+      "loss": 0.5473,
       "rewards/accuracies": 0.731249988079071,
-      "rewards/chosen": -1.4471591711044312,
-      "rewards/margins": 0.7098628282546997,
-      "rewards/rejected": -2.157021999359131,
+      "rewards/chosen": -2.1484389305114746,
+      "rewards/margins": 0.6641989946365356,
+      "rewards/rejected": -2.8126378059387207,
       "step": 290
     },
     {
       "epoch": 0.68,
-      "grad_norm": 45.964008321418135,
+      "grad_norm": 23.65537068910306,
       "learning_rate": 2.7566622175067443e-08,
-      "logits/chosen": -1.776922583580017,
-      "logits/rejected": -1.7095201015472412,
-      "logps/chosen": -1.5357799530029297,
-      "logps/rejected": -1.7406812906265259,
-      "loss": 0.5452,
-      "rewards/accuracies": 0.706250011920929,
-      "rewards/chosen": -1.6359821557998657,
-      "rewards/margins": 0.6345733404159546,
-      "rewards/rejected": -2.2705554962158203,
+      "logits/chosen": -1.7356256246566772,
+      "logits/rejected": -1.6781995296478271,
+      "logps/chosen": -2.3750412464141846,
+      "logps/rejected": -2.7629873752593994,
+      "loss": 0.5497,
+      "rewards/accuracies": 0.7562500238418579,
+      "rewards/chosen": -2.3329153060913086,
+      "rewards/margins": 0.6199190616607666,
+      "rewards/rejected": -2.9528346061706543,
       "step": 300
     },
     {
       "epoch": 0.71,
-      "grad_norm": 57.60996565680647,
+      "grad_norm": 27.623729083981935,
       "learning_rate": 2.408731201945432e-08,
-      "logits/chosen": -1.7720359563827515,
-      "logits/rejected": -1.7233737707138062,
-      "logps/chosen": -1.4881327152252197,
-      "logps/rejected": -1.6220896244049072,
-      "loss": 0.5332,
-      "rewards/accuracies": 0.7562500238418579,
-      "rewards/chosen": -1.4711856842041016,
-      "rewards/margins": 0.5690120458602905,
-      "rewards/rejected": -2.0401976108551025,
+      "logits/chosen": -1.7345046997070312,
+      "logits/rejected": -1.6898084878921509,
+      "logps/chosen": -2.304429531097412,
+      "logps/rejected": -2.555560827255249,
+      "loss": 0.5555,
+      "rewards/accuracies": 0.675000011920929,
+      "rewards/chosen": -2.2210681438446045,
+      "rewards/margins": 0.46195346117019653,
+      "rewards/rejected": -2.6830217838287354,
       "step": 310
     },
     {
       "epoch": 0.73,
-      "grad_norm": 47.81844840381173,
+      "grad_norm": 26.649151417142534,
       "learning_rate": 2.0771830220378112e-08,
-      "logits/chosen": -1.7288305759429932,
-      "logits/rejected": -1.6700308322906494,
-      "logps/chosen": -1.5263229608535767,
-      "logps/rejected": -1.6329742670059204,
-      "loss": 0.5383,
-      "rewards/accuracies": 0.7124999761581421,
-      "rewards/chosen": -1.5171924829483032,
-      "rewards/margins": 0.5725088715553284,
-      "rewards/rejected": -2.0897011756896973,
+      "logits/chosen": -1.68793523311615,
+      "logits/rejected": -1.6301701068878174,
+      "logps/chosen": -2.3206005096435547,
+      "logps/rejected": -2.559023380279541,
+      "loss": 0.5553,
+      "rewards/accuracies": 0.625,
+      "rewards/chosen": -2.195432424545288,
+      "rewards/margins": 0.4925463795661926,
+      "rewards/rejected": -2.687978744506836,
       "step": 320
     },
     {
       "epoch": 0.75,
-      "grad_norm": 42.404801937182704,
+      "grad_norm": 24.875498436093228,
       "learning_rate": 1.7641138321260257e-08,
-      "logits/chosen": -1.7647384405136108,
-      "logits/rejected": -1.6919339895248413,
-      "logps/chosen": -1.4824151992797852,
