End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:add0d70ea78084f7e522da6b7059d2467c69ade7c790bc93398b051dc18c908a
+oid sha256:619030f187f1a7ed66ae988b67627c6461f61d9ad09ec49864892babd1bae79f
 size 35237104

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 4.873177704212139e-06,
-    "train_runtime": 673.5653,
-    "train_samples": 14,
-    "train_samples_per_second": 0.475,
-    "train_steps_per_second": 0.03
+    "train_loss": 2.9364487659222505e-06,
+    "train_runtime": 628.8319,
+    "train_samples": 44,
+    "train_samples_per_second": 0.509,
+    "train_steps_per_second": 0.032
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 4.873177704212139e-06,
-    "train_runtime": 673.5653,
-    "train_samples": 14,
-    "train_samples_per_second": 0.475,
-    "train_steps_per_second": 0.03
+    "train_loss": 2.9364487659222505e-06,
+    "train_runtime": 628.8319,
+    "train_samples": 44,
+    "train_samples_per_second": 0.509,
+    "train_steps_per_second": 0.032
 }

trainer_state.json

@@ -1,7 +1,7 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 5.0,
+  "epoch": 1.8181818181818183,
   "eval_steps": 500,
   "global_step": 20,
   "is_hyper_param_search": false,
@@ -9,209 +9,209 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 319.21875,
-      "epoch": 0.5714285714285714,
-      "grad_norm": 50.99075698852539,
-      "kl": 0.010051651486719493,
+      "completion_length": 372.25,
+      "epoch": 0.18181818181818182,
+      "grad_norm": 1.9300328493118286,
+      "kl": 0.002685736268176697,
       "learning_rate": 4.965903258506806e-07,
       "loss": 0.0,
-      "reward": 0.1684375638142228,
-      "reward_std": 0.36661340156570077,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.0,
+      "reward": 1.9535507448017597,
+      "reward_std": 2.3369685858488083,
+      "rewards/concensus_correctness_reward_func": 0.75,
+      "rewards/consensus_reward_func": 0.0625,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.11190631706267595,
+      "rewards/final_correctness_reward_func": 0.375,
+      "rewards/question_recreation_reward_func": 0.4927695244550705,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.05653125233948231,
+      "rewards/xmlcount_reward_func": 0.27328124828636646,
       "step": 2
     },
     {
-      "completion_length": 437.75,
-      "epoch": 1.0,
-      "grad_norm": 0.9677923321723938,
-      "kl": 0.005021085050733139,
+      "completion_length": 343.03125,
+      "epoch": 0.36363636363636365,
+      "grad_norm": 3.323760747909546,
+      "kl": 0.0038706637133145705,
       "learning_rate": 4.698684378016222e-07,
       "loss": 0.0,
-      "reward": 1.2924380178252857,
-      "reward_std": 2.3202759933968387,
-      "rewards/concensus_correctness_reward_func": 0.8530833336214224,
-      "rewards/consensus_reward_func": 0.08333333333333333,
+      "reward": 1.1179207004606724,
+      "reward_std": 0.9569476544857025,
+      "rewards/concensus_correctness_reward_func": 0.09912500157952309,
+      "rewards/consensus_reward_func": 0.125,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.33897967388232547,
+      "rewards/final_correctness_reward_func": 0.25,
+      "rewards/question_recreation_reward_func": 0.36013949010521173,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.017041688164075215,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.2680312469601631,
       "step": 4
     },
     {
-      "completion_length": 274.5,
-      "epoch": 1.5714285714285714,
-      "grad_norm": 2.3173298835754395,
-      "kl": 0.009283999257604592,
+      "completion_length": 275.65625,
+      "epoch": 0.5454545454545454,
+      "grad_norm": 1.9294134378433228,
+      "kl": 0.0021318396175047383,
       "learning_rate": 4.193203929064353e-07,
       "loss": 0.0,
-      "reward": 0.38673332217149436,
-      "reward_std": 0.6992381454911083,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.0,
+      "reward": 1.1408963352441788,
+      "reward_std": 0.967832338064909,
+      "rewards/concensus_correctness_reward_func": 0.125,
+      "rewards/consensus_reward_func": 0.125,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.26610831159632653,
+      "rewards/final_correctness_reward_func": 0.0625,
+      "rewards/question_recreation_reward_func": 0.4269276261329651,
