End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5b0aca02560bbc514c3b066ec1ca723bbef591fa46de062d4f9bb4ffbbfaeaa2
+oid sha256:ded1be1d8f4fb77fb5bcb942b465102bb922be1aeb409d15615c0487a4ddcff2
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 7.440462638896861e-07,
-    "train_runtime": 1056.8449,
-    "train_samples": 20,
-    "train_samples_per_second": 0.303,
-    "train_steps_per_second": 0.019
+    "train_loss": 6.568556671027181e-07,
+    "train_runtime": 565.9009,
+    "train_samples": 11,
+    "train_samples_per_second": 0.565,
+    "train_steps_per_second": 0.035
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 7.440462638896861e-07,
-    "train_runtime": 1056.8449,
-    "train_samples": 20,
-    "train_samples_per_second": 0.303,
-    "train_steps_per_second": 0.019
+    "train_loss": 6.568556671027181e-07,
+    "train_runtime": 565.9009,
+    "train_samples": 11,
+    "train_samples_per_second": 0.565,
+    "train_steps_per_second": 0.035
 }

trainer_state.json

@@ -2,7 +2,7 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 4.0,
+  "epoch": 6.7272727272727275,
   "eval_steps": 500,
   "global_step": 20,
   "is_hyper_param_search": false,
@@ -10,209 +10,209 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 524.71875,
-      "epoch": 0.4,
-      "grad_norm": 0.32683268189430237,
-      "kl": 0.0007783704058965668,
+      "completion_length": 469.0,
+      "epoch": 0.7272727272727273,
+      "grad_norm": 4.352033615112305,
+      "kl": 0.0007249983573274221,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 1.8962685018777847,
-      "reward_std": 1.202414907515049,
-      "rewards/concensus_correctness_reward_func": 0.0078125,
-      "rewards/consensus_reward_func": 0.75,
+      "reward": 1.4192078448832035,
+      "reward_std": 1.0422990955412388,
+      "rewards/concensus_correctness_reward_func": 0.07124999910593033,
+      "rewards/consensus_reward_func": 0.25,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5507372403517365,
+      "rewards/question_recreation_reward_func": 0.49339537788182497,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.5720937680453062,
+      "rewards/strict_format_reward_func": 0.046875,
+      "rewards/xmlcount_reward_func": 0.5576874963007867,
       "step": 2
     },
     {
-      "completion_length": 493.8125,
-      "epoch": 0.8,
-      "grad_norm": 0.43430647253990173,
-      "kl": 0.0006869099743198603,
+      "completion_length": 420.4642857142857,
+      "epoch": 1.3636363636363638,
+      "grad_norm": 0.4544900953769684,
+      "kl": 0.0007280591276607343,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 1.277583060786128,
-      "reward_std": 1.1487413346767426,
+      "reward": 1.2708178418023246,
+      "reward_std": 0.9965992761509759,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.5,
+      "rewards/consensus_reward_func": 0.42857142857142855,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.3125,
-      "rewards/question_recreation_reward_func": 0.5429580342024565,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.6406035221048764,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": -0.09349999623373151,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.18378571953092301,
       "step": 4
     },
     {
-      "completion_length": 430.0,
-      "epoch": 1.2,
-      "grad_norm": 0.7402258515357971,
-      "kl": 0.0007218625542009249,
+      "completion_length": 447.64285714285717,
+      "epoch": 2.0,
+      "grad_norm": 0.5355516076087952,
+      "kl": 0.0006103158022077489,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 1.834559328854084,
-      "reward_std": 1.5259039457887411,
-      "rewards/concensus_correctness_reward_func": 0.11962500214576721,
-      "rewards/consensus_reward_func": 0.5,
+      "reward": 1.648884313447135,
+      "reward_std": 0.8219013725008283,
+      "rewards/concensus_correctness_reward_func": 0.18214285054377147,
+      "rewards/consensus_reward_func": 0.35714285714285715,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.25,
-      "rewards/question_recreation_reward_func": 0.5861218366771936,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.7413843359266009,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.03125,
-      "rewards/xmlcount_reward_func": 0.34756250120699406,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.35035714294229237,
       "step": 6
     },
     {
-      "completion_length": 483.25,
-      "epoch": 1.6,
-      "grad_norm": 0.5379654765129089,
-      "kl": 0.0007469515039701946,
+      "completion_length": 479.625,
+      "epoch": 2.7272727272727275,
+      "grad_norm": 0.5460347533226013,
+      "kl": 0.0006964662425161805,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 1.4211411476135254,
-      "reward_std": 1.0882117003202438,
-      "rewards/concensus_correctness_reward_func": 0.015062499791383743,
-      "rewards/consensus_reward_func": 0.4375,
+      "reward": 1.2303326353430748,
+      "reward_std": 0.737933162599802,
+      "rewards/concensus_correctness_reward_func": 0.046875,
+      "rewards/consensus_reward_func": 0.1875,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0625,
-      "rewards/question_recreation_reward_func": 0.5641099270433187,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5855201371014118,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.341968750115484,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.3948124935850501,
       "step": 8
     },
     {
-      "completion_length": 494.1875,
-      "epoch": 2.0,
-      "grad_norm": 0.4419167637825012,
-      "kl": 0.0006833942279627081,
+      "completion_length": 441.89285714285717,
+      "epoch": 3.3636363636363638,
+      "grad_norm": 0.5250329971313477,
+      "kl": 0.0006974046035403651,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 1.5143687799572945,
-      "reward_std": 0.7561689466238022,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.375,
+      "reward": 1.554992641721453,
+      "reward_std": 1.1143607326916285,
+      "rewards/concensus_correctness_reward_func": 0.10971428666796003,
