End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cd98e00b0eb5f1b79ef312cd4553a048ec68874e8fb6ac1f417df77ac2a102c8
+oid sha256:ff0fd43627d1cb5e51232d3ab4599c3caa848f9e95e32e15886f5685fb223110
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 6.243158054530796e-07,
-    "train_runtime": 662.6617,
+    "train_loss": 6.091722354995e-07,
+    "train_runtime": 676.4725,
     "train_samples": 11,
-    "train_samples_per_second": 0.483,
+    "train_samples_per_second": 0.473,
     "train_steps_per_second": 0.03
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 6.243158054530796e-07,
-    "train_runtime": 662.6617,
+    "train_loss": 6.091722354995e-07,
+    "train_runtime": 676.4725,
     "train_samples": 11,
-    "train_samples_per_second": 0.483,
+    "train_samples_per_second": 0.473,
     "train_steps_per_second": 0.03
 }

trainer_state.json

@@ -10,202 +10,202 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 518.59375,
+      "completion_length": 574.125,
       "epoch": 0.7272727272727273,
-      "grad_norm": 0.3101458251476288,
-      "kl": 0.0006586692034034058,
+      "grad_norm": 0.2845724821090698,
+      "kl": 0.0005749856500187889,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 1.5320980101823807,
-      "reward_std": 1.391620872542262,
+      "reward": 0.7415162026882172,
+      "reward_std": 0.8856677860021591,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.5,
+      "rewards/consensus_reward_func": 0.125,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5687229856848717,
+      "rewards/question_recreation_reward_func": 0.6607661917805672,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.03125,
-      "rewards/xmlcount_reward_func": 0.4321250058710575,
+      "rewards/xmlcount_reward_func": -0.07549997000023723,
       "step": 2
     },
     {
-      "completion_length": 607.6785714285714,
+      "completion_length": 508.0,
       "epoch": 1.3636363636363638,
-      "grad_norm": 0.41752761602401733,
-      "kl": 0.0005801213389661695,
+      "grad_norm": 0.35529419779777527,
+      "kl": 0.0007047405898837107,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 0.7936271301337651,
-      "reward_std": 0.8227866802896772,
-      "rewards/concensus_correctness_reward_func": 0.0,
+      "reward": 1.0851636188370841,
+      "reward_std": 0.9851899061884198,
+      "rewards/concensus_correctness_reward_func": 0.02914285659790039,
       "rewards/consensus_reward_func": 0.14285714285714285,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5946271462099892,
+      "rewards/question_recreation_reward_func": 0.5143064622368131,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.03828570778880801,
+      "rewards/strict_format_reward_func": 0.03571428571428571,
+      "rewards/xmlcount_reward_func": 0.3631428522723062,
       "step": 4
     },
     {
-      "completion_length": 419.9642857142857,
+      "completion_length": 542.7142857142857,
       "epoch": 2.0,
-      "grad_norm": 0.6341163516044617,
-      "kl": 0.0007463138733458306,
+      "grad_norm": 0.3987818956375122,
+      "kl": 0.0006694004363712988,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 1.005787574819156,
-      "reward_std": 0.9503834183726992,
-      "rewards/concensus_correctness_reward_func": 0.0,
+      "reward": 1.1537523886987142,
+      "reward_std": 1.110692071063178,
+      "rewards/concensus_correctness_reward_func": 0.02914285659790039,
       "rewards/consensus_reward_func": 0.35714285714285715,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.6678590263639178,
+      "rewards/question_recreation_reward_func": 0.6857523918151855,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.01921428952898298,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.0638571275132043,
       "step": 6
     },
     {
-      "completion_length": 458.625,
+      "completion_length": 572.6875,
       "epoch": 2.7272727272727275,
-      "grad_norm": 0.337610125541687,
-      "kl": 0.0006834228988736868,
+      "grad_norm": 0.45522359013557434,
+      "kl": 0.0006443273268814664,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 1.112733330577612,
-      "reward_std": 0.8835000544786453,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.1875,
+      "reward": 0.7862032204866409,
+      "reward_std": 0.8552147410809994,
+      "rewards/concensus_correctness_reward_func": 0.026624999940395355,
+      "rewards/consensus_reward_func": 0.125,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.6689833588898182,
+      "rewards/question_recreation_reward_func": 0.6428594682365656,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.24062499403953552,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": -0.008281249552965164,
       "step": 8
     },
     {
-      "completion_length": 511.0357142857143,
+      "completion_length": 477.07142857142856,
       "epoch": 3.3636363636363638,
-      "grad_norm": 0.4655108153820038,
-      "kl": 0.0006738068186677992,
+      "grad_norm": 0.7522605657577515,
+      "kl": 0.0008250060907032873,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 0.7448970739330564,
-      "reward_std": 0.9845571688243321,
-      "rewards/concensus_correctness_reward_func": 0.04607142720903669,
+      "reward": 0.9303557702473232,
+      "reward_std": 0.7904164152485984,
+      "rewards/concensus_correctness_reward_func": 0.0,
