End of training

README.md

@@ -4,6 +4,7 @@ license: llama3.1
 base_model: meta-llama/Meta-Llama-3.1-8B
 tags:
 - llama-factory
+- full
 - generated_from_trainer
 model-index:
 - name: oh-dcft-v3.1-gpt-4o-mini
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # oh-dcft-v3.1-gpt-4o-mini
 
-This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B) on an unknown dataset.
+This model is a fine-tuned version of [meta-llama/Meta-Llama-3.1-8B](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B) on the mlfoundations-dev/oh-dcft-v3.1-gpt-4o-mini dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.6408
 

all_results.json

@@ -1,12 +1,12 @@
 {
-    "epoch": 2.9946777054997042,
-    "eval_loss": 0.6413019895553589,
-    "eval_runtime": 226.4353,
-    "eval_samples_per_second": 50.297,
-    "eval_steps_per_second": 0.393,
-    "total_flos": 2120178393415680.0,
-    "train_loss": 0.6201754979801027,
-    "train_runtime": 38082.1567,
-    "train_samples_per_second": 17.045,
-    "train_steps_per_second": 0.033
+    "epoch": 2.9961549837326236,
+    "eval_loss": 0.6407967805862427,
+    "eval_runtime": 283.7814,
+    "eval_samples_per_second": 40.133,
+    "eval_steps_per_second": 0.627,
+    "total_flos": 2120387773071360.0,
+    "train_loss": 0.6192890667425701,
+    "train_runtime": 41807.2835,
+    "train_samples_per_second": 15.526,
+    "train_steps_per_second": 0.03
 }

eval_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 2.9946777054997042,
-    "eval_loss": 0.6413019895553589,
-    "eval_runtime": 226.4353,
-    "eval_samples_per_second": 50.297,
-    "eval_steps_per_second": 0.393
+    "epoch": 2.9961549837326236,
+    "eval_loss": 0.6407967805862427,
+    "eval_runtime": 283.7814,
+    "eval_samples_per_second": 40.133,
+    "eval_steps_per_second": 0.627
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 2.9946777054997042,
-    "total_flos": 2120178393415680.0,
-    "train_loss": 0.6201754979801027,
-    "train_runtime": 38082.1567,
-    "train_samples_per_second": 17.045,
-    "train_steps_per_second": 0.033
+    "epoch": 2.9961549837326236,
+    "total_flos": 2120387773071360.0,
+    "train_loss": 0.6192890667425701,
+    "train_runtime": 41807.2835,
+    "train_samples_per_second": 15.526,
+    "train_steps_per_second": 0.03
 }

trainer_state.json

@@ -1,7 +1,7 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.9946777054997042,
+  "epoch": 2.9961549837326236,
   "eval_steps": 500,
   "global_step": 1266,
   "is_hyper_param_search": false,
@@ -9,919 +9,919 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.02365464222353637,
-      "grad_norm": 1.5665423397587683,
+      "epoch": 0.023661638568470866,
+      "grad_norm": 5.716220186648855,
       "learning_rate": 5e-06,
-      "loss": 0.8891,
+      "loss": 0.8897,
       "step": 10
     },
     {
-      "epoch": 0.04730928444707274,
-      "grad_norm": 1.0589767378232773,
+      "epoch": 0.04732327713694173,
+      "grad_norm": 0.8490804690852378,
       "learning_rate": 5e-06,
-      "loss": 0.7949,
+      "loss": 0.7877,
       "step": 20
     },
     {
-      "epoch": 0.0709639266706091,
-      "grad_norm": 0.8838983419870171,
+      "epoch": 0.0709849157054126,
+      "grad_norm": 0.8359497191663584,
       "learning_rate": 5e-06,
-      "loss": 0.7599,
+      "loss": 0.7533,
       "step": 30
     },
     {
-      "epoch": 0.09461856889414548,
-      "grad_norm": 0.9501620192903069,
+      "epoch": 0.09464655427388347,
+      "grad_norm": 2.872573622069312,
       "learning_rate": 5e-06,
-      "loss": 0.7391,
+      "loss": 0.7354,
       "step": 40
     },
     {
-      "epoch": 0.11827321111768184,
-      "grad_norm": 0.7634636002964827,
+      "epoch": 0.11830819284235433,
+      "grad_norm": 1.2235372630281218,
       "learning_rate": 5e-06,
-      "loss": 0.7272,
+      "loss": 0.726,
       "step": 50
     },
     {
-      "epoch": 0.1419278533412182,
-      "grad_norm": 0.6753406016779676,
+      "epoch": 0.1419698314108252,
+      "grad_norm": 0.7914495021160732,
       "learning_rate": 5e-06,
-      "loss": 0.7159,
+      "loss": 0.7152,
       "step": 60
     },
     {
-      "epoch": 0.16558249556475457,
-      "grad_norm": 0.6614325908691252,
+      "epoch": 0.16563146997929606,
+      "grad_norm": 0.6440056541571992,
       "learning_rate": 5e-06,
-      "loss": 0.7118,
+      "loss": 0.7101,
       "step": 70
     },
     {
-      "epoch": 0.18923713778829096,
-      "grad_norm": 0.7719184686061289,
+      "epoch": 0.18929310854776693,
+      "grad_norm": 1.2255311186058144,
       "learning_rate": 5e-06,
-      "loss": 0.7019,
+      "loss": 0.7007,
       "step": 80
     },
     {
-      "epoch": 0.21289178001182732,
