End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:97f3e0e9007b7a33e091fbec9a578c3bfc22238bbfa34d9c50b452b5c2145725
+oid sha256:c565861640a22211bc382c8877c90f879aa9be966d5d102a078727d2cb69b0b6
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 7.064127771627682e-07,
-    "train_runtime": 609.7652,
+    "train_loss": 8.487639661325374e-07,
+    "train_runtime": 596.4258,
     "train_samples": 11,
-    "train_samples_per_second": 0.525,
-    "train_steps_per_second": 0.033
+    "train_samples_per_second": 0.537,
+    "train_steps_per_second": 0.034
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 7.064127771627682e-07,
-    "train_runtime": 609.7652,
+    "train_loss": 8.487639661325374e-07,
+    "train_runtime": 596.4258,
     "train_samples": 11,
-    "train_samples_per_second": 0.525,
-    "train_steps_per_second": 0.033
+    "train_samples_per_second": 0.537,
+    "train_steps_per_second": 0.034
 }

trainer_state.json

@@ -10,203 +10,203 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 499.40625,
+      "completion_length": 489.90625,
       "epoch": 0.7272727272727273,
-      "grad_norm": 0.49710243940353394,
-      "kl": 0.000682185356708942,
+      "grad_norm": 0.3609657883644104,
+      "kl": 0.0006489891966339201,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 1.1417980957776308,
-      "reward_std": 1.2042934596538544,
-      "rewards/concensus_correctness_reward_func": 0.0598750002682209,
-      "rewards/consensus_reward_func": 0.3125,
+      "reward": 0.828802197240293,
+      "reward_std": 0.7736977692693472,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.0625,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5572980614379048,
+      "rewards/question_recreation_reward_func": 0.5462084300816059,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.196500014513731,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.22009374480694532,
       "step": 2
     },
     {
-      "completion_length": 373.9642857142857,
+      "completion_length": 511.5,
       "epoch": 1.3636363636363638,
-      "grad_norm": 0.5117027163505554,
-      "kl": 0.0008518625316875321,
+      "grad_norm": 0.42430993914604187,
+      "kl": 0.0007115530482094203,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 0.9260532217366355,
-      "reward_std": 1.0265512253556932,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "reward": 0.6244428072656903,
+      "reward_std": 1.1362362929752894,
+      "rewards/concensus_correctness_reward_func": 0.017857142857142856,
+      "rewards/consensus_reward_func": 0.14285714285714285,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.07142857142857142,
-      "rewards/question_recreation_reward_func": 0.5391603516680854,
-      "rewards/soft_format_reward_func": 0.017857142857142856,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.4853356395448957,
+      "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.011892857828310557,
+      "rewards/xmlcount_reward_func": -0.021607141941785812,
       "step": 4
     },
     {
-      "completion_length": 478.35714285714283,
+      "completion_length": 464.39285714285717,
       "epoch": 2.0,
-      "grad_norm": 0.2743651866912842,
-      "kl": 0.0006004835512222988,
+      "grad_norm": 0.6451336741447449,
+      "kl": 0.000845598169170054,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 1.1801103949546814,
-      "reward_std": 1.2289031105382102,
-      "rewards/concensus_correctness_reward_func": 0.11828571557998657,
-      "rewards/consensus_reward_func": 0.7142857142857143,
+      "reward": 1.5903914826256889,
+      "reward_std": 2.294496755514826,
+      "rewards/concensus_correctness_reward_func": 0.7411428572876113,
+      "rewards/consensus_reward_func": 0.21428571428571427,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5983246862888336,
+      "rewards/final_correctness_reward_func": 0.14285714285714285,
+      "rewards/question_recreation_reward_func": 0.5562128743955067,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.25078571907111574,
+      "rewards/xmlcount_reward_func": -0.06410714345318931,
       "step": 6
     },
     {
-      "completion_length": 434.21875,
+      "completion_length": 420.1875,
       "epoch": 2.7272727272727275,
-      "grad_norm": 0.34855952858924866,
-      "kl": 0.0007185739377746359,
+      "grad_norm": 0.4900331199169159,
+      "kl": 0.0007078720445861109,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 0.9275249540805817,
-      "reward_std": 0.889457019045949,
-      "rewards/concensus_correctness_reward_func": 0.0,
+      "reward": 0.8949470296502113,
+      "reward_std": 0.9619810990989208,
+      "rewards/concensus_correctness_reward_func": 0.03125,
       "rewards/consensus_reward_func": 0.25,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.6967124193906784,
+      "rewards/question_recreation_reward_func": 0.6426657810807228,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": -0.034812500700354576,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": -0.028968743979930878,
       "step": 8
     },
     {
-      "completion_length": 473.7142857142857,
+      "completion_length": 521.5,
       "epoch": 3.3636363636363638,
-      "grad_norm": 0.3439553678035736,
-      "kl": 0.0006509673382554736,
+      "grad_norm": 0.43673744797706604,
+      "kl": 0.0006336758794662144,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 0.5664068056004388,
-      "reward_std": 0.974337100982666,
-      "rewards/concensus_correctness_reward_func": 0.059142857789993286,
-      "rewards/consensus_reward_func": 0.14285714285714285,
+      "reward": 0.8506239099161965,
+      "reward_std": 0.8674267274992806,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.07142857142857142,
