End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:65ce21fef9d7dca68bcddcd5e892adcd2982504af98c531975fd508708db2202
+oid sha256:795d6e022daec5a365cfd2ad82e799c7c4be105ea7594f507d81eba5e20695e8
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 1.760407212714199e-06,
-    "train_runtime": 560.1249,
+    "train_loss": 2.1270964680297764e-06,
+    "train_runtime": 526.1289,
     "train_samples": 11,
-    "train_samples_per_second": 0.571,
-    "train_steps_per_second": 0.036
+    "train_samples_per_second": 0.608,
+    "train_steps_per_second": 0.038
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 1.760407212714199e-06,
-    "train_runtime": 560.1249,
+    "train_loss": 2.1270964680297764e-06,
+    "train_runtime": 526.1289,
     "train_samples": 11,
-    "train_samples_per_second": 0.571,
-    "train_steps_per_second": 0.036
+    "train_samples_per_second": 0.608,
+    "train_steps_per_second": 0.038
 }

trainer_state.json

@@ -10,203 +10,203 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 495.34375,
+      "completion_length": 390.28125,
       "epoch": 0.7272727272727273,
-      "grad_norm": 0.3851245939731598,
-      "kl": 0.0015201627466012724,
+      "grad_norm": 0.46063151955604553,
+      "kl": 0.0021709811844630167,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 1.23531174659729,
-      "reward_std": 1.1557490825653076,
-      "rewards/concensus_correctness_reward_func": 0.0008750000270083547,
-      "rewards/consensus_reward_func": 0.375,
+      "reward": 1.1787039148621261,
+      "reward_std": 0.7971708327531815,
+      "rewards/concensus_correctness_reward_func": 0.005375000182539225,
+      "rewards/consensus_reward_func": 0.25,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0625,
-      "rewards/question_recreation_reward_func": 0.6263742372393608,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5472664330154657,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.1705624982714653,
+      "rewards/xmlcount_reward_func": 0.3760625086724758,
       "step": 2
     },
     {
-      "completion_length": 362.7142857142857,
+      "completion_length": 400.64285714285717,
       "epoch": 1.3636363636363638,
-      "grad_norm": 0.46032318472862244,
-      "kl": 0.0023172974719532897,
+      "grad_norm": 0.9065832495689392,
+      "kl": 0.0019683726963454057,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 1.392972162791661,
-      "reward_std": 1.4584640605109078,
-      "rewards/concensus_correctness_reward_func": 0.14257142799241201,
-      "rewards/consensus_reward_func": 0.5,
+      "reward": 0.7512879967689514,
+      "reward_std": 0.9332555000271116,
+      "rewards/concensus_correctness_reward_func": 0.0014285716627325331,
+      "rewards/consensus_reward_func": 0.21428571428571427,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.3609007064785276,
+      "rewards/question_recreation_reward_func": 0.3517880014010838,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.3895000049046108,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.1659285788025175,
       "step": 4
     },
     {
-      "completion_length": 377.7857142857143,
+      "completion_length": 367.07142857142856,
       "epoch": 2.0,
-      "grad_norm": 3.0854783058166504,
-      "kl": 0.00172044662758708,
+      "grad_norm": 0.521306037902832,
+      "kl": 0.0022118387111861792,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 0.6036367884704045,
-      "reward_std": 0.5871127205235618,
+      "reward": 1.1688547900744848,
+      "reward_std": 0.7610821894236973,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_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": 0.5374582058617047,
+      "rewards/question_recreation_reward_func": 0.6109976725918906,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.0661785677075386,
+      "rewards/xmlcount_reward_func": 0.3435714329992022,
       "step": 6
     },
     {
-      "completion_length": 432.21875,
+      "completion_length": 403.28125,
       "epoch": 2.7272727272727275,
-      "grad_norm": 0.4442577660083771,
-      "kl": 0.0016866580190253444,
+      "grad_norm": 0.4673061668872833,
+      "kl": 0.002159006107831374,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 1.3614656776189804,
-      "reward_std": 1.080071672797203,
-      "rewards/concensus_correctness_reward_func": 0.03099999949336052,
-      "rewards/consensus_reward_func": 0.4375,
+      "reward": 0.89167794957757,
+      "reward_std": 1.1015209294855595,
+      "rewards/concensus_correctness_reward_func": 0.036625000182539225,
+      "rewards/consensus_reward_func": 0.1875,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5839969431981444,
+      "rewards/question_recreation_reward_func": 0.5902717527933419,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.30896875262260437,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.06165625900030136,
       "step": 8
     },
     {
-      "completion_length": 349.7142857142857,
+      "completion_length": 390.89285714285717,
       "epoch": 3.3636363636363638,
-      "grad_norm": 0.5599241852760315,
-      "kl": 0.0020511089491524865,
+      "grad_norm": 0.5723989605903625,
+      "kl": 0.0016518193796010955,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 1.660039212022509,
-      "reward_std": 1.1306958198547363,
-      "rewards/concensus_correctness_reward_func": 0.30899999822889057,
-      "rewards/consensus_reward_func": 0.5714285714285714,
+      "reward": 0.9222419304507119,
+      "reward_std": 1.0598461883408683,
+      "rewards/concensus_correctness_reward_func": 0.05135714103068624,
+      "rewards/consensus_reward_func": 0.42857142857142855,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.34428915807179045,
