End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d3361331a01d4cdd6f6d18e077bd40664b7e6facb8c36f9f47d7bfba2c0762e5
+oid sha256:1713873e51b1409fa90eed104a59fca153d610dde9d9fd7cf9dfe80f8bdbdff5
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 7.388806182007102e-07,
-    "train_runtime": 1494.2718,
+    "train_loss": 8.672385831687279e-07,
+    "train_runtime": 1475.0626,
     "train_samples": 14,
-    "train_samples_per_second": 0.214,
-    "train_steps_per_second": 0.013
+    "train_samples_per_second": 0.217,
+    "train_steps_per_second": 0.014
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 7.388806182007102e-07,
-    "train_runtime": 1494.2718,
+    "train_loss": 8.672385831687279e-07,
+    "train_runtime": 1475.0626,
     "train_samples": 14,
-    "train_samples_per_second": 0.214,
-    "train_steps_per_second": 0.013
+    "train_samples_per_second": 0.217,
+    "train_steps_per_second": 0.014
 }

trainer_state.json

@@ -10,203 +10,203 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 327.5625,
+      "completion_length": 349.75,
       "epoch": 0.5714285714285714,
-      "grad_norm": 0.5660309195518494,
-      "kl": 0.0007402191331493668,
+      "grad_norm": 0.5519270300865173,
+      "kl": 0.0008171074368874542,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 1.9583971425890923,
-      "reward_std": 1.6289257481694221,
-      "rewards/concensus_correctness_reward_func": 0.30149999633431435,
-      "rewards/consensus_reward_func": 0.1875,
+      "reward": 3.053681083023548,
+      "reward_std": 3.1452729776501656,
+      "rewards/concensus_correctness_reward_func": 1.453124986961484,
+      "rewards/consensus_reward_func": 0.3125,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.625,
-      "rewards/question_recreation_reward_func": 0.49680337868630886,
+      "rewards/final_correctness_reward_func": 0.4375,
+      "rewards/question_recreation_reward_func": 0.4569310378283262,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0625,
-      "rewards/xmlcount_reward_func": 0.2850937619805336,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.393624986987561,
       "step": 2
     },
     {
-      "completion_length": 249.66666666666666,
+      "completion_length": 255.375,
       "epoch": 1.0,
-      "grad_norm": 0.5000850558280945,
-      "kl": 0.0008946122834458947,
+      "grad_norm": 0.5439753532409668,
+      "kl": 0.001034095068462193,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 4.727487623691559,
-      "reward_std": 4.456474117934704,
-      "rewards/concensus_correctness_reward_func": 2.944833322583387,
-      "rewards/consensus_reward_func": 0.5,
+      "reward": 3.095159411430359,
+      "reward_std": 3.0661760171254477,
+      "rewards/concensus_correctness_reward_func": 0.9074166642191509,
+      "rewards/consensus_reward_func": 0.5833333333333334,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.3333333333333333,
-      "rewards/question_recreation_reward_func": 0.49019574125607807,
+      "rewards/final_correctness_reward_func": 0.5833333333333334,
+      "rewards/question_recreation_reward_func": 0.539076047639052,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.45912500595053035,
+      "rewards/xmlcount_reward_func": 0.4820000057419141,
       "step": 4
     },
     {
-      "completion_length": 314.6875,
+      "completion_length": 280.5625,
       "epoch": 1.5714285714285714,
-      "grad_norm": 0.49750667810440063,
-      "kl": 0.0009101929572352674,
+      "grad_norm": 0.622399091720581,
+      "kl": 0.0009089883824344724,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 3.238606572151184,
-      "reward_std": 3.600623197853565,
-      "rewards/concensus_correctness_reward_func": 1.3313124999403954,
-      "rewards/consensus_reward_func": 0.1875,
+      "reward": 2.3543258160352707,
+      "reward_std": 2.355586849153042,
+      "rewards/concensus_correctness_reward_func": 0.7060624975711107,
+      "rewards/consensus_reward_func": 0.3125,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 1.0,
-      "rewards/question_recreation_reward_func": 0.440856559202075,
+      "rewards/final_correctness_reward_func": 0.3125,
+      "rewards/question_recreation_reward_func": 0.3981695678085089,
