{"step": 0, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.8160546875, "mean_total_reward": -0.2160546875, "std_total_reward": 0.6499251853729252, "total_tokens": 83564, "mean_completion_length": 1671.28, "median_completion_length": 1812, "min_completion_length": 992, "max_completion_length": 2048, "frac_has_think": 0.7, "frac_has_boxed": 0.7, "mean_reward_std_per_prompt": 0.2667112492320983, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": null, "avg_step_time_sec": null, "gpu_allocated_gb": null, "gpu_reserved_gb": null, "gpu_peak_gb": null, "elapsed_sec": null, "eta_sec": null, "timestamp": "2026-02-15T11:33:11.590004"} {"step": 1, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.22, "mean_correctness_reward": 0.22, "mean_length_penalty": -0.94798828125, "mean_total_reward": -0.72798828125, "std_total_reward": 0.4868252299068701, "total_tokens": 97074, "mean_completion_length": 1941.48, "median_completion_length": 2048, "min_completion_length": 1079, "max_completion_length": 2048, "frac_has_think": 0.46, "frac_has_boxed": 0.46, "mean_reward_std_per_prompt": 0.31793526641545145, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 27.895130157470703, "avg_step_time_sec": 27.895130157470703, "gpu_allocated_gb": 32.365621248, "gpu_reserved_gb": 79.92246272, "gpu_peak_gb": 73.906566656, "elapsed_sec": 27.90002417564392, "eta_sec": 5551.13090133667, "timestamp": "2026-02-15T11:33:41.060132"} {"step": 2, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.78, "mean_correctness_reward": 0.78, "mean_length_penalty": -0.795869140625, "mean_total_reward": -0.015869140625, "std_total_reward": 0.5432612294152535, "total_tokens": 81497, "mean_completion_length": 1629.94, "median_completion_length": 1713, "min_completion_length": 809, "max_completion_length": 2048, "frac_has_think": 0.86, "frac_has_boxed": 0.86, "mean_reward_std_per_prompt": 0.4748324743306036, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 29.455275058746338, "avg_step_time_sec": 28.67520260810852, "gpu_allocated_gb": 32.365654016, "gpu_reserved_gb": 79.95392, "gpu_peak_gb": 74.127090176, "elapsed_sec": 57.3616578578949, "eta_sec": 5677.690116405487, "timestamp": "2026-02-15T11:34:08.743714"} {"step": 3, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.74, "mean_correctness_reward": 0.74, "mean_length_penalty": -0.770419921875, "mean_total_reward": -0.030419921875, "std_total_reward": 0.5936957424871508, "total_tokens": 78891, "mean_completion_length": 1577.82, "median_completion_length": 1617, "min_completion_length": 655, "max_completion_length": 2048, "frac_has_think": 0.76, "frac_has_boxed": 0.76, "mean_reward_std_per_prompt": 0.2356193544143079, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 27.598248720169067, "avg_step_time_sec": 28.31621797879537, "gpu_allocated_gb": 32.365626368, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.127090176, "elapsed_sec": 84.96695590019226, "eta_sec": 5578.294941822687, "timestamp": "2026-02-15T11:34:36.612027"} {"step": 4, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.22, "mean_correctness_reward": 0.22, "mean_length_penalty": -0.937734375, "mean_total_reward": -0.717734375, "std_total_reward": 0.5349577450331449, "total_tokens": 96024, "mean_completion_length": 1920.48, "median_completion_length": 2048, "min_completion_length": 1134, "max_completion_length": 2048, "frac_has_think": 0.26, "frac_has_boxed": 0.26, "mean_reward_std_per_prompt": 0.23754178444649257, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.6, "step_time_sec": 27.87776780128479, "avg_step_time_sec": 28.206605434417725, "gpu_allocated_gb": 32.365638656, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.127090176, "elapsed_sec": 112.85106658935547, "eta_sec": 5528.494665145874, "timestamp": "2026-02-15T11:35:05.596114"} {"step": 5, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.54, "mean_correctness_reward": 0.54, "mean_length_penalty": -0.797822265625, "mean_total_reward": -0.257822265625, "std_total_reward": 0.7000454989360313, "total_tokens": 81697, "mean_completion_length": 1633.94, "median_completion_length": 1947, "min_completion_length": 701, "max_completion_length": 2048, "frac_has_think": 0.6, "frac_has_boxed": 0.6, "mean_reward_std_per_prompt": 0.33186226805859126, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 29.506404161453247, "avg_step_time_sec": 28.46656517982483, "gpu_allocated_gb": 32.365667328, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.156304384, "elapsed_sec": 142.36412286758423, "eta_sec": 5550.980210065842, "timestamp": "2026-02-15T11:35:33.971004"} {"step": 6, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.66, "mean_correctness_reward": 0.66, "mean_length_penalty": -0.8316015625, "mean_total_reward": -0.1716015625, "std_total_reward": 0.6099003266370717, "total_tokens": 85156, "mean_completion_length": 1703.12, "median_completion_length": 1831, "min_completion_length": 786, "max_completion_length": 2048, "frac_has_think": 0.8, "frac_has_boxed": 0.8, "mean_reward_std_per_prompt": 0.5046512660097489, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 27.82741069793701, "avg_step_time_sec": 28.360039432843525, "gpu_allocated_gb": 32.365632512, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.156304384, "elapsed_sec": 