End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fa7590fd90bfde3b1335b7433d356b270696d6765a8d4aaee2338a4785cfad02
+oid sha256:44fe9e051dbe4238468b279f1a709f9ac7fa916c0dc84e280a1485614e79d70e
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 7.007378655998764e-07,
-    "train_runtime": 1568.7671,
+    "train_loss": 7.011475133822387e-07,
+    "train_runtime": 1646.5405,
     "train_samples": 11,
-    "train_samples_per_second": 0.204,
-    "train_steps_per_second": 0.013
+    "train_samples_per_second": 0.194,
+    "train_steps_per_second": 0.012
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 7.007378655998764e-07,
-    "train_runtime": 1568.7671,
+    "train_loss": 7.011475133822387e-07,
+    "train_runtime": 1646.5405,
     "train_samples": 11,
-    "train_samples_per_second": 0.204,
-    "train_steps_per_second": 0.013
+    "train_samples_per_second": 0.194,
+    "train_steps_per_second": 0.012
 }

trainer_state.json

@@ -10,203 +10,203 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 479.21875,
+      "completion_length": 497.8125,
       "epoch": 0.7272727272727273,
-      "grad_norm": 0.497830331325531,
-      "kl": 0.0006894778707646765,
+      "grad_norm": 0.5344477891921997,
+      "kl": 0.0006605069211218506,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 0.8361779022961855,
-      "reward_std": 0.9613135643303394,
-      "rewards/concensus_correctness_reward_func": 0.0010000000474974513,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 1.339249111711979,
+      "reward_std": 1.1557906121015549,
+      "rewards/concensus_correctness_reward_func": 0.0886249989271164,
+      "rewards/consensus_reward_func": 0.4375,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5334903616458178,
+      "rewards/final_correctness_reward_func": 0.1875,
+      "rewards/question_recreation_reward_func": 0.46371785760857165,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.05168749764561653,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.14628126472234726,
       "step": 2
     },
     {
-      "completion_length": 331.42857142857144,
+      "completion_length": 377.75,
       "epoch": 1.3636363636363638,
-      "grad_norm": 0.5410258173942566,
-      "kl": 0.0008775076047251267,
+      "grad_norm": 0.3888702392578125,
+      "kl": 0.0007420627932463374,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 1.1655291829790388,
-      "reward_std": 1.250451820237296,
-      "rewards/concensus_correctness_reward_func": 0.03278571353959186,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "reward": 1.165995559522084,
+      "reward_std": 0.7460833532469613,
+      "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": 0.5209220349788666,
+      "rewards/question_recreation_reward_func": 0.5369955897331238,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.05357142857142857,
-      "rewards/xmlcount_reward_func": 0.2725357147199767,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.4861428567341396,
       "step": 4
     },
     {
-      "completion_length": 436.0,
+      "completion_length": 458.2142857142857,
       "epoch": 2.0,
-      "grad_norm": 0.41036298871040344,
-      "kl": 0.0007396088829929275,
+      "grad_norm": 0.2860358655452728,
+      "kl": 0.0007417477567547134,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 1.134751796722412,
-      "reward_std": 1.0689653626510076,
-      "rewards/concensus_correctness_reward_func": 0.06400000197546822,
-      "rewards/consensus_reward_func": 0.21428571428571427,
+      "reward": 1.2034938867603029,
+      "reward_std": 1.1629417070320673,
+      "rewards/concensus_correctness_reward_func": 0.015857143061501638,
+      "rewards/consensus_reward_func": 0.2857142857142857,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.4305018394121102,
+      "rewards/final_correctness_reward_func": 0.07142857142857142,
+      "rewards/question_recreation_reward_func": 0.40042244110788616,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.05357142857142857,
-      "rewards/xmlcount_reward_func": 0.3723928523915155,
+      "rewards/xmlcount_reward_func": 0.37650000410420553,
       "step": 6
     },
     {
-      "completion_length": 388.8125,
+      "completion_length": 388.75,
       "epoch": 2.7272727272727275,
-      "grad_norm": 0.613455593585968,
-      "kl": 0.00076409201574279,
+      "grad_norm": 4.042172908782959,
+      "kl": 0.0010465572268003598,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 1.0549859162420034,
-      "reward_std": 0.9957783780992031,
-      "rewards/concensus_correctness_reward_func": 0.0020000000949949026,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 0.9157519098371267,
+      "reward_std": 1.1724815517663956,
+      "rewards/concensus_correctness_reward_func": 0.045125000178813934,
+      "rewards/consensus_reward_func": 0.375,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5592359164729714,
+      "rewards/question_recreation_reward_func": 0.4620331604965031,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.24375000409781933,
+      "rewards/xmlcount_reward_func": 0.03359374776482582,
       "step": 8
     },
     {
-      "completion_length": 389.2857142857143,
+      "completion_length": 493.60714285714283,
       "epoch": 3.3636363636363638,
-      "grad_norm": 0.4000310003757477,
-      "kl": 0.0007634159077757172,
+      "grad_norm": 0.3916516602039337,
+      "kl": 0.0007648904242419771,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 1.1786940097808838,
-      "reward_std": 0.9912613758019039,
-      "rewards/concensus_correctness_reward_func": 0.045428569827760966,
-      "rewards/consensus_reward_func": 0.35714285714285715,
+      "reward": 1.1886889593941825,
+      "reward_std": 1.203088402748108,
+      "rewards/concensus_correctness_reward_func": 0.05357142857142857,
+      "rewards/consensus_reward_func": 0.07142857142857142,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.41876543100391117,
+      "rewards/final_correctness_reward_func": 0.14285714285714285,
