End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:729df86178e951ac406b64872637f6342e6e324130eec2c547f842497eb36d7d
+oid sha256:9bef461733c9bf37484c2069176747480fe26c92ba92cfddafda4ead2f5e72ed
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 6.59394788726786e-07,
-    "train_runtime": 657.9092,
+    "train_loss": 7.355144987286622e-07,
+    "train_runtime": 621.2835,
     "train_samples": 11,
-    "train_samples_per_second": 0.486,
-    "train_steps_per_second": 0.03
+    "train_samples_per_second": 0.515,
+    "train_steps_per_second": 0.032
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 6.59394788726786e-07,
-    "train_runtime": 657.9092,
+    "train_loss": 7.355144987286622e-07,
+    "train_runtime": 621.2835,
     "train_samples": 11,
-    "train_samples_per_second": 0.486,
-    "train_steps_per_second": 0.03
+    "train_samples_per_second": 0.515,
+    "train_steps_per_second": 0.032
 }

trainer_state.json

@@ -10,203 +10,203 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 529.5625,
+      "completion_length": 407.5,
       "epoch": 0.7272727272727273,
-      "grad_norm": 0.3786744177341461,
-      "kl": 0.0006517168221762404,
+      "grad_norm": 0.5940650701522827,
+      "kl": 0.0007805308450770099,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 1.925076533574611,
-      "reward_std": 2.4438512101769447,
-      "rewards/concensus_correctness_reward_func": 0.7248124815523624,
-      "rewards/consensus_reward_func": 0.375,
+      "reward": 1.089016579091549,
+      "reward_std": 1.0153970010578632,
+      "rewards/concensus_correctness_reward_func": 0.008999999612569809,
+      "rewards/consensus_reward_func": 0.25,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.25,
-      "rewards/question_recreation_reward_func": 0.5201703049242496,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.43173531256616116,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.039468761533498764,
+      "rewards/xmlcount_reward_func": 0.3826562473550439,
       "step": 2
     },
     {
-      "completion_length": 518.6428571428571,
+      "completion_length": 438.57142857142856,
       "epoch": 1.3636363636363638,
-      "grad_norm": 0.32368478178977966,
-      "kl": 0.0008078359200486115,
+      "grad_norm": 0.528326153755188,
+      "kl": 0.0007703895846913968,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 0.5248787983187607,
-      "reward_std": 0.7695371742759433,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.07142857142857142,
+      "reward": 0.8055335581302643,
+      "reward_std": 0.7478888928890228,
+      "rewards/concensus_correctness_reward_func": 0.008928571428571428,
+      "rewards/consensus_reward_func": 0.14285714285714285,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.41720023325511385,
+      "rewards/question_recreation_reward_func": 0.35178354595388683,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.036249990441969464,
+      "rewards/xmlcount_reward_func": 0.30196428298950195,
       "step": 4
     },
     {
-      "completion_length": 521.8214285714286,
+      "completion_length": 402.2857142857143,
       "epoch": 2.0,
-      "grad_norm": 0.33299964666366577,
-      "kl": 0.0006702714350207575,
+      "grad_norm": 0.3969787061214447,
+      "kl": 0.0007904843743225294,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 2.068307025091989,
-      "reward_std": 2.345199533871242,
-      "rewards/concensus_correctness_reward_func": 0.7142857142857143,
-      "rewards/consensus_reward_func": 0.5,
+      "reward": 1.0549773488725935,
+      "reward_std": 0.95536869764328,
+      "rewards/concensus_correctness_reward_func": 0.03142857125827244,
+      "rewards/consensus_reward_func": 0.21428571428571427,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.21428571428571427,
-      "rewards/question_recreation_reward_func": 0.5130927264690399,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.502727325473513,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.12664285727909633,
+      "rewards/xmlcount_reward_func": 0.30653572188956396,
       "step": 6
     },
     {
-      "completion_length": 606.15625,
+      "completion_length": 449.4375,
       "epoch": 2.7272727272727275,
-      "grad_norm": 0.32098329067230225,
-      "kl": 0.0006415432690118905,
+      "grad_norm": 0.4158034920692444,
+      "kl": 0.0006508094738819636,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 1.0850283484905958,
-      "reward_std": 1.1896289810538292,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.5,
+      "reward": 0.7934029828757048,
+      "reward_std": 1.0302056893706322,
+      "rewards/concensus_correctness_reward_func": 0.0044999998062849045,
+      "rewards/consensus_reward_func": 0.25,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.39380961563438177,
-      "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.1912187566049397,
+      "rewards/question_recreation_reward_func": 0.481402987614274,
+      "rewards/soft_format_reward_func": 0.015625,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.02625000011175871,
       "step": 8
     },
     {
-      "completion_length": 472.10714285714283,
+      "completion_length": 547.4285714285714,
       "epoch": 3.3636363636363638,
-      "grad_norm": 0.3529147505760193,
-      "kl": 0.0007211144007409789,
+      "grad_norm": 0.7092744708061218,
+      "kl": 0.0007078879072131323,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 2.2373720088175366,
-      "reward_std": 2.4381445475987027,
-      "rewards/concensus_correctness_reward_func": 0.8887142624173846,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "reward": 0.6819984870297569,
+      "reward_std": 1.1465651563235693,
+      "rewards/concensus_correctness_reward_func": -0.016357143010411943,
+      "rewards/consensus_reward_func": 0.14285714285714285,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.42857142857142855,