-      "grad_norm": 0.7524231413666251,
+      "epoch": 0.2129547471162378,
+      "grad_norm": 0.5863287056241707,
       "learning_rate": 5e-06,
-      "loss": 0.6872,
+      "loss": 0.6856,
       "step": 90
     },
     {
-      "epoch": 0.23654642223536368,
-      "grad_norm": 0.8110229180711027,
+      "epoch": 0.23661638568470866,
+      "grad_norm": 0.6673942726540693,
       "learning_rate": 5e-06,
-      "loss": 0.6853,
+      "loss": 0.685,
       "step": 100
     },
     {
-      "epoch": 0.26020106445890007,
-      "grad_norm": 0.4971482406443161,
+      "epoch": 0.26027802425317953,
+      "grad_norm": 0.6052355221282993,
       "learning_rate": 5e-06,
-      "loss": 0.6909,
+      "loss": 0.6898,
       "step": 110
     },
     {
-      "epoch": 0.2838557066824364,
-      "grad_norm": 0.9464051864732146,
+      "epoch": 0.2839396628216504,
+      "grad_norm": 0.6230907972132269,
       "learning_rate": 5e-06,
-      "loss": 0.6782,
+      "loss": 0.6772,
       "step": 120
     },
     {
-      "epoch": 0.3075103489059728,
-      "grad_norm": 0.5071907565316945,
+      "epoch": 0.30760130139012126,
+      "grad_norm": 0.5418727176057381,
       "learning_rate": 5e-06,
-      "loss": 0.6744,
+      "loss": 0.6733,
       "step": 130
     },
     {
-      "epoch": 0.33116499112950915,
-      "grad_norm": 0.9224817650868851,
+      "epoch": 0.33126293995859213,
+      "grad_norm": 0.6020619371732497,
       "learning_rate": 5e-06,
-      "loss": 0.6828,
+      "loss": 0.6819,
       "step": 140
     },
     {
-      "epoch": 0.35481963335304556,
-      "grad_norm": 0.8885554899363501,
+      "epoch": 0.354924578527063,
+      "grad_norm": 0.8518067973190109,
       "learning_rate": 5e-06,
-      "loss": 0.6661,
+      "loss": 0.6648,
       "step": 150
     },
     {
-      "epoch": 0.3784742755765819,
-      "grad_norm": 0.4781037463107871,
+      "epoch": 0.37858621709553386,
+      "grad_norm": 0.5115922350127938,
       "learning_rate": 5e-06,
-      "loss": 0.6701,
+      "loss": 0.6691,
       "step": 160
     },
     {
-      "epoch": 0.4021289178001183,
-      "grad_norm": 0.6023227514998295,
+      "epoch": 0.4022478556640047,
+      "grad_norm": 0.7022530591693269,
       "learning_rate": 5e-06,
-      "loss": 0.6677,
+      "loss": 0.667,
       "step": 170
     },
     {
-      "epoch": 0.42578356002365464,
-      "grad_norm": 0.5603961021476538,
+      "epoch": 0.4259094942324756,
+      "grad_norm": 0.5834704070363212,
       "learning_rate": 5e-06,
-      "loss": 0.6742,
+      "loss": 0.6739,
       "step": 180
     },
     {
-      "epoch": 0.449438202247191,
-      "grad_norm": 0.4984868681428728,
+      "epoch": 0.44957113280094646,
+      "grad_norm": 0.4653603355749383,
       "learning_rate": 5e-06,
-      "loss": 0.6711,
+      "loss": 0.6698,
       "step": 190
     },
     {
-      "epoch": 0.47309284447072736,
-      "grad_norm": 0.6312594557804502,
+      "epoch": 0.4732327713694173,
+      "grad_norm": 0.568232089382732,
       "learning_rate": 5e-06,
-      "loss": 0.6689,
+      "loss": 0.6688,
       "step": 200
     },
     {
-      "epoch": 0.4967474866942638,
-      "grad_norm": 0.9550758404721197,
+      "epoch": 0.4968944099378882,
+      "grad_norm": 1.0573970656294764,
       "learning_rate": 5e-06,
-      "loss": 0.6707,
+      "loss": 0.6697,
       "step": 210
     },
     {
-      "epoch": 0.5204021289178001,
-      "grad_norm": 0.6169594889049308,
+      "epoch": 0.5205560485063591,
+      "grad_norm": 0.7158198022991291,
       "learning_rate": 5e-06,
-      "loss": 0.6699,
+      "loss": 0.6681,
       "step": 220
     },
     {
-      "epoch": 0.5440567711413364,
-      "grad_norm": 0.46636130254439206,
+      "epoch": 0.54421768707483,
+      "grad_norm": 0.43788884341448125,
       "learning_rate": 5e-06,
-      "loss": 0.6615,
+      "loss": 0.6603,
       "step": 230
     },
     {
-      "epoch": 0.5677114133648729,
-      "grad_norm": 0.5215033819505257,
+      "epoch": 0.5678793256433008,
+      "grad_norm": 0.4725698050393869,
       "learning_rate": 5e-06,
-      "loss": 0.6657,
+      "loss": 0.6645,
       "step": 240
     },
     {
-      "epoch": 0.5913660555884093,
-      "grad_norm": 0.5111456784576656,
+      "epoch": 0.5915409642117717,
+      "grad_norm": 0.6197309721248793,
       "learning_rate": 5e-06,
-      "loss": 0.6633,
+      "loss": 0.6616,
       "step": 250
     },
     {
-      "epoch": 0.6150206978119456,
-      "grad_norm": 0.517604760100503,
+      "epoch": 0.6152026027802425,
+      "grad_norm": 0.5419735846310992,
       "learning_rate": 5e-06,
-      "loss": 0.6531,
+      "loss": 0.652,
       "step": 260
