End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f77adb9ccddf1e5b1ea5456eb0795bf809475e84f293f157c70def872159d88d
+oid sha256:6a33fe903d76ef2cab0c995b96392c4dec27c685433144c7a7e36986957ec34c
 size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 1.1607659075707489e-06,
-    "train_runtime": 1310.2582,
+    "train_loss": 9.804715489281079e-07,
+    "train_runtime": 1429.7914,
     "train_samples": 11,
-    "train_samples_per_second": 0.244,
-    "train_steps_per_second": 0.015
+    "train_samples_per_second": 0.224,
+    "train_steps_per_second": 0.014
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 1.1607659075707489e-06,
-    "train_runtime": 1310.2582,
+    "train_loss": 9.804715489281079e-07,
+    "train_runtime": 1429.7914,
     "train_samples": 11,
-    "train_samples_per_second": 0.244,
-    "train_steps_per_second": 0.015
+    "train_samples_per_second": 0.224,
+    "train_steps_per_second": 0.014
 }

trainer_state.json

@@ -10,203 +10,203 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 506.15625,
+      "completion_length": 546.25,
       "epoch": 0.7272727272727273,
-      "grad_norm": 0.339279443025589,
-      "kl": 0.0009709163859952241,
+      "grad_norm": 0.3629768490791321,
+      "kl": 0.0008834904219838791,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 1.0586679205298424,
-      "reward_std": 0.9162167347967625,
+      "reward": 0.7893820339813828,
+      "reward_std": 0.9789414554834366,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.3125,
+      "rewards/consensus_reward_func": 0.0625,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.42026167176663876,
+      "rewards/final_correctness_reward_func": 0.125,
+      "rewards/question_recreation_reward_func": 0.477319560945034,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.31028125807642937,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.12456250749528408,
       "step": 2
     },
     {
-      "completion_length": 541.4642857142857,
+      "completion_length": 497.75,
       "epoch": 1.3636363636363638,
-      "grad_norm": 0.3777475357055664,
-      "kl": 0.0009490517882763275,
+      "grad_norm": 0.2759566009044647,
+      "kl": 0.0008876241294534079,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 1.024622346673693,
-      "reward_std": 1.3441601055009025,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.35714285714285715,
+      "reward": 1.334993405001504,
+      "reward_std": 1.241017405475889,
+      "rewards/concensus_correctness_reward_func": 0.11914285591670445,
+      "rewards/consensus_reward_func": 0.5,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.6680866394724164,
+      "rewards/final_correctness_reward_func": 0.07142857142857142,
+      "rewards/question_recreation_reward_func": 0.7142433992453984,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.0006071478128433228,
+      "rewards/xmlcount_reward_func": -0.06982142903975078,
       "step": 4
     },
     {
-      "completion_length": 394.0357142857143,
+      "completion_length": 531.1428571428571,
       "epoch": 2.0,
-      "grad_norm": 9.476052284240723,
-      "kl": 0.0027206389433039086,
+      "grad_norm": 0.4373629093170166,
+      "kl": 0.000981977126295013,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 0.8097397450889859,
-      "reward_std": 1.0565592816897802,
+      "reward": 0.42511291801929474,
+      "reward_std": 0.9857849138123649,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.42857142857142855,
+      "rewards/consensus_reward_func": 0.14285714285714285,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.46263259862150463,
+      "rewards/question_recreation_reward_func": 0.5162914799792426,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.08146428423268455,
+      "rewards/xmlcount_reward_func": -0.23403571120330266,
       "step": 6
     },
     {
-      "completion_length": 504.125,
+      "completion_length": 551.4375,
       "epoch": 2.7272727272727275,
-      "grad_norm": 2.0453450679779053,
-      "kl": 0.0009930663873092271,
+      "grad_norm": 0.3953864872455597,
+      "kl": 0.0010158523655263707,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 1.2847755327820778,
-      "reward_std": 0.8506403658539057,
-      "rewards/concensus_correctness_reward_func": 0.023874999955296516,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 0.48503089137375355,
+      "reward_std": 1.0153761729598045,
+      "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": 0.6057443059980869,
+      "rewards/final_correctness_reward_func": 0.0625,
+      "rewards/question_recreation_reward_func": 0.5571871381253004,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.3895312398672104,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": -0.19715623185038567,
       "step": 8
     },
     {
-      "completion_length": 510.7142857142857,
+      "completion_length": 450.9642857142857,
       "epoch": 3.3636363636363638,
-      "grad_norm": 0.5151574015617371,
-      "kl": 0.0010096280685892062,
+      "grad_norm": 0.48153209686279297,
+      "kl": 0.001320448945209916,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 0.8320176963295255,
-      "reward_std": 1.1409024851662772,
-      "rewards/concensus_correctness_reward_func": 0.02728571423462459,
-      "rewards/consensus_reward_func": 0.21428571428571427,
+      "reward": 1.4874113542693002,
+      "reward_std": 0.8693422717707497,
+      "rewards/concensus_correctness_reward_func": 0.15057142823934555,
+      "rewards/consensus_reward_func": 0.42857142857142855,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.5239819682070187,