-      "logps/rejected": -1.6646735668182373,
-      "loss": 0.5341,
-      "rewards/accuracies": 0.824999988079071,
-      "rewards/chosen": -1.4928021430969238,
-      "rewards/margins": 0.7736164927482605,
-      "rewards/rejected": -2.266418695449829,
+      "logits/chosen": -1.722652792930603,
+      "logits/rejected": -1.6583467721939087,
+      "logps/chosen": -2.2422232627868652,
+      "logps/rejected": -2.6411871910095215,
+      "loss": 0.5479,
+      "rewards/accuracies": 0.831250011920929,
+      "rewards/chosen": -2.116737127304077,
+      "rewards/margins": 0.7428576946258545,
+      "rewards/rejected": -2.8595950603485107,
       "step": 330
     },
     {
       "epoch": 0.77,
-      "grad_norm": 41.58228712777898,
+      "grad_norm": 26.307918213457313,
       "learning_rate": 1.4715029564277793e-08,
-      "logits/chosen": -1.833791971206665,
-      "logits/rejected": -1.7838408946990967,
-      "logps/chosen": -1.4812287092208862,
-      "logps/rejected": -1.657918930053711,
-      "loss": 0.5469,
+      "logits/chosen": -1.7855126857757568,
+      "logits/rejected": -1.7428696155548096,
+      "logps/chosen": -2.2222747802734375,
+      "logps/rejected": -2.578981399536133,
+      "loss": 0.5581,
       "rewards/accuracies": 0.75,
-      "rewards/chosen": -1.3828407526016235,
-      "rewards/margins": 0.6981338262557983,
-      "rewards/rejected": -2.080974578857422,
+      "rewards/chosen": -2.0352282524108887,
+      "rewards/margins": 0.6392861604690552,
+      "rewards/rejected": -2.6745142936706543,
       "step": 340
     },
     {
       "epoch": 0.8,
-      "grad_norm": 50.79187416231153,
+      "grad_norm": 25.341211676404754,
       "learning_rate": 1.2012003751113343e-08,
-      "logits/chosen": -1.8199939727783203,
-      "logits/rejected": -1.7557872533798218,
-      "logps/chosen": -1.4952068328857422,
-      "logps/rejected": -1.6920936107635498,
-      "loss": 0.5205,
-      "rewards/accuracies": 0.75,
-      "rewards/chosen": -1.6651843786239624,
-      "rewards/margins": 0.7334816455841064,
-      "rewards/rejected": -2.3986661434173584,
+      "logits/chosen": -1.7763115167617798,
+      "logits/rejected": -1.7206395864486694,
+      "logps/chosen": -2.328850269317627,
+      "logps/rejected": -2.70664119720459,
+      "loss": 0.5366,
+      "rewards/accuracies": 0.71875,
+      "rewards/chosen": -2.3333606719970703,
+      "rewards/margins": 0.6552005410194397,
+      "rewards/rejected": -2.9885611534118652,
       "step": 350
     },
     {
       "epoch": 0.82,
-      "grad_norm": 46.03762157927165,
+      "grad_norm": 31.147010925293188,
       "learning_rate": 9.549150281252633e-09,
-      "logits/chosen": -1.7812334299087524,
-      "logits/rejected": -1.7282968759536743,
-      "logps/chosen": -1.479620337486267,
-      "logps/rejected": -1.663819670677185,
-      "loss": 0.5227,
+      "logits/chosen": -1.725211501121521,
+      "logits/rejected": -1.6820595264434814,
+      "logps/chosen": -2.2657017707824707,
+      "logps/rejected": -2.6082816123962402,
+      "loss": 0.5395,
       "rewards/accuracies": 0.6937500238418579,
-      "rewards/chosen": -1.5374261140823364,
-      "rewards/margins": 0.6623435020446777,
-      "rewards/rejected": -2.1997694969177246,
+      "rewards/chosen": -2.1871337890625,
+      "rewards/margins": 0.5816982388496399,
+      "rewards/rejected": -2.768831729888916,
       "step": 360
     },
     {
       "epoch": 0.84,
-      "grad_norm": 42.572961776589516,
+      "grad_norm": 24.868596217446985,
       "learning_rate": 7.3420401072985306e-09,
