| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.004997251511668582, |
| "eval_steps": 500, |
| "global_step": 1200, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 134.45, |
| "correct_distill_rate": 0.0875, |
| "distill_active_tokens": 21.8, |
| "distill_lambda": 0.5225, |
| "distill_loss": 0.39805216789245607, |
| "distill_pairs": 675.8, |
| "distill_skipped_no_correct_rate": 0.3, |
| "eligible_block_rate": 0.6, |
| "epoch": 4.1643762597238186e-05, |
| "grad_norm": 0.053834181278944016, |
| "grpo_loss": -0.0009230197872966528, |
| "kl": 0.08882212191820145, |
| "learning_rate": 2e-06, |
| "loss": 0.2047, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5500000007450581, |
| "reward_std": 0.2305372968316078, |
| "rewards/countdown_reward_func": 0.5500000007450581, |
| "step": 10, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 0.3 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 131.7375, |
| "correct_distill_rate": 0.1, |
| "distill_active_tokens": 25.325, |
| "distill_lambda": 0.5725, |
| "distill_loss": 0.41235754191875457, |
| "distill_pairs": 785.075, |
| "distill_skipped_no_correct_rate": 0.2, |
| "eligible_block_rate": 0.8, |
| "epoch": 8.328752519447637e-05, |
| "grad_norm": 1.2181833982467651, |
| "grpo_loss": 0.0015296985395252705, |
| "kl": 0.07489410489797592, |
| "learning_rate": 2e-06, |
| "loss": 0.2385, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.6437500007450581, |
| "reward_std": 0.14525448456406592, |
| "rewards/countdown_reward_func": 0.6437500007450581, |
| "step": 20, |
| "teacher_correct_rate": 0.8, |
| "teacher_step_mean": 25.6, |
| "teacher_step_min": 25.6, |
| "zero_std_ratio": 0.6 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 136.2125, |
| "correct_distill_rate": 0.025, |
| "distill_active_tokens": 6.45, |
| "distill_lambda": 0.6224999999999999, |
| "distill_loss": 0.10351597368717194, |
| "distill_pairs": 199.95, |
| "distill_skipped_no_correct_rate": 0.8, |
| "eligible_block_rate": 0.2, |
| "epoch": 0.00012493128779171456, |
| "grad_norm": 0.06615928560495377, |
| "grpo_loss": 0.003594341850839555, |
| "kl": 0.08914288803935051, |
| "learning_rate": 2e-06, |
| "loss": 0.068, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.2787500016391277, |
| "reward_std": 0.00353553369641304, |
| "rewards/countdown_reward_func": 0.2787500016391277, |
| "step": 30, |
| "teacher_correct_rate": 0.2, |
| "teacher_step_mean": 6.4, |
| "teacher_step_min": 6.4, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.2625, |
| "correct_distill_rate": 0.1, |
| "distill_active_tokens": 25.15, |
| "distill_lambda": 0.6725, |
| "distill_loss": 0.3509261578321457, |
| "distill_pairs": 779.65, |
| "distill_skipped_no_correct_rate": 0.2, |
| "eligible_block_rate": 0.7, |
| "epoch": 0.00016657505038895275, |
| "grad_norm": 0.06280245631933212, |
| "grpo_loss": 0.0034127396531403063, |
| "kl": 0.07255078330636025, |
| "learning_rate": 2e-06, |
| "loss": 0.2393, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.748750002682209, |
| "reward_std": 0.09584256410598754, |
| "rewards/countdown_reward_func": 0.748750002682209, |
| "step": 40, |
| "teacher_correct_rate": 0.7, |
| "teacher_step_mean": 22.4, |
| "teacher_step_min": 22.4, |
| "zero_std_ratio": 0.8 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 130.05, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 19.425, |
| "distill_lambda": 0.7224999999999999, |
| "distill_loss": 0.29210796356201174, |
| "distill_pairs": 602.175, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.00020821881298619093, |
| "grad_norm": 1.322534203529358, |
| "grpo_loss": 0.004616522812284529, |
| "kl": 0.07806920409202575, |
| "learning_rate": 2e-06, |
| "loss": 0.2173, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.44875000417232513, |
| "reward_std": 0.12158873081207275, |
| "rewards/countdown_reward_func": 0.44875000417232513, |
| "step": 50, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 0.7 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.9, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.9, |
| "distill_lambda": 0.7725000000000001, |
| "distill_loss": 0.20438768863677978, |
| "distill_pairs": 492.9, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.0002498625755834291, |
| "grad_norm": 2.4953360557556152, |
| "grpo_loss": 0.0034515597857534886, |
| "kl": 0.08414197862148284, |
| "learning_rate": 2e-06, |
| "loss": 0.1619, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.40375000312924386, |
| "reward_std": 0.07348170876502991, |
| "rewards/countdown_reward_func": 0.40375000312924386, |
| "step": 60, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 0.8 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 128.0375, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.9, |
| "distill_lambda": 0.8225, |
| "distill_loss": 0.1732427805662155, |
| "distill_pairs": 492.9, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.0002915063381806673, |
| "grad_norm": 1.0679490566253662, |
| "grpo_loss": 0.00284015997312963, |
| "kl": 0.07100400105118751, |
| "learning_rate": 2e-06, |
| "loss": 0.1468, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5500000007450581, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.5500000007450581, |
| "step": 70, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.625, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 19.0, |
| "distill_lambda": 0.8724999999999999, |
| "distill_loss": 0.22138834297657012, |
| "distill_pairs": 589.0, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.0003331501007779055, |
| "grad_norm": 1.7115774154663086, |
| "grpo_loss": 0.0028273086762055756, |
| "kl": 0.07396577894687653, |
| "learning_rate": 2e-06, |
| "loss": 0.1965, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5612500011920929, |
| "reward_std": 0.03181980550289154, |
| "rewards/countdown_reward_func": 0.5612500011920929, |
| "step": 80, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.7875, |
| "correct_distill_rate": 0.0375, |
| "distill_active_tokens": 9.45, |
| "distill_lambda": 0.9225, |
| "distill_loss": 0.11074086129665375, |
| "distill_pairs": 292.95, |
| "distill_skipped_no_correct_rate": 0.7, |
| "eligible_block_rate": 0.3, |
| "epoch": 0.0003747938633751437, |
| "grad_norm": 0.02468368038535118, |
| "grpo_loss": 0.0027098105056211352, |
| "kl": 0.06996863931417466, |
| "learning_rate": 2e-06, |
| "loss": 0.1058, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.34749999865889547, |
| "reward_std": 0.04166190326213837, |
| "rewards/countdown_reward_func": 0.34749999865889547, |
| "step": 90, |
| "teacher_correct_rate": 0.3, |
| "teacher_step_mean": 9.6, |
| "teacher_step_min": 9.6, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.0125, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.7, |
| "distill_lambda": 0.9724999999999998, |
| "distill_loss": 0.11106297373771667, |
| "distill_pairs": 393.7, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.3, |
| "epoch": 0.00041643762597238186, |
| "grad_norm": 1.4299242496490479, |
| "grpo_loss": 0.0029246833291836083, |
| "kl": 0.07311708219349385, |
| "learning_rate": 2e-06, |
| "loss": 0.112, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.3812500014901161, |
| "reward_std": 0.03181980550289154, |
| "rewards/countdown_reward_func": 0.3812500014901161, |
| "step": 100, |
| "teacher_correct_rate": 0.3, |
| "teacher_step_mean": 9.6, |
| "teacher_step_min": 9.6, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.9875, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.7, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.10228514820337295, |
| "distill_pairs": 393.7, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.00045808138856962005, |
| "grad_norm": 0.05337980389595032, |
| "grpo_loss": 0.002648683008737862, |
| "kl": 0.06621707826852799, |
| "learning_rate": 2e-06, |
| "loss": 0.1049, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.46000000089406967, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.46000000089406967, |
| "step": 110, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.7625, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.75, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.1441916659474373, |
| "distill_pairs": 488.25, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.3, |
| "epoch": 0.0004997251511668582, |
| "grad_norm": 0.03301816061139107, |
| "grpo_loss": 0.002710840804502368, |
| "kl": 0.062121478095650676, |
| "learning_rate": 2e-06, |
| "loss": 0.1469, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.38000000193715094, |
| "reward_std": 0.0989949494600296, |
| "rewards/countdown_reward_func": 0.38000000193715094, |
| "step": 120, |
| "teacher_correct_rate": 0.3, |
| "teacher_step_mean": 9.6, |
| "teacher_step_min": 9.6, |
| "zero_std_ratio": 0.7 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.0125, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.6, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.0893730416893959, |
| "distill_pairs": 390.6, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.0005413689137640964, |
| "grad_norm": 0.045211318880319595, |
| "grpo_loss": 0.0025563356932252644, |
| "kl": 0.06390839330852031, |
| "learning_rate": 2e-06, |
| "loss": 0.0919, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.46000000089406967, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.46000000089406967, |
| "step": 130, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.375, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 19.175, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.1658264935016632, |
| "distill_pairs": 594.425, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.0005830126763613346, |
| "grad_norm": 0.030579833313822746, |
| "grpo_loss": 0.002399870869703591, |
| "kl": 0.0632798433303833, |
| "learning_rate": 2e-06, |
| "loss": 0.1682, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5837500005960464, |
| "reward_std": 0.04657942354679108, |
| "rewards/countdown_reward_func": 0.5837500005960464, |
| "step": 140, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.4, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.7, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.1175243616104126, |
| "distill_pairs": 486.7, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.0006246564389585727, |
