End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:889f45978a06f5e80f03d967759ea94725b29262665b6ffce37ea1e7701dc642
+oid sha256:e0ad4988bbd0f213db4e3f22376d40e86d7a83e0c1f15cc1769d34e37ccd2426
 size 35237104

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 2.8422014338502776e-06,
-    "train_runtime": 747.7694,
-    "train_samples": 18,
-    "train_samples_per_second": 0.428,
-    "train_steps_per_second": 0.027
+    "train_loss": 2.944767936696735e-06,
+    "train_runtime": 330.048,
+    "train_samples": 11,
+    "train_samples_per_second": 0.97,
+    "train_steps_per_second": 0.061
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 2.8422014338502776e-06,
-    "train_runtime": 747.7694,
-    "train_samples": 18,
-    "train_samples_per_second": 0.428,
-    "train_steps_per_second": 0.027
+    "train_loss": 2.944767936696735e-06,
+    "train_runtime": 330.048,
+    "train_samples": 11,
+    "train_samples_per_second": 0.97,
+    "train_steps_per_second": 0.061
 }

trainer_state.json

@@ -2,7 +2,7 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 4.0,
+  "epoch": 6.7272727272727275,
   "eval_steps": 500,
   "global_step": 20,
   "is_hyper_param_search": false,
@@ -10,209 +10,209 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 236.0,
-      "epoch": 0.4444444444444444,
-      "grad_norm": 10.439835548400879,
-      "kl": 0.004431007546372712,
+      "completion_length": 347.21875,
+      "epoch": 0.7272727272727273,
+      "grad_norm": 3.9205262660980225,
+      "kl": 0.0013051503119640984,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 0.25179125973954797,
-      "reward_std": 0.2632170447614044,
+      "reward": 0.11110634659416974,
+      "reward_std": 0.3795590694062412,
       "rewards/concensus_correctness_reward_func": 0.0,
       "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.1523850136436522,
+      "rewards/question_recreation_reward_func": 0.13291884120553732,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.09940624982118607,
+      "rewards/xmlcount_reward_func": -0.021812498569488525,
       "step": 2
     },
     {
-      "completion_length": 201.125,
-      "epoch": 0.8888888888888888,
-      "grad_norm": 19.111417770385742,
-      "kl": 0.003833223097899463,
+      "completion_length": 310.14285714285717,
+      "epoch": 1.3636363636363638,
+      "grad_norm": 35.81222152709961,
+      "kl": 0.003564855191923146,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 0.10808797710342333,
-      "reward_std": 0.10177052326616831,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.0,
+      "reward": 0.33923356447901043,
+      "reward_std": 0.5153605126376662,
+      "rewards/concensus_correctness_reward_func": 0.008928571428571428,
+      "rewards/consensus_reward_func": 0.07142857142857142,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.0700567233725451,
+      "rewards/question_recreation_reward_func": 0.135019274694579,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.03803124977275729,
+      "rewards/xmlcount_reward_func": 0.1238571406075997,
       "step": 4
     },
     {
-      "completion_length": 355.9166666666667,
-      "epoch": 1.2222222222222223,
-      "grad_norm": 36.298274993896484,
-      "kl": 0.003768470574868843,
+      "completion_length": 499.42857142857144,
+      "epoch": 2.0,
+      "grad_norm": 1.1778466701507568,
+      "kl": 0.0012288401443843863,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 0.29789741709828377,
-      "reward_std": 0.51212003454566,
+      "reward": 0.2827788378511156,
+      "reward_std": 0.5379786475428513,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.16666666666666666,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.11677242318789165,
+      "rewards/question_recreation_reward_func": 0.20110026374459267,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.014458333452542623,
+      "rewards/xmlcount_reward_func": 0.08167856931686401,
       "step": 6
     },
     {
-      "completion_length": 259.03125,
-      "epoch": 1.6666666666666665,
-      "grad_norm": 32.289794921875,
-      "kl": 0.004386888962471858,
+      "completion_length": 417.8125,
+      "epoch": 2.7272727272727275,
+      "grad_norm": 85.56259155273438,
+      "kl": 0.012747703374770936,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 0.10565200773999095,
-      "reward_std": 0.27231678273528814,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.0,
+      "reward": 0.3929629949852824,
+      "reward_std": 0.3982247731764801,
+      "rewards/concensus_correctness_reward_func": 0.0078125,
+      "rewards/consensus_reward_func": 0.0625,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.13174574682489038,
+      "rewards/question_recreation_reward_func": 0.17696297960355878,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.026093745604157448,
+      "rewards/xmlcount_reward_func": 0.1456875056028366,
       "step": 8
     },
     {
-      "completion_length": 304.1666666666667,
-      "epoch": 2.0,
-      "grad_norm": 24.74953842163086,
-      "kl": 0.0017361478239763528,
+      "completion_length": 433.82142857142856,
+      "epoch": 3.3636363636363638,
+      "grad_norm": 2.4649531841278076,
+      "kl": 0.0011016575237070875,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 0.09029453371961911,
-      "reward_std": 0.0767441635640959,
+      "reward": 0.23983536288142204,
