End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dcee684657599e0415179ab3726d9ac27ee1927d8110384fecce84b1c8bccc6d
+oid sha256:7fb378f4ba903893b3467fc73d9638e12ded461e7d67ff61c062573a6cf22a76
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 6.14172711266292e-07,
-    "train_runtime": 1676.0262,
+    "train_loss": 6.621092552450136e-07,
+    "train_runtime": 1630.9123,
     "train_samples": 11,
-    "train_samples_per_second": 0.191,
+    "train_samples_per_second": 0.196,
     "train_steps_per_second": 0.012
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 6.14172711266292e-07,
-    "train_runtime": 1676.0262,
+    "train_loss": 6.621092552450136e-07,
+    "train_runtime": 1630.9123,
     "train_samples": 11,
-    "train_samples_per_second": 0.191,
+    "train_samples_per_second": 0.196,
     "train_steps_per_second": 0.012
 }

trainer_state.json

@@ -10,202 +10,202 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 613.0,
+      "completion_length": 430.96875,
       "epoch": 0.7272727272727273,
-      "grad_norm": 0.24522292613983154,
-      "kl": 0.0005619159528578166,
+      "grad_norm": 0.545572817325592,
+      "kl": 0.000742794818506809,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 0.6703863288275898,
-      "reward_std": 1.126855794340372,
-      "rewards/concensus_correctness_reward_func": 0.06362499902024865,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 0.9380801543593407,
+      "reward_std": 1.09766436368227,
+      "rewards/concensus_correctness_reward_func": 0.003874999936670065,
+      "rewards/consensus_reward_func": 0.375,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.44601133465766907,
+      "rewards/question_recreation_reward_func": 0.466673893854022,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.08924999088048935,
+      "rewards/xmlcount_reward_func": 0.09253125265240669,
       "step": 2
     },
     {
-      "completion_length": 589.4285714285714,
+      "completion_length": 417.67857142857144,
       "epoch": 1.3636363636363638,
-      "grad_norm": 0.27994006872177124,
-      "kl": 0.0006406649232043751,
+      "grad_norm": 0.40460318326950073,
+      "kl": 0.0007159239945134946,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 0.6909585297107697,
-      "reward_std": 1.2162732226508004,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.21428571428571427,
+      "reward": 1.2944229756082808,
+      "reward_std": 1.0728516238076347,
+      "rewards/concensus_correctness_reward_func": 0.03157142869063786,
+      "rewards/consensus_reward_func": 0.42857142857142855,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.14285714285714285,
-      "rewards/question_recreation_reward_func": 0.5334942394069263,
-      "rewards/soft_format_reward_func": 0.0,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5914943771702903,
+      "rewards/soft_format_reward_func": 0.017857142857142856,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.19967857003211975,
+      "rewards/xmlcount_reward_func": 0.2249285727739334,
       "step": 4
     },
     {
-      "completion_length": 429.35714285714283,
+      "completion_length": 383.2142857142857,
       "epoch": 2.0,
-      "grad_norm": 0.649292528629303,
-      "kl": 0.0007167610373081905,
+      "grad_norm": 0.4155687689781189,
+      "kl": 0.0007597279784801815,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 0.8096832718167987,
-      "reward_std": 1.2684654082570757,
-      "rewards/concensus_correctness_reward_func": 0.0061428573514734,
-      "rewards/consensus_reward_func": 0.14285714285714285,
+      "reward": 1.204940242426736,
+      "reward_std": 1.215496735913413,
+      "rewards/concensus_correctness_reward_func": 0.013285713536398751,
+      "rewards/consensus_reward_func": 0.35714285714285715,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.21428571428571427,
-      "rewards/question_recreation_reward_func": 0.3149689648832594,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.38661879513944897,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.11357141818319048,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.4478928723505565,
       "step": 6
     },
     {
-      "completion_length": 528.28125,
+      "completion_length": 539.875,
       "epoch": 2.7272727272727275,
-      "grad_norm": 0.4247061312198639,
-      "kl": 0.0006532615007017739,
+      "grad_norm": 0.3224160969257355,
+      "kl": 0.0006412690163415391,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 0.608462642878294,
-      "reward_std": 0.904325507581234,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.0,
+      "reward": 0.7731022760272026,
+      "reward_std": 0.9987973496317863,
+      "rewards/concensus_correctness_reward_func": 0.003874999936670065,
+      "rewards/consensus_reward_func": 0.3125,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.1875,
-      "rewards/question_recreation_reward_func": 0.42793136183172464,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.4241334795951843,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": -0.022593742236495018,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.03259375039488077,
       "step": 8
     },
     {
-      "completion_length": 515.9642857142857,
+      "completion_length": 411.10714285714283,
       "epoch": 3.3636363636363638,
-      "grad_norm": 0.6112687587738037,
-      "kl": 0.0007279971780787621,
+      "grad_norm": 0.3444700837135315,
+      "kl": 0.0007552724231832794,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 0.7809470061744962,
-      "reward_std": 1.25520954813276,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.14285714285714285,
+      "reward": 1.462347882134574,
+      "reward_std": 1.1783929892948695,
+      "rewards/concensus_correctness_reward_func": 0.0565714282648904,
