7231 lines
257 KiB
JSON
7231 lines
257 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.32,
|
|
"eval_steps": 500,
|
|
"global_step": 200,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"calib/answer_extract_rate": 0.390625,
|
|
"calib/avg_num_step_conf": 2.19921875,
|
|
"calib/ece": 0.6720437956204381,
|
|
"calib/final_conf_rate": 0.53515625,
|
|
"calib/format_rate": 0.30859375,
|
|
"calib/frac_conf_gt_0.9": 0.4744525547445255,
|
|
"calib/gap": 0.09598290598290582,
|
|
"calib/mean_conf": 0.8180291970802919,
|
|
"calib/mu_c": 0.9,
|
|
"calib/mu_w": 0.8040170940170942,
|
|
"calib/nonempty_final_conf_rate": 0.53515625,
|
|
"calib/nonempty_reasoning_rate": 0.51171875,
|
|
"calib/nonempty_step_conf_rate": 0.4375,
|
|
"calib/pce": 0.6720437956204381,
|
|
"calib/std_conf": 0.25329605140720096,
|
|
"calib/step_conf_rate": 0.4375,
|
|
"calib/step_q_c": 0.7754022988505747,
|
|
"calib/step_q_c_n": 87.0,
|
|
"calib/step_q_gap": -0.031173331401526228,
|
|
"calib/step_q_w": 0.8065756302521009,
|
|
"calib/step_q_w_n": 476.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 2501.0,
|
|
"completions/max_terminated_length": 2501.0,
|
|
"completions/mean_length": 332.60546875,
|
|
"completions/mean_terminated_length": 333.9098205566406,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.0016,
|
|
"grad_norm": 47.9720344543457,
|
|
"learning_rate": 0.0,
|
|
"loss": -0.133,
|
|
"num_tokens": 338475.0,
|
|
"reward": 0.232421875,
|
|
"reward_std": 0.3509902358055115,
|
|
"rewards/accuracy_reward_step": 0.078125,
|
|
"rewards/format_reward_step": 0.30859375,
|
|
"step": 1
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.40625,
|
|
"calib/avg_num_step_conf": 1.8828125,
|
|
"calib/ece": 0.657346153846154,
|
|
"calib/final_conf_rate": 0.5078125,
|
|
"calib/format_rate": 0.328125,
|
|
"calib/frac_conf_gt_0.9": 0.5538461538461539,
|
|
"calib/gap": 0.09180690399136981,
|
|
"calib/mean_conf": 0.8650384615384616,
|
|
"calib/mu_c": 0.9377777777777778,
|
|
"calib/mu_w": 0.845970873786408,
|
|
"calib/nonempty_final_conf_rate": 0.5078125,
|
|
"calib/nonempty_reasoning_rate": 0.4921875,
|
|
"calib/nonempty_step_conf_rate": 0.421875,
|
|
"calib/pce": 0.657346153846154,
|
|
"calib/std_conf": 0.1902169774775902,
|
|
"calib/step_conf_rate": 0.421875,
|
|
"calib/step_q_c": 0.8529457364341085,
|
|
"calib/step_q_c_n": 129.0,
|
|
"calib/step_q_gap": 0.04926301688736623,
|
|
"calib/step_q_w": 0.8036827195467423,
|
|
"calib/step_q_w_n": 353.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 2868.0,
|
|
"completions/max_terminated_length": 2868.0,
|
|
"completions/mean_length": 369.18359375,
|
|
"completions/mean_terminated_length": 370.63140869140625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0032,
|
|
"grad_norm": 17.44351577758789,
|
|
"learning_rate": 5e-08,
|
|
"loss": -0.1458,
|
|
"num_tokens": 693922.0,
|
|
"reward": 0.26953125,
|
|
"reward_std": 0.4115494191646576,
|
|
"rewards/accuracy_reward_step": 0.10546875,
|
|
"rewards/format_reward_step": 0.328125,
|
|
"step": 2
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.4453125,
|
|
"calib/avg_num_step_conf": 2.08984375,
|
|
"calib/ece": 0.6778014184397165,
|
|
"calib/final_conf_rate": 0.55078125,
|
|
"calib/format_rate": 0.32421875,
|
|
"calib/frac_conf_gt_0.9": 0.45390070921985815,
|
|
"calib/gap": 0.043058216654384696,
|
|
"calib/mean_conf": 0.8409219858156027,
|
|
"calib/mu_c": 0.8769565217391305,
|
|
"calib/mu_w": 0.8338983050847458,
|
|
"calib/nonempty_final_conf_rate": 0.55078125,
|
|
"calib/nonempty_reasoning_rate": 0.52734375,
|
|
"calib/nonempty_step_conf_rate": 0.4140625,
|
|
"calib/pce": 0.6778014184397165,
|
|
"calib/std_conf": 0.1980575914137701,
|
|
"calib/step_conf_rate": 0.4140625,
|
|
"calib/step_q_c": 0.7861038961038961,
|
|
"calib/step_q_c_n": 77.0,
|
|
"calib/step_q_gap": 0.025776385186865425,
|
|
"calib/step_q_w": 0.7603275109170307,
|
|
"calib/step_q_w_n": 458.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2866.0,
|
|
"completions/max_terminated_length": 2866.0,
|
|
"completions/mean_length": 308.23046875,
|
|
"completions/mean_terminated_length": 310.657470703125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0048,
|
|
"grad_norm": 39.398468017578125,
|
|
"learning_rate": 1e-07,
|
|
"loss": -0.0842,
|
|
"num_tokens": 1029013.0,
|
|
"reward": 0.251953125,
|
|
"reward_std": 0.37410488724708557,
|
|
"rewards/accuracy_reward_step": 0.08984375,
|
|
"rewards/format_reward_step": 0.32421875,
|
|
"step": 3
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.4453125,
|
|
"calib/avg_num_step_conf": 2.72265625,
|
|
"calib/ece": 0.6189041095890411,
|
|
"calib/final_conf_rate": 0.5703125,
|
|
"calib/format_rate": 0.33984375,
|
|
"calib/frac_conf_gt_0.9": 0.4178082191780822,
|
|
"calib/gap": 0.11410256410256403,
|
|
"calib/mean_conf": 0.7969863013698631,
|
|
"calib/mu_c": 0.8907692307692306,
|
|
"calib/mu_w": 0.7766666666666666,
|
|
"calib/nonempty_final_conf_rate": 0.5703125,
|
|
"calib/nonempty_reasoning_rate": 0.58984375,
|
|
"calib/nonempty_step_conf_rate": 0.5,
|
|
"calib/pce": 0.6189041095890411,
|
|
"calib/std_conf": 0.2410262586246447,
|
|
"calib/step_conf_rate": 0.5,
|
|
"calib/step_q_c": 0.8159722222222222,
|
|
"calib/step_q_c_n": 72.0,
|
|
"calib/step_q_gap": 0.04115622222222226,
|
|
"calib/step_q_w": 0.774816,
|
|
"calib/step_q_w_n": 625.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 2706.0,
|
|
"completions/max_terminated_length": 2706.0,
|
|
"completions/mean_length": 356.44140625,
|
|
"completions/mean_terminated_length": 357.8392333984375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0064,
|
|
"grad_norm": 15.860779762268066,
|
|
"learning_rate": 1.5e-07,
|
|
"loss": -0.0715,
|
|
"num_tokens": 1378878.0,
|
|
"reward": 0.271484375,
|
|
"reward_std": 0.3853558599948883,
|
|
"rewards/accuracy_reward_step": 0.1015625,
|
|
"rewards/format_reward_step": 0.33984375,
|
|
"step": 4
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.4140625,
|
|
"calib/avg_num_step_conf": 2.06640625,
|
|
"calib/ece": 0.6523021582733812,
|
|
"calib/final_conf_rate": 0.54296875,
|
|
"calib/format_rate": 0.3203125,
|
|
"calib/frac_conf_gt_0.9": 0.5107913669064749,
|
|
"calib/gap": 0.024090909090909052,
|
|
"calib/mean_conf": 0.8609352517985612,
|
|
"calib/mu_c": 0.88,
|
|
"calib/mu_w": 0.855909090909091,
|
|
"calib/nonempty_final_conf_rate": 0.54296875,
|
|
"calib/nonempty_reasoning_rate": 0.51953125,
|
|
"calib/nonempty_step_conf_rate": 0.4453125,
|
|
"calib/pce": 0.6523021582733812,
|
|
"calib/std_conf": 0.17768654028664443,
|
|
"calib/step_conf_rate": 0.4453125,
|
|
"calib/step_q_c": 0.8280733944954128,
|
|
"calib/step_q_c_n": 109.0,
|
|
"calib/step_q_gap": 0.05551969521140332,
|
|
"calib/step_q_w": 0.7725536992840095,
|
|
"calib/step_q_w_n": 419.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2552.0,
|
|
"completions/max_terminated_length": 2552.0,
|
|
"completions/mean_length": 311.7421875,
|
|
"completions/mean_terminated_length": 314.19683837890625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.008,
|
|
"grad_norm": 21.699703216552734,
|
|
"learning_rate": 2e-07,
|
|
"loss": -0.0928,
|
|
"num_tokens": 1720028.0,
|
|
"reward": 0.27734375,
|
|
"reward_std": 0.39154917001724243,
|
|
"rewards/accuracy_reward_step": 0.1171875,
|
|
"rewards/format_reward_step": 0.3203125,
|
|
"step": 5
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.42578125,
|
|
"calib/avg_num_step_conf": 2.34375,
|
|
"calib/ece": 0.689857142857143,
|
|
"calib/final_conf_rate": 0.546875,
|
|
"calib/format_rate": 0.33203125,
|
|
"calib/frac_conf_gt_0.9": 0.5857142857142857,
|
|
"calib/gap": 0.06312499999999999,
|
|
"calib/mean_conf": 0.8898571428571428,
|
|
"calib/mu_c": 0.9403571428571429,
|
|
"calib/mu_w": 0.8772321428571429,
|
|
"calib/nonempty_final_conf_rate": 0.546875,
|
|
"calib/nonempty_reasoning_rate": 0.53515625,
|
|
"calib/nonempty_step_conf_rate": 0.45703125,
|
|
"calib/pce": 0.689857142857143,
|
|
"calib/std_conf": 0.15006658386142047,
|
|
"calib/step_conf_rate": 0.45703125,
|
|
"calib/step_q_c": 0.7554385964912281,
|
|
"calib/step_q_c_n": 114.0,
|
|
"calib/step_q_gap": -0.06482889322070595,
|
|
"calib/step_q_w": 0.8202674897119341,
|
|
"calib/step_q_w_n": 486.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2843.0,
|
|
"completions/max_terminated_length": 2843.0,
|
|
"completions/mean_length": 334.62890625,
|
|
"completions/mean_terminated_length": 337.2637634277344,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0096,
|
|
"grad_norm": 60.161705017089844,
|
|
"learning_rate": 2.5e-07,
|
|
"loss": -0.1094,
|
|
"num_tokens": 2067637.0,
|
|
"reward": 0.279296875,
|
|
"reward_std": 0.4157405495643616,
|
|
"rewards/accuracy_reward_step": 0.11328125,
|
|
"rewards/format_reward_step": 0.33203125,
|
|
"step": 6
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.4609375,
|
|
"calib/avg_num_step_conf": 2.84375,
|
|
"calib/ece": 0.6351578947368423,
|
|
"calib/final_conf_rate": 0.59375,
|
|
"calib/format_rate": 0.33984375,
|
|
"calib/frac_conf_gt_0.9": 0.46710526315789475,
|
|
"calib/gap": 0.07446974140229257,
|
|
"calib/mean_conf": 0.8391052631578948,
|
|
"calib/mu_c": 0.8983870967741934,
|
|
"calib/mu_w": 0.8239173553719008,
|
|
"calib/nonempty_final_conf_rate": 0.59375,
|
|
"calib/nonempty_reasoning_rate": 0.546875,
|
|
"calib/nonempty_step_conf_rate": 0.43359375,
|
|
"calib/pce": 0.6351578947368423,
|
|
"calib/std_conf": 0.19963740677244204,
|
|
"calib/step_conf_rate": 0.43359375,
|
|
"calib/step_q_c": 0.835945945945946,
|
|
"calib/step_q_c_n": 111.0,
|
|
"calib/step_q_gap": 0.06098646458451962,
|
|
"calib/step_q_w": 0.7749594813614263,
|
|
"calib/step_q_w_n": 617.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2738.0,
|
|
"completions/max_terminated_length": 2738.0,
|
|
"completions/mean_length": 373.12109375,
|
|
"completions/mean_terminated_length": 377.54547119140625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0112,
|
|
"grad_norm": 17.492589950561523,
|
|
"learning_rate": 3e-07,
|
|
"loss": -0.0997,
|
|
"num_tokens": 2424132.0,
|
|
"reward": 0.294921875,
|
|
"reward_std": 0.3952005207538605,
|
|
"rewards/accuracy_reward_step": 0.125,
|
|
"rewards/format_reward_step": 0.33984375,
|
|
"step": 7
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.47265625,
|
|
"calib/avg_num_step_conf": 2.17578125,
|
|
"calib/ece": 0.6398734177215191,
|
|
"calib/final_conf_rate": 0.6171875,
|
|
"calib/format_rate": 0.34375,
|
|
"calib/frac_conf_gt_0.9": 0.5063291139240507,
|
|
"calib/gap": 0.09216724738675963,
|
|
"calib/mean_conf": 0.8613924050632913,
|
|
"calib/mu_c": 0.9331428571428573,
|
|
"calib/mu_w": 0.8409756097560976,
|
|
"calib/nonempty_final_conf_rate": 0.6171875,
|
|
"calib/nonempty_reasoning_rate": 0.58203125,
|
|
"calib/nonempty_step_conf_rate": 0.46875,
|
|
"calib/pce": 0.6398734177215191,
|
|
"calib/std_conf": 0.16952080024135338,
|
|
"calib/step_conf_rate": 0.46875,
|
|
"calib/step_q_c": 0.8074380165289258,
|
|
"calib/step_q_c_n": 121.0,
|
|
"calib/step_q_gap": 0.02860774129956789,
|
|
"calib/step_q_w": 0.7788302752293579,
|
|
"calib/step_q_w_n": 436.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2909.0,
|
|
"completions/max_terminated_length": 2909.0,
|
|
"completions/mean_length": 349.1953125,
|
|
"completions/mean_terminated_length": 351.94488525390625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0128,
|
|
"grad_norm": 11.481454849243164,
|
|
"learning_rate": 3.5e-07,
|
|
"loss": 0.008,
|
|
"num_tokens": 2775366.0,
|
|
"reward": 0.3125,
|
|
"reward_std": 0.4172360301017761,
|
|
"rewards/accuracy_reward_step": 0.140625,
|
|
"rewards/format_reward_step": 0.34375,
|
|
"step": 8
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.4765625,
|
|
"calib/avg_num_step_conf": 2.3203125,
|
|
"calib/ece": 0.7459019108280255,
|
|
"calib/final_conf_rate": 0.61328125,
|
|
"calib/format_rate": 0.375,
|
|
"calib/frac_conf_gt_0.9": 0.39490445859872614,
|
|
"calib/gap": 0.018681028368794372,
|
|
"calib/mean_conf": 0.8363477707006369,
|
|
"calib/mu_c": 0.853125,
|
|
"calib/mu_w": 0.8344439716312056,
|
|
"calib/nonempty_final_conf_rate": 0.61328125,
|
|
"calib/nonempty_reasoning_rate": 0.59765625,
|
|
"calib/nonempty_step_conf_rate": 0.51171875,
|
|
"calib/pce": 0.7401694267515924,
|
|
"calib/std_conf": 0.19627053621659557,
|
|
"calib/step_conf_rate": 0.51171875,
|
|
"calib/step_q_c": 0.8187037037037037,
|
|
"calib/step_q_c_n": 54.0,
|
|
"calib/step_q_gap": 0.026277777777777844,
|
|
"calib/step_q_w": 0.7924259259259259,
|
|
"calib/step_q_w_n": 540.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 2760.0,
|
|
"completions/max_terminated_length": 2760.0,
|
|
"completions/mean_length": 354.1796875,
|
|
"completions/mean_terminated_length": 355.5686340332031,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0144,
|
|
"grad_norm": 17.99668312072754,
|
|
"learning_rate": 4e-07,
|
|
"loss": -0.0382,
|
|
"num_tokens": 3127356.0,
|
|
"reward": 0.25390625,
|
|
"reward_std": 0.328166127204895,
|
|
"rewards/accuracy_reward_step": 0.06640625,
|
|
"rewards/format_reward_step": 0.375,
|
|
"step": 9
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.56640625,
|
|
"calib/avg_num_step_conf": 2.5234375,
|
|
"calib/ece": 0.7055191256830601,
|
|
"calib/final_conf_rate": 0.71484375,
|
|
"calib/format_rate": 0.47265625,
|
|
"calib/frac_conf_gt_0.9": 0.5027322404371585,
|
|
"calib/gap": 0.05723646723646714,
|
|
"calib/mean_conf": 0.8530601092896175,
|
|
"calib/mu_c": 0.9018518518518519,
|
|
"calib/mu_w": 0.8446153846153848,
|
|
"calib/nonempty_final_conf_rate": 0.71484375,
|
|
"calib/nonempty_reasoning_rate": 0.62109375,
|
|
"calib/nonempty_step_conf_rate": 0.55078125,
|
|
"calib/pce": 0.7055191256830601,
|
|
"calib/std_conf": 0.188480932618246,
|
|
"calib/step_conf_rate": 0.55078125,
|
|
"calib/step_q_c": 0.8683146067415732,
|
|
"calib/step_q_c_n": 89.0,
|
|
"calib/step_q_gap": 0.07136667137353003,
|
|
"calib/step_q_w": 0.7969479353680432,
|
|
"calib/step_q_w_n": 557.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 2785.0,
|
|
"completions/max_terminated_length": 2785.0,
|
|
"completions/mean_length": 322.42578125,
|
|
"completions/mean_terminated_length": 323.6902160644531,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.016,
|
|
"grad_norm": 10.932899475097656,
|
|
"learning_rate": 4.5e-07,
|
|
"loss": -0.0028,
|
|
"num_tokens": 3470017.0,
|
|
"reward": 0.345703125,
|
|
"reward_std": 0.3697780966758728,
|
|
"rewards/accuracy_reward_step": 0.109375,
|
|
"rewards/format_reward_step": 0.47265625,
|
|
"step": 10
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.515625,
|
|
"calib/avg_num_step_conf": 2.28125,
|
|
"calib/ece": 0.6898159509202453,
|
|
"calib/final_conf_rate": 0.63671875,
|
|
"calib/format_rate": 0.37890625,
|
|
"calib/frac_conf_gt_0.9": 0.5276073619631901,
|
|
"calib/gap": 0.06452937649880097,
|
|
"calib/mean_conf": 0.8370552147239264,
|
|
"calib/mu_c": 0.8920833333333333,
|
|
"calib/mu_w": 0.8275539568345324,
|
|
"calib/nonempty_final_conf_rate": 0.63671875,
|
|
"calib/nonempty_reasoning_rate": 0.609375,
|
|
"calib/nonempty_step_conf_rate": 0.48828125,
|
|
"calib/pce": 0.6898159509202453,
|
|
"calib/std_conf": 0.2278276512391015,
|
|
"calib/step_conf_rate": 0.48828125,
|
|
"calib/step_q_c": 0.833896103896104,
|
|
"calib/step_q_c_n": 77.0,
|
|
"calib/step_q_gap": 0.056894131509516166,
|
|
"calib/step_q_w": 0.7770019723865879,
|
|
"calib/step_q_w_n": 507.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2631.0,
|
|
"completions/max_terminated_length": 2631.0,
|
|
"completions/mean_length": 315.140625,
|
|
"completions/mean_terminated_length": 317.6220397949219,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0176,
|
|
"grad_norm": 11.573603630065918,
|
|
"learning_rate": 5e-07,
|
|
"loss": -0.0437,
|
|
"num_tokens": 3807693.0,
|
|
"reward": 0.283203125,
|
|
"reward_std": 0.37644994258880615,
|
|
"rewards/accuracy_reward_step": 0.09375,
|
|
"rewards/format_reward_step": 0.37890625,
|
|
"step": 11
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.50390625,
|
|
"calib/avg_num_step_conf": 2.5234375,
|
|
"calib/ece": 0.6290196078431373,
|
|
"calib/final_conf_rate": 0.6640625,
|
|
"calib/format_rate": 0.36328125,
|
|
"calib/frac_conf_gt_0.9": 0.47058823529411764,
|
|
"calib/gap": 0.05849759533970078,
|
|
"calib/mean_conf": 0.8466666666666667,
|
|
"calib/mu_c": 0.8924324324324325,
|
|
"calib/mu_w": 0.8339348370927318,
|
|
"calib/nonempty_final_conf_rate": 0.6640625,
|
|
"calib/nonempty_reasoning_rate": 0.6484375,
|
|
"calib/nonempty_step_conf_rate": 0.5234375,
|
|
"calib/pce": 0.6290196078431373,
|
|
"calib/std_conf": 0.18737121939137066,
|
|
"calib/step_conf_rate": 0.5234375,
|
|
"calib/step_q_c": 0.7443396226415094,
|
|
"calib/step_q_c_n": 159.0,
|
|
"calib/step_q_gap": -0.07715524388826467,
|
|
"calib/step_q_w": 0.8214948665297741,
|
|
"calib/step_q_w_n": 487.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 2822.0,
|
|
"completions/max_terminated_length": 2822.0,
|
|
"completions/mean_length": 367.65234375,
|
|
"completions/mean_terminated_length": 369.0941467285156,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0192,
|
|
"grad_norm": 13.031610488891602,
|
|
"learning_rate": 5.5e-07,
|
|
"loss": -0.0015,
|
|
"num_tokens": 4162716.0,
|
|
"reward": 0.337890625,
|
|
"reward_std": 0.39855268597602844,
|
|
"rewards/accuracy_reward_step": 0.15625,
|
|
"rewards/format_reward_step": 0.36328125,
|
|
"step": 12
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.55078125,
|
|
"calib/avg_num_step_conf": 2.5625,
|
|
"calib/ece": 0.6973684210526316,
|
|
"calib/final_conf_rate": 0.66796875,
|
|
"calib/format_rate": 0.453125,
|
|
"calib/frac_conf_gt_0.9": 0.5146198830409356,
|
|
"calib/gap": 0.045510638297872275,
|
|
"calib/mean_conf": 0.8728070175438597,
|
|
"calib/mu_c": 0.9103333333333332,
|
|
"calib/mu_w": 0.8648226950354609,
|
|
"calib/nonempty_final_conf_rate": 0.66796875,
|
|
"calib/nonempty_reasoning_rate": 0.66015625,
|
|
"calib/nonempty_step_conf_rate": 0.57421875,
|
|
"calib/pce": 0.6973684210526316,
|
|
"calib/std_conf": 0.16450266411356892,
|
|
"calib/step_conf_rate": 0.57421875,
|
|
"calib/step_q_c": 0.8602912621359223,
|
|
"calib/step_q_c_n": 103.0,
|
|
"calib/step_q_gap": 0.08663845924261304,
|
|
"calib/step_q_w": 0.7736528028933093,
|
|
"calib/step_q_w_n": 553.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2592.0,
|
|
"completions/max_terminated_length": 2592.0,
|
|
"completions/mean_length": 302.15625,
|
|
"completions/mean_terminated_length": 304.5354309082031,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0208,
|
|
"grad_norm": 20.21171760559082,
|
|
"learning_rate": 6e-07,
|
|
"loss": -0.0033,
|
|
"num_tokens": 4502212.0,
|
|
"reward": 0.3515625,
|
|
"reward_std": 0.4020528793334961,
|
|
"rewards/accuracy_reward_step": 0.125,
|
|
"rewards/format_reward_step": 0.453125,
|
|
"step": 13
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.55078125,
|
|
"calib/avg_num_step_conf": 2.96875,
|
|
"calib/ece": 0.5569461077844313,
|
|
"calib/final_conf_rate": 0.65234375,
|
|
"calib/format_rate": 0.4453125,
|
|
"calib/frac_conf_gt_0.9": 0.48502994011976047,
|
|
"calib/gap": 0.10652485994397765,
|
|
"calib/mean_conf": 0.8326347305389222,
|
|
"calib/mu_c": 0.9085416666666667,
|
|
"calib/mu_w": 0.802016806722689,
|
|
"calib/nonempty_final_conf_rate": 0.65234375,
|
|
"calib/nonempty_reasoning_rate": 0.65234375,
|
|
"calib/nonempty_step_conf_rate": 0.5625,
|
|
"calib/pce": 0.5510778443113774,
|
|
"calib/std_conf": 0.2198589738323819,
|
|
"calib/step_conf_rate": 0.5625,
|
|
"calib/step_q_c": 0.8128994082840236,
|
|
"calib/step_q_c_n": 169.0,
|
|
"calib/step_q_gap": 0.08107199711651092,
|
|
"calib/step_q_w": 0.7318274111675127,
|
|
"calib/step_q_w_n": 591.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2400.0,
|
|
