End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:122e249fed41a092cb24bb483a0b0f8c62c1e3e43bc170a767988cefe7623ea8
+oid sha256:910074061cdef71a393790c732d21330a84d21e156fdd8dec0aa625eff4cff20
 size 35237104

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 2.4479934097598743e-06,
-    "train_runtime": 565.5412,
+    "train_loss": 4.995642120775301e-06,
+    "train_runtime": 610.5749,
     "train_samples": 14,
-    "train_samples_per_second": 0.566,
-    "train_steps_per_second": 0.035
+    "train_samples_per_second": 0.524,
+    "train_steps_per_second": 0.033
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 2.4479934097598743e-06,
-    "train_runtime": 565.5412,
+    "train_loss": 4.995642120775301e-06,
+    "train_runtime": 610.5749,
     "train_samples": 14,
-    "train_samples_per_second": 0.566,
-    "train_steps_per_second": 0.035
+    "train_samples_per_second": 0.524,
+    "train_steps_per_second": 0.033
 }

trainer_state.json

@@ -10,203 +10,203 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 287.3125,
+      "completion_length": 352.40625,
       "epoch": 0.5714285714285714,
-      "grad_norm": 2.8832011222839355,
-      "kl": 0.00394943701394368,
+      "grad_norm": 3.7579925060272217,
+      "kl": 0.004185061530733947,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 0.4581340132281184,
-      "reward_std": 0.5015503789763898,
-      "rewards/concensus_correctness_reward_func": 0.042249999940395355,
+      "reward": 0.30802640365436673,
+      "reward_std": 0.4295234101591632,
+      "rewards/concensus_correctness_reward_func": 0.014437500387430191,
       "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.16038401192054152,
-      "rewards/soft_format_reward_func": 0.015625,
+      "rewards/question_recreation_reward_func": 0.25730764446780086,
+      "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.2398749990388751,
+      "rewards/xmlcount_reward_func": 0.03628124576061964,
       "step": 2
     },
     {
-      "completion_length": 317.8333333333333,
+      "completion_length": 391.5,
       "epoch": 1.0,
-      "grad_norm": 1.4717791080474854,
-      "kl": 0.0013749226927757263,
+      "grad_norm": 0.7946363091468811,
+      "kl": 0.009476465502909074,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 0.3418625301371018,
-      "reward_std": 0.8531895938018957,
-      "rewards/concensus_correctness_reward_func": 0.0620000014702479,
-      "rewards/consensus_reward_func": 0.08333333333333333,
+      "reward": 1.3049255851656199,
+      "reward_std": 2.243508925040563,
+      "rewards/concensus_correctness_reward_func": 0.8333333333333334,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.3050708717977007,
+      "rewards/final_correctness_reward_func": 0.25,
+      "rewards/question_recreation_reward_func": 0.23875894149144491,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.10854166342566411,
+      "rewards/xmlcount_reward_func": -0.017166662340362866,
       "step": 4
     },
     {
-      "completion_length": 329.53125,
+      "completion_length": 453.28125,
       "epoch": 1.5714285714285714,
-      "grad_norm": 1.9261950254440308,
-      "kl": 0.0017050012611434795,
+      "grad_norm": 2.741084575653076,
+      "kl": 0.001483817322878167,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 0.3830822017043829,
-      "reward_std": 0.4381517507135868,
-      "rewards/concensus_correctness_reward_func": 0.042249999940395355,
+      "reward": 0.26920943846926093,
+      "reward_std": 0.48570725601166487,
+      "rewards/concensus_correctness_reward_func": 0.0,
       "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.16058220900595188,
+      "rewards/question_recreation_reward_func": 0.1770219358149916,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.18025000113993883,
+      "rewards/xmlcount_reward_func": 0.09218749543651938,
       "step": 6
     },
     {
-      "completion_length": 356.25,
+      "completion_length": 420.7083333333333,
       "epoch": 2.0,
-      "grad_norm": 1.444770336151123,
-      "kl": 0.001178021537877309,
+      "grad_norm": 0.7737870812416077,
+      "kl": 0.001691168814431876,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 0.5916601779560248,
-      "reward_std": 0.7444767082730929,
-      "rewards/concensus_correctness_reward_func": 0.038333334028720856,
-      "rewards/consensus_reward_func": 0.08333333333333333,
+      "reward": 0.46565589557091397,
+      "reward_std": 0.8166446760296822,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.31041015684604645,
+      "rewards/final_correctness_reward_func": 0.08333333333333333,
+      "rewards/question_recreation_reward_func": 0.332822581132253,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.1595833276708921,
+      "rewards/xmlcount_reward_func": 0.04949999786913395,
       "step": 8
     },
     {
-      "completion_length": 345.875,
+      "completion_length": 415.375,
       "epoch": 2.571428571428571,
-      "grad_norm": 2.997659206390381,
-      "kl": 0.0028061752454959787,
+      "grad_norm": 2.5258235931396484,
+      "kl": 0.0015403052238980308,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 0.4578349385410547,
-      "reward_std": 0.7899111276492476,
-      "rewards/concensus_correctness_reward_func": 0.12674999237060547,
+      "reward": 0.5891070775687695,
+      "reward_std": 0.8081069309264421,
