End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1073d76a18d971079c06d227040aace0e4dfda72904e1c626faa46f5a1bab622
+oid sha256:f614f898694cc3f5cae13335d2929ed92430f98827c497e272326c6a2ea30af1
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 2.193293255459139e-06,
-    "train_runtime": 577.4662,
+    "train_loss": 2.1295968963386256e-06,
+    "train_runtime": 640.5844,
     "train_samples": 11,
-    "train_samples_per_second": 0.554,
-    "train_steps_per_second": 0.035
+    "train_samples_per_second": 0.5,
+    "train_steps_per_second": 0.031
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 2.193293255459139e-06,
-    "train_runtime": 577.4662,
+    "train_loss": 2.1295968963386256e-06,
+    "train_runtime": 640.5844,
     "train_samples": 11,
-    "train_samples_per_second": 0.554,
-    "train_steps_per_second": 0.035
+    "train_samples_per_second": 0.5,
+    "train_steps_per_second": 0.031
 }

trainer_state.json

@@ -10,203 +10,203 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 430.6875,
+      "completion_length": 455.84375,
       "epoch": 0.7272727272727273,
-      "grad_norm": 9.133371353149414,
-      "kl": 0.0027560016023926437,
+      "grad_norm": 0.5454344153404236,
+      "kl": 0.002561649336712435,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 0.9409511927515268,
-      "reward_std": 0.9327438697218895,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.3125,
+      "reward": 1.2996555045247078,
+      "reward_std": 1.4193213507533073,
+      "rewards/concensus_correctness_reward_func": 0.057624999433755875,
+      "rewards/consensus_reward_func": 0.375,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.125,
-      "rewards/question_recreation_reward_func": 0.5155761726200581,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.6668429747223854,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.012124991044402122,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": 0.16893750615417957,
       "step": 2
     },
     {
-      "completion_length": 423.89285714285717,
+      "completion_length": 489.2142857142857,
       "epoch": 1.3636363636363638,
-      "grad_norm": 0.40043893456459045,
-      "kl": 0.0019687856048611657,
+      "grad_norm": 0.5793511271476746,
+      "kl": 0.0020495095273612867,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 2.2995907238551547,
-      "reward_std": 2.9735161747251238,
-      "rewards/concensus_correctness_reward_func": 0.8704999951379639,
-      "rewards/consensus_reward_func": 0.5714285714285714,
+      "reward": 1.3453438154288702,
+      "reward_std": 1.085130172116416,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.42857142857142855,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.614340718303408,
+      "rewards/question_recreation_reward_func": 0.5988795331546238,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.24332144004958017,
+      "rewards/xmlcount_reward_func": 0.3178928515740803,
       "step": 4
     },
     {
-      "completion_length": 372.7857142857143,
+      "completion_length": 571.5357142857143,
       "epoch": 2.0,
-      "grad_norm": 0.4292924404144287,
-      "kl": 0.002527345537341067,
+      "grad_norm": 0.3269042670726776,
+      "kl": 0.0019223600060545973,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 2.434863882405417,
-      "reward_std": 2.0036334012235915,
-      "rewards/concensus_correctness_reward_func": 1.0627143042428153,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "reward": 1.0687090839658464,
+      "reward_std": 1.128563838345664,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.21428571428571427,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.14285714285714285,
-      "rewards/question_recreation_reward_func": 0.62836383496012,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5492448168141502,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.05357142857142857,
-      "rewards/xmlcount_reward_func": 0.26164285838603973,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.2873214270387377,
       "step": 6
     },
     {
-      "completion_length": 492.6875,
+      "completion_length": 556.78125,
       "epoch": 2.7272727272727275,
-      "grad_norm": 0.38836348056793213,
-      "kl": 0.002220321781351231,
+      "grad_norm": 0.3594064712524414,
+      "kl": 0.00206602067919448,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 1.011883558705449,
-      "reward_std": 0.976864617317915,
-      "rewards/concensus_correctness_reward_func": 0.01575000025331974,
-      "rewards/consensus_reward_func": 0.5,
+      "reward": 1.2085097059607506,
+      "reward_std": 1.1704649776220322,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.375,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5699460301548243,
+      "rewards/question_recreation_reward_func": 0.7188222259283066,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.0738124898634851,
+      "rewards/xmlcount_reward_func": 0.11468750610947609,
       "step": 8
     },
     {
-      "completion_length": 389.39285714285717,
+      "completion_length": 471.0357142857143,
       "epoch": 3.3636363636363638,
-      "grad_norm": 0.610636293888092,
-      "kl": 0.002784155109631164,
+      "grad_norm": 0.4684782326221466,
+      "kl": 0.0025888860026108368,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 2.689215055533818,
-      "reward_std": 2.76099171595914,
-      "rewards/concensus_correctness_reward_func": 1.1207857483199664,
-      "rewards/consensus_reward_func": 0.35714285714285715,
+      "reward": 1.1132340942110335,
+      "reward_std": 1.0355860250336784,
+      "rewards/concensus_correctness_reward_func": 0.07142857142857142,
+      "rewards/consensus_reward_func": 0.07142857142857142,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.21428571428571427,
-      "rewards/question_recreation_reward_func": 0.5914651155471802,
-      "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.05357142857142857,
-      "rewards/xmlcount_reward_func": 0.351964282138007,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5203769547598702,