"completions/max_terminated_length": 2400.0,
|
|
"completions/mean_length": 316.21875,
|
|
"completions/mean_terminated_length": 318.7086486816406,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0224,
|
|
"grad_norm": 7.095739364624023,
|
|
"learning_rate": 6.5e-07,
|
|
"loss": -0.0308,
|
|
"num_tokens": 4844676.0,
|
|
"reward": 0.4140625,
|
|
"reward_std": 0.40704041719436646,
|
|
"rewards/accuracy_reward_step": 0.19140625,
|
|
"rewards/format_reward_step": 0.4453125,
|
|
"step": 14
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.53515625,
|
|
"calib/avg_num_step_conf": 2.64453125,
|
|
"calib/ece": 0.7221173184357543,
|
|
"calib/final_conf_rate": 0.69921875,
|
|
"calib/format_rate": 0.4375,
|
|
"calib/frac_conf_gt_0.9": 0.45251396648044695,
|
|
"calib/gap": 0.047625942684766054,
|
|
"calib/mean_conf": 0.867368715083799,
|
|
"calib/mu_c": 0.9080769230769229,
|
|
"calib/mu_w": 0.8604509803921568,
|
|
"calib/nonempty_final_conf_rate": 0.69921875,
|
|
"calib/nonempty_reasoning_rate": 0.6484375,
|
|
"calib/nonempty_step_conf_rate": 0.55859375,
|
|
"calib/pce": 0.7221173184357543,
|
|
"calib/std_conf": 0.15471728884704963,
|
|
"calib/step_conf_rate": 0.55859375,
|
|
"calib/step_q_c": 0.8536585365853658,
|
|
"calib/step_q_c_n": 82.0,
|
|
"calib/step_q_gap": 0.033087108013937216,
|
|
"calib/step_q_w": 0.8205714285714286,
|
|
"calib/step_q_w_n": 595.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2624.0,
|
|
"completions/max_terminated_length": 2624.0,
|
|
"completions/mean_length": 330.5859375,
|
|
"completions/mean_terminated_length": 333.18896484375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 3.0,
|
|
"epoch": 0.024,
|
|
"grad_norm": 15.420191764831543,
|
|
"learning_rate": 7e-07,
|
|
"loss": -0.0287,
|
|
"num_tokens": 5190482.0,
|
|
"reward": 0.32421875,
|
|
"reward_std": 0.38361692428588867,
|
|
"rewards/accuracy_reward_step": 0.10546875,
|
|
"rewards/format_reward_step": 0.4375,
|
|
"step": 15
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.64453125,
|
|
"calib/avg_num_step_conf": 3.0078125,
|
|
"calib/ece": 0.6353367875647669,
|
|
"calib/final_conf_rate": 0.75390625,
|
|
"calib/format_rate": 0.5,
|
|
"calib/frac_conf_gt_0.9": 0.5077720207253886,
|
|
"calib/gap": 0.021999408459035807,
|
|
"calib/mean_conf": 0.8736787564766839,
|
|
"calib/mu_c": 0.8904347826086957,
|
|
"calib/mu_w": 0.8684353741496599,
|
|
"calib/nonempty_final_conf_rate": 0.75390625,
|
|
"calib/nonempty_reasoning_rate": 0.75390625,
|
|
"calib/nonempty_step_conf_rate": 0.63671875,
|
|
"calib/pce": 0.6353367875647669,
|
|
"calib/std_conf": 0.15623630440125438,
|
|
"calib/step_conf_rate": 0.63671875,
|
|
"calib/step_q_c": 0.8060789473684209,
|
|
"calib/step_q_c_n": 190.0,
|
|
"calib/step_q_gap": -0.03312794918330331,
|
|
"calib/step_q_w": 0.8392068965517242,
|
|
"calib/step_q_w_n": 580.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 2530.0,
|
|
"completions/max_terminated_length": 2530.0,
|
|
"completions/mean_length": 299.28515625,
|
|
"completions/mean_terminated_length": 300.4588317871094,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 4.0,
|
|
"epoch": 0.0256,
|
|
"grad_norm": 30.280681610107422,
|
|
"learning_rate": 7.5e-07,
|
|
"loss": -0.049,
|
|
"num_tokens": 5528243.0,
|
|
"reward": 0.4296875,
|
|
"reward_std": 0.421233594417572,
|
|
"rewards/accuracy_reward_step": 0.1796875,
|
|
"rewards/format_reward_step": 0.5,
|
|
"step": 16
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.5859375,
|
|
"calib/avg_num_step_conf": 3.3828125,
|
|
"calib/ece": 0.680291208791209,
|
|
"calib/final_conf_rate": 0.7109375,
|
|
"calib/format_rate": 0.48046875,
|
|
"calib/frac_conf_gt_0.9": 0.4945054945054945,
|
|
"calib/gap": 0.054295505389464926,
|
|
"calib/mean_conf": 0.8616098901098903,
|
|
"calib/mu_c": 0.906060606060606,
|
|
"calib/mu_w": 0.8517651006711411,
|
|
"calib/nonempty_final_conf_rate": 0.7109375,
|
|
"calib/nonempty_reasoning_rate": 0.70703125,
|
|
"calib/nonempty_step_conf_rate": 0.6171875,
|
|
"calib/pce": 0.680291208791209,
|
|
"calib/std_conf": 0.17178502573571683,
|
|
"calib/step_conf_rate": 0.6171875,
|
|
"calib/step_q_c": 0.8084076433121017,
|
|
"calib/step_q_c_n": 157.0,
|
|
"calib/step_q_gap": 0.051073369687277936,
|
|
"calib/step_q_w": 0.7573342736248238,
|
|
"calib/step_q_w_n": 709.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2685.0,
|
|
"completions/max_terminated_length": 2685.0,
|
|
"completions/mean_length": 325.03125,
|
|
"completions/mean_terminated_length": 327.5905456542969,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0272,
|
|
"grad_norm": 35.424591064453125,
|
|
"learning_rate": 8e-07,
|
|
"loss": -0.0219,
|
|
"num_tokens": 5862803.0,
|
|
"reward": 0.369140625,
|
|
"reward_std": 0.41405531764030457,
|
|
"rewards/accuracy_reward_step": 0.12890625,
|
|
"rewards/format_reward_step": 0.48046875,
|
|
"step": 17
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.59375,
|
|
"calib/avg_num_step_conf": 3.09765625,
|
|
"calib/ece": 0.7170000000000001,
|
|
"calib/final_conf_rate": 0.703125,
|
|
"calib/format_rate": 0.50390625,
|
|
"calib/frac_conf_gt_0.9": 0.48333333333333334,
|
|
"calib/gap": 0.039259259259259216,
|
|
"calib/mean_conf": 0.8603333333333334,
|
|
"calib/mu_c": 0.8937037037037038,
|
|
"calib/mu_w": 0.8544444444444446,
|
|
"calib/nonempty_final_conf_rate": 0.703125,
|
|
"calib/nonempty_reasoning_rate": 0.6953125,
|
|
"calib/nonempty_step_conf_rate": 0.6328125,
|
|
"calib/pce": 0.7136666666666668,
|
|
"calib/std_conf": 0.16531081835943667,
|
|
"calib/step_conf_rate": 0.6328125,
|
|
"calib/step_q_c": 0.8222962962962964,
|
|
"calib/step_q_c_n": 135.0,
|
|
"calib/step_q_gap": 0.037478667116965014,
|
|
"calib/step_q_w": 0.7848176291793314,
|
|
"calib/step_q_w_n": 658.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2649.0,
|
|
"completions/max_terminated_length": 2649.0,
|
|
"completions/mean_length": 348.58203125,
|
|
"completions/mean_terminated_length": 351.3267822265625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0288,
|
|
"grad_norm": 98.14373779296875,
|
|
"learning_rate": 8.499999999999999e-07,
|
|
"loss": -0.0782,
|
|
"num_tokens": 6213768.0,
|
|
"reward": 0.369140625,
|
|
"reward_std": 0.40399375557899475,
|
|
"rewards/accuracy_reward_step": 0.1171875,
|
|
"rewards/format_reward_step": 0.50390625,
|
|
"step": 18
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.53515625,
|
|
"calib/avg_num_step_conf": 2.47265625,
|
|
"calib/ece": 0.6429113924050631,
|
|
"calib/final_conf_rate": 0.6171875,
|
|
"calib/format_rate": 0.41015625,
|
|
"calib/frac_conf_gt_0.9": 0.5,
|
|
"calib/gap": 0.08930424242424251,
|
|
"calib/mean_conf": 0.8517721518987342,
|
|
"calib/mu_c": 0.9224242424242424,
|
|
"calib/mu_w": 0.8331199999999999,
|
|
"calib/nonempty_final_conf_rate": 0.6171875,
|
|
"calib/nonempty_reasoning_rate": 0.6328125,
|
|
"calib/nonempty_step_conf_rate": 0.5234375,
|
|
"calib/pce": 0.6429113924050631,
|
|
"calib/std_conf": 0.18752727729803034,
|
|
"calib/step_conf_rate": 0.5234375,
|
|
"calib/step_q_c": 0.8583458646616541,
|
|
"calib/step_q_c_n": 133.0,
|
|
"calib/step_q_gap": 0.08488586466165404,
|
|
"calib/step_q_w": 0.77346,
|
|
"calib/step_q_w_n": 500.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2535.0,
|
|
"completions/max_terminated_length": 2535.0,
|
|
"completions/mean_length": 258.1328125,
|
|
"completions/mean_terminated_length": 260.16534423828125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0304,
|
|
"grad_norm": 32.44593811035156,
|
|
"learning_rate": 9e-07,
|
|
"loss": -0.1044,
|
|
"num_tokens": 6540946.0,
|
|
"reward": 0.333984375,
|
|
"reward_std": 0.4141427278518677,
|
|
"rewards/accuracy_reward_step": 0.12890625,
|
|
"rewards/format_reward_step": 0.41015625,
|
|
"step": 19
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.57421875,
|
|
"calib/avg_num_step_conf": 2.9765625,
|
|
"calib/ece": 0.7031005917159765,
|
|
"calib/final_conf_rate": 0.66015625,
|
|
"calib/format_rate": 0.46875,
|
|
"calib/frac_conf_gt_0.9": 0.5739644970414202,
|
|
"calib/gap": -0.00785918003565067,
|
|
"calib/mean_conf": 0.885112426035503,
|
|
"calib/mu_c": 0.8787878787878787,
|
|
"calib/mu_w": 0.8866470588235293,
|
|
"calib/nonempty_final_conf_rate": 0.66015625,
|
|
"calib/nonempty_reasoning_rate": 0.6953125,
|
|
"calib/nonempty_step_conf_rate": 0.60546875,
|
|
"calib/pce": 0.6964733727810652,
|
|
"calib/std_conf": 0.15612767616829498,
|
|
"calib/step_conf_rate": 0.60546875,
|
|
"calib/step_q_c": 0.80135593220339,
|
|
"calib/step_q_c_n": 118.0,
|
|
"calib/step_q_gap": -0.021159595746920612,
|
|
"calib/step_q_w": 0.8225155279503106,
|
|
"calib/step_q_w_n": 644.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2500.0,
|
|
"completions/max_terminated_length": 2500.0,
|
|
"completions/mean_length": 297.640625,
|
|
"completions/mean_terminated_length": 299.9842529296875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.032,
|
|
"grad_norm": 7.341494560241699,
|
|
"learning_rate": 9.499999999999999e-07,
|
|
"loss": -0.0809,
|
|
"num_tokens": 6878446.0,
|
|
"reward": 0.36328125,
|
|
"reward_std": 0.38229066133499146,
|
|
"rewards/accuracy_reward_step": 0.12890625,
|
|
"rewards/format_reward_step": 0.46875,
|
|
"step": 20
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.6640625,
|
|
"calib/avg_num_step_conf": 2.97265625,
|
|
"calib/ece": 0.6631351351351351,
|
|
"calib/final_conf_rate": 0.72265625,
|
|
"calib/format_rate": 0.52734375,
|
|
"calib/frac_conf_gt_0.9": 0.5297297297297298,
|
|
"calib/gap": 0.06178123881131381,
|
|
"calib/mean_conf": 0.8685405405405405,
|
|
"calib/mu_c": 0.9176315789473684,
|
|
"calib/mu_w": 0.8558503401360545,
|
|
"calib/nonempty_final_conf_rate": 0.72265625,
|
|
"calib/nonempty_reasoning_rate": 0.734375,
|
|
"calib/nonempty_step_conf_rate": 0.609375,
|
|
"calib/pce": 0.6631351351351351,
|
|
"calib/std_conf": 0.1669844780931214,
|
|
"calib/step_conf_rate": 0.609375,
|
|
"calib/step_q_c": 0.8349693251533742,
|
|
"calib/step_q_c_n": 163.0,
|
|
"calib/step_q_gap": 0.027761967293842393,
|
|
"calib/step_q_w": 0.8072073578595318,
|
|
"calib/step_q_w_n": 598.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 2633.0,
|
|
"completions/max_terminated_length": 2633.0,
|
|
"completions/mean_length": 319.4296875,
|
|
"completions/mean_terminated_length": 320.682373046875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0336,
|
|
"grad_norm": 12.134476661682129,
|
|
"learning_rate": 1e-06,
|
|
"loss": -0.1004,
|
|
"num_tokens": 7218700.0,
|
|
"reward": 0.419921875,
|
|
"reward_std": 0.4154573082923889,
|
|
"rewards/accuracy_reward_step": 0.15625,
|
|
"rewards/format_reward_step": 0.52734375,
|
|
"step": 21
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.71484375,
|
|
"calib/avg_num_step_conf": 3.5625,
|
|
"calib/ece": 0.7179272727272729,
|
|
"calib/final_conf_rate": 0.7734375,
|
|
"calib/format_rate": 0.62109375,
|
|
"calib/frac_conf_gt_0.9": 0.5909090909090909,
|
|
"calib/gap": 0.031584218077475024,
|
|
"calib/mean_conf": 0.8856040404040405,
|
|
"calib/mu_c": 0.9117647058823529,
|
|
"calib/mu_w": 0.8801804878048779,
|
|
"calib/nonempty_final_conf_rate": 0.7734375,
|
|
"calib/nonempty_reasoning_rate": 0.78125,
|
|
"calib/nonempty_step_conf_rate": 0.703125,
|
|
"calib/pce": 0.7159070707070709,
|
|
"calib/std_conf": 0.15533752622718763,
|
|
"calib/step_conf_rate": 0.703125,
|
|
"calib/step_q_c": 0.8201428571428571,
|
|
"calib/step_q_c_n": 140.0,
|
|
"calib/step_q_gap": -0.0040799407846040525,
|
|
"calib/step_q_w": 0.8242227979274611,
|
|
"calib/step_q_w_n": 772.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2740.0,
|
|
"completions/max_terminated_length": 2740.0,
|
|
"completions/mean_length": 298.83203125,
|
|
"completions/mean_terminated_length": 301.1850280761719,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0352,
|
|
"grad_norm": 25.838075637817383,
|
|
"learning_rate": 9.944444444444444e-07,
|
|
"loss": -0.0722,
|
|
"num_tokens": 7557345.0,
|
|
"reward": 0.447265625,
|
|
"reward_std": 0.4125175476074219,
|
|
"rewards/accuracy_reward_step": 0.13671875,
|
|
"rewards/format_reward_step": 0.62109375,
|
|
"step": 22
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.6953125,
|
|
"calib/avg_num_step_conf": 3.41015625,
|
|
"calib/ece": 0.6334895833333334,
|
|
"calib/final_conf_rate": 0.75,
|
|
"calib/format_rate": 0.58984375,
|
|
"calib/frac_conf_gt_0.9": 0.640625,
|
|
"calib/gap": 0.04990498167503732,
|
|
"calib/mean_conf": 0.90953125,
|
|
"calib/mu_c": 0.9456603773584905,
|
|
"calib/mu_w": 0.8957553956834532,
|
|
"calib/nonempty_final_conf_rate": 0.75,
|
|
"calib/nonempty_reasoning_rate": 0.75390625,
|
|
"calib/nonempty_step_conf_rate": 0.671875,
|
|
"calib/pce": 0.6334895833333334,
|
|
"calib/std_conf": 0.11774482058008964,
|
|
"calib/step_conf_rate": 0.671875,
|
|
"calib/step_q_c": 0.881795918367347,
|
|
"calib/step_q_c_n": 245.0,
|
|
"calib/step_q_gap": 0.036557064864162436,
|
|
"calib/step_q_w": 0.8452388535031846,
|
|
"calib/step_q_w_n": 628.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 2497.0,
|
|
"completions/max_terminated_length": 2497.0,
|
|
"completions/mean_length": 287.34375,
|
|
"completions/mean_terminated_length": 288.4706115722656,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0368,
|
|
"grad_norm": 32.517234802246094,
|
|
"learning_rate": 9.88888888888889e-07,
|
|
"loss": -0.0535,
|
|
"num_tokens": 7891513.0,
|
|
"reward": 0.505859375,
|
|
"reward_std": 0.4474440813064575,
|
|
"rewards/accuracy_reward_step": 0.2109375,
|
|
"rewards/format_reward_step": 0.58984375,
|
|
"step": 23
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.71875,
|
|
"calib/avg_num_step_conf": 3.5390625,
|
|
"calib/ece": 0.6369543147208123,
|
|
"calib/final_conf_rate": 0.76953125,
|
|
"calib/format_rate": 0.63671875,
|
|
"calib/frac_conf_gt_0.9": 0.6649746192893401,
|
|
"calib/gap": 0.05553333333333332,
|
|
"calib/mean_conf": 0.8907614213197971,
|
|
"calib/mu_c": 0.9322,
|
|
"calib/mu_w": 0.8766666666666667,
|
|
"calib/nonempty_final_conf_rate": 0.76953125,
|
|
"calib/nonempty_reasoning_rate": 0.79296875,
|
|
"calib/nonempty_step_conf_rate": 0.71875,
|
|
"calib/pce": 0.6369543147208123,
|
|
"calib/std_conf": 0.1714729527857561,
|
|
"calib/step_conf_rate": 0.71875,
|
|
"calib/step_q_c": 0.8439673913043478,
|
|
"calib/step_q_c_n": 184.0,
|
|
"calib/step_q_gap": 0.03170977357581595,
|
|
"calib/step_q_w": 0.8122576177285319,
|
|
"calib/step_q_w_n": 722.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2468.0,
|
|
"completions/max_terminated_length": 2468.0,
|
|
"completions/mean_length": 285.83203125,
|
|
"completions/mean_terminated_length": 288.0826721191406,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0384,
|
|
"grad_norm": 28.229576110839844,
|
|
"learning_rate": 9.833333333333332e-07,
|
|
"loss": -0.052,
|
|
"num_tokens": 8224238.0,
|
|
"reward": 0.513671875,
|
|
"reward_std": 0.42473044991493225,
|
|
"rewards/accuracy_reward_step": 0.1953125,
|
|
"rewards/format_reward_step": 0.63671875,
|
|
"step": 24
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.72265625,
|
|
"calib/avg_num_step_conf": 3.98046875,
|
|
"calib/ece": 0.6131979695431471,
|
|
"calib/final_conf_rate": 0.76953125,
|
|
"calib/format_rate": 0.6484375,
|
|
"calib/frac_conf_gt_0.9": 0.5634517766497462,
|
|
"calib/gap": 0.05753561253561279,
|
|
"calib/mean_conf": 0.8873096446700508,
|
|
"calib/mu_c": 0.9290740740740744,
|
|
"calib/mu_w": 0.8715384615384616,
|
|
"calib/nonempty_final_conf_rate": 0.76953125,
|
|
"calib/nonempty_reasoning_rate": 0.78515625,
|
|
"calib/nonempty_step_conf_rate": 0.72265625,
|
|
"calib/pce": 0.6131979695431471,
|
|
"calib/std_conf": 0.14102010713951482,
|
|
"calib/step_conf_rate": 0.72265625,
|
|
"calib/step_q_c": 0.8701724137931034,
|
|
"calib/step_q_c_n": 232.0,
|
|
"calib/step_q_gap": 0.05313302370415807,
|
|
"calib/step_q_w": 0.8170393900889453,
|
|
"calib/step_q_w_n": 787.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2438.0,
|
|
"completions/max_terminated_length": 2438.0,
|
|
"completions/mean_length": 258.58203125,
|
|
"completions/mean_terminated_length": 260.61810302734375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.04,
|
|
"grad_norm": 22.257802963256836,
|
|
"learning_rate": 9.777777777777778e-07,
|
|
"loss": -0.085,
|
|
"num_tokens": 8552115.0,
|
|
"reward": 0.5390625,
|
|
"reward_std": 0.4408751428127289,
|
|
"rewards/accuracy_reward_step": 0.21484375,
|
|
"rewards/format_reward_step": 0.6484375,
|
|
"step": 25
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.76171875,
|
|
"calib/avg_num_step_conf": 3.8515625,
|
|
"calib/ece": 0.7327488151658769,
|
|
"calib/final_conf_rate": 0.82421875,
|
|
"calib/format_rate": 0.66015625,
|
|
"calib/frac_conf_gt_0.9": 0.6161137440758294,
|
|
"calib/gap": 0.041826984126984046,
|
|
"calib/mean_conf": 0.9033649289099527,
|
|
"calib/mu_c": 0.9380555555555556,
|
|
"calib/mu_w": 0.8962285714285716,
|
|
"calib/nonempty_final_conf_rate": 0.82421875,
|
|
"calib/nonempty_reasoning_rate": 0.80859375,
|
|
"calib/nonempty_step_conf_rate": 0.7265625,
|
|
"calib/pce": 0.7327488151658769,
|
|
"calib/std_conf": 0.1236090554144201,
|
|
"calib/step_conf_rate": 0.7265625,
|
|
"calib/step_q_c": 0.8673885350318472,
|
|
"calib/step_q_c_n": 157.0,
|
|
"calib/step_q_gap": 0.028088173150061957,
|
|
"calib/step_q_w": 0.8393003618817853,
|
|
"calib/step_q_w_n": 829.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 2222.0,
|
|
"completions/max_terminated_length": 2222.0,
|
|
"completions/mean_length": 257.65234375,
|
|
"completions/mean_terminated_length": 258.6627502441406,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.0416,
|
|
"grad_norm": 25.09646987915039,
|
|
"learning_rate": 9.722222222222222e-07,
|
|
"loss": -0.1375,
|
|
"num_tokens": 8878442.0,
|
|
"reward": 0.470703125,
|
|
"reward_std": 0.407234251499176,
|
|
"rewards/accuracy_reward_step": 0.140625,
|
|
"rewards/format_reward_step": 0.66015625,
|
|
"step": 26
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.73828125,
|
|
"calib/avg_num_step_conf": 3.6875,
|
|
"calib/ece": 0.70248730964467,
|
|
"calib/final_conf_rate": 0.76953125,
|
|
"calib/format_rate": 0.6171875,
|
|
"calib/frac_conf_gt_0.9": 0.5076142131979695,
|
|
"calib/gap": 0.06638036809815939,
|
|
"calib/mean_conf": 0.8750761421319798,
|
|
"calib/mu_c": 0.9299999999999999,
|
|
"calib/mu_w": 0.8636196319018405,
|
|
"calib/nonempty_final_conf_rate": 0.76953125,
|
|
"calib/nonempty_reasoning_rate": 0.796875,
|
|
"calib/nonempty_step_conf_rate": 0.703125,
|
|
"calib/pce": 0.70248730964467,
|
|
"calib/std_conf": 0.162401455600176,
|
|
"calib/step_conf_rate": 0.703125,
|
|
"calib/step_q_c": 0.8438848920863309,
|
|
"calib/step_q_c_n": 139.0,
|
|
"calib/step_q_gap": 0.02372340140310103,
|
|
"calib/step_q_w": 0.8201614906832299,
|
|
"calib/step_q_w_n": 805.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2806.0,
|
|
"completions/max_terminated_length": 2806.0,
|
|
"completions/mean_length": 278.7109375,
|
|
"completions/mean_terminated_length": 280.905517578125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0432,
|
|
"grad_norm": 45.72233963012695,
|
|
"learning_rate": 9.666666666666666e-07,
|
|
"loss": -0.0227,
|
|
"num_tokens": 9211936.0,
|
|
"reward": 0.44140625,
|
|
"reward_std": 0.37693342566490173,
|
|
"rewards/accuracy_reward_step": 0.1328125,
|
|
"rewards/format_reward_step": 0.6171875,
|
|
"step": 27
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.79296875,
|
|
"calib/avg_num_step_conf": 3.9375,
|
|
"calib/ece": 0.6911004784688995,
|
|
"calib/final_conf_rate": 0.81640625,
|
|
"calib/format_rate": 0.69140625,
|
|
"calib/frac_conf_gt_0.9": 0.5933014354066986,
|
|
"calib/gap": 0.030397793263646777,
|
|
"calib/mean_conf": 0.8872727272727272,
|
|
"calib/mu_c": 0.9117073170731707,
|
|
"calib/mu_w": 0.8813095238095239,
|
|
"calib/nonempty_final_conf_rate": 0.81640625,