+      "rewards/question_recreation_reward_func": 0.5925541860716683,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.05357142857142857,
-      "rewards/xmlcount_reward_func": 0.012892857726131166,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": 0.29785714617797304,
       "step": 10
     },
     {
-      "completion_length": 543.9285714285714,
+      "completion_length": 525.8928571428571,
       "epoch": 4.0,
-      "grad_norm": 0.3699837327003479,
-      "kl": 0.0009374998376837798,
+      "grad_norm": 0.38143840432167053,
+      "kl": 0.001059339556377381,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 1.5399821060044425,
-      "reward_std": 1.3528829557555062,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.42857142857142855,
+      "reward": 1.0610307965959822,
+      "reward_std": 1.0929101343665804,
+      "rewards/concensus_correctness_reward_func": 0.05957142795835223,
+      "rewards/consensus_reward_func": 0.14285714285714285,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.07142857142857142,
-      "rewards/question_recreation_reward_func": 0.451982102223805,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.4176021920783179,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.05357142857142857,
-      "rewards/xmlcount_reward_func": 0.534428562436785,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.44100002305848257,
       "step": 12
     },
     {
-      "completion_length": 442.84375,
+      "completion_length": 559.90625,
       "epoch": 4.7272727272727275,
-      "grad_norm": 0.5262240171432495,
-      "kl": 0.0011314730218145996,
+      "grad_norm": 0.4833117127418518,
+      "kl": 0.0010816850699484348,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 1.073042456060648,
-      "reward_std": 0.958334431052208,
+      "reward": 0.8927601464092731,
+      "reward_std": 0.8923898264765739,
       "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": 0.6101674325764179,
+      "rewards/final_correctness_reward_func": 0.0625,
+      "rewards/question_recreation_reward_func": 0.5541351325809956,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.2753749964758754,
+      "rewards/xmlcount_reward_func": 0.08862500078976154,
       "step": 14
     },
     {
-      "completion_length": 503.64285714285717,
+      "completion_length": 512.6785714285714,
       "epoch": 5.363636363636363,
-      "grad_norm": 0.4339660108089447,
-      "kl": 0.0009521031835382539,
+      "grad_norm": 0.43158963322639465,
+      "kl": 0.001266635738180152,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 1.0212854487555367,
-      "reward_std": 0.5305720716714859,
+      "reward": 0.5066124935235296,
+      "reward_std": 0.7872398836272103,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.2857142857142857,
+      "rewards/consensus_reward_func": 0.07142857142857142,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.47414258122444153,
+      "rewards/question_recreation_reward_func": 0.4602196104824543,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.26142857330186026,
+      "rewards/strict_format_reward_func": 0.017857142857142856,
+      "rewards/xmlcount_reward_func": -0.04289285678948675,
       "step": 16
     },
     {
-      "completion_length": 487.5357142857143,
+      "completion_length": 494.85714285714283,
       "epoch": 6.0,
-      "grad_norm": 0.28034958243370056,
-      "kl": 0.0009002718475780316,
+      "grad_norm": 0.3546796143054962,
+      "kl": 0.0010320868459530175,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 1.2305413314274378,
-      "reward_std": 1.281937335218702,
-      "rewards/concensus_correctness_reward_func": 0.02728571423462459,
-      "rewards/consensus_reward_func": 0.35714285714285715,
+      "reward": 1.2014339140483312,
+      "reward_std": 1.0514784412724631,
+      "rewards/concensus_correctness_reward_func": 0.05957142795835223,
+      "rewards/consensus_reward_func": 0.2857142857142857,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.6880412825516292,
+      "rewards/question_recreation_reward_func": 0.5871481852872031,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.03571428571428571,
-      "rewards/xmlcount_reward_func": 0.122357149209295,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.2690000044448035,
       "step": 18
     },
     {
-      "completion_length": 467.875,
+      "completion_length": 581.9375,
       "epoch": 6.7272727272727275,
-      "grad_norm": 5.054487228393555,
-      "kl": 0.0019134482718072832,
+      "grad_norm": 0.35805544257164,
+      "kl": 0.0009703972173156217,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 1.316497080028057,
-      "reward_std": 1.0599600225687027,
-      "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.375,
+      "reward": 0.6761963879689574,
+      "reward_std": 1.0061921868473291,
+      "rewards/concensus_correctness_reward_func": 0.012625000439584255,
+      "rewards/consensus_reward_func": 0.0625,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.39609083347022533,
+      "rewards/question_recreation_reward_func": 0.6275088861584663,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.5297812481876463,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": -0.05768749490380287,
       "step": 20
     },
     {
       "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 1.1607659075707489e-06,
-      "train_runtime": 1310.2582,
-      "train_samples_per_second": 0.244,
-      "train_steps_per_second": 0.015
+      "train_loss": 9.804715489281079e-07,
+      "train_runtime": 1429.7914,
+      "train_samples_per_second": 0.224,
+      "train_steps_per_second": 0.014
     }
   ],
   "logging_steps": 2,