     },
     {
-      "epoch": 0.638675340035482,
-      "grad_norm": 0.6357411311158894,
+      "epoch": 0.6388642413487134,
+      "grad_norm": 0.7492458117282386,
       "learning_rate": 5e-06,
-      "loss": 0.6652,
+      "loss": 0.6641,
       "step": 270
     },
     {
-      "epoch": 0.6623299822590183,
-      "grad_norm": 0.6476086780436829,
+      "epoch": 0.6625258799171843,
+      "grad_norm": 0.6330397603875498,
       "learning_rate": 5e-06,
-      "loss": 0.6676,
+      "loss": 0.6652,
       "step": 280
     },
     {
-      "epoch": 0.6859846244825547,
-      "grad_norm": 0.46209126279187906,
+      "epoch": 0.6861875184856552,
+      "grad_norm": 0.5065865148226276,
       "learning_rate": 5e-06,
-      "loss": 0.6601,
+      "loss": 0.6597,
       "step": 290
     },
     {
-      "epoch": 0.7096392667060911,
-      "grad_norm": 0.521950153282793,
+      "epoch": 0.709849157054126,
+      "grad_norm": 0.6357867453691719,
       "learning_rate": 5e-06,
-      "loss": 0.6554,
+      "loss": 0.6535,
       "step": 300
     },
     {
-      "epoch": 0.7332939089296274,
-      "grad_norm": 0.47211201942387604,
+      "epoch": 0.7335107956225969,
+      "grad_norm": 0.5149227689012754,
       "learning_rate": 5e-06,
-      "loss": 0.6617,
+      "loss": 0.6619,
       "step": 310
     },
     {
-      "epoch": 0.7569485511531638,
-      "grad_norm": 0.5341122235375386,
+      "epoch": 0.7571724341910677,
+      "grad_norm": 0.44037598215405893,
       "learning_rate": 5e-06,
-      "loss": 0.6563,
+      "loss": 0.6556,
       "step": 320
     },
     {
-      "epoch": 0.7806031933767001,
-      "grad_norm": 0.5708931772664904,
+      "epoch": 0.7808340727595386,
+      "grad_norm": 0.5181894268105267,
       "learning_rate": 5e-06,
-      "loss": 0.654,
+      "loss": 0.6541,
       "step": 330
     },
     {
-      "epoch": 0.8042578356002366,
-      "grad_norm": 0.6195560307499095,
+      "epoch": 0.8044957113280095,
+      "grad_norm": 0.6098945421576618,
       "learning_rate": 5e-06,
-      "loss": 0.6543,
+      "loss": 0.6538,
       "step": 340
     },
     {
-      "epoch": 0.8279124778237729,
-      "grad_norm": 0.5625110791293699,
+      "epoch": 0.8281573498964804,
+      "grad_norm": 0.5384169075272255,
       "learning_rate": 5e-06,
-      "loss": 0.6621,
+      "loss": 0.6602,
       "step": 350
     },
     {
-      "epoch": 0.8515671200473093,
-      "grad_norm": 0.4510061091403496,
+      "epoch": 0.8518189884649512,
+      "grad_norm": 0.47214684121487954,
       "learning_rate": 5e-06,
-      "loss": 0.6546,
+      "loss": 0.6538,
       "step": 360
     },
     {
-      "epoch": 0.8752217622708457,
-      "grad_norm": 0.46670984016010486,
+      "epoch": 0.8754806270334221,
+      "grad_norm": 0.4904597024034496,
       "learning_rate": 5e-06,
-      "loss": 0.6576,
+      "loss": 0.6567,
       "step": 370
     },
     {
-      "epoch": 0.898876404494382,
-      "grad_norm": 0.5044563574335732,
+      "epoch": 0.8991422656018929,
+      "grad_norm": 0.4775889282770322,
       "learning_rate": 5e-06,
-      "loss": 0.6489,
+      "loss": 0.6478,
       "step": 380
     },
     {
-      "epoch": 0.9225310467179184,
-      "grad_norm": 0.6276485740632245,
+      "epoch": 0.9228039041703638,
+      "grad_norm": 0.6785347844213175,
       "learning_rate": 5e-06,
-      "loss": 0.6548,
+      "loss": 0.6553,
       "step": 390
     },
     {
-      "epoch": 0.9461856889414547,
-      "grad_norm": 0.4849232290323015,
+      "epoch": 0.9464655427388347,
+      "grad_norm": 0.48131943447081366,
       "learning_rate": 5e-06,
-      "loss": 0.6473,
+      "loss": 0.6466,
       "step": 400
     },
     {
-      "epoch": 0.9698403311649911,
-      "grad_norm": 0.47066212467927304,
+      "epoch": 0.9701271813073056,
+      "grad_norm": 0.4819882000845211,
       "learning_rate": 5e-06,
-      "loss": 0.6504,
+      "loss": 0.6491,
       "step": 410
     },
     {
-      "epoch": 0.9934949733885275,
-      "grad_norm": 0.5498376915549745,
+      "epoch": 0.9937888198757764,
+      "grad_norm": 0.48326887732044893,
       "learning_rate": 5e-06,
-      "loss": 0.6489,
+      "loss": 0.648,
       "step": 420
     },
     {
-      "epoch": 0.9982259018332348,
-      "eval_loss": 0.6508141756057739,
-      "eval_runtime": 226.6795,
-      "eval_samples_per_second": 50.243,
-      "eval_steps_per_second": 0.393,
+      "epoch": 0.9985211475894705,
+      "eval_loss": 0.6503860950469971,
+      "eval_runtime": 295.5868,
+      "eval_samples_per_second": 38.53,
+      "eval_steps_per_second": 0.602,
       "step": 422
     },
     {
-      "epoch": 1.0171496156120639,