|
|
"calib/nonempty_reasoning_rate": 0.83984375,
|
|
"calib/nonempty_step_conf_rate": 0.75390625,
|
|
"calib/pce": 0.6911004784688995,
|
|
"calib/std_conf": 0.1610543231267796,
|
|
"calib/step_conf_rate": 0.75390625,
|
|
"calib/step_q_c": 0.8209389671361503,
|
|
"calib/step_q_c_n": 213.0,
|
|
"calib/step_q_gap": -0.023287447958189267,
|
|
"calib/step_q_w": 0.8442264150943396,
|
|
"calib/step_q_w_n": 795.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 1805.0,
|
|
"completions/max_terminated_length": 1805.0,
|
|
"completions/mean_length": 260.96484375,
|
|
"completions/mean_terminated_length": 261.9882507324219,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0448,
|
|
"grad_norm": 27.64651107788086,
|
|
"learning_rate": 9.61111111111111e-07,
|
|
"loss": -0.1104,
|
|
"num_tokens": 9535415.0,
|
|
"reward": 0.505859375,
|
|
"reward_std": 0.41910263895988464,
|
|
"rewards/accuracy_reward_step": 0.16015625,
|
|
"rewards/format_reward_step": 0.69140625,
|
|
"step": 28
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.7265625,
|
|
"calib/avg_num_step_conf": 3.7578125,
|
|
"calib/ece": 0.7486528497409327,
|
|
"calib/final_conf_rate": 0.75390625,
|
|
"calib/format_rate": 0.62890625,
|
|
"calib/frac_conf_gt_0.9": 0.694300518134715,
|
|
"calib/gap": 0.031744953416149024,
|
|
"calib/mean_conf": 0.9144559585492228,
|
|
"calib/mu_c": 0.9409375,
|
|
"calib/mu_w": 0.909192546583851,
|
|
"calib/nonempty_final_conf_rate": 0.75390625,
|
|
"calib/nonempty_reasoning_rate": 0.80078125,
|
|
"calib/nonempty_step_conf_rate": 0.73046875,
|
|
"calib/pce": 0.7486528497409327,
|
|
"calib/std_conf": 0.11464222055563011,
|
|
"calib/step_conf_rate": 0.73046875,
|
|
"calib/step_q_c": 0.8688235294117648,
|
|
"calib/step_q_c_n": 153.0,
|
|
"calib/step_q_gap": 0.028020068348723948,
|
|
"calib/step_q_w": 0.8408034610630408,
|
|
"calib/step_q_w_n": 809.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2102.0,
|
|
"completions/max_terminated_length": 2102.0,
|
|
"completions/mean_length": 298.55859375,
|
|
"completions/mean_terminated_length": 300.9094543457031,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0464,
|
|
"grad_norm": 28.71442413330078,
|
|
"learning_rate": 9.555555555555556e-07,
|
|
"loss": -0.08,
|
|
"num_tokens": 9873230.0,
|
|
"reward": 0.447265625,
|
|
"reward_std": 0.38781794905662537,
|
|
"rewards/accuracy_reward_step": 0.1328125,
|
|
"rewards/format_reward_step": 0.62890625,
|
|
"step": 29
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.80078125,
|
|
"calib/avg_num_step_conf": 4.0703125,
|
|
"calib/ece": 0.5768047619047619,
|
|
"calib/final_conf_rate": 0.8203125,
|
|
"calib/format_rate": 0.72265625,
|
|
"calib/frac_conf_gt_0.9": 0.6142857142857143,
|
|
"calib/gap": 0.04852389762565534,
|
|
"calib/mean_conf": 0.9053761904761906,
|
|
"calib/mu_c": 0.9379565217391306,
|
|
"calib/mu_w": 0.8894326241134752,
|
|
"calib/nonempty_final_conf_rate": 0.8203125,
|
|
"calib/nonempty_reasoning_rate": 0.8359375,
|
|
"calib/nonempty_step_conf_rate": 0.765625,
|
|
"calib/pce": 0.5768047619047619,
|
|
"calib/std_conf": 0.10518082572300542,
|
|
"calib/step_conf_rate": 0.765625,
|
|
"calib/step_q_c": 0.8512857142857142,
|
|
"calib/step_q_c_n": 315.0,
|
|
"calib/step_q_gap": 0.030652977009235527,
|
|
"calib/step_q_w": 0.8206327372764787,
|
|
"calib/step_q_w_n": 727.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2438.0,
|
|
"completions/max_terminated_length": 2438.0,
|
|
"completions/mean_length": 250.2265625,
|
|
"completions/mean_terminated_length": 252.1968536376953,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.048,
|
|
"grad_norm": 48.1530647277832,
|
|
"learning_rate": 9.499999999999999e-07,
|
|
"loss": -0.1343,
|
|
"num_tokens": 10198696.0,
|
|
"reward": 0.630859375,
|
|
"reward_std": 0.416718989610672,
|
|
"rewards/accuracy_reward_step": 0.26953125,
|
|
"rewards/format_reward_step": 0.72265625,
|
|
"step": 30
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.7890625,
|
|
"calib/avg_num_step_conf": 3.76171875,
|
|
"calib/ece": 0.5877522935779818,
|
|
"calib/final_conf_rate": 0.8515625,
|
|
"calib/format_rate": 0.71875,
|
|
"calib/frac_conf_gt_0.9": 0.6559633027522935,
|
|
"calib/gap": 0.04655675517945701,
|
|
"calib/mean_conf": 0.9042660550458715,
|
|
"calib/mu_c": 0.936086956521739,
|
|
"calib/mu_w": 0.889530201342282,
|
|
"calib/nonempty_final_conf_rate": 0.8515625,
|
|
"calib/nonempty_reasoning_rate": 0.81640625,
|
|
"calib/nonempty_step_conf_rate": 0.7578125,
|
|
"calib/pce": 0.5877522935779818,
|
|
"calib/std_conf": 0.13482119604419016,
|
|
"calib/step_conf_rate": 0.7578125,
|
|
"calib/step_q_c": 0.85013201320132,
|
|
"calib/step_q_c_n": 303.0,
|
|
"calib/step_q_gap": 0.004071407140713856,
|
|
"calib/step_q_w": 0.8460606060606062,
|
|
"calib/step_q_w_n": 660.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0,
|
|
"completions/max_length": 1196.0,
|
|
"completions/max_terminated_length": 1196.0,
|
|
"completions/mean_length": 233.57421875,
|
|
"completions/mean_terminated_length": 233.57421875,
|
|
"completions/min_length": 1.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.0496,
|
|
"grad_norm": 30.72689437866211,
|
|
"learning_rate": 9.444444444444444e-07,
|
|
"loss": -0.1023,
|
|
"num_tokens": 10518179.0,
|
|
"reward": 0.62890625,
|
|
"reward_std": 0.42506512999534607,
|
|
"rewards/accuracy_reward_step": 0.26953125,
|
|
"rewards/format_reward_step": 0.71875,
|
|
"step": 31
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.81640625,
|
|
"calib/avg_num_step_conf": 4.25,
|
|
"calib/ece": 0.6657651162790699,
|
|
"calib/final_conf_rate": 0.83984375,
|
|
"calib/format_rate": 0.75,
|
|
"calib/frac_conf_gt_0.9": 0.6186046511627907,
|
|
"calib/gap": 0.027463636363636357,
|
|
"calib/mean_conf": 0.8983232558139534,
|
|
"calib/mu_c": 0.9194000000000001,
|
|
"calib/mu_w": 0.8919363636363637,
|
|
"calib/nonempty_final_conf_rate": 0.83984375,
|
|
"calib/nonempty_reasoning_rate": 0.8671875,
|
|
"calib/nonempty_step_conf_rate": 0.80859375,
|
|
"calib/pce": 0.6657651162790699,
|
|
"calib/std_conf": 0.11949634194500217,
|
|
"calib/step_conf_rate": 0.80859375,
|
|
"calib/step_q_c": 0.8575824175824176,
|
|
"calib/step_q_c_n": 273.0,
|
|
"calib/step_q_gap": 0.03659836850266285,
|
|
"calib/step_q_w": 0.8209840490797548,
|
|
"calib/step_q_w_n": 815.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 819.0,
|
|
"completions/max_terminated_length": 819.0,
|
|
"completions/mean_length": 249.30859375,
|
|
"completions/mean_terminated_length": 251.2716522216797,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.0512,
|
|
"grad_norm": 56.229164123535156,
|
|
"learning_rate": 9.388888888888888e-07,
|
|
"loss": -0.1638,
|
|
"num_tokens": 10844042.0,
|
|
"reward": 0.57421875,
|
|
"reward_std": 0.3676033020019531,
|
|
"rewards/accuracy_reward_step": 0.19921875,
|
|
"rewards/format_reward_step": 0.75,
|
|
"step": 32
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.91796875,
|
|
"calib/avg_num_step_conf": 4.984375,
|
|
"calib/ece": 0.6375208333333335,
|
|
"calib/final_conf_rate": 0.9375,
|
|
"calib/format_rate": 0.8125,
|
|
"calib/frac_conf_gt_0.9": 0.5833333333333334,
|
|
"calib/gap": 0.04897517216382752,
|
|
"calib/mean_conf": 0.8958541666666666,
|
|
"calib/mu_c": 0.9321774193548388,
|
|
"calib/mu_w": 0.8832022471910113,
|
|
"calib/nonempty_final_conf_rate": 0.9375,
|
|
"calib/nonempty_reasoning_rate": 0.953125,
|
|
"calib/nonempty_step_conf_rate": 0.8671875,
|
|
"calib/pce": 0.6375208333333335,
|
|
"calib/std_conf": 0.13658025564714282,
|
|
"calib/step_conf_rate": 0.8671875,
|
|
"calib/step_q_c": 0.8521617647058823,
|
|
"calib/step_q_c_n": 340.0,
|
|
"calib/step_q_gap": 0.028700226244343874,
|
|
"calib/step_q_w": 0.8234615384615385,
|
|
"calib/step_q_w_n": 936.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 893.0,
|
|
"completions/max_terminated_length": 893.0,
|
|
"completions/mean_length": 258.7109375,
|
|
"completions/mean_terminated_length": 260.7480163574219,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0528,
|
|
"grad_norm": 33.666343688964844,
|
|
"learning_rate": 9.333333333333333e-07,
|
|
"loss": -0.0623,
|
|
"num_tokens": 11171512.0,
|
|
"reward": 0.6484375,
|
|
"reward_std": 0.3970039486885071,
|
|
"rewards/accuracy_reward_step": 0.2421875,
|
|
"rewards/format_reward_step": 0.8125,
|
|
"step": 33
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.9140625,
|
|
"calib/avg_num_step_conf": 4.79296875,
|
|
"calib/ece": 0.7196583333333335,
|
|
"calib/final_conf_rate": 0.9375,
|
|
"calib/format_rate": 0.84765625,
|
|
"calib/frac_conf_gt_0.9": 0.6416666666666667,
|
|
"calib/gap": 0.04740976739058533,
|
|
"calib/mean_conf": 0.9154916666666666,
|
|
"calib/mu_c": 0.9536170212765959,
|
|
"calib/mu_w": 0.9062072538860105,
|
|
"calib/nonempty_final_conf_rate": 0.9375,
|
|
"calib/nonempty_reasoning_rate": 0.9375,
|
|
"calib/nonempty_step_conf_rate": 0.890625,
|
|
"calib/pce": 0.7196583333333335,
|
|
"calib/std_conf": 0.09544143019266958,
|
|
"calib/step_conf_rate": 0.890625,
|
|
"calib/step_q_c": 0.87369918699187,
|
|
"calib/step_q_c_n": 246.0,
|
|
"calib/step_q_gap": 0.038581959672807864,
|
|
"calib/step_q_w": 0.8351172273190621,
|
|
"calib/step_q_w_n": 981.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 1319.0,
|
|
"completions/max_terminated_length": 1319.0,
|
|
"completions/mean_length": 261.5859375,
|
|
"completions/mean_terminated_length": 262.6117858886719,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.0544,
|
|
"grad_norm": 18.248090744018555,
|
|
"learning_rate": 9.277777777777777e-07,
|
|
"loss": -0.0844,
|
|
"num_tokens": 11499734.0,
|
|
"reward": 0.607421875,
|
|
"reward_std": 0.3113514184951782,
|
|
"rewards/accuracy_reward_step": 0.18359375,
|
|
"rewards/format_reward_step": 0.84765625,
|
|
"step": 34
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.9453125,
|
|
"calib/avg_num_step_conf": 4.90625,
|
|
"calib/ece": 0.5461216326530611,
|
|
"calib/final_conf_rate": 0.95703125,
|
|
"calib/format_rate": 0.8828125,
|
|
"calib/frac_conf_gt_0.9": 0.6530612244897959,
|
|
"calib/gap": 0.02993523728331904,
|
|
"calib/mean_conf": 0.9216318367346938,
|
|
"calib/mu_c": 0.9403260869565218,
|
|
"calib/mu_w": 0.9103908496732027,
|
|
"calib/nonempty_final_conf_rate": 0.95703125,
|
|
"calib/nonempty_reasoning_rate": 0.96484375,
|
|
"calib/nonempty_step_conf_rate": 0.91796875,
|
|
"calib/pce": 0.5461216326530611,
|
|
"calib/std_conf": 0.0916808045623392,
|
|
"calib/step_conf_rate": 0.91796875,
|
|
"calib/step_q_c": 0.8607843137254901,
|
|
"calib/step_q_c_n": 408.0,
|
|
"calib/step_q_gap": 0.011596224102848662,
|
|
"calib/step_q_w": 0.8491880896226415,
|
|
"calib/step_q_w_n": 848.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 1752.0,
|
|
"completions/max_terminated_length": 1752.0,
|
|
"completions/mean_length": 253.953125,
|
|
"completions/mean_terminated_length": 254.94903564453125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 19.0,
|
|
"epoch": 0.056,
|
|
"grad_norm": 10.243550300598145,
|
|
"learning_rate": 9.222222222222222e-07,
|
|
"loss": -0.0095,
|
|
"num_tokens": 11826890.0,
|
|
"reward": 0.80078125,
|
|
"reward_std": 0.39892858266830444,
|
|
"rewards/accuracy_reward_step": 0.359375,
|
|
"rewards/format_reward_step": 0.8828125,
|
|
"step": 35
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.9296875,
|
|
"calib/avg_num_step_conf": 4.875,
|
|
"calib/ece": 0.5691666666666667,
|
|
"calib/final_conf_rate": 0.9375,
|
|
"calib/format_rate": 0.88671875,
|
|
"calib/frac_conf_gt_0.9": 0.7208333333333333,
|
|
"calib/gap": 0.04074383301707796,
|
|
"calib/mean_conf": 0.9233333333333332,
|
|
"calib/mu_c": 0.9496470588235295,
|
|
"calib/mu_w": 0.9089032258064516,
|
|
"calib/nonempty_final_conf_rate": 0.9375,
|
|
"calib/nonempty_reasoning_rate": 0.953125,
|
|
"calib/nonempty_step_conf_rate": 0.921875,
|
|
"calib/pce": 0.5691666666666667,
|
|
"calib/std_conf": 0.09852946541799339,
|
|
"calib/step_conf_rate": 0.921875,
|
|
"calib/step_q_c": 0.8680975609756097,
|
|
"calib/step_q_c_n": 410.0,
|
|
"calib/step_q_gap": 0.022238372431456987,
|
|
"calib/step_q_w": 0.8458591885441528,
|
|
"calib/step_q_w_n": 838.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 898.0,
|
|
"completions/max_terminated_length": 898.0,
|
|
"completions/mean_length": 238.45703125,
|
|
"completions/mean_terminated_length": 239.3921661376953,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.0576,
|
|
"grad_norm": 20.548439025878906,
|
|
"learning_rate": 9.166666666666665e-07,
|
|
"loss": -0.0271,
|
|
"num_tokens": 12141415.0,
|
|
"reward": 0.779296875,
|
|
"reward_std": 0.4399545192718506,
|
|
"rewards/accuracy_reward_step": 0.3359375,
|
|
"rewards/format_reward_step": 0.88671875,
|
|
"step": 36
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.9296875,
|
|
"calib/avg_num_step_conf": 4.98046875,
|
|
"calib/ece": 0.644792531120332,
|
|
"calib/final_conf_rate": 0.94140625,
|
|
"calib/format_rate": 0.8671875,
|
|
"calib/frac_conf_gt_0.9": 0.7053941908713693,
|
|
"calib/gap": 0.04945061188811184,
|
|
"calib/mean_conf": 0.9145020746887966,
|
|
"calib/mu_c": 0.9506153846153846,
|
|
"calib/mu_w": 0.9011647727272728,
|
|
"calib/nonempty_final_conf_rate": 0.94140625,
|
|
"calib/nonempty_reasoning_rate": 0.95703125,
|
|
"calib/nonempty_step_conf_rate": 0.91796875,
|
|
"calib/pce": 0.644792531120332,
|
|
"calib/std_conf": 0.1280647568455238,
|
|
"calib/step_conf_rate": 0.91796875,
|
|
"calib/step_q_c": 0.8601038062283737,
|
|
"calib/step_q_c_n": 289.0,
|
|
"calib/step_q_gap": 0.0183695263095095,
|
|
"calib/step_q_w": 0.8417342799188642,
|
|
"calib/step_q_w_n": 986.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 845.0,
|
|
"completions/max_terminated_length": 845.0,
|
|
"completions/mean_length": 247.51953125,
|
|
"completions/mean_terminated_length": 249.468505859375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.0592,
|
|
"grad_norm": 16.210430145263672,
|
|
"learning_rate": 9.11111111111111e-07,
|
|
"loss": -0.0462,
|
|
"num_tokens": 12466204.0,
|
|
"reward": 0.6875,
|
|
"reward_std": 0.4034990072250366,
|
|
"rewards/accuracy_reward_step": 0.25390625,
|
|
"rewards/format_reward_step": 0.8671875,
|
|
"step": 37
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.9296875,
|
|
"calib/avg_num_step_conf": 5.1640625,
|
|
"calib/ece": 0.6612863070539419,
|
|
"calib/final_conf_rate": 0.94140625,
|
|
"calib/format_rate": 0.875,
|
|
"calib/frac_conf_gt_0.9": 0.7012448132780082,
|
|
"calib/gap": 0.04826795580110521,
|
|
"calib/mean_conf": 0.9102489626556016,
|
|
"calib/mu_c": 0.9465000000000001,
|
|
"calib/mu_w": 0.8982320441988949,
|
|
"calib/nonempty_final_conf_rate": 0.94140625,
|
|
"calib/nonempty_reasoning_rate": 0.95703125,
|
|
"calib/nonempty_step_conf_rate": 0.9296875,
|
|
"calib/pce": 0.6612863070539419,
|
|
"calib/std_conf": 0.13660958488150282,
|
|
"calib/step_conf_rate": 0.9296875,
|
|
"calib/step_q_c": 0.84876582278481,
|
|
"calib/step_q_c_n": 316.0,
|
|
"calib/step_q_gap": 0.006747930140674763,
|
|
"calib/step_q_w": 0.8420178926441353,
|
|
"calib/step_q_w_n": 1006.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 583.0,
|
|
"completions/max_terminated_length": 583.0,
|
|
"completions/mean_length": 237.515625,
|
|
"completions/mean_terminated_length": 239.3858184814453,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.0608,
|
|
"grad_norm": 10.232465744018555,
|
|
"learning_rate": 9.055555555555556e-07,
|
|
"loss": -0.0463,
|
|
"num_tokens": 12789000.0,
|
|
"reward": 0.671875,
|
|
"reward_std": 0.30752283334732056,
|
|
"rewards/accuracy_reward_step": 0.234375,
|
|
"rewards/format_reward_step": 0.875,
|
|
"step": 38
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.94140625,
|
|
"calib/avg_num_step_conf": 5.71875,
|
|
"calib/ece": 0.7620164609053499,
|
|
"calib/final_conf_rate": 0.94921875,
|
|
"calib/format_rate": 0.88671875,
|
|
"calib/frac_conf_gt_0.9": 0.6337448559670782,
|
|
"calib/gap": 0.04123720808186837,
|
|
"calib/mean_conf": 0.9109876543209877,
|
|
"calib/mu_c": 0.9459459459459462,
|
|
"calib/mu_w": 0.9047087378640778,
|
|
"calib/nonempty_final_conf_rate": 0.94921875,
|
|
"calib/nonempty_reasoning_rate": 0.96875,
|
|
"calib/nonempty_step_conf_rate": 0.93359375,
|
|
"calib/pce": 0.7603703703703705,
|
|
"calib/std_conf": 0.11771422049891216,
|
|
"calib/step_conf_rate": 0.93359375,
|
|
"calib/step_q_c": 0.8639444444444444,
|
|
"calib/step_q_c_n": 180.0,
|
|
"calib/step_q_gap": 0.023850913657227024,
|
|
"calib/step_q_w": 0.8400935307872174,
|
|
"calib/step_q_w_n": 1283.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 800.0,
|
|
"completions/max_terminated_length": 800.0,
|
|
"completions/mean_length": 272.58203125,
|
|
"completions/mean_terminated_length": 274.72833251953125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0624,
|
|
"grad_norm": 63.97758483886719,
|
|
"learning_rate": 9e-07,
|
|
"loss": -0.0187,
|
|
"num_tokens": 13119877.0,
|
|
"reward": 0.591796875,
|
|
"reward_std": 0.2877320349216461,
|
|
"rewards/accuracy_reward_step": 0.1484375,
|
|
"rewards/format_reward_step": 0.88671875,
|
|
"step": 39
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.92578125,
|
|
"calib/avg_num_step_conf": 5.28125,
|
|
"calib/ece": 0.6455833333333334,
|
|
"calib/final_conf_rate": 0.9375,
|
|
"calib/format_rate": 0.85546875,
|
|
"calib/frac_conf_gt_0.9": 0.6583333333333333,
|
|
"calib/gap": 0.03305120167189124,
|
|
"calib/mean_conf": 0.9205833333333333,
|
|
"calib/mu_c": 0.9445454545454546,
|
|
"calib/mu_w": 0.9114942528735633,
|
|
"calib/nonempty_final_conf_rate": 0.9375,
|
|
"calib/nonempty_reasoning_rate": 0.9609375,
|
|
"calib/nonempty_step_conf_rate": 0.91796875,
|
|
"calib/pce": 0.6455833333333334,
|
|
"calib/std_conf": 0.09383758871345509,
|
|
"calib/step_conf_rate": 0.91796875,
|
|
"calib/step_q_c": 0.8645266272189348,
|
|
"calib/step_q_c_n": 338.0,
|
|
"calib/step_q_gap": 0.01657790927021685,
|
|
"calib/step_q_w": 0.847948717948718,
|
|
"calib/step_q_w_n": 1014.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 1705.0,
|
|
"completions/max_terminated_length": 1705.0,
|
|
"completions/mean_length": 262.23046875,
|
|
"completions/mean_terminated_length": 263.25885009765625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.064,
|
|
"grad_norm": 25.175500869750977,
|
|
"learning_rate": 8.944444444444445e-07,
|
|
"loss": -0.0202,
|
|
"num_tokens": 13449152.0,
|
|
"reward": 0.685546875,
|
|
"reward_std": 0.3665532171726227,
|
|
"rewards/accuracy_reward_step": 0.2578125,
|
|
"rewards/format_reward_step": 0.85546875,
|
|
"step": 40
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.9375,
|
|
"calib/avg_num_step_conf": 5.37890625,
|
|
"calib/ece": 0.7596265145228215,
|
|
"calib/final_conf_rate": 0.94140625,
|
|
"calib/format_rate": 0.8984375,
|
|
"calib/frac_conf_gt_0.9": 0.7427385892116183,
|
|
"calib/gap": 0.02382273510409194,
|
|
"calib/mean_conf": 0.9339003734439835,
|
|
"calib/mu_c": 0.9535714285714286,
|
|
"calib/mu_w": 0.9297486934673367,
|
|
"calib/nonempty_final_conf_rate": 0.94140625,
|
|
"calib/nonempty_reasoning_rate": 0.95703125,
|
|
"calib/nonempty_step_conf_rate": 0.9375,
|
|
"calib/pce": 0.7596265145228215,
|
|
"calib/std_conf": 0.06590362283175459,
|
|
"calib/step_conf_rate": 0.9375,
|
|
"calib/step_q_c": 0.8828634361233482,
|
|
"calib/step_q_c_n": 227.0,
|
|
"calib/step_q_gap": 0.030634436123348152,
|
|
"calib/step_q_w": 0.852229,
|
|
"calib/step_q_w_n": 1150.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 590.0,
|
|
"completions/max_terminated_length": 590.0,
|
|
"completions/mean_length": 244.40234375,
|
|
"completions/mean_terminated_length": 246.32676696777344,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.0656,
|
|
"grad_norm": 14.0829496383667,
|
|
"learning_rate": 8.888888888888888e-07,
|
|
"loss": -0.0629,
|
|
"num_tokens": 13773863.0,
