End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6a493fcd4f8490e6da17c81266ac2321fb8c0b900cea4e713346926af3f7089b
+oid sha256:b9b946e8d9ad5bb62b05a8f77ecba797394efe30a23ba8d70d8bb4a71d4bd6dc
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 7.334376395995435e-07,
-    "train_runtime": 1213.9576,
-    "train_samples": 15,
-    "train_samples_per_second": 0.264,
-    "train_steps_per_second": 0.016
+    "train_loss": 7.458811126070942e-07,
+    "train_runtime": 620.7238,
+    "train_samples": 11,
+    "train_samples_per_second": 0.516,
+    "train_steps_per_second": 0.032
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 7.334376395995435e-07,
-    "train_runtime": 1213.9576,
-    "train_samples": 15,
-    "train_samples_per_second": 0.264,
-    "train_steps_per_second": 0.016
+    "train_loss": 7.458811126070942e-07,
+    "train_runtime": 620.7238,
+    "train_samples": 11,
+    "train_samples_per_second": 0.516,
+    "train_steps_per_second": 0.032
 }

trainer_state.json

@@ -2,7 +2,7 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 5.0,
+  "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": 370.5,
-      "epoch": 0.5333333333333333,
-      "grad_norm": 0.6320868134498596,
-      "kl": 0.0,
+      "completion_length": 487.875,
+      "epoch": 0.7272727272727273,
+      "grad_norm": 0.3863629102706909,
+      "kl": 0.000673882823321037,
       "learning_rate": 5e-07,
-      "loss": -0.0,
-      "reward": 1.178517110645771,
-      "reward_std": 0.9825490638613701,
-      "rewards/concensus_correctness_reward_func": 0.04531250335276127,
-      "rewards/consensus_reward_func": 0.125,
+      "loss": 0.0,
+      "reward": 1.0874615386128426,
+      "reward_std": 1.007015310227871,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.3125,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.25,
-      "rewards/question_recreation_reward_func": 0.420610835775733,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.4903052784502506,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.3219687547534704,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.28465624526143074,
       "step": 2
     },
     {
-      "completion_length": 328.92857142857144,
-      "epoch": 1.0,
-      "grad_norm": 0.47581061720848083,
-      "kl": 0.0009534425917081535,
+      "completion_length": 390.85714285714283,
+      "epoch": 1.3636363636363638,
+      "grad_norm": 0.35562217235565186,
+      "kl": 0.0007252936234830745,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 1.2720517558710915,
-      "reward_std": 1.0473234908921378,
+      "reward": 0.9998211647783007,
+      "reward_std": 1.2654829919338226,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.14285714285714285,
+      "rewards/consensus_reward_func": 0.5714285714285714,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.42857142857142855,
-      "rewards/question_recreation_reward_func": 0.4950517469218799,
-      "rewards/soft_format_reward_func": 0.017857142857142856,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.49107117312295095,
+      "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.18771429998534067,
+      "rewards/xmlcount_reward_func": -0.06267857338700976,
       "step": 4
     },
     {
-      "completion_length": 390.96875,
-      "epoch": 1.5333333333333332,
-      "grad_norm": 0.5493902564048767,
-      "kl": 0.0007869710389059037,
+      "completion_length": 450.57142857142856,
+      "epoch": 2.0,
+      "grad_norm": 0.27575239539146423,
+      "kl": 0.0007243190484587103,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 1.5517104975879192,
-      "reward_std": 1.0216393284499645,
-      "rewards/concensus_correctness_reward_func": 0.06012500077486038,
-      "rewards/consensus_reward_func": 0.1875,
+      "reward": 1.2959521668297904,
+      "reward_std": 1.229691309588296,
+      "rewards/concensus_correctness_reward_func": 0.060857142720903666,
+      "rewards/consensus_reward_func": 0.5714285714285714,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.375,
-      "rewards/question_recreation_reward_func": 0.4212729502469301,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5307021524224963,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.03125,
-      "rewards/xmlcount_reward_func": 0.4765625037252903,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.13296428908194816,
       "step": 6
     },
     {
-      "completion_length": 387.5,
-      "epoch": 2.0,
-      "grad_norm": 0.38475939631462097,
-      "kl": 0.0007659006265125104,
+      "completion_length": 429.15625,
+      "epoch": 2.7272727272727275,
+      "grad_norm": 4.421164512634277,
+      "kl": 0.0010074221136164851,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 0.7905604477439608,
-      "reward_std": 0.9036691273961749,
+      "reward": 0.8775190860033035,
+      "reward_std": 0.9346675388514996,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.07142857142857142,
+      "rewards/consensus_reward_func": 0.1875,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.21428571428571427,
-      "rewards/question_recreation_reward_func": 0.4089176186493465,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5945190861821175,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.0780714319220611,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": 0.06425000168383121,
       "step": 8
     },
     {
-      "completion_length": 298.375,
-      "epoch": 2.533333333333333,
-      "grad_norm": 4.775766372680664,
-      "kl": 0.0009655183021095581,
+      "completion_length": 440.60714285714283,
+      "epoch": 3.3636363636363638,
+      "grad_norm": 0.47421273589134216,
+      "kl": 0.0007379744783975184,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 1.925378292798996,
-      "reward_std": 1.332388810813427,
-      "rewards/concensus_correctness_reward_func": 0.09481250308454037,
-      "rewards/consensus_reward_func": 0.3125,
+      "reward": 1.2224792901958739,
+      "reward_std": 1.154342634337289,
+      "rewards/concensus_correctness_reward_func": 0.030428571360451833,
+      "rewards/consensus_reward_func": 0.35714285714285715,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.5625,
