End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:669ce87026e41b9709a798f018818fac0b441f3df0b751d78ad13943242a72ff
+oid sha256:c45f449b54f13973513de973c7f898a9444bab8665bd536fc9d0bc2f8a11f075
 size 35237104

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 2.5205041884746038e-06,
-    "train_runtime": 2309.381,
-    "train_samples": 59,
-    "train_samples_per_second": 0.139,
-    "train_steps_per_second": 0.009
+    "train_loss": 1.4121161370894697e-06,
+    "train_runtime": 1835.0928,
+    "train_samples": 13,
+    "train_samples_per_second": 0.174,
+    "train_steps_per_second": 0.011
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 2.5205041884746038e-06,
-    "train_runtime": 2309.381,
-    "train_samples": 59,
-    "train_samples_per_second": 0.139,
-    "train_steps_per_second": 0.009
+    "train_loss": 1.4121161370894697e-06,
+    "train_runtime": 1835.0928,
+    "train_samples": 13,
+    "train_samples_per_second": 0.174,
+    "train_steps_per_second": 0.011
 }

trainer_state.json

@@ -2,7 +2,7 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 1.3389830508474576,
+  "epoch": 5.0,
   "eval_steps": 500,
   "global_step": 20,
   "is_hyper_param_search": false,
@@ -10,209 +10,209 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 365.375,
-      "epoch": 0.13559322033898305,
-      "grad_norm": 11.633384704589844,
-      "kl": 0.0,
+      "completion_length": 437.34375,
+      "epoch": 0.6153846153846154,
+      "grad_norm": 2.043631076812744,
+      "kl": 0.0017710743995849043,
       "learning_rate": 4.965903258506806e-07,
-      "loss": -0.0,
-      "reward": 0.42032941058278084,
-      "reward_std": 0.5112706171348691,
+      "loss": 0.0,
+      "reward": 0.4190886758733541,
+      "reward_std": 0.6645759915700182,
       "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": 0.23526690888684243,
+      "rewards/question_recreation_reward_func": 0.21440116595476866,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.12256249971687794,
+      "rewards/xmlcount_reward_func": 0.14218749850988388,
       "step": 2
     },
     {
-      "completion_length": 539.03125,
-      "epoch": 0.2711864406779661,
-      "grad_norm": 1.1121511459350586,
-      "kl": 0.0012746257416438311,
+      "completion_length": 236.05,
+      "epoch": 1.0,
+      "grad_norm": 1.026563048362732,
+      "kl": 0.0026961707742884757,
       "learning_rate": 4.698684378016222e-07,
       "loss": 0.0,
-      "reward": 0.48772054351866245,
-      "reward_std": 0.5359725765883923,
-      "rewards/concensus_correctness_reward_func": 0.024937499314546585,
-      "rewards/consensus_reward_func": 0.125,
+      "reward": 0.15675089359283448,
+      "reward_std": 0.2769328474998474,
+      "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": 0.3020643088966608,
+      "rewards/question_recreation_reward_func": 0.13105088621377944,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.03571874834597111,
+      "rewards/xmlcount_reward_func": 0.025700002908706665,
       "step": 4
     },
     {
-      "completion_length": 376.4375,
-      "epoch": 0.4067796610169492,
-      "grad_norm": 28.250625610351562,
-      "kl": 0.0014740825135959312,
+      "completion_length": 295.625,
+      "epoch": 1.6153846153846154,
+      "grad_norm": 24.10466766357422,
+      "kl": 0.0014241181343095377,
       "learning_rate": 4.193203929064353e-07,
       "loss": 0.0,
-      "reward": 0.268473110627383,
-      "reward_std": 0.4494456851389259,
+      "reward": 0.24623784562572837,
+      "reward_std": 0.21599767965381034,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.0625,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0625,
-      "rewards/question_recreation_reward_func": 0.1787543643731624,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.14892533840611577,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.03528125211596489,
+      "rewards/xmlcount_reward_func": 0.09731249744072556,
       "step": 6
     },
     {
-      "completion_length": 505.90625,
-      "epoch": 0.5423728813559322,
-      "grad_norm": 3.9899539947509766,
-      "kl": 0.00813884272793075,
+      "completion_length": 409.0,
+      "epoch": 2.0,
+      "grad_norm": 0.6894894242286682,
+      "kl": 0.0011497939005494118,
       "learning_rate": 3.5042385616324236e-07,
       "loss": 0.0,
-      "reward": 0.2546458290889859,
-      "reward_std": 0.35357041633687913,
+      "reward": 0.2617270529270172,
+      "reward_std": 0.415588341653347,
       "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": 0.16873957961797714,
+      "rewards/question_recreation_reward_func": 0.18262706324458122,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.0859062522649765,
+      "rewards/xmlcount_reward_func": 0.07909999787807465,
       "step": 8
     },
     {
-      "completion_length": 440.6875,
-      "epoch": 0.6779661016949152,
-      "grad_norm": 2.0828049182891846,
-      "kl": 0.0054728194227209315,
+      "completion_length": 478.5,
+      "epoch": 2.6153846153846154,
+      "grad_norm": 1.2047410011291504,
+      "kl": 0.0014263475895859301,
       "learning_rate": 2.706448363680831e-07,
       "loss": 0.0,
-      "reward": 0.467122764326632,
-      "reward_std": 0.6823233440518379,
-      "rewards/concensus_correctness_reward_func": 0.0,
+      "reward": 0.28225671174004674,
+      "reward_std": 0.4247669519390911,
+      "rewards/concensus_correctness_reward_func": 0.0078125,
       "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.125,