170.19834899902344, "eta_sec": 5501.8476499716435, "timestamp": "2026-02-15T11:36:02.067425"} {"step": 7, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.38, "mean_correctness_reward": 0.38, "mean_length_penalty": -0.873984375, "mean_total_reward": -0.493984375, "std_total_reward": 0.6557377037439972, "total_tokens": 89496, "mean_completion_length": 1789.92, "median_completion_length": 2048, "min_completion_length": 663, "max_completion_length": 2048, "frac_has_think": 0.48, "frac_has_boxed": 0.48, "mean_reward_std_per_prompt": 0.11828978718448858, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.6, "step_time_sec": 28.08332586288452, "avg_step_time_sec": 28.32050892284938, "gpu_allocated_gb": 32.36562432, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.156304384, "elapsed_sec": 198.28924441337585, "eta_sec": 5465.8582221099305, "timestamp": "2026-02-15T11:36:30.941797"} {"step": 8, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.52, "mean_correctness_reward": 0.52, "mean_length_penalty": -0.81548828125, "mean_total_reward": -0.29548828125, "std_total_reward": 0.6218682474507763, "total_tokens": 83506, "mean_completion_length": 1670.12, "median_completion_length": 1729, "min_completion_length": 750, "max_completion_length": 2048, "frac_has_think": 0.76, "frac_has_boxed": 0.76, "mean_reward_std_per_prompt": 0.43986133642550646, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 28.73879623413086, "avg_step_time_sec": 28.372794836759567, "gpu_allocated_gb": 32.365616128, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.156304384, "elapsed_sec": 227.0348563194275, "eta_sec": 5447.576608657837, "timestamp": "2026-02-15T11:36:58.756452"} {"step": 9, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.72, "mean_correctness_reward": 0.72, "mean_length_penalty": -0.70578125, "mean_total_reward": 0.01421875, "std_total_reward": 0.6304753121149165, "total_tokens": 72272, "mean_completion_length": 1445.44, "median_completion_length": 1526, "min_completion_length": 621, "max_completion_length": 2048, "frac_has_think": 0.82, "frac_has_boxed": 0.82, "mean_reward_std_per_prompt": 0.29746215175230795, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 27.96594548225403, "avg_step_time_sec": 28.327589352925617, "gpu_allocated_gb": 32.365641728, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.156304384, "elapsed_sec": 255.01069450378418, "eta_sec": 5410.5695664087925, "timestamp": "2026-02-15T11:37:26.508605"} {"step": 10, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.52, "mean_correctness_reward": 0.52, "mean_length_penalty": -0.91453125, "mean_total_reward": -0.39453125, "std_total_reward": 0.5869108208235144, "total_tokens": 93648, "mean_completion_length": 1872.96, "median_completion_length": 1947, "min_completion_length": 1080, "max_completion_length": 2048, "frac_has_think": 0.7, "frac_has_boxed": 0.7, "mean_reward_std_per_prompt": 0.4797255706256569, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 28.081921577453613, "avg_step_time_sec": 28.303022575378417, "gpu_allocated_gb": 32.365712384, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 283.10052490234375, "eta_sec": 5377.574289321899, "timestamp": "2026-02-15T11:37:55.692820"} {"step": 11, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.83880859375, "mean_total_reward": -0.23880859375, "std_total_reward": 0.6303528015839397, "total_tokens": 85894, "mean_completion_length": 1717.88, "median_completion_length": 1828, "min_completion_length": 845, "max_completion_length": 2048, "frac_has_think": 0.64, "frac_has_boxed": 0.66, "mean_reward_std_per_prompt": 0.45607858102872345, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 28.838776111602783, "avg_step_time_sec": 28.351727442307904, "gpu_allocated_gb": 32.36563456, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 311.94597601890564, "eta_sec": 5358.476486596194, "timestamp": "2026-02-15T11:38:24.370395"} {"step": 12, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.52, "mean_correctness_reward": 0.52, "mean_length_penalty": -0.80587890625, "mean_total_reward": -0.28587890625, "std_total_reward": 0.6024453735913499, "total_tokens": 82522, "mean_completion_length": 1650.44, "median_completion_length": 1663, "min_completion_length": 922, "max_completion_length": 2048, "frac_has_think": 0.72, "frac_has_boxed": 0.72, "mean_reward_std_per_prompt": 0.35063350203155275, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 28.66943073272705, "avg_step_time_sec": 28.378202716509502, "gpu_allocated_gb": 32.365643776, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 340.62236499786377, "eta_sec": 5335.102110703787, "timestamp": "2026-02-15T11:38:52.271447"} {"step": 13, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.54, "mean_correctness_reward": 0.54, "mean_length_penalty": -0.85796875, "mean_total_reward": -0.31796875, "std_total_reward": 0.634743982503643, "total_tokens": 87856, "mean_completion_length": 1757.12, "median_completion_length": 2048, "min_completion_length": 906, "max_completion_length": 2048, "frac_has_think": 0.6, "frac_has_boxed": 0.6, "mean_reward_std_per_prompt": 0.45123571367514254, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 27.895596265792847, "avg_step_time_sec": 