End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:204c882b54379fa8abce3df2950329e1d4eeccea90480384a4401dc31626349f
+oid sha256:8d400e22463117128c5687405661f6c0c1de57bf15b029d0d86b95dbeef93166
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 6.219378576588497e-07,
-    "train_runtime": 634.3838,
+    "train_loss": 7.993006249762403e-07,
+    "train_runtime": 660.6931,
     "train_samples": 11,
-    "train_samples_per_second": 0.504,
-    "train_steps_per_second": 0.032
+    "train_samples_per_second": 0.484,
+    "train_steps_per_second": 0.03
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 6.219378576588497e-07,
-    "train_runtime": 634.3838,
+    "train_loss": 7.993006249762403e-07,
+    "train_runtime": 660.6931,
     "train_samples": 11,
-    "train_samples_per_second": 0.504,
-    "train_steps_per_second": 0.032
+    "train_samples_per_second": 0.484,
+    "train_steps_per_second": 0.03
 }

trainer_state.json

@@ -10,203 +10,203 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 466.9375,
+      "completion_length": 518.5,
       "epoch": 0.7272727272727273,
-      "grad_norm": 0.4590955376625061,
-      "kl": 0.0006995353105594404,
+      "grad_norm": 0.555176317691803,
+      "kl": 0.0007004447579674888,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 1.0615508595947176,
-      "reward_std": 1.0327275833114982,
-      "rewards/concensus_correctness_reward_func": 0.0,
+      "reward": 0.8274347987025976,
+      "reward_std": 1.1000676304101944,
+      "rewards/concensus_correctness_reward_func": 0.015625,
       "rewards/consensus_reward_func": 0.3125,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5619883388280869,
+      "rewards/question_recreation_reward_func": 0.49177856743335724,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.046875,
-      "rewards/xmlcount_reward_func": 0.14018749445676804,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.007531255949288607,
       "step": 2
     },
     {
-      "completion_length": 527.7857142857143,
+      "completion_length": 380.92857142857144,
       "epoch": 1.3636363636363638,
-      "grad_norm": 0.3488618731498718,
-      "kl": 0.0006287240789138845,
+      "grad_norm": 0.43665412068367004,
+      "kl": 0.0008776169935507434,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 1.063449433871678,
-      "reward_std": 1.1212820240429469,
-      "rewards/concensus_correctness_reward_func": 0.009285713945116316,
-      "rewards/consensus_reward_func": 0.35714285714285715,
+      "reward": 0.7335536139351981,
+      "reward_std": 0.7739071760858808,
+      "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": 0.26569944407258717,
-      "rewards/soft_format_reward_func": 0.017857142857142856,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.41346428649766104,
+      "rewards/question_recreation_reward_func": 0.38558934203216005,
+      "rewards/soft_format_reward_func": 0.0,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.2586785948702267,
       "step": 4
     },
     {
-      "completion_length": 484.60714285714283,
+      "completion_length": 544.2142857142857,
       "epoch": 2.0,
-      "grad_norm": 0.3382681906223297,
-      "kl": 0.0006566861369979701,
+      "grad_norm": 0.31843292713165283,
+      "kl": 0.0007293318132204669,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 0.45333464443683624,
-      "reward_std": 0.9945533658776965,
-      "rewards/concensus_correctness_reward_func": 0.018928570406777517,
-      "rewards/consensus_reward_func": 0.14285714285714285,
+      "reward": 0.8059655938829694,
+      "reward_std": 1.1535072326660156,
+      "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": 0.44633465153830393,
+      "rewards/question_recreation_reward_func": 0.5153227469750813,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.1547857118504388,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.058500008923666816,
       "step": 6
     },
     {
-      "completion_length": 515.65625,
+      "completion_length": 562.4375,
       "epoch": 2.7272727272727275,
-      "grad_norm": 0.38815051317214966,
-      "kl": 0.000585481491725659,
+      "grad_norm": 4.0187530517578125,
+      "kl": 0.0014460544553003274,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 0.9084305539727211,
-      "reward_std": 1.0081006679683924,
+      "reward": 1.044090647250414,
+      "reward_std": 0.9590163063257933,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.375,
+      "rewards/consensus_reward_func": 0.5625,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.3833993040025234,
+      "rewards/question_recreation_reward_func": 0.44065316766500473,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.15003126417286694,
+      "rewards/xmlcount_reward_func": 0.04093749704770744,
       "step": 8
     },
     {
-      "completion_length": 466.39285714285717,
+      "completion_length": 495.75,
       "epoch": 3.3636363636363638,
-      "grad_norm": 0.5429649353027344,
-      "kl": 0.0007274033975720938,
+      "grad_norm": 3.4272704124450684,
+      "kl": 0.000784852820548362,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 0.8720454105309078,
-      "reward_std": 0.9343861000878471,
-      "rewards/concensus_correctness_reward_func": 0.03785714081355503,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "reward": 0.6675052131925311,
+      "reward_std": 1.1034608014992304,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.35714285714285715,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.43379540528569904,
