End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f9b2cea86f052ae6665697fc4d8835b4f0c9e02b29529451bd6259d6d4eccc95
+oid sha256:363c30094536b554a184c76bcd696b68cd055032abb29f7a5ed5669c70521c2a
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 6.444601922339643e-07,
-    "train_runtime": 1608.3431,
+    "train_loss": 7.198085086201899e-07,
+    "train_runtime": 1584.7401,
     "train_samples": 11,
-    "train_samples_per_second": 0.199,
-    "train_steps_per_second": 0.012
+    "train_samples_per_second": 0.202,
+    "train_steps_per_second": 0.013
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 6.444601922339643e-07,
-    "train_runtime": 1608.3431,
+    "train_loss": 7.198085086201899e-07,
+    "train_runtime": 1584.7401,
     "train_samples": 11,
-    "train_samples_per_second": 0.199,
-    "train_steps_per_second": 0.012
+    "train_samples_per_second": 0.202,
+    "train_steps_per_second": 0.013
 }

trainer_state.json

@@ -10,203 +10,203 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 467.90625,
+      "completion_length": 561.59375,
       "epoch": 0.7272727272727273,
-      "grad_norm": 0.485609233379364,
-      "kl": 0.0007106798075255938,
+      "grad_norm": 0.3222217261791229,
+      "kl": 0.0006156760464364197,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 1.2300263158977032,
-      "reward_std": 1.1693743616342545,
+      "reward": 0.7097736224532127,
+      "reward_std": 1.075104609131813,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.4375,
+      "rewards/consensus_reward_func": 0.125,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.6245888117700815,
+      "rewards/question_recreation_reward_func": 0.5884298821911216,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.16793749667704105,
+      "rewards/strict_format_reward_func": 0.046875,
+      "rewards/xmlcount_reward_func": -0.050531258806586266,
       "step": 2
     },
     {
-      "completion_length": 461.39285714285717,
+      "completion_length": 413.14285714285717,
       "epoch": 1.3636363636363638,
-      "grad_norm": 0.4931257367134094,
-      "kl": 0.0007476569485983678,
+      "grad_norm": 0.39414873719215393,
+      "kl": 0.0007785040652379394,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 0.2825267655508859,
-      "reward_std": 1.0751322656869888,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.14285714285714285,
+      "reward": 0.8490615487098694,
+      "reward_std": 1.1400364807673864,
+      "rewards/concensus_correctness_reward_func": 0.0282857141324452,
+      "rewards/consensus_reward_func": 0.2857142857142857,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.3599196289266859,
+      "rewards/question_recreation_reward_func": 0.4676329472235271,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.22024999346051896,
+      "rewards/xmlcount_reward_func": 0.06742857609476362,
       "step": 4
     },
     {
-      "completion_length": 568.2142857142857,
+      "completion_length": 492.4642857142857,
       "epoch": 2.0,
-      "grad_norm": 0.22118231654167175,
-      "kl": 0.000631775295395138,
+      "grad_norm": 0.3449578285217285,
+      "kl": 0.0006918981796064015,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 1.604168461901801,
-      "reward_std": 1.2598716957228524,
-      "rewards/concensus_correctness_reward_func": 0.054285715733255656,
-      "rewards/consensus_reward_func": 0.6428571428571429,
+      "reward": 0.9568400574581963,
+      "reward_std": 0.7401026104177747,
+      "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": 0.5840612916009766,
-      "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.30510713905096054,
+      "rewards/question_recreation_reward_func": 0.5246257483959198,
+      "rewards/soft_format_reward_func": 0.017857142857142856,
+      "rewards/strict_format_reward_func": 0.03571428571428571,
+      "rewards/xmlcount_reward_func": 0.3072142941611154,
       "step": 6
     },
     {
-      "completion_length": 471.34375,
+      "completion_length": 470.9375,
       "epoch": 2.7272727272727275,
-      "grad_norm": 0.4210323393344879,
-      "kl": 0.0007207773451227695,
+      "grad_norm": 0.39589089155197144,
+      "kl": 0.0007375327404588461,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 1.0432898998260498,
-      "reward_std": 0.9795822128653526,
+      "reward": 0.8644137904047966,
+      "reward_std": 1.089905098080635,
       "rewards/concensus_correctness_reward_func": 0.0,
       "rewards/consensus_reward_func": 0.3125,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.4576649102382362,
-      "rewards/soft_format_reward_func": 0.015625,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.2575000021606684,
+      "rewards/question_recreation_reward_func": 0.5000075455754995,
+      "rewards/soft_format_reward_func": 0.0,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": 0.02065625786781311,
       "step": 8
     },
     {
-      "completion_length": 477.5357142857143,
+      "completion_length": 561.7857142857143,
       "epoch": 3.3636363636363638,
-      "grad_norm": 0.36557313799858093,
-      "kl": 0.0007496414140665106,
+      "grad_norm": 2.330106735229492,
+      "kl": 0.0006754749130258071,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 0.5420836593423571,
-      "reward_std": 0.947016784123012,
-      "rewards/concensus_correctness_reward_func": 0.02257142961025238,
-      "rewards/consensus_reward_func": 0.14285714285714285,
+      "reward": 0.37959482840129305,
+      "reward_std": 0.7886925105537687,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.37286937875407083,