| "grad_norm": 0.06928855180740356, |
| "grpo_loss": 0.0024270519148558377, |
| "kl": 0.060676299035549164, |
| "learning_rate": 2e-06, |
| "loss": 0.12, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5500000007450581, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.5500000007450581, |
| "step": 150, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.9375, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 19.225, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.14448934346437453, |
| "distill_pairs": 595.975, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.000666300201555811, |
| "grad_norm": 4.191497325897217, |
| "grpo_loss": 0.0024056359892711044, |
| "kl": 0.06014093905687332, |
| "learning_rate": 2e-06, |
| "loss": 0.1469, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.55, |
| "reward_std": 0.10530151426792145, |
| "rewards/countdown_reward_func": 0.55, |
| "step": 160, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 0.7 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.45, |
| "correct_distill_rate": 0.1, |
| "distill_active_tokens": 25.175, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.2028204768896103, |
| "distill_pairs": 780.425, |
| "distill_skipped_no_correct_rate": 0.2, |
| "eligible_block_rate": 0.8, |
| "epoch": 0.0007079439641530491, |
| "grad_norm": 0.03161434456706047, |
| "grpo_loss": 0.0023555899737402798, |
| "kl": 0.05888980589807034, |
| "learning_rate": 2e-06, |
| "loss": 0.2052, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.7974999979138374, |
| "reward_std": 0.04166190326213837, |
| "rewards/countdown_reward_func": 0.7974999979138374, |
| "step": 170, |
| "teacher_correct_rate": 0.8, |
| "teacher_step_mean": 25.6, |
| "teacher_step_min": 25.6, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.6, |
| "correct_distill_rate": 0.0375, |
| "distill_active_tokens": 9.4, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.059993934631347653, |
| "distill_pairs": 291.4, |
| "distill_skipped_no_correct_rate": 0.7, |
| "eligible_block_rate": 0.3, |
| "epoch": 0.0007495877267502874, |
| "grad_norm": 0.056213609874248505, |
| "grpo_loss": 0.0024489407893270253, |
| "kl": 0.06122352182865143, |
| "learning_rate": 2e-06, |
| "loss": 0.0624, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.3700000010430813, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.3700000010430813, |
| "step": 180, |
| "teacher_correct_rate": 0.3, |
| "teacher_step_mean": 9.6, |
| "teacher_step_min": 9.6, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.4, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 18.7, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.12998489588499068, |
| "distill_pairs": 579.7, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.6, |
| "epoch": 0.0007912314893475255, |
| "grad_norm": 2.3939881324768066, |
| "grpo_loss": 0.0022017126553691923, |
| "kl": 0.05504281930625439, |
| "learning_rate": 2e-06, |
| "loss": 0.1322, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.6400000005960464, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.6400000005960464, |
| "step": 190, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.2, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 16.0, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.07298972904682159, |
| "distill_pairs": 496.0, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.0008328752519447637, |
| "grad_norm": 0.03724989667534828, |
| "grpo_loss": 0.002455541817471385, |
| "kl": 0.061388547718524936, |
| "learning_rate": 2e-06, |
| "loss": 0.0754, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5500000007450581, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.5500000007450581, |
| "step": 200, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.4, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 18.9, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.09762635082006454, |
| "distill_pairs": 585.9, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.6, |
| "epoch": 0.0008745190145420019, |
| "grad_norm": 2.339521646499634, |
| "grpo_loss": 0.0026678973343223335, |
| "kl": 0.06669746115803718, |
| "learning_rate": 2e-06, |
| "loss": 0.1003, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.6287499994039536, |
| "reward_std": 0.03181980550289154, |
| "rewards/countdown_reward_func": 0.6287499994039536, |
| "step": 210, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.2, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 18.6, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.10003485530614853, |
| "distill_pairs": 576.6, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.0009161627771392401, |
| "grad_norm": 0.055458247661590576, |
| "grpo_loss": 0.0025825670454651117, |
| "kl": 0.06456419043242931, |
| "learning_rate": 2e-06, |
| "loss": 0.1026, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5612500011920929, |
| "reward_std": 0.03181980550289154, |
| "rewards/countdown_reward_func": 0.5612500011920929, |
| "step": 220, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.1625, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 19.125, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.11488302350044251, |
| "distill_pairs": 592.875, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.0009578065397364782, |
| "grad_norm": 2.3541901111602783, |
| "grpo_loss": 0.0020325953490100803, |
| "kl": 0.059603889286518094, |
| "learning_rate": 2e-06, |
| "loss": 0.1169, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5950000017881394, |
| "reward_std": 0.04810702204704285, |
| "rewards/countdown_reward_func": 0.5950000017881394, |
| "step": 230, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.0875, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.6, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.07333092391490936, |
| "distill_pairs": 390.6, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.0009994503023337165, |
| "grad_norm": 3.1169795989990234, |
| "grpo_loss": 0.0025994319119490685, |
| "kl": 0.06498579904437066, |
| "learning_rate": 2e-06, |
| "loss": 0.0759, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.46000000089406967, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.46000000089406967, |
| "step": 240, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.1, |
| "correct_distill_rate": 0.0375, |
| "distill_active_tokens": 9.4, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.05240741074085235, |
| "distill_pairs": 291.4, |
| "distill_skipped_no_correct_rate": 0.7, |
| "eligible_block_rate": 0.3, |
| "epoch": 0.0010410940649309547, |
| "grad_norm": 0.05559613183140755, |
| "grpo_loss": 0.0024914028821513058, |
| "kl": 0.06228507347404957, |
| "learning_rate": 2e-06, |
| "loss": 0.0549, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.3700000010430813, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.3700000010430813, |
| "step": 250, |
| "teacher_correct_rate": 0.3, |
| "teacher_step_mean": 9.6, |
| "teacher_step_min": 9.6, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.1625, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 19.0, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.10929710865020752, |
| "distill_pairs": 589.0, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.6, |
| "epoch": 0.0010827378275281927, |
| "grad_norm": 1.4984177350997925, |
| "grpo_loss": 0.0020767614827491344, |
| "kl": 0.05536299608647823, |
| "learning_rate": 2e-06, |
| "loss": 0.1114, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.6174999982118606, |
| "reward_std": 0.06363961100578308, |
| "rewards/countdown_reward_func": 0.6174999982118606, |
| "step": 260, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 0.8 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.0875, |
| "correct_distill_rate": 0.0875, |
| "distill_active_tokens": 21.7, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.1794736258685589, |
| "distill_pairs": 672.7, |
| "distill_skipped_no_correct_rate": 0.3, |
| "eligible_block_rate": 0.7, |
| "epoch": 0.001124381590125431, |
| "grad_norm": 0.0466265007853508, |
| "grpo_loss": 0.002448706584982574, |
| "kl": 0.054203113168478013, |
| "learning_rate": 2e-06, |
| "loss": 0.1819, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.669999998062849, |
| "reward_std": 0.12330824732780457, |
| "rewards/countdown_reward_func": 0.669999998062849, |
| "step": 270, |
| "teacher_correct_rate": 0.7, |
| "teacher_step_mean": 22.4, |
| "teacher_step_min": 22.4, |
| "zero_std_ratio": 0.7 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.575, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 19.0, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.08347213044762611, |
| "distill_pairs": 589.0, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.0011660253527226692, |
| "grad_norm": 3.082209825515747, |
| "grpo_loss": 0.0022777387872338294, |
| "kl": 0.056943460181355474, |
| "learning_rate": 2e-06, |
| "loss": 0.0857, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.48250000178813934, |
| "reward_std": 0.06363961100578308, |
| "rewards/countdown_reward_func": 0.48250000178813934, |
| "step": 280, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 0.8 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.8, |
| "correct_distill_rate": 0.0875, |
| "distill_active_tokens": 22.05, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.09772423505783082, |
| "distill_pairs": 683.55, |
| "distill_skipped_no_correct_rate": 0.3, |
| "eligible_block_rate": 0.7, |
| "epoch": 0.0012076691153199075, |
| "grad_norm": 2.1087024211883545, |
| "grpo_loss": 0.0022489023162052036, |
| "kl": 0.05622255988419056, |
| "learning_rate": 2e-06, |
| "loss": 0.1, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.7300000004470348, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.7300000004470348, |
| "step": 290, |
| "teacher_correct_rate": 0.7, |
| "teacher_step_mean": 22.4, |
| "teacher_step_min": 22.4, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.6125, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 19.0, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.07141144275665283, |
| "distill_pairs": 589.0, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.6, |
| "epoch": 0.0012493128779171455, |
