End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5ef448650e684c51b0a72c11862e4ad207ff4f6bbeffa6777e76881bc14fd7df
+oid sha256:ef427109c12e9c134ef312ddff3602d498dc4299f20d5ed75d136d215a012939
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 1.5633871726095094e-06,
-    "train_runtime": 669.1206,
+    "train_loss": 1.5715185440967617e-06,
+    "train_runtime": 692.3425,
     "train_samples": 11,
-    "train_samples_per_second": 0.478,
-    "train_steps_per_second": 0.03
+    "train_samples_per_second": 0.462,
+    "train_steps_per_second": 0.029
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 1.5633871726095094e-06,
-    "train_runtime": 669.1206,
+    "train_loss": 1.5715185440967617e-06,
+    "train_runtime": 692.3425,
     "train_samples": 11,
-    "train_samples_per_second": 0.478,
-    "train_steps_per_second": 0.03
+    "train_samples_per_second": 0.462,
+    "train_steps_per_second": 0.029
 }

trainer_state.json

@@ -10,203 +10,203 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 508.84375,
+      "completion_length": 517.1875,
       "epoch": 0.7272727272727273,
-      "grad_norm": 0.50871741771698,
-      "kl": 0.002379992227361072,
+      "grad_norm": 0.6420091390609741,
+      "kl": 0.0019688690153998323,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 0.948925256729126,
-      "reward_std": 1.3094163816422224,
-      "rewards/concensus_correctness_reward_func": 0.046875,
-      "rewards/consensus_reward_func": 0.375,
+      "reward": 1.5067992843687534,
+      "reward_std": 0.986472899094224,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.4375,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5915502645075321,
+      "rewards/question_recreation_reward_func": 0.6470805648714304,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.06449999287724495,
+      "rewards/strict_format_reward_func": 0.0625,
+      "rewards/xmlcount_reward_func": 0.3597187642008066,
       "step": 2
     },
     {
-      "completion_length": 492.2142857142857,
+      "completion_length": 547.2142857142857,
       "epoch": 1.3636363636363638,
-      "grad_norm": 0.5408371090888977,
-      "kl": 0.001495275238994509,
+      "grad_norm": 0.4010975658893585,
+      "kl": 0.0015678029657075448,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 0.8839784115552902,
-      "reward_std": 1.1927723671708788,
-      "rewards/concensus_correctness_reward_func": 0.026285714336803982,
-      "rewards/consensus_reward_func": 0.21428571428571427,
+      "reward": 1.2731680827481406,
+      "reward_std": 1.2122740660394942,
+      "rewards/concensus_correctness_reward_func": 0.019857142652784075,
+      "rewards/consensus_reward_func": 0.2857142857142857,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.07142857142857142,
-      "rewards/question_recreation_reward_func": 0.5831212401390076,
+      "rewards/question_recreation_reward_func": 0.6732038089207241,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.03571428571428571,
-      "rewards/xmlcount_reward_func": -0.04685715053762708,
+      "rewards/xmlcount_reward_func": 0.18725000321865082,
       "step": 4
     },
     {
-      "completion_length": 487.75,
+      "completion_length": 573.75,
       "epoch": 2.0,
-      "grad_norm": 0.3275243937969208,
-      "kl": 0.001916582729401333,
+      "grad_norm": 0.4096458852291107,
+      "kl": 0.0014084425305814616,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 1.378743290901184,
-      "reward_std": 1.215172827243805,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.42857142857142855,
+      "reward": 1.5097219433103288,
+      "reward_std": 1.2099641816956657,
+      "rewards/concensus_correctness_reward_func": 0.052571428673607964,
+      "rewards/consensus_reward_func": 0.6428571428571429,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.14285714285714285,
-      "rewards/question_recreation_reward_func": 0.5664218557732446,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.6305076128670147,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.2408928551844188,
+      "rewards/xmlcount_reward_func": 0.1837857280458723,
       "step": 6
     },
     {
-      "completion_length": 494.1875,
+      "completion_length": 541.46875,
       "epoch": 2.7272727272727275,
-      "grad_norm": 0.39771324396133423,
-      "kl": 0.0015884681488387287,
+      "grad_norm": 0.45404794812202454,
+      "kl": 0.0016080429923022166,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 1.2462583221495152,
-      "reward_std": 1.4062652960419655,
-      "rewards/concensus_correctness_reward_func": 0.0625,
-      "rewards/consensus_reward_func": 0.5625,
+      "reward": 0.8502027094364166,
+      "reward_std": 1.3040681928396225,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.25,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5716645419597626,
+      "rewards/question_recreation_reward_func": 0.5663901790976524,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.04959375970065594,
+      "rewards/xmlcount_reward_func": 0.03381251357495785,
       "step": 8
     },
     {
-      "completion_length": 530.1071428571429,
+      "completion_length": 601.5357142857143,
       "epoch": 3.3636363636363638,
-      "grad_norm": 0.7798460125923157,
-      "kl": 0.0016198848440710986,
+      "grad_norm": 0.3453723192214966,
+      "kl": 0.001553945301566273,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 1.0500428378582,
-      "reward_std": 0.8941459315163749,
-      "rewards/concensus_correctness_reward_func": 0.07935714295932225,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "reward": 1.2584942494119917,
+      "reward_std": 1.4121184519359045,
+      "rewards/concensus_correctness_reward_func": 0.019857142652784075,
+      "rewards/consensus_reward_func": 0.42857142857142855,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.695364270891462,
+      "rewards/final_correctness_reward_func": 0.07142857142857142,