28.34107914337745, "gpu_allocated_gb": 32.365641728, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 368.5245831012726, "eta_sec": 5299.781799811583, "timestamp": "2026-02-15T11:39:21.158828"} {"step": 14, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.5, "mean_correctness_reward": 0.5, "mean_length_penalty": -0.787451171875, "mean_total_reward": -0.287451171875, "std_total_reward": 0.6197439841085798, "total_tokens": 80635, "mean_completion_length": 1612.7, "median_completion_length": 1617, "min_completion_length": 833, "max_completion_length": 2048, "frac_has_think": 0.82, "frac_has_boxed": 0.84, "mean_reward_std_per_prompt": 0.2942299902299244, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 28.928817749023438, "avg_step_time_sec": 28.383060472352163, "gpu_allocated_gb": 32.365654016, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 397.4595789909363, "eta_sec": 5279.249247857502, "timestamp": "2026-02-15T11:39:48.912731"} {"step": 15, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.7, "mean_correctness_reward": 0.7, "mean_length_penalty": -0.737373046875, "mean_total_reward": -0.037373046875, "std_total_reward": 0.6515994289553758, "total_tokens": 75507, "mean_completion_length": 1510.14, "median_completion_length": 1565, "min_completion_length": 580, "max_completion_length": 2048, "frac_has_think": 0.76, "frac_has_boxed": 0.76, "mean_reward_std_per_prompt": 0.26526107913764346, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 27.844939470291138, "avg_step_time_sec": 28.34718573888143, "gpu_allocated_gb": 32.36562944, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 425.3108820915222, "eta_sec": 5244.229361693065, "timestamp": "2026-02-15T11:40:16.904182"} {"step": 16, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.46, "mean_correctness_reward": 0.46, "mean_length_penalty": -0.859501953125, "mean_total_reward": -0.399501953125, "std_total_reward": 0.6616736003871163, "total_tokens": 88013, "mean_completion_length": 1760.26, "median_completion_length": 2048, "min_completion_length": 847, "max_completion_length": 2048, "frac_has_think": 0.62, "frac_has_boxed": 0.62, "mean_reward_std_per_prompt": 0.34877354565188656, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 27.8202805519104, "avg_step_time_sec": 28.31425416469574, "gpu_allocated_gb": 32.365674496, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 453.1375062465668, "eta_sec": 5209.822766304016, "timestamp": "2026-02-15T11:40:45.272925"} {"step": 17, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.56, "mean_correctness_reward": 0.56, "mean_length_penalty": -0.82986328125, "mean_total_reward": -0.26986328125, "std_total_reward": 0.6456395496010395, "total_tokens": 84978, "mean_completion_length": 1699.56, "median_completion_length": 1885, "min_completion_length": 968, "max_completion_length": 2048, "frac_has_think": 0.8, "frac_has_boxed": 0.82, "mean_reward_std_per_prompt": 0.25155920043674007, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 28.360896587371826, "avg_step_time_sec": 28.316997836617862, "gpu_allocated_gb": 32.365631488, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 481.50518465042114, "eta_sec": 5182.010604101069, "timestamp": "2026-02-15T11:41:13.826555"} {"step": 18, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.54, "mean_correctness_reward": 0.54, "mean_length_penalty": -0.85298828125, "mean_total_reward": -0.31298828125, "std_total_reward": 0.650792951352729, "total_tokens": 87346, "mean_completion_length": 1746.92, "median_completion_length": 2029, "min_completion_length": 742, "max_completion_length": 2048, "frac_has_think": 0.62, "frac_has_boxed": 0.62, "mean_reward_std_per_prompt": 0.42383651901291136, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 28.556214570999146, "avg_step_time_sec": 28.3302876551946, "gpu_allocated_gb": 32.36563456, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 510.06799244880676, "eta_sec": 5156.112353245417, "timestamp": "2026-02-15T11:41:42.215912"} {"step": 19, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.62, "mean_correctness_reward": 0.62, "mean_length_penalty": -0.86095703125, "mean_total_reward": -0.24095703125, "std_total_reward": 0.6018231527431745, "total_tokens": 88162, "mean_completion_length": 1763.24, "median_completion_length": 1814, "min_completion_length": 682, "max_completion_length": 2048, "frac_has_think": 0.72, "frac_has_boxed": 0.72, "mean_reward_std_per_prompt": 0.35473516107749953, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 28.44787621498108, "avg_step_time_sec": 28.33647652676231, "gpu_allocated_gb": 32.365650944, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 538.5218114852905, "eta_sec": 5128.902251343979, "timestamp": "2026-02-15T11:42:11.211907"} {"step": 20, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.66, "mean_correctness_reward": 0.66, "mean_length_penalty": -0.839736328125, "mean_total_reward": -0.179736328125, "std_total_reward": 0.6002260517425583, "total_tokens": 85989, "mean_completion_length": 1719.78, "median_completion_length": 1807, "min_completion_length": 870, "max_completion_length": 2048, "frac_has_think": 0.72, "frac_has_boxed": 0.72, "mean_reward_std_per_prompt": 0.39151552259366923, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 