-      "logits/chosen": -1.8132097721099854,
-      "logits/rejected": -1.7593128681182861,
-      "logps/chosen": -1.5283488035202026,
-      "logps/rejected": -1.730432152748108,
-      "loss": 0.5242,
-      "rewards/accuracies": 0.75,
-      "rewards/chosen": -1.5686765909194946,
-      "rewards/margins": 0.7230969667434692,
-      "rewards/rejected": -2.291773557662964,
+      "logits/chosen": -1.7670882940292358,
+      "logits/rejected": -1.7162415981292725,
+      "logps/chosen": -2.3288838863372803,
+      "logps/rejected": -2.694032669067383,
+      "loss": 0.5377,
+      "rewards/accuracies": 0.6812499761581421,
+      "rewards/chosen": -2.2285408973693848,
+      "rewards/margins": 0.6153696179389954,
+      "rewards/rejected": -2.8439104557037354,
       "step": 370
     },
     {
       "epoch": 0.86,
-      "grad_norm": 52.728611883061056,
+      "grad_norm": 28.264277931849893,
       "learning_rate": 5.404627290395369e-09,
-      "logits/chosen": -1.7789500951766968,
-      "logits/rejected": -1.7105478048324585,
-      "logps/chosen": -1.4771665334701538,
-      "logps/rejected": -1.659226417541504,
-      "loss": 0.5267,
-      "rewards/accuracies": 0.7562500238418579,
-      "rewards/chosen": -1.437803864479065,
-      "rewards/margins": 0.710318386554718,
-      "rewards/rejected": -2.1481223106384277,
+      "logits/chosen": -1.7286021709442139,
+      "logits/rejected": -1.6685943603515625,
+      "logps/chosen": -2.2551302909851074,
+      "logps/rejected": -2.6050727367401123,
+      "loss": 0.537,
+      "rewards/accuracies": 0.737500011920929,
+      "rewards/chosen": -2.1310482025146484,
+      "rewards/margins": 0.619892954826355,
+      "rewards/rejected": -2.750941514968872,
       "step": 380
     },
     {
       "epoch": 0.89,
-      "grad_norm": 69.05994145197278,
+      "grad_norm": 32.571402526944446,
       "learning_rate": 3.74916077816162e-09,
-      "logits/chosen": -1.7877365350723267,
-      "logits/rejected": -1.7246090173721313,
-      "logps/chosen": -1.4689042568206787,
-      "logps/rejected": -1.6005665063858032,
-      "loss": 0.5424,
-      "rewards/accuracies": 0.71875,
-      "rewards/chosen": -1.5955479145050049,
-      "rewards/margins": 0.5776661038398743,
-      "rewards/rejected": -2.1732141971588135,
+      "logits/chosen": -1.734569787979126,
+      "logits/rejected": -1.678733468055725,
+      "logps/chosen": -2.2776572704315186,
+      "logps/rejected": -2.573931932449341,
+      "loss": 0.5505,
+      "rewards/accuracies": 0.6812499761581421,
+      "rewards/chosen": -2.255725145339966,
+      "rewards/margins": 0.5602914690971375,
+      "rewards/rejected": -2.816016674041748,
       "step": 390
     },
     {
       "epoch": 0.91,
-      "grad_norm": 47.34880971893926,
+      "grad_norm": 30.76786815731744,
       "learning_rate": 2.386106962899165e-09,
-      "logits/chosen": -1.7117360830307007,
-      "logits/rejected": -1.6371567249298096,
-      "logps/chosen": -1.5578153133392334,
-      "logps/rejected": -1.7342666387557983,
-      "loss": 0.5379,
-      "rewards/accuracies": 0.7124999761581421,
-      "rewards/chosen": -1.6569716930389404,
-      "rewards/margins": 0.6005586981773376,
-      "rewards/rejected": -2.257530450820923,
+      "logits/chosen": -1.6618547439575195,
+      "logits/rejected": -1.599841594696045,
+      "logps/chosen": -2.417301654815674,
+      "logps/rejected": -2.7463181018829346,
+      "loss": 0.5418,
+      "rewards/accuracies": 0.6937500238418579,
+      "rewards/chosen": -2.381761312484741,