| "grad_norm": 0.020935993641614914, |
| "grpo_loss": 0.0024958384805358947, |
| "kl": 0.060207242146134377, |
| "learning_rate": 2e-06, |
| "loss": 0.0739, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.55, |
| "reward_std": 0.08332380652427673, |
| "rewards/countdown_reward_func": 0.55, |
| "step": 300, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 0.8 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 124.0, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 18.55, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.136562230437994, |
| "distill_pairs": 575.05, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.6, |
| "epoch": 0.0012909566405143837, |
| "grad_norm": 2.875196933746338, |
| "grpo_loss": 0.0021033661207184194, |
| "kl": 0.05258412063121796, |
| "learning_rate": 2e-06, |
| "loss": 0.1387, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5950000017881394, |
| "reward_std": 0.04810702204704285, |
| "rewards/countdown_reward_func": 0.5950000017881394, |
| "step": 310, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.125, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.5, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.09108708947896957, |
| "distill_pairs": 480.5, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.001332600403111622, |
| "grad_norm": 0.04492371156811714, |
| "grpo_loss": 0.0026152728823944926, |
| "kl": 0.06538174822926521, |
| "learning_rate": 2e-06, |
| "loss": 0.0937, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.4825000025331974, |
| "reward_std": 0.04166190326213837, |
| "rewards/countdown_reward_func": 0.4825000025331974, |
| "step": 320, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.8, |
| "correct_distill_rate": 0.0375, |
| "distill_active_tokens": 9.375, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.055923382937908175, |
| "distill_pairs": 290.625, |
| "distill_skipped_no_correct_rate": 0.7, |
| "eligible_block_rate": 0.3, |
| "epoch": 0.0013742441657088602, |
| "grad_norm": 0.03447745740413666, |
| "grpo_loss": 0.0023592984769493343, |
| "kl": 0.05898246318101883, |
| "learning_rate": 2e-06, |
| "loss": 0.0583, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.3700000010430813, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.3700000010430813, |
| "step": 330, |
| "teacher_correct_rate": 0.3, |
| "teacher_step_mean": 9.6, |
| "teacher_step_min": 9.6, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.15, |
| "correct_distill_rate": 0.0875, |
| "distill_active_tokens": 22.15, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.11356632858514785, |
| "distill_pairs": 686.65, |
| "distill_skipped_no_correct_rate": 0.3, |
| "eligible_block_rate": 0.7, |
| "epoch": 0.0014158879283060982, |
| "grad_norm": 1.270885705947876, |
| "grpo_loss": 0.002127381146419793, |
| "kl": 0.05428750142455101, |
| "learning_rate": 2e-06, |
| "loss": 0.1157, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.718749999254942, |
| "reward_std": 0.03181980550289154, |
| "rewards/countdown_reward_func": 0.718749999254942, |
| "step": 340, |
| "teacher_correct_rate": 0.7, |
| "teacher_step_mean": 22.4, |
| "teacher_step_min": 22.4, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.2375, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 19.1, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.05533134564757347, |
| "distill_pairs": 592.1, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.6, |
| "epoch": 0.0014575316909033365, |
| "grad_norm": 1.8522121906280518, |
| "grpo_loss": 0.0022492191521450875, |
| "kl": 0.05623047985136509, |
| "learning_rate": 2e-06, |
| "loss": 0.0576, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.6400000005960464, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.6400000005960464, |
| "step": 350, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.875, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.875, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.10500670894980431, |
| "distill_pairs": 492.125, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.0014991754535005747, |
| "grad_norm": 2.168140172958374, |
| "grpo_loss": 0.002285866404417902, |
| "kl": 0.0593788281083107, |
| "learning_rate": 2e-06, |
| "loss": 0.1073, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5387499995529652, |
| "reward_std": 0.03181980550289154, |
| "rewards/countdown_reward_func": 0.5387499995529652, |
| "step": 360, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.7, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.6, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.07073100209236145, |
| "distill_pairs": 390.6, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.001540819216097813, |
| "grad_norm": 4.3752288818359375, |
| "grpo_loss": 0.002176740684080869, |
| "kl": 0.05441855676472187, |
| "learning_rate": 2e-06, |
| "loss": 0.0729, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.3924999997019768, |
| "reward_std": 0.08976892530918121, |
| "rewards/countdown_reward_func": 0.3924999997019768, |
| "step": 370, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 0.8 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.8875, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.6, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.03389553502202034, |
| "distill_pairs": 390.6, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.001582462978695051, |
| "grad_norm": 0.02600081078708172, |
| "grpo_loss": 0.0019671379821375014, |
| "kl": 0.0491784505546093, |
| "learning_rate": 2e-06, |
| "loss": 0.0359, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.46000000089406967, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.46000000089406967, |
| "step": 380, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.6, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.8, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.038029097765684125, |
| "distill_pairs": 396.8, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.0016241067412922892, |
| "grad_norm": 0.030630774796009064, |
| "grpo_loss": 0.0023428802145645023, |
| "kl": 0.058572006225585935, |
| "learning_rate": 2e-06, |
| "loss": 0.0404, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.46000000089406967, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.46000000089406967, |
| "step": 390, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.8, |
| "correct_distill_rate": 0.025, |
| "distill_active_tokens": 6.375, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.03722453340888023, |
| "distill_pairs": 197.625, |
| "distill_skipped_no_correct_rate": 0.8, |
| "eligible_block_rate": 0.2, |
| "epoch": 0.0016657505038895275, |
| "grad_norm": 0.0389198400080204, |
| "grpo_loss": 0.001956176327075809, |
| "kl": 0.04890440851449966, |
| "learning_rate": 2e-06, |
| "loss": 0.0392, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.2800000011920929, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.2800000011920929, |
| "step": 400, |
| "teacher_correct_rate": 0.2, |
| "teacher_step_mean": 6.4, |
| "teacher_step_min": 6.4, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.0, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.775, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.045875276625156405, |
| "distill_pairs": 396.025, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.0017073942664867657, |
| "grad_norm": 0.028869755566120148, |
| "grpo_loss": 0.002037958812434226, |
| "kl": 0.05094897225499153, |
| "learning_rate": 2e-06, |
| "loss": 0.0479, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.46000000089406967, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.46000000089406967, |
| "step": 410, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.725, |
| "correct_distill_rate": 0.0375, |
| "distill_active_tokens": 9.475, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.04254463836550713, |
| "distill_pairs": 293.725, |
| "distill_skipped_no_correct_rate": 0.7, |
| "eligible_block_rate": 0.2, |
| "epoch": 0.0017490380290840037, |
| "grad_norm": 0.04496864229440689, |
| "grpo_loss": 0.0021196508081629873, |
| "kl": 0.05299126654863358, |
| "learning_rate": 2e-06, |
| "loss": 0.0447, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.29125000163912773, |
| "reward_std": 0.03181980550289154, |
| "rewards/countdown_reward_func": 0.29125000163912773, |
| "step": 420, |
| "teacher_correct_rate": 0.2, |
| "teacher_step_mean": 6.4, |
| "teacher_step_min": 6.4, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.4, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.7, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.0385009765625, |
| "distill_pairs": 393.7, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.3, |
| "epoch": 0.001790681791681242, |
| "grad_norm": 0.039158210158348083, |
| "grpo_loss": 0.0019648080691695214, |
| "kl": 0.04912020564079285, |
| "learning_rate": 2e-06, |
| "loss": 0.0405, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.3812500014901161, |
| "reward_std": 0.03181980550289154, |
| "rewards/countdown_reward_func": 0.3812500014901161, |
| "step": 430, |
| "teacher_correct_rate": 0.3, |
| "teacher_step_mean": 9.6, |
| "teacher_step_min": 9.6, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.6375, |
| "correct_distill_rate": 0.025, |
| "distill_active_tokens": 6.4, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.015114924311637879, |
| "distill_pairs": 198.4, |
| "distill_skipped_no_correct_rate": 0.8, |
| "eligible_block_rate": 0.2, |
| "epoch": 0.0018323255542784802, |
| "grad_norm": 1.9908090829849243, |
| "grpo_loss": 0.0018898562062531709, |
| "kl": 0.0472464069724083, |
| "learning_rate": 2e-06, |
| "loss": 0.017, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.2800000011920929, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.2800000011920929, |
| "step": 440, |
| "teacher_correct_rate": 0.2, |
| "teacher_step_mean": 6.4, |
| "teacher_step_min": 6.4, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.4625, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.65, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.09938333183526993, |
| "distill_pairs": 485.15, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.0018739693168757184, |
| "grad_norm": 3.268312454223633, |
