End of training

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d5a5db1ca33542874d8002e36eb5419408164060f93f435109f44e4e741a65be
+oid sha256:f7c8bfaa87f7466064ee1d3ff39f365bfe09318d58214f24e6e37e5d826ade5d
 size 35237104

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 2.6350203826908115e-06,
-    "train_runtime": 2184.1674,
-    "train_samples": 22,
-    "train_samples_per_second": 0.147,
+    "train_loss": 2.0231228631928387e-06,
+    "train_runtime": 2157.233,
+    "train_samples": 11,
+    "train_samples_per_second": 0.148,
     "train_steps_per_second": 0.009
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 2.6350203826908115e-06,
-    "train_runtime": 2184.1674,
-    "train_samples": 22,
-    "train_samples_per_second": 0.147,
+    "train_loss": 2.0231228631928387e-06,
+    "train_runtime": 2157.233,
+    "train_samples": 11,
+    "train_samples_per_second": 0.148,
     "train_steps_per_second": 0.009
 }

trainer_state.json

@@ -2,7 +2,7 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 3.3636363636363638,
+  "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": 459.4375,
-      "epoch": 0.36363636363636365,
-      "grad_norm": 1.3302592039108276,
-      "kl": 0.0,
+      "completion_length": 438.84375,
+      "epoch": 0.7272727272727273,
+      "grad_norm": 19.830360412597656,
+      "kl": 0.003334177119540982,
       "learning_rate": 4.965903258506806e-07,
       "loss": 0.0,
-      "reward": 0.04405219480395317,
-      "reward_std": 0.6433118600398302,
+      "reward": 0.17543404083698988,
+      "reward_std": 0.22624898492358625,
       "rewards/concensus_correctness_reward_func": 0.0,
       "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.2288334424374625,
+      "rewards/question_recreation_reward_func": 0.1097465381026268,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.18478125147521496,
+      "rewards/xmlcount_reward_func": 0.06568749900907278,
       "step": 2
     },
     {
-      "completion_length": 409.90625,
-      "epoch": 0.7272727272727273,
-      "grad_norm": 1.2176474332809448,
-      "kl": 0.0020845978724537417,
+      "completion_length": 285.39285714285717,
+      "epoch": 1.3636363636363638,
+      "grad_norm": 2.6216137409210205,
+      "kl": 0.0015792557637074164,
       "learning_rate": 4.698684378016222e-07,
       "loss": 0.0,
-      "reward": 0.7678184271790087,
-      "reward_std": 0.7591042327694595,
-      "rewards/concensus_correctness_reward_func": 0.013437500223517418,
-      "rewards/consensus_reward_func": 0.375,
+      "reward": 0.21063038440687315,
+      "reward_std": 0.2409371439633625,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.26688094018027186,
+      "rewards/question_recreation_reward_func": 0.12295180878468923,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.11249999701976776,
+      "rewards/xmlcount_reward_func": 0.08767857375953879,
       "step": 4
     },
     {
-      "completion_length": 367.9583333333333,
-      "epoch": 1.0,
-      "grad_norm": 1.5351356267929077,
-      "kl": 0.004026121634524316,
+      "completion_length": 294.5,
+      "epoch": 2.0,
+      "grad_norm": 9.230982780456543,
+      "kl": 0.0015167358298120753,
       "learning_rate": 4.193203929064353e-07,
       "loss": 0.0,
-      "reward": 0.5474015871683756,
-      "reward_std": 0.8145603885253271,
+      "reward": 0.10711567582828659,
+      "reward_std": 0.08755659498274326,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.08333333333333333,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.3452765668431918,
+      "rewards/question_recreation_reward_func": 0.07575853328619685,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.11879166960716248,
+      "rewards/xmlcount_reward_func": 0.031357143606458394,
       "step": 6
     },
     {
-      "completion_length": 340.90625,
-      "epoch": 1.3636363636363638,
-      "grad_norm": 2.2870802879333496,
-      "kl": 0.0028898825694341213,
+      "completion_length": 347.28125,
+      "epoch": 2.7272727272727275,
+      "grad_norm": 4.538873672485352,
+      "kl": 0.0013625815699924715,
       "learning_rate": 3.5042385616324236e-07,
       "loss": 0.0,
-      "reward": 0.7129990463145077,
-      "reward_std": 1.1731511121615767,
-      "rewards/concensus_correctness_reward_func": 0.013437500223517418,
-      "rewards/consensus_reward_func": 0.375,
+      "reward": 0.17055620742030442,
+      "reward_std": 0.23858435079455376,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0625,
-      "rewards/question_recreation_reward_func": 0.25137403002008796,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.10461870580911636,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.010687506292015314,
+      "rewards/xmlcount_reward_func": 0.06593749951571226,
       "step": 8
     },
     {
-      "completion_length": 415.125,
-      "epoch": 1.7272727272727273,
-      "grad_norm": 5.562981605529785,
-      "kl": 0.0016170966700883582,
+      "completion_length": 363.2142857142857,
+      "epoch": 3.3636363636363638,
+      "grad_norm": 6.419910430908203,
+      "kl": 0.0016153148608282208,
       "learning_rate": 2.706448363680831e-07,
       "loss": 0.0,
-      "reward": 0.295339772477746,
-      "reward_std": 0.6460105914156884,
+      "reward": 0.07476867362856865,
+      "reward_std": 0.2939557205619557,
       "rewards/concensus_correctness_reward_func": 0.0,
       "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.27162102051079273,