+      "rewards/margins": 0.5453529357910156,
+      "rewards/rejected": -2.927114725112915,
       "step": 400
     },
     {
       "epoch": 0.93,
-      "grad_norm": 50.11568741978605,
+      "grad_norm": 31.25431184934112,
       "learning_rate": 1.3240835096913706e-09,
-      "logits/chosen": -1.746082067489624,
-      "logits/rejected": -1.6423866748809814,
-      "logps/chosen": -1.4619499444961548,
-      "logps/rejected": -1.6878178119659424,
-      "loss": 0.5428,
-      "rewards/accuracies": 0.75,
-      "rewards/chosen": -1.451996088027954,
-      "rewards/margins": 0.7822669744491577,
-      "rewards/rejected": -2.2342631816864014,
+      "logits/chosen": -1.6889197826385498,
+      "logits/rejected": -1.5975219011306763,
+      "logps/chosen": -2.266620397567749,
+      "logps/rejected": -2.7067856788635254,
+      "loss": 0.5607,
+      "rewards/accuracies": 0.7124999761581421,
+      "rewards/chosen": -2.190139055252075,
+      "rewards/margins": 0.7415019273757935,
+      "rewards/rejected": -2.931641101837158,
       "step": 410
     },
     {
       "epoch": 0.96,
-      "grad_norm": 102.373067174958,
+      "grad_norm": 23.39234165392022,
       "learning_rate": 5.698048727497462e-10,
-      "logits/chosen": -1.769753098487854,
-      "logits/rejected": -1.696448564529419,
-      "logps/chosen": -1.5222785472869873,
-      "logps/rejected": -1.7370811700820923,
-      "loss": 0.5332,
-      "rewards/accuracies": 0.768750011920929,
-      "rewards/chosen": -1.6549043655395508,
-      "rewards/margins": 0.7392269968986511,
-      "rewards/rejected": -2.3941311836242676,
+      "logits/chosen": -1.7234703302383423,
+      "logits/rejected": -1.65776789188385,
+      "logps/chosen": -2.363593578338623,
+      "logps/rejected": -2.7682502269744873,
+      "loss": 0.5431,
+      "rewards/accuracies": 0.7250000238418579,
+      "rewards/chosen": -2.3445918560028076,
+      "rewards/margins": 0.6753993034362793,
+      "rewards/rejected": -3.019990921020508,
       "step": 420
     },
     {
       "epoch": 0.98,
-      "grad_norm": 44.58592035367784,
+      "grad_norm": 24.387678586587597,
       "learning_rate": 1.2803984447259387e-10,
-      "logits/chosen": -1.781581163406372,
-      "logits/rejected": -1.724043846130371,
-      "logps/chosen": -1.5100823640823364,
-      "logps/rejected": -1.7025842666625977,
-      "loss": 0.5259,
-      "rewards/accuracies": 0.737500011920929,
-      "rewards/chosen": -1.638715386390686,
-      "rewards/margins": 0.7242453694343567,
-      "rewards/rejected": -2.3629605770111084,
+      "logits/chosen": -1.7296276092529297,
+      "logits/rejected": -1.68634033203125,
+      "logps/chosen": -2.347764730453491,
+      "logps/rejected": -2.731468677520752,
+      "loss": 0.5333,
+      "rewards/accuracies": 0.7250000238418579,
+      "rewards/chosen": -2.330850601196289,
+      "rewards/margins": 0.6721019744873047,
+      "rewards/rejected": -3.002953052520752,
       "step": 430
     },
     {
       "epoch": 1.0,
       "step": 439,
       "total_flos": 0.0,
-      "train_loss": 0.4228534948309896,
-      "train_runtime": 5290.3981,
-      "train_samples_per_second": 10.63,
+      "train_loss": 0.4350618153878388,
+      "train_runtime": 5287.1692,
+      "train_samples_per_second": 10.636,
       "train_steps_per_second": 0.083
     }
   ],

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9a53011c0d49d2ed2a349f4e82d07898fc20dfe797eca2ac6cb079ef749dab9f
+oid sha256:0716202a831bf05f44c63f92c313ac4b7bae8886fe6e5852b8e205d38a7569ae
 size 6328