+      "rewards/question_recreation_reward_func": 0.4380633703299931,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.03571428571428571,
-      "rewards/xmlcount_reward_func": 0.39960715600422453,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.004250013402530125,
       "step": 10
     },
     {
-      "completion_length": 495.92857142857144,
+      "completion_length": 402.75,
       "epoch": 4.0,
-      "grad_norm": 0.38454505801200867,
-      "kl": 0.0019243384096106248,
+      "grad_norm": 0.5822679400444031,
+      "kl": 0.0021718702711431043,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 1.0013103996004378,
-      "reward_std": 0.9705695403473718,
-      "rewards/concensus_correctness_reward_func": 0.12135714292526245,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "reward": 0.867274991103581,
+      "reward_std": 0.5567212679556438,
+      "rewards/concensus_correctness_reward_func": -0.0023571428443704334,
+      "rewards/consensus_reward_func": 0.07142857142857142,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5770603716373444,
-      "rewards/soft_format_reward_func": 0.017857142857142856,
+      "rewards/question_recreation_reward_func": 0.546417840889522,
+      "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.000678571207182748,
+      "rewards/xmlcount_reward_func": 0.25178571896893637,
       "step": 12
     },
     {
-      "completion_length": 403.53125,
+      "completion_length": 440.25,
       "epoch": 4.7272727272727275,
-      "grad_norm": 0.4447067677974701,
-      "kl": 0.0018273314417456277,
+      "grad_norm": 47.66981506347656,
+      "kl": 0.0034808746131602675,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 1.136631891131401,
-      "reward_std": 0.8802363304421306,
-      "rewards/concensus_correctness_reward_func": 0.07368750125169754,
-      "rewards/consensus_reward_func": 0.375,
+      "reward": 0.7478947602212429,
+      "reward_std": 0.9147473722696304,
+      "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": 0.37563188280910254,
+      "rewards/question_recreation_reward_func": 0.4533947594463825,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.3123125210404396,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.15387500450015068,
       "step": 14
     },
     {
-      "completion_length": 439.35714285714283,
+      "completion_length": 426.2142857142857,
       "epoch": 5.363636363636363,
-      "grad_norm": 0.4888899028301239,
-      "kl": 0.0017850026072535133,
+      "grad_norm": 0.5740916728973389,
+      "kl": 0.0023087138875520657,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 0.7386083092008319,
-      "reward_std": 0.9949612553630557,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.07142857142857142,
+      "reward": 0.937805770763329,
+      "reward_std": 0.6349426976272038,
+      "rewards/concensus_correctness_reward_func": 0.03571428571428571,
+      "rewards/consensus_reward_func": 0.14285714285714285,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.4658940147076334,
-      "rewards/soft_format_reward_func": 0.017857142857142856,
+      "rewards/question_recreation_reward_func": 0.45016290673187803,
+      "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.16557144692965917,
+      "rewards/xmlcount_reward_func": 0.29121429632817,
       "step": 16
     },
     {
-      "completion_length": 390.10714285714283,
+      "completion_length": 402.64285714285717,
       "epoch": 6.0,
-      "grad_norm": 0.4372193217277527,
-      "kl": 0.0022854185164240854,
+      "grad_norm": 0.6819967031478882,
+      "kl": 0.0021416779574272887,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 1.2345900194985526,
-      "reward_std": 1.5980539066450936,
-      "rewards/concensus_correctness_reward_func": 0.11921428782599312,
-      "rewards/consensus_reward_func": 0.42857142857142855,
+      "reward": 1.0460299679211207,
+      "reward_std": 0.9950524738856724,
+      "rewards/concensus_correctness_reward_func": 0.04185714306575911,
+      "rewards/consensus_reward_func": 0.14285714285714285,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.49841142126492094,
+      "rewards/question_recreation_reward_func": 0.48249426909855436,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.18839287012815475,
+      "rewards/xmlcount_reward_func": 0.37882144323417116,
       "step": 18
     },
     {
-      "completion_length": 411.34375,
+      "completion_length": 382.25,
       "epoch": 6.7272727272727275,
-      "grad_norm": 0.7081403136253357,
-      "kl": 0.0019678706958075054,
+      "grad_norm": 0.4199608266353607,
+      "kl": 0.002551245881477371,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 0.8901518955826759,
-      "reward_std": 0.9801672622561455,
-      "rewards/concensus_correctness_reward_func": 0.04268750175833702,
-      "rewards/consensus_reward_func": 0.125,
+      "reward": 0.9639232084155083,
+      "reward_std": 0.9224265478551388,
+      "rewards/concensus_correctness_reward_func": 0.023499999195337296,
+      "rewards/consensus_reward_func": 0.1875,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.4092768579721451,
+      "rewards/question_recreation_reward_func": 0.46442320942878723,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.29756249696947634,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.2884999979287386,
       "step": 20
     },
     {
       "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 1.760407212714199e-06,
-      "train_runtime": 560.1249,
-      "train_samples_per_second": 0.571,
-      "train_steps_per_second": 0.036
+      "train_loss": 2.1270964680297764e-06,
+      "train_runtime": 526.1289,
+      "train_samples_per_second": 0.608,
+      "train_steps_per_second": 0.038
     }
   ],
   "logging_steps": 2,