-      "rewards/question_recreation_reward_func": 0.524846994318068,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5376935941832406,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.41509374789893627,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.27935714381081717,
       "step": 10
     },
     {
-      "completion_length": 359.85714285714283,
-      "epoch": 3.0,
-      "grad_norm": 0.6231438517570496,
-      "kl": 0.0008056558934705598,
+      "completion_length": 499.5357142857143,
+      "epoch": 4.0,
+      "grad_norm": 0.41844943165779114,
+      "kl": 0.0007170024826856596,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 1.1937377282551356,
-      "reward_std": 1.1367330593722207,
+      "reward": 0.9398225631032672,
+      "reward_std": 0.8540037316935403,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "rewards/consensus_reward_func": 0.5714285714285714,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.21428571428571427,
-      "rewards/question_recreation_reward_func": 0.4232020058802196,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.4725368299654552,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.25267857419592993,
+      "rewards/xmlcount_reward_func": -0.1220000045640128,
       "step": 12
     },
     {
-      "completion_length": 391.75,
-      "epoch": 3.533333333333333,
-      "grad_norm": 0.671714723110199,
-      "kl": 0.0007653082866454497,
+      "completion_length": 403.3125,
+      "epoch": 4.7272727272727275,
+      "grad_norm": 5.868327617645264,
+      "kl": 0.0012472297603380866,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 2.090083334594965,
-      "reward_std": 1.2006787694990635,
-      "rewards/concensus_correctness_reward_func": 0.427249975502491,
-      "rewards/consensus_reward_func": 0.375,
+      "reward": 0.8139126822352409,
+      "reward_std": 0.9305730424821377,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.125,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.4375,
-      "rewards/question_recreation_reward_func": 0.4738645385950804,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.4786001555621624,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0625,
-      "rewards/xmlcount_reward_func": 0.3139687478542328,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.2103125024586916,
       "step": 14
     },
     {
-      "completion_length": 279.60714285714283,
-      "epoch": 4.0,
-      "grad_norm": 0.6454631090164185,
-      "kl": 0.0009609005646780133,
+      "completion_length": 507.57142857142856,
+      "epoch": 5.363636363636363,
+      "grad_norm": 0.411307692527771,
+      "kl": 0.0006107560891125884,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 2.0080738365650177,
-      "reward_std": 1.9522871311221803,
-      "rewards/concensus_correctness_reward_func": 0.7142857142857143,
-      "rewards/consensus_reward_func": 0.07142857142857142,
+      "reward": 0.7328580894640514,
+      "reward_std": 0.6605191060474941,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.2857142857142857,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.2857142857142857,
-      "rewards/question_recreation_reward_func": 0.5558595657348633,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.48507237966571537,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.38078572281769346,
+      "rewards/xmlcount_reward_func": -0.03792857698031834,
       "step": 16
     },
     {
-      "completion_length": 352.21875,
-      "epoch": 4.533333333333333,
-      "grad_norm": 0.7853459119796753,
-      "kl": 0.0008694051721249707,
+      "completion_length": 514.0,
+      "epoch": 6.0,
+      "grad_norm": 0.4124680459499359,
+      "kl": 0.0006932900786133749,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 1.630488459020853,
-      "reward_std": 1.1447128355503082,
-      "rewards/concensus_correctness_reward_func": 0.07325000315904617,
-      "rewards/consensus_reward_func": 0.125,
+      "reward": 0.98552576984678,
+      "reward_std": 1.170050778559276,
+      "rewards/concensus_correctness_reward_func": 0.030428571360451833,
+      "rewards/consensus_reward_func": 0.5714285714285714,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.6875,
-      "rewards/question_recreation_reward_func": 0.4198634847998619,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.4463829163994108,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.32487500412389636,
+      "rewards/xmlcount_reward_func": -0.06271427656923022,
       "step": 18
     },
     {
-      "completion_length": 347.42857142857144,
-      "epoch": 5.0,
-      "grad_norm": 0.8366572856903076,
-      "kl": 0.0008741397344108139,
+      "completion_length": 467.21875,
+      "epoch": 6.7272727272727275,
+      "grad_norm": 0.4149317741394043,
+      "kl": 0.0006911894888617098,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 1.2023059981209892,
-      "reward_std": 0.7625741745744433,
-      "rewards/concensus_correctness_reward_func": -0.018285714089870453,
-      "rewards/consensus_reward_func": 0.35714285714285715,
+      "reward": 0.851905420422554,
+      "reward_std": 1.0343977212905884,
+      "rewards/concensus_correctness_reward_func": 0.026624999940395355,
+      "rewards/consensus_reward_func": 0.25,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.4690203091927937,
+      "rewards/question_recreation_reward_func": 0.46009293384850025,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.39442857993500574,
+      "rewards/xmlcount_reward_func": 0.11518749874085188,
       "step": 20
     },
     {
-      "epoch": 5.0,
+      "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 7.334376395995435e-07,
-      "train_runtime": 1213.9576,
-      "train_samples_per_second": 0.264,
-      "train_steps_per_second": 0.016
+      "train_loss": 7.458811126070942e-07,
+      "train_runtime": 620.7238,
+      "train_samples_per_second": 0.516,
+      "train_steps_per_second": 0.032
     }
   ],
   "logging_steps": 2,
   "max_steps": 20,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 7,
+  "num_train_epochs": 10,
   "save_steps": 25,
   "stateful_callbacks": {
     "TrainerControl": {