+      "reward_std": 0.26847671078784124,
       "rewards/concensus_correctness_reward_func": 0.0,
       "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.09512786411990722,
+      "rewards/question_recreation_reward_func": 0.1622282170823642,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.004833333194255829,
+      "rewards/xmlcount_reward_func": 0.07760714207376752,
       "step": 10
     },
     {
-      "completion_length": 247.96875,
-      "epoch": 2.4444444444444446,
-      "grad_norm": 12.753266334533691,
-      "kl": 0.0014171793518471532,
+      "completion_length": 341.5,
+      "epoch": 4.0,
+      "grad_norm": 1.8753986358642578,
+      "kl": 0.0011920631181315652,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 0.1847613099962473,
-      "reward_std": 0.12021742109209299,
+      "reward": 0.2205039480967181,
+      "reward_std": 0.32309810045574394,
       "rewards/concensus_correctness_reward_func": 0.0,
       "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.15773006528615952,
+      "rewards/question_recreation_reward_func": 0.13439680955239705,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.027031248901039362,
+      "rewards/xmlcount_reward_func": 0.08610714333398002,
       "step": 12
     },
     {
-      "completion_length": 258.3125,
-      "epoch": 2.888888888888889,
-      "grad_norm": 31.789426803588867,
-      "kl": 0.0035024683456867933,
+      "completion_length": 439.0625,
+      "epoch": 4.7272727272727275,
+      "grad_norm": 65.4488754272461,
+      "kl": 0.00342925704899244,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 0.25030552195676137,
-      "reward_std": 0.39126282584038563,
+      "reward": 0.3297867029905319,
+      "reward_std": 0.38429783983156085,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.0625,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0625,
-      "rewards/question_recreation_reward_func": 0.11771176448382903,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.13841170771047473,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.07009374909102917,
+      "rewards/xmlcount_reward_func": 0.1288749990053475,
       "step": 14
     },
     {
-      "completion_length": 248.45833333333334,
-      "epoch": 3.2222222222222223,
-      "grad_norm": 13.452702522277832,
-      "kl": 0.0031033945463908217,
+      "completion_length": 269.82142857142856,
+      "epoch": 5.363636363636363,
+      "grad_norm": 14.348326683044434,
+      "kl": 0.0022195359592192937,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 0.16759652675439915,
-      "reward_std": 0.21643815096467733,
+      "reward": 0.18191001085298403,
+      "reward_std": 0.2123505531677178,
       "rewards/concensus_correctness_reward_func": 0.0,
       "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.07597152112672727,
+      "rewards/question_recreation_reward_func": 0.10833857900329999,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.09162500003973643,
+      "rewards/xmlcount_reward_func": 0.07357142652784075,
       "step": 16
     },
     {
-      "completion_length": 312.78125,
-      "epoch": 3.6666666666666665,
-      "grad_norm": 12.385773658752441,
-      "kl": 0.001501579157775268,
+      "completion_length": 391.42857142857144,
+      "epoch": 6.0,
+      "grad_norm": 2.8804738521575928,
+      "kl": 0.002775567941301103,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 0.09900575468782336,
-      "reward_std": 0.49123311004950665,
+      "reward": 0.1302676254085132,
+      "reward_std": 0.11047785675951413,
       "rewards/concensus_correctness_reward_func": 0.0,
       "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.17381824657786638,
+      "rewards/question_recreation_reward_func": 0.12073191388377122,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.07481250632554293,
+      "rewards/xmlcount_reward_func": 0.009535713919571467,
       "step": 18
     },
     {
-      "completion_length": 248.95833333333334,
-      "epoch": 4.0,
-      "grad_norm": 4.445725440979004,
-      "kl": 0.0037293038330972195,
+      "completion_length": 350.5,
+      "epoch": 6.7272727272727275,
+      "grad_norm": 27.853628158569336,
+      "kl": 0.0014230193482944742,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 0.15316738188266754,
-      "reward_std": 0.12911921491225561,
+      "reward": 0.1323570990934968,
+      "reward_std": 0.3228283408097923,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.0625,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.1028340496122837,
+      "rewards/question_recreation_reward_func": 0.1812945973360911,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.05033333351214727,
+      "rewards/xmlcount_reward_func": -0.11143749579787254,
       "step": 20
     },
     {
-      "epoch": 4.0,
+      "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 2.8422014338502776e-06,
-      "train_runtime": 747.7694,
-      "train_samples_per_second": 0.428,
-      "train_steps_per_second": 0.027
+      "train_loss": 2.944767936696735e-06,
+      "train_runtime": 330.048,
+      "train_samples_per_second": 0.97,
+      "train_steps_per_second": 0.061
     }
   ],
   "logging_steps": 2,
   "max_steps": 20,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 5,
+  "num_train_epochs": 10,
   "save_steps": 25,
   "stateful_callbacks": {
     "TrainerControl": {