End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e25743e03f6466f34092031960afc5fa3e0464d89d4e67dc778ddba86a84e3b7
+oid sha256:f84591a70c42fed524a2b93d587a6e9848d5a88de9815883061ebbd5da41f3be
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 8.599709872214589e-07,
-    "train_runtime": 984.9167,
+    "train_loss": 8.411427756982448e-07,
+    "train_runtime": 1527.9925,
     "train_samples": 14,
-    "train_samples_per_second": 0.325,
-    "train_steps_per_second": 0.02
+    "train_samples_per_second": 0.209,
+    "train_steps_per_second": 0.013
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 8.599709872214589e-07,
-    "train_runtime": 984.9167,
+    "train_loss": 8.411427756982448e-07,
+    "train_runtime": 1527.9925,
     "train_samples": 14,
-    "train_samples_per_second": 0.325,
-    "train_steps_per_second": 0.02
+    "train_samples_per_second": 0.209,
+    "train_steps_per_second": 0.013
 }

trainer_state.json

@@ -10,203 +10,203 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 309.6875,
+      "completion_length": 276.40625,
       "epoch": 0.5714285714285714,
-      "grad_norm": 0.6436980962753296,
-      "kl": 0.0009541559702483937,
+      "grad_norm": 0.47430989146232605,
+      "kl": 0.0008302802962134592,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 1.4780212081968784,
-      "reward_std": 0.9956347960978746,
-      "rewards/concensus_correctness_reward_func": 0.06074999924749136,
-      "rewards/consensus_reward_func": 0.375,
+      "reward": 2.714479446411133,
+      "reward_std": 2.4156916365027428,
+      "rewards/concensus_correctness_reward_func": 0.9347499855794013,
+      "rewards/consensus_reward_func": 0.5,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.25,
-      "rewards/question_recreation_reward_func": 0.39670873060822487,
+      "rewards/final_correctness_reward_func": 0.4375,
+      "rewards/question_recreation_reward_func": 0.5122918896377087,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.046875,
-      "rewards/xmlcount_reward_func": 0.3486875002272427,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.3299375083297491,
       "step": 2
     },
     {
-      "completion_length": 297.1666666666667,
+      "completion_length": 276.2083333333333,
       "epoch": 1.0,
-      "grad_norm": 0.6005321741104126,
-      "kl": 0.00102247311345612,
+      "grad_norm": 0.394611656665802,
+      "kl": 0.0009083122325440248,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 1.332326278090477,
-      "reward_std": 1.1768109301726024,
-      "rewards/concensus_correctness_reward_func": 0.07725000195205212,
-      "rewards/consensus_reward_func": 0.4166666666666667,
+      "reward": 4.962588508923848,
+      "reward_std": 4.682466775178909,
+      "rewards/concensus_correctness_reward_func": 2.7731666415929794,
+      "rewards/consensus_reward_func": 0.5,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.49520127723614377,
+      "rewards/final_correctness_reward_func": 0.5833333333333334,
+      "rewards/question_recreation_reward_func": 0.6173801670471827,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.3432083452741305,
+      "rewards/xmlcount_reward_func": 0.48870832721392315,
       "step": 4
     },
     {
-      "completion_length": 350.09375,
+      "completion_length": 299.03125,
       "epoch": 1.5714285714285714,
-      "grad_norm": 0.6561959981918335,
-      "kl": 0.0015157189263845794,
+      "grad_norm": 0.47179076075553894,
+      "kl": 0.0008499017931171693,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 3.1877787858247757,
-      "reward_std": 2.6128285750746727,
-      "rewards/concensus_correctness_reward_func": 1.4440625570714474,
-      "rewards/consensus_reward_func": 0.5625,
+      "reward": 2.335354745388031,
+      "reward_std": 2.4642272479832172,
+      "rewards/concensus_correctness_reward_func": 0.9320624866522849,
+      "rewards/consensus_reward_func": 0.5,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.25,
-      "rewards/question_recreation_reward_func": 0.5010287947952747,
+      "rewards/final_correctness_reward_func": 0.1875,