-      "rewards/soft_format_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.42725521538938793,
+      "rewards/soft_format_reward_func": 0.017857142857142856,
       "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.09682144224643707,
+      "rewards/xmlcount_reward_func": -0.15260715250458037,
       "step": 10
     },
     {
-      "completion_length": 495.42857142857144,
+      "completion_length": 512.6428571428571,
       "epoch": 4.0,
-      "grad_norm": 0.2401263564825058,
-      "kl": 0.0006068038305134646,
+      "grad_norm": 0.3071759045124054,
+      "kl": 0.0008274220578771617,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 0.8263501333338874,
-      "reward_std": 1.0283640282494682,
+      "reward": 1.309668013027736,
+      "reward_std": 1.1936028088842119,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.35714285714285715,
+      "rewards/consensus_reward_func": 0.5,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.39677869847842623,
-      "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.0724285798413413,
+      "rewards/question_recreation_reward_func": 0.5594894502844129,
+      "rewards/soft_format_reward_func": 0.017857142857142856,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.21446428767272405,
       "step": 12
     },
     {
-      "completion_length": 440.5625,
+      "completion_length": 550.09375,
       "epoch": 4.7272727272727275,
-      "grad_norm": 1.4231760501861572,
-      "kl": 0.0007035199232632294,
+      "grad_norm": 0.3173958361148834,
+      "kl": 0.0008341810971614905,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 1.0565725192427635,
-      "reward_std": 1.2860574126243591,
-      "rewards/concensus_correctness_reward_func": 0.01656249910593033,
-      "rewards/consensus_reward_func": 0.375,
+      "reward": 0.7159233158454299,
+      "reward_std": 1.0219199433922768,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.125,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.37782250344753265,
+      "rewards/question_recreation_reward_func": 0.4055483331903815,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.27156251296401024,
+      "rewards/xmlcount_reward_func": 0.16975000500679016,
       "step": 14
     },
     {
-      "completion_length": 540.5357142857143,
+      "completion_length": 479.0357142857143,
       "epoch": 5.363636363636363,
-      "grad_norm": 1.7067158222198486,
-      "kl": 0.000782927913990404,
+      "grad_norm": 0.4549387991428375,
+      "kl": 0.0008137404719101531,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 0.7602636175496238,
-      "reward_std": 1.1363796251160758,
+      "reward": 1.1487190595694952,
+      "reward_std": 0.8686993377549308,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "rewards/consensus_reward_func": 0.35714285714285715,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.43315647542476654,
+      "rewards/question_recreation_reward_func": 0.5325047671794891,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.02353573271206447,
+      "rewards/xmlcount_reward_func": 0.2412142881325313,
       "step": 16
     },
     {
-      "completion_length": 508.39285714285717,
+      "completion_length": 450.35714285714283,
       "epoch": 6.0,
-      "grad_norm": 0.33809056878089905,
-      "kl": 0.0006033167924865016,
+      "grad_norm": 0.9689341187477112,
+      "kl": 0.0009170615979071174,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 1.1679890028067998,
-      "reward_std": 1.0983782325472151,
-      "rewards/concensus_correctness_reward_func": 0.018928570406777517,
-      "rewards/consensus_reward_func": 0.5,
+      "reward": 0.4135749212333134,
+      "reward_std": 1.0793337864535195,
+      "rewards/concensus_correctness_reward_func": 0.014714285731315613,
+      "rewards/consensus_reward_func": 0.2857142857142857,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.07142857142857142,
-      "rewards/question_recreation_reward_func": 0.4114890034709658,
-      "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.1661428383418492,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.3206820615700313,
+      "rewards/soft_format_reward_func": 0.017857142857142856,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": -0.2432499953678676,
       "step": 18
     },
     {
-      "completion_length": 471.4375,
+      "completion_length": 572.3125,
       "epoch": 6.7272727272727275,
-      "grad_norm": 0.43950626254081726,
-      "kl": 0.0006802630596212111,
+      "grad_norm": 0.3049357235431671,
+      "kl": 0.0007123384966689628,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 1.1322370530106127,
-      "reward_std": 0.9111562184989452,
-      "rewards/concensus_correctness_reward_func": 0.015625,
-      "rewards/consensus_reward_func": 0.5,
+      "reward": 0.6959485113620758,
+      "reward_std": 1.0689484477043152,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.125,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.45245583169162273,
+      "rewards/question_recreation_reward_func": 0.49332351610064507,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.16415624879300594,
+      "rewards/xmlcount_reward_func": 0.07762499898672104,
       "step": 20
     },
     {
       "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 6.219378576588497e-07,
-      "train_runtime": 634.3838,
-      "train_samples_per_second": 0.504,
-      "train_steps_per_second": 0.032
+      "train_loss": 7.993006249762403e-07,
+      "train_runtime": 660.6931,
+      "train_samples_per_second": 0.484,
+      "train_steps_per_second": 0.03
     }
   ],
   "logging_steps": 2,