End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ef0c9fc2a22680be6b9b8fa9540f1875dc6c049693303b0d3390ad2c07c40f9d
+oid sha256:689db96dd937d9a9be7b853a60f125f06315cd2e043483523dd59590deab5b3d
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 6.641451051336844e-07,
-    "train_runtime": 718.6924,
-    "train_samples": 17,
-    "train_samples_per_second": 0.445,
-    "train_steps_per_second": 0.028
+    "train_loss": 6.728446294346213e-07,
+    "train_runtime": 730.6569,
+    "train_samples": 11,
+    "train_samples_per_second": 0.438,
+    "train_steps_per_second": 0.027
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 6.641451051336844e-07,
-    "train_runtime": 718.6924,
-    "train_samples": 17,
-    "train_samples_per_second": 0.445,
-    "train_steps_per_second": 0.028
+    "train_loss": 6.728446294346213e-07,
+    "train_runtime": 730.6569,
+    "train_samples": 11,
+    "train_samples_per_second": 0.438,
+    "train_steps_per_second": 0.027
 }

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": 427.875,
-      "epoch": 0.47058823529411764,
-      "grad_norm": 0.5876138806343079,
-      "kl": 0.0007815321441739798,
+      "completion_length": 538.03125,
+      "epoch": 0.7272727272727273,
+      "grad_norm": 0.4673408269882202,
+      "kl": 0.0006751879664079752,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 1.6895362362265587,
-      "reward_std": 1.607804922387004,
-      "rewards/concensus_correctness_reward_func": 0.30674999952316284,
-      "rewards/consensus_reward_func": 0.375,
+      "reward": 0.6300931870937347,
+      "reward_std": 0.7030879985541105,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.0625,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.3125,
-      "rewards/question_recreation_reward_func": 0.5297549953684211,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.4146869406104088,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.14990626135841012,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.15290625393390656,
       "step": 2
     },
     {
-      "completion_length": 340.0,
-      "epoch": 0.9411764705882353,
-      "grad_norm": 0.98947674036026,
-      "kl": 0.0008097236932371743,
+      "completion_length": 711.4642857142857,
+      "epoch": 1.3636363636363638,
+      "grad_norm": 0.3041817545890808,
+      "kl": 0.0005705223551818303,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 1.1686192080378532,
-      "reward_std": 0.8920873831957579,
-      "rewards/concensus_correctness_reward_func": 0.0078125,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 0.960112737757819,
+      "reward_std": 1.290339435849871,
+      "rewards/concensus_correctness_reward_func": 0.0447142847946712,
+      "rewards/consensus_reward_func": 0.35714285714285715,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.562056683935225,
-      "rewards/soft_format_reward_func": 0.03125,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.301875002682209,
+      "rewards/question_recreation_reward_func": 0.3419698966400964,
+      "rewards/soft_format_reward_func": 0.0,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.21628572046756744,
       "step": 4
     },
     {
-      "completion_length": 538.9,
-      "epoch": 1.2352941176470589,
-      "grad_norm": 0.518266499042511,
-      "kl": 0.0006855448475107551,
+      "completion_length": 592.1428571428571,
+      "epoch": 2.0,
+      "grad_norm": 0.321912944316864,
+      "kl": 0.0006380494700611702,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 2.2191823825240133,
-      "reward_std": 1.331932282447815,
-      "rewards/concensus_correctness_reward_func": 0.525,
-      "rewards/consensus_reward_func": 0.9,
+      "reward": 0.8947670417172569,
+      "reward_std": 1.1553706526756287,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.35714285714285715,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.3,
-      "rewards/question_recreation_reward_func": 0.5259323239326477,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.42183847086770193,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.031749993562698364,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.09792858362197876,
       "step": 6
     },
     {
-      "completion_length": 329.9375,
-      "epoch": 1.7058823529411766,
-      "grad_norm": 0.4131653308868408,
-      "kl": 0.0007813851116225123,
+      "completion_length": 634.625,
+      "epoch": 2.7272727272727275,
+      "grad_norm": 0.33803796768188477,
+      "kl": 0.0006307054318313021,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 1.6284910812973976,
-      "reward_std": 0.8130084574222565,
-      "rewards/concensus_correctness_reward_func": 0.11649999767541885,
-      "rewards/consensus_reward_func": 0.3125,
+      "reward": 0.5797345377504826,
+      "reward_std": 1.2402313686907291,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.1875,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.4375,
-      "rewards/question_recreation_reward_func": 0.4998348420485854,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.42010954208672047,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.26215624529868364,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": -0.04350000247359276,
       "step": 8
     },
     {
-      "completion_length": 403.2,
-      "epoch": 2.0,
-      "grad_norm": 0.13836334645748138,
-      "kl": 0.0006139612174592912,
+      "completion_length": 491.17857142857144,
+      "epoch": 3.3636363636363638,
+      "grad_norm": 0.42815327644348145,
+      "kl": 0.0007511325987122421,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 2.0887853801250458,
-      "reward_std": 0.6115315139293671,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 1.1,
+      "reward": 0.4952862433024815,
+      "reward_std": 0.8029870901788984,
+      "rewards/concensus_correctness_reward_func": -0.02121428506714957,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5645354151725769,
+      "rewards/question_recreation_reward_func": 0.3294648144926344,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.4242500215768814,