+      "rewards/question_recreation_reward_func": 0.10076867416501045,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.023718755692243576,
+      "rewards/xmlcount_reward_func": -0.025999996811151505,
       "step": 10
     },
     {
-      "completion_length": 523.875,
-      "epoch": 2.0,
-      "grad_norm": 1.5765122175216675,
-      "kl": 0.0012822053298198928,
+      "completion_length": 288.64285714285717,
+      "epoch": 4.0,
+      "grad_norm": 42.38132858276367,
+      "kl": 0.005553228639265788,
       "learning_rate": 1.886286282148002e-07,
       "loss": 0.0,
-      "reward": 0.5160771167526642,
-      "reward_std": 0.8615054544061422,
+      "reward": 0.2218315941946847,
+      "reward_std": 0.3740644904651812,
       "rewards/concensus_correctness_reward_func": 0.0,
-      "rewards/consensus_reward_func": 0.3333333333333333,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.1667854206946989,
+      "rewards/question_recreation_reward_func": 0.16290302335151605,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.01595833556105693,
+      "rewards/xmlcount_reward_func": 0.058928570577076504,
       "step": 12
     },
     {
-      "completion_length": 423.1875,
-      "epoch": 2.3636363636363638,
-      "grad_norm": 3.4827518463134766,
-      "kl": 0.004907177797576878,
+      "completion_length": 311.84375,
+      "epoch": 4.7272727272727275,
+      "grad_norm": 20.04547882080078,
+      "kl": 0.001355127285933122,
       "learning_rate": 1.1326296046939333e-07,
       "loss": 0.0,
-      "reward": 0.1884107068181038,
-      "reward_std": 0.3662752130185254,
+      "reward": 0.18036225740797818,
+      "reward_std": 0.21330193988978863,
       "rewards/concensus_correctness_reward_func": 0.0,
       "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.12262945214752108,
-      "rewards/soft_format_reward_func": 0.015625,
+      "rewards/question_recreation_reward_func": 0.12386225792579353,
+      "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.050156254321336746,
+      "rewards/xmlcount_reward_func": 0.05649999948218465,
       "step": 14
     },
     {
-      "completion_length": 346.875,
-      "epoch": 2.7272727272727275,
-      "grad_norm": 4.646650314331055,
-      "kl": 0.006251230763155036,
+      "completion_length": 342.64285714285717,
+      "epoch": 5.363636363636363,
+      "grad_norm": 3.181459903717041,
+      "kl": 0.0011706448027065822,
       "learning_rate": 5.271487265090163e-08,
       "loss": 0.0,
-      "reward": 0.8924960913136601,
-      "reward_std": 1.0580921154469252,
-      "rewards/concensus_correctness_reward_func": 0.05974999815225601,
-      "rewards/consensus_reward_func": 0.5,
+      "reward": -0.033251527696847916,
+      "reward_std": 0.32666803231196745,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.264589830301702,
+      "rewards/question_recreation_reward_func": 0.06167703467820372,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.0681562414392829,
+      "rewards/xmlcount_reward_func": -0.09492856530206543,
       "step": 16
     },
     {
-      "completion_length": 447.875,
-      "epoch": 3.0,
-      "grad_norm": 2.753358840942383,
-      "kl": 0.0035982539411634207,
+      "completion_length": 322.85714285714283,
+      "epoch": 6.0,
+      "grad_norm": 1.6145594120025635,
+      "kl": 0.0013051111184592759,
       "learning_rate": 1.3545689574841341e-08,
       "loss": 0.0,
-      "reward": 0.31809323467314243,
-      "reward_std": 0.6482827632377545,
+      "reward": 0.1403887660375663,
+      "reward_std": 0.1475449800491333,
       "rewards/concensus_correctness_reward_func": 0.0,
       "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.08333333333333333,
-      "rewards/question_recreation_reward_func": 0.2697598912442724,
-      "rewards/soft_format_reward_func": 0.020833333333333332,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.08678162470459938,
+      "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": -0.055833334724108376,
+      "rewards/xmlcount_reward_func": 0.05360714292951992,
       "step": 18
     },
     {
-      "completion_length": 371.9375,
-      "epoch": 3.3636363636363638,
-      "grad_norm": 37.19569778442383,
-      "kl": 0.0018885131285060197,
+      "completion_length": 259.34375,
+      "epoch": 6.7272727272727275,
+      "grad_norm": 5.390585899353027,
+      "kl": 0.0029644069654750638,
       "learning_rate": 0.0,
       "loss": 0.0,
-      "reward": 0.9909486873075366,
-      "reward_std": 1.7817546646110713,
-      "rewards/concensus_correctness_reward_func": 0.625,
-      "rewards/consensus_reward_func": 0.0625,
+      "reward": 0.052817828953266144,
+      "reward_std": 0.3324423659360036,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.0,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0625,
-      "rewards/question_recreation_reward_func": 0.15935491654090583,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.11581782391294837,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.08159375563263893,
+      "rewards/xmlcount_reward_func": -0.06300000031478703,
       "step": 20
     },
     {
-      "epoch": 3.3636363636363638,
+      "epoch": 6.7272727272727275,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 2.6350203826908115e-06,
-      "train_runtime": 2184.1674,
-      "train_samples_per_second": 0.147,
+      "train_loss": 2.0231228631928387e-06,
+      "train_runtime": 2157.233,
+      "train_samples_per_second": 0.148,
       "train_steps_per_second": 0.009
     }
   ],
   "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": {