End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:06297c81f85c777ce5474785196cf01778d44aab7c10e0597e1e3bbd96534d64
+oid sha256:4335d492132d14f7396aff2f237b7c63752b0998caf2d5dfb15bf7b673feccc8
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 1.6401619859607309e-06,
-    "train_runtime": 1223.4771,
-    "train_samples": 11,
-    "train_samples_per_second": 0.262,
+    "train_loss": 2.083803985897248e-06,
+    "train_runtime": 1271.7904,
+    "train_samples": 20,
+    "train_samples_per_second": 0.252,
     "train_steps_per_second": 0.016
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 1.6401619859607309e-06,
-    "train_runtime": 1223.4771,
-    "train_samples": 11,
-    "train_samples_per_second": 0.262,
+    "train_loss": 2.083803985897248e-06,
+    "train_runtime": 1271.7904,
+    "train_samples": 20,
+    "train_samples_per_second": 0.252,
     "train_steps_per_second": 0.016
 }

trainer_state.json

@@ -2,7 +2,7 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 6.7272727272727275,
+  "epoch": 4.0,
   "eval_steps": 500,
   "global_step": 20,
   "is_hyper_param_search": false,
@@ -10,209 +10,209 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 473.125,
-      "epoch": 0.7272727272727273,
-      "grad_norm": 0.5780012011528015,
-      "kl": 0.0018195659504272044,
+      "completion_length": 358.03125,
+      "epoch": 0.4,
+      "grad_norm": 0.4435279965400696,
+      "kl": 0.0019558896019589156,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 0.7852539010345936,
-      "reward_std": 0.6700366213917732,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.0,
+      "reward": 2.21342371404171,
+      "reward_std": 1.5274258591234684,
+      "rewards/concensus_correctness_reward_func": 0.5029999986290932,
+      "rewards/consensus_reward_func": 0.625,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.6348789054900408,
+      "rewards/final_correctness_reward_func": 0.125,
+      "rewards/question_recreation_reward_func": 0.5663924645632505,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.15037500904873013,
+      "rewards/xmlcount_reward_func": 0.39403124898672104,
       "step": 2
     },
     {
-      "completion_length": 461.32142857142856,
-      "epoch": 1.3636363636363638,
-      "grad_norm": 0.43347495794296265,
-      "kl": 0.0015060585823708347,
+      "completion_length": 332.375,
+      "epoch": 0.8,
+      "grad_norm": 0.6189950704574585,
+      "kl": 0.002220589100033976,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 1.1546246686152049,
-      "reward_std": 0.9197601931435722,
-      "rewards/concensus_correctness_reward_func": 0.03864285775593349,
-      "rewards/consensus_reward_func": 0.35714285714285715,
+      "reward": 1.9861630648374557,
+      "reward_std": 0.9714729022234678,
+      "rewards/concensus_correctness_reward_func": 0.12475000228732824,
+      "rewards/consensus_reward_func": 0.6875,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.41933891496488024,
+      "rewards/question_recreation_reward_func": 0.5830693021416664,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.3395000100135803,
+      "rewards/strict_format_reward_func": 0.046875,
+      "rewards/xmlcount_reward_func": 0.54396877810359,
       "step": 4
     },
     {
-      "completion_length": 553.6785714285714,
-      "epoch": 2.0,
-      "grad_norm": 0.24650119245052338,
-      "kl": 0.001656832073682121,
+      "completion_length": 412.09375,
+      "epoch": 1.2,
+      "grad_norm": 0.6348840594291687,
+      "kl": 0.0016933643346419558,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 1.446252771786281,
-      "reward_std": 1.6744095257350378,
-      "rewards/concensus_correctness_reward_func": 0.19264285692146846,
-      "rewards/consensus_reward_func": 0.7857142857142857,
+      "reward": 1.5512390583753586,
+      "reward_std": 1.3353641331195831,
+      "rewards/concensus_correctness_reward_func": 0.2669375017285347,
+      "rewards/consensus_reward_func": 0.1875,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5666455977729389,
+      "rewards/final_correctness_reward_func": 0.0625,