+      "rewards/xmlcount_reward_func": 0.18703571813447134,
       "step": 10
     },
     {
-      "completion_length": 444.5625,
-      "epoch": 2.4705882352941178,
-      "grad_norm": 0.40184539556503296,
-      "kl": 0.0006372530478984118,
+      "completion_length": 463.64285714285717,
+      "epoch": 4.0,
+      "grad_norm": 0.4289427101612091,
+      "kl": 0.0007512595122014838,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 1.8296705558896065,
-      "reward_std": 1.1184727065265179,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.875,
+      "reward": 0.8499365193503243,
+      "reward_std": 1.04093365584101,
+      "rewards/concensus_correctness_reward_func": 0.026857141937528337,
+      "rewards/consensus_reward_func": 0.21428571428571427,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0625,
-      "rewards/question_recreation_reward_func": 0.5306080486625433,
+      "rewards/final_correctness_reward_func": 0.07142857142857142,
+      "rewards/question_recreation_reward_func": 0.3951150540794645,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.36156250536441803,
+      "rewards/xmlcount_reward_func": 0.14225000249488012,
       "step": 12
     },
     {
-      "completion_length": 349.3125,
-      "epoch": 2.9411764705882355,
-      "grad_norm": 0.648057758808136,
-      "kl": 0.0008520848932676017,
+      "completion_length": 461.9375,
+      "epoch": 4.7272727272727275,
+      "grad_norm": 1.1750767230987549,
+      "kl": 0.0008141517610056326,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 1.988405354321003,
-      "reward_std": 1.3347271010279655,
-      "rewards/concensus_correctness_reward_func": 0.2265625,
-      "rewards/consensus_reward_func": 0.5,
+      "reward": 0.48488674685359,
+      "reward_std": 0.6717841476202011,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.0625,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.3125,
-      "rewards/question_recreation_reward_func": 0.5402803681790829,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.35335549898445606,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.03125,
-      "rewards/xmlcount_reward_func": 0.37781250290572643,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.0690312460064888,
       "step": 14
     },
     {
-      "completion_length": 408.5,
-      "epoch": 3.235294117647059,
-      "grad_norm": 0.6705062389373779,
-      "kl": 0.0009508899878710508,
+      "completion_length": 501.0357142857143,
+      "epoch": 5.363636363636363,
+      "grad_norm": 3.0457513332366943,
+      "kl": 0.0008896194464926209,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 1.2143768668174744,
-      "reward_std": 0.8865768074989319,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.4,
+      "reward": 1.0617176720074244,
+      "reward_std": 1.299524188041687,
+      "rewards/concensus_correctness_reward_func": 0.017857142857142856,
+      "rewards/consensus_reward_func": 0.42857142857142855,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.6206269025802612,
+      "rewards/question_recreation_reward_func": 0.43861049839428495,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.19375,
+      "rewards/xmlcount_reward_func": 0.17667856173855917,
       "step": 16
     },
     {
-      "completion_length": 320.40625,
-      "epoch": 3.7058823529411766,
-      "grad_norm": 1.4544986486434937,
-      "kl": 0.0009329580789199099,
+      "completion_length": 508.67857142857144,
+      "epoch": 6.0,
+      "grad_norm": 0.4187767803668976,
+      "kl": 0.0007303775720564383,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 2.1640911400318146,
-      "reward_std": 1.306618556380272,
-      "rewards/concensus_correctness_reward_func": 0.27806249633431435,
-      "rewards/consensus_reward_func": 0.625,
+      "reward": 0.8850597121885845,
+      "reward_std": 0.9508567367281232,
+      "rewards/concensus_correctness_reward_func": 0.017857142857142856,
+      "rewards/consensus_reward_func": 0.2857142857142857,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.3125,
-      "rewards/question_recreation_reward_func": 0.5214036731049418,
-      "rewards/soft_format_reward_func": 0.015625,
-      "rewards/strict_format_reward_func": 0.03125,
-      "rewards/xmlcount_reward_func": 0.3802500069141388,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.35448830681187765,
+      "rewards/soft_format_reward_func": 0.0,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.2091428658791951,
       "step": 18
     },
     {
-      "completion_length": 395.4,
-      "epoch": 4.0,
-      "grad_norm": 0.2986851930618286,
-      "kl": 0.0007005963008850813,
+      "completion_length": 513.375,
+      "epoch": 6.7272727272727275,
+      "grad_norm": 0.36682870984077454,
+      "kl": 0.0008761969875195064,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 1.7007000684738158,
-      "reward_std": 0.8157501250505448,
-      "rewards/concensus_correctness_reward_func": 0.028499999642372133,
-      "rewards/consensus_reward_func": 0.9,
+      "reward": 0.7504257028922439,
+      "reward_std": 0.8404918480664492,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.3125,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.2,
-      "rewards/question_recreation_reward_func": 0.4101000681519508,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.42698819376528263,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.16210000067949296,
+      "rewards/xmlcount_reward_func": 0.01093749888241291,
       "step": 20
     },
     {
-      "epoch": 4.0,
+      "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 6.641451051336844e-07,
-      "train_runtime": 718.6924,
-      "train_samples_per_second": 0.445,
-      "train_steps_per_second": 0.028
+      "train_loss": 6.728446294346213e-07,
+      "train_runtime": 730.6569,
+      "train_samples_per_second": 0.438,
+      "train_steps_per_second": 0.027
     }
   ],
   "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": {