|
|
"reward": 0.61328125,
|
|
"reward_std": 0.25447797775268555,
|
|
"rewards/accuracy_reward_step": 0.1640625,
|
|
"rewards/format_reward_step": 0.8984375,
|
|
"step": 41
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.953125,
|
|
"calib/avg_num_step_conf": 5.6640625,
|
|
"calib/ece": 0.5372064777327934,
|
|
"calib/final_conf_rate": 0.96484375,
|
|
"calib/format_rate": 0.91015625,
|
|
"calib/frac_conf_gt_0.9": 0.6923076923076923,
|
|
"calib/gap": 0.021238272626931343,
|
|
"calib/mean_conf": 0.9258704453441295,
|
|
"calib/mu_c": 0.9388541666666667,
|
|
"calib/mu_w": 0.9176158940397353,
|
|
"calib/nonempty_final_conf_rate": 0.96484375,
|
|
"calib/nonempty_reasoning_rate": 0.96484375,
|
|
"calib/nonempty_step_conf_rate": 0.94140625,
|
|
"calib/pce": 0.5372064777327934,
|
|
"calib/std_conf": 0.08506220287114333,
|
|
"calib/step_conf_rate": 0.94140625,
|
|
"calib/step_q_c": 0.8620292504570385,
|
|
"calib/step_q_c_n": 547.0,
|
|
"calib/step_q_gap": 0.0074999038346685865,
|
|
"calib/step_q_w": 0.8545293466223699,
|
|
"calib/step_q_w_n": 903.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 779.0,
|
|
"completions/max_terminated_length": 779.0,
|
|
"completions/mean_length": 261.890625,
|
|
"completions/mean_terminated_length": 263.9527587890625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.0672,
|
|
"grad_norm": 17.002763748168945,
|
|
"learning_rate": 8.833333333333333e-07,
|
|
"loss": -0.0442,
|
|
"num_tokens": 14099395.0,
|
|
"reward": 0.830078125,
|
|
"reward_std": 0.3919033706188202,
|
|
"rewards/accuracy_reward_step": 0.375,
|
|
"rewards/format_reward_step": 0.91015625,
|
|
"step": 42
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.94921875,
|
|
"calib/avg_num_step_conf": 6.32421875,
|
|
"calib/ece": 0.6638367346938774,
|
|
"calib/final_conf_rate": 0.95703125,
|
|
"calib/format_rate": 0.88671875,
|
|
"calib/frac_conf_gt_0.9": 0.7061224489795919,
|
|
"calib/gap": 0.029949930939226488,
|
|
"calib/mean_conf": 0.9250612244897959,
|
|
"calib/mu_c": 0.9471875000000001,
|
|
"calib/mu_w": 0.9172375690607736,
|
|
"calib/nonempty_final_conf_rate": 0.95703125,
|
|
"calib/nonempty_reasoning_rate": 0.9765625,
|
|
"calib/nonempty_step_conf_rate": 0.953125,
|
|
"calib/pce": 0.6638367346938774,
|
|
"calib/std_conf": 0.09329181956903092,
|
|
"calib/step_conf_rate": 0.953125,
|
|
"calib/step_q_c": 0.858123324396783,
|
|
"calib/step_q_c_n": 373.0,
|
|
"calib/step_q_gap": 0.01197565184461602,
|
|
"calib/step_q_w": 0.8461476725521669,
|
|
"calib/step_q_w_n": 1246.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 2620.0,
|
|
"completions/max_terminated_length": 2620.0,
|
|
"completions/mean_length": 312.546875,
|
|
"completions/mean_terminated_length": 313.7725524902344,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.0688,
|
|
"grad_norm": 7.766086578369141,
|
|
"learning_rate": 8.777777777777777e-07,
|
|
"loss": -0.0113,
|
|
"num_tokens": 14441551.0,
|
|
"reward": 0.693359375,
|
|
"reward_std": 0.30214864015579224,
|
|
"rewards/accuracy_reward_step": 0.25,
|
|
"rewards/format_reward_step": 0.88671875,
|
|
"step": 43
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.921875,
|
|
"calib/avg_num_step_conf": 7.21875,
|
|
"calib/ece": 0.6732780082987552,
|
|
"calib/final_conf_rate": 0.94140625,
|
|
"calib/format_rate": 0.87890625,
|
|
"calib/frac_conf_gt_0.9": 0.6804979253112033,
|
|
"calib/gap": 0.03413479052823298,
|
|
"calib/mean_conf": 0.9197510373443984,
|
|
"calib/mu_c": 0.9452459016393441,
|
|
"calib/mu_w": 0.9111111111111111,
|
|
"calib/nonempty_final_conf_rate": 0.94140625,
|
|
"calib/nonempty_reasoning_rate": 0.9609375,
|
|
"calib/nonempty_step_conf_rate": 0.94140625,
|
|
"calib/pce": 0.6699585062240664,
|
|
"calib/std_conf": 0.10817584160242431,
|
|
"calib/step_conf_rate": 0.94140625,
|
|
"calib/step_q_c": 0.8699236641221374,
|
|
"calib/step_q_c_n": 393.0,
|
|
"calib/step_q_gap": 0.0193394716822749,
|
|
"calib/step_q_w": 0.8505841924398625,
|
|
"calib/step_q_w_n": 1455.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 1817.0,
|
|
"completions/max_terminated_length": 1817.0,
|
|
"completions/mean_length": 301.06640625,
|
|
"completions/mean_terminated_length": 303.43701171875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0704,
|
|
"grad_norm": 22.831974029541016,
|
|
"learning_rate": 8.722222222222222e-07,
|
|
"loss": -0.0944,
|
|
"num_tokens": 14778840.0,
|
|
"reward": 0.681640625,
|
|
"reward_std": 0.35068172216415405,
|
|
"rewards/accuracy_reward_step": 0.2421875,
|
|
"rewards/format_reward_step": 0.87890625,
|
|
"step": 44
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.90234375,
|
|
"calib/avg_num_step_conf": 7.21875,
|
|
"calib/ece": 0.6819246861924687,
|
|
"calib/final_conf_rate": 0.93359375,
|
|
"calib/format_rate": 0.83203125,
|
|
"calib/frac_conf_gt_0.9": 0.6569037656903766,
|
|
"calib/gap": 0.02893930523028876,
|
|
"calib/mean_conf": 0.916234309623431,
|
|
"calib/mu_c": 0.9383928571428571,
|
|
"calib/mu_w": 0.9094535519125684,
|
|
"calib/nonempty_final_conf_rate": 0.93359375,
|
|
"calib/nonempty_reasoning_rate": 0.96875,
|
|
"calib/nonempty_step_conf_rate": 0.9453125,
|
|
"calib/pce": 0.6819246861924687,
|
|
"calib/std_conf": 0.10689240824307658,
|
|
"calib/step_conf_rate": 0.9453125,
|
|
"calib/step_q_c": 0.8586650485436893,
|
|
"calib/step_q_c_n": 412.0,
|
|
"calib/step_q_gap": 0.028442207318062573,
|
|
"calib/step_q_w": 0.8302228412256267,
|
|
"calib/step_q_w_n": 1436.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2664.0,
|
|
"completions/max_terminated_length": 2664.0,
|
|
"completions/mean_length": 333.71875,
|
|
"completions/mean_terminated_length": 336.3464660644531,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.072,
|
|
"grad_norm": 33.78303527832031,
|
|
"learning_rate": 8.666666666666667e-07,
|
|
"loss": -0.0129,
|
|
"num_tokens": 15119352.0,
|
|
"reward": 0.634765625,
|
|
"reward_std": 0.34433940052986145,
|
|
"rewards/accuracy_reward_step": 0.21875,
|
|
"rewards/format_reward_step": 0.83203125,
|
|
"step": 45
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.78125,
|
|
"calib/avg_num_step_conf": 8.2578125,
|
|
"calib/ece": 0.7925333333333335,
|
|
"calib/final_conf_rate": 0.87890625,
|
|
"calib/format_rate": 0.71484375,
|
|
"calib/frac_conf_gt_0.9": 0.6355555555555555,
|
|
"calib/gap": 0.008950000000000014,
|
|
"calib/mean_conf": 0.9036444444444444,
|
|
"calib/mu_c": 0.9116,
|
|
"calib/mu_w": 0.90265,
|
|
"calib/nonempty_final_conf_rate": 0.87890625,
|
|
"calib/nonempty_reasoning_rate": 0.921875,
|
|
"calib/nonempty_step_conf_rate": 0.89453125,
|
|
"calib/pce": 0.7925333333333335,
|
|
"calib/std_conf": 0.11397488135658186,
|
|
"calib/step_conf_rate": 0.89453125,
|
|
"calib/step_q_c": 0.8361176470588234,
|
|
"calib/step_q_c_n": 170.0,
|
|
"calib/step_q_gap": 0.022300774630839904,
|
|
"calib/step_q_w": 0.8138168724279835,
|
|
"calib/step_q_w_n": 1944.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2833.0,
|
|
"completions/max_terminated_length": 2833.0,
|
|
"completions/mean_length": 390.0625,
|
|
"completions/mean_terminated_length": 393.13385009765625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.0736,
|
|
"grad_norm": 54.98508071899414,
|
|
"learning_rate": 8.611111111111111e-07,
|
|
"loss": -0.0464,
|
|
"num_tokens": 15475712.0,
|
|
"reward": 0.458984375,
|
|
"reward_std": 0.34532803297042847,
|
|
"rewards/accuracy_reward_step": 0.1015625,
|
|
"rewards/format_reward_step": 0.71484375,
|
|
"step": 46
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.51953125,
|
|
"calib/avg_num_step_conf": 8.9296875,
|
|
"calib/ece": 0.8392434210526315,
|
|
"calib/final_conf_rate": 0.59375,
|
|
"calib/format_rate": 0.40625,
|
|
"calib/frac_conf_gt_0.9": 0.6118421052631579,
|
|
"calib/gap": 0.013880281690141105,
|
|
"calib/mean_conf": 0.9050328947368422,
|
|
"calib/mu_c": 0.9180000000000001,
|
|
"calib/mu_w": 0.904119718309859,
|
|
"calib/nonempty_final_conf_rate": 0.59375,
|
|
"calib/nonempty_reasoning_rate": 0.73046875,
|
|
"calib/nonempty_step_conf_rate": 0.69921875,
|
|
"calib/pce": 0.8392434210526315,
|
|
"calib/std_conf": 0.1176657519463254,
|
|
"calib/step_conf_rate": 0.69921875,
|
|
"calib/step_q_c": 0.8893055555555556,
|
|
"calib/step_q_c_n": 72.0,
|
|
"calib/step_q_gap": 0.09709462511291789,
|
|
"calib/step_q_w": 0.7922109304426377,
|
|
"calib/step_q_w_n": 2214.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2930.0,
|
|
"completions/max_terminated_length": 2930.0,
|
|
"completions/mean_length": 585.80859375,
|
|
"completions/mean_terminated_length": 590.4212646484375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.0752,
|
|
"grad_norm": 21.259794235229492,
|
|
"learning_rate": 8.555555555555555e-07,
|
|
"loss": -0.0241,
|
|
"num_tokens": 15886935.0,
|
|
"reward": 0.24609375,
|
|
"reward_std": 0.3053395748138428,
|
|
"rewards/accuracy_reward_step": 0.04296875,
|
|
"rewards/format_reward_step": 0.40625,
|
|
"step": 47
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.1640625,
|
|
"calib/avg_num_step_conf": 2.48046875,
|
|
"calib/ece": 0.8633076923076923,
|
|
"calib/final_conf_rate": 0.15234375,
|
|
"calib/format_rate": 0.09765625,
|
|
"calib/frac_conf_gt_0.9": 0.6153846153846154,
|
|
"calib/gap": 0.005702702702702567,
|
|
"calib/mean_conf": 0.9145897435897435,
|
|
"calib/mu_c": 0.9199999999999999,
|
|
"calib/mu_w": 0.9142972972972974,
|
|
"calib/nonempty_final_conf_rate": 0.15234375,
|
|
"calib/nonempty_reasoning_rate": 0.29296875,
|
|
"calib/nonempty_step_conf_rate": 0.26953125,
|
|
"calib/pce": 0.8633076923076923,
|
|
"calib/std_conf": 0.07260702726063477,
|
|
"calib/step_conf_rate": 0.26953125,
|
|
"calib/step_q_c": 0.76,
|
|
"calib/step_q_c_n": 10.0,
|
|
"calib/step_q_gap": -0.0807663999999999,
|
|
"calib/step_q_w": 0.8407663999999999,
|
|
"calib/step_q_w_n": 625.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01953125,
|
|
"completions/max_length": 3006.0,
|
|
"completions/max_terminated_length": 3006.0,
|
|
"completions/mean_length": 958.6015625,
|
|
"completions/mean_terminated_length": 977.697265625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.0768,
|
|
"grad_norm": 26.788244247436523,
|
|
"learning_rate": 8.499999999999999e-07,
|
|
"loss": 0.0328,
|
|
"num_tokens": 16386185.0,
|
|
"reward": 0.060546875,
|
|
"reward_std": 0.12249716371297836,
|
|
"rewards/accuracy_reward_step": 0.01171875,
|
|
"rewards/format_reward_step": 0.09765625,
|
|
"step": 48
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0234375,
|
|
"calib/avg_num_step_conf": 0.984375,
|
|
"calib/ece": 0.9020000000000001,
|
|
"calib/final_conf_rate": 0.01953125,
|
|
"calib/format_rate": 0.015625,
|
|
"calib/frac_conf_gt_0.9": 0.8,
|
|
"calib/mean_conf": 0.9020000000000001,
|
|
"calib/mu_c": NaN,
|
|
"calib/mu_w": 0.9020000000000001,
|
|
"calib/nonempty_final_conf_rate": 0.01953125,
|
|
"calib/nonempty_reasoning_rate": 0.0859375,
|
|
"calib/nonempty_step_conf_rate": 0.078125,
|
|
"calib/pce": 0.9020000000000001,
|
|
"calib/std_conf": 0.12734205903785284,
|
|
"calib/step_conf_rate": 0.078125,
|
|
"calib/step_q_w": 0.7507420634920635,
|
|
"calib/step_q_w_n": 252.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.02734375,
|
|
"completions/max_length": 2898.0,
|
|
"completions/max_terminated_length": 2898.0,
|
|
"completions/mean_length": 1243.87890625,
|
|
"completions/mean_terminated_length": 1278.8472900390625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.0784,
|
|
"grad_norm": 5.6303815841674805,
|
|
"learning_rate": 8.444444444444444e-07,
|
|
"loss": -0.0025,
|
|
"num_tokens": 16963690.0,
|
|
"reward": 0.0078125,
|
|
"reward_std": 0.018281511962413788,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.015625,
|
|
"step": 49
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.03515625,
|
|
"calib/avg_num_step_conf": 0.49609375,
|
|
"calib/ece": 0.8109999999999999,
|
|
"calib/final_conf_rate": 0.01953125,
|
|
"calib/format_rate": 0.0078125,
|
|
"calib/frac_conf_gt_0.9": 0.2,
|
|
"calib/mean_conf": 0.8109999999999999,
|
|
"calib/mu_c": NaN,
|
|
"calib/mu_w": 0.8109999999999999,
|
|
"calib/nonempty_final_conf_rate": 0.01953125,
|
|
"calib/nonempty_reasoning_rate": 0.0625,
|
|
"calib/nonempty_step_conf_rate": 0.046875,
|
|
"calib/pce": 0.8109999999999999,
|
|
"calib/std_conf": 0.11226753760548948,
|
|
"calib/step_conf_rate": 0.046875,
|
|
"calib/step_q_w": 0.7346020997375329,
|
|
"calib/step_q_w_n": 127.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.015625,
|
|
"completions/max_length": 3000.0,
|
|
"completions/max_terminated_length": 3000.0,
|
|
"completions/mean_length": 1146.8671875,
|
|
"completions/mean_terminated_length": 1165.071533203125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.08,
|
|
"grad_norm": 0.5443186163902283,
|
|
"learning_rate": 8.388888888888888e-07,
|
|
"loss": 0.0011,
|
|
"num_tokens": 17519288.0,
|
|
"reward": 0.00390625,
|
|
"reward_std": 0.007232969626784325,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0078125,
|
|
"step": 50
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0390625,
|
|
"calib/avg_num_step_conf": 0.5859375,
|
|
"calib/ece": 0.7766666666666667,
|
|
"calib/final_conf_rate": 0.046875,
|
|
"calib/format_rate": 0.02734375,
|
|
"calib/frac_conf_gt_0.9": 0.5,
|
|
"calib/mean_conf": 0.7766666666666667,
|
|
"calib/mu_c": NaN,
|
|
"calib/mu_w": 0.7766666666666667,
|
|
"calib/nonempty_final_conf_rate": 0.046875,
|
|
"calib/nonempty_reasoning_rate": 0.11328125,
|
|
"calib/nonempty_step_conf_rate": 0.10546875,
|
|
"calib/pce": 0.7766666666666667,
|
|
"calib/std_conf": 0.2907843798341918,
|
|
"calib/step_conf_rate": 0.10546875,
|
|
"calib/step_q_w": 0.8009333333333334,
|
|
"calib/step_q_w_n": 150.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.015625,
|
|
"completions/max_length": 3005.0,
|
|
"completions/max_terminated_length": 3005.0,
|
|
"completions/mean_length": 1167.09765625,
|
|
"completions/mean_terminated_length": 1185.623046875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 3.0,
|
|
"epoch": 0.0816,
|
|
"grad_norm": 13.37551498413086,
|
|
"learning_rate": 8.333333333333333e-07,
|
|
"loss": 0.0419,
|
|
"num_tokens": 18075633.0,
|
|
"reward": 0.013671875,
|
|
"reward_std": 0.03485432639718056,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.02734375,
|
|
"step": 51
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.03125,
|
|
"calib/avg_num_step_conf": 0.27734375,
|
|
"calib/ece": 0.7934549999999999,
|
|
"calib/final_conf_rate": 0.0234375,
|
|
"calib/format_rate": 0.00390625,
|
|
"calib/frac_conf_gt_0.9": 0.3333333333333333,
|
|
"calib/mean_conf": 0.7934549999999999,
|
|
"calib/mu_c": NaN,
|
|
"calib/mu_w": 0.7934549999999999,
|
|
"calib/nonempty_final_conf_rate": 0.0234375,
|
|
"calib/nonempty_reasoning_rate": 0.0703125,
|
|
"calib/nonempty_step_conf_rate": 0.046875,
|
|
"calib/pce": 0.7934549999999999,
|
|
"calib/std_conf": 0.25778263154254594,
|
|
"calib/step_conf_rate": 0.046875,
|
|
"calib/step_q_w": 0.771830985915493,
|
|
"calib/step_q_w_n": 71.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 3034.0,
|
|
"completions/max_terminated_length": 3034.0,
|
|
"completions/mean_length": 1342.6015625,
|
|
"completions/mean_terminated_length": 1353.1732177734375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.0832,
|
|
"grad_norm": 0.47544148564338684,
|
|
"learning_rate": 8.277777777777777e-07,
|
|
"loss": 0.0118,
|
|
"num_tokens": 18681411.0,
|
|
"reward": 0.001953125,
|
|
"reward_std": 0.005524271633476019,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.00390625,
|
|
"step": 52
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0078125,
|
|
"calib/avg_num_step_conf": 0.15234375,
|
|
"calib/ece": 0.98,
|
|
"calib/final_conf_rate": 0.00390625,
|
|
"calib/format_rate": 0.00390625,
|
|
"calib/frac_conf_gt_0.9": 1.0,
|
|
"calib/mean_conf": 0.98,
|
|
"calib/mu_c": NaN,
|
|
"calib/mu_w": 0.98,
|
|
"calib/nonempty_final_conf_rate": 0.00390625,
|
|
"calib/nonempty_reasoning_rate": 0.04296875,
|
|
"calib/nonempty_step_conf_rate": 0.04296875,
|
|
"calib/pce": 0.98,
|
|
"calib/std_conf": 0.0,
|
|
"calib/step_conf_rate": 0.04296875,
|
|
"calib/step_q_w": 0.7835897435897436,
|
|
"calib/step_q_w_n": 39.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.02734375,
|
|
"completions/max_length": 2905.0,
|
|
"completions/max_terminated_length": 2905.0,
|
|
"completions/mean_length": 1348.04296875,
|
|
"completions/mean_terminated_length": 1385.939697265625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.0848,
|
|
"grad_norm": 0.9600926637649536,
|
|
"learning_rate": 8.222222222222221e-07,
|
|
"loss": 0.0017,
|
|
"num_tokens": 19288406.0,
|
|
"reward": 0.001953125,
|
|
"reward_std": 0.005524271633476019,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.00390625,
|
|
"step": 53
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.03125,
|
|
"calib/avg_num_step_conf": 0.328125,
|
|
"calib/ece": 0.8422222222222223,
|
|
"calib/final_conf_rate": 0.03515625,
|
|
"calib/format_rate": 0.015625,
|
|
"calib/frac_conf_gt_0.9": 0.5555555555555556,
|
|
"calib/mean_conf": 0.8422222222222222,
|
|
"calib/mu_c": NaN,
|
|
"calib/mu_w": 0.8422222222222222,
|
|
"calib/nonempty_final_conf_rate": 0.03515625,
|
|
"calib/nonempty_reasoning_rate": 0.09765625,
|
|
"calib/nonempty_step_conf_rate": 0.09375,
|
|
"calib/pce": 0.8422222222222223,
|
|
"calib/std_conf": 0.15324474091231954,
|
|
"calib/step_conf_rate": 0.09375,
|
|
"calib/step_q_w": 0.8053571428571425,
|
|
"calib/step_q_w_n": 84.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.03515625,
|
|
"completions/max_length": 2953.0,
|
|
"completions/max_terminated_length": 2953.0,
|
|
"completions/mean_length": 1340.8671875,
|
|
"completions/mean_terminated_length": 1389.7247314453125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 5.0,
|
|
"epoch": 0.0864,
|
|
"grad_norm": 1.3612425327301025,
|
|
"learning_rate": 8.166666666666666e-07,
|
|
"loss": -0.0042,
|
|
"num_tokens": 19893140.0,
|
|
"reward": 0.0078125,
|
|
"reward_std": 0.018281511962413788,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.015625,
|
|
"step": 54
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0078125,
|
|
"calib/avg_num_step_conf": 0.06640625,
|
|
"calib/ece": 0.9,
|
|
"calib/final_conf_rate": 0.00390625,
|
|
"calib/format_rate": 0.0,
|
|
"calib/frac_conf_gt_0.9": 0.0,
|
|
"calib/mean_conf": 0.9,
|
|
"calib/mu_c": NaN,
|
|
"calib/mu_w": 0.9,
|
|
"calib/nonempty_final_conf_rate": 0.00390625,
|
|
"calib/nonempty_reasoning_rate": 0.0390625,
|
|
"calib/nonempty_step_conf_rate": 0.03515625,
|
|
"calib/pce": 0.9,
|
|
"calib/std_conf": 0.0,
|
|
"calib/step_conf_rate": 0.03515625,
|
|
"calib/step_q_w": 0.7994117647058824,
|
|
"calib/step_q_w_n": 17.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0234375,
|
|
"completions/max_length": 2943.0,
|
|
"completions/max_terminated_length": 2943.0,
|
|
"completions/mean_length": 1346.0625,
|
|
"completions/mean_terminated_length": 1378.3680419921875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 3.0,
|
|
"epoch": 0.088,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 8.11111111111111e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 20499572.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.03125,
|
|
"calib/ece": 0.75,
|
|
"calib/final_conf_rate": 0.00390625,
|
|
"calib/format_rate": 0.0,
|
|
"calib/frac_conf_gt_0.9": 0.0,
|
|
"calib/mean_conf": 0.75,
|
|
"calib/mu_c": NaN,
|
|
"calib/mu_w": 0.75,
|
|
"calib/nonempty_final_conf_rate": 0.00390625,
|
|
"calib/nonempty_reasoning_rate": 0.0078125,
|
|
"calib/nonempty_step_conf_rate": 0.0078125,
|
|
"calib/pce": 0.75,
|
|
"calib/std_conf": 0.0,
|
|
"calib/step_conf_rate": 0.0078125,
|
|
"calib/step_q_w": 0.577815,
|
|
"calib/step_q_w_n": 8.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2960.0,
|
|
"completions/max_terminated_length": 2960.0,