| "grpo_loss": 0.0028587776585482063, |
| "kl": 0.05145539902150631, |
| "learning_rate": 2e-06, |
| "loss": 0.1022, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.4712500013411045, |
| "reward_std": 0.03181980550289154, |
| "rewards/countdown_reward_func": 0.4712500013411045, |
| "step": 450, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.7, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 18.975, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.07450196370482445, |
| "distill_pairs": 588.225, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.6, |
| "epoch": 0.0019156130794729565, |
| "grad_norm": 3.746710777282715, |
| "grpo_loss": 0.00205158325843513, |
| "kl": 0.051289581879973414, |
| "learning_rate": 2e-06, |
| "loss": 0.0766, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.6400000005960464, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.6400000005960464, |
| "step": 460, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.6, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.75, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.07725294530391694, |
| "distill_pairs": 488.25, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.001957256842070195, |
| "grad_norm": 0.023627346381545067, |
| "grpo_loss": 0.0019426416838541627, |
| "kl": 0.04856603592634201, |
| "learning_rate": 2e-06, |
| "loss": 0.0792, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.49375000074505804, |
| "reward_std": 0.04657942354679108, |
| "rewards/countdown_reward_func": 0.49375000074505804, |
| "step": 470, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.75, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 18.8, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.05614905580878258, |
| "distill_pairs": 582.8, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.6, |
| "epoch": 0.001998900604667433, |
| "grad_norm": 0.049967773258686066, |
| "grpo_loss": 0.0020881981821730734, |
| "kl": 0.052204956859350206, |
| "learning_rate": 2e-06, |
| "loss": 0.0582, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.6400000005960464, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.6400000005960464, |
| "step": 480, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.8, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 18.95, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.0735462661832571, |
| "distill_pairs": 587.45, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.6, |
| "epoch": 0.002040544367264671, |
| "grad_norm": 0.02936868742108345, |
| "grpo_loss": 0.0021301842760294678, |
| "kl": 0.053254608809947965, |
| "learning_rate": 2e-06, |
| "loss": 0.0757, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.6400000005960464, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.6400000005960464, |
| "step": 490, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.8875, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.65, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.08831988796591758, |
| "distill_pairs": 485.15, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.0020821881298619094, |
| "grad_norm": 5.8246941566467285, |
| "grpo_loss": 0.001958867406938225, |
| "kl": 0.048971685394644734, |
| "learning_rate": 2e-06, |
| "loss": 0.0903, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5500000007450581, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.5500000007450581, |
| "step": 500, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.4, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.7, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.02966131865978241, |
| "distill_pairs": 393.7, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.0021238318924591475, |
| "grad_norm": 3.822479486465454, |
| "grpo_loss": 0.0024712653481401502, |
| "kl": 0.06178163699805737, |
| "learning_rate": 2e-06, |
| "loss": 0.0321, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.46000000089406967, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.46000000089406967, |
| "step": 510, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.1625, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 18.95, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.07386690117418766, |
| "distill_pairs": 587.45, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.0021654756550563855, |
| "grad_norm": 0.030225517228245735, |
| "grpo_loss": 0.002001090080011636, |
| "kl": 0.050027239695191385, |
| "learning_rate": 2e-06, |
| "loss": 0.0759, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5612500011920929, |
| "reward_std": 0.03181980550289154, |
| "rewards/countdown_reward_func": 0.5612500011920929, |
| "step": 520, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.575, |
| "correct_distill_rate": 0.1, |
| "distill_active_tokens": 25.3, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.0799015786498785, |
| "distill_pairs": 784.3, |
| "distill_skipped_no_correct_rate": 0.2, |
| "eligible_block_rate": 0.8, |
| "epoch": 0.002207119417653624, |
| "grad_norm": 1.040562629699707, |
| "grpo_loss": 0.0025180481374263764, |
| "kl": 0.06295120380818844, |
| "learning_rate": 2e-06, |
| "loss": 0.0824, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.8200000002980232, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.8200000002980232, |
| "step": 530, |
| "teacher_correct_rate": 0.8, |
| "teacher_step_mean": 25.6, |
| "teacher_step_min": 25.6, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.0, |
| "correct_distill_rate": 0.0375, |
| "distill_active_tokens": 9.3, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.040867003798484805, |
| "distill_pairs": 288.3, |
| "distill_skipped_no_correct_rate": 0.7, |
| "eligible_block_rate": 0.2, |
| "epoch": 0.002248763180250862, |
| "grad_norm": 8.516297340393066, |
| "grpo_loss": 0.0021306542679667474, |
| "kl": 0.053266337513923644, |
| "learning_rate": 2e-06, |
| "loss": 0.043, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.29125000163912773, |
| "reward_std": 0.03181980550289154, |
| "rewards/countdown_reward_func": 0.29125000163912773, |
| "step": 540, |
| "teacher_correct_rate": 0.2, |
| "teacher_step_mean": 6.4, |
| "teacher_step_min": 6.4, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.3875, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.725, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.08069004192948341, |
| "distill_pairs": 394.475, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.0022904069428481004, |
| "grad_norm": 6.605107307434082, |
| "grpo_loss": 0.0019996401271782814, |
| "kl": 0.049990979582071306, |
| "learning_rate": 2e-06, |
| "loss": 0.0827, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.40375000089406965, |
| "reward_std": 0.04657942354679108, |
| "rewards/countdown_reward_func": 0.40375000089406965, |
| "step": 550, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.7, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 19.075, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.10508804470300674, |
| "distill_pairs": 591.325, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.0023320507054453384, |
| "grad_norm": 2.5305211544036865, |
| "grpo_loss": 0.0020563705125823616, |
| "kl": 0.05140921957790852, |
| "learning_rate": 2e-06, |
| "loss": 0.1071, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5612500041723252, |
| "reward_std": 0.09468644559383392, |
| "rewards/countdown_reward_func": 0.5612500041723252, |
| "step": 560, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 0.8 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.3375, |
| "correct_distill_rate": 0.0875, |
| "distill_active_tokens": 21.975, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.09298599511384964, |
| "distill_pairs": 681.225, |
| "distill_skipped_no_correct_rate": 0.3, |
| "eligible_block_rate": 0.7, |
| "epoch": 0.0023736944680425765, |
| "grad_norm": 8.474016189575195, |
| "grpo_loss": 0.0018267488572746514, |
| "kl": 0.04794607944786548, |
| "learning_rate": 2e-06, |
| "loss": 0.0948, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.6737500004470348, |
| "reward_std": 0.08824132680892945, |
| "rewards/countdown_reward_func": 0.6737500004470348, |
| "step": 570, |
| "teacher_correct_rate": 0.7, |
| "teacher_step_mean": 22.4, |
| "teacher_step_min": 22.4, |
| "zero_std_ratio": 0.8 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.325, |
| "correct_distill_rate": 0.0375, |
| "distill_active_tokens": 9.4, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.035434228181838986, |
| "distill_pairs": 291.4, |
| "distill_skipped_no_correct_rate": 0.7, |
| "eligible_block_rate": 0.3, |
| "epoch": 0.002415338230639815, |
| "grad_norm": 0.031069383025169373, |
| "grpo_loss": 0.001945614768192172, |
| "kl": 0.0486403688788414, |
| "learning_rate": 2e-06, |
| "loss": 0.0374, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.3700000010430813, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.3700000010430813, |
| "step": 580, |
| "teacher_correct_rate": 0.3, |
| "teacher_step_mean": 9.6, |
| "teacher_step_min": 9.6, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.875, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.8, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.04654240682721138, |
| "distill_pairs": 396.8, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.002456981993237053, |
| "grad_norm": 0.043531909584999084, |
| "grpo_loss": 0.0020854957867413757, |
| "kl": 0.05213739536702633, |
| "learning_rate": 2e-06, |
| "loss": 0.0486, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.46000000089406967, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.46000000089406967, |
| "step": 590, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.175, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.8, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.021402059122920036, |
| "distill_pairs": 396.8, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.002498625755834291, |
| "grad_norm": 0.053084440529346466, |
| "grpo_loss": 0.0018724173889495433, |
| "kl": 0.046810435876250264, |
| "learning_rate": 2e-06, |
| "loss": 0.0233, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.46000000089406967, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.46000000089406967, |