+      "rewards/soft_format_reward_func": 0.017857142857142856,
+      "rewards/strict_format_reward_func": 0.03571428571428571,
+      "rewards/xmlcount_reward_func": 0.39642857015132904,
       "step": 10
     },
     {
-      "completion_length": 556.7857142857143,
+      "completion_length": 480.17857142857144,
       "epoch": 4.0,
-      "grad_norm": 0.457033634185791,
-      "kl": 0.0024600966717116535,
+      "grad_norm": 0.6803588271141052,
+      "kl": 0.0026295380500544396,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 2.6462377224649702,
-      "reward_std": 2.9253767217908586,
-      "rewards/concensus_correctness_reward_func": 1.4776428554739272,
+      "reward": 1.3360323650496346,
+      "reward_std": 1.5073168448039465,
+      "rewards/concensus_correctness_reward_func": 0.03571428571428571,
       "rewards/consensus_reward_func": 0.5714285714285714,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.14285714285714285,
-      "rewards/question_recreation_reward_func": 0.566273399761745,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5627466653074537,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.11196428963116237,
+      "rewards/strict_format_reward_func": 0.03571428571428571,
+      "rewards/xmlcount_reward_func": 0.1304285717861993,
       "step": 12
     },
     {
-      "completion_length": 425.65625,
+      "completion_length": 452.5,
       "epoch": 4.7272727272727275,
-      "grad_norm": 0.6648972630500793,
-      "kl": 0.0025974765594583005,
+      "grad_norm": 0.6019690632820129,
+      "kl": 0.0020850498985964805,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 3.8004759550094604,
-      "reward_std": 3.797370210289955,
-      "rewards/concensus_correctness_reward_func": 1.8828750001266599,
-      "rewards/consensus_reward_func": 0.5625,
+      "reward": 1.0868228152394295,
+      "reward_std": 1.2246942669153214,
+      "rewards/concensus_correctness_reward_func": 0.0859375,
+      "rewards/consensus_reward_func": 0.25,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.25,
-      "rewards/question_recreation_reward_func": 0.6825697161257267,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5152916051447392,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.4225312490016222,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.21996874269098043,
       "step": 14
     },
     {
-      "completion_length": 416.82142857142856,
+      "completion_length": 478.64285714285717,
       "epoch": 5.363636363636363,
-      "grad_norm": 0.4645010828971863,
-      "kl": 0.0028217563605202095,
+      "grad_norm": 0.3849664628505707,
+      "kl": 0.0019824893014239414,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 1.8712751705731665,
-      "reward_std": 2.635926459516798,
-      "rewards/concensus_correctness_reward_func": 0.7633571411882129,
-      "rewards/consensus_reward_func": 0.42857142857142855,
+      "reward": 1.5740687847137451,
+      "reward_std": 1.4999604054859705,
+      "rewards/concensus_correctness_reward_func": 0.07142857142857142,
+      "rewards/consensus_reward_func": 0.7142857142857143,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5541323167937142,
+      "rewards/question_recreation_reward_func": 0.6577473367963519,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.1073571400982993,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.13060714730194636,
       "step": 16
     },
     {
-      "completion_length": 451.7857142857143,
+      "completion_length": 522.3214285714286,
       "epoch": 6.0,
-      "grad_norm": 0.5214642286300659,
-      "kl": 0.0018680449408878172,
+      "grad_norm": 0.4006715714931488,
+      "kl": 0.002713062268282686,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 2.39990365930966,
-      "reward_std": 2.4969028319631303,
-      "rewards/concensus_correctness_reward_func": 0.8394286004560334,
+      "reward": 1.184859233243125,
+      "reward_std": 1.088052702801568,
+      "rewards/concensus_correctness_reward_func": 0.0,
       "rewards/consensus_reward_func": 0.35714285714285715,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.2857142857142857,
-      "rewards/question_recreation_reward_func": 0.5529751644602844,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5907520651817322,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.3467857028756823,
+      "rewards/strict_format_reward_func": 0.05357142857142857,
+      "rewards/xmlcount_reward_func": 0.1833928610597338,
       "step": 18
     },
     {
-      "completion_length": 535.65625,
+      "completion_length": 505.46875,
       "epoch": 6.7272727272727275,
-      "grad_norm": 0.4509134292602539,
-      "kl": 0.0018071198574034497,
+      "grad_norm": 0.37507492303848267,
+      "kl": 0.0024967321514850482,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 1.5683128461241722,
-      "reward_std": 1.0115205124020576,
-      "rewards/concensus_correctness_reward_func": 0.023625001311302185,
-      "rewards/consensus_reward_func": 0.4375,
+      "reward": 1.5403007566928864,
+      "reward_std": 1.1960295662283897,
+      "rewards/concensus_correctness_reward_func": 0.125,
+      "rewards/consensus_reward_func": 0.3125,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.25,
-      "rewards/question_recreation_reward_func": 0.5450003277510405,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.6121757589280605,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.31218750961124897,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": 0.45937499683350325,
       "step": 20
     },
     {
       "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 2.193293255459139e-06,
-      "train_runtime": 577.4662,
-      "train_samples_per_second": 0.554,
-      "train_steps_per_second": 0.035
+      "train_loss": 2.1295968963386256e-06,
+      "train_runtime": 640.5844,
+      "train_samples_per_second": 0.5,
+      "train_steps_per_second": 0.031
     }
   ],
   "logging_steps": 2,