28.80611562728882, "avg_step_time_sec": 28.359958481788635, "gpu_allocated_gb": 32.365620224, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 567.3346104621887, "eta_sec": 5104.792526721954, "timestamp": "2026-02-15T11:42:39.318033"} {"step": 21, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.74, "mean_correctness_reward": 0.74, "mean_length_penalty": -0.727158203125, "mean_total_reward": 0.012841796875, "std_total_reward": 0.6186490853215542, "total_tokens": 74461, "mean_completion_length": 1489.22, "median_completion_length": 1385, "min_completion_length": 755, "max_completion_length": 2048, "frac_has_think": 0.8, "frac_has_boxed": 0.8, "mean_reward_std_per_prompt": 0.3571375865478872, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 28.27785038948059, "avg_step_time_sec": 28.35604857263111, "gpu_allocated_gb": 32.365652992, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 595.6196579933167, "eta_sec": 5075.732694500969, "timestamp": "2026-02-15T11:43:07.405106"} {"step": 22, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.5, "mean_correctness_reward": 0.5, "mean_length_penalty": -0.77875, "mean_total_reward": -0.27875, "std_total_reward": 0.6636473644955244, "total_tokens": 79744, "mean_completion_length": 1594.88, "median_completion_length": 1625, "min_completion_length": 802, "max_completion_length": 2048, "frac_has_think": 0.74, "frac_has_boxed": 0.74, "mean_reward_std_per_prompt": 0.36452368225594967, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 28.601853609085083, "avg_step_time_sec": 28.367221528833564, "gpu_allocated_gb": 32.365769728, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.392025088, "elapsed_sec": 624.227977514267, "eta_sec": 5049.365432132375, "timestamp": "2026-02-15T11:43:35.975429"} {"step": 23, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.48, "mean_correctness_reward": 0.48, "mean_length_penalty": -0.77822265625, "mean_total_reward": -0.29822265625, "std_total_reward": 0.69280067691858, "total_tokens": 79690, "mean_completion_length": 1593.8, "median_completion_length": 1829, "min_completion_length": 764, "max_completion_length": 2048, "frac_has_think": 0.74, "frac_has_boxed": 0.74, "mean_reward_std_per_prompt": 0.35191251051618105, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 28.907505750656128, "avg_step_time_sec": 28.390712147173673, "gpu_allocated_gb": 32.365832192, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 653.1425502300262, "eta_sec": 5025.15605004974, "timestamp": "2026-02-15T11:44:05.141927"} {"step": 24, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.74, "mean_correctness_reward": 0.74, "mean_length_penalty": -0.6832421875, "mean_total_reward": 0.0567578125, "std_total_reward": 0.6436294262012068, "total_tokens": 69964, "mean_completion_length": 1399.28, "median_completion_length": 1307, "min_completion_length": 640, "max_completion_length": 2048, "frac_has_think": 0.8, "frac_has_boxed": 0.82, "mean_reward_std_per_prompt": 0.23246234590634401, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 28.163193702697754, "avg_step_time_sec": 28.38123221198718, "gpu_allocated_gb": 32.365621248, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 681.313001871109, "eta_sec": 4995.096869309744, "timestamp": "2026-02-15T11:44:31.930589"} {"step": 25, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.58, "mean_correctness_reward": 0.58, "mean_length_penalty": -0.603046875, "mean_total_reward": -0.023046875, "std_total_reward": 0.4768262989779115, "total_tokens": 61752, "mean_completion_length": 1235.04, "median_completion_length": 1172, "min_completion_length": 760, "max_completion_length": 2048, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.16762759130700117, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 26.82032084465027, "avg_step_time_sec": 28.3187957572937, "gpu_allocated_gb": 32.365627392, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 708.1406071186066, "eta_sec": 4955.789257526398, "timestamp": "2026-02-15T11:44:57.940953"} {"step": 26, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.68, "mean_correctness_reward": 0.68, "mean_length_penalty": -0.73138671875, "mean_total_reward": -0.05138671875, "std_total_reward": 0.6731955998622006, "total_tokens": 74894, "mean_completion_length": 1497.88, "median_completion_length": 1671, "min_completion_length": 626, "max_completion_length": 2048, "frac_has_think": 0.78, "frac_has_boxed": 0.78, "mean_reward_std_per_prompt": 0.21955946526319498, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 26.51204562187195, "avg_step_time_sec": 28.249305367469788, "gpu_allocated_gb": 32.365635584, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 734.6594171524048, "eta_sec": 4915.379133939743, "timestamp": "2026-02-15T11:45:25.673274"} {"step": 27, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.8, "mean_correctness_reward": 0.8, "mean_length_penalty": -0.563310546875, "mean_total_reward": 0.236689453125, "std_total_reward": 0.4473467470655598, "total_tokens": 57683, "mean_completion_length": 1153.66, "median_completion_length": 1040, "min_completion_length": 681, "max_completion_length": 2018, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.08264824204477633, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 