| "step": 600, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.2, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.6, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.04758702777326107, |
| "distill_pairs": 390.6, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.0025402695184315294, |
| "grad_norm": 0.01820826716721058, |
| "grpo_loss": 0.0019454847322776913, |
| "kl": 0.04863711819052696, |
| "learning_rate": 2e-06, |
| "loss": 0.0495, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.46000000089406967, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.46000000089406967, |
| "step": 610, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.8125, |
| "correct_distill_rate": 0.0875, |
| "distill_active_tokens": 22.15, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.05205099135637283, |
| "distill_pairs": 686.65, |
| "distill_skipped_no_correct_rate": 0.3, |
| "eligible_block_rate": 0.7, |
| "epoch": 0.0025819132810287674, |
| "grad_norm": 1.40639066696167, |
| "grpo_loss": 0.0020130555029027166, |
| "kl": 0.050326353311538695, |
| "learning_rate": 2e-06, |
| "loss": 0.0541, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.6850000016391278, |
| "reward_std": 0.04810702204704285, |
| "rewards/countdown_reward_func": 0.6850000016391278, |
| "step": 620, |
| "teacher_correct_rate": 0.7, |
| "teacher_step_mean": 22.4, |
| "teacher_step_min": 22.4, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.2875, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 18.825, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.06753165014088154, |
| "distill_pairs": 583.575, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.6, |
| "epoch": 0.002623557043626006, |
| "grad_norm": 2.0643532276153564, |
| "grpo_loss": 0.0020374794374220075, |
| "kl": 0.05093698911368847, |
| "learning_rate": 2e-06, |
| "loss": 0.0696, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5387500017881394, |
| "reward_std": 0.1363483488559723, |
| "rewards/countdown_reward_func": 0.5387500017881394, |
| "step": 630, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 0.7 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 124.4, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.5, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.02291668504476547, |
| "distill_pairs": 387.5, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.3, |
| "epoch": 0.002665200806223244, |
| "grad_norm": 1.110180139541626, |
| "grpo_loss": 0.0022120778798125685, |
| "kl": 0.057579219341278076, |
| "learning_rate": 2e-06, |
| "loss": 0.0251, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.392500002682209, |
| "reward_std": 0.04166190326213837, |
| "rewards/countdown_reward_func": 0.392500002682209, |
| "step": 640, |
| "teacher_correct_rate": 0.3, |
| "teacher_step_mean": 9.6, |
| "teacher_step_min": 9.6, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 124.6, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.35, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.06112914122641087, |
| "distill_pairs": 475.85, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.002706844568820482, |
| "grad_norm": 0.022418711334466934, |
| "grpo_loss": 0.0016894275788217783, |
| "kl": 0.042235690355300906, |
| "learning_rate": 2e-06, |
| "loss": 0.0628, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5500000007450581, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.5500000007450581, |
| "step": 650, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.825, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.7, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.05616593211889267, |
| "distill_pairs": 393.7, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.0027484883314177204, |
| "grad_norm": 0.03228927031159401, |
| "grpo_loss": 0.004499219870194793, |
| "kl": 0.045708934962749484, |
| "learning_rate": 2e-06, |
| "loss": 0.0607, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.40375000089406965, |
| "reward_std": 0.08824132680892945, |
| "rewards/countdown_reward_func": 0.40375000089406965, |
| "step": 660, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 0.8 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.8, |
| "correct_distill_rate": 0.0875, |
| "distill_active_tokens": 22.0, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.0397821057587862, |
| "distill_pairs": 682.0, |
| "distill_skipped_no_correct_rate": 0.3, |
| "eligible_block_rate": 0.7, |
| "epoch": 0.0027901320940149584, |
| "grad_norm": 2.9696037769317627, |
| "grpo_loss": 0.0018949230085127055, |
| "kl": 0.04737307578325271, |
| "learning_rate": 2e-06, |
| "loss": 0.0417, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.7300000004470348, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.7300000004470348, |
| "step": 670, |
| "teacher_correct_rate": 0.7, |
| "teacher_step_mean": 22.4, |
| "teacher_step_min": 22.4, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.8, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.6, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.0299550898373127, |
| "distill_pairs": 390.6, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.0028317758566121965, |
| "grad_norm": 1.5643595457077026, |
| "grpo_loss": 0.0019430788233876228, |
| "kl": 0.04857697077095509, |
| "learning_rate": 2e-06, |
| "loss": 0.0319, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.46000000089406967, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.46000000089406967, |
| "step": 680, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.0, |
| "correct_distill_rate": 0.0375, |
| "distill_active_tokens": 9.4, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.013994240760803222, |
| "distill_pairs": 291.4, |
| "distill_skipped_no_correct_rate": 0.7, |
| "eligible_block_rate": 0.3, |
| "epoch": 0.002873419619209435, |
| "grad_norm": 0.040210120379924774, |
| "grpo_loss": 0.0021145820734091102, |
| "kl": 0.05286455228924751, |
| "learning_rate": 2e-06, |
| "loss": 0.0161, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.3700000010430813, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.3700000010430813, |
| "step": 690, |
| "teacher_correct_rate": 0.3, |
| "teacher_step_mean": 9.6, |
| "teacher_step_min": 9.6, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.7125, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.7, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.062433788925409316, |
| "distill_pairs": 393.7, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.002915063381806673, |
| "grad_norm": 3.5091564655303955, |
| "grpo_loss": 0.0018852821667678654, |
| "kl": 0.04713212437927723, |
| "learning_rate": 2e-06, |
| "loss": 0.0643, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.43749999850988386, |
| "reward_std": 0.04166190326213837, |
| "rewards/countdown_reward_func": 0.43749999850988386, |
| "step": 700, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.825, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.85, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.11187820881605148, |
| "distill_pairs": 398.35, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.0029567071444039114, |
| "grad_norm": 0.0350663997232914, |
| "grpo_loss": 0.0018463104148395359, |
| "kl": 0.04833796359598637, |
| "learning_rate": 2e-06, |
| "loss": 0.1137, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.42625000327825546, |
| "reward_std": 0.04657942354679108, |
| "rewards/countdown_reward_func": 0.42625000327825546, |
| "step": 710, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.15, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.8, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.041693848371505735, |
| "distill_pairs": 489.8, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.0029983509070011494, |
| "grad_norm": 0.9875336289405823, |
| "grpo_loss": 0.0018766400637105108, |
| "kl": 0.04691600389778614, |
| "learning_rate": 2e-06, |
| "loss": 0.0436, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5500000007450581, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.5500000007450581, |
| "step": 720, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.9875, |
| "correct_distill_rate": 0.0875, |
| "distill_active_tokens": 22.0, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.06945269349962473, |
| "distill_pairs": 682.0, |
| "distill_skipped_no_correct_rate": 0.3, |
| "eligible_block_rate": 0.7, |
| "epoch": 0.0030399946695983874, |
| "grad_norm": 0.041114483028650284, |
| "grpo_loss": 0.0018398920190520584, |
| "kl": 0.0459972757846117, |
| "learning_rate": 2e-06, |
| "loss": 0.0713, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.6962500028312206, |
| "reward_std": 0.04657942354679108, |
| "rewards/countdown_reward_func": 0.6962500028312206, |
| "step": 730, |
| "teacher_correct_rate": 0.7, |
| "teacher_step_mean": 22.4, |
| "teacher_step_min": 22.4, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.3625, |
| "correct_distill_rate": 0.0375, |
| "distill_active_tokens": 9.5, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.03527157679200173, |
| "distill_pairs": 294.5, |
| "distill_skipped_no_correct_rate": 0.7, |
| "eligible_block_rate": 0.3, |
| "epoch": 0.003081638432195626, |
| "grad_norm": 0.027781084179878235, |
| "grpo_loss": 0.0019762673880904915, |
| "kl": 0.0494066845625639, |
| "learning_rate": 2e-06, |
| "loss": 0.0372, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.3700000010430813, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.3700000010430813, |
| "step": 740, |
| "teacher_correct_rate": 0.3, |
| "teacher_step_mean": 9.6, |
| "teacher_step_min": 9.6, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.1, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.7, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.014785934984683991, |
| "distill_pairs": 393.7, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.003123282194792864, |
| "grad_norm": 1.2722299098968506, |
| "grpo_loss": 0.0018116761348210275, |
| "kl": 0.04529190361499787, |
| "learning_rate": 2e-06, |
| "loss": 0.0166, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.46000000089406967, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.46000000089406967, |
| "step": 750, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.525, |
| "correct_distill_rate": 0.0375, |
