End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5d32dd4fa0e751b196507ea002f3a61ee806451e62b7fed3ee2d6d652a8366af
+oid sha256:1e9c75ea0086cab697668ce441630306c0a73ae69fa8ac8086cbd3418c5f76fd
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 7.206053567188064e-07,
-    "train_runtime": 1797.6879,
+    "train_loss": 6.466318950515415e-07,
+    "train_runtime": 1576.8638,
     "train_samples": 11,
-    "train_samples_per_second": 0.178,
-    "train_steps_per_second": 0.011
+    "train_samples_per_second": 0.203,
+    "train_steps_per_second": 0.013
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 7.206053567188064e-07,
-    "train_runtime": 1797.6879,
+    "train_loss": 6.466318950515415e-07,
+    "train_runtime": 1576.8638,
     "train_samples": 11,
-    "train_samples_per_second": 0.178,
-    "train_steps_per_second": 0.011
+    "train_samples_per_second": 0.203,
+    "train_steps_per_second": 0.013
 }

trainer_state.json

@@ -10,203 +10,203 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 379.1875,
+      "completion_length": 476.0,
       "epoch": 0.7272727272727273,
-      "grad_norm": 0.47303661704063416,
-      "kl": 0.0008382251326111145,
+      "grad_norm": 0.5276138186454773,
+      "kl": 0.0006918963335920125,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 1.2963236942887306,
-      "reward_std": 1.3006894960999489,
-      "rewards/concensus_correctness_reward_func": 0.09331249631941319,
-      "rewards/consensus_reward_func": 0.4375,
+      "reward": 0.8905868269503117,
+      "reward_std": 0.9817413352429867,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.1875,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.530417438596487,
+      "rewards/question_recreation_reward_func": 0.5417430754750967,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.2194687519222498,
+      "rewards/xmlcount_reward_func": 0.1457187570631504,
       "step": 2
     },
     {
-      "completion_length": 407.42857142857144,
+      "completion_length": 604.5357142857143,
       "epoch": 1.3636363636363638,
-      "grad_norm": 0.44756442308425903,
-      "kl": 0.0008138769605596151,
+      "grad_norm": 0.45205041766166687,
+      "kl": 0.0006221245400541063,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 0.8841377794742584,
-      "reward_std": 0.7753841685397285,
-      "rewards/concensus_correctness_reward_func": -0.01764285670859473,
+      "reward": 0.9950631601469857,
+      "reward_std": 0.8837312459945679,
+      "rewards/concensus_correctness_reward_func": 0.016285714294229234,
       "rewards/consensus_reward_func": 0.21428571428571427,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.3659592185701643,
+      "rewards/final_correctness_reward_func": 0.07142857142857142,
+      "rewards/question_recreation_reward_func": 0.3275988719293049,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.3215357033269746,
+      "rewards/xmlcount_reward_func": 0.3654642839516912,
       "step": 4
     },
     {
-      "completion_length": 413.14285714285717,
+      "completion_length": 566.5357142857143,
       "epoch": 2.0,
-      "grad_norm": 0.4722234010696411,
-      "kl": 0.0007242053148469754,
+      "grad_norm": 0.45169490575790405,
+      "kl": 0.0007057673730222243,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 1.5978770766939436,
-      "reward_std": 1.196230207170759,
-      "rewards/concensus_correctness_reward_func": 0.08271428304059165,
-      "rewards/consensus_reward_func": 0.42857142857142855,
+      "reward": 1.136856577226094,
+      "reward_std": 0.7043021747044155,
+      "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": 0.5249485075473785,
+      "rewards/question_recreation_reward_func": 0.5205351114273071,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.5437857189348766,
+      "rewards/xmlcount_reward_func": 0.45560715028217863,
       "step": 6
     },
     {
-      "completion_length": 475.90625,
+      "completion_length": 435.03125,
       "epoch": 2.7272727272727275,
-      "grad_norm": 1.3227853775024414,
-      "kl": 0.0009269550064345822,
+      "grad_norm": 0.4807434678077698,
+      "kl": 0.0007996001113497186,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 0.8639741316437721,
-      "reward_std": 0.8438533209264278,
-      "rewards/concensus_correctness_reward_func": 0.010812499560415745,
-      "rewards/consensus_reward_func": 0.0625,
+      "reward": 0.895195696502924,
+      "reward_std": 0.9222910553216934,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.25,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.4206928927451372,
-      "rewards/soft_format_reward_func": 0.015625,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.35434375051409006,
+      "rewards/question_recreation_reward_func": 0.4469144530594349,
+      "rewards/soft_format_reward_func": 0.0,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.18265624903142452,
       "step": 8
     },
     {
-      "completion_length": 366.5357142857143,
+      "completion_length": 514.0,
       "epoch": 3.3636363636363638,
-      "grad_norm": 0.4223344326019287,
-      "kl": 0.0009213967756035604,
+      "grad_norm": 0.45199769735336304,
+      "kl": 0.0006535403995907732,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 1.9471516268593925,
-      "reward_std": 1.1154484408242362,
-      "rewards/concensus_correctness_reward_func": 0.17528570762702397,
-      "rewards/consensus_reward_func": 0.7142857142857143,
+      "reward": 0.5129805675574711,
+      "reward_std": 1.017007691519601,
+      "rewards/concensus_correctness_reward_func": 0.0005714285985699721,
