End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fe7bf06a6ca4c04ae3b5008fc18ae51b9a54194ab47f3d27299cd1228d565d7d
+oid sha256:e94029ca89d6cd8c93031968bdff730e3617a20881d7731054248002ad6615ee
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 7.303286452042812e-07,
-    "train_runtime": 670.9955,
-    "train_samples": 13,
-    "train_samples_per_second": 0.477,
-    "train_steps_per_second": 0.03
+    "train_loss": 5.868451239621208e-07,
+    "train_runtime": 647.9125,
+    "train_samples": 11,
+    "train_samples_per_second": 0.494,
+    "train_steps_per_second": 0.031
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 7.303286452042812e-07,
-    "train_runtime": 670.9955,
-    "train_samples": 13,
-    "train_samples_per_second": 0.477,
-    "train_steps_per_second": 0.03
+    "train_loss": 5.868451239621208e-07,
+    "train_runtime": 647.9125,
+    "train_samples": 11,
+    "train_samples_per_second": 0.494,
+    "train_steps_per_second": 0.031
 }

trainer_state.json

@@ -2,7 +2,7 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 5.0,
+  "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": 380.53125,
-      "epoch": 0.6153846153846154,
-      "grad_norm": 0.4489820599555969,
-      "kl": 0.0008734324583201669,
+      "completion_length": 431.5,
+      "epoch": 0.7272727272727273,
+      "grad_norm": 0.5174108147621155,
+      "kl": 0.000712571847543586,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 1.6363564729690552,
-      "reward_std": 1.459476925432682,
-      "rewards/concensus_correctness_reward_func": 0.06562500074505806,
-      "rewards/consensus_reward_func": 0.4375,
+      "reward": 1.0435986891388893,
+      "reward_std": 1.0856736898422241,
+      "rewards/concensus_correctness_reward_func": 0.020874999463558197,
+      "rewards/consensus_reward_func": 0.1875,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.25,
-      "rewards/question_recreation_reward_func": 0.48848147690296173,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.46972369588911533,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.046875,
-      "rewards/xmlcount_reward_func": 0.34787500463426113,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": 0.3342499975115061,
       "step": 2
     },
     {
-      "completion_length": 282.05,
-      "epoch": 1.0,
-      "grad_norm": 0.2521287798881531,
-      "kl": 0.0008821029914543033,
+      "completion_length": 429.32142857142856,
+      "epoch": 1.3636363636363638,
+      "grad_norm": 0.40975213050842285,
+      "kl": 0.0007477319034348641,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 2.128608989715576,
-      "reward_std": 1.7205048680305481,
-      "rewards/concensus_correctness_reward_func": 0.17389999628067015,
-      "rewards/consensus_reward_func": 0.6,
+      "reward": 0.7117581112044198,
+      "reward_std": 0.8257838019302913,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.4,
-      "rewards/question_recreation_reward_func": 0.5337589919567108,
-      "rewards/soft_format_reward_func": 0.025,
-      "rewards/strict_format_reward_func": 0.05,
-      "rewards/xmlcount_reward_func": 0.34595000743865967,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5281509671892438,
+      "rewards/soft_format_reward_func": 0.0,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.1836071557232312,
       "step": 4
     },
     {
-      "completion_length": 315.0,
-      "epoch": 1.6153846153846154,
-      "grad_norm": 5.095241069793701,
-      "kl": 0.0009411159189767204,
+      "completion_length": 614.4642857142857,
+      "epoch": 2.0,
+      "grad_norm": 0.2254764884710312,
+      "kl": 0.000554072969992246,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 1.6255136616528034,
-      "reward_std": 1.489594727754593,
-      "rewards/concensus_correctness_reward_func": 0.01718750037252903,
-      "rewards/consensus_reward_func": 0.5,
+      "reward": 0.917183348110744,
+      "reward_std": 1.2978510005133492,
+      "rewards/concensus_correctness_reward_func": 0.02385714224406651,
+      "rewards/consensus_reward_func": 0.42857142857142855,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.375,
-      "rewards/question_recreation_reward_func": 0.4625761993229389,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.6112905059541974,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.27074999921023846,
+      "rewards/xmlcount_reward_func": -0.14653571854744638,
       "step": 6
     },
     {
-      "completion_length": 341.7,
-      "epoch": 2.0,
-      "grad_norm": 0.2078591287136078,
-      "kl": 0.0007837335928343236,
+      "completion_length": 477.46875,
+      "epoch": 2.7272727272727275,
+      "grad_norm": 0.357846200466156,
+      "kl": 0.0006728375228703953,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 2.0344817638397217,
-      "reward_std": 1.0058663219213486,
-      "rewards/concensus_correctness_reward_func": 0.1,
-      "rewards/consensus_reward_func": 0.5,
+      "reward": 1.1117506809532642,
+      "reward_std": 1.0734700448811054,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.3125,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.4,
-      "rewards/question_recreation_reward_func": 0.5607817590236663,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.4936569482088089,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.4736999936401844,
+      "rewards/xmlcount_reward_func": 0.3055937457829714,
       "step": 8
     },
     {
-      "completion_length": 361.0,
-      "epoch": 2.6153846153846154,
-      "grad_norm": 0.6822924017906189,
-      "kl": 0.0008542858995497227,
+      "completion_length": 573.6071428571429,
+      "epoch": 3.3636363636363638,
+      "grad_norm": 0.29670479893684387,
+      "kl": 0.0005498271576860654,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 1.6393529465422034,
-      "reward_std": 1.3287841528654099,
-      "rewards/concensus_correctness_reward_func": 0.12550000101327896,
-      "rewards/consensus_reward_func": 0.5,
+      "reward": 0.70088278395789,
+      "reward_std": 1.0947566969054086,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.1875,