+      "rewards/question_recreation_reward_func": 0.5625203074887395,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.09874998139483589,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.45615623891353607,
       "step": 6
     },
     {
-      "completion_length": 489.90625,
-      "epoch": 2.7272727272727275,
-      "grad_norm": 0.36353105306625366,
-      "kl": 0.0016505489984410815,
+      "completion_length": 283.375,
+      "epoch": 1.6,
+      "grad_norm": 0.7926058173179626,
+      "kl": 0.0018770112801576033,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 1.2562414929270744,
-      "reward_std": 1.0721860602498055,
-      "rewards/concensus_correctness_reward_func": 0.05324999988079071,
-      "rewards/consensus_reward_func": 0.375,
+      "reward": 1.7348872683942318,
+      "reward_std": 1.064245454967022,
+      "rewards/concensus_correctness_reward_func": 0.11574999988079071,
+      "rewards/consensus_reward_func": 0.25,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5846789591014385,
+      "rewards/final_correctness_reward_func": 0.3125,
+      "rewards/question_recreation_reward_func": 0.6253872429952025,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.2276875041425228,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": 0.40000001416774467,
       "step": 8
     },
     {
-      "completion_length": 500.7142857142857,
-      "epoch": 3.3636363636363638,
-      "grad_norm": 0.5545067191123962,
-      "kl": 0.0017322281741404108,
+      "completion_length": 374.5,
+      "epoch": 2.0,
+      "grad_norm": 0.6137144565582275,
+      "kl": 0.0022025189246051013,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 1.230509843145098,
-      "reward_std": 1.23474839755467,
-      "rewards/concensus_correctness_reward_func": 0.0247142847095217,
-      "rewards/consensus_reward_func": 0.5,
+      "reward": 1.6147940009832382,
+      "reward_std": 1.4487705081701279,
+      "rewards/concensus_correctness_reward_func": 0.11381249874830246,
+      "rewards/consensus_reward_func": 0.1875,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.07142857142857142,
-      "rewards/question_recreation_reward_func": 0.3760812836033957,
+      "rewards/final_correctness_reward_func": 0.3125,
+      "rewards/question_recreation_reward_func": 0.4762627389281988,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.25828570978982107,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": 0.4934687502682209,
       "step": 10
     },
     {
-      "completion_length": 540.7142857142857,
-      "epoch": 4.0,
-      "grad_norm": 0.2851618528366089,
-      "kl": 0.0018666927935555577,
+      "completion_length": 441.34375,
+      "epoch": 2.4,
+      "grad_norm": 0.4752722978591919,
+      "kl": 0.002051696108537726,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 0.9000063027654376,
-      "reward_std": 0.8522530368396214,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.14285714285714285,
+      "reward": 1.3832528349012136,
+      "reward_std": 0.9370621740818024,
+      "rewards/concensus_correctness_reward_func": 0.054249998182058334,
+      "rewards/consensus_reward_func": 0.125,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5019705896931035,
+      "rewards/final_correctness_reward_func": 0.125,
+      "rewards/question_recreation_reward_func": 0.5265341196209192,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.2373214289546013,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.5368437469005585,
       "step": 12
     },
     {
-      "completion_length": 490.59375,
-      "epoch": 4.7272727272727275,
-      "grad_norm": 0.43547067046165466,
-      "kl": 0.0016540642027393915,
+      "completion_length": 408.8125,
+      "epoch": 2.8,
+      "grad_norm": 0.7636963129043579,
+      "kl": 0.002045114611973986,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 0.7657471336424351,
-      "reward_std": 0.8933027181774378,
-      "rewards/concensus_correctness_reward_func": 0.02162499912083149,
-      "rewards/consensus_reward_func": 0.1875,
+      "reward": 1.9599885269999504,
+      "reward_std": 2.0160218477249146,
+      "rewards/concensus_correctness_reward_func": 0.33500000089406967,
+      "rewards/consensus_reward_func": 0.375,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.44059090316295624,