27.114254236221313, "avg_step_time_sec": 28.207266436682808, "gpu_allocated_gb": 32.3656192, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 761.7795751094818, "eta_sec": 4879.857093546126, "timestamp": "2026-02-15T11:45:51.035893"} {"step": 28, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.64, "mean_correctness_reward": 0.64, "mean_length_penalty": -0.7408203125, "mean_total_reward": -0.1008203125, "std_total_reward": 0.6723905367909817, "total_tokens": 75860, "mean_completion_length": 1517.2, "median_completion_length": 1582, "min_completion_length": 632, "max_completion_length": 2048, "frac_has_think": 0.76, "frac_has_boxed": 0.76, "mean_reward_std_per_prompt": 0.45298756497533404, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 25.320915460586548, "avg_step_time_sec": 28.1041824732508, "gpu_allocated_gb": 32.36559616, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 787.1061275005341, "eta_sec": 4833.919385399137, "timestamp": "2026-02-15T11:46:18.901238"} {"step": 29, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.9, "mean_correctness_reward": 0.9, "mean_length_penalty": -0.59404296875, "mean_total_reward": 0.30595703125, "std_total_reward": 0.457200041029854, "total_tokens": 60830, "mean_completion_length": 1216.6, "median_completion_length": 1106, "min_completion_length": 685, "max_completion_length": 2048, "frac_has_think": 0.94, "frac_has_boxed": 0.94, "mean_reward_std_per_prompt": 0.2732193972052843, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 28.28618049621582, "avg_step_time_sec": 28.110458267146143, "gpu_allocated_gb": 32.365682688, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 815.3987131118774, "eta_sec": 4806.88836368199, "timestamp": "2026-02-15T11:46:45.149276"} {"step": 30, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.44, "mean_correctness_reward": 0.44, "mean_length_penalty": -0.725361328125, "mean_total_reward": -0.285361328125, "std_total_reward": 0.47564340573362013, "total_tokens": 74277, "mean_completion_length": 1485.54, "median_completion_length": 1469, "min_completion_length": 562, "max_completion_length": 2048, "frac_has_think": 0.98, "frac_has_boxed": 0.98, "mean_reward_std_per_prompt": 0.34299835588909067, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 25.83580708503723, "avg_step_time_sec": 28.034636561075846, "gpu_allocated_gb": 32.365612032, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 841.241907119751, "eta_sec": 4765.888215382894, "timestamp": "2026-02-15T11:47:12.875705"} {"step": 31, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.86, "mean_correctness_reward": 0.86, "mean_length_penalty": -0.56138671875, "mean_total_reward": 0.29861328125, "std_total_reward": 0.4815187487571622, "total_tokens": 57486, "mean_completion_length": 1149.72, "median_completion_length": 1045, "min_completion_length": 620, "max_completion_length": 2048, "frac_has_think": 0.92, "frac_has_boxed": 0.92, "mean_reward_std_per_prompt": 0.35798366157276595, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 27.951727151870728, "avg_step_time_sec": 28.031962064004713, "gpu_allocated_gb": 32.365669376, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 869.1998245716095, "eta_sec": 4737.401588816796, "timestamp": "2026-02-15T11:47:38.658382"} {"step": 32, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.64, "mean_correctness_reward": 0.64, "mean_length_penalty": -0.67712890625, "mean_total_reward": -0.03712890625, "std_total_reward": 0.6180919188623678, "total_tokens": 69338, "mean_completion_length": 1386.76, "median_completion_length": 1178, "min_completion_length": 736, "max_completion_length": 2048, "frac_has_think": 0.92, "frac_has_boxed": 0.92, "mean_reward_std_per_prompt": 0.26507048748980905, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 25.537105083465576, "avg_step_time_sec": 27.953997783362865, "gpu_allocated_gb": 32.365612032, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 894.7432882785797, "eta_sec": 4696.271627604961, "timestamp": "2026-02-15T11:48:05.373568"} {"step": 33, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.78, "mean_correctness_reward": 0.78, "mean_length_penalty": -0.642978515625, "mean_total_reward": 0.137021484375, "std_total_reward": 0.5982777346101518, "total_tokens": 65841, "mean_completion_length": 1316.82, "median_completion_length": 1225, "min_completion_length": 397, "max_completion_length": 2048, "frac_has_think": 0.8, "frac_has_boxed": 0.8, "mean_reward_std_per_prompt": 0.12083424502131015, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 26.94904112815857, "avg_step_time_sec": 27.92354455138698, "gpu_allocated_gb": 32.365663232, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 921.6993505954742, "eta_sec": 4663.231940081625, "timestamp": "2026-02-15T11:48:32.582262"} {"step": 34, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.92, "mean_correctness_reward": 0.92, "mean_length_penalty": -0.5986328125, "mean_total_reward": 0.3213671875, "std_total_reward": 0.41499067227095526, "total_tokens": 61300, "mean_completion_length": 1226, "median_completion_length": 1124, "min_completion_length": 714, "max_completion_length": 2048, "frac_has_think": 0.94, "frac_has_boxed": 0.94, "mean_reward_std_per_prompt": 