+      "rewards/consensus_reward_func": 0.07142857142857142,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.49286590942314695,
+      "rewards/question_recreation_reward_func": 0.5081234255007335,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.03571428571428571,
-      "rewards/xmlcount_reward_func": 0.5290000140666962,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": -0.08499999557222639,
       "step": 10
     },
     {
-      "completion_length": 454.89285714285717,
+      "completion_length": 497.10714285714283,
       "epoch": 4.0,
-      "grad_norm": 0.5085252523422241,
-      "kl": 0.0006998284786407437,
+      "grad_norm": 0.38202306628227234,
+      "kl": 0.0007072919480768698,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 1.1383702797549111,
-      "reward_std": 0.8994386494159698,
-      "rewards/concensus_correctness_reward_func": 0.00021428614854812622,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "reward": 0.6489094985382897,
+      "reward_std": 0.9131356051989964,
+      "rewards/concensus_correctness_reward_func": 0.029857143759727478,
+      "rewards/consensus_reward_func": 0.14285714285714285,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.43279886884348734,
+      "rewards/question_recreation_reward_func": 0.39351664057799746,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.40178572280066355,
+      "rewards/xmlcount_reward_func": 0.06482143008283206,
       "step": 12
     },
     {
-      "completion_length": 467.09375,
+      "completion_length": 387.15625,
       "epoch": 4.7272727272727275,
-      "grad_norm": 0.6773208379745483,
-      "kl": 0.0006632515505771153,
+      "grad_norm": 0.8552425503730774,
+      "kl": 0.0007712517472100444,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 1.4484406746923923,
-      "reward_std": 1.0123861022293568,
-      "rewards/concensus_correctness_reward_func": 0.07981249876320362,
-      "rewards/consensus_reward_func": 0.375,
+      "reward": 0.8629500679671764,
+      "reward_std": 0.7735467366874218,
+      "rewards/concensus_correctness_reward_func": 0.0005000000237487257,
+      "rewards/consensus_reward_func": 0.0625,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.47909694258123636,
+      "rewards/question_recreation_reward_func": 0.42723130341619253,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.4989062584936619,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.37271876260638237,
       "step": 14
     },
     {
-      "completion_length": 438.0357142857143,
+      "completion_length": 518.0,
       "epoch": 5.363636363636363,
-      "grad_norm": 0.4543914496898651,
-      "kl": 0.000652431951103998,
+      "grad_norm": 0.41807857155799866,
+      "kl": 0.000663631537463516,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 1.1117327766759055,
-      "reward_std": 1.0020887894289834,
-      "rewards/concensus_correctness_reward_func": 0.017857142857142856,
-      "rewards/consensus_reward_func": 0.14285714285714285,
+      "reward": 0.6208901362759727,
+      "reward_std": 0.9498651453426906,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.07142857142857142,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.4449828032936369,
+      "rewards/question_recreation_reward_func": 0.5296044392245156,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.4881785735487938,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.01985713094472885,
       "step": 16
     },
     {
-      "completion_length": 423.75,
+      "completion_length": 579.1785714285714,
       "epoch": 6.0,
-      "grad_norm": 0.46868741512298584,
-      "kl": 0.0008392483869101852,
+      "grad_norm": 0.4610684812068939,
+      "kl": 0.000690574080051322,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 1.3158689992768424,
-      "reward_std": 1.046988070011139,
-      "rewards/concensus_correctness_reward_func": 0.040928571351936886,
-      "rewards/consensus_reward_func": 0.5,
+      "reward": 0.5093637534550258,
+      "reward_std": 0.892940702182906,
+      "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": 0.4287618654114859,
+      "rewards/question_recreation_reward_func": 0.4273637448038374,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.34617857741458075,
+      "rewards/xmlcount_reward_func": -0.060857132875493596,
       "step": 18
     },
     {
-      "completion_length": 450.96875,
+      "completion_length": 560.03125,
       "epoch": 6.7272727272727275,
-      "grad_norm": 2.5736513137817383,
-      "kl": 0.0007858428652980365,
+      "grad_norm": 0.38961368799209595,
+      "kl": 0.0006178304938657675,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 0.8791648773476481,
-      "reward_std": 1.0170073471963406,
-      "rewards/concensus_correctness_reward_func": 0.043999998830258846,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 1.1620018668472767,
+      "reward_std": 1.1849149875342846,
+      "rewards/concensus_correctness_reward_func": 0.0010000000474974513,
+      "rewards/consensus_reward_func": 0.3125,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.4682898800820112,
+      "rewards/question_recreation_reward_func": 0.6256268601864576,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.1168750049546361,
+      "rewards/xmlcount_reward_func": 0.2228750018402934,
       "step": 20
     },
     {
       "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 7.206053567188064e-07,
-      "train_runtime": 1797.6879,
-      "train_samples_per_second": 0.178,
-      "train_steps_per_second": 0.011
+      "train_loss": 6.466318950515415e-07,
+      "train_runtime": 1576.8638,
+      "train_samples_per_second": 0.203,
+      "train_steps_per_second": 0.013
     }
   ],
   "logging_steps": 2,