-      "grad_norm": 0.5216981625298186,
+      "epoch": 1.0177462289263532,
+      "grad_norm": 0.5034024020075702,
       "learning_rate": 5e-06,
-      "loss": 0.6225,
+      "loss": 0.6519,
       "step": 430
     },
     {
-      "epoch": 1.0408042578356003,
-      "grad_norm": 0.5287278292361843,
+      "epoch": 1.041407867494824,
+      "grad_norm": 0.5181852823487637,
       "learning_rate": 5e-06,
-      "loss": 0.6077,
+      "loss": 0.6054,
       "step": 440
     },
     {
-      "epoch": 1.0644589000591367,
-      "grad_norm": 0.7304842680236713,
+      "epoch": 1.0650695060632949,
+      "grad_norm": 0.8086692425168592,
       "learning_rate": 5e-06,
-      "loss": 0.6058,
+      "loss": 0.6036,
       "step": 450
     },
     {
-      "epoch": 1.0881135422826729,
-      "grad_norm": 0.5644902109246774,
+      "epoch": 1.0887311446317658,
+      "grad_norm": 0.7087824451639375,
       "learning_rate": 5e-06,
       "loss": 0.6138,
       "step": 460
     },
     {
-      "epoch": 1.1117681845062093,
-      "grad_norm": 0.4753652134651053,
+      "epoch": 1.1123927832002367,
+      "grad_norm": 0.6293561803711789,
       "learning_rate": 5e-06,
-      "loss": 0.6129,
+      "loss": 0.611,
       "step": 470
     },
     {
-      "epoch": 1.1354228267297457,
-      "grad_norm": 0.7165345582019033,
+      "epoch": 1.1360544217687074,
+      "grad_norm": 0.8003707364488958,
       "learning_rate": 5e-06,
-      "loss": 0.6125,
+      "loss": 0.6114,
       "step": 480
     },
     {
-      "epoch": 1.1590774689532821,
-      "grad_norm": 0.4641812019487026,
+      "epoch": 1.1597160603371783,
+      "grad_norm": 0.5556436203083284,
       "learning_rate": 5e-06,
-      "loss": 0.6106,
+      "loss": 0.6083,
       "step": 490
     },
     {
-      "epoch": 1.1827321111768185,
-      "grad_norm": 0.4846409340358244,
+      "epoch": 1.1833776989056493,
+      "grad_norm": 0.5222890058260059,
       "learning_rate": 5e-06,
-      "loss": 0.6065,
+      "loss": 0.6038,
       "step": 500
     },
     {
-      "epoch": 1.2063867534003547,
-      "grad_norm": 0.5624538054911821,
+      "epoch": 1.2070393374741202,
+      "grad_norm": 0.5834395725334883,
       "learning_rate": 5e-06,
-      "loss": 0.6058,
+      "loss": 0.6038,
       "step": 510
     },
     {
-      "epoch": 1.2300413956238911,
-      "grad_norm": 0.5069531857984526,
+      "epoch": 1.2307009760425909,
+      "grad_norm": 0.6523486232667094,
       "learning_rate": 5e-06,
-      "loss": 0.6114,
+      "loss": 0.6103,
       "step": 520
     },
     {
-      "epoch": 1.2536960378474276,
-      "grad_norm": 0.5605749208808489,
+      "epoch": 1.2543626146110618,
+      "grad_norm": 0.509097134236641,
       "learning_rate": 5e-06,
-      "loss": 0.6117,
+      "loss": 0.6129,
       "step": 530
     },
     {
-      "epoch": 1.277350680070964,
-      "grad_norm": 0.5081677690669225,
+      "epoch": 1.2780242531795327,
+      "grad_norm": 0.6109585275321274,
       "learning_rate": 5e-06,
-      "loss": 0.6133,
+      "loss": 0.6125,
       "step": 540
     },
     {
-      "epoch": 1.3010053222945004,
-      "grad_norm": 0.5493923133675747,
+      "epoch": 1.3016858917480034,
+      "grad_norm": 0.4671830560905703,
       "learning_rate": 5e-06,
-      "loss": 0.6101,
+      "loss": 0.6094,
       "step": 550
     },
     {
-      "epoch": 1.3246599645180366,
-      "grad_norm": 0.44898345952125707,
+      "epoch": 1.3253475303164743,
+      "grad_norm": 0.43619649641699987,
       "learning_rate": 5e-06,
-      "loss": 0.6096,
+      "loss": 0.6076,
       "step": 560
     },
     {
-      "epoch": 1.348314606741573,
-      "grad_norm": 0.6655768726116028,
+      "epoch": 1.3490091688849453,
+      "grad_norm": 0.9334517894855655,
       "learning_rate": 5e-06,
-      "loss": 0.6094,
+      "loss": 0.6086,
       "step": 570
     },
     {
-      "epoch": 1.3719692489651094,
-      "grad_norm": 0.4649052455972333,
+      "epoch": 1.3726708074534162,
+      "grad_norm": 0.6158441071105143,
       "learning_rate": 5e-06,
-      "loss": 0.614,
+      "loss": 0.6114,
       "step": 580
     },
     {
-      "epoch": 1.3956238911886458,
-      "grad_norm": 0.43539888089438844,
+      "epoch": 1.396332446021887,
+      "grad_norm": 0.4177935072924951,
       "learning_rate": 5e-06,
-      "loss": 0.6138,
+      "loss": 0.6118,
       "step": 590
     },
     {
-      "epoch": 1.4192785334121822,
-      "grad_norm": 0.5826398325628406,
+      "epoch": 1.4199940845903578,
+      "grad_norm": 0.5261498840709943,
       "learning_rate": 5e-06,
-      "loss": 0.601,
+      "loss": 0.6012,
       "step": 600
     },
     {
-      "epoch": 1.4429331756357184,