0.23719866495743402, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 27.409321308135986, "avg_step_time_sec": 27.908420338350183, "gpu_allocated_gb": 32.365696, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 949.1156570911407, "eta_sec": 4632.7977761661305, "timestamp": "2026-02-15T11:48:58.964308"} {"step": 35, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.8, "mean_correctness_reward": 0.8, "mean_length_penalty": -0.55421875, "mean_total_reward": 0.24578125, "std_total_reward": 0.6085592180085025, "total_tokens": 56752, "mean_completion_length": 1135.04, "median_completion_length": 975, "min_completion_length": 494, "max_completion_length": 2048, "frac_has_think": 0.98, "frac_has_boxed": 0.98, "mean_reward_std_per_prompt": 0.09970469469976802, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 26.056549310684204, "avg_step_time_sec": 27.85550973755973, "gpu_allocated_gb": 32.365632512, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 975.1788766384125, "eta_sec": 4596.159106697355, "timestamp": "2026-02-15T11:49:24.714770"} {"step": 36, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.58, "mean_correctness_reward": 0.58, "mean_length_penalty": -0.601611328125, "mean_total_reward": -0.021611328125, "std_total_reward": 0.6383807572642441, "total_tokens": 61605, "mean_completion_length": 1232.1, "median_completion_length": 1037, "min_completion_length": 620, "max_completion_length": 2048, "frac_has_think": 0.78, "frac_has_boxed": 0.78, "mean_reward_std_per_prompt": 0.1322656322303718, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 25.82071876525879, "avg_step_time_sec": 27.798987766106922, "gpu_allocated_gb": 32.635115008, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1001.0065834522247, "eta_sec": 4559.033993641535, "timestamp": "2026-02-15T11:49:51.561038"} {"step": 37, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.92, "mean_correctness_reward": 0.92, "mean_length_penalty": -0.631396484375, "mean_total_reward": 0.288603515625, "std_total_reward": 0.38808606540656704, "total_tokens": 64655, "mean_completion_length": 1293.1, "median_completion_length": 1211, "min_completion_length": 714, "max_completion_length": 2048, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.200593917321634, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 27.21919012069702, "avg_step_time_sec": 27.783317559474224, "gpu_allocated_gb": 32.36573184, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1028.2330887317657, "eta_sec": 4528.680762194298, "timestamp": "2026-02-15T11:50:18.436599"} {"step": 38, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.62, "mean_correctness_reward": 0.62, "mean_length_penalty": -0.573134765625, "mean_total_reward": 0.046865234375, "std_total_reward": 0.7023044969157552, "total_tokens": 58689, "mean_completion_length": 1173.78, "median_completion_length": 934, "min_completion_length": 581, "max_completion_length": 2048, "frac_has_think": 0.82, "frac_has_boxed": 0.82, "mean_reward_std_per_prompt": 0.1388156224599236, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 26.501450777053833, "avg_step_time_sec": 27.74958422309474, "gpu_allocated_gb": 32.365658112, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1054.7414832115173, "eta_sec": 4495.432644141348, "timestamp": "2026-02-15T11:50:44.403212"} {"step": 39, "num_prompts": 5, "num_rollouts": 50, "accuracy": 1.0, "mean_correctness_reward": 1.0, "mean_length_penalty": -0.424970703125, "mean_total_reward": 0.575029296875, "std_total_reward": 0.08401665270451533, "total_tokens": 43517, "mean_completion_length": 870.34, "median_completion_length": 840, "min_completion_length": 618, "max_completion_length": 1440, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.06636853595792905, "frac_prompts_all_correct": 1.0, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 26.291970014572144, "avg_step_time_sec": 27.71220949979929, "gpu_allocated_gb": 32.365620224, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1081.040296792984, "eta_sec": 4461.665729467685, "timestamp": "2026-02-15T11:51:05.412623"} {"step": 40, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.54, "mean_correctness_reward": 0.54, "mean_length_penalty": -0.55349609375, "mean_total_reward": -0.01349609375, "std_total_reward": 0.5499442138123322, "total_tokens": 56678, "mean_completion_length": 1133.56, "median_completion_length": 1131, "min_completion_length": 565, "max_completion_length": 2048, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.20259070300683873, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 17.985809326171875, "avg_step_time_sec": 27.469049495458602, "gpu_allocated_gb": 32.364769792, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1099.033228635788, "eta_sec": 4395.0479192733765, "timestamp": "2026-02-15T11:51:28.547110"} {"step": 41, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.86, "mean_correctness_reward": 0.86, "mean_length_penalty": -0.559169921875, "mean_total_reward": 0.300830078125, "std_total_reward": 0.5584069960192812, "total_tokens": 57259, "mean_completion_length": 1145.18, "median_completion_length": 1080, "min_completion_length": 453, "max_completion_length": 2048, "frac_has_think": 0.92, "frac_has_boxed": 0.92, "mean_reward_std_per_prompt": 