| "distill_active_tokens": 9.4, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.020367415994405745, |
| "distill_pairs": 291.4, |
| "distill_skipped_no_correct_rate": 0.7, |
| "eligible_block_rate": 0.3, |
| "epoch": 0.003164925957390102, |
| "grad_norm": 0.023068469017744064, |
| "grpo_loss": 0.001675213617272675, |
| "kl": 0.041880404204130174, |
| "learning_rate": 2e-06, |
| "loss": 0.022, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.3587499998509884, |
| "reward_std": 0.03181980550289154, |
| "rewards/countdown_reward_func": 0.3587499998509884, |
| "step": 760, |
| "teacher_correct_rate": 0.3, |
| "teacher_step_mean": 9.6, |
| "teacher_step_min": 9.6, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.0125, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 19.175, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.1046835258603096, |
| "distill_pairs": 594.425, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.6, |
| "epoch": 0.0032065697199873404, |
| "grad_norm": 2.8684561252593994, |
| "grpo_loss": 0.0018690836033783853, |
| "kl": 0.04672709070146084, |
| "learning_rate": 2e-06, |
| "loss": 0.1066, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.6400000005960464, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.6400000005960464, |
| "step": 770, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.1125, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.9, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.0902952967211604, |
| "distill_pairs": 492.9, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.0032482134825845784, |
| "grad_norm": 0.051981739699840546, |
| "grpo_loss": 0.0019321527099236846, |
| "kl": 0.04708785824477672, |
| "learning_rate": 2e-06, |
| "loss": 0.0922, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5375000007450581, |
| "reward_std": 0.03535533845424652, |
| "rewards/countdown_reward_func": 0.5375000007450581, |
| "step": 780, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.8875, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.8, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.088916876912117, |
| "distill_pairs": 396.8, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.3, |
| "epoch": 0.0032898572451818165, |
| "grad_norm": 4.031944274902344, |
| "grpo_loss": 0.0018385385279543697, |
| "kl": 0.050358007475733754, |
| "learning_rate": 2e-06, |
| "loss": 0.0908, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.3700000002980232, |
| "reward_std": 0.06363961100578308, |
| "rewards/countdown_reward_func": 0.3700000002980232, |
| "step": 790, |
| "teacher_correct_rate": 0.3, |
| "teacher_step_mean": 9.6, |
| "teacher_step_min": 9.6, |
| "zero_std_ratio": 0.8 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.0, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.7, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.02560471072793007, |
| "distill_pairs": 486.7, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.003331501007779055, |
| "grad_norm": 0.027767013758420944, |
| "grpo_loss": 0.001708028349094093, |
| "kl": 0.04270070940256119, |
| "learning_rate": 2e-06, |
| "loss": 0.0273, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5500000007450581, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.5500000007450581, |
| "step": 800, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.55, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.7, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.026517413929104804, |
| "distill_pairs": 486.7, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.003373144770376293, |
| "grad_norm": 1.448878526687622, |
| "grpo_loss": 0.0016887560370378197, |
| "kl": 0.042218901216983795, |
| "learning_rate": 2e-06, |
| "loss": 0.0282, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5500000007450581, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.5500000007450581, |
| "step": 810, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.1, |
| "correct_distill_rate": 0.1, |
| "distill_active_tokens": 25.4, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.0639851501211524, |
| "distill_pairs": 787.4, |
| "distill_skipped_no_correct_rate": 0.2, |
| "eligible_block_rate": 0.8, |
| "epoch": 0.0034147885329735314, |
| "grad_norm": 0.05175022408366203, |
| "grpo_loss": 0.001742440788075328, |
| "kl": 0.043561020866036415, |
| "learning_rate": 2e-06, |
| "loss": 0.0657, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.8200000002980232, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.8200000002980232, |
| "step": 820, |
| "teacher_correct_rate": 0.8, |
| "teacher_step_mean": 25.6, |
| "teacher_step_min": 25.6, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.1875, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.75, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.07595743983983994, |
| "distill_pairs": 488.25, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.0034564322955707694, |
| "grad_norm": 0.033084504306316376, |
| "grpo_loss": 0.0018853740533813835, |
| "kl": 0.04713435247540474, |
| "learning_rate": 2e-06, |
| "loss": 0.0778, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5500000007450581, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.5500000007450581, |
| "step": 830, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.775, |
| "correct_distill_rate": 0.0875, |
| "distill_active_tokens": 22.15, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.06020920127630234, |
| "distill_pairs": 686.65, |
| "distill_skipped_no_correct_rate": 0.3, |
| "eligible_block_rate": 0.7, |
| "epoch": 0.0034980760581680074, |
| "grad_norm": 0.03143149986863136, |
| "grpo_loss": 0.0017725960467942059, |
| "kl": 0.04652954116463661, |
| "learning_rate": 2e-06, |
| "loss": 0.062, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.707499998062849, |
| "reward_std": 0.04166190326213837, |
| "rewards/countdown_reward_func": 0.707499998062849, |
| "step": 840, |
| "teacher_correct_rate": 0.7, |
| "teacher_step_mean": 22.4, |
| "teacher_step_min": 22.4, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.2, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.65, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.08109580650925637, |
| "distill_pairs": 485.15, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.003539719820765246, |
| "grad_norm": 7.128382205963135, |
| "grpo_loss": 0.0017045807209797204, |
| "kl": 0.0426146661862731, |
| "learning_rate": 2e-06, |
| "loss": 0.0828, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.43749999776482584, |
| "reward_std": 0.10530151426792145, |
| "rewards/countdown_reward_func": 0.43749999776482584, |
| "step": 850, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 0.7 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.4, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 18.65, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.041441801376640795, |
| "distill_pairs": 578.15, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.6, |
| "epoch": 0.003581363583362484, |
| "grad_norm": 2.0089380741119385, |
| "grpo_loss": 0.00152717997552827, |
| "kl": 0.03817950077354908, |
| "learning_rate": 2e-06, |
| "loss": 0.043, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.6400000005960464, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.6400000005960464, |
| "step": 860, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 124.2, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.7, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.019977014884352683, |
| "distill_pairs": 486.7, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.003623007345959722, |
| "grad_norm": 0.9501437544822693, |
| "grpo_loss": 0.0017192380037158728, |
| "kl": 0.042980996891856194, |
| "learning_rate": 2e-06, |
| "loss": 0.0217, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5387499995529652, |
| "reward_std": 0.03181980550289154, |
| "rewards/countdown_reward_func": 0.5387499995529652, |
| "step": 870, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.4, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.7, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.014679186046123505, |
| "distill_pairs": 393.7, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.0036646511085569604, |
| "grad_norm": 1.2250865697860718, |
| "grpo_loss": 0.0017130593187175691, |
| "kl": 0.04282648488879204, |
| "learning_rate": 2e-06, |
| "loss": 0.0164, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.46000000089406967, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.46000000089406967, |
| "step": 880, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.3, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 18.975, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.029483658075332642, |
| "distill_pairs": 588.225, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.6, |
| "epoch": 0.0037062948711541984, |
| "grad_norm": 1.2688686847686768, |
| "grpo_loss": 0.0019477210007607937, |
| "kl": 0.048693026229739186, |
| "learning_rate": 2e-06, |
| "loss": 0.0314, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.6400000005960464, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.6400000005960464, |
| "step": 890, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.2625, |
| "correct_distill_rate": 0.0375, |
| "distill_active_tokens": 9.525, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.06054073497653008, |
| "distill_pairs": 295.275, |
| "distill_skipped_no_correct_rate": 0.7, |
| "eligible_block_rate": 0.3, |
| "epoch": 0.003747938633751437, |
| "grad_norm": 0.02274780534207821, |
| "grpo_loss": 0.001566296920645982, |
| "kl": 0.04244049154222011, |
| "learning_rate": 2e-06, |
| "loss": 0.0621, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.31375000104308126, |
| "reward_std": 0.04657942354679108, |
| "rewards/countdown_reward_func": 0.31375000104308126, |
| "step": 900, |
| "teacher_correct_rate": 0.3, |
| "teacher_step_mean": 9.6, |
| "teacher_step_min": 9.6, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.625, |
| "correct_distill_rate": 0.0875, |
| "distill_active_tokens": 22.15, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.07372288815677167, |
| "distill_pairs": 686.65, |
| "distill_skipped_no_correct_rate": 0.3, |
| "eligible_block_rate": 0.7, |
| "epoch": 0.003789582396348675, |