+      "rewards/question_recreation_reward_func": 0.6241728067398071,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.03571428571428571,
-      "rewards/xmlcount_reward_func": -0.04610714529241834,
+      "rewards/xmlcount_reward_func": 0.07875001643385206,
       "step": 10
     },
     {
-      "completion_length": 522.0,
+      "completion_length": 450.0,
       "epoch": 4.0,
-      "grad_norm": 0.3984813392162323,
-      "kl": 0.0013336356579592185,
+      "grad_norm": 0.46400099992752075,
+      "kl": 0.0015125636543546403,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 1.0651543481009347,
-      "reward_std": 0.872273291860308,
+      "reward": 0.887412177664893,
+      "reward_std": 0.8893355514322009,
       "rewards/concensus_correctness_reward_func": 0.0,
       "rewards/consensus_reward_func": 0.14285714285714285,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.14285714285714285,
-      "rewards/question_recreation_reward_func": 0.5091900740350995,
+      "rewards/final_correctness_reward_func": 0.21428571428571427,
+      "rewards/question_recreation_reward_func": 0.5103407331875393,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.270249998995236,
+      "rewards/xmlcount_reward_func": 0.01992857881954738,
       "step": 12
     },
     {
-      "completion_length": 534.03125,
+      "completion_length": 586.8125,
       "epoch": 4.7272727272727275,
-      "grad_norm": 0.5504440069198608,
-      "kl": 0.0016409848467446864,
+      "grad_norm": 0.28827542066574097,
+      "kl": 0.0011791825891123153,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 1.3454676195979118,
-      "reward_std": 1.3569010496139526,
-      "rewards/concensus_correctness_reward_func": 0.1484375,
-      "rewards/consensus_reward_func": 0.4375,
+      "reward": 0.6554815284907818,
+      "reward_std": 0.9318620897829533,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.125,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.6023426447063684,
+      "rewards/question_recreation_reward_func": 0.5917002856731415,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.1571875074878335,
+      "rewards/xmlcount_reward_func": -0.06121874740347266,
       "step": 14
     },
     {
-      "completion_length": 520.8928571428571,
+      "completion_length": 466.39285714285717,
       "epoch": 5.363636363636363,
-      "grad_norm": 0.5635874271392822,
-      "kl": 0.001528095329246883,
+      "grad_norm": 1.2276360988616943,
+      "kl": 0.0022477747074195315,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 0.8026008499520165,
-      "reward_std": 0.8021095607961927,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.35714285714285715,
+      "reward": 1.2439466140099935,
+      "reward_std": 1.1877283496516091,
+      "rewards/concensus_correctness_reward_func": 0.008928571428571428,
+      "rewards/consensus_reward_func": 0.2857142857142857,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.14285714285714285,
-      "rewards/question_recreation_reward_func": 0.5305651660476413,
+      "rewards/final_correctness_reward_func": 0.07142857142857142,
+      "rewards/question_recreation_reward_func": 0.5682680138519832,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.22796428203582764,
+      "rewards/strict_format_reward_func": 0.03571428571428571,
+      "rewards/xmlcount_reward_func": 0.27389287097113474,
       "step": 16
     },
     {
-      "completion_length": 497.42857142857144,
+      "completion_length": 483.9642857142857,
       "epoch": 6.0,
-      "grad_norm": 0.35303908586502075,
-      "kl": 0.0022293299054061727,
+      "grad_norm": 4.336320400238037,
+      "kl": 0.0021773939869100495,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 1.2344707633767809,
-      "reward_std": 1.1399089055401939,
-      "rewards/concensus_correctness_reward_func": 0.11607142857142858,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "reward": 1.505273767880031,
+      "reward_std": 1.1624436804226466,
+      "rewards/concensus_correctness_reward_func": 0.019857142652784075,
+      "rewards/consensus_reward_func": 0.21428571428571427,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.07142857142857142,
-      "rewards/question_recreation_reward_func": 0.5813993364572525,
+      "rewards/question_recreation_reward_func": 0.7239166242735726,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.017857142857142856,
-      "rewards/xmlcount_reward_func": 0.16199999621936254,
+      "rewards/xmlcount_reward_func": 0.45792857238224577,
       "step": 18
     },
     {
-      "completion_length": 629.84375,
+      "completion_length": 524.40625,
       "epoch": 6.7272727272727275,
-      "grad_norm": 0.3519224524497986,
-      "kl": 0.0012083955953130499,
+      "grad_norm": 0.5881478190422058,
+      "kl": 0.0016475072770845145,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 0.8868681024760008,
-      "reward_std": 1.070122130215168,
-      "rewards/concensus_correctness_reward_func": 0.140625,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 0.6808819696307182,
+      "reward_std": 1.0567774400115013,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.1875,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.7396180965006351,
-      "rewards/soft_format_reward_func": 0.015625,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": -0.2746249958872795,
+      "rewards/question_recreation_reward_func": 0.6210694797337055,
+      "rewards/soft_format_reward_func": 0.0,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": -0.1276874914765358,
       "step": 20
     },
     {
       "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 1.5633871726095094e-06,
-      "train_runtime": 669.1206,
-      "train_samples_per_second": 0.478,
-      "train_steps_per_second": 0.03
+      "train_loss": 1.5715185440967617e-06,
+      "train_runtime": 692.3425,
+      "train_samples_per_second": 0.462,
+      "train_steps_per_second": 0.029
     }
   ],
   "logging_steps": 2,