       "rewards/soft_format_reward_func": 0.015625,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.2633125074207783,
+      "rewards/xmlcount_reward_func": 0.6094687469303608,
       "step": 6
     },
     {
-      "completion_length": 276.4166666666667,
+      "completion_length": 245.125,
       "epoch": 2.0,
-      "grad_norm": 0.3782678544521332,
-      "kl": 0.0007823882527494183,
+      "grad_norm": 0.6636540293693542,
+      "kl": 0.0008719059987924993,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 1.9467649857203166,
-      "reward_std": 1.3353716631730397,
-      "rewards/concensus_correctness_reward_func": 0.29350000619888306,
-      "rewards/consensus_reward_func": 0.4166666666666667,
+      "reward": 4.6469279527664185,
+      "reward_std": 4.8226219316323595,
+      "rewards/concensus_correctness_reward_func": 2.503499999952813,
+      "rewards/consensus_reward_func": 0.6666666666666666,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.5833333333333334,
-      "rewards/question_recreation_reward_func": 0.44318165878454846,
+      "rewards/question_recreation_reward_func": 0.470219482978185,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.21008333067099252,
+      "rewards/xmlcount_reward_func": 0.4232083422442277,
       "step": 8
     },
     {
-      "completion_length": 292.40625,
+      "completion_length": 326.25,
       "epoch": 2.571428571428571,
-      "grad_norm": 0.533460259437561,
-      "kl": 0.0008978317273431458,
+      "grad_norm": 0.4317708909511566,
+      "kl": 0.0007528728965553455,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 2.953047474846244,
-      "reward_std": 3.533324506133795,
-      "rewards/concensus_correctness_reward_func": 1.3231874704360962,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 5.892444729804993,
+      "reward_std": 3.724656067788601,
+      "rewards/concensus_correctness_reward_func": 3.8065624833106995,
+      "rewards/consensus_reward_func": 0.75,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.625,
-      "rewards/question_recreation_reward_func": 0.5515474192798138,
+      "rewards/final_correctness_reward_func": 0.5,
+      "rewards/question_recreation_reward_func": 0.5077260434627533,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.20331249106675386,
+      "rewards/xmlcount_reward_func": 0.3281562621705234,
       "step": 10
     },
     {
-      "completion_length": 331.2916666666667,
+      "completion_length": 299.0,
       "epoch": 3.0,
-      "grad_norm": 0.38148120045661926,
-      "kl": 0.0007997689438828578,
+      "grad_norm": 0.46557772159576416,
+      "kl": 0.0008803623107572397,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 2.8399418791135154,
-      "reward_std": 2.7565351873636246,
-      "rewards/concensus_correctness_reward_func": 1.1515833387772243,
-      "rewards/consensus_reward_func": 0.16666666666666666,
+      "reward": 3.283529301484426,
+      "reward_std": 3.281624217828115,
+      "rewards/concensus_correctness_reward_func": 1.7380000008270144,
+      "rewards/consensus_reward_func": 0.5,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.75,
-      "rewards/question_recreation_reward_func": 0.4849003131190936,
-      "rewards/soft_format_reward_func": 0.020833333333333332,
+      "rewards/final_correctness_reward_func": 0.08333333333333333,
+      "rewards/question_recreation_reward_func": 0.45998769005139667,
+      "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.26595834642648697,
+      "rewards/xmlcount_reward_func": 0.5022083173195521,
       "step": 12
     },
     {
-      "completion_length": 330.875,
+      "completion_length": 329.59375,
       "epoch": 3.571428571428571,
-      "grad_norm": 0.5229682922363281,
-      "kl": 0.0009297839205828495,
+      "grad_norm": 0.5926380753517151,
+      "kl": 0.0017208340723300353,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 3.057193100452423,
-      "reward_std": 3.6767283603549004,
-      "rewards/concensus_correctness_reward_func": 1.349062496679835,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 3.203888300806284,
+      "reward_std": 3.8441673293709755,
+      "rewards/concensus_correctness_reward_func": 1.2763749994337559,
+      "rewards/consensus_reward_func": 0.5,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.625,