-      "rewards/question_recreation_reward_func": 0.5220139494964054,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.6670881807804108,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.22903570319925035,
+      "rewards/strict_format_reward_func": 0.03571428571428571,
+      "rewards/xmlcount_reward_func": 0.1478214327778135,
       "step": 10
     },
     {
-      "completion_length": 440.67857142857144,
+      "completion_length": 420.42857142857144,
       "epoch": 4.0,
-      "grad_norm": 0.23475508391857147,
-      "kl": 0.0007548628408195717,
+      "grad_norm": 0.4140712022781372,
+      "kl": 0.0008527751030799534,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 0.7998639941215515,
-      "reward_std": 1.201057676758085,
-      "rewards/concensus_correctness_reward_func": 0.029571428894996643,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "reward": 0.8269651840840068,
+      "reward_std": 0.8055251027856555,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.14285714285714285,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.431828288095338,
+      "rewards/question_recreation_reward_func": 0.4500009013073785,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.03571428571428571,
-      "rewards/xmlcount_reward_func": 0.01703571847506932,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.21624999386923655,
       "step": 12
     },
     {
-      "completion_length": 503.53125,
+      "completion_length": 481.0625,
       "epoch": 4.7272727272727275,
-      "grad_norm": 0.7081426382064819,
-      "kl": 0.0012828102844650857,
+      "grad_norm": 0.42424920201301575,
+      "kl": 0.0006703631843265612,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 1.2126230709254742,
-      "reward_std": 1.1269454210996628,
-      "rewards/concensus_correctness_reward_func": 0.07537500001490116,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 1.1690492890775204,
+      "reward_std": 0.9069762676954269,
+      "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": 0.6234667934477329,
+      "rewards/question_recreation_reward_func": 0.6187680251896381,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.26378126442432404,
+      "rewards/xmlcount_reward_func": 0.22215624619275331,
       "step": 14
     },
     {
-      "completion_length": 505.57142857142856,
+      "completion_length": 423.75,
       "epoch": 5.363636363636363,
-      "grad_norm": 0.3318561911582947,
-      "kl": 0.0006546001615268844,
+      "grad_norm": 0.4367072582244873,
+      "kl": 0.0024985303919363233,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 0.8020758735282081,
-      "reward_std": 0.9012836643627712,
-      "rewards/concensus_correctness_reward_func": 0.06528571460928236,
+      "reward": 0.7280806686197009,
+      "reward_std": 0.8712436748402459,
+      "rewards/concensus_correctness_reward_func": 0.0,
       "rewards/consensus_reward_func": 0.14285714285714285,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.48907589486667086,
+      "rewards/final_correctness_reward_func": 0.07142857142857142,
+      "rewards/question_recreation_reward_func": 0.4392235449382237,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.10485714673995972,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.05671428915645395,
       "step": 16
     },
     {
-      "completion_length": 451.0357142857143,
+      "completion_length": 596.2857142857143,
       "epoch": 6.0,
-      "grad_norm": 5.338632106781006,
-      "kl": 0.0007286149110378963,
+      "grad_norm": 0.19180385768413544,
+      "kl": 0.0006549531244672835,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 0.9245169460773468,
-      "reward_std": 1.1597868374415807,
-      "rewards/concensus_correctness_reward_func": 0.0,
+      "reward": 2.444173582962581,
+      "reward_std": 2.860402743731226,
+      "rewards/concensus_correctness_reward_func": 1.4642857142857142,
       "rewards/consensus_reward_func": 0.35714285714285715,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.6596240486417498,
+      "rewards/final_correctness_reward_func": 0.07142857142857142,
+      "rewards/question_recreation_reward_func": 0.6413163840770721,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.0922499852521079,
+      "rewards/strict_format_reward_func": 0.03571428571428571,
+      "rewards/xmlcount_reward_func": -0.12571429354803904,
       "step": 18
     },
     {
-      "completion_length": 544.96875,
+      "completion_length": 451.1875,
       "epoch": 6.7272727272727275,
-      "grad_norm": 0.4531197249889374,
-      "kl": 0.0006524953714688309,
+      "grad_norm": 0.5098544955253601,
+      "kl": 0.0009375976696901489,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 0.9392223320901394,
-      "reward_std": 1.3096578195691109,
-      "rewards/concensus_correctness_reward_func": 0.051750000566244125,
-      "rewards/consensus_reward_func": 0.3125,
+      "reward": 1.6025606133043766,
+      "reward_std": 2.088516917079687,
+      "rewards/concensus_correctness_reward_func": 0.6485000001266599,
+      "rewards/consensus_reward_func": 0.1875,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0625,
-      "rewards/question_recreation_reward_func": 0.6449410542845726,
-      "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": -0.14809374418109655,
+      "rewards/question_recreation_reward_func": 0.555873055011034,
+      "rewards/soft_format_reward_func": 0.015625,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": 0.10131250135600567,
       "step": 20
     },
     {
       "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 7.064127771627682e-07,
-      "train_runtime": 609.7652,
-      "train_samples_per_second": 0.525,
-      "train_steps_per_second": 0.033
+      "train_loss": 8.487639661325374e-07,
+      "train_runtime": 596.4258,
+      "train_samples_per_second": 0.537,
+      "train_steps_per_second": 0.034
     }
   ],
   "logging_steps": 2,