-      "rewards/question_recreation_reward_func": 0.5479791164398193,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.4367127673966544,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.06853571001972471,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.11878570701394763,
       "step": 10
     },
     {
-      "completion_length": 477.85714285714283,
+      "completion_length": 526.7857142857143,
       "epoch": 4.0,
-      "grad_norm": 0.3556641936302185,
-      "kl": 0.0007031152739987842,
+      "grad_norm": 0.25688636302948,
+      "kl": 0.0006741188657802663,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 1.0147725492715836,
-      "reward_std": 1.1198050209454127,
-      "rewards/concensus_correctness_reward_func": 0.053714283875056675,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "reward": 0.9941186138561794,
+      "reward_std": 0.8434865730149406,
+      "rewards/concensus_correctness_reward_func": -0.05057142887796674,
+      "rewards/consensus_reward_func": 0.35714285714285715,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.4988082711185728,
+      "rewards/question_recreation_reward_func": 0.5689757402454104,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.1586785614490509,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.11857144641024726,
       "step": 12
     },
     {
-      "completion_length": 544.625,
+      "completion_length": 546.96875,
       "epoch": 4.7272727272727275,
-      "grad_norm": 0.4013236463069916,
-      "kl": 0.0006297369363892358,
+      "grad_norm": 0.35739797353744507,
+      "kl": 0.0007387452606053557,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 1.5754604525864124,
-      "reward_std": 2.408501971513033,
-      "rewards/concensus_correctness_reward_func": 0.6484999991953373,
-      "rewards/consensus_reward_func": 0.3125,
+      "reward": 1.4755905121564865,
+      "reward_std": 0.9978351555764675,
+      "rewards/concensus_correctness_reward_func": 0.07062499970197678,
+      "rewards/consensus_reward_func": 0.375,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.25,
-      "rewards/question_recreation_reward_func": 0.3719604127109051,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5299967229366302,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.0074999975040555,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": 0.4687187559902668,
       "step": 14
     },
     {
-      "completion_length": 515.2142857142857,
+      "completion_length": 474.5,
       "epoch": 5.363636363636363,
-      "grad_norm": 0.3888353705406189,
-      "kl": 0.0007684821279586426,
+      "grad_norm": 0.5380047559738159,
+      "kl": 0.0007827934750821441,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 0.8480535106999534,
-      "reward_std": 1.0383951578821455,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.5,
+      "reward": 1.1178060003689356,
+      "reward_std": 0.9631217760699136,
+      "rewards/concensus_correctness_reward_func": -0.007428571581840515,
+      "rewards/consensus_reward_func": 0.35714285714285715,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.4994821037564959,
-      "rewards/soft_format_reward_func": 0.017857142857142856,
+      "rewards/question_recreation_reward_func": 0.5305917114019394,
+      "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.16928571249757493,
+      "rewards/xmlcount_reward_func": 0.23750000806259258,
       "step": 16
     },
     {
-      "completion_length": 504.17857142857144,
+      "completion_length": 377.2857142857143,
       "epoch": 6.0,
-      "grad_norm": 1.9306660890579224,
-      "kl": 0.0008064009598456323,
+      "grad_norm": 0.6083970665931702,
+      "kl": 0.0009674839176503676,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 1.2755470446177892,
-      "reward_std": 1.1879023483821325,
-      "rewards/concensus_correctness_reward_func": 0.08721428258078438,
-      "rewards/consensus_reward_func": 0.21428571428571427,
+      "reward": 1.7046442798205785,
+      "reward_std": 0.9649685876710075,
+      "rewards/concensus_correctness_reward_func": 0.015428570764405387,
+      "rewards/consensus_reward_func": 0.6428571428571429,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.21428571428571427,
-      "rewards/question_recreation_reward_func": 0.423475586942264,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.43110853007861544,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.33628570820604053,
+      "rewards/strict_format_reward_func": 0.03571428571428571,
+      "rewards/xmlcount_reward_func": 0.5795357120888573,
       "step": 18
     },
     {
-      "completion_length": 481.875,
+      "completion_length": 503.5625,
       "epoch": 6.7272727272727275,
-      "grad_norm": 0.5555954575538635,
-      "kl": 0.0007709507881372701,
+      "grad_norm": 0.5003705620765686,
+      "kl": 0.0009899184442474507,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 1.3486508503556252,
-      "reward_std": 1.145274255424738,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 0.9532081484794617,
+      "reward_std": 1.5441483110189438,
+      "rewards/concensus_correctness_reward_func": 0.005374999716877937,
+      "rewards/consensus_reward_func": 0.4375,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.25,
-      "rewards/question_recreation_reward_func": 0.4009946119040251,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.3971456475555897,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.4320312598720193,
+      "rewards/xmlcount_reward_func": 0.0975625105202198,
       "step": 20
     },
     {
       "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 6.59394788726786e-07,
-      "train_runtime": 657.9092,
-      "train_samples_per_second": 0.486,
-      "train_steps_per_second": 0.03
+      "train_loss": 7.355144987286622e-07,
+      "train_runtime": 621.2835,
+      "train_samples_per_second": 0.515,
+      "train_steps_per_second": 0.032
     }
   ],
   "logging_steps": 2,