End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7fcd3b9b1c9c9768d486b7760202b5bbd038d77de16f063a8908737019a83e2f
+oid sha256:e1728362b9231b39a86fabbcf8b3429f649f7197724fb04438c2b8f86f576410
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 1.4845098007754133e-06,
-    "train_runtime": 531.6911,
+    "train_loss": 1.7891535435410333e-06,
+    "train_runtime": 610.1098,
     "train_samples": 11,
-    "train_samples_per_second": 0.602,
-    "train_steps_per_second": 0.038
+    "train_samples_per_second": 0.524,
+    "train_steps_per_second": 0.033
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 1.4845098007754133e-06,
-    "train_runtime": 531.6911,
+    "train_loss": 1.7891535435410333e-06,
+    "train_runtime": 610.1098,
     "train_samples": 11,
-    "train_samples_per_second": 0.602,
-    "train_steps_per_second": 0.038
+    "train_samples_per_second": 0.524,
+    "train_steps_per_second": 0.033
 }

trainer_state.json

@@ -10,203 +10,203 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 364.9375,
+      "completion_length": 432.3125,
       "epoch": 0.7272727272727273,
-      "grad_norm": 0.4663001000881195,
-      "kl": 0.0017626049666432664,
+      "grad_norm": 0.41115620732307434,
+      "kl": 0.001654699444770813,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 1.173146827146411,
-      "reward_std": 0.9561699032783508,
-      "rewards/concensus_correctness_reward_func": 0.09787499904632568,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 1.0831282511353493,
+      "reward_std": 1.0800125412642956,
+      "rewards/concensus_correctness_reward_func": 0.06375000067055225,
+      "rewards/consensus_reward_func": 0.4375,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.25,
-      "rewards/question_recreation_reward_func": 0.37455306435003877,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.493065744638443,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.03125,
-      "rewards/xmlcount_reward_func": 0.1694687455892563,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.08881250536069274,
       "step": 2
     },
     {
-      "completion_length": 498.75,
+      "completion_length": 495.32142857142856,
       "epoch": 1.3636363636363638,
-      "grad_norm": 0.39783263206481934,
-      "kl": 0.001358998814664249,
+      "grad_norm": 0.4448104798793793,
+      "kl": 0.001431963880479868,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 1.123238938195365,
-      "reward_std": 1.1782569800104414,
-      "rewards/concensus_correctness_reward_func": 0.03664285796029227,
-      "rewards/consensus_reward_func": 0.5,
+      "reward": 0.7858572985444751,
+      "reward_std": 0.8992064595222473,
+      "rewards/concensus_correctness_reward_func": 0.04171428403684071,
+      "rewards/consensus_reward_func": 0.07142857142857142,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5083817797047752,
+      "rewards/question_recreation_reward_func": 0.5004644564219883,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.06035715075475829,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.17224999623639242,
       "step": 4
     },
     {
-      "completion_length": 409.7857142857143,
+      "completion_length": 427.42857142857144,
       "epoch": 2.0,
-      "grad_norm": 0.4471225440502167,
-      "kl": 0.0017250077549501189,
+      "grad_norm": 0.35848864912986755,
+      "kl": 0.0016491883101740054,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 3.185895102364676,
-      "reward_std": 2.5625331572123935,
-      "rewards/concensus_correctness_reward_func": 1.423785714166505,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "reward": 1.1533292021070207,
+      "reward_std": 0.9785437073026385,
+      "rewards/concensus_correctness_reward_func": 0.00964285752602986,
+      "rewards/consensus_reward_func": 0.35714285714285715,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.35714285714285715,
-      "rewards/question_recreation_reward_func": 0.5010379723140171,
+      "rewards/final_correctness_reward_func": 0.07142857142857142,
+      "rewards/question_recreation_reward_func": 0.5021148749760219,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.03571428571428571,
-      "rewards/xmlcount_reward_func": 0.5825000064713615,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.2130000085702964,
       "step": 6
     },
     {
-      "completion_length": 453.375,
+      "completion_length": 396.03125,
       "epoch": 2.7272727272727275,
-      "grad_norm": 0.4004893898963928,
-      "kl": 0.0012358660096651874,
+      "grad_norm": 0.4652974009513855,
+      "kl": 0.0018128370720660314,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 1.0723718777298927,
-      "reward_std": 1.1871207617223263,
-      "rewards/concensus_correctness_reward_func": -0.004187500104308128,
-      "rewards/consensus_reward_func": 0.4375,
+      "reward": 1.1249517127871513,
+      "reward_std": 0.972180338576436,
+      "rewards/concensus_correctness_reward_func": 0.0426874989643693,
+      "rewards/consensus_reward_func": 0.25,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.125,
-      "rewards/question_recreation_reward_func": 0.531996876001358,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5647642109543085,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.017937500029802322,
+      "rewards/strict_format_reward_func": 0.046875,
+      "rewards/xmlcount_reward_func": 0.2206249968148768,
       "step": 8
     },
     {
-      "completion_length": 419.64285714285717,
+      "completion_length": 507.0357142857143,
       "epoch": 3.3636363636363638,
-      "grad_norm": 0.5154035687446594,
-      "kl": 0.0019263403207462812,
+      "grad_norm": 0.4943205416202545,
+      "kl": 0.0017367783168862974,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 1.688667967915535,
-      "reward_std": 0.8806300418717521,
-      "rewards/concensus_correctness_reward_func": 0.06171428731509617,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "reward": 1.1879901460238866,
+      "reward_std": 1.009871048586709,
+      "rewards/concensus_correctness_reward_func": 0.07971428547586713,
+      "rewards/consensus_reward_func": 0.21428571428571427,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.5714285714285714,