+      "rewards/question_recreation_reward_func": 0.459581800869533,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.05357142857142857,
-      "rewards/xmlcount_reward_func": 0.3037857264280319,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.4433928866471563,
       "step": 10
     },
     {
-      "completion_length": 454.42857142857144,
+      "completion_length": 382.60714285714283,
       "epoch": 4.0,
-      "grad_norm": 0.46324434876441956,
-      "kl": 0.0006662574347241648,
+      "grad_norm": 0.5440981984138489,
+      "kl": 0.000703945814166218,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 1.3409278094768524,
-      "reward_std": 1.3151888506753104,
-      "rewards/concensus_correctness_reward_func": 0.08785714408648866,
-      "rewards/consensus_reward_func": 0.35714285714285715,
+      "reward": 2.2168113525424684,
+      "reward_std": 2.4800874079976762,
+      "rewards/concensus_correctness_reward_func": 0.7758571420397077,
+      "rewards/consensus_reward_func": 0.5714285714285714,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5007849793348994,
+      "rewards/final_correctness_reward_func": 0.07142857142857142,
+      "rewards/question_recreation_reward_func": 0.5068114027380943,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.03571428571428571,
-      "rewards/xmlcount_reward_func": 0.35942857393196653,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.29128572397998403,
       "step": 12
     },
     {
-      "completion_length": 427.84375,
+      "completion_length": 500.5625,
       "epoch": 4.7272727272727275,
-      "grad_norm": 0.36418354511260986,
-      "kl": 0.0008765329112065956,
+      "grad_norm": 0.41056036949157715,
+      "kl": 0.0006735397255397402,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 1.168391801416874,
-      "reward_std": 1.1325892731547356,
-      "rewards/concensus_correctness_reward_func": 0.051187500124797225,
-      "rewards/consensus_reward_func": 0.3125,
+      "reward": 1.2925249487161636,
+      "reward_std": 1.3193300478160381,
+      "rewards/concensus_correctness_reward_func": 0.013875000178813934,
+      "rewards/consensus_reward_func": 0.4375,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0625,
-      "rewards/question_recreation_reward_func": 0.5072980597615242,
+      "rewards/question_recreation_reward_func": 0.5702437162399292,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.03125,
-      "rewards/xmlcount_reward_func": 0.20365623105317354,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.20840624906122684,
       "step": 14
     },
     {
-      "completion_length": 434.5,
+      "completion_length": 454.39285714285717,
       "epoch": 5.363636363636363,
-      "grad_norm": 0.6217504143714905,
-      "kl": 0.0008525813214613923,
+      "grad_norm": 0.5643100738525391,
+      "kl": 0.0006845367315690964,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 0.936353611094611,
-      "reward_std": 0.8693441791193826,
-      "rewards/concensus_correctness_reward_func": 0.022714284913880483,
+      "reward": 0.9351854633007731,
+      "reward_std": 0.9433658080441611,
+      "rewards/concensus_correctness_reward_func": 0.031714286123003275,
       "rewards/consensus_reward_func": 0.14285714285714285,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5036750175058842,
+      "rewards/final_correctness_reward_func": 0.07142857142857142,
+      "rewards/question_recreation_reward_func": 0.5786854858909335,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.07142857142857142,
-      "rewards/xmlcount_reward_func": 0.1956785704408373,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.11050001212528773,
       "step": 16
     },
     {
-      "completion_length": 435.32142857142856,
+      "completion_length": 410.32142857142856,
       "epoch": 6.0,
-      "grad_norm": 0.32569819688796997,
-      "kl": 0.000702967617793807,
+      "grad_norm": 0.3228893280029297,
+      "kl": 0.0008362333506478795,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 1.2448477915355138,
-      "reward_std": 1.0681161028998238,
-      "rewards/concensus_correctness_reward_func": 0.011214285822851318,
-      "rewards/consensus_reward_func": 0.35714285714285715,
+      "reward": 0.9021126500197819,
+      "reward_std": 1.0241698324680328,
+      "rewards/concensus_correctness_reward_func": 0.017857142857142856,
+      "rewards/consensus_reward_func": 0.2857142857142857,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.07142857142857142,
-      "rewards/question_recreation_reward_func": 0.4135977657777922,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.32804122247866224,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.3736071437597275,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.2705000032271658,
       "step": 18
     },
     {
-      "completion_length": 396.90625,
+      "completion_length": 483.6875,
       "epoch": 6.7272727272727275,
-      "grad_norm": 0.40409594774246216,
-      "kl": 0.0007001970298006199,
+      "grad_norm": 0.3410959839820862,
+      "kl": 0.0007752301244181581,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 1.109384085983038,
-      "reward_std": 0.938243392854929,
-      "rewards/concensus_correctness_reward_func": 0.009812500094994903,
-      "rewards/consensus_reward_func": 0.3125,
+      "reward": 1.2597090499475598,
+      "reward_std": 1.0453971847891808,
+      "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": 0.5014465823769569,
+      "rewards/final_correctness_reward_func": 0.125,
+      "rewards/question_recreation_reward_func": 0.49870902951806784,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.27000000420957804,
+      "rewards/xmlcount_reward_func": 0.2453750055283308,
       "step": 20
     },
     {
       "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 7.007378655998764e-07,
-      "train_runtime": 1568.7671,
-      "train_samples_per_second": 0.204,
-      "train_steps_per_second": 0.013
+      "train_loss": 7.011475133822387e-07,
+      "train_runtime": 1646.5405,
+      "train_samples_per_second": 0.194,
+      "train_steps_per_second": 0.012
     }
   ],
   "logging_steps": 2,