End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a06e0735fce1268b71e714768fb80177e2a7389b998688a2e3c17e5df5ba5960
+oid sha256:dea64d5935127c39f89e419b465822408942dfdf60a8ee2e2204d22492e27d7e
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 6.380430704666651e-07,
-    "train_runtime": 1093.548,
+    "train_loss": 6.35625968925524e-07,
+    "train_runtime": 597.6366,
     "train_samples": 11,
-    "train_samples_per_second": 0.293,
-    "train_steps_per_second": 0.018
+    "train_samples_per_second": 0.535,
+    "train_steps_per_second": 0.033
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 6.380430704666651e-07,
-    "train_runtime": 1093.548,
+    "train_loss": 6.35625968925524e-07,
+    "train_runtime": 597.6366,
     "train_samples": 11,
-    "train_samples_per_second": 0.293,
-    "train_steps_per_second": 0.018
+    "train_samples_per_second": 0.535,
+    "train_steps_per_second": 0.033
 }

trainer_state.json

@@ -10,203 +10,203 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 529.125,
+      "completion_length": 376.28125,
       "epoch": 0.7272727272727273,
-      "grad_norm": 0.4646318852901459,
-      "kl": 0.0006358221507980488,
+      "grad_norm": 0.3437902331352234,
+      "kl": 0.000827149415272288,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 0.39380765706300735,
-      "reward_std": 0.9743207413703203,
+      "reward": 0.9650179520249367,
+      "reward_std": 0.9399564154446125,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.0625,
+      "rewards/consensus_reward_func": 0.125,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.4490889124572277,
+      "rewards/final_correctness_reward_func": 0.0625,
+      "rewards/question_recreation_reward_func": 0.48545546270906925,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": -0.13340624701231718,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": 0.26081249862909317,
       "step": 2
     },
     {
-      "completion_length": 501.5,
+      "completion_length": 451.2857142857143,
       "epoch": 1.3636363636363638,
-      "grad_norm": 0.4398326277732849,
-      "kl": 0.0006624749346104052,
+      "grad_norm": 0.5616311430931091,
+      "kl": 0.0006403406815869468,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 1.6800819635391235,
-      "reward_std": 1.4839196843760354,
-      "rewards/concensus_correctness_reward_func": 0.11485714146069118,
-      "rewards/consensus_reward_func": 0.35714285714285715,
+      "reward": 1.5441367796489172,
+      "reward_std": 1.3162878240857805,
+      "rewards/concensus_correctness_reward_func": 0.0816428576196943,
+      "rewards/consensus_reward_func": 0.5,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.21428571428571427,
-      "rewards/question_recreation_reward_func": 0.5619391628674099,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.6427796483039856,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.03571428571428571,
-      "rewards/xmlcount_reward_func": 0.39614285874579636,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.3197142779827118,
       "step": 4
     },
     {
-      "completion_length": 515.0357142857143,
+      "completion_length": 462.10714285714283,
       "epoch": 2.0,
-      "grad_norm": 0.3059370517730713,
-      "kl": 0.0007214588612051946,
+      "grad_norm": 0.4168936312198639,
+      "kl": 0.000719171848946384,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 0.8787845202854702,
-      "reward_std": 1.4067505257470267,
-      "rewards/concensus_correctness_reward_func": 0.0821428554398673,
+      "reward": 0.8503530536379132,
+      "reward_std": 1.232443732874734,
+      "rewards/concensus_correctness_reward_func": 0.1277142869574683,
       "rewards/consensus_reward_func": 0.2857142857142857,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.14285714285714285,
-      "rewards/question_recreation_reward_func": 0.45535593373434885,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5647101657731193,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.08728571874754769,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": -0.1456428606595312,
       "step": 6
     },
     {
-      "completion_length": 450.46875,
+      "completion_length": 422.0625,
       "epoch": 2.7272727272727275,
-      "grad_norm": 0.4287567734718323,
-      "kl": 0.0006250313490454573,
+      "grad_norm": 0.5831401944160461,
+      "kl": 0.0006976408658374567,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 2.1336237341165543,
-      "reward_std": 2.356996573507786,
-      "rewards/concensus_correctness_reward_func": 0.7309999987483025,
-      "rewards/consensus_reward_func": 0.4375,
+      "reward": 1.1317045465111732,
+      "reward_std": 1.075602475553751,
+      "rewards/concensus_correctness_reward_func": 0.05581250041723251,
+      "rewards/consensus_reward_func": 0.125,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.125,
-      "rewards/question_recreation_reward_func": 0.5192175172269344,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5921108238399029,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.3209062535315752,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.3431562567129731,
       "step": 8
     },
     {
-      "completion_length": 415.60714285714283,
+      "completion_length": 443.64285714285717,
       "epoch": 3.3636363636363638,
-      "grad_norm": 0.4459122121334076,
-      "kl": 0.0009690781423289861,
+      "grad_norm": 0.4295225143432617,
+      "kl": 0.0007067695010586508,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 1.4117277903216225,
-      "reward_std": 1.060990469796317,
-      "rewards/concensus_correctness_reward_func": 0.12857142516544887,
-      "rewards/consensus_reward_func": 0.21428571428571427,
+      "reward": 1.6077298362340247,
+      "reward_std": 1.0551166534423828,
+      "rewards/concensus_correctness_reward_func": 0.1914999932050705,