-      "grad_norm": 0.4128719638682357,
+      "epoch": 1.4436557231588287,
+      "grad_norm": 0.4118403315344034,
       "learning_rate": 5e-06,
-      "loss": 0.6022,
+      "loss": 0.6006,
       "step": 610
     },
     {
-      "epoch": 1.4665878178592548,
-      "grad_norm": 0.45529767094349805,
+      "epoch": 1.4673173617272997,
+      "grad_norm": 0.4500509679639768,
       "learning_rate": 5e-06,
-      "loss": 0.6052,
+      "loss": 0.6041,
       "step": 620
     },
     {
-      "epoch": 1.4902424600827913,
-      "grad_norm": 0.4330632579708222,
+      "epoch": 1.4909790002957704,
+      "grad_norm": 0.47480411557819086,
       "learning_rate": 5e-06,
-      "loss": 0.6156,
+      "loss": 0.6161,
       "step": 630
     },
     {
-      "epoch": 1.5138971023063275,
-      "grad_norm": 0.44938881612128456,
+      "epoch": 1.5146406388642415,
+      "grad_norm": 0.44192140481751196,
       "learning_rate": 5e-06,
-      "loss": 0.6152,
+      "loss": 0.6104,
       "step": 640
     },
     {
-      "epoch": 1.537551744529864,
-      "grad_norm": 0.46628155511438146,
+      "epoch": 1.5383022774327122,
+      "grad_norm": 0.43019141037435493,
       "learning_rate": 5e-06,
-      "loss": 0.6041,
+      "loss": 0.6044,
       "step": 650
     },
     {
-      "epoch": 1.5612063867534003,
-      "grad_norm": 0.5843372770531808,
+      "epoch": 1.5619639160011831,
+      "grad_norm": 0.5928638480003088,
       "learning_rate": 5e-06,
-      "loss": 0.6091,
+      "loss": 0.6093,
       "step": 660
     },
     {
-      "epoch": 1.5848610289769367,
-      "grad_norm": 0.5865805864061058,
+      "epoch": 1.585625554569654,
+      "grad_norm": 0.680185794067505,
       "learning_rate": 5e-06,
-      "loss": 0.6055,
+      "loss": 0.6018,
       "step": 670
     },
     {
-      "epoch": 1.6085156712004731,
-      "grad_norm": 0.6180152508932449,
+      "epoch": 1.6092871931381247,
+      "grad_norm": 0.4446498601791125,
       "learning_rate": 5e-06,
-      "loss": 0.6098,
+      "loss": 0.6103,
       "step": 680
     },
     {
-      "epoch": 1.6321703134240093,
-      "grad_norm": 0.5250880775614019,
+      "epoch": 1.6329488317065957,
+      "grad_norm": 0.5328567150499224,
       "learning_rate": 5e-06,
-      "loss": 0.6035,
+      "loss": 0.6013,
       "step": 690
     },
     {
-      "epoch": 1.655824955647546,
-      "grad_norm": 0.478978441549023,
+      "epoch": 1.6566104702750666,
+      "grad_norm": 0.5840475503667699,
       "learning_rate": 5e-06,
-      "loss": 0.6107,
+      "loss": 0.6109,
       "step": 700
     },
     {
-      "epoch": 1.6794795978710821,
-      "grad_norm": 0.5161048108398052,
+      "epoch": 1.6802721088435373,
+      "grad_norm": 0.5235562677113405,
       "learning_rate": 5e-06,
-      "loss": 0.6106,
+      "loss": 0.6094,
       "step": 710
     },
     {
-      "epoch": 1.7031342400946186,
-      "grad_norm": 0.5444636992966723,
+      "epoch": 1.7039337474120084,
+      "grad_norm": 0.6484686160336842,
       "learning_rate": 5e-06,
-      "loss": 0.6123,
+      "loss": 0.6122,
       "step": 720
     },
     {
-      "epoch": 1.726788882318155,
-      "grad_norm": 0.49014325863511293,
+      "epoch": 1.7275953859804791,
+      "grad_norm": 0.5923644814057659,
       "learning_rate": 5e-06,
-      "loss": 0.6131,
+      "loss": 0.612,
       "step": 730
     },
     {
-      "epoch": 1.7504435245416912,
-      "grad_norm": 0.6313993012312799,
+      "epoch": 1.75125702454895,
+      "grad_norm": 0.4896255525659027,
       "learning_rate": 5e-06,
-      "loss": 0.6059,
+      "loss": 0.6042,
       "step": 740
     },
     {
-      "epoch": 1.7740981667652278,
-      "grad_norm": 0.5158091820023666,
+      "epoch": 1.774918663117421,
+      "grad_norm": 0.47313474988505516,
       "learning_rate": 5e-06,
-      "loss": 0.6116,
+      "loss": 0.6105,
       "step": 750
     },
     {
-      "epoch": 1.797752808988764,
-      "grad_norm": 0.47029237511942096,
+      "epoch": 1.7985803016858917,
+      "grad_norm": 0.4971344193994014,
       "learning_rate": 5e-06,
-      "loss": 0.617,
+      "loss": 0.6146,
       "step": 760
     },
     {
-      "epoch": 1.8214074512123004,
-      "grad_norm": 0.4280088668307211,
+      "epoch": 1.8222419402543626,
+      "grad_norm": 0.42650481882946567,
       "learning_rate": 5e-06,
-      "loss": 0.6167,
+      "loss": 0.6161,
       "step": 770
     },
     {
-      "epoch": 1.8450620934358368,
-      "grad_norm": 0.42280798454326274,
+      "epoch": 1.8459035788228335,
+      "grad_norm": 0.45409425789151847,
       "learning_rate": 5e-06,
-      "loss": 0.6094,
+      "loss": 0.607,
       "step": 780
     },
     {