+      "rewards/consensus_reward_func": 0.35714285714285715,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.6822125148028135,
+      "rewards/question_recreation_reward_func": 0.5665283918912921,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.45715625397861004,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.5037500091961452,
       "step": 10
     },
     {
-      "completion_length": 407.59375,
-      "epoch": 2.4,
-      "grad_norm": 1.2626070976257324,
-      "kl": 0.0008575995125283953,
+      "completion_length": 490.60714285714283,
+      "epoch": 4.0,
+      "grad_norm": 0.37077268958091736,
+      "kl": 0.0006115809207715626,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 1.3198797106742859,
-      "reward_std": 1.149541474878788,
-      "rewards/concensus_correctness_reward_func": 0.0234375,
-      "rewards/consensus_reward_func": 0.5625,
+      "reward": 0.7811040367398944,
+      "reward_std": 0.9223696504320417,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.07142857142857142,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.34706723038107157,
+      "rewards/question_recreation_reward_func": 0.5047111681529454,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.386875007301569,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.18710715004376002,
       "step": 12
     },
     {
-      "completion_length": 627.4375,
-      "epoch": 2.8,
-      "grad_norm": 0.2844941020011902,
-      "kl": 0.0007131319762265775,
+      "completion_length": 506.15625,
+      "epoch": 4.7272727272727275,
+      "grad_norm": 0.48777249455451965,
+      "kl": 0.0007442439818987623,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 1.8757458478212357,
-      "reward_std": 1.1824609991163015,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.375,
+      "reward": 1.5310741290450096,
+      "reward_std": 1.209854133427143,
+      "rewards/concensus_correctness_reward_func": 0.11924999952316284,
+      "rewards/consensus_reward_func": 0.5625,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.3125,
-      "rewards/question_recreation_reward_func": 0.7107458561658859,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.4672929239459336,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.47750001959502697,
+      "rewards/xmlcount_reward_func": 0.38203124795109034,
       "step": 14
     },
     {
-      "completion_length": 389.15625,
-      "epoch": 3.2,
-      "grad_norm": 7.29202127456665,
-      "kl": 0.00080794016685104,
+      "completion_length": 375.14285714285717,
+      "epoch": 5.363636363636363,
+      "grad_norm": 0.4743534326553345,
+      "kl": 0.0007699355981977922,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 1.5777490362524986,
-      "reward_std": 1.2104695066809654,
-      "rewards/concensus_correctness_reward_func": 0.06606250116601586,
-      "rewards/consensus_reward_func": 0.5,
+      "reward": 0.9784166387149266,
+      "reward_std": 0.6934289123330798,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.07142857142857142,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0625,
-      "rewards/question_recreation_reward_func": 0.6006240528076887,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.6064166405371257,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.3329374957829714,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.3005714310067041,
       "step": 16
     },
     {
-      "completion_length": 504.71875,
-      "epoch": 3.6,
-      "grad_norm": 0.5127859115600586,
-      "kl": 0.0007266185748449061,
+      "completion_length": 391.5,
+      "epoch": 6.0,
+      "grad_norm": 0.41280442476272583,
+      "kl": 0.0007403797353617847,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 2.157407224178314,
-      "reward_std": 1.291458360850811,
-      "rewards/concensus_correctness_reward_func": 0.059812501072883606,
-      "rewards/consensus_reward_func": 0.8125,
+      "reward": 1.3840059041976929,
+      "reward_std": 0.7487279496022633,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.2857142857142857,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.1875,
-      "rewards/question_recreation_reward_func": 0.6287197517231107,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5329701921769551,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.03125,
-      "rewards/xmlcount_reward_func": 0.4376250107306987,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.5474643004792077,
       "step": 18
     },
     {
-      "completion_length": 508.5625,
-      "epoch": 4.0,
-      "grad_norm": 0.4247759282588959,
-      "kl": 0.0006629491217609029,
+      "completion_length": 413.71875,
+      "epoch": 6.7272727272727275,
+      "grad_norm": 0.4090598523616791,
+      "kl": 0.000847268798679579,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 1.6457609236240387,
-      "reward_std": 1.067463479936123,
-      "rewards/concensus_correctness_reward_func": 0.04062499850988388,
-      "rewards/consensus_reward_func": 0.375,
+      "reward": 0.9405753538012505,
+      "reward_std": 0.917273748666048,
+      "rewards/concensus_correctness_reward_func": 0.015375000424683094,
+      "rewards/consensus_reward_func": 0.125,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.125,
-      "rewards/question_recreation_reward_func": 0.6141359470784664,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.3791066133417189,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.03125,
-      "rewards/xmlcount_reward_func": 0.4597499929368496,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.421093737706542,
       "step": 20
     },
     {
-      "epoch": 4.0,
+      "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 7.440462638896861e-07,
-      "train_runtime": 1056.8449,
-      "train_samples_per_second": 0.303,
-      "train_steps_per_second": 0.019
+      "train_loss": 6.568556671027181e-07,
+      "train_runtime": 565.9009,
+      "train_samples_per_second": 0.565,
+      "train_steps_per_second": 0.035
     }
   ],
   "logging_steps": 2,
   "max_steps": 20,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 4,
+  "num_train_epochs": 10,
   "save_steps": 25,
   "stateful_callbacks": {
     "TrainerControl": {