       "rewards/soft_format_reward_func": 0.015625,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.10499999672174454,
+      "rewards/xmlcount_reward_func": 0.3858437556773424,
       "step": 6
     },
     {
-      "completion_length": 444.125,
-      "epoch": 2.0,
-      "grad_norm": 0.7961806654930115,
-      "kl": 0.003991736409564813,
+      "completion_length": 319.96875,
+      "epoch": 0.7272727272727273,
+      "grad_norm": 1.2534875869750977,
+      "kl": 0.0027487699990160763,
       "learning_rate": 3.5042385616324236e-07,
       "loss": 0.0,
-      "reward": 1.1846553509434063,
-      "reward_std": 2.380153588950634,
-      "rewards/concensus_correctness_reward_func": 0.8333333333333334,
-      "rewards/consensus_reward_func": 0.0,
+      "reward": 0.9228200912475586,
+      "reward_std": 1.5564873665571213,
+      "rewards/concensus_correctness_reward_func": 0.125,
+      "rewards/consensus_reward_func": 0.1875,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.16666666666666666,
-      "rewards/question_recreation_reward_func": 0.2876970109840234,
+      "rewards/final_correctness_reward_func": 0.125,
+      "rewards/question_recreation_reward_func": 0.3794763386249542,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.10304167121648788,
+      "rewards/xmlcount_reward_func": 0.10584374889731407,
       "step": 8
     },
     {
-      "completion_length": 479.03125,
-      "epoch": 2.571428571428571,
-      "grad_norm": 1.8381205797195435,
-      "kl": 0.0032171533894143067,
+      "completion_length": 435.375,
+      "epoch": 0.9090909090909091,
+      "grad_norm": 1.2784631252288818,
+      "kl": 0.002754525456111878,
       "learning_rate": 2.706448363680831e-07,
       "loss": 0.0,
-      "reward": 0.36011022701859474,
-      "reward_std": 0.7439824137836695,
-      "rewards/concensus_correctness_reward_func": 0.0533749982714653,
-      "rewards/consensus_reward_func": 0.0,
+      "reward": 0.8444329872727394,
+      "reward_std": 1.0790864313021302,
+      "rewards/concensus_correctness_reward_func": 0.125,
+      "rewards/consensus_reward_func": 0.125,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.1875,
-      "rewards/question_recreation_reward_func": 0.33779770066030324,
-      "rewards/soft_format_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.28580794762820005,
+      "rewards/soft_format_reward_func": 0.015625,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.21856248658150434,
+      "rewards/xmlcount_reward_func": 0.1055000051856041,
       "step": 10
     },
     {
-      "completion_length": 425.0416666666667,
-      "epoch": 3.0,
-      "grad_norm": 0.9481332898139954,
-      "kl": 0.002326692493321995,
+      "completion_length": 383.21875,
+      "epoch": 1.0909090909090908,
+      "grad_norm": 3.725076913833618,
+      "kl": 0.0038445138634415343,
       "learning_rate": 1.886286282148002e-07,
       "loss": 0.0,
-      "reward": 0.49296411996086437,
-      "reward_std": 0.8617748295267423,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.0,
+      "reward": 1.3737196251749992,
+      "reward_std": 1.564310446381569,
+      "rewards/concensus_correctness_reward_func": 0.24268750101327896,
+      "rewards/consensus_reward_func": 0.375,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.08333333333333333,
-      "rewards/question_recreation_reward_func": 0.3043391207853953,
+      "rewards/final_correctness_reward_func": 0.125,
+      "rewards/question_recreation_reward_func": 0.37809463776648045,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.10529166584213574,
+      "rewards/xmlcount_reward_func": 0.2529374994337559,
       "step": 12
     },
     {
-      "completion_length": 352.90625,
-      "epoch": 3.571428571428571,
-      "grad_norm": 18.12985610961914,
-      "kl": 0.0036522214286378585,
+      "completion_length": 408.4375,
+      "epoch": 1.2727272727272727,
+      "grad_norm": 3.1979212760925293,
+      "kl": 0.002077052035019733,
       "learning_rate": 1.1326296046939333e-07,
       "loss": 0.0,
-      "reward": 0.9987665724474937,
-      "reward_std": 1.963424475863576,
-      "rewards/concensus_correctness_reward_func": 0.625,
-      "rewards/consensus_reward_func": 0.0,
+      "reward": 4.569394081830978,
+      "reward_std": 5.111470676958561,
+      "rewards/concensus_correctness_reward_func": 3.208000000566244,