-      "epoch": 1.868716735659373,
-      "grad_norm": 0.5015010241124669,
+      "epoch": 1.8695652173913042,
+      "grad_norm": 0.47548804917566284,
       "learning_rate": 5e-06,
-      "loss": 0.606,
+      "loss": 0.6057,
       "step": 790
     },
     {
-      "epoch": 1.8923713778829097,
-      "grad_norm": 0.4647527250682549,
+      "epoch": 1.8932268559597754,
+      "grad_norm": 0.49936009431407224,
       "learning_rate": 5e-06,
       "loss": 0.614,
       "step": 800
     },
     {
-      "epoch": 1.9160260201064458,
-      "grad_norm": 0.45335597096154956,
+      "epoch": 1.916888494528246,
+      "grad_norm": 0.42095008482708873,
       "learning_rate": 5e-06,
-      "loss": 0.6142,
+      "loss": 0.6118,
       "step": 810
     },
     {
-      "epoch": 1.9396806623299823,
-      "grad_norm": 0.4377674877634198,
+      "epoch": 1.940550133096717,
+      "grad_norm": 0.469345021146904,
       "learning_rate": 5e-06,
-      "loss": 0.6009,
+      "loss": 0.6,
       "step": 820
     },
     {
-      "epoch": 1.9633353045535187,
-      "grad_norm": 0.4922578419992401,
+      "epoch": 1.964211771665188,
+      "grad_norm": 0.4821117072104332,
       "learning_rate": 5e-06,
-      "loss": 0.6108,
+      "loss": 0.6112,
       "step": 830
     },
     {
-      "epoch": 1.9869899467770549,
-      "grad_norm": 0.4440378004846697,
+      "epoch": 1.9878734102336586,
+      "grad_norm": 0.42444873739826294,
       "learning_rate": 5e-06,
-      "loss": 0.5988,
+      "loss": 0.5984,
       "step": 840
     },
     {
-      "epoch": 1.9988172678888232,
-      "eval_loss": 0.6402984857559204,
-      "eval_runtime": 227.8867,
-      "eval_samples_per_second": 49.977,
-      "eval_steps_per_second": 0.391,
+      "epoch": 1.999704229517894,
+      "eval_loss": 0.640017032623291,
+      "eval_runtime": 291.0784,
+      "eval_samples_per_second": 39.127,
+      "eval_steps_per_second": 0.612,
       "step": 845
     },
     {
-      "epoch": 2.0106445890005915,
-      "grad_norm": 0.6249254373352294,
+      "epoch": 2.0118308192842353,
+      "grad_norm": 0.743615658405185,
       "learning_rate": 5e-06,
-      "loss": 0.5864,
+      "loss": 0.612,
       "step": 850
     },
     {
-      "epoch": 2.0342992312241277,
-      "grad_norm": 0.5229366411844095,
+      "epoch": 2.0354924578527065,
+      "grad_norm": 0.6307060227030052,
       "learning_rate": 5e-06,
-      "loss": 0.5668,
+      "loss": 0.5665,
       "step": 860
     },
     {
-      "epoch": 2.057953873447664,
-      "grad_norm": 0.5273129182521392,
+      "epoch": 2.059154096421177,
+      "grad_norm": 0.4909469687109977,
       "learning_rate": 5e-06,
-      "loss": 0.569,
+      "loss": 0.5669,
       "step": 870
     },
     {
-      "epoch": 2.0816085156712005,
-      "grad_norm": 0.5807006714823286,
+      "epoch": 2.082815734989648,
+      "grad_norm": 0.4982651206960173,
       "learning_rate": 5e-06,
-      "loss": 0.5605,
+      "loss": 0.5576,
       "step": 880
     },
     {
-      "epoch": 2.1052631578947367,
-      "grad_norm": 0.5316074498783832,
+      "epoch": 2.106477373558119,
+      "grad_norm": 0.5189496162921208,
       "learning_rate": 5e-06,
-      "loss": 0.5644,
+      "loss": 0.5632,
       "step": 890
     },
     {
-      "epoch": 2.1289178001182734,
-      "grad_norm": 0.571954438502547,
+      "epoch": 2.1301390121265897,
+      "grad_norm": 0.5290452662480181,
       "learning_rate": 5e-06,
-      "loss": 0.5674,
+      "loss": 0.5673,
       "step": 900
     },
     {
-      "epoch": 2.1525724423418096,
-      "grad_norm": 0.4503797360191138,
+      "epoch": 2.1538006506950604,
+      "grad_norm": 0.4298930748962063,
       "learning_rate": 5e-06,
-      "loss": 0.5625,
+      "loss": 0.5608,
       "step": 910
     },
     {
-      "epoch": 2.1762270845653457,
-      "grad_norm": 0.49334190713186854,
+      "epoch": 2.1774622892635316,
+      "grad_norm": 0.46653862184558975,
       "learning_rate": 5e-06,
-      "loss": 0.5648,
+      "loss": 0.5622,
       "step": 920
     },
     {
-      "epoch": 2.1998817267888824,
-      "grad_norm": 0.5335672716582398,
+      "epoch": 2.2011239278320023,
+      "grad_norm": 0.47565404277144174,
       "learning_rate": 5e-06,
-      "loss": 0.5647,
+      "loss": 0.5634,
       "step": 930
     },
     {
-      "epoch": 2.2235363690124186,
-      "grad_norm": 0.5933317390318844,
+      "epoch": 2.2247855664004734,
+      "grad_norm": 0.45977306537518353,
       "learning_rate": 5e-06,
-      "loss": 0.5662,
+      "loss": 0.5642,
       "step": 940
     },
     {
-      "epoch": 2.247191011235955,
-      "grad_norm": 0.5891028799938252,
+      "epoch": 2.248447204968944,