+      "rewards/final_correctness_reward_func": 0.375,
+      "rewards/question_recreation_reward_func": 0.5672072526067495,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.1160312567371875,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.2921562595292926,
       "step": 14
     },
     {
-      "completion_length": 456.0,
-      "epoch": 5.363636363636363,
-      "grad_norm": 0.427194744348526,
-      "kl": 0.0021815421558650477,
+      "completion_length": 317.5625,
+      "epoch": 3.2,
+      "grad_norm": 0.5858137011528015,
+      "kl": 0.0020121160996495746,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 1.1721123542104448,
-      "reward_std": 0.9155192162309375,
-      "rewards/concensus_correctness_reward_func": 0.0684285717351096,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "reward": 1.6474761962890625,
+      "reward_std": 0.9893582053482533,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.375,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.43671950165714535,
-      "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.38124998978206087,
+      "rewards/final_correctness_reward_func": 0.125,
+      "rewards/question_recreation_reward_func": 0.7839136756956577,
+      "rewards/soft_format_reward_func": 0.015625,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.3323125075548887,
       "step": 16
     },
     {
-      "completion_length": 458.92857142857144,
-      "epoch": 6.0,
-      "grad_norm": 0.5137476325035095,
-      "kl": 0.00196937275385218,
+      "completion_length": 286.9375,
+      "epoch": 3.6,
+      "grad_norm": 1.1306827068328857,
+      "kl": 0.002668315835762769,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 1.2816025529588972,
-      "reward_std": 1.0332180857658386,
-      "rewards/concensus_correctness_reward_func": 0.031071428741727556,
-      "rewards/consensus_reward_func": 0.35714285714285715,
+      "reward": 1.638163760304451,
+      "reward_std": 1.1374807581305504,
+      "rewards/concensus_correctness_reward_func": 0.1132500022649765,
+      "rewards/consensus_reward_func": 0.3125,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5437096644725118,
-      "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.3318214310067041,
+      "rewards/final_correctness_reward_func": 0.125,
+      "rewards/question_recreation_reward_func": 0.4999137260019779,
+      "rewards/soft_format_reward_func": 0.015625,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": 0.5406250059604645,
       "step": 18
     },
     {
-      "completion_length": 501.9375,
-      "epoch": 6.7272727272727275,
-      "grad_norm": 0.6245306730270386,
-      "kl": 0.0016426051588496193,
+      "completion_length": 352.75,
+      "epoch": 4.0,
+      "grad_norm": 0.6708921194076538,
+      "kl": 0.0019438412127783522,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 0.8476437460631132,
-      "reward_std": 0.9873409513384104,
-      "rewards/concensus_correctness_reward_func": 0.0521249994635582,
-      "rewards/consensus_reward_func": 0.1875,
+      "reward": 1.4137624818831682,
+      "reward_std": 0.9389824829995632,
+      "rewards/concensus_correctness_reward_func": 0.16868750005960464,
+      "rewards/consensus_reward_func": 0.0625,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5099562611430883,
+      "rewards/final_correctness_reward_func": 0.0625,
+      "rewards/question_recreation_reward_func": 0.5511687491089106,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.0824374994263053,
+      "rewards/xmlcount_reward_func": 0.5532812587916851,
       "step": 20
     },
     {
-      "epoch": 6.7272727272727275,
+      "epoch": 4.0,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 1.6401619859607309e-06,
-      "train_runtime": 1223.4771,
-      "train_samples_per_second": 0.262,
+      "train_loss": 2.083803985897248e-06,
+      "train_runtime": 1271.7904,
+      "train_samples_per_second": 0.252,
       "train_steps_per_second": 0.016
     }
   ],
   "logging_steps": 2,
   "max_steps": 20,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 10,
+  "num_train_epochs": 4,
   "save_steps": 25,
   "stateful_callbacks": {
     "TrainerControl": {