End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5479cd0cf4efc6cf294a93caa62162b3275f9b923ef72308a9efe07b78857b8c
+oid sha256:ee98810309e6116d19eac59d945d06eca9f20b708ff4029a5fd4a02a535e6df2
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 6.795386809699267e-07,
-    "train_runtime": 594.5988,
+    "train_loss": 6.468071546805732e-07,
+    "train_runtime": 564.7638,
     "train_samples": 11,
-    "train_samples_per_second": 0.538,
-    "train_steps_per_second": 0.034
+    "train_samples_per_second": 0.567,
+    "train_steps_per_second": 0.035
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 6.795386809699267e-07,
-    "train_runtime": 594.5988,
+    "train_loss": 6.468071546805732e-07,
+    "train_runtime": 564.7638,
     "train_samples": 11,
-    "train_samples_per_second": 0.538,
-    "train_steps_per_second": 0.034
+    "train_samples_per_second": 0.567,
+    "train_steps_per_second": 0.035
 }

trainer_state.json

@@ -10,203 +10,203 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 450.6875,
+      "completion_length": 419.1875,
       "epoch": 0.7272727272727273,
-      "grad_norm": 0.6235963106155396,
-      "kl": 0.0008548437472200021,
+      "grad_norm": 0.4025554955005646,
+      "kl": 0.0007115886437532026,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 0.9764094315469265,
-      "reward_std": 1.0595479533076286,
-      "rewards/concensus_correctness_reward_func": 0.010562499985098839,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 1.2856132220476866,
+      "reward_std": 0.9523135796189308,
+      "rewards/concensus_correctness_reward_func": 0.07937500067055225,
+      "rewards/consensus_reward_func": 0.4375,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5234719412401319,
+      "rewards/question_recreation_reward_func": 0.5550507390871644,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.19237499311566353,
+      "rewards/xmlcount_reward_func": 0.21368752419948578,
       "step": 2
     },
     {
-      "completion_length": 443.92857142857144,
+      "completion_length": 474.5,
       "epoch": 1.3636363636363638,
-      "grad_norm": 0.45450910925865173,
-      "kl": 0.0006687534956394561,
+      "grad_norm": 0.38703691959381104,
+      "kl": 0.0005440224652245108,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 0.6286563383681434,
-      "reward_std": 0.8178789062159402,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.07142857142857142,
+      "reward": 1.2591932330812727,
+      "reward_std": 0.9980412061725344,
+      "rewards/concensus_correctness_reward_func": 0.05500000076634543,
+      "rewards/consensus_reward_func": 0.42857142857142855,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.4648349327700479,
+      "rewards/question_recreation_reward_func": 0.630478927067348,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.09239285439252853,
+      "rewards/xmlcount_reward_func": 0.1451428553887776,
       "step": 4
     },
     {
-      "completion_length": 502.75,
+      "completion_length": 391.2142857142857,
       "epoch": 2.0,
-      "grad_norm": 0.3585743010044098,
-      "kl": 0.0007040688568460089,
+      "grad_norm": 0.34758260846138,
+      "kl": 0.0006707894499413669,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 0.7069842017122677,
-      "reward_std": 1.0232832048620497,
-      "rewards/concensus_correctness_reward_func": 0.013999999633857183,
-      "rewards/consensus_reward_func": 0.14285714285714285,
+      "reward": 1.0978498458862305,
+      "reward_std": 0.7003298039947238,
+      "rewards/concensus_correctness_reward_func": 0.03571428571428571,
+      "rewards/consensus_reward_func": 0.07142857142857142,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.4452699050307274,
+      "rewards/question_recreation_reward_func": 0.5435998290777206,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.0870000102690288,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.4471071479575975,
       "step": 6
     },
     {
-      "completion_length": 479.0,
+      "completion_length": 390.40625,
       "epoch": 2.7272727272727275,
-      "grad_norm": 0.5422738790512085,
-      "kl": 0.0007109935322660021,
+      "grad_norm": 0.48531419038772583,
+      "kl": 0.0007366152312897611,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 0.8491019681096077,
-      "reward_std": 1.0652108862996101,
-      "rewards/concensus_correctness_reward_func": 0.021124999970197678,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 0.70655282959342,
+      "reward_std": 0.9273619055747986,
+      "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": 0.3934769369661808,
+      "rewards/question_recreation_reward_func": 0.48299031239002943,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.03125,
-      "rewards/xmlcount_reward_func": 0.15325000137090683,
+      "rewards/strict_format_reward_func": 0.046875,
+      "rewards/xmlcount_reward_func": 0.1141875023022294,
       "step": 8
     },
     {
-      "completion_length": 472.5,
+      "completion_length": 419.7142857142857,
       "epoch": 3.3636363636363638,
-      "grad_norm": 0.37737125158309937,
-      "kl": 0.000669495011347213,
+      "grad_norm": 0.4628700911998749,
+      "kl": 0.0006924757409641254,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 0.9145918956824711,
-      "reward_std": 0.6341695764235088,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.14285714285714285,
+      "reward": 1.183305595602308,
+      "reward_std": 1.4018447484288896,
+      "rewards/concensus_correctness_reward_func": 0.1294285718883787,
+      "rewards/consensus_reward_func": 0.42857142857142855,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.6267347585942064,
