End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:afa5bda84ec50d2ca44812b3a98b368965ce8979d131b8fb84c91fda704667b5
+oid sha256:7fcd3b9b1c9c9768d486b7760202b5bbd038d77de16f063a8908737019a83e2f
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 1.3469746477312584e-06,
-    "train_runtime": 606.9964,
+    "train_loss": 1.4845098007754133e-06,
+    "train_runtime": 531.6911,
     "train_samples": 11,
-    "train_samples_per_second": 0.527,
-    "train_steps_per_second": 0.033
+    "train_samples_per_second": 0.602,
+    "train_steps_per_second": 0.038
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 1.3469746477312584e-06,
-    "train_runtime": 606.9964,
+    "train_loss": 1.4845098007754133e-06,
+    "train_runtime": 531.6911,
     "train_samples": 11,
-    "train_samples_per_second": 0.527,
-    "train_steps_per_second": 0.033
+    "train_samples_per_second": 0.602,
+    "train_steps_per_second": 0.038
 }

trainer_state.json

@@ -10,203 +10,203 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 414.25,
+      "completion_length": 364.9375,
       "epoch": 0.7272727272727273,
-      "grad_norm": 0.49510616064071655,
-      "kl": 0.0014182917075231671,
+      "grad_norm": 0.4663001000881195,
+      "kl": 0.0017626049666432664,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 0.6955308392643929,
-      "reward_std": 0.7927929870784283,
-      "rewards/concensus_correctness_reward_func": 0.0,
+      "reward": 1.173146827146411,
+      "reward_std": 0.9561699032783508,
+      "rewards/concensus_correctness_reward_func": 0.09787499904632568,
       "rewards/consensus_reward_func": 0.25,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.4437808468937874,
+      "rewards/final_correctness_reward_func": 0.25,
+      "rewards/question_recreation_reward_func": 0.37455306435003877,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.0017500128597021103,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": 0.1694687455892563,
       "step": 2
     },
     {
-      "completion_length": 573.5714285714286,
+      "completion_length": 498.75,
       "epoch": 1.3636363636363638,
-      "grad_norm": 1.2850693464279175,
-      "kl": 0.0013809393193306668,
+      "grad_norm": 0.39783263206481934,
+      "kl": 0.001358998814664249,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 0.8315435009343284,
-      "reward_std": 0.9889501077788216,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.35714285714285715,
+      "reward": 1.123238938195365,
+      "reward_std": 1.1782569800104414,
+      "rewards/concensus_correctness_reward_func": 0.03664285796029227,
+      "rewards/consensus_reward_func": 0.5,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5804363297564643,
+      "rewards/question_recreation_reward_func": 0.5083817797047752,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": -0.12389285436698369,
+      "rewards/xmlcount_reward_func": 0.06035715075475829,
       "step": 4
     },
     {
-      "completion_length": 485.4642857142857,
+      "completion_length": 409.7857142857143,
       "epoch": 2.0,
-      "grad_norm": 0.3890193700790405,
-      "kl": 0.0013459432188288442,
+      "grad_norm": 0.4471225440502167,
+      "kl": 0.0017250077549501189,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 1.2041999101638794,
-      "reward_std": 1.2372096351214819,
-      "rewards/concensus_correctness_reward_func": 0.0,
+      "reward": 3.185895102364676,
+      "reward_std": 2.5625331572123935,
+      "rewards/concensus_correctness_reward_func": 1.423785714166505,
       "rewards/consensus_reward_func": 0.2857142857142857,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.41419985571077894,
+      "rewards/final_correctness_reward_func": 0.35714285714285715,
+      "rewards/question_recreation_reward_func": 0.5010379723140171,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.5042857144560132,
+      "rewards/strict_format_reward_func": 0.03571428571428571,
+      "rewards/xmlcount_reward_func": 0.5825000064713615,
       "step": 6
     },
     {
-      "completion_length": 440.84375,
+      "completion_length": 453.375,
       "epoch": 2.7272727272727275,
-      "grad_norm": 0.4658561944961548,
-      "kl": 0.0015507572607020847,
+      "grad_norm": 0.4004893898963928,
+      "kl": 0.0012358660096651874,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 1.0474483966827393,
-      "reward_std": 0.9702736120671034,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.375,
+      "reward": 1.0723718777298927,
+      "reward_std": 1.1871207617223263,
+      "rewards/concensus_correctness_reward_func": -0.004187500104308128,
+      "rewards/consensus_reward_func": 0.4375,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.47438590601086617,
-      "rewards/soft_format_reward_func": 0.03125,
+      "rewards/final_correctness_reward_func": 0.125,
+      "rewards/question_recreation_reward_func": 0.531996876001358,
+      "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.16681249998509884,
+      "rewards/xmlcount_reward_func": -0.017937500029802322,
       "step": 8
     },
     {
-      "completion_length": 507.32142857142856,
+      "completion_length": 419.64285714285717,
       "epoch": 3.3636363636363638,
-      "grad_norm": 0.4284559488296509,
-      "kl": 0.0012049856512541218,
+      "grad_norm": 0.5154035687446594,
+      "kl": 0.0019263403207462812,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 0.7789082548448018,
-      "reward_std": 0.8491070440837315,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.14285714285714285,
+      "reward": 1.688667967915535,
+      "reward_std": 0.8806300418717521,
+      "rewards/concensus_correctness_reward_func": 0.06171428731509617,
+      "rewards/consensus_reward_func": 0.2857142857142857,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.4714439809322357,