       "rewards/consensus_reward_func": 0.14285714285714285,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5173256482396807,
+      "rewards/question_recreation_reward_func": 0.5055700625692096,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.038642864142145426,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.2640714320753302,
       "step": 10
     },
     {
-      "completion_length": 581.4642857142857,
+      "completion_length": 579.2142857142857,
       "epoch": 4.0,
-      "grad_norm": 0.3229527473449707,
-      "kl": 0.0006438559919063534,
+      "grad_norm": 0.4029735028743744,
+      "kl": 0.0005824401658693594,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 0.901876015322549,
-      "reward_std": 0.5733589742864881,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.14285714285714285,
+      "reward": 1.1292478569916315,
+      "reward_std": 0.8686830273696354,
+      "rewards/concensus_correctness_reward_func": 0.030428571360451833,
+      "rewards/consensus_reward_func": 0.2857142857142857,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5391260130064828,
+      "rewards/question_recreation_reward_func": 0.5652121348040444,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.21989286371639796,
+      "rewards/xmlcount_reward_func": 0.2478928565979004,
       "step": 12
     },
     {
-      "completion_length": 471.3125,
+      "completion_length": 491.71875,
       "epoch": 4.7272727272727275,
-      "grad_norm": 0.45552772283554077,
-      "kl": 0.0006461980010499246,
+      "grad_norm": 0.4175528883934021,
+      "kl": 0.0006781894153391477,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 1.0484812445938587,
-      "reward_std": 0.9308878369629383,
+      "reward": 0.8274601846933365,
+      "reward_std": 0.827268935739994,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.1875,
+      "rewards/consensus_reward_func": 0.3125,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5975437425076962,
+      "rewards/question_recreation_reward_func": 0.5699914516881108,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.24781250103842467,
+      "rewards/xmlcount_reward_func": -0.07065626233816147,
       "step": 14
     },
     {
-      "completion_length": 495.10714285714283,
+      "completion_length": 584.6428571428571,
       "epoch": 5.363636363636363,
-      "grad_norm": 16.87542152404785,
-      "kl": 0.0007952758543459433,
+      "grad_norm": 0.6693198680877686,
+      "kl": 0.0005940364166495524,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 1.2276427149772644,
-      "reward_std": 0.8300601371697017,
+      "reward": 1.452887875693185,
+      "reward_std": 1.118608525821141,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "rewards/consensus_reward_func": 0.42857142857142855,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5341070124081203,
+      "rewards/final_correctness_reward_func": 0.21428571428571427,
+      "rewards/question_recreation_reward_func": 0.6081378630229405,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.03571428571428571,
-      "rewards/xmlcount_reward_func": 0.372107134865863,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.20189285916941507,
       "step": 16
     },
     {
-      "completion_length": 681.3571428571429,
+      "completion_length": 631.2142857142857,
       "epoch": 6.0,
-      "grad_norm": 0.4771121144294739,
-      "kl": 0.000520679762952828,
+      "grad_norm": 0.3829941749572754,
+      "kl": 0.0006050001913016396,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 0.602795483810561,
-      "reward_std": 0.9616657410349164,
+      "reward": 1.1395895055362157,
+      "reward_std": 1.3548330409186227,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.14285714285714285,
+      "rewards/consensus_reward_func": 0.5,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5815811795847756,
+      "rewards/question_recreation_reward_func": 0.4571609433208193,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.1216428450175694,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.16457142042262213,
       "step": 18
     },
     {
-      "completion_length": 476.34375,
+      "completion_length": 467.9375,
       "epoch": 6.7272727272727275,
-      "grad_norm": 0.4769397974014282,
-      "kl": 0.000715115420462098,
+      "grad_norm": 0.7331616878509521,
+      "kl": 0.0006881270383019,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 0.5274622952565551,
-      "reward_std": 0.5800921861082315,
+      "reward": 1.052180115133524,
+      "reward_std": 1.0876893140375614,
       "rewards/concensus_correctness_reward_func": 0.025499999523162842,
-      "rewards/consensus_reward_func": 0.125,
+      "rewards/consensus_reward_func": 0.25,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5373685359954834,
-      "rewards/soft_format_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5559613723307848,
+      "rewards/soft_format_reward_func": 0.015625,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.16040625050663948,
+      "rewards/xmlcount_reward_func": 0.2050937619060278,
       "step": 20
     },
     {
       "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 6.243158054530796e-07,
-      "train_runtime": 662.6617,
-      "train_samples_per_second": 0.483,
+      "train_loss": 6.091722354995e-07,
+      "train_runtime": 676.4725,
+      "train_samples_per_second": 0.473,
       "train_steps_per_second": 0.03
     }
   ],