+      "rewards/question_recreation_reward_func": 0.5398234911262989,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.41456250473856926,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.17596874572336674,
       "step": 6
     },
     {
-      "completion_length": 314.25,
+      "completion_length": 311.5416666666667,
       "epoch": 2.0,
-      "grad_norm": 0.5843239426612854,
-      "kl": 0.000984727618439744,
+      "grad_norm": 0.44331005215644836,
+      "kl": 0.0007898747959795097,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 1.2161661287148793,
-      "reward_std": 1.1077017411589622,
-      "rewards/concensus_correctness_reward_func": 0.04599999853720268,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 3.7824546794096627,
+      "reward_std": 4.247074256340663,
+      "rewards/concensus_correctness_reward_func": 1.9034999758005142,
+      "rewards/consensus_reward_func": 0.3333333333333333,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.08333333333333333,
-      "rewards/question_recreation_reward_func": 0.588124468922615,
+      "rewards/final_correctness_reward_func": 0.6666666666666666,
+      "rewards/question_recreation_reward_func": 0.5416630903879801,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.2487083338201046,
+      "rewards/xmlcount_reward_func": 0.337291660408179,
       "step": 8
     },
     {
-      "completion_length": 314.53125,
+      "completion_length": 266.3125,
       "epoch": 2.571428571428571,
-      "grad_norm": 0.5834755897521973,
-      "kl": 0.0009463745809625834,
+      "grad_norm": 0.7045848369598389,
+      "kl": 0.0009941780881490558,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 1.455033615231514,
-      "reward_std": 0.8989755287766457,
-      "rewards/concensus_correctness_reward_func": 0.05224999971687794,
-      "rewards/consensus_reward_func": 0.4375,
+      "reward": 3.563333647325635,
+      "reward_std": 4.363310743123293,
+      "rewards/concensus_correctness_reward_func": 1.9877500161528587,
+      "rewards/consensus_reward_func": 0.375,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0625,
-      "rewards/question_recreation_reward_func": 0.5025336071848869,
+      "rewards/final_correctness_reward_func": 0.375,
+      "rewards/question_recreation_reward_func": 0.44967750646173954,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.40024999622255564,
+      "rewards/xmlcount_reward_func": 0.37590624997392297,
       "step": 10
     },
     {
-      "completion_length": 255.375,
+      "completion_length": 256.8333333333333,
       "epoch": 3.0,
-      "grad_norm": 0.4639264941215515,
-      "kl": 0.0009347472708517065,
+      "grad_norm": 0.34191426634788513,
+      "kl": 0.0008086791591874013,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 1.8883747557799022,
-      "reward_std": 0.9045937011639277,
-      "rewards/concensus_correctness_reward_func": 0.11074999844034512,
-      "rewards/consensus_reward_func": 0.4166666666666667,
+      "reward": 1.2910330494244893,
+      "reward_std": 1.2320965280135472,
+      "rewards/concensus_correctness_reward_func": 0.2633333336561918,
+      "rewards/consensus_reward_func": 0.25,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.4166666666666667,
-      "rewards/question_recreation_reward_func": 0.3912080451846123,
+      "rewards/final_correctness_reward_func": 0.25,
+      "rewards/question_recreation_reward_func": 0.3598247393965721,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.5530833502610525,
+      "rewards/xmlcount_reward_func": 0.16787500182787576,
       "step": 12
     },
     {
-      "completion_length": 299.3125,
+      "completion_length": 255.59375,
       "epoch": 3.571428571428571,
-      "grad_norm": 0.518161416053772,
-      "kl": 0.0008556295215385035,
+      "grad_norm": 5.995476722717285,
+      "kl": 0.0015220589339151047,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 1.5839294716715813,
-      "reward_std": 1.1322771236300468,
-      "rewards/concensus_correctness_reward_func": 0.10881250165402889,
-      "rewards/consensus_reward_func": 0.5625,
+      "reward": 2.9619270265102386,
+      "reward_std": 3.1472322661429644,