+      "rewards/consensus_reward_func": 0.5,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.2857142857142857,
-      "rewards/question_recreation_reward_func": 0.5194827147892543,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.6305621351514544,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.1671071414436613,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.25735713115760256,
       "step": 10
     },
     {
-      "completion_length": 520.0357142857143,
+      "completion_length": 435.85714285714283,
       "epoch": 4.0,
-      "grad_norm": 0.29373371601104736,
-      "kl": 0.0006241699115240148,
+      "grad_norm": 0.2986714243888855,
+      "kl": 0.0006829753069074026,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 0.7070630788803101,
-      "reward_std": 0.9244653582572937,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.14285714285714285,
+      "reward": 0.9998348184994289,
+      "reward_std": 0.8780212189470019,
+      "rewards/concensus_correctness_reward_func": 0.03328571521810123,
+      "rewards/consensus_reward_func": 0.21428571428571427,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5393130736691611,
+      "rewards/question_recreation_reward_func": 0.5905490645340511,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.024892862354006087,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.1438571416905948,
       "step": 12
     },
     {
-      "completion_length": 595.1875,
+      "completion_length": 485.96875,
       "epoch": 4.7272727272727275,
-      "grad_norm": 0.2765561640262604,
-      "kl": 0.0006751239561708644,
+      "grad_norm": 0.3659514784812927,
+      "kl": 0.00071150277290144,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 0.6923840567469597,
-      "reward_std": 0.9703349322080612,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.0625,
+      "reward": 1.6259073615074158,
+      "reward_std": 1.6027247682213783,
+      "rewards/concensus_correctness_reward_func": 0.02474999986588955,
+      "rewards/consensus_reward_func": 0.625,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0625,
-      "rewards/question_recreation_reward_func": 0.5641340762376785,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.6519073890522122,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.0032499851658940315,
+      "rewards/strict_format_reward_func": 0.046875,
+      "rewards/xmlcount_reward_func": 0.2773750014603138,
       "step": 14
     },
     {
-      "completion_length": 496.7142857142857,
+      "completion_length": 416.25,
       "epoch": 5.363636363636363,
-      "grad_norm": 0.3044429421424866,
-      "kl": 0.0006677072919306479,
+      "grad_norm": 0.45220518112182617,
+      "kl": 0.0006384395861199924,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 1.6827466019562312,
-      "reward_std": 2.3171258632625853,
-      "rewards/concensus_correctness_reward_func": 0.7142857142857143,
-      "rewards/consensus_reward_func": 0.14285714285714285,
+      "reward": 1.3776842228003912,
+      "reward_std": 1.10369587796075,
+      "rewards/concensus_correctness_reward_func": 0.00885714271238872,
+      "rewards/consensus_reward_func": 0.6428571428571429,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.14285714285714285,
-      "rewards/question_recreation_reward_func": 0.5389608804668699,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.6524342170783451,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.14378570978130614,
+      "rewards/xmlcount_reward_func": 0.07353571057319641,
       "step": 16
     },
     {
-      "completion_length": 486.39285714285717,
+      "completion_length": 446.64285714285717,
       "epoch": 6.0,
-      "grad_norm": 0.32584676146507263,
-      "kl": 0.0007059235857533557,
+      "grad_norm": 0.3257089853286743,
+      "kl": 0.0007032139941916934,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 0.6391109717743737,
-      "reward_std": 0.8559697781290326,
-      "rewards/concensus_correctness_reward_func": 0.0,
+      "reward": 1.0999488064220972,
+      "reward_std": 0.9497513175010681,
+      "rewards/concensus_correctness_reward_func": 0.0282857141324452,
       "rewards/consensus_reward_func": 0.14285714285714285,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.39850386125700815,
+      "rewards/question_recreation_reward_func": 0.5714488093342099,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.09774999746254512,
+      "rewards/xmlcount_reward_func": 0.357357137969562,
       "step": 18
     },
     {
-      "completion_length": 527.46875,
+      "completion_length": 450.65625,
       "epoch": 6.7272727272727275,
-      "grad_norm": 0.3653308153152466,
-      "kl": 0.0006879912543809041,
+      "grad_norm": 0.4257511496543884,
+      "kl": 0.0007408625024254434,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 0.6358731538057327,
-      "reward_std": 1.2101131454110146,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.125,
+      "reward": 1.1619919557124376,
+      "reward_std": 1.1545419916510582,
+      "rewards/concensus_correctness_reward_func": 0.02474999986588955,
+      "rewards/consensus_reward_func": 0.25,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.125,
-      "rewards/question_recreation_reward_func": 0.45649817399680614,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5668044406920671,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": -0.08625001832842827,
+      "rewards/strict_format_reward_func": 0.0625,
+      "rewards/xmlcount_reward_func": 0.2579374946653843,
       "step": 20
     },
     {
       "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 6.14172711266292e-07,
-      "train_runtime": 1676.0262,
-      "train_samples_per_second": 0.191,
+      "train_loss": 6.621092552450136e-07,
+      "train_runtime": 1630.9123,
+      "train_samples_per_second": 0.196,
       "train_steps_per_second": 0.012
     }
   ],