-      "rewards/question_recreation_reward_func": 0.5319154784083366,
+      "rewards/final_correctness_reward_func": 0.07142857142857142,
+      "rewards/question_recreation_reward_func": 0.5665256423609597,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.27881250716745853,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.04507142411811011,
       "step": 10
     },
     {
-      "completion_length": 381.4,
-      "epoch": 3.0,
-      "grad_norm": 0.1449594348669052,
-      "kl": 0.0007726986717898398,
+      "completion_length": 496.82142857142856,
+      "epoch": 4.0,
+      "grad_norm": 0.8005714416503906,
+      "kl": 0.0005984455263907355,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 2.0257710933685305,
-      "reward_std": 1.5671683251857758,
-      "rewards/concensus_correctness_reward_func": 0.17389999628067015,
-      "rewards/consensus_reward_func": 0.6,
+      "reward": 0.8457667657307216,
+      "reward_std": 0.7227382106440408,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.7,
-      "rewards/question_recreation_reward_func": 0.5558210909366608,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5543382252965655,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.003949987888336182,
+      "rewards/xmlcount_reward_func": 0.29142857236521585,
       "step": 12
     },
     {
-      "completion_length": 338.15625,
-      "epoch": 3.6153846153846154,
-      "grad_norm": 0.5234334468841553,
-      "kl": 0.0008029253149288706,
+      "completion_length": 489.84375,
+      "epoch": 4.7272727272727275,
+      "grad_norm": 0.4241688549518585,
+      "kl": 0.0006718316217302345,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 2.26827435195446,
-      "reward_std": 0.9868108257651329,
-      "rewards/concensus_correctness_reward_func": 0.10868749767541885,
-      "rewards/consensus_reward_func": 0.4375,
+      "reward": 0.789498545229435,
+      "reward_std": 0.7724240645766258,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.0625,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.5625,
-      "rewards/question_recreation_reward_func": 0.6076806150376797,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.6531860418617725,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.5519062541425228,
+      "rewards/xmlcount_reward_func": 0.07381249964237213,
       "step": 14
     },
     {
-      "completion_length": 317.9,
-      "epoch": 4.0,
-      "grad_norm": 0.22867438197135925,
-      "kl": 0.0009389304323121905,
+      "completion_length": 573.25,
+      "epoch": 5.363636363636363,
+      "grad_norm": 0.24242204427719116,
+      "kl": 0.0005594768694468907,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 1.5226385742425919,
-      "reward_std": 1.0005009174346924,
-      "rewards/concensus_correctness_reward_func": 0.05,
-      "rewards/consensus_reward_func": 0.5,
+      "reward": 0.7660920662539346,
+      "reward_std": 0.8827205130032131,
+      "rewards/concensus_correctness_reward_func": 0.01957142778805324,
+      "rewards/consensus_reward_func": 0.07142857142857142,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5472385793924331,
+      "rewards/question_recreation_reward_func": 0.6511277769293103,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.42539999485015867,
+      "rewards/xmlcount_reward_func": 0.023964287979262217,
       "step": 16
     },
     {
-      "completion_length": 297.5,
-      "epoch": 4.615384615384615,
-      "grad_norm": 0.5817599296569824,
-      "kl": 0.0009179217449855059,
+      "completion_length": 505.75,
+      "epoch": 6.0,
+      "grad_norm": 0.33538007736206055,
+      "kl": 0.0007109953190333077,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 2.037886768579483,
-      "reward_std": 1.2188720293343067,
-      "rewards/concensus_correctness_reward_func": 0.06562500074505806,
-      "rewards/consensus_reward_func": 0.6875,
+      "reward": 0.9219183283192771,
+      "reward_std": 0.9126663931778499,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.21428571428571427,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.1875,
-      "rewards/question_recreation_reward_func": 0.559730526059866,
-      "rewards/soft_format_reward_func": 0.0,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5539540350437164,
+      "rewards/soft_format_reward_func": 0.017857142857142856,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.5375312566757202,
+      "rewards/xmlcount_reward_func": 0.13582143081086023,
       "step": 18
     },
     {
-      "completion_length": 356.75,
-      "epoch": 5.0,
-      "grad_norm": 0.4133365750312805,
-      "kl": 0.0011233275523409247,
+      "completion_length": 491.5,
+      "epoch": 6.7272727272727275,
+      "grad_norm": 0.4001785218715668,
+      "kl": 0.000623058349447092,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 1.7879921913146972,
-      "reward_std": 1.4964043855667115,
+      "reward": 0.7281233966350555,
+      "reward_std": 0.9216739386320114,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.6,
+      "rewards/consensus_reward_func": 0.125,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.5,
-      "rewards/question_recreation_reward_func": 0.5599421709775925,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5827796421945095,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.025,
-      "rewards/xmlcount_reward_func": 0.1030500203371048,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.020343752577900887,
       "step": 20
     },
     {
-      "epoch": 5.0,
+      "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 7.303286452042812e-07,
-      "train_runtime": 670.9955,
-      "train_samples_per_second": 0.477,
-      "train_steps_per_second": 0.03
+      "train_loss": 5.868451239621208e-07,
+      "train_runtime": 647.9125,
+      "train_samples_per_second": 0.494,
+      "train_steps_per_second": 0.031
     }
   ],
   "logging_steps": 2,
   "max_steps": 20,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 7,
+  "num_train_epochs": 10,
   "save_steps": 25,
   "stateful_callbacks": {
     "TrainerControl": {