0.22983745846496048, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 26.090973615646362, "avg_step_time_sec": 27.435437888633913, "gpu_allocated_gb": 32.365696, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1125.1315779685974, "eta_sec": 4362.234624292792, "timestamp": "2026-02-15T11:51:54.733016"} {"step": 42, "num_prompts": 5, "num_rollouts": 50, "accuracy": 1.0, "mean_correctness_reward": 1.0, "mean_length_penalty": -0.396904296875, "mean_total_reward": 0.603095703125, "std_total_reward": 0.1505976852406278, "total_tokens": 40643, "mean_completion_length": 812.86, "median_completion_length": 696, "min_completion_length": 445, "max_completion_length": 1704, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.09656291981743591, "frac_prompts_all_correct": 1.0, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 26.359044075012207, "avg_step_time_sec": 27.409809464500064, "gpu_allocated_gb": 32.365651968, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1151.4968285560608, "eta_sec": 4330.74989539101, "timestamp": "2026-02-15T11:52:17.278434"} {"step": 43, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.96, "mean_correctness_reward": 0.96, "mean_length_penalty": -0.539453125, "mean_total_reward": 0.420546875, "std_total_reward": 0.3527527058887583, "total_tokens": 55240, "mean_completion_length": 1104.8, "median_completion_length": 950, "min_completion_length": 478, "max_completion_length": 2048, "frac_has_think": 0.96, "frac_has_boxed": 0.96, "mean_reward_std_per_prompt": 0.1648852138515499, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 20.39082360267639, "avg_step_time_sec": 27.24657723515533, "gpu_allocated_gb": 32.365125632, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1171.8938722610474, "eta_sec": 4277.7126259193865, "timestamp": "2026-02-15T11:52:41.272727"} {"step": 44, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.74, "mean_correctness_reward": 0.74, "mean_length_penalty": -0.5462109375, "mean_total_reward": 0.1937890625, "std_total_reward": 0.6393972747591926, "total_tokens": 55932, "mean_completion_length": 1118.64, "median_completion_length": 963, "min_completion_length": 467, "max_completion_length": 2048, "frac_has_think": 0.9, "frac_has_boxed": 0.9, "mean_reward_std_per_prompt": 0.24499852338902528, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 25.606101036071777, "avg_step_time_sec": 27.209293685176156, "gpu_allocated_gb": 32.365645824, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1197.5064723491669, "eta_sec": 4244.649814887481, "timestamp": "2026-02-15T11:53:06.979846"} {"step": 45, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.8, "mean_correctness_reward": 0.8, "mean_length_penalty": -0.531748046875, "mean_total_reward": 0.268251953125, "std_total_reward": 0.6393863063942439, "total_tokens": 54451, "mean_completion_length": 1089.02, "median_completion_length": 924, "min_completion_length": 454, "max_completion_length": 2048, "frac_has_think": 0.82, "frac_has_boxed": 0.82, "mean_reward_std_per_prompt": 0.03912884350483527, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 25.69822359085083, "avg_step_time_sec": 27.175714349746706, "gpu_allocated_gb": 32.365638656, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1223.2116866111755, "eta_sec": 4212.235724210739, "timestamp": "2026-02-15T11:53:33.155752"} {"step": 46, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.9, "mean_correctness_reward": 0.9, "mean_length_penalty": -0.448798828125, "mean_total_reward": 0.451201171875, "std_total_reward": 0.4933565180824905, "total_tokens": 45957, "mean_completion_length": 919.14, "median_completion_length": 717, "min_completion_length": 525, "max_completion_length": 2048, "frac_has_think": 0.94, "frac_has_boxed": 0.94, "mean_reward_std_per_prompt": 0.14402638767385248, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 26.245737075805664, "avg_step_time_sec": 27.155497452487115, "gpu_allocated_gb": 32.3656704, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1249.463711977005, "eta_sec": 4181.946607683016, "timestamp": "2026-02-15T11:53:58.080014"} {"step": 47, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.66, "mean_correctness_reward": 0.66, "mean_length_penalty": -0.51080078125, "mean_total_reward": 0.14919921875, "std_total_reward": 0.6273734547956477, "total_tokens": 52306, "mean_completion_length": 1046.12, "median_completion_length": 866, "min_completion_length": 437, "max_completion_length": 2048, "frac_has_think": 0.96, "frac_has_boxed": 0.96, "mean_reward_std_per_prompt": 0.22265059484375693, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 24.84592914581299, "avg_step_time_sec": 27.106357701281283, "gpu_allocated_gb": 32.36563456, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1274.3162083625793, "eta_sec": 4147.272728296036, "timestamp": "2026-02-15T11:54:23.487254"} {"step": 48, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.8, "mean_correctness_reward": 0.8, "mean_length_penalty": -0.477705078125, "mean_total_reward": 0.322294921875, "std_total_reward": 0.6672947846751791, "total_tokens": 48917, "mean_completion_length": 978.34, "median_completion_length": 711, "min_completion_length": 369, "max_completion_length": 2048, "frac_has_think": 