| "grad_norm": 1.4420886039733887, |
| "grpo_loss": 0.0019108039909042419, |
| "kl": 0.04556427001953125, |
| "learning_rate": 2e-06, |
| "loss": 0.0756, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.707499998062849, |
| "reward_std": 0.04166190326213837, |
| "rewards/countdown_reward_func": 0.707499998062849, |
| "step": 910, |
| "teacher_correct_rate": 0.7, |
| "teacher_step_mean": 22.4, |
| "teacher_step_min": 22.4, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.8, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 18.9, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.05120294746011496, |
| "distill_pairs": 585.9, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.6, |
| "epoch": 0.003831226158945913, |
| "grad_norm": 0.07518143206834793, |
| "grpo_loss": 0.0017321450752206147, |
| "kl": 0.04330362752079964, |
| "learning_rate": 2e-06, |
| "loss": 0.0529, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.6400000005960464, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.6400000005960464, |
| "step": 920, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.625, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.5, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.015241103060543538, |
| "distill_pairs": 387.5, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.0038728699215431514, |
| "grad_norm": 1.555560827255249, |
| "grpo_loss": 0.0016179736121557653, |
| "kl": 0.04044934064149856, |
| "learning_rate": 2e-06, |
| "loss": 0.0169, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.46000000089406967, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.46000000089406967, |
| "step": 930, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.325, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.925, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.07077029719948769, |
| "distill_pairs": 493.675, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.00391451368414039, |
| "grad_norm": 3.617987632751465, |
| "grpo_loss": 0.001815111073665321, |
| "kl": 0.04756661541759968, |
| "learning_rate": 2e-06, |
| "loss": 0.0726, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5162499971687794, |
| "reward_std": 0.07348170876502991, |
| "rewards/countdown_reward_func": 0.5162499971687794, |
| "step": 940, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 0.8 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.4, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.8, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.017609160020947458, |
| "distill_pairs": 489.8, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.003956157446737628, |
| "grad_norm": 1.1723670959472656, |
| "grpo_loss": 0.001807892112992704, |
| "kl": 0.045197305455803874, |
| "learning_rate": 2e-06, |
| "loss": 0.0194, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5500000007450581, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.5500000007450581, |
| "step": 950, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.95, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.7, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.010681399516761303, |
| "distill_pairs": 393.7, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.003997801209334866, |
| "grad_norm": 0.029792888090014458, |
| "grpo_loss": 0.0015423979260958732, |
| "kl": 0.0385599497705698, |
| "learning_rate": 2e-06, |
| "loss": 0.0122, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.46000000089406967, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.46000000089406967, |
| "step": 960, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.475, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 18.95, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.08463714867830277, |
| "distill_pairs": 587.45, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.6, |
| "epoch": 0.004039444971932104, |
| "grad_norm": 0.05133519321680069, |
| "grpo_loss": 0.0014427412359509617, |
| "kl": 0.03834581784904003, |
| "learning_rate": 2e-06, |
| "loss": 0.0861, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5950000017881394, |
| "reward_std": 0.04810702204704285, |
| "rewards/countdown_reward_func": 0.5950000017881394, |
| "step": 970, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.4875, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.75, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.051837844960391524, |
| "distill_pairs": 488.25, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.004081088734529342, |
| "grad_norm": 0.059578362852334976, |
| "grpo_loss": 0.0016698208870366215, |
| "kl": 0.041745524108409884, |
| "learning_rate": 2e-06, |
| "loss": 0.0535, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5500000007450581, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.5500000007450581, |
| "step": 980, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.125, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.7, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.05289398673921823, |
| "distill_pairs": 486.7, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.00412273249712658, |
| "grad_norm": 8.03367805480957, |
| "grpo_loss": 0.001515911938622594, |
| "kl": 0.037897751480340955, |
| "learning_rate": 2e-06, |
| "loss": 0.0544, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5162500031292439, |
| "reward_std": 0.04657942354679108, |
| "rewards/countdown_reward_func": 0.5162500031292439, |
| "step": 990, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.7, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.925, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.08010751288384199, |
| "distill_pairs": 493.675, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.004164376259723819, |
| "grad_norm": 0.023414593189954758, |
| "grpo_loss": 0.0016792724374681712, |
| "kl": 0.04198181107640266, |
| "learning_rate": 2e-06, |
| "loss": 0.0818, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5500000007450581, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.5500000007450581, |
| "step": 1000, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.7875, |
| "correct_distill_rate": 0.0375, |
| "distill_active_tokens": 9.6, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.012844983115792274, |
| "distill_pairs": 297.6, |
| "distill_skipped_no_correct_rate": 0.7, |
| "eligible_block_rate": 0.3, |
| "epoch": 0.004206020022321057, |
| "grad_norm": 0.030537134036421776, |
| "grpo_loss": 0.0016840341733768583, |
| "kl": 0.04210085496306419, |
| "learning_rate": 2e-06, |
| "loss": 0.0145, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.3700000010430813, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.3700000010430813, |
| "step": 1010, |
| "teacher_correct_rate": 0.3, |
| "teacher_step_mean": 9.6, |
| "teacher_step_min": 9.6, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.8, |
| "correct_distill_rate": 0.0875, |
| "distill_active_tokens": 22.3, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.026153293903917074, |
| "distill_pairs": 691.3, |
| "distill_skipped_no_correct_rate": 0.3, |
| "eligible_block_rate": 0.7, |
| "epoch": 0.004247663784918295, |
| "grad_norm": 0.318440318107605, |
| "grpo_loss": 0.0016528526786714793, |
| "kl": 0.041321360133588315, |
| "learning_rate": 2e-06, |
| "loss": 0.0278, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.718749999254942, |
| "reward_std": 0.03181980550289154, |
| "rewards/countdown_reward_func": 0.718749999254942, |
| "step": 1020, |
| "teacher_correct_rate": 0.7, |
| "teacher_step_mean": 22.4, |
| "teacher_step_min": 22.4, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.675, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 19.0, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.012605499103665352, |
| "distill_pairs": 589.0, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.6, |
| "epoch": 0.004289307547515533, |
| "grad_norm": 0.03095727600157261, |
| "grpo_loss": 0.0018430021358653904, |
| "kl": 0.0460751049220562, |
| "learning_rate": 2e-06, |
| "loss": 0.0144, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5724999994039536, |
| "reward_std": 0.08976892530918121, |
| "rewards/countdown_reward_func": 0.5724999994039536, |
| "step": 1030, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 0.8 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.2, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 18.95, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.057916640490293506, |
| "distill_pairs": 587.45, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.004330951310112771, |
| "grad_norm": 3.0107522010803223, |
| "grpo_loss": 0.001734669110737741, |
| "kl": 0.043366647139191625, |
| "learning_rate": 2e-06, |
| "loss": 0.0597, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5725000023841857, |
| "reward_std": 0.04166190326213837, |
| "rewards/countdown_reward_func": 0.5725000023841857, |
| "step": 1040, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.3, |
| "correct_distill_rate": 0.0375, |
| "distill_active_tokens": 9.375, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.02286108545958996, |
| "distill_pairs": 290.625, |
| "distill_skipped_no_correct_rate": 0.7, |
| "eligible_block_rate": 0.3, |
| "epoch": 0.00437259507271001, |
| "grad_norm": 0.05123801529407501, |
| "grpo_loss": 0.0016572008840739726, |
| "kl": 0.04143002331256866, |
| "learning_rate": 2e-06, |
| "loss": 0.0245, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.3700000010430813, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.3700000010430813, |
| "step": 1050, |
| "teacher_correct_rate": 0.3, |
| "teacher_step_mean": 9.6, |
| "teacher_step_min": 9.6, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.1125, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.45, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.06458345837891102, |
| "distill_pairs": 385.95, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.004414238835307248, |
| "grad_norm": 0.025315623730421066, |
| "grpo_loss": 0.0016826133127324284, |
| "kl": 0.04206533432006836, |
| "learning_rate": 2e-06, |
| "loss": 0.0663, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.46000000089406967, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.46000000089406967, |
