End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8f85f337a1620968694c92e857ecb522061d8d81c326e309c462ec3c8e2895f1
+oid sha256:abc291953a1ca46c43d32d3561d030cc2a688c089e52141100f08f43d4671b18
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 1.6110261071844434e-06,
-    "train_runtime": 1261.9596,
+    "train_loss": 1.8898989651461306e-06,
+    "train_runtime": 1302.145,
     "train_samples": 11,
-    "train_samples_per_second": 0.254,
-    "train_steps_per_second": 0.016
+    "train_samples_per_second": 0.246,
+    "train_steps_per_second": 0.015
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 1.6110261071844434e-06,
-    "train_runtime": 1261.9596,
+    "train_loss": 1.8898989651461306e-06,
+    "train_runtime": 1302.145,
     "train_samples": 11,
-    "train_samples_per_second": 0.254,
-    "train_steps_per_second": 0.016
+    "train_samples_per_second": 0.246,
+    "train_steps_per_second": 0.015
 }

trainer_state.json

@@ -10,203 +10,203 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 444.15625,
+      "completion_length": 511.46875,
       "epoch": 0.7272727272727273,
-      "grad_norm": 0.4610167145729065,
-      "kl": 0.001777610603312496,
+      "grad_norm": 0.36133483052253723,
+      "kl": 0.001603923701622989,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 0.8878671452403069,
-      "reward_std": 0.906554289162159,
-      "rewards/concensus_correctness_reward_func": 0.040312498807907104,
-      "rewards/consensus_reward_func": 0.125,
+      "reward": 0.5120482891798019,
+      "reward_std": 0.8564222641289234,
+      "rewards/concensus_correctness_reward_func": 0.00787500012665987,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5808983948081732,
+      "rewards/question_recreation_reward_func": 0.49826704151928425,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.14165623672306538,
+      "rewards/xmlcount_reward_func": 0.00590624799951911,
       "step": 2
     },
     {
-      "completion_length": 492.2142857142857,
+      "completion_length": 441.5357142857143,
       "epoch": 1.3636363636363638,
-      "grad_norm": 0.4195440113544464,
-      "kl": 0.0014599822627912676,
+      "grad_norm": 0.5908653140068054,
+      "kl": 0.0015872263715469412,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 2.1185888137136186,
-      "reward_std": 2.380754053592682,
-      "rewards/concensus_correctness_reward_func": 0.7713571574006762,
+      "reward": 0.9753610406603131,
+      "reward_std": 1.0440780009542192,
+      "rewards/concensus_correctness_reward_func": 0.021071427634784153,
       "rewards/consensus_reward_func": 0.2857142857142857,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.2857142857142857,
-      "rewards/question_recreation_reward_func": 0.5301245748996735,
+      "rewards/final_correctness_reward_func": 0.07142857142857142,
+      "rewards/question_recreation_reward_func": 0.3123253264597484,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.24567856533186777,
+      "rewards/xmlcount_reward_func": 0.28482144325971603,
       "step": 4
     },
     {
-      "completion_length": 353.4642857142857,
+      "completion_length": 438.2857142857143,
       "epoch": 2.0,
-      "grad_norm": 0.49842774868011475,
-      "kl": 0.0021343231367479476,
+      "grad_norm": 0.5404853224754333,
+      "kl": 0.0018828680040314794,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 2.063119207109724,
-      "reward_std": 2.3219028881617954,
-      "rewards/concensus_correctness_reward_func": 0.7687142842582294,
-      "rewards/consensus_reward_func": 0.35714285714285715,
+      "reward": 1.176893949508667,
+      "reward_std": 1.0374990786824907,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.21428571428571427,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.07142857142857142,
-      "rewards/question_recreation_reward_func": 0.476083442568779,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.4331439584493637,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.3897500102009092,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.5116071360451835,
       "step": 6
     },
     {
-      "completion_length": 481.28125,
+      "completion_length": 497.875,
       "epoch": 2.7272727272727275,
-      "grad_norm": 0.5632760524749756,
-      "kl": 0.001908284189994447,
+      "grad_norm": 0.555487871170044,
+      "kl": 0.0018082119495375082,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 2.734431181102991,
-      "reward_std": 2.4757192246615887,
-      "rewards/concensus_correctness_reward_func": 1.2863125117437448,
-      "rewards/consensus_reward_func": 0.375,
+      "reward": 1.0806040093302727,
+      "reward_std": 1.3192631155252457,
+      "rewards/concensus_correctness_reward_func": 0.06018749997019768,
+      "rewards/consensus_reward_func": 0.4375,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.25,
-      "rewards/question_recreation_reward_func": 0.5605562962591648,
-      "rewards/soft_format_reward_func": 0.015625,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.4723540134727955,
+      "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.24693750590085983,
+      "rewards/xmlcount_reward_func": 0.1105625107884407,
       "step": 8
     },
     {
-      "completion_length": 519.1071428571429,
+      "completion_length": 475.32142857142856,
       "epoch": 3.3636363636363638,
-      "grad_norm": 0.475861132144928,
-      "kl": 0.0015336510924888508,
+      "grad_norm": 0.3771866261959076,
+      "kl": 0.0019980433280579746,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 1.3614673763513565,
-      "reward_std": 0.8124520140034812,
-      "rewards/concensus_correctness_reward_func": 0.01792857050895691,
-      "rewards/consensus_reward_func": 0.07142857142857142,
+      "reward": 0.9557304978370667,
+      "reward_std": 1.2536362324442183,
+      "rewards/concensus_correctness_reward_func": 0.06321428503308978,