+      "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": 0.2588036877568811,
+      "rewards/final_correctness_reward_func": 0.125,
+      "rewards/question_recreation_reward_func": 0.3199820921290666,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.009781254804693162,
+      "rewards/xmlcount_reward_func": 0.08162500150501728,
       "step": 10
     },
     {
-      "completion_length": 361.5833333333333,
+      "completion_length": 382.2916666666667,
       "epoch": 3.0,
-      "grad_norm": 1.055799126625061,
-      "kl": 0.0021580825559794903,
+      "grad_norm": 0.891693115234375,
+      "kl": 0.00470517436042428,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 0.4883204946915309,
-      "reward_std": 0.9803517982363701,
-      "rewards/concensus_correctness_reward_func": 0.0620000014702479,
-      "rewards/consensus_reward_func": 0.08333333333333333,
+      "reward": 1.087148867547512,
+      "reward_std": 2.083983345578114,
+      "rewards/concensus_correctness_reward_func": 0.8333333333333334,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.34465381503105164,
+      "rewards/question_recreation_reward_func": 0.2563988597442706,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.001666662593682607,
+      "rewards/xmlcount_reward_func": -0.0025833348433176675,
       "step": 12
     },
     {
-      "completion_length": 383.78125,
+      "completion_length": 405.5,
       "epoch": 3.571428571428571,
-      "grad_norm": 3.506430149078369,
-      "kl": 0.0016388819349231198,
+      "grad_norm": 10.450708389282227,
+      "kl": 0.0039054710723576136,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 0.5316155850887299,
-      "reward_std": 0.7287016687914729,
-      "rewards/concensus_correctness_reward_func": 0.04650000110268593,
-      "rewards/consensus_reward_func": 0.0625,
+      "reward": 2.2041085474193096,
+      "reward_std": 3.35321690980345,
+      "rewards/concensus_correctness_reward_func": 1.875,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.3055530800484121,
+      "rewards/final_correctness_reward_func": 0.125,
+      "rewards/question_recreation_reward_func": 0.26617118320427835,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.1170624983496964,
+      "rewards/xmlcount_reward_func": -0.06206250889226794,
       "step": 14
     },
     {
-      "completion_length": 332.2083333333333,
+      "completion_length": 417.25,
       "epoch": 4.0,
-      "grad_norm": 4.015366077423096,
-      "kl": 0.0015126650784319888,
+      "grad_norm": 1.009475588798523,
+      "kl": 0.0014468229298169415,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 0.24088450521230698,
-      "reward_std": 0.6172617338597775,
-      "rewards/concensus_correctness_reward_func": 0.056333333253860474,
+      "reward": 0.44937922308842343,
+      "reward_std": 0.701083225508531,
+      "rewards/concensus_correctness_reward_func": 0.0,
       "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.13805115719636282,
+      "rewards/question_recreation_reward_func": 0.34487922737995785,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.04650000482797623,
+      "rewards/xmlcount_reward_func": 0.1044999963293473,
       "step": 16
     },
     {
-      "completion_length": 377.4375,
+      "completion_length": 419.90625,
       "epoch": 4.571428571428571,
-      "grad_norm": 3.1088175773620605,
-      "kl": 0.0013925813509558793,
+      "grad_norm": 4.776589870452881,
+      "kl": 0.002067406428977847,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 0.2849379200488329,
-      "reward_std": 0.5267030587419868,
-      "rewards/concensus_correctness_reward_func": 0.0,
+      "reward": 1.5636410352308303,
+      "reward_std": 2.9666851684451103,
+      "rewards/concensus_correctness_reward_func": 1.2618124997243285,
       "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.23284417716786265,
+      "rewards/question_recreation_reward_func": 0.23548481822945178,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.05209375359117985,
+      "rewards/xmlcount_reward_func": 0.0663437508046627,
       "step": 18
     },
     {
-      "completion_length": 420.0,
+      "completion_length": 408.2083333333333,
       "epoch": 5.0,
-      "grad_norm": 1.3363821506500244,
-      "kl": 0.010974116526388874,
+      "grad_norm": 0.9863994121551514,
+      "kl": 0.031827634976555906,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 0.2139842833081881,
-      "reward_std": 0.5290305142601331,
+      "reward": 0.6483268290758133,
+      "reward_std": 0.9140989532073339,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.08333333333333333,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.24231762439012527,
+      "rewards/question_recreation_reward_func": 0.3939934764057398,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.02833333859841029,
+      "rewards/xmlcount_reward_func": 0.1710000013311704,
       "step": 20
     },
     {
       "epoch": 5.0,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 2.4479934097598743e-06,
-      "train_runtime": 565.5412,
-      "train_samples_per_second": 0.566,
-      "train_steps_per_second": 0.035
+      "train_loss": 4.995642120775301e-06,
+      "train_runtime": 610.5749,
+      "train_samples_per_second": 0.524,
+      "train_steps_per_second": 0.033
     }
   ],
   "logging_steps": 2,