| "step": 1060, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.9, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.575, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.04755713753402233, |
| "distill_pairs": 389.825, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.004455882597904486, |
| "grad_norm": 1.6858030557632446, |
| "grpo_loss": 0.0017302519991062582, |
| "kl": 0.04325630068778992, |
| "learning_rate": 2e-06, |
| "loss": 0.0493, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.46000000089406967, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.46000000089406967, |
| "step": 1070, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.9, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.675, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.0340701088309288, |
| "distill_pairs": 392.925, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.004497526360501724, |
| "grad_norm": 0.6195261478424072, |
| "grpo_loss": 0.0016782286344096064, |
| "kl": 0.04195571616292, |
| "learning_rate": 2e-06, |
| "loss": 0.0357, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.46000000089406967, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.46000000089406967, |
| "step": 1080, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.6, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 19.1, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.023949543945491314, |
| "distill_pairs": 592.1, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.6, |
| "epoch": 0.004539170123098962, |
| "grad_norm": 0.6692307591438293, |
| "grpo_loss": 0.001612464804202318, |
| "kl": 0.040311620384454724, |
| "learning_rate": 2e-06, |
| "loss": 0.0256, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.6400000005960464, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.6400000005960464, |
| "step": 1090, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.9, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.75, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.04136073831468821, |
| "distill_pairs": 395.25, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.004580813885696201, |
| "grad_norm": 0.020976591855287552, |
| "grpo_loss": 0.001682728691957891, |
| "kl": 0.04206821881234646, |
| "learning_rate": 2e-06, |
| "loss": 0.043, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.46000000089406967, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.46000000089406967, |
| "step": 1100, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.5125, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 18.95, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.039055611938238144, |
| "distill_pairs": 587.45, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.6, |
| "epoch": 0.004622457648293439, |
| "grad_norm": 2.2283780574798584, |
| "grpo_loss": 0.0015541882952675224, |
| "kl": 0.043409249372780324, |
| "learning_rate": 2e-06, |
| "loss": 0.0406, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5275000035762787, |
| "reward_std": 0.08976892530918121, |
| "rewards/countdown_reward_func": 0.5275000035762787, |
| "step": 1110, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 0.8 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.6125, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.6, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.02048632614314556, |
| "distill_pairs": 390.6, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.004664101410890677, |
| "grad_norm": 0.023302381858229637, |
| "grpo_loss": 0.0015883660409599543, |
| "kl": 0.039709235355257985, |
| "learning_rate": 2e-06, |
| "loss": 0.0221, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.426249997317791, |
| "reward_std": 0.07348170876502991, |
| "rewards/countdown_reward_func": 0.426249997317791, |
| "step": 1120, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 0.8 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.1375, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 19.175, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.09878965616226196, |
| "distill_pairs": 594.425, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.6, |
| "epoch": 0.004705745173487915, |
| "grad_norm": 20.61865997314453, |
| "grpo_loss": 0.0016930088866502046, |
| "kl": 0.04232514686882496, |
| "learning_rate": 2e-06, |
| "loss": 0.1005, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5725000023841857, |
| "reward_std": 0.04166190326213837, |
| "rewards/countdown_reward_func": 0.5725000023841857, |
| "step": 1130, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.575, |
| "correct_distill_rate": 0.0625, |
| "distill_active_tokens": 15.9, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.011602642387151718, |
| "distill_pairs": 492.9, |
| "distill_skipped_no_correct_rate": 0.5, |
| "eligible_block_rate": 0.5, |
| "epoch": 0.004747388936085153, |
| "grad_norm": 0.02838357351720333, |
| "grpo_loss": 0.001829492545221001, |
| "kl": 0.04573731385171413, |
| "learning_rate": 2e-06, |
| "loss": 0.0134, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.5500000007450581, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.5500000007450581, |
| "step": 1140, |
| "teacher_correct_rate": 0.5, |
| "teacher_step_mean": 16.0, |
| "teacher_step_min": 16.0, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.3375, |
| "correct_distill_rate": 0.0875, |
| "distill_active_tokens": 22.175, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.035655363276600835, |
| "distill_pairs": 687.425, |
| "distill_skipped_no_correct_rate": 0.3, |
| "eligible_block_rate": 0.7, |
| "epoch": 0.004789032698682391, |
| "grad_norm": 1.563031792640686, |
| "grpo_loss": 0.001887683931272477, |
| "kl": 0.04388321377336979, |
| "learning_rate": 2e-06, |
| "loss": 0.0375, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.6962500028312206, |
| "reward_std": 0.04657942354679108, |
| "rewards/countdown_reward_func": 0.6962500028312206, |
| "step": 1150, |
| "teacher_correct_rate": 0.7, |
| "teacher_step_mean": 22.4, |
| "teacher_step_min": 22.4, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 126.4, |
| "correct_distill_rate": 0.0875, |
| "distill_active_tokens": 22.1, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.01679153088480234, |
| "distill_pairs": 685.1, |
| "distill_skipped_no_correct_rate": 0.3, |
| "eligible_block_rate": 0.7, |
| "epoch": 0.00483067646127963, |
| "grad_norm": 0.4330219030380249, |
| "grpo_loss": 0.0016536793322302402, |
| "kl": 0.041342003270983696, |
| "learning_rate": 2e-06, |
| "loss": 0.0184, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.7175000004470349, |
| "reward_std": 0.03535533845424652, |
| "rewards/countdown_reward_func": 0.7175000004470349, |
| "step": 1160, |
| "teacher_correct_rate": 0.7, |
| "teacher_step_mean": 22.4, |
| "teacher_step_min": 22.4, |
| "zero_std_ratio": 0.9 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.95, |
| "correct_distill_rate": 0.05, |
| "distill_active_tokens": 12.5, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.008152574859559536, |
| "distill_pairs": 387.5, |
| "distill_skipped_no_correct_rate": 0.6, |
| "eligible_block_rate": 0.4, |
| "epoch": 0.004872320223876868, |
| "grad_norm": 0.0251710657030344, |
| "grpo_loss": 0.001681314269080758, |
| "kl": 0.04203285649418831, |
| "learning_rate": 2e-06, |
| "loss": 0.0098, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.46000000089406967, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.46000000089406967, |
| "step": 1170, |
| "teacher_correct_rate": 0.4, |
| "teacher_step_mean": 12.8, |
| "teacher_step_min": 12.8, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 125.6375, |
| "correct_distill_rate": 0.025, |
| "distill_active_tokens": 6.175, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.044068896770477296, |
| "distill_pairs": 191.425, |
| "distill_skipped_no_correct_rate": 0.8, |
| "eligible_block_rate": 0.2, |
| "epoch": 0.004913963986474106, |
| "grad_norm": 0.025189882144331932, |
| "grpo_loss": 0.0016091925092041493, |
| "kl": 0.04022981449961662, |
| "learning_rate": 2e-06, |
| "loss": 0.0457, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.2800000011920929, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.2800000011920929, |
| "step": 1180, |
| "teacher_correct_rate": 0.2, |
| "teacher_step_mean": 6.4, |
| "teacher_step_min": 6.4, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.2, |
| "correct_distill_rate": 0.075, |
| "distill_active_tokens": 19.1, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.0418258648365736, |
| "distill_pairs": 592.1, |
| "distill_skipped_no_correct_rate": 0.4, |
| "eligible_block_rate": 0.6, |
| "epoch": 0.004955607749071344, |
| "grad_norm": 0.02630944736301899, |
| "grpo_loss": 0.001519427145831287, |
| "kl": 0.03798567913472652, |
| "learning_rate": 2e-06, |
| "loss": 0.0433, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.6400000005960464, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.6400000005960464, |
| "step": 1190, |
| "teacher_correct_rate": 0.6, |
| "teacher_step_mean": 19.2, |
| "teacher_step_min": 19.2, |
| "zero_std_ratio": 1.0 |
| }, |
| { |
| "clip_ratio": 0.0, |
| "completion_length": 127.5, |
| "correct_distill_rate": 0.0375, |
| "distill_active_tokens": 9.6, |
| "distill_lambda": 1.0, |
| "distill_loss": 0.005412598140537739, |
| "distill_pairs": 297.6, |
| "distill_skipped_no_correct_rate": 0.7, |
| "eligible_block_rate": 0.3, |
| "epoch": 0.004997251511668582, |
| "grad_norm": 0.8026695251464844, |
| "grpo_loss": 0.0014662531320936978, |
| "kl": 0.036656329222023484, |
| "learning_rate": 2e-06, |
| "loss": 0.0069, |
| "productive_jump_progress_mean": 0.0, |
| "productive_jump_score_mean": 0.0, |
| "productive_jump_selected_rate": 0.0, |
| "productive_jump_span_mean": 0.0, |
| "productive_jump_tail_entropy_mean": 0.0, |
| "reward": 0.3700000010430813, |
| "reward_std": 0.0, |
| "rewards/countdown_reward_func": 0.3700000010430813, |
| "step": 1200, |
| "teacher_correct_rate": 0.3, |
| "teacher_step_mean": 9.6, |
| "teacher_step_min": 9.6, |
| "zero_std_ratio": 1.0 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 1200, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 1, |
| "save_steps": 200, |
| "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": 2, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