+      "rewards/consensus_reward_func": 0.42857142857142855,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.2857142857142857,
-      "rewards/question_recreation_reward_func": 0.5030745204005923,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.393909110554627,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.48332143468516214,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.052178574459893365,
       "step": 10
     },
     {
-      "completion_length": 539.2142857142857,
+      "completion_length": 470.9642857142857,
       "epoch": 4.0,
-      "grad_norm": 0.47353699803352356,
-      "kl": 0.001682647121404963,
+      "grad_norm": 0.6351341009140015,
+      "kl": 0.0018639524849796934,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 0.8666789616857257,
-      "reward_std": 1.181697815656662,
-      "rewards/concensus_correctness_reward_func": 0.00021428571614835943,
-      "rewards/consensus_reward_func": 0.14285714285714285,
+      "reward": 0.788248428276607,
+      "reward_std": 0.9311599561146328,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.21428571428571427,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.4841432571411133,
+      "rewards/question_recreation_reward_func": 0.47521270385810305,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.23946427660329,
+      "rewards/xmlcount_reward_func": 0.09875001864773887,
       "step": 12
     },
     {
-      "completion_length": 488.59375,
+      "completion_length": 461.78125,
       "epoch": 4.7272727272727275,
-      "grad_norm": 0.4984962046146393,
-      "kl": 0.0016166545101441443,
+      "grad_norm": 1.127291202545166,
+      "kl": 0.002080975675198715,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 1.6124419197440147,
-      "reward_std": 1.9673981852829456,
-      "rewards/concensus_correctness_reward_func": 0.6563749983906746,
-      "rewards/consensus_reward_func": 0.0625,
+      "reward": 1.060232674703002,
+      "reward_std": 0.9556848704814911,
+      "rewards/concensus_correctness_reward_func": 0.03687499836087227,
+      "rewards/consensus_reward_func": 0.1875,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.1875,
-      "rewards/question_recreation_reward_func": 0.40244188345968723,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.43473270907998085,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.2880000047152862,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": 0.3698749952018261,
       "step": 14
     },
     {
-      "completion_length": 447.67857142857144,
+      "completion_length": 438.0,
       "epoch": 5.363636363636363,
-      "grad_norm": 0.5061261653900146,
-      "kl": 0.0017929897377533571,
+      "grad_norm": 0.6246246695518494,
+      "kl": 0.00327364047656634,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 1.1384226637227195,
-      "reward_std": 0.9844975428921836,
-      "rewards/concensus_correctness_reward_func": 0.0337142859186445,
+      "reward": 0.6977638048785073,
+      "reward_std": 0.9071963259152004,
+      "rewards/concensus_correctness_reward_func": 0.017857142857142856,
       "rewards/consensus_reward_func": 0.21428571428571427,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5523512001548495,
+      "rewards/question_recreation_reward_func": 0.3870138313089098,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.3380714314324515,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.06074999272823334,
       "step": 16
     },
     {
-      "completion_length": 522.5,
+      "completion_length": 523.8214285714286,
       "epoch": 6.0,
-      "grad_norm": 0.611285388469696,
-      "kl": 0.0016157111302683397,
+      "grad_norm": 0.5964250564575195,
+      "kl": 0.0018691010773181915,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 1.05314644745418,
-      "reward_std": 0.7948270993573325,
-      "rewards/concensus_correctness_reward_func": 0.01792857050895691,
-      "rewards/consensus_reward_func": 0.07142857142857142,
+      "reward": 0.9746680898325784,
+      "reward_std": 1.0538555532693863,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.2857142857142857,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.07142857142857142,
-      "rewards/question_recreation_reward_func": 0.4600749909877777,
-      "rewards/soft_format_reward_func": 0.017857142857142856,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5008823488439832,
+      "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.4144285704408373,
+      "rewards/xmlcount_reward_func": 0.18807143824441092,
       "step": 18
     },
     {
-      "completion_length": 488.09375,
+      "completion_length": 451.21875,
       "epoch": 6.7272727272727275,
-      "grad_norm": 0.5456557869911194,
-      "kl": 0.0017919444653671235,
+      "grad_norm": 0.5534034371376038,
+      "kl": 0.0024654374283272773,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 1.3167592827230692,
-      "reward_std": 0.8622551411390305,
-      "rewards/concensus_correctness_reward_func": 0.015687499195337296,
-      "rewards/consensus_reward_func": 0.1875,
+      "reward": 1.1108720749616623,
+      "reward_std": 1.2174674794077873,
+      "rewards/concensus_correctness_reward_func": 0.05249999836087227,
+      "rewards/consensus_reward_func": 0.3125,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.25,
-      "rewards/question_recreation_reward_func": 0.5698843225836754,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.46424705907702446,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.2780624986626208,
+      "rewards/xmlcount_reward_func": 0.26600001007318497,
       "step": 20
     },
     {
       "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 1.6110261071844434e-06,
-      "train_runtime": 1261.9596,
-      "train_samples_per_second": 0.254,
-      "train_steps_per_second": 0.016
+      "train_loss": 1.8898989651461306e-06,
+      "train_runtime": 1302.145,
+      "train_samples_per_second": 0.246,
+      "train_steps_per_second": 0.015
     }
   ],
   "logging_steps": 2,