+      "rewards/consensus_reward_func": 0.3125,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.125,
-      "rewards/question_recreation_reward_func": 0.2873290863353759,
-      "rewards/soft_format_reward_func": 0.0,
+      "rewards/final_correctness_reward_func": 0.375,
+      "rewards/question_recreation_reward_func": 0.46067542023956776,
+      "rewards/soft_format_reward_func": 0.015625,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.03856250084936619,
+      "rewards/xmlcount_reward_func": 0.19759375415742397,
       "step": 14
     },
     {
-      "completion_length": 406.1666666666667,
-      "epoch": 4.0,
-      "grad_norm": 67.13907623291016,
-      "kl": 0.010182096312443415,
+      "completion_length": 436.34375,
+      "epoch": 1.4545454545454546,
+      "grad_norm": 1.4435375928878784,
+      "kl": 0.0014565634628525004,
       "learning_rate": 5.271487265090163e-08,
       "loss": 0.0,
-      "reward": 0.15596704526493946,
-      "reward_std": 0.725780225203683,
-      "rewards/concensus_correctness_reward_func": 0.010416666666666666,
+      "reward": 1.052122674882412,
+      "reward_std": 0.8749217577278614,
+      "rewards/concensus_correctness_reward_func": 0.03125,
       "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.19955036928877234,
+      "rewards/final_correctness_reward_func": 0.0625,
+      "rewards/question_recreation_reward_func": 0.4829976772889495,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.05400000140070915,
+      "rewards/xmlcount_reward_func": 0.4753750041127205,
       "step": 16
     },
     {
-      "completion_length": 361.0,
-      "epoch": 4.571428571428571,
-      "grad_norm": 3.9042296409606934,
-      "kl": 0.003631016006693244,
+      "completion_length": 434.21875,
+      "epoch": 1.6363636363636362,
+      "grad_norm": 1.8603307008743286,
+      "kl": 0.004560476292681415,
       "learning_rate": 1.3545689574841341e-08,
       "loss": 0.0,
-      "reward": 0.6882577603682876,
-      "reward_std": 0.9258282171795145,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.0,
+      "reward": 1.5926124211400747,
+      "reward_std": 2.343985839514062,
+      "rewards/concensus_correctness_reward_func": 0.65625,
+      "rewards/consensus_reward_func": 0.125,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.1875,
-      "rewards/question_recreation_reward_func": 0.40460149571299553,
+      "rewards/question_recreation_reward_func": 0.3674874210264534,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.03125,
-      "rewards/xmlcount_reward_func": 0.06490625068545341,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.2407499998807907,
       "step": 18
     },
     {
-      "completion_length": 458.125,
-      "epoch": 5.0,
-      "grad_norm": 1.765657901763916,
-      "kl": 0.003590759743625919,
+      "completion_length": 310.65625,
+      "epoch": 1.8181818181818183,
+      "grad_norm": 1.8424596786499023,
+      "kl": 0.0032100926473503932,
       "learning_rate": 0.0,
       "loss": 0.0,
-      "reward": 0.02644654798010985,
-      "reward_std": 0.7460031639784575,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.0,
+      "reward": 0.7539666602388024,
+      "reward_std": 0.9992944076657295,
+      "rewards/concensus_correctness_reward_func": 0.03506249934434891,
+      "rewards/consensus_reward_func": 0.125,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.08333333333333333,
-      "rewards/question_recreation_reward_func": 0.16532153760393462,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.31849788781255484,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.22220832606156668,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.25978125259280205,
       "step": 20
     },
     {
-      "epoch": 5.0,
+      "epoch": 1.8181818181818183,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 4.873177704212139e-06,
-      "train_runtime": 673.5653,
-      "train_samples_per_second": 0.475,
-      "train_steps_per_second": 0.03
+      "train_loss": 2.9364487659222505e-06,
+      "train_runtime": 628.8319,
+      "train_samples_per_second": 0.509,
+      "train_steps_per_second": 0.032
     }
   ],
   "logging_steps": 2,
   "max_steps": 20,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 7,
+  "num_train_epochs": 2,
   "save_steps": 25,
   "stateful_callbacks": {
     "TrainerControl": {