End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cd2902d26acc7b4cff3552506fdc9a4114787211e7cbd63a03bb054a42c87baa
+oid sha256:ac8f97193e7d1baad5f9e22e97feb9d29168e048d11bb78a796abd3ac92ab77e
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 8.949041671257873e-07,
-    "train_runtime": 663.6389,
+    "train_loss": 9.848059619343985e-07,
+    "train_runtime": 706.4449,
     "train_samples": 11,
-    "train_samples_per_second": 0.482,
-    "train_steps_per_second": 0.03
+    "train_samples_per_second": 0.453,
+    "train_steps_per_second": 0.028
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 8.949041671257873e-07,
-    "train_runtime": 663.6389,
+    "train_loss": 9.848059619343985e-07,
+    "train_runtime": 706.4449,
     "train_samples": 11,
-    "train_samples_per_second": 0.482,
-    "train_steps_per_second": 0.03
+    "train_samples_per_second": 0.453,
+    "train_steps_per_second": 0.028
 }

trainer_state.json

@@ -10,203 +10,203 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 465.59375,
+      "completion_length": 504.25,
       "epoch": 0.7272727272727273,
-      "grad_norm": 0.5124723315238953,
-      "kl": 0.0009230868017766625,
+      "grad_norm": 0.6294240951538086,
+      "kl": 0.0008423507533734664,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 1.036515474319458,
-      "reward_std": 1.5581095591187477,
-      "rewards/concensus_correctness_reward_func": 0.027374999597668648,
-      "rewards/consensus_reward_func": 0.4375,
+      "reward": 1.2705396972596645,
+      "reward_std": 0.8699165377765894,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.1875,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0625,
-      "rewards/question_recreation_reward_func": 0.6265154760330915,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.6449146866798401,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": -0.132999992929399,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": 0.406875004991889,
       "step": 2
     },
     {
-      "completion_length": 423.0,
+      "completion_length": 460.60714285714283,
       "epoch": 1.3636363636363638,
-      "grad_norm": 0.5446125864982605,
-      "kl": 0.0010588045032428844,
+      "grad_norm": 0.39938339591026306,
+      "kl": 0.0009498974208586983,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 1.0730863894735063,
-      "reward_std": 1.0504036588328225,
-      "rewards/concensus_correctness_reward_func": 0.036857144108840396,
-      "rewards/consensus_reward_func": 0.42857142857142855,
+      "reward": 1.149305330855506,
+      "reward_std": 1.067141784088952,
+      "rewards/concensus_correctness_reward_func": 0.0497857140643256,
+      "rewards/consensus_reward_func": 0.2857142857142857,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5810149524893079,
+      "rewards/question_recreation_reward_func": 0.6258052842957633,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.00878570122378213,
+      "rewards/strict_format_reward_func": 0.03571428571428571,
+      "rewards/xmlcount_reward_func": 0.15228570784841264,
       "step": 4
     },
     {
-      "completion_length": 681.2857142857143,
+      "completion_length": 584.25,
       "epoch": 2.0,
-      "grad_norm": 4.352319240570068,
-      "kl": 0.0014373474529877836,
+      "grad_norm": 1.4154448509216309,
+      "kl": 0.0009216495839479778,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 0.9819575761045728,
-      "reward_std": 0.9538520510707583,
+      "reward": 0.5566322782209941,
+      "reward_std": 1.250387300338064,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.42857142857142855,
+      "rewards/consensus_reward_func": 0.14285714285714285,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5134575941732952,
+      "rewards/final_correctness_reward_func": 0.07142857142857142,
+      "rewards/question_recreation_reward_func": 0.48673940343516214,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.03992855229548046,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": -0.16224999300071172,
       "step": 6
     },
     {
-      "completion_length": 504.125,
+      "completion_length": 540.21875,
       "epoch": 2.7272727272727275,
-      "grad_norm": 0.5981078743934631,
-      "kl": 0.0008505457299179398,
+      "grad_norm": 0.7249035239219666,
+      "kl": 0.0008380638319067657,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 0.9509530607610941,
-      "reward_std": 1.2510951459407806,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 1.1796844955533743,
+      "reward_std": 1.1831015311181545,
+      "rewards/concensus_correctness_reward_func": 0.03125,
+      "rewards/consensus_reward_func": 0.5,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0625,
-      "rewards/question_recreation_reward_func": 0.5736718215048313,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.6402782127261162,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.06478124856948853,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": -0.007468747906386852,
       "step": 8
     },
     {
-      "completion_length": 557.8214285714286,
+      "completion_length": 617.6428571428571,
       "epoch": 3.3636363636363638,
-      "grad_norm": 3.4286811351776123,
-      "kl": 0.0011433123561021472,
+      "grad_norm": 0.4157549738883972,
+      "kl": 0.0008354247339801597,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 1.5194084729467119,
-      "reward_std": 1.0138834714889526,
+      "reward": 1.2782049434525626,
+      "reward_std": 1.1169604999678475,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.7857142857142857,