|
|
"completions/mean_length": 1211.46484375,
|
|
"completions/mean_terminated_length": 1221.00390625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 3.0,
|
|
"epoch": 0.0896,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 8.055555555555556e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 21071851.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 56
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.015625,
|
|
"calib/avg_num_step_conf": 0.09375,
|
|
"calib/ece": 0.8666666666666667,
|
|
"calib/final_conf_rate": 0.01171875,
|
|
"calib/format_rate": 0.0,
|
|
"calib/frac_conf_gt_0.9": 0.3333333333333333,
|
|
"calib/mean_conf": 0.8666666666666667,
|
|
"calib/mu_c": NaN,
|
|
"calib/mu_w": 0.8666666666666667,
|
|
"calib/nonempty_final_conf_rate": 0.01171875,
|
|
"calib/nonempty_reasoning_rate": 0.0546875,
|
|
"calib/nonempty_step_conf_rate": 0.046875,
|
|
"calib/pce": 0.8666666666666667,
|
|
"calib/std_conf": 0.08498365855987973,
|
|
"calib/step_conf_rate": 0.046875,
|
|
"calib/step_q_w": 0.7087499999999999,
|
|
"calib/step_q_w_n": 24.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2996.0,
|
|
"completions/max_terminated_length": 2996.0,
|
|
"completions/mean_length": 1441.05078125,
|
|
"completions/mean_terminated_length": 1458.138427734375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 9.0,
|
|
"epoch": 0.0912,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 8e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 21697264.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 57
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.00390625,
|
|
"calib/avg_num_step_conf": 0.00390625,
|
|
"calib/ece": 0.89,
|
|
"calib/final_conf_rate": 0.00390625,
|
|
"calib/format_rate": 0.00390625,
|
|
"calib/frac_conf_gt_0.9": 0.0,
|
|
"calib/mean_conf": 0.89,
|
|
"calib/mu_c": NaN,
|
|
"calib/mu_w": 0.89,
|
|
"calib/nonempty_final_conf_rate": 0.00390625,
|
|
"calib/nonempty_reasoning_rate": 0.00390625,
|
|
"calib/nonempty_step_conf_rate": 0.00390625,
|
|
"calib/pce": 0.89,
|
|
"calib/std_conf": 0.0,
|
|
"calib/step_conf_rate": 0.00390625,
|
|
"calib/step_q_w": 0.83,
|
|
"calib/step_q_w_n": 1.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 3042.0,
|
|
"completions/max_terminated_length": 3042.0,
|
|
"completions/mean_length": 1447.109375,
|
|
"completions/mean_terminated_length": 1464.268798828125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 5.0,
|
|
"epoch": 0.0928,
|
|
"grad_norm": 0.07886432111263275,
|
|
"learning_rate": 7.944444444444444e-07,
|
|
"loss": 0.0072,
|
|
"num_tokens": 22327828.0,
|
|
"reward": 0.001953125,
|
|
"reward_std": 0.005524271633476019,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.00390625,
|
|
"step": 58
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2983.0,
|
|
"completions/max_terminated_length": 2983.0,
|
|
"completions/mean_length": 1470.5390625,
|
|
"completions/mean_terminated_length": 1487.976318359375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 4.0,
|
|
"epoch": 0.0944,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 7.888888888888889e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 22966430.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 59
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.00390625,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/ece": 0.982,
|
|
"calib/final_conf_rate": 0.00390625,
|
|
"calib/format_rate": 0.0,
|
|
"calib/frac_conf_gt_0.9": 1.0,
|
|
"calib/mean_conf": 0.982,
|
|
"calib/mu_c": NaN,
|
|
"calib/mu_w": 0.982,
|
|
"calib/nonempty_final_conf_rate": 0.00390625,
|
|
"calib/nonempty_reasoning_rate": 0.00390625,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/pce": 0.982,
|
|
"calib/std_conf": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.015625,
|
|
"completions/max_length": 2929.0,
|
|
"completions/max_terminated_length": 2929.0,
|
|
"completions/mean_length": 1242.18359375,
|
|
"completions/mean_terminated_length": 1261.90087890625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 7.0,
|
|
"epoch": 0.096,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 7.833333333333333e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 23546573.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.015625,
|
|
"calib/avg_num_step_conf": 0.05078125,
|
|
"calib/ece": 0.9033333333333333,
|
|
"calib/final_conf_rate": 0.01171875,
|
|
"calib/format_rate": 0.0078125,
|
|
"calib/frac_conf_gt_0.9": 0.6666666666666666,
|
|
"calib/mean_conf": 0.9033333333333333,
|
|
"calib/mu_c": NaN,
|
|
"calib/mu_w": 0.9033333333333333,
|
|
"calib/nonempty_final_conf_rate": 0.01171875,
|
|
"calib/nonempty_reasoning_rate": 0.0234375,
|
|
"calib/nonempty_step_conf_rate": 0.01953125,
|
|
"calib/pce": 0.9033333333333333,
|
|
"calib/std_conf": 0.06599663291074438,
|
|
"calib/step_conf_rate": 0.01953125,
|
|
"calib/step_q_w": 0.7952777777777778,
|
|
"calib/step_q_w_n": 12.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2940.0,
|
|
"completions/max_terminated_length": 2940.0,
|
|
"completions/mean_length": 1263.74609375,
|
|
"completions/mean_terminated_length": 1273.6968994140625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 7.0,
|
|
"epoch": 0.0976,
|
|
"grad_norm": 2.4402859210968018,
|
|
"learning_rate": 7.777777777777778e-07,
|
|
"loss": 0.0114,
|
|
"num_tokens": 24132108.0,
|
|
"reward": 0.00390625,
|
|
"reward_std": 0.011048543266952038,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0078125,
|
|
"step": 61
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/ece": 0.99,
|
|
"calib/final_conf_rate": 0.00390625,
|
|
"calib/format_rate": 0.0,
|
|
"calib/frac_conf_gt_0.9": 1.0,
|
|
"calib/mean_conf": 0.99,
|
|
"calib/mu_c": NaN,
|
|
"calib/mu_w": 0.99,
|
|
"calib/nonempty_final_conf_rate": 0.00390625,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/pce": 0.99,
|
|
"calib/std_conf": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01953125,
|
|
"completions/max_length": 3004.0,
|
|
"completions/max_terminated_length": 3004.0,
|
|
"completions/mean_length": 1362.46484375,
|
|
"completions/mean_terminated_length": 1389.6055908203125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 3.0,
|
|
"epoch": 0.0992,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 7.722222222222222e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 24742483.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 62
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.02734375,
|
|
"calib/avg_num_step_conf": 0.33984375,
|
|
"calib/ece": 0.716,
|
|
"calib/final_conf_rate": 0.01953125,
|
|
"calib/format_rate": 0.00390625,
|
|
"calib/frac_conf_gt_0.9": 0.0,
|
|
"calib/mean_conf": 0.716,
|
|
"calib/mu_c": NaN,
|
|
"calib/mu_w": 0.716,
|
|
"calib/nonempty_final_conf_rate": 0.01953125,
|
|
"calib/nonempty_reasoning_rate": 0.0625,
|
|
"calib/nonempty_step_conf_rate": 0.046875,
|
|
"calib/pce": 0.716,
|
|
"calib/std_conf": 0.23711600536446292,
|
|
"calib/step_conf_rate": 0.046875,
|
|
"calib/step_q_w": 0.6549770114942529,
|
|
"calib/step_q_w_n": 87.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01953125,
|
|
"completions/max_length": 2984.0,
|
|
"completions/max_terminated_length": 2984.0,
|
|
"completions/mean_length": 1299.87890625,
|
|
"completions/mean_terminated_length": 1325.77294921875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 3.0,
|
|
"epoch": 0.1008,
|
|
"grad_norm": 3.530916929244995,
|
|
"learning_rate": 7.666666666666667e-07,
|
|
"loss": 0.0124,
|
|
"num_tokens": 25330612.0,
|
|
"reward": 0.005859375,
|
|
"reward_std": 0.01657281443476677,
|
|
"rewards/accuracy_reward_step": 0.00390625,
|
|
"rewards/format_reward_step": 0.00390625,
|
|
"step": 63
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.00390625,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.00390625,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.015625,
|
|
"completions/max_length": 3001.0,
|
|
"completions/max_terminated_length": 3001.0,
|
|
"completions/mean_length": 1257.80859375,
|
|
"completions/mean_terminated_length": 1277.77392578125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 11.0,
|
|
"epoch": 0.1024,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 7.61111111111111e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 25914755.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 64
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.01953125,
|
|
"calib/avg_num_step_conf": 0.0859375,
|
|
"calib/ece": 0.65,
|
|
"calib/final_conf_rate": 0.0078125,
|
|
"calib/format_rate": 0.00390625,
|
|
"calib/frac_conf_gt_0.9": 0.0,
|
|
"calib/mean_conf": 0.65,
|
|
"calib/mu_c": NaN,
|
|
"calib/mu_w": 0.65,
|
|
"calib/nonempty_final_conf_rate": 0.0078125,
|
|
"calib/nonempty_reasoning_rate": 0.02734375,
|
|
"calib/nonempty_step_conf_rate": 0.01171875,
|
|
"calib/pce": 0.65,
|
|
"calib/std_conf": 0.15000000000000002,
|
|
"calib/step_conf_rate": 0.01171875,
|
|
"calib/step_q_w": 0.7795121212121212,
|
|
"calib/step_q_w_n": 22.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.03125,
|
|
"completions/max_length": 2931.0,
|
|
"completions/max_terminated_length": 2931.0,
|
|
"completions/mean_length": 1273.6953125,
|
|
"completions/mean_terminated_length": 1314.7822265625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.104,
|
|
"grad_norm": 3.730851173400879,
|
|
"learning_rate": 7.555555555555555e-07,
|
|
"loss": 0.0111,
|
|
"num_tokens": 26498589.0,
|
|
"reward": 0.001953125,
|
|
"reward_std": 0.005524271633476019,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.00390625,
|
|
"step": 65
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.015625,
|
|
"calib/avg_num_step_conf": 0.015625,
|
|
"calib/ece": 0.3,
|
|
"calib/final_conf_rate": 0.00390625,
|
|
"calib/format_rate": 0.0,
|
|
"calib/frac_conf_gt_0.9": 0.0,
|
|
"calib/mean_conf": 0.3,
|
|
"calib/mu_c": NaN,
|
|
"calib/mu_w": 0.3,
|
|
"calib/nonempty_final_conf_rate": 0.00390625,
|
|
"calib/nonempty_reasoning_rate": 0.02734375,
|
|
"calib/nonempty_step_conf_rate": 0.01171875,
|
|
"calib/pce": 0.3,
|
|
"calib/std_conf": 0.0,
|
|
"calib/step_conf_rate": 0.01171875,
|
|
"calib/step_q_w": 0.5225,
|
|
"calib/step_q_w_n": 4.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01953125,
|
|
"completions/max_length": 3018.0,
|
|
"completions/max_terminated_length": 3018.0,
|
|
"completions/mean_length": 1236.9765625,
|
|
"completions/mean_terminated_length": 1261.6175537109375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.1056,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 7.5e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 27076695.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 66
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0078125,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.00390625,
|
|
"calib/nonempty_step_conf_rate": 0.00390625,
|
|
"calib/step_conf_rate": 0.00390625,
|
|
"calib/step_q_w": 1.0,
|
|
"calib/step_q_w_n": 2.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01953125,
|
|
"completions/max_length": 2984.0,
|
|
"completions/max_terminated_length": 2984.0,
|
|
"completions/mean_length": 1154.9296875,
|
|
"completions/mean_terminated_length": 1177.936279296875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.1072,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 7.444444444444444e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 27634109.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 67
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.00390625,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.00390625,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 3021.0,
|
|
"completions/max_terminated_length": 3021.0,
|
|
"completions/mean_length": 1266.05859375,
|
|
"completions/mean_terminated_length": 1271.0235595703125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 9.0,
|
|
"epoch": 0.1088,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 7.388888888888889e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 28219700.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 68
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 3004.0,
|
|
"completions/max_terminated_length": 3004.0,
|
|
"completions/mean_length": 1211.1015625,
|
|
"completions/mean_terminated_length": 1225.4625244140625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 13.0,
|
|
"epoch": 0.1104,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 7.333333333333332e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 28791558.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 69
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2970.0,
|
|
"completions/max_terminated_length": 2970.0,
|
|
"completions/mean_length": 1100.359375,
|
|
"completions/mean_terminated_length": 1109.0235595703125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 4.0,
|
|
"epoch": 0.112,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 7.277777777777777e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 29334970.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01953125,
|
|
"completions/max_length": 2980.0,
|
|
"completions/max_terminated_length": 2980.0,
|
|
"completions/mean_length": 1057.43359375,
|
|
"completions/mean_terminated_length": 1078.498046875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 4.0,
|
|
"epoch": 0.1136,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 7.222222222222221e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 29863345.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 71
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.00390625,
|
|
"calib/avg_num_step_conf": 0.01171875,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.01171875,
|
|
"calib/nonempty_step_conf_rate": 0.0078125,
|
|
"calib/step_conf_rate": 0.0078125,
|
|
"calib/step_q_w": 0.7032999999999999,
|
|
"calib/step_q_w_n": 3.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01953125,
|
|
"completions/max_length": 2981.0,
|
|
"completions/max_terminated_length": 2981.0,
|
|
"completions/mean_length": 1086.828125,
|
|
"completions/mean_terminated_length": 1108.4781494140625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 22.0,
|
|
"epoch": 0.1152,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 7.166666666666667e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 30399453.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 72
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01953125,
|
|
"completions/max_length": 2895.0,
|
|
"completions/max_terminated_length": 2895.0,
|
|
"completions/mean_length": 1068.90234375,
|
|
"completions/mean_terminated_length": 1090.1953125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.1168,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 7.111111111111111e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 30933660.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 73
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.00390625,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.00390625,
|
|
"calib/nonempty_step_conf_rate": 0.00390625,
|
|
"calib/step_conf_rate": 0.00390625,
|
|
"calib/step_q_w": 0.45,
|
|
"calib/step_q_w_n": 1.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0234375,
|
|
"completions/max_length": 2999.0,
|
|
"completions/max_terminated_length": 2999.0,
|
|
"completions/mean_length": 1020.0390625,
|
|
"completions/mean_terminated_length": 1044.52001953125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 8.0,
|
|
"epoch": 0.1184,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 7.055555555555556e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 31451646.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 74
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 3006.0,
|
|
"completions/max_terminated_length": 3006.0,
|
|
"completions/mean_length": 1156.68359375,
|
|
"completions/mean_terminated_length": 1170.3992919921875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 3.0,
|
|
"epoch": 0.12,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 7e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 32007253.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2926.0,
|
|
"completions/max_terminated_length": 2926.0,
|
|
"completions/mean_length": 1067.0234375,
|
|
"completions/mean_terminated_length": 1075.4251708984375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 17.0,
|
|
"epoch": 0.1216,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 6.944444444444444e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 32541139.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 76
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2959.0,
|
|
"completions/max_terminated_length": 2959.0,
|
|
"completions/mean_length": 1015.3671875,
|
|
"completions/mean_terminated_length": 1027.4071044921875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 17.0,
|
|
"epoch": 0.1232,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 6.888888888888889e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 33061681.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 77
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.00390625,
|
|
"calib/avg_num_step_conf": 0.00390625,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0078125,
|
|
"calib/nonempty_step_conf_rate": 0.00390625,
|
|
"calib/step_conf_rate": 0.00390625,
|
|
"calib/step_q_w": 0.9,
|
|
"calib/step_q_w_n": 1.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 3017.0,
|
|
"completions/max_terminated_length": 3017.0,
|
|
"completions/mean_length": 1039.0,
|
|
"completions/mean_terminated_length": 1051.3201904296875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 8.0,
|
|
"epoch": 0.1248,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 6.833333333333333e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 33584465.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 78
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0234375,
|
|
"completions/max_length": 2929.0,
|
|
"completions/max_terminated_length": 2929.0,
|
|
"completions/mean_length": 1074.19140625,
|
|
"completions/mean_terminated_length": 1099.9720458984375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 11.0,
|
|
"epoch": 0.1264,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 6.777777777777778e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 34114298.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 79
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2836.0,
|
|
"completions/max_terminated_length": 2836.0,
|
|
"completions/mean_length": 1014.15625,
|
|
"completions/mean_terminated_length": 1022.1417236328125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 5.0,
|
|
"epoch": 0.128,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 6.722222222222222e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 34635874.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2940.0,
|
|
"completions/max_terminated_length": 2940.0,
|
|
"completions/mean_length": 1188.2421875,
|
|
"completions/mean_terminated_length": 1197.598388671875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 8.0,
|
|
"epoch": 0.1296,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 6.666666666666666e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 35196288.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 81
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.00390625,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.00390625,
|
|
"calib/nonempty_step_conf_rate": 0.00390625,
|
|
"calib/step_conf_rate": 0.00390625,
|
|
"calib/step_q_w": 0.8,
|
|
"calib/step_q_w_n": 1.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01953125,
|
|
"completions/max_length": 2950.0,
|
|
"completions/max_terminated_length": 2950.0,
|
|
"completions/mean_length": 1147.44921875,
|
|