+      "rewards/final_correctness_reward_func": 0.07142857142857142,
+      "rewards/question_recreation_reward_func": 0.40823769356523243,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.003785711313996996,
+      "rewards/xmlcount_reward_func": -0.10007142807756152,
       "step": 10
     },
     {
-      "completion_length": 544.0714285714286,
+      "completion_length": 534.4642857142857,
       "epoch": 4.0,
-      "grad_norm": 0.4929836690425873,
-      "kl": 0.0007356132014787623,
+      "grad_norm": 0.2411220520734787,
+      "kl": 0.0007088499343288797,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 1.1885977642876762,
-      "reward_std": 1.199118435382843,
+      "reward": 0.9174802005290985,
+      "reward_std": 0.8755575588771275,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.42857142857142855,
+      "rewards/consensus_reward_func": 0.21428571428571427,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.496383473277092,
+      "rewards/question_recreation_reward_func": 0.5952659291880471,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.26364285392420633,
+      "rewards/strict_format_reward_func": 0.03571428571428571,
+      "rewards/xmlcount_reward_func": 0.07221428198473794,
       "step": 12
     },
     {
-      "completion_length": 503.9375,
+      "completion_length": 612.6875,
       "epoch": 4.7272727272727275,
-      "grad_norm": 0.4013039171695709,
-      "kl": 0.0006397504039341584,
+      "grad_norm": 0.26711583137512207,
+      "kl": 0.0006393189869413618,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 0.9641091949306428,
-      "reward_std": 0.9753384664654732,
-      "rewards/concensus_correctness_reward_func": 0.05925000458955765,
-      "rewards/consensus_reward_func": 0.4375,
+      "reward": 0.5164269562810659,
+      "reward_std": 1.0301097724586725,
+      "rewards/concensus_correctness_reward_func": 0.02474999986588955,
+      "rewards/consensus_reward_func": 0.1875,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.397171713411808,
+      "rewards/final_correctness_reward_func": 0.0625,
+      "rewards/question_recreation_reward_func": 0.4383957292884588,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.07018749974668026,
+      "rewards/xmlcount_reward_func": -0.1967187589034438,
       "step": 14
     },
     {
-      "completion_length": 563.1428571428571,
+      "completion_length": 530.25,
       "epoch": 5.363636363636363,
-      "grad_norm": 0.5356454253196716,
-      "kl": 0.0006699909738797162,
+      "grad_norm": 0.2758399248123169,
+      "kl": 0.0006760963130675789,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 0.7650677136012486,
-      "reward_std": 0.9540353034223829,
+      "reward": 0.7838593018906457,
+      "reward_std": 1.0056123180048806,
       "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": 0.5341748382364001,
+      "rewards/question_recreation_reward_func": 0.5464664612497602,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.15946428797074727,
+      "rewards/xmlcount_reward_func": 0.16596427985600062,
       "step": 16
     },
     {
-      "completion_length": 521.3928571428571,
+      "completion_length": 423.67857142857144,
       "epoch": 6.0,
-      "grad_norm": 0.5471972227096558,
-      "kl": 0.0007351575976437223,
+      "grad_norm": 0.3440936803817749,
+      "kl": 0.0007996462684656892,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 1.1135373541287013,
-      "reward_std": 1.0160424539021082,
-      "rewards/concensus_correctness_reward_func": 0.04514285922050476,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "reward": 0.6838892570563725,
+      "reward_std": 0.9769647972924369,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.21428571428571427,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5741445156080383,
+      "rewards/final_correctness_reward_func": 0.07142857142857142,
+      "rewards/question_recreation_reward_func": 0.4169606491923332,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.03571428571428571,
-      "rewards/xmlcount_reward_func": 0.17282143022332871,
+      "rewards/xmlcount_reward_func": -0.054499996559960504,
       "step": 18
     },
     {
-      "completion_length": 502.125,
+      "completion_length": 449.3125,
       "epoch": 6.7272727272727275,
-      "grad_norm": 0.4341804087162018,
-      "kl": 0.0006616544378630351,
+      "grad_norm": 0.476371705532074,
+      "kl": 0.0014172161172609776,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 1.3489604517817497,
-      "reward_std": 0.9606379866600037,
-      "rewards/concensus_correctness_reward_func": 0.05925000458955765,
-      "rewards/consensus_reward_func": 0.5625,
+      "reward": 0.9342581760138273,
+      "reward_std": 0.7589832246303558,
+      "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": 0.5194604564458132,
-      "rewards/soft_format_reward_func": 0.015625,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.19212500751018524,
+      "rewards/final_correctness_reward_func": 0.0625,
+      "rewards/question_recreation_reward_func": 0.5328519130125642,
+      "rewards/soft_format_reward_func": 0.0,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.0732812499627471,
       "step": 20
     },
     {
       "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 6.444601922339643e-07,
-      "train_runtime": 1608.3431,
-      "train_samples_per_second": 0.199,
-      "train_steps_per_second": 0.012
+      "train_loss": 7.198085086201899e-07,
+      "train_runtime": 1584.7401,
+      "train_samples_per_second": 0.202,
+      "train_steps_per_second": 0.013
     }
   ],
   "logging_steps": 2,