-      "rewards/question_recreation_reward_func": 0.27787277568131685,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.2094442001543939,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.06425000075250864,
+      "rewards/xmlcount_reward_func": 0.06500000273808837,
       "step": 10
     },
     {
-      "completion_length": 479.71875,
-      "epoch": 0.8135593220338984,
-      "grad_norm": 1.8237814903259277,
-      "kl": 0.001228883207659237,
+      "completion_length": 424.1,
+      "epoch": 3.0,
+      "grad_norm": 0.4744141697883606,
+      "kl": 0.002176445326767862,
       "learning_rate": 1.886286282148002e-07,
       "loss": 0.0,
-      "reward": 0.6772353190463036,
-      "reward_std": 0.9963489768560976,
-      "rewards/concensus_correctness_reward_func": 0.09187500178813934,
-      "rewards/consensus_reward_func": 0.1875,
+      "reward": 0.2148798704147339,
+      "reward_std": 0.341147843003273,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0625,
-      "rewards/question_recreation_reward_func": 0.22079780069179833,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.1381298743188381,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.11456249840557575,
+      "rewards/xmlcount_reward_func": 0.07674999311566352,
       "step": 12
     },
     {
-      "completion_length": 412.375,
-      "epoch": 0.9491525423728814,
-      "grad_norm": 2.8275840282440186,
-      "kl": 0.0016695725716999732,
+      "completion_length": 353.375,
+      "epoch": 3.6153846153846154,
+      "grad_norm": 1.9431273937225342,
+      "kl": 0.002492671621439513,
       "learning_rate": 1.1326296046939333e-07,
       "loss": 0.0,
-      "reward": 0.3181928126141429,
-      "reward_std": 0.6266264107543975,
+      "reward": 0.41643676720559597,
+      "reward_std": 0.5340236783958972,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.125,
+      "rewards/consensus_reward_func": 0.0625,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.29797407519072294,
+      "rewards/question_recreation_reward_func": 0.2364055346697569,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.10478125279769301,
+      "rewards/xmlcount_reward_func": 0.11753124604001641,
       "step": 14
     },
     {
-      "completion_length": 489.17857142857144,
-      "epoch": 1.0677966101694916,
-      "grad_norm": 4.222405433654785,
-      "kl": 0.003275405873344945,
+      "completion_length": 318.95,
+      "epoch": 4.0,
+      "grad_norm": 1.447885274887085,
+      "kl": 0.001752735092304647,
       "learning_rate": 5.271487265090163e-08,
       "loss": 0.0,
-      "reward": 0.5893184766173363,
-      "reward_std": 0.6519519694681678,
+      "reward": 0.21283867582678795,
+      "reward_std": 0.23432353511452675,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.21428571428571427,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.29463991842099596,
+      "rewards/question_recreation_reward_func": 0.1704386778175831,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.08039285881178719,
+      "rewards/xmlcount_reward_func": 0.042400000244379045,
       "step": 16
     },
     {
-      "completion_length": 417.5625,
-      "epoch": 1.2033898305084745,
-      "grad_norm": 2.7506484985351562,
-      "kl": 0.001554385518829804,
+      "completion_length": 373.5625,
+      "epoch": 4.615384615384615,
+      "grad_norm": 2.1303718090057373,
+      "kl": 0.0012529623854788952,
       "learning_rate": 1.3545689574841341e-08,
       "loss": 0.0,
-      "reward": 0.5497422339394689,
-      "reward_std": 0.7734244260936975,
-      "rewards/concensus_correctness_reward_func": 0.07824999839067459,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 0.16706466767936945,
+      "reward_std": 0.3705905247479677,
+      "rewards/concensus_correctness_reward_func": 0.0078125,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.20064851082861423,
+      "rewards/question_recreation_reward_func": 0.10893965675495565,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.020843751030042768,
+      "rewards/xmlcount_reward_func": 0.05031249951571226,
       "step": 18
     },
     {
-      "completion_length": 424.25,
-      "epoch": 1.3389830508474576,
-      "grad_norm": 7.309051513671875,
-      "kl": 0.0016514052113052458,
+      "completion_length": 331.7,
+      "epoch": 5.0,
+      "grad_norm": 1.0344990491867065,
+      "kl": 0.0013469671830534935,
       "learning_rate": 0.0,
       "loss": 0.0,
-      "reward": 0.3152765342965722,
-      "reward_std": 0.42165259015746415,
+      "reward": 0.4605413466691971,
+      "reward_std": 0.728150698542595,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.1,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.2161827760282904,
+      "rewards/question_recreation_reward_func": 0.23284133970737458,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.09909375011920929,
+      "rewards/xmlcount_reward_func": 0.12770000137388707,
       "step": 20
     },
     {
-      "epoch": 1.3389830508474576,
+      "epoch": 5.0,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 2.5205041884746038e-06,
-      "train_runtime": 2309.381,
-      "train_samples_per_second": 0.139,
-      "train_steps_per_second": 0.009
+      "train_loss": 1.4121161370894697e-06,
+      "train_runtime": 1835.0928,
+      "train_samples_per_second": 0.174,
+      "train_steps_per_second": 0.011
     }
   ],
   "logging_steps": 2,
   "max_steps": 20,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 2,
+  "num_train_epochs": 7,
   "save_steps": 25,
   "stateful_callbacks": {
     "TrainerControl": {