+      "rewards/final_correctness_reward_func": 0.5714285714285714,
+      "rewards/question_recreation_reward_func": 0.5533822476863861,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.16460714595658438,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.19857142014162882,
       "step": 10
     },
     {
-      "completion_length": 361.7857142857143,
+      "completion_length": 511.57142857142856,
       "epoch": 4.0,
-      "grad_norm": 0.33758416771888733,
-      "kl": 0.0019130518360595619,
+      "grad_norm": 0.27420559525489807,
+      "kl": 0.0013052445137873292,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 1.1142860225268774,
-      "reward_std": 1.1628771935190474,
+      "reward": 0.343838948224272,
+      "reward_std": 1.0978365540504456,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.42857142857142855,
+      "rewards/consensus_reward_func": 0.07142857142857142,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.07142857142857142,
-      "rewards/question_recreation_reward_func": 0.4090003105146544,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.46316037114177433,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.05357142857142857,
-      "rewards/xmlcount_reward_func": 0.15171429302011216,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": -0.19074999221733638,
       "step": 12
     },
     {
-      "completion_length": 431.96875,
+      "completion_length": 344.1875,
       "epoch": 4.7272727272727275,
-      "grad_norm": 0.8539683818817139,
-      "kl": 0.0014734029246028513,
+      "grad_norm": 0.5283384919166565,
+      "kl": 0.0016884810465853661,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 0.9644927456974983,
-      "reward_std": 1.040159910917282,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 3.800176238641143,
+      "reward_std": 3.459130048751831,
+      "rewards/concensus_correctness_reward_func": 2.5270000007003546,
+      "rewards/consensus_reward_func": 0.4375,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.46711775101721287,
+      "rewards/final_correctness_reward_func": 0.1875,
+      "rewards/question_recreation_reward_func": 0.5017388071864843,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.23174998909235,
+      "rewards/xmlcount_reward_func": 0.13081249222159386,
       "step": 14
     },
     {
-      "completion_length": 378.82142857142856,
+      "completion_length": 438.60714285714283,
       "epoch": 5.363636363636363,
-      "grad_norm": 0.5045291781425476,
-      "kl": 0.0014974439150786825,
+      "grad_norm": 0.4809930920600891,
+      "kl": 0.0013627653658789182,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 0.830702930688858,
-      "reward_std": 1.0524823750768388,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "reward": 1.5581893447254385,
+      "reward_std": 2.1124278477260043,
+      "rewards/concensus_correctness_reward_func": 0.8574285901018551,
+      "rewards/consensus_reward_func": 0.14285714285714285,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.3943457656673023,
+      "rewards/final_correctness_reward_func": 0.07142857142857142,
+      "rewards/question_recreation_reward_func": 0.3227250405720302,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.1506428654704775,
+      "rewards/xmlcount_reward_func": 0.16374999284744263,
       "step": 16
     },
     {
-      "completion_length": 500.17857142857144,
+      "completion_length": 492.67857142857144,
       "epoch": 6.0,
-      "grad_norm": 0.324450820684433,
-      "kl": 0.001282154640648514,
+      "grad_norm": 0.4203881025314331,
+      "kl": 0.0020158881255026373,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 0.8862546469484057,
-      "reward_std": 0.8207760836396899,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.14285714285714285,
+      "reward": 2.326434467520033,
+      "reward_std": 2.6343056857585907,
+      "rewards/concensus_correctness_reward_func": 0.8878571433680398,
+      "rewards/consensus_reward_func": 0.35714285714285715,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5371832198330334,
+      "rewards/final_correctness_reward_func": 0.2857142857142857,
+      "rewards/question_recreation_reward_func": 0.6165773357663836,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.03571428571428571,
-      "rewards/xmlcount_reward_func": 0.17049999109336308,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.17914285830089025,
       "step": 18
     },
     {
-      "completion_length": 502.4375,
+      "completion_length": 381.59375,
       "epoch": 6.7272727272727275,
-      "grad_norm": 0.40171298384666443,
-      "kl": 0.0014211268862709403,
+      "grad_norm": 0.49792689085006714,
+      "kl": 0.0017203347160830162,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 0.9943891875445843,
-      "reward_std": 1.0806648414582014,
-      "rewards/concensus_correctness_reward_func": 0.0,
+      "reward": 1.8499253150075674,
+      "reward_std": 2.047252554446459,
+      "rewards/concensus_correctness_reward_func": 0.7549374997615814,
       "rewards/consensus_reward_func": 0.25,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5157329179346561,
+      "rewards/final_correctness_reward_func": 0.125,
+      "rewards/question_recreation_reward_func": 0.41011281311511993,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.21303124958649278,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.3098749965429306,
       "step": 20
     },
     {
       "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 1.3469746477312584e-06,
-      "train_runtime": 606.9964,
-      "train_samples_per_second": 0.527,
-      "train_steps_per_second": 0.033
+      "train_loss": 1.4845098007754133e-06,
+      "train_runtime": 531.6911,
+      "train_samples_per_second": 0.602,
+      "train_steps_per_second": 0.038
     }
   ],
   "logging_steps": 2,