-      "rewards/question_recreation_reward_func": 0.5533822476863861,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5498829739434379,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.19857142014162882,
+      "rewards/strict_format_reward_func": 0.03571428571428571,
+      "rewards/xmlcount_reward_func": 0.30839285786662785,
       "step": 10
     },
     {
-      "completion_length": 511.57142857142856,
+      "completion_length": 424.85714285714283,
       "epoch": 4.0,
-      "grad_norm": 0.27420559525489807,
-      "kl": 0.0013052445137873292,
+      "grad_norm": 0.541883111000061,
+      "kl": 0.0019839617057836483,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 0.343838948224272,
-      "reward_std": 1.0978365540504456,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.07142857142857142,
+      "reward": 1.0200607010296412,
+      "reward_std": 0.9637793579271862,
+      "rewards/concensus_correctness_reward_func": 0.0825000022138868,
+      "rewards/consensus_reward_func": 0.42857142857142855,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.46316037114177433,
+      "rewards/question_recreation_reward_func": 0.4172749881233488,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.19074999221733638,
+      "rewards/xmlcount_reward_func": 0.09171428637845176,
       "step": 12
     },
     {
-      "completion_length": 344.1875,
+      "completion_length": 405.125,
       "epoch": 4.7272727272727275,
-      "grad_norm": 0.5283384919166565,
-      "kl": 0.0016884810465853661,
+      "grad_norm": 0.7639334797859192,
+      "kl": 0.003259157674619928,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 3.800176238641143,
-      "reward_std": 3.459130048751831,
-      "rewards/concensus_correctness_reward_func": 2.5270000007003546,
-      "rewards/consensus_reward_func": 0.4375,
+      "reward": 1.3091236352920532,
+      "reward_std": 1.4509413577616215,
+      "rewards/concensus_correctness_reward_func": 0.1508749984204769,
+      "rewards/consensus_reward_func": 0.375,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.1875,
-      "rewards/question_recreation_reward_func": 0.5017388071864843,
-      "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.13081249222159386,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.44602989312261343,
+      "rewards/soft_format_reward_func": 0.015625,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": 0.2903437493368983,
       "step": 14
     },
     {
-      "completion_length": 438.60714285714283,
+      "completion_length": 403.42857142857144,
       "epoch": 5.363636363636363,
-      "grad_norm": 0.4809930920600891,
-      "kl": 0.0013627653658789182,
+      "grad_norm": 0.4199073314666748,
+      "kl": 0.0016617131186649203,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 1.5581893447254385,
-      "reward_std": 2.1124278477260043,
-      "rewards/concensus_correctness_reward_func": 0.8574285901018551,
-      "rewards/consensus_reward_func": 0.14285714285714285,
+      "reward": 0.9939025385039193,
+      "reward_std": 1.1395008031811034,
+      "rewards/concensus_correctness_reward_func": 0.05021428742579052,
+      "rewards/consensus_reward_func": 0.35714285714285715,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.07142857142857142,
-      "rewards/question_recreation_reward_func": 0.3227250405720302,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.3302239828876087,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.16374999284744263,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.23846428734915598,
       "step": 16
     },
     {
-      "completion_length": 492.67857142857144,
+      "completion_length": 414.17857142857144,
       "epoch": 6.0,
-      "grad_norm": 0.4203881025314331,
-      "kl": 0.0020158881255026373,
+      "grad_norm": 0.9216535091400146,
+      "kl": 0.0019826509690444383,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 2.326434467520033,
-      "reward_std": 2.6343056857585907,
-      "rewards/concensus_correctness_reward_func": 0.8878571433680398,
-      "rewards/consensus_reward_func": 0.35714285714285715,
+      "reward": 0.7129433985267367,
+      "reward_std": 0.8727979681321553,
+      "rewards/concensus_correctness_reward_func": 0.01928571505205972,
+      "rewards/consensus_reward_func": 0.2857142857142857,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.2857142857142857,
-      "rewards/question_recreation_reward_func": 0.6165773357663836,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.37819340346114977,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.17914285830089025,
+      "rewards/xmlcount_reward_func": 0.0297500022820064,
       "step": 18
     },
     {
-      "completion_length": 381.59375,
+      "completion_length": 460.0,
       "epoch": 6.7272727272727275,
-      "grad_norm": 0.49792689085006714,
-      "kl": 0.0017203347160830162,
+      "grad_norm": 0.48125284910202026,
+      "kl": 0.001955981715582311,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 1.8499253150075674,
-      "reward_std": 2.047252554446459,
-      "rewards/concensus_correctness_reward_func": 0.7549374997615814,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 1.39322067797184,
+      "reward_std": 1.2507520020008087,
+      "rewards/concensus_correctness_reward_func": 0.08237500116229057,
+      "rewards/consensus_reward_func": 0.5,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.125,
-      "rewards/question_recreation_reward_func": 0.41011281311511993,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5450331903994083,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.3098749965429306,
+      "rewards/xmlcount_reward_func": 0.26581248827278614,
       "step": 20
     },
     {
       "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 1.4845098007754133e-06,
-      "train_runtime": 531.6911,
-      "train_samples_per_second": 0.602,
-      "train_steps_per_second": 0.038
+      "train_loss": 1.7891535435410333e-06,
+      "train_runtime": 610.1098,
+      "train_samples_per_second": 0.524,
+      "train_steps_per_second": 0.033
     }
   ],
   "logging_steps": 2,