+      "grad_norm": 0.6195624778560548,
       "learning_rate": 5e-06,
-      "loss": 0.5716,
+      "loss": 0.5701,
       "step": 950
     },
     {
-      "epoch": 2.2708456534594914,
-      "grad_norm": 0.5459106423210626,
+      "epoch": 2.272108843537415,
+      "grad_norm": 0.5625286018926492,
       "learning_rate": 5e-06,
-      "loss": 0.5651,
+      "loss": 0.5624,
       "step": 960
     },
     {
-      "epoch": 2.2945002956830276,
-      "grad_norm": 0.5290952299819603,
+      "epoch": 2.295770482105886,
+      "grad_norm": 0.5516894450156284,
       "learning_rate": 5e-06,
-      "loss": 0.568,
+      "loss": 0.5688,
       "step": 970
     },
     {
-      "epoch": 2.3181549379065642,
-      "grad_norm": 0.612560828315047,
+      "epoch": 2.3194321206743567,
+      "grad_norm": 0.46703213745702465,
       "learning_rate": 5e-06,
-      "loss": 0.5695,
+      "loss": 0.5687,
       "step": 980
     },
     {
-      "epoch": 2.3418095801301004,
-      "grad_norm": 0.4941237525745503,
+      "epoch": 2.3430937592428274,
+      "grad_norm": 0.4529577298482224,
       "learning_rate": 5e-06,
-      "loss": 0.5689,
+      "loss": 0.5664,
       "step": 990
     },
     {
-      "epoch": 2.365464222353637,
-      "grad_norm": 0.5650925849128444,
+      "epoch": 2.3667553978112985,
+      "grad_norm": 0.5294804903105876,
       "learning_rate": 5e-06,
-      "loss": 0.5688,
+      "loss": 0.5679,
       "step": 1000
     },
     {
-      "epoch": 2.3891188645771733,
-      "grad_norm": 0.4960493342953518,
+      "epoch": 2.390417036379769,
+      "grad_norm": 0.476104510415757,
       "learning_rate": 5e-06,
-      "loss": 0.568,
+      "loss": 0.5661,
       "step": 1010
     },
     {
-      "epoch": 2.4127735068007095,
-      "grad_norm": 0.5675556976300212,
+      "epoch": 2.4140786749482404,
+      "grad_norm": 0.50214160643677,
       "learning_rate": 5e-06,
-      "loss": 0.5656,
+      "loss": 0.5632,
       "step": 1020
     },
     {
-      "epoch": 2.436428149024246,
-      "grad_norm": 0.4679420975237042,
+      "epoch": 2.437740313516711,
+      "grad_norm": 0.4770392958453329,
       "learning_rate": 5e-06,
-      "loss": 0.5658,
+      "loss": 0.5646,
       "step": 1030
     },
     {
-      "epoch": 2.4600827912477823,
-      "grad_norm": 0.5934681961848346,
+      "epoch": 2.4614019520851818,
+      "grad_norm": 0.4746961197857643,
       "learning_rate": 5e-06,
-      "loss": 0.5699,
+      "loss": 0.5663,
       "step": 1040
     },
     {
-      "epoch": 2.483737433471319,
-      "grad_norm": 0.4994164056392902,
+      "epoch": 2.485063590653653,
+      "grad_norm": 0.5075465690483602,
       "learning_rate": 5e-06,
       "loss": 0.5705,
       "step": 1050
     },
     {
-      "epoch": 2.507392075694855,
-      "grad_norm": 0.553203563454583,
+      "epoch": 2.5087252292221236,
+      "grad_norm": 0.5735236324477879,
       "learning_rate": 5e-06,
       "loss": 0.5692,
       "step": 1060
     },
     {
-      "epoch": 2.5310467179183913,
-      "grad_norm": 0.4828644604519897,
+      "epoch": 2.5323868677905947,
+      "grad_norm": 0.7357338280472365,
       "learning_rate": 5e-06,
-      "loss": 0.5741,
+      "loss": 0.5693,
       "step": 1070
     },
     {
-      "epoch": 2.554701360141928,
-      "grad_norm": 0.49982783159481065,
+      "epoch": 2.5560485063590654,
+      "grad_norm": 0.6312728091303266,
       "learning_rate": 5e-06,
-      "loss": 0.5729,
+      "loss": 0.5721,
       "step": 1080
     },
     {
-      "epoch": 2.578356002365464,
-      "grad_norm": 0.5028480314612812,
+      "epoch": 2.579710144927536,
+      "grad_norm": 0.6063258814160167,
       "learning_rate": 5e-06,
-      "loss": 0.5688,
+      "loss": 0.5668,
       "step": 1090
     },
     {
-      "epoch": 2.6020106445890008,
-      "grad_norm": 0.47037814252850013,
+      "epoch": 2.603371783496007,
+      "grad_norm": 0.512061567286246,
       "learning_rate": 5e-06,
-      "loss": 0.5693,
+      "loss": 0.5706,
       "step": 1100
     },
     {
-      "epoch": 2.625665286812537,
-      "grad_norm": 0.5106941109306207,
+      "epoch": 2.627033422064478,
+      "grad_norm": 0.4805538505321553,
       "learning_rate": 5e-06,
-      "loss": 0.5671,
+      "loss": 0.5643,
       "step": 1110
     },
     {
-      "epoch": 2.649319929036073,
-      "grad_norm": 0.6215547951000225,
+      "epoch": 2.6506950606329487,
+      "grad_norm": 0.47434096978031415,
       "learning_rate": 5e-06,
-      "loss": 0.5792,
+      "loss": 0.5784,
       "step": 1120
     },
     {
-      "epoch": 2.67297457125961,
-      "grad_norm": 0.5708225359288293,
+      "epoch": 2.67435669920142,