"completions/mean_terminated_length": 1170.3067626953125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 9.0,
|
|
"epoch": 0.1312,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 6.611111111111111e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 35750803.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 82
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.015625,
|
|
"completions/max_length": 2885.0,
|
|
"completions/max_terminated_length": 2885.0,
|
|
"completions/mean_length": 1118.296875,
|
|
"completions/mean_terminated_length": 1136.0477294921875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 4.0,
|
|
"epoch": 0.1328,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 6.555555555555555e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 36295047.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 83
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2844.0,
|
|
"completions/max_terminated_length": 2844.0,
|
|
"completions/mean_length": 895.1328125,
|
|
"completions/mean_terminated_length": 902.1810913085938,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 5.0,
|
|
"epoch": 0.1344,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 6.5e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 36785433.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 84
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2964.0,
|
|
"completions/max_terminated_length": 2964.0,
|
|
"completions/mean_length": 1084.10546875,
|
|
"completions/mean_terminated_length": 1096.9605712890625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.136,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 6.444444444444444e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 37322388.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 3048.0,
|
|
"completions/max_terminated_length": 3048.0,
|
|
"completions/mean_length": 1126.015625,
|
|
"completions/mean_terminated_length": 1139.36767578125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 7.0,
|
|
"epoch": 0.1376,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 6.388888888888888e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 37871904.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 86
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 3023.0,
|
|
"completions/max_terminated_length": 3023.0,
|
|
"completions/mean_length": 1228.83984375,
|
|
"completions/mean_terminated_length": 1238.5157470703125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 10.0,
|
|
"epoch": 0.1392,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 6.333333333333332e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 38445119.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 87
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2965.0,
|
|
"completions/max_terminated_length": 2965.0,
|
|
"completions/mean_length": 1121.98046875,
|
|
"completions/mean_terminated_length": 1130.81494140625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 11.0,
|
|
"epoch": 0.1408,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 6.277777777777777e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 38992386.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 88
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2963.0,
|
|
"completions/max_terminated_length": 2963.0,
|
|
"completions/mean_length": 1134.3515625,
|
|
"completions/mean_terminated_length": 1147.8023681640625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 6.0,
|
|
"epoch": 0.1424,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 6.222222222222223e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 39544348.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 89
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2895.0,
|
|
"completions/max_terminated_length": 2895.0,
|
|
"completions/mean_length": 1057.27734375,
|
|
"completions/mean_terminated_length": 1069.8143310546875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 8.0,
|
|
"epoch": 0.144,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 6.166666666666667e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 40077155.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 3025.0,
|
|
"completions/max_terminated_length": 3025.0,
|
|
"completions/mean_length": 1146.1484375,
|
|
"completions/mean_terminated_length": 1155.1732177734375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 35.0,
|
|
"epoch": 0.1456,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 6.111111111111112e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 40626425.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 91
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.015625,
|
|
"completions/max_length": 2963.0,
|
|
"completions/max_terminated_length": 2963.0,
|
|
"completions/mean_length": 1144.28125,
|
|
"completions/mean_terminated_length": 1162.4444580078125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 8.0,
|
|
"epoch": 0.1472,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 6.055555555555555e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 41180273.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 92
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2994.0,
|
|
"completions/max_terminated_length": 2994.0,
|
|
"completions/mean_length": 1154.6796875,
|
|
"completions/mean_terminated_length": 1163.7716064453125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 12.0,
|
|
"epoch": 0.1488,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 6e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 41737943.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 93
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.015625,
|
|
"completions/max_length": 2941.0,
|
|
"completions/max_terminated_length": 2941.0,
|
|
"completions/mean_length": 1111.44921875,
|
|
"completions/mean_terminated_length": 1129.09130859375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.1504,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5.944444444444444e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 42279618.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 94
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01953125,
|
|
"completions/max_length": 2985.0,
|
|
"completions/max_terminated_length": 2985.0,
|
|
"completions/mean_length": 1065.7578125,
|
|
"completions/mean_terminated_length": 1086.988037109375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 17.0,
|
|
"epoch": 0.152,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5.888888888888889e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 42813884.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2945.0,
|
|
"completions/max_terminated_length": 2945.0,
|
|
"completions/mean_length": 1039.0,
|
|
"completions/mean_terminated_length": 1047.18115234375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 4.0,
|
|
"epoch": 0.1536,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5.833333333333334e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 43341308.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 96
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2984.0,
|
|
"completions/max_terminated_length": 2984.0,
|
|
"completions/mean_length": 1063.22265625,
|
|
"completions/mean_terminated_length": 1071.594482421875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 21.0,
|
|
"epoch": 0.1552,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5.777777777777777e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 43874309.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 97
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 3007.0,
|
|
"completions/max_terminated_length": 3007.0,
|
|
"completions/mean_length": 1113.98828125,
|
|
"completions/mean_terminated_length": 1127.1976318359375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 6.0,
|
|
"epoch": 0.1568,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5.722222222222222e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 44420186.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 98
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2906.0,
|
|
"completions/max_terminated_length": 2906.0,
|
|
"completions/mean_length": 1128.171875,
|
|
"completions/mean_terminated_length": 1141.5494384765625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.1584,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5.666666666666666e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 44971142.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 99
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.015625,
|
|
"completions/max_length": 2922.0,
|
|
"completions/max_terminated_length": 2922.0,
|
|
"completions/mean_length": 1054.0546875,
|
|
"completions/mean_terminated_length": 1070.7857666015625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 15.0,
|
|
"epoch": 0.16,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5.611111111111111e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 45503124.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2832.0,
|
|
"completions/max_terminated_length": 2832.0,
|
|
"completions/mean_length": 938.03125,
|
|
"completions/mean_terminated_length": 949.1541748046875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 7.0,
|
|
"epoch": 0.1616,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5.555555555555555e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 46005404.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 101
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2908.0,
|
|
"completions/max_terminated_length": 2908.0,
|
|
"completions/mean_length": 1130.16015625,
|
|
"completions/mean_terminated_length": 1143.561279296875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 11.0,
|
|
"epoch": 0.1632,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5.5e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 46556869.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 102
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2822.0,
|
|
"completions/max_terminated_length": 2822.0,
|
|
"completions/mean_length": 972.79296875,
|
|
"completions/mean_terminated_length": 984.328125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 8.0,
|
|
"epoch": 0.1648,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5.444444444444443e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 47067784.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 103
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.00390625,
|
|
"calib/avg_num_step_conf": 0.00390625,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0078125,
|
|
"calib/nonempty_step_conf_rate": 0.00390625,
|
|
"calib/step_conf_rate": 0.00390625,
|
|
"calib/step_q_w": 0.8,
|
|
"calib/step_q_w_n": 1.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 3061.0,
|
|
"completions/max_terminated_length": 3061.0,
|
|
"completions/mean_length": 1063.31640625,
|
|
"completions/mean_terminated_length": 1071.68896484375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 8.0,
|
|
"epoch": 0.1664,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5.388888888888888e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 47595849.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 104
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 2919.0,
|
|
"completions/max_terminated_length": 2919.0,
|
|
"completions/mean_length": 1130.78515625,
|
|
"completions/mean_terminated_length": 1135.2197265625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 14.0,
|
|
"epoch": 0.168,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5.333333333333333e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 48147322.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2956.0,
|
|
"completions/max_terminated_length": 2956.0,
|
|
"completions/mean_length": 993.98828125,
|
|
"completions/mean_terminated_length": 1001.81494140625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 16.0,
|
|
"epoch": 0.1696,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5.277777777777777e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 48663927.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 106
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2848.0,
|
|
"completions/max_terminated_length": 2848.0,
|
|
"completions/mean_length": 1050.23046875,
|
|
"completions/mean_terminated_length": 1062.683837890625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 18.0,
|
|
"epoch": 0.1712,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5.222222222222223e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 49194834.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 107
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 3048.0,
|
|
"completions/max_terminated_length": 3048.0,
|
|
"completions/mean_length": 1131.703125,
|
|
"completions/mean_terminated_length": 1145.12255859375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.1728,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5.166666666666667e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 49745774.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 108
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.00390625,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.00390625,
|
|
"calib/nonempty_step_conf_rate": 0.00390625,
|
|
"calib/step_conf_rate": 0.00390625,
|
|
"calib/step_q_w": 0.98,
|
|
"calib/step_q_w_n": 1.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01953125,
|
|
"completions/max_length": 2955.0,
|
|
"completions/max_terminated_length": 2955.0,
|
|
"completions/mean_length": 1086.59765625,
|
|
"completions/mean_terminated_length": 1108.2430419921875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.1744,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5.111111111111111e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 50281599.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 109
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2829.0,
|
|
"completions/max_terminated_length": 2829.0,
|
|
"completions/mean_length": 987.63671875,
|
|
"completions/mean_terminated_length": 999.3478393554688,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 20.0,
|
|
"epoch": 0.176,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5.055555555555555e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 50796578.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2823.0,
|
|
"completions/max_terminated_length": 2823.0,
|
|
"completions/mean_length": 1035.81640625,
|
|
"completions/mean_terminated_length": 1043.972412109375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 25.0,
|
|
"epoch": 0.1776,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 51322723.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 111
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2861.0,
|
|
"completions/max_terminated_length": 2861.0,
|
|
"completions/mean_length": 1091.7734375,
|
|
"completions/mean_terminated_length": 1100.3701171875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 4.0,
|
|
"epoch": 0.1792,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 4.944444444444445e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 51864361.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 112
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2962.0,
|
|
"completions/max_terminated_length": 2962.0,
|
|
"completions/mean_length": 1121.0234375,
|
|
"completions/mean_terminated_length": 1129.850341796875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.1808,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 4.888888888888889e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 52408191.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 113
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0234375,
|
|
"completions/max_length": 2949.0,
|
|
"completions/max_terminated_length": 2949.0,
|
|
"completions/mean_length": 1183.1328125,
|
|
"completions/mean_terminated_length": 1211.528076171875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 17.0,
|
|
"epoch": 0.1824,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 4.833333333333333e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 52972689.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 114
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 2974.0,
|
|
"completions/max_terminated_length": 2974.0,
|
|
"completions/mean_length": 1078.66796875,
|
|
"completions/mean_terminated_length": 1082.8980712890625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 15.0,
|
|
"epoch": 0.184,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 4.777777777777778e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 53510836.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2849.0,
|
|
"completions/max_terminated_length": 2849.0,
|
|
"completions/mean_length": 1002.87890625,
|
|
"completions/mean_terminated_length": 1010.7755737304688,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 7.0,
|
|
"epoch": 0.1856,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 4.722222222222222e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 54028253.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 116
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01953125,
|
|
"completions/max_length": 3010.0,
|
|
"completions/max_terminated_length": 3010.0,
|
|
"completions/mean_length": 1162.40625,
|
|
"completions/mean_terminated_length": 1185.561767578125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 12.0,
|
|
"epoch": 0.1872,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 4.6666666666666666e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 54583869.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 117
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.02734375,
|
|
"completions/max_length": 3053.0,
|
|
"completions/max_terminated_length": 3053.0,
|
|
"completions/mean_length": 1087.0,
|
|
"completions/mean_terminated_length": 1117.5582275390625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 15.0,
|
|
"epoch": 0.1888,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 4.611111111111111e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 55122637.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 118
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.00390625,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.00390625,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 3070.0,
|
|
"completions/max_terminated_length": 3070.0,
|
|
"completions/mean_length": 1022.1484375,
|
|
"completions/mean_terminated_length": 1034.268798828125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.1904,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 4.555555555555555e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 55640435.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 119
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.015625,
|
|
"completions/max_length": 3062.0,
|
|
"completions/max_terminated_length": 3062.0,
|
|
"completions/mean_length": 1080.94921875,
|
|
"completions/mean_terminated_length": 1098.107177734375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.192,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 4.5e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 56174926.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 3038.0,
|
|
"completions/max_terminated_length": 3038.0,
|
|
"completions/mean_length": 1094.5078125,
|
|
"completions/mean_terminated_length": 1107.4862060546875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 8.0,
|
|
"epoch": 0.1936,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 4.444444444444444e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 56712568.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 121
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2988.0,
|
|
"completions/max_terminated_length": 2988.0,
|
|
"completions/mean_length": 1165.0625,
|
|