-      "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.14499999902078084,
+      "rewards/question_recreation_reward_func": 0.6321269784654889,
+      "rewards/soft_format_reward_func": 0.03571428571428571,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": -0.060392867241586955,
       "step": 10
     },
     {
-      "completion_length": 426.14285714285717,
+      "completion_length": 386.57142857142856,
       "epoch": 4.0,
-      "grad_norm": 0.3590288460254669,
-      "kl": 0.0008327030476980976,
+      "grad_norm": 0.5434279441833496,
+      "kl": 0.0006527610489034227,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 0.974457072360175,
-      "reward_std": 1.0593456966536385,
-      "rewards/concensus_correctness_reward_func": 0.013999999633857183,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "reward": 0.7653476042406899,
+      "reward_std": 0.8373887177024569,
+      "rewards/concensus_correctness_reward_func": 0.030714286225182668,
+      "rewards/consensus_reward_func": 0.21428571428571427,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.550064218895776,
+      "rewards/question_recreation_reward_func": 0.5036690256425312,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.12467857343809945,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": -0.001178561576775142,
       "step": 12
     },
     {
-      "completion_length": 553.09375,
+      "completion_length": 386.3125,
       "epoch": 4.7272727272727275,
-      "grad_norm": 0.3092457354068756,
-      "kl": 0.0006055341618775856,
+      "grad_norm": 0.4393087327480316,
+      "kl": 0.0008711709633644205,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 0.6260320618748665,
-      "reward_std": 0.9468649551272392,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.125,
+      "reward": 1.2212899252772331,
+      "reward_std": 1.1120201498270035,
+      "rewards/concensus_correctness_reward_func": 0.048125000670552254,
+      "rewards/consensus_reward_func": 0.375,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5082195810973644,
+      "rewards/question_recreation_reward_func": 0.5600711535662413,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.007187509152572602,
+      "rewards/xmlcount_reward_func": 0.23809375427663326,
       "step": 14
     },
     {
-      "completion_length": 441.75,
+      "completion_length": 499.17857142857144,
       "epoch": 5.363636363636363,
-      "grad_norm": 0.300152987241745,
-      "kl": 0.0007428586873824575,
+      "grad_norm": 0.4070427715778351,
+      "kl": 0.000641536241995969,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 1.0600772287164415,
-      "reward_std": 0.9918473405497414,
-      "rewards/concensus_correctness_reward_func": 0.006999999816928591,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "reward": 0.622774315731866,
+      "reward_std": 1.1508670066084181,
+      "rewards/concensus_correctness_reward_func": 0.015357143112591334,
+      "rewards/consensus_reward_func": 0.07142857142857142,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5070772149733135,
+      "rewards/question_recreation_reward_func": 0.5319528803229332,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.26028572235788616,
+      "rewards/xmlcount_reward_func": 0.004035707030977521,
       "step": 16
     },
     {
-      "completion_length": 391.39285714285717,
+      "completion_length": 438.75,
       "epoch": 6.0,
-      "grad_norm": 0.4337644577026367,
-      "kl": 0.0008002218175014215,
+      "grad_norm": 0.38713639974594116,
+      "kl": 0.0006755495851393789,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 1.319738385932786,
-      "reward_std": 0.9078400475638253,
-      "rewards/concensus_correctness_reward_func": 0.042714285531214306,
+      "reward": 0.9099430825029101,
+      "reward_std": 1.0716484870229448,
+      "rewards/concensus_correctness_reward_func": 0.045357143240315576,
       "rewards/consensus_reward_func": 0.21428571428571427,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5893455360616956,
+      "rewards/question_recreation_reward_func": 0.5349787814276559,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.4733928569725582,
+      "rewards/xmlcount_reward_func": 0.11532143184116908,
       "step": 18
     },
     {
-      "completion_length": 460.75,
+      "completion_length": 378.78125,
       "epoch": 6.7272727272727275,
-      "grad_norm": 0.40713295340538025,
-      "kl": 0.0007180436950875446,
+      "grad_norm": 0.44257575273513794,
+      "kl": 0.0006841454087407328,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 0.8723896625451744,
-      "reward_std": 0.9429914578795433,
-      "rewards/concensus_correctness_reward_func": 0.04368750052526593,
+      "reward": 1.4425083473324776,
+      "reward_std": 1.0375258475542068,
+      "rewards/concensus_correctness_reward_func": 0.048125000670552254,
       "rewards/consensus_reward_func": 0.3125,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5706396587193012,
-      "rewards/soft_format_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.6988520696759224,
+      "rewards/soft_format_reward_func": 0.015625,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.05443750321865082,
+      "rewards/xmlcount_reward_func": 0.36740625742822886,
       "step": 20
     },
     {
       "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 6.795386809699267e-07,
-      "train_runtime": 594.5988,
-      "train_samples_per_second": 0.538,
-      "train_steps_per_second": 0.034
+      "train_loss": 6.468071546805732e-07,
+      "train_runtime": 564.7638,
+      "train_samples_per_second": 0.567,
+      "train_steps_per_second": 0.035
     }
   ],
   "logging_steps": 2,