+      "rewards/consensus_reward_func": 0.5714285714285714,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.21428571428571427,
-      "rewards/question_recreation_reward_func": 0.5573349041598183,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.4601941300289972,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.27939286615167347,
+      "rewards/xmlcount_reward_func": 0.366749997649874,
       "step": 10
     },
     {
-      "completion_length": 584.9642857142857,
+      "completion_length": 512.8571428571429,
       "epoch": 4.0,
-      "grad_norm": 0.36204561591148376,
-      "kl": 0.000589396118552291,
+      "grad_norm": 0.42578762769699097,
+      "kl": 0.0006562869945940163,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 0.9562124311923981,
-      "reward_std": 1.1778249740600586,
-      "rewards/concensus_correctness_reward_func": 0.08471428496497017,
-      "rewards/consensus_reward_func": 0.35714285714285715,
+      "reward": 0.4705148552145277,
+      "reward_std": 0.6741891162736076,
+      "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": 0.481891040291105,
+      "rewards/question_recreation_reward_func": 0.43669344059058596,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.03246427859578814,
+      "rewards/xmlcount_reward_func": -0.03760716106210436,
       "step": 12
     },
     {
-      "completion_length": 432.65625,
+      "completion_length": 460.1875,
       "epoch": 4.7272727272727275,
-      "grad_norm": 0.5002179741859436,
-      "kl": 0.000725990132195875,
+      "grad_norm": 0.5053578615188599,
+      "kl": 0.0006863677699584514,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 2.022418200969696,
-      "reward_std": 2.153670124709606,
-      "rewards/concensus_correctness_reward_func": 0.7094374895095825,
-      "rewards/consensus_reward_func": 0.3125,
+      "reward": 1.2665400207042694,
+      "reward_std": 1.1816745065152645,
+      "rewards/concensus_correctness_reward_func": 0.1038750009611249,
+      "rewards/consensus_reward_func": 0.375,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.1875,
-      "rewards/question_recreation_reward_func": 0.5288244183175266,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.6119150072336197,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.2685312498360872,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.17575000412762165,
       "step": 14
     },
     {
-      "completion_length": 540.3214285714286,
+      "completion_length": 596.8928571428571,
       "epoch": 5.363636363636363,
-      "grad_norm": 3.299898862838745,
-      "kl": 0.0007763790033225502,
+      "grad_norm": 0.40903177857398987,
+      "kl": 0.0006001410407147237,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 0.654516909803663,
-      "reward_std": 1.215036724294935,
-      "rewards/concensus_correctness_reward_func": 0.022142857313156128,
-      "rewards/consensus_reward_func": 0.21428571428571427,
+      "reward": 0.8541636381830487,
+      "reward_std": 1.013209513255528,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.14285714285714285,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.14285714285714285,
-      "rewards/question_recreation_reward_func": 0.40419550240039825,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.48498509185654776,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.1289642857653754,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.20846428615706308,
       "step": 16
     },
     {
-      "completion_length": 492.92857142857144,
+      "completion_length": 388.64285714285717,
       "epoch": 6.0,
-      "grad_norm": 4.3900957107543945,
-      "kl": 0.000726213857498286,
+      "grad_norm": 0.3834454119205475,
+      "kl": 0.0007274486041361732,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 1.6323457615716117,
-      "reward_std": 1.1951770058700018,
-      "rewards/concensus_correctness_reward_func": 0.12685714023453848,
+      "reward": 1.0557954971279417,
+      "reward_std": 1.018987434250968,
+      "rewards/concensus_correctness_reward_func": 0.009000000144754137,
       "rewards/consensus_reward_func": 0.2857142857142857,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5524172101702008,
-      "rewards/soft_format_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5256883757455009,
+      "rewards/soft_format_reward_func": 0.017857142857142856,
       "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.6495000209127154,
+      "rewards/xmlcount_reward_func": 0.19967857216085708,
       "step": 18
     },
     {
-      "completion_length": 525.34375,
+      "completion_length": 399.28125,
       "epoch": 6.7272727272727275,
-      "grad_norm": 0.6854709386825562,
-      "kl": 0.0005694236060662661,
+      "grad_norm": 0.4910911023616791,
+      "kl": 0.0007835296382836532,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 1.6998868212103844,
-      "reward_std": 2.261262886226177,
-      "rewards/concensus_correctness_reward_func": 0.7890624888241291,
-      "rewards/consensus_reward_func": 0.3125,
+      "reward": 0.9040417112410069,
+      "reward_std": 0.9552651233971119,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.0625,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.1875,
-      "rewards/question_recreation_reward_func": 0.5786056481301785,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5213229823857546,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": -0.1834062561392784,
+      "rewards/strict_format_reward_func": 0.0625,
+      "rewards/xmlcount_reward_func": 0.25771874852944165,
       "step": 20
     },
     {
       "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 6.380430704666651e-07,
-      "train_runtime": 1093.548,
-      "train_samples_per_second": 0.293,
-      "train_steps_per_second": 0.018
+      "train_loss": 6.35625968925524e-07,
+      "train_runtime": 597.6366,
+      "train_samples_per_second": 0.535,
+      "train_steps_per_second": 0.033
     }
   ],
   "logging_steps": 2,