"completions/mean_terminated_length": 1174.2362060546875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 12.0,
|
|
"epoch": 0.1952,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 4.3888888888888884e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 57272272.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 122
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0234375,
|
|
"completions/max_length": 3016.0,
|
|
"completions/max_terminated_length": 3016.0,
|
|
"completions/mean_length": 1099.51953125,
|
|
"completions/mean_terminated_length": 1125.9080810546875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.1968,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 4.3333333333333335e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 57810981.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 123
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2870.0,
|
|
"completions/max_terminated_length": 2870.0,
|
|
"completions/mean_length": 968.6015625,
|
|
"completions/mean_terminated_length": 976.2283325195312,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 3.0,
|
|
"epoch": 0.1984,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 4.2777777777777775e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 58319463.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 124
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2826.0,
|
|
"completions/max_terminated_length": 2826.0,
|
|
"completions/mean_length": 1017.83203125,
|
|
"completions/mean_terminated_length": 1025.846435546875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 4.0,
|
|
"epoch": 0.2,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 4.222222222222222e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 58841228.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2903.0,
|
|
"completions/max_terminated_length": 2903.0,
|
|
"completions/mean_length": 1040.59375,
|
|
"completions/mean_terminated_length": 1052.932861328125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 3.0,
|
|
"epoch": 0.2016,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 4.1666666666666667e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 59369388.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 126
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.015625,
|
|
"completions/max_length": 2942.0,
|
|
"completions/max_terminated_length": 2942.0,
|
|
"completions/mean_length": 1140.87109375,
|
|
"completions/mean_terminated_length": 1158.980224609375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 10.0,
|
|
"epoch": 0.2032,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 4.1111111111111107e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 59923595.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 127
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2932.0,
|
|
"completions/max_terminated_length": 2932.0,
|
|
"completions/mean_length": 1121.69921875,
|
|
"completions/mean_terminated_length": 1130.531494140625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 6.0,
|
|
"epoch": 0.2048,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 4.055555555555555e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 60471974.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 128
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2974.0,
|
|
"completions/max_terminated_length": 2974.0,
|
|
"completions/mean_length": 1096.81640625,
|
|
"completions/mean_terminated_length": 1109.8221435546875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.2064,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 4e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 61011023.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 129
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 3031.0,
|
|
"completions/max_terminated_length": 3031.0,
|
|
"completions/mean_length": 1042.7421875,
|
|
"completions/mean_terminated_length": 1055.1068115234375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 9.0,
|
|
"epoch": 0.208,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 3.9444444444444444e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 61535589.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.015625,
|
|
"completions/max_length": 2905.0,
|
|
"completions/max_terminated_length": 2905.0,
|
|
"completions/mean_length": 1132.5625,
|
|
"completions/mean_terminated_length": 1150.539794921875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 8.0,
|
|
"epoch": 0.2096,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 3.888888888888889e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 62087669.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 131
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2914.0,
|
|
"completions/max_terminated_length": 2914.0,
|
|
"completions/mean_length": 1072.91796875,
|
|
"completions/mean_terminated_length": 1081.3660888671875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.2112,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 3.8333333333333335e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 62623632.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 132
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01953125,
|
|
"completions/max_length": 3017.0,
|
|
"completions/max_terminated_length": 3017.0,
|
|
"completions/mean_length": 958.64453125,
|
|
"completions/mean_terminated_length": 977.7410888671875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.2128,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 3.7777777777777775e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 63127029.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 133
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2972.0,
|
|
"completions/max_terminated_length": 2972.0,
|
|
"completions/mean_length": 1142.4140625,
|
|
"completions/mean_terminated_length": 1151.409423828125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 3.0,
|
|
"epoch": 0.2144,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 3.722222222222222e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 63681503.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 134
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.015625,
|
|
"completions/max_length": 2936.0,
|
|
"completions/max_terminated_length": 2936.0,
|
|
"completions/mean_length": 1090.01171875,
|
|
"completions/mean_terminated_length": 1107.3135986328125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 15.0,
|
|
"epoch": 0.216,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 3.666666666666666e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 64222690.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2860.0,
|
|
"completions/max_terminated_length": 2860.0,
|
|
"completions/mean_length": 961.3125,
|
|
"completions/mean_terminated_length": 968.8818969726562,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.2176,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 3.6111111111111107e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 64730930.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 136
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2900.0,
|
|
"completions/max_terminated_length": 2900.0,
|
|
"completions/mean_length": 1112.10546875,
|
|
"completions/mean_terminated_length": 1125.29248046875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.2192,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 3.5555555555555553e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 65277773.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 137
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2918.0,
|
|
"completions/max_terminated_length": 2918.0,
|
|
"completions/mean_length": 1041.1484375,
|
|
"completions/mean_terminated_length": 1049.346435546875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 6.0,
|
|
"epoch": 0.2208,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 3.5e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 65805883.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 138
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2932.0,
|
|
"completions/max_terminated_length": 2932.0,
|
|
"completions/mean_length": 998.8046875,
|
|
"completions/mean_terminated_length": 1006.6693115234375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 3.0,
|
|
"epoch": 0.2224,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 3.4444444444444444e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 66323385.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 139
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2837.0,
|
|
"completions/max_terminated_length": 2837.0,
|
|
"completions/mean_length": 1101.73046875,
|
|
"completions/mean_terminated_length": 1114.7945556640625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 23.0,
|
|
"epoch": 0.224,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 3.388888888888889e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 66867004.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 2936.0,
|
|
"completions/max_terminated_length": 2936.0,
|
|
"completions/mean_length": 1091.921875,
|
|
"completions/mean_terminated_length": 1096.2039794921875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 14.0,
|
|
"epoch": 0.2256,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 3.333333333333333e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 67408680.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 141
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 3022.0,
|
|
"completions/max_terminated_length": 3022.0,
|
|
"completions/mean_length": 1006.89453125,
|
|
"completions/mean_terminated_length": 1014.8228149414062,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 18.0,
|
|
"epoch": 0.2272,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 3.2777777777777776e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 67927597.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 142
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2758.0,
|
|
"completions/max_terminated_length": 2758.0,
|
|
"completions/mean_length": 1120.359375,
|
|
"completions/mean_terminated_length": 1133.644287109375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 5.0,
|
|
"epoch": 0.2288,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 3.222222222222222e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 68474529.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 143
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 3004.0,
|
|
"completions/max_terminated_length": 3004.0,
|
|
"completions/mean_length": 1132.0234375,
|
|
"completions/mean_terminated_length": 1145.4466552734375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 5.0,
|
|
"epoch": 0.2304,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 3.166666666666666e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 69024839.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 144
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 3041.0,
|
|
"completions/max_terminated_length": 3041.0,
|
|
"completions/mean_length": 1088.08984375,
|
|
"completions/mean_terminated_length": 1096.657470703125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 9.0,
|
|
"epoch": 0.232,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 3.111111111111111e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 69565406.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2842.0,
|
|
"completions/max_terminated_length": 2842.0,
|
|
"completions/mean_length": 991.71484375,
|
|
"completions/mean_terminated_length": 1003.474365234375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 25.0,
|
|
"epoch": 0.2336,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 3.055555555555556e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 70081429.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 146
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 3022.0,
|
|
"completions/max_terminated_length": 3022.0,
|
|
"completions/mean_length": 1031.70703125,
|
|
"completions/mean_terminated_length": 1039.8306884765625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.2352,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 3e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 70606626.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 147
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.015625,
|
|
"completions/max_length": 2918.0,
|
|
"completions/max_terminated_length": 2918.0,
|
|
"completions/mean_length": 1101.3203125,
|
|
"completions/mean_terminated_length": 1118.8016357421875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 5.0,
|
|
"epoch": 0.2368,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 2.9444444444444444e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 71146876.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 148
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2798.0,
|
|
"completions/max_terminated_length": 2798.0,
|
|
"completions/mean_length": 1010.1484375,
|
|
"completions/mean_terminated_length": 1018.1023559570312,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 22.0,
|
|
"epoch": 0.2384,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 2.8888888888888885e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 71667082.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 149
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2975.0,
|
|
"completions/max_terminated_length": 2975.0,
|
|
"completions/mean_length": 1086.33203125,
|
|
"completions/mean_terminated_length": 1094.8858642578125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 7.0,
|
|
"epoch": 0.24,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 2.833333333333333e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 72205943.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2901.0,
|
|
"completions/max_terminated_length": 2901.0,
|
|
"completions/mean_length": 1060.27734375,
|
|
"completions/mean_terminated_length": 1068.6259765625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 5.0,
|
|
"epoch": 0.2416,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 2.7777777777777776e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 72739254.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 151
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2894.0,
|
|
"completions/max_terminated_length": 2894.0,
|
|
"completions/mean_length": 1052.00390625,
|
|
"completions/mean_terminated_length": 1060.287353515625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 3.0,
|
|
"epoch": 0.2432,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 2.7222222222222216e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 73268375.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 152
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2955.0,
|
|
"completions/max_terminated_length": 2955.0,
|
|
"completions/mean_length": 1053.90234375,
|
|
"completions/mean_terminated_length": 1062.2008056640625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 9.0,
|
|
"epoch": 0.2448,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 2.6666666666666667e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 73799910.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 153
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2934.0,
|
|
"completions/max_terminated_length": 2934.0,
|
|
"completions/mean_length": 1075.3828125,
|
|
"completions/mean_terminated_length": 1083.850341796875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 6.0,
|
|
"epoch": 0.2464,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 2.6111111111111113e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 74337352.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 154
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2837.0,
|
|
"completions/max_terminated_length": 2837.0,
|
|
"completions/mean_length": 1098.609375,
|
|
"completions/mean_terminated_length": 1107.2598876953125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 9.0,
|
|
"epoch": 0.248,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 2.5555555555555553e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 74880468.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2976.0,
|
|
"completions/max_terminated_length": 2976.0,
|
|
"completions/mean_length": 1070.65234375,
|
|
"completions/mean_terminated_length": 1083.347900390625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.2496,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 2.5e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 75416699.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 156
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01953125,
|
|
"completions/max_length": 2859.0,
|
|
"completions/max_terminated_length": 2859.0,
|
|
"completions/mean_length": 1022.59375,
|
|
"completions/mean_terminated_length": 1042.9642333984375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 4.0,
|
|
"epoch": 0.2512,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 2.4444444444444445e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 75939691.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 157
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2893.0,
|
|
"completions/max_terminated_length": 2893.0,
|
|
"completions/mean_length": 1023.33984375,
|
|
"completions/mean_terminated_length": 1031.3975830078125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 9.0,
|
|
"epoch": 0.2528,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 2.388888888888889e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 76463618.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 158
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2825.0,
|
|
"completions/max_terminated_length": 2825.0,
|
|
"completions/mean_length": 1109.16796875,
|
|
"completions/mean_terminated_length": 1117.901611328125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 5.0,
|
|
"epoch": 0.2544,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 2.3333333333333333e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 77009709.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 159
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2907.0,
|
|
"completions/max_terminated_length": 2907.0,
|
|
"completions/mean_length": 1154.60546875,
|
|
"completions/mean_terminated_length": 1163.6968994140625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 9.0,
|
|
"epoch": 0.256,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 2.2777777777777776e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 77563432.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.015625,
|
|
"completions/max_length": 3018.0,
|
|
"completions/max_terminated_length": 3018.0,
|
|
"completions/mean_length": 1136.26953125,
|
|
"completions/mean_terminated_length": 1154.3056640625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 22.0,
|
|
"epoch": 0.2576,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 2.222222222222222e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 78112093.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 161
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01953125,
|
|
"completions/max_length": 3016.0,
|
|
"completions/max_terminated_length": 3016.0,
|
|
"completions/mean_length": 1101.1328125,
|
|
"completions/mean_terminated_length": 1123.0677490234375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 4.0,
|
|
"epoch": 0.2592,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 2.1666666666666667e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 78651087.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 162
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 3014.0,
|
|
"completions/max_terminated_length": 3014.0,
|
|
"completions/mean_length": 1056.65625,
|
|