+      "grad_norm": 0.48174002692295587,
       "learning_rate": 5e-06,
-      "loss": 0.5667,
+      "loss": 0.5642,
       "step": 1130
     },
     {
-      "epoch": 2.696629213483146,
-      "grad_norm": 0.4694908632163934,
+      "epoch": 2.6980183377698905,
+      "grad_norm": 0.5206174302871078,
       "learning_rate": 5e-06,
-      "loss": 0.5666,
+      "loss": 0.5647,
       "step": 1140
     },
     {
-      "epoch": 2.7202838557066826,
-      "grad_norm": 0.5487886870493401,
+      "epoch": 2.7216799763383612,
+      "grad_norm": 0.5231066430924299,
       "learning_rate": 5e-06,
-      "loss": 0.569,
+      "loss": 0.5717,
       "step": 1150
     },
     {
-      "epoch": 2.743938497930219,
-      "grad_norm": 0.5178726800202738,
+      "epoch": 2.7453416149068324,
+      "grad_norm": 0.5493755325406757,
       "learning_rate": 5e-06,
-      "loss": 0.5755,
+      "loss": 0.5722,
       "step": 1160
     },
     {
-      "epoch": 2.767593140153755,
-      "grad_norm": 0.572400968003655,
+      "epoch": 2.769003253475303,
+      "grad_norm": 0.4710457798970187,
       "learning_rate": 5e-06,
-      "loss": 0.5687,
+      "loss": 0.5673,
       "step": 1170
     },
     {
-      "epoch": 2.7912477823772917,
-      "grad_norm": 0.4505102109858482,
+      "epoch": 2.792664892043774,
+      "grad_norm": 0.6017685441197285,
       "learning_rate": 5e-06,
-      "loss": 0.5694,
+      "loss": 0.5673,
       "step": 1180
     },
     {
-      "epoch": 2.814902424600828,
-      "grad_norm": 0.4721423964551259,
+      "epoch": 2.816326530612245,
+      "grad_norm": 0.48667322523587175,
       "learning_rate": 5e-06,
-      "loss": 0.5711,
+      "loss": 0.5712,
       "step": 1190
     },
     {
-      "epoch": 2.8385570668243645,
-      "grad_norm": 0.45313707571101325,
+      "epoch": 2.8399881691807156,
+      "grad_norm": 0.6136328812041214,
       "learning_rate": 5e-06,
-      "loss": 0.568,
+      "loss": 0.5689,
       "step": 1200
     },
     {
-      "epoch": 2.8622117090479007,
-      "grad_norm": 0.4905572612804298,
+      "epoch": 2.8636498077491868,
+      "grad_norm": 0.4848336832061156,
       "learning_rate": 5e-06,
-      "loss": 0.5708,
+      "loss": 0.5683,
       "step": 1210
     },
     {
-      "epoch": 2.885866351271437,
-      "grad_norm": 0.4488283221129483,
+      "epoch": 2.8873114463176575,
+      "grad_norm": 0.6143451558754445,
       "learning_rate": 5e-06,
-      "loss": 0.5644,
+      "loss": 0.5611,
       "step": 1220
     },
     {
-      "epoch": 2.9095209934949735,
-      "grad_norm": 0.5086777707270197,
+      "epoch": 2.9109730848861286,
+      "grad_norm": 0.45186971294615735,
       "learning_rate": 5e-06,
-      "loss": 0.57,
+      "loss": 0.5713,
       "step": 1230
     },
     {
-      "epoch": 2.9331756357185097,
-      "grad_norm": 0.5014035211864805,
+      "epoch": 2.9346347234545993,
+      "grad_norm": 0.45118967147278705,
       "learning_rate": 5e-06,
-      "loss": 0.5719,
+      "loss": 0.5703,
       "step": 1240
     },
     {
-      "epoch": 2.9568302779420463,
-      "grad_norm": 0.46427442721882733,
+      "epoch": 2.95829636202307,
+      "grad_norm": 0.5082137766462498,
       "learning_rate": 5e-06,
-      "loss": 0.5812,
+      "loss": 0.5791,
       "step": 1250
     },
     {
-      "epoch": 2.9804849201655825,
-      "grad_norm": 0.4933592342573123,
+      "epoch": 2.9819580005915407,
+      "grad_norm": 0.5446211548606755,
       "learning_rate": 5e-06,
-      "loss": 0.5728,
+      "loss": 0.5714,
       "step": 1260
     },
     {
-      "epoch": 2.9946777054997042,
-      "eval_loss": 0.6413019895553589,
-      "eval_runtime": 227.2584,
-      "eval_samples_per_second": 50.115,
-      "eval_steps_per_second": 0.392,
+      "epoch": 2.9961549837326236,
+      "eval_loss": 0.6407967805862427,
+      "eval_runtime": 281.5097,
+      "eval_samples_per_second": 40.457,
+      "eval_steps_per_second": 0.632,
       "step": 1266
     },
     {
-      "epoch": 2.9946777054997042,
+      "epoch": 2.9961549837326236,
       "step": 1266,
-      "total_flos": 2120178393415680.0,
-      "train_loss": 0.6201754979801027,
-      "train_runtime": 38082.1567,
-      "train_samples_per_second": 17.045,
-      "train_steps_per_second": 0.033
+      "total_flos": 2120387773071360.0,
+      "train_loss": 0.6192890667425701,
+      "train_runtime": 41807.2835,
+      "train_samples_per_second": 15.526,
+      "train_steps_per_second": 0.03
     }
   ],
   "logging_steps": 10,
@@ -941,7 +941,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 2120178393415680.0,
+  "total_flos": 2120387773071360.0,
   "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null