+      "rewards/consensus_reward_func": 0.42857142857142855,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.14285714285714285,
-      "rewards/question_recreation_reward_func": 0.6914084553718567,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.597062064068658,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.10057141552014011,
+      "rewards/strict_format_reward_func": 0.03571428571428571,
+      "rewards/xmlcount_reward_func": 0.21685715871197836,
       "step": 10
     },
     {
-      "completion_length": 479.39285714285717,
+      "completion_length": 592.6428571428571,
       "epoch": 4.0,
-      "grad_norm": 0.37685057520866394,
-      "kl": 0.0010118477927920008,
+      "grad_norm": 3.815882682800293,
+      "kl": 0.0014564325267981207,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 0.9312462338379451,
-      "reward_std": 1.1156473415238517,
-      "rewards/concensus_correctness_reward_func": 0.043571428528853824,
-      "rewards/consensus_reward_func": 0.21428571428571427,
+      "reward": 0.8755534759589604,
+      "reward_std": 1.5114734172821045,
+      "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": 0.5092462343829018,
+      "rewards/question_recreation_reward_func": 0.5599820060389382,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.14628572123391287,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": -0.04157141702515738,
       "step": 12
     },
     {
-      "completion_length": 512.5,
+      "completion_length": 576.09375,
       "epoch": 4.7272727272727275,
-      "grad_norm": 0.41625791788101196,
-      "kl": 0.00083249084855197,
+      "grad_norm": 0.4694238305091858,
+      "kl": 0.0019940473794122227,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 0.9675329588353634,
-      "reward_std": 1.2559568583965302,
+      "reward": 1.0448056980967522,
+      "reward_std": 1.276182696223259,
       "rewards/concensus_correctness_reward_func": 0.0,
       "rewards/consensus_reward_func": 0.4375,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5535642243921757,
+      "rewards/question_recreation_reward_func": 0.6383682060986757,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": -0.03915624879300594,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": -0.03106249589473009,
       "step": 14
     },
     {
-      "completion_length": 441.0,
+      "completion_length": 582.3214285714286,
       "epoch": 5.363636363636363,
-      "grad_norm": 4.590785980224609,
-      "kl": 0.0008904532795505864,
+      "grad_norm": 0.788042426109314,
+      "kl": 0.0007741098192387394,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 1.1700707503727503,
-      "reward_std": 0.806686418397086,
+      "reward": 0.9345985140119281,
+      "reward_std": 0.7620645059006554,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.35714285714285715,
+      "rewards/consensus_reward_func": 0.07142857142857142,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.6240707827465874,
+      "rewards/question_recreation_reward_func": 0.5092770840440478,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.1710000102009092,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.3538928670542581,
       "step": 16
     },
     {
-      "completion_length": 515.4642857142857,
+      "completion_length": 601.0714285714286,
       "epoch": 6.0,
-      "grad_norm": 0.49765175580978394,
-      "kl": 0.0008838894926676792,
+      "grad_norm": 0.3955620229244232,
+      "kl": 0.0008603796637284436,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 1.1322413640362876,
-      "reward_std": 1.2885967322758265,
+      "reward": 1.498078465461731,
+      "reward_std": 1.03599237544196,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.35714285714285715,
+      "rewards/consensus_reward_func": 0.2857142857142857,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.07142857142857142,
-      "rewards/question_recreation_reward_func": 0.7196342263902936,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.669364162853786,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": -0.033821433782577515,
+      "rewards/xmlcount_reward_func": 0.5251428782939911,
       "step": 18
     },
     {
-      "completion_length": 554.21875,
+      "completion_length": 555.21875,
       "epoch": 6.7272727272727275,
-      "grad_norm": 0.38015037775039673,
-      "kl": 0.0008534809603588656,
+      "grad_norm": 0.38449040055274963,
+      "kl": 0.000928697467315942,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 1.5363270118832588,
-      "reward_std": 1.0051772482693195,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.5625,
+      "reward": 1.5113182216882706,
+      "reward_std": 1.1425050217658281,
+      "rewards/concensus_correctness_reward_func": 0.03125,
+      "rewards/consensus_reward_func": 0.375,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0625,
-      "rewards/question_recreation_reward_func": 0.6325770374387503,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.6907557286322117,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.26312499307096004,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": 0.3830625033006072,
       "step": 20
     },
     {
       "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 8.949041671257873e-07,
-      "train_runtime": 663.6389,
-      "train_samples_per_second": 0.482,
-      "train_steps_per_second": 0.03
+      "train_loss": 9.848059619343985e-07,
+      "train_runtime": 706.4449,
+      "train_samples_per_second": 0.453,
+      "train_steps_per_second": 0.028
     }
   ],
   "logging_steps": 2,