"completions/mean_terminated_length": 1069.185791015625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.2608,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 2.111111111111111e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 79180167.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 163
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2888.0,
|
|
"completions/max_terminated_length": 2888.0,
|
|
"completions/mean_length": 1163.8125,
|
|
"completions/mean_terminated_length": 1172.976318359375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 7.0,
|
|
"epoch": 0.2624,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 2.0555555555555553e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 79739663.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 164
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2948.0,
|
|
"completions/max_terminated_length": 2948.0,
|
|
"completions/mean_length": 1224.3828125,
|
|
"completions/mean_terminated_length": 1238.9012451171875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 11.0,
|
|
"epoch": 0.264,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 2e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 80314225.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 3022.0,
|
|
"completions/max_terminated_length": 3022.0,
|
|
"completions/mean_length": 1024.66015625,
|
|
"completions/mean_terminated_length": 1032.7283935546875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 25.0,
|
|
"epoch": 0.2656,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 1.9444444444444445e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 80835002.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 166
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2798.0,
|
|
"completions/max_terminated_length": 2798.0,
|
|
"completions/mean_length": 1034.26171875,
|
|
"completions/mean_terminated_length": 1042.405517578125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 9.0,
|
|
"epoch": 0.2672,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 1.8888888888888888e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 81361653.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 167
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.015625,
|
|
"completions/max_length": 2841.0,
|
|
"completions/max_terminated_length": 2841.0,
|
|
"completions/mean_length": 1116.234375,
|
|
"completions/mean_terminated_length": 1133.952392578125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 18.0,
|
|
"epoch": 0.2688,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 1.833333333333333e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 81909553.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 168
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2944.0,
|
|
"completions/max_terminated_length": 2944.0,
|
|
"completions/mean_length": 1004.765625,
|
|
"completions/mean_terminated_length": 1016.6798706054688,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 3.0,
|
|
"epoch": 0.2704,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 1.7777777777777776e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 82428069.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 169
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.00390625,
|
|
"calib/avg_num_step_conf": 0.00390625,
|
|
"calib/ece": 0.8,
|
|
"calib/final_conf_rate": 0.00390625,
|
|
"calib/format_rate": 0.0,
|
|
"calib/frac_conf_gt_0.9": 0.0,
|
|
"calib/mean_conf": 0.8,
|
|
"calib/mu_c": NaN,
|
|
"calib/mu_w": 0.8,
|
|
"calib/nonempty_final_conf_rate": 0.00390625,
|
|
"calib/nonempty_reasoning_rate": 0.0078125,
|
|
"calib/nonempty_step_conf_rate": 0.00390625,
|
|
"calib/pce": 0.8,
|
|
"calib/std_conf": 0.0,
|
|
"calib/step_conf_rate": 0.00390625,
|
|
"calib/step_q_w": 0.75,
|
|
"calib/step_q_w_n": 1.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2963.0,
|
|
"completions/max_terminated_length": 2963.0,
|
|
"completions/mean_length": 1091.875,
|
|
"completions/mean_terminated_length": 1104.8221435546875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.272,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 1.7222222222222222e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 82960605.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 3012.0,
|
|
"completions/max_terminated_length": 3012.0,
|
|
"completions/mean_length": 960.78515625,
|
|
"completions/mean_terminated_length": 968.3504028320312,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 9.0,
|
|
"epoch": 0.2736,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 1.6666666666666665e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 83465222.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 171
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 3059.0,
|
|
"completions/max_terminated_length": 3059.0,
|
|
"completions/mean_length": 1178.6875,
|
|
"completions/mean_terminated_length": 1192.6640625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 5.0,
|
|
"epoch": 0.2752,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 1.611111111111111e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 84027150.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 172
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 3009.0,
|
|
"completions/max_terminated_length": 3009.0,
|
|
"completions/mean_length": 1210.18359375,
|
|
"completions/mean_terminated_length": 1224.53369140625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 7.0,
|
|
"epoch": 0.2768,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 1.5555555555555556e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 84597773.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 173
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2929.0,
|
|
"completions/max_terminated_length": 2929.0,
|
|
"completions/mean_length": 1143.296875,
|
|
"completions/mean_terminated_length": 1156.853759765625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 2.0,
|
|
"epoch": 0.2784,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 1.5e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 85152601.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 174
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.015625,
|
|
"completions/max_length": 2915.0,
|
|
"completions/max_terminated_length": 2915.0,
|
|
"completions/mean_length": 1112.11328125,
|
|
"completions/mean_terminated_length": 1129.7659912109375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 14.0,
|
|
"epoch": 0.28,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 1.4444444444444442e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 85697078.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0234375,
|
|
"completions/max_length": 3021.0,
|
|
"completions/max_terminated_length": 3021.0,
|
|
"completions/mean_length": 1048.34375,
|
|
"completions/mean_terminated_length": 1073.5040283203125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 8.0,
|
|
"epoch": 0.2816,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 1.3888888888888888e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 86224110.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 176
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2962.0,
|
|
"completions/max_terminated_length": 2962.0,
|
|
"completions/mean_length": 1140.140625,
|
|
"completions/mean_terminated_length": 1149.1181640625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 1.0,
|
|
"epoch": 0.2832,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 1.3333333333333334e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 86778130.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 177
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2815.0,
|
|
"completions/max_terminated_length": 2815.0,
|
|
"completions/mean_length": 1183.52734375,
|
|
"completions/mean_terminated_length": 1192.846435546875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 3.0,
|
|
"epoch": 0.2848,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 1.2777777777777777e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 87343257.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 178
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.015625,
|
|
"completions/max_length": 2986.0,
|
|
"completions/max_terminated_length": 2986.0,
|
|
"completions/mean_length": 1141.56640625,
|
|
"completions/mean_terminated_length": 1159.6865234375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 9.0,
|
|
"epoch": 0.2864,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 1.2222222222222222e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 87891170.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 179
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2974.0,
|
|
"completions/max_terminated_length": 2974.0,
|
|
"completions/mean_length": 1196.1484375,
|
|
"completions/mean_terminated_length": 1210.33203125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 5.0,
|
|
"epoch": 0.288,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 1.1666666666666667e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 88455328.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2921.0,
|
|
"completions/max_terminated_length": 2921.0,
|
|
"completions/mean_length": 1195.22265625,
|
|
"completions/mean_terminated_length": 1204.6339111328125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 28.0,
|
|
"epoch": 0.2896,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 1.111111111111111e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 89018161.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 181
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2984.0,
|
|
"completions/max_terminated_length": 2984.0,
|
|
"completions/mean_length": 1099.7421875,
|
|
"completions/mean_terminated_length": 1108.401611328125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 9.0,
|
|
"epoch": 0.2912,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 1.0555555555555555e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 89561495.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 182
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 3012.0,
|
|
"completions/max_terminated_length": 3012.0,
|
|
"completions/mean_length": 1136.4921875,
|
|
"completions/mean_terminated_length": 1140.9490966796875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 4.0,
|
|
"epoch": 0.2928,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 1e-07,
|
|
"loss": 0.0,
|
|
"num_tokens": 90113261.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 183
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2984.0,
|
|
"completions/max_terminated_length": 2984.0,
|
|
"completions/mean_length": 1074.80078125,
|
|
"completions/mean_terminated_length": 1083.2637939453125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 23.0,
|
|
"epoch": 0.2944,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 9.444444444444444e-08,
|
|
"loss": 0.0,
|
|
"num_tokens": 90648722.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 184
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2894.0,
|
|
"completions/max_terminated_length": 2894.0,
|
|
"completions/mean_length": 1060.1875,
|
|
"completions/mean_terminated_length": 1072.7589111328125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 8.0,
|
|
"epoch": 0.296,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 8.888888888888888e-08,
|
|
"loss": 0.0,
|
|
"num_tokens": 91181714.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2904.0,
|
|
"completions/max_terminated_length": 2904.0,
|
|
"completions/mean_length": 1102.87890625,
|
|
"completions/mean_terminated_length": 1115.95654296875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 6.0,
|
|
"epoch": 0.2976,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 8.333333333333333e-08,
|
|
"loss": 0.0,
|
|
"num_tokens": 91726195.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 186
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2932.0,
|
|
"completions/max_terminated_length": 2932.0,
|
|
"completions/mean_length": 1163.03515625,
|
|
"completions/mean_terminated_length": 1172.19287109375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 10.0,
|
|
"epoch": 0.2992,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 7.777777777777778e-08,
|
|
"loss": 0.0,
|
|
"num_tokens": 92285100.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 187
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.015625,
|
|
"completions/max_length": 2896.0,
|
|
"completions/max_terminated_length": 2896.0,
|
|
"completions/mean_length": 1062.48828125,
|
|
"completions/mean_terminated_length": 1079.353271484375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 10.0,
|
|
"epoch": 0.3008,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 7.222222222222221e-08,
|
|
"loss": 0.0,
|
|
"num_tokens": 92814521.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 188
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 3057.0,
|
|
"completions/max_terminated_length": 3057.0,
|
|
"completions/mean_length": 1209.0078125,
|
|
"completions/mean_terminated_length": 1213.7491455078125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 7.0,
|
|
"epoch": 0.3024,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 6.666666666666667e-08,
|
|
"loss": 0.0,
|
|
"num_tokens": 93383731.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 189
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.015625,
|
|
"completions/max_length": 2900.0,
|
|
"completions/max_terminated_length": 2900.0,
|
|
"completions/mean_length": 1079.34375,
|
|
"completions/mean_terminated_length": 1096.4761962890625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 12.0,
|
|
"epoch": 0.304,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 6.111111111111111e-08,
|
|
"loss": 0.0,
|
|
"num_tokens": 93911275.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 3032.0,
|
|
"completions/max_terminated_length": 3032.0,
|
|
"completions/mean_length": 1036.34765625,
|
|
"completions/mean_terminated_length": 1044.5078125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 19.0,
|
|
"epoch": 0.3056,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5.555555555555555e-08,
|
|
"loss": 0.0,
|
|
"num_tokens": 94438436.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 191
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2859.0,
|
|
"completions/max_terminated_length": 2859.0,
|
|
"completions/mean_length": 1163.59765625,
|
|
"completions/mean_terminated_length": 1177.395263671875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 9.0,
|
|
"epoch": 0.3072,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5e-08,
|
|
"loss": 0.0,
|
|
"num_tokens": 94998461.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 192
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.0078125,
|
|
"completions/max_length": 2913.0,
|
|
"completions/max_terminated_length": 2913.0,
|
|
"completions/mean_length": 1036.12890625,
|
|
"completions/mean_terminated_length": 1044.287353515625,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 9.0,
|
|
"epoch": 0.3088,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 4.444444444444444e-08,
|
|
"loss": 0.0,
|
|
"num_tokens": 95525406.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 193
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.00390625,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.00390625,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 2952.0,
|
|
"completions/max_terminated_length": 2952.0,
|
|
"completions/mean_length": 1050.99609375,
|
|
"completions/mean_terminated_length": 1055.11767578125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 12.0,
|
|
"epoch": 0.3104,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 3.888888888888889e-08,
|
|
"loss": 0.0,
|
|
"num_tokens": 96055093.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 194
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2955.0,
|
|
"completions/max_terminated_length": 2955.0,
|
|
"completions/mean_length": 1149.95703125,
|
|
"completions/mean_terminated_length": 1163.5928955078125,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 3.0,
|
|
"epoch": 0.312,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 3.3333333333333334e-08,
|
|
"loss": 0.0,
|
|
"num_tokens": 96608714.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.00390625,
|
|
"completions/max_length": 3033.0,
|
|
"completions/max_terminated_length": 3033.0,
|
|
"completions/mean_length": 1134.48828125,
|
|
"completions/mean_terminated_length": 1138.9373779296875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 6.0,
|
|
"epoch": 0.3136,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 2.7777777777777774e-08,
|
|
"loss": 0.0,
|
|
"num_tokens": 97161287.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 196
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2964.0,
|
|
"completions/max_terminated_length": 2964.0,
|
|
"completions/mean_length": 1116.45703125,
|
|
"completions/mean_terminated_length": 1129.6956787109375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 3.0,
|
|
"epoch": 0.3152,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 2.222222222222222e-08,
|
|
"loss": 0.0,
|
|
"num_tokens": 97707196.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 197
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 2785.0,
|
|
"completions/max_terminated_length": 2785.0,
|
|
"completions/mean_length": 1100.80859375,
|
|
"completions/mean_terminated_length": 1113.8616943359375,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 5.0,
|
|
"epoch": 0.3168,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 1.6666666666666667e-08,
|
|
"loss": 0.0,
|
|
"num_tokens": 98251147.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 198
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.015625,
|
|
"completions/max_length": 2881.0,
|
|
"completions/max_terminated_length": 2881.0,
|
|
"completions/mean_length": 1063.796875,
|
|
"completions/mean_terminated_length": 1080.6826171875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 11.0,
|
|
"epoch": 0.3184,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 1.111111111111111e-08,
|
|
"loss": 0.0,
|
|
"num_tokens": 98785623.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 199
|
|
},
|
|
{
|
|
"calib/answer_extract_rate": 0.0,
|
|
"calib/avg_num_step_conf": 0.0,
|
|
"calib/final_conf_rate": 0.0,
|
|
"calib/format_rate": 0.0,
|
|
"calib/nonempty_final_conf_rate": 0.0,
|
|
"calib/nonempty_reasoning_rate": 0.0,
|
|
"calib/nonempty_step_conf_rate": 0.0,
|
|
"calib/step_conf_rate": 0.0,
|
|
"clip_ratio/high_max": 0.0,
|
|
"clip_ratio/high_mean": 0.0,
|
|
"clip_ratio/low_mean": 0.0,
|
|
"clip_ratio/low_min": 0.0,
|
|
"clip_ratio/region_mean": 0.0,
|
|
"completions/clipped_ratio": 0.01171875,
|
|
"completions/max_length": 3065.0,
|
|
"completions/max_terminated_length": 3065.0,
|
|
"completions/mean_length": 1158.99609375,
|
|
"completions/mean_terminated_length": 1172.7391357421875,
|
|
"completions/min_length": 0.0,
|
|
"completions/min_terminated_length": 8.0,
|
|
"epoch": 0.32,
|
|
"grad_norm": 0.0,
|
|
"learning_rate": 5.555555555555555e-09,
|
|
"loss": 0.0,
|
|
"num_tokens": 99341446.0,
|
|
"reward": 0.0,
|
|
"reward_std": 0.0,
|
|
"rewards/accuracy_reward_step": 0.0,
|
|
"rewards/format_reward_step": 0.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 0.32,
|
|
"step": 200,
|
|
"total_flos": 0.0,
|
|
"train_loss": -0.01405992767540738,
|
|
"train_runtime": 12749.9065,
|
|
"train_samples_per_second": 4.016,
|
|
"train_steps_per_second": 0.016
|
|
}
|
|
],
|
|
"logging_steps": 1,
|
|
"max_steps": 200,
|
|
"num_input_tokens_seen": 99341446,
|
|
"num_train_epochs": 1,
|
|
"save_steps": 25,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": true
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 0.0,
|
|
"train_batch_size": 4,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|