-      "rewards/question_recreation_reward_func": 0.6216618604958057,
+      "rewards/final_correctness_reward_func": 0.5,
+      "rewards/question_recreation_reward_func": 0.6332946009933949,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.19584372825920582,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.29421874321997166,
       "step": 14
     },
     {
-      "completion_length": 322.25,
+      "completion_length": 306.75,
       "epoch": 4.0,
-      "grad_norm": 0.2658565044403076,
-      "kl": 0.000788417411968112,
+      "grad_norm": 0.3222191035747528,
+      "kl": 0.0009899105158789705,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 4.762086749076843,
-      "reward_std": 6.355244348446528,
-      "rewards/concensus_correctness_reward_func": 2.646250009536743,
-      "rewards/consensus_reward_func": 0.5833333333333334,
+      "reward": 1.3310787379741669,
+      "reward_std": 1.0895682473977406,
+      "rewards/concensus_correctness_reward_func": 0.11124999821186066,
+      "rewards/consensus_reward_func": 0.16666666666666666,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.8333333333333334,
-      "rewards/question_recreation_reward_func": 0.4211700198551019,
+      "rewards/final_correctness_reward_func": 0.16666666666666666,
+      "rewards/question_recreation_reward_func": 0.39641206959883374,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.2779999999329448,
+      "rewards/xmlcount_reward_func": 0.4900833247229457,
       "step": 16
     },
     {
-      "completion_length": 301.125,
+      "completion_length": 311.9375,
       "epoch": 4.571428571428571,
-      "grad_norm": 0.4315449595451355,
-      "kl": 0.0009034630784299225,
+      "grad_norm": 0.4399689733982086,
+      "kl": 0.0008669333474244922,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 2.29608665779233,
-      "reward_std": 2.6273415572941303,
-      "rewards/concensus_correctness_reward_func": 0.8200625046156347,
-      "rewards/consensus_reward_func": 0.1875,
+      "reward": 3.1671593114733696,
+      "reward_std": 2.7661515697836876,
+      "rewards/concensus_correctness_reward_func": 1.3582499958574772,
+      "rewards/consensus_reward_func": 0.5625,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.4375,
-      "rewards/question_recreation_reward_func": 0.561367915943265,
+      "rewards/final_correctness_reward_func": 0.375,
+      "rewards/question_recreation_reward_func": 0.5038155522197485,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.2896562432870269,
+      "rewards/xmlcount_reward_func": 0.36759375035762787,
       "step": 18
     },
     {
-      "completion_length": 278.0,
+      "completion_length": 275.6666666666667,
       "epoch": 5.0,
-      "grad_norm": 0.3298706114292145,
-      "kl": 0.0008178624557331204,
+      "grad_norm": 0.46918413043022156,
+      "kl": 0.0009636022441554815,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 4.6252261698246,
-      "reward_std": 4.7473550538222,
-      "rewards/concensus_correctness_reward_func": 2.7185000005799034,
-      "rewards/consensus_reward_func": 0.4166666666666667,
+      "reward": 2.517007460196813,
+      "reward_std": 2.4441120252013206,
+      "rewards/concensus_correctness_reward_func": 0.9087499976158142,
+      "rewards/consensus_reward_func": 0.25,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.75,
-      "rewards/question_recreation_reward_func": 0.380059281984965,
+      "rewards/final_correctness_reward_func": 0.5,
+      "rewards/question_recreation_reward_func": 0.4217992251118024,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.020833333333333332,
-      "rewards/xmlcount_reward_func": 0.33916666979591054,
+      "rewards/strict_format_reward_func": 0.041666666666666664,
+      "rewards/xmlcount_reward_func": 0.39479164717098075,
       "step": 20
     },
     {
       "epoch": 5.0,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 7.388806182007102e-07,
-      "train_runtime": 1494.2718,
-      "train_samples_per_second": 0.214,
-      "train_steps_per_second": 0.013
+      "train_loss": 8.672385831687279e-07,
+      "train_runtime": 1475.0626,
+      "train_samples_per_second": 0.217,
+      "train_steps_per_second": 0.014
     }
   ],
   "logging_steps": 2,