+      "rewards/concensus_correctness_reward_func": 1.25,
+      "rewards/consensus_reward_func": 0.1875,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.601585753262043,
+      "rewards/final_correctness_reward_func": 0.5625,
+      "rewards/question_recreation_reward_func": 0.5498332381248474,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.3110312633216381,
+      "rewards/xmlcount_reward_func": 0.4120937641710043,
       "step": 14
     },
     {
-      "completion_length": 327.0416666666667,
+      "completion_length": 283.25,
       "epoch": 4.0,
-      "grad_norm": 0.39999398589134216,
-      "kl": 0.00091908725638253,
+      "grad_norm": 0.40561026334762573,
+      "kl": 0.0009338054514955729,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 1.5979528526465099,
-      "reward_std": 1.0697652734816074,
-      "rewards/concensus_correctness_reward_func": 0.424083324149251,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 1.7401005364954472,
+      "reward_std": 1.7101846138636272,
+      "rewards/concensus_correctness_reward_func": 0.4287500015149514,
+      "rewards/consensus_reward_func": 0.5833333333333334,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.3333333333333333,
-      "rewards/question_recreation_reward_func": 0.36491114397843677,
+      "rewards/question_recreation_reward_func": 0.33676718175411224,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.22562500089406967,
+      "rewards/xmlcount_reward_func": 0.0579166654497385,
       "step": 16
     },
     {
-      "completion_length": 304.84375,
+      "completion_length": 287.0625,
       "epoch": 4.571428571428571,
-      "grad_norm": 0.6128489971160889,
-      "kl": 0.0008732353162486106,
+      "grad_norm": 0.47521623969078064,
+      "kl": 0.0007609435997437686,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 1.1475956365466118,
-      "reward_std": 0.8980563282966614,
-      "rewards/concensus_correctness_reward_func": 0.05843749828636646,
+      "reward": 2.659770254045725,
+      "reward_std": 3.5067169405519962,
+      "rewards/concensus_correctness_reward_func": 1.2612500004470348,
       "rewards/consensus_reward_func": 0.25,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0625,
-      "rewards/question_recreation_reward_func": 0.5500018913298845,
+      "rewards/final_correctness_reward_func": 0.375,
+      "rewards/question_recreation_reward_func": 0.556770320981741,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.21103125251829624,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.21674999967217445,
       "step": 18
     },
     {
-      "completion_length": 352.9583333333333,
+      "completion_length": 250.41666666666666,
       "epoch": 5.0,
-      "grad_norm": 0.44116389751434326,
-      "kl": 0.0008763076912146062,
+      "grad_norm": 0.4530916213989258,
+      "kl": 0.0010295834314698975,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 2.2834275166193643,
-      "reward_std": 2.6374321778615317,
-      "rewards/concensus_correctness_reward_func": 0.8485833331942558,
-      "rewards/consensus_reward_func": 0.3333333333333333,
+      "reward": 2.3395425279935202,
+      "reward_std": 1.2531716078519821,
+      "rewards/concensus_correctness_reward_func": 0.38941664807498455,
+      "rewards/consensus_reward_func": 0.5833333333333334,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.4166666666666667,
-      "rewards/question_recreation_reward_func": 0.37088586886723834,
+      "rewards/question_recreation_reward_func": 0.4011259178320567,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.313958336909612,
+      "rewards/strict_format_reward_func": 0.020833333333333332,
+      "rewards/xmlcount_reward_func": 0.5281666616598765,
       "step": 20
     },
     {
       "epoch": 5.0,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 8.599709872214589e-07,
-      "train_runtime": 984.9167,
-      "train_samples_per_second": 0.325,
-      "train_steps_per_second": 0.02
+      "train_loss": 8.411427756982448e-07,
+      "train_runtime": 1527.9925,
+      "train_samples_per_second": 0.209,
+      "train_steps_per_second": 0.013
     }
   ],
   "logging_steps": 2,