0.8, "frac_has_boxed": 0.8, "mean_reward_std_per_prompt": 0.04145151893004691, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 25.387911796569824, "avg_step_time_sec": 27.07055674493313, "gpu_allocated_gb": 32.3656448, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1299.7105057239532, "eta_sec": 4114.7246252298355, "timestamp": "2026-02-15T11:54:49.112183"} {"step": 49, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.45556640625, "mean_total_reward": 0.14443359375, "std_total_reward": 0.717756018602415, "total_tokens": 46650, "mean_completion_length": 933, "median_completion_length": 779, "min_completion_length": 282, "max_completion_length": 2048, "frac_has_think": 0.8, "frac_has_boxed": 0.8, "mean_reward_std_per_prompt": 0.03798106356449531, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 25.616735458374023, "avg_step_time_sec": 27.040886922758453, "gpu_allocated_gb": 32.365645824, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1325.333322763443, "eta_sec": 4083.1739253365263, "timestamp": "2026-02-15T11:55:14.092960"} {"step": 50, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.98, "mean_correctness_reward": 0.98, "mean_length_penalty": -0.39953125, "mean_total_reward": 0.58046875, "std_total_reward": 0.2000757074116922, "total_tokens": 40912, "mean_completion_length": 818.24, "median_completion_length": 774, "min_completion_length": 426, "max_completion_length": 1480, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.09826456706783719, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 24.98887014389038, "avg_step_time_sec": 26.999846587181093, "gpu_allocated_gb": 32.365627392, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1350.3286204338074, "eta_sec": 4049.976988077164, "timestamp": "2026-02-15T11:55:35.382843"} {"step": 51, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.78, "mean_correctness_reward": 0.78, "mean_length_penalty": -0.38716796875, "mean_total_reward": 0.39283203125, "std_total_reward": 0.4586914503273345, "total_tokens": 39646, "mean_completion_length": 792.92, "median_completion_length": 706, "min_completion_length": 387, "max_completion_length": 2048, "frac_has_think": 0.98, "frac_has_boxed": 0.98, "mean_reward_std_per_prompt": 0.11768351185763873, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 18.304606437683105, "avg_step_time_sec": 26.829351682288973, "gpu_allocated_gb": 32.36482816, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1369.2116520404816, "eta_sec": 3997.573400661057, "timestamp": "2026-02-15T11:55:57.642804"} {"step": 52, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.74, "mean_correctness_reward": 0.74, "mean_length_penalty": -0.425302734375, "mean_total_reward": 0.314697265625, "std_total_reward": 0.6002479696465869, "total_tokens": 43551, "mean_completion_length": 871.02, "median_completion_length": 825, "min_completion_length": 376, "max_completion_length": 2048, "frac_has_think": 0.94, "frac_has_boxed": 0.94, "mean_reward_std_per_prompt": 0.2442390709480528, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 24.509216785430908, "avg_step_time_sec": 26.784733703503242, "gpu_allocated_gb": 32.365621248, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1393.7266104221344, "eta_sec": 3964.14058811848, "timestamp": "2026-02-15T11:56:22.137081"} {"step": 53, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.8, "mean_correctness_reward": 0.8, "mean_length_penalty": -0.58818359375, "mean_total_reward": 0.21181640625, "std_total_reward": 0.5997703510121579, "total_tokens": 60230, "mean_completion_length": 1204.6, "median_completion_length": 1102, "min_completion_length": 566, "max_completion_length": 2048, "frac_has_think": 0.92, "frac_has_boxed": 0.92, "mean_reward_std_per_prompt": 0.2440713224042606, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 24.621791124343872, "avg_step_time_sec": 26.743923466160613, "gpu_allocated_gb": 32.365645824, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1418.3545923233032, "eta_sec": 3931.35674952561, "timestamp": "2026-02-15T11:56:48.193120"} {"step": 54, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.8, "mean_correctness_reward": 0.8, "mean_length_penalty": -0.53560546875, "mean_total_reward": 0.26439453125, "std_total_reward": 0.6315660817659599, "total_tokens": 54846, "mean_completion_length": 1096.92, "median_completion_length": 894, "min_completion_length": 448, "max_completion_length": 2048, "frac_has_think": 0.9, "frac_has_boxed": 0.9, "mean_reward_std_per_prompt": 0.26023449587384323, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 26.124768495559692, "avg_step_time_sec": 26.732457633371705, "gpu_allocated_gb": 32.365657088, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1444.485324382782, "eta_sec": 3902.938814472269, "timestamp": "2026-02-15T11:57:14.457080"} {"step": 55, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.58, "mean_correctness_reward": 0.58, "mean_length_penalty": -0.4503515625, "mean_total_reward": 0.1296484375, "std_total_reward": 0.684819981664961, "total_tokens": 46116, "mean_completion_length": 922.32, "median_completion_length": 630, "min_completion_length": 360, "max_completion_length": 2048, "frac_has_think": 0.8, "frac_has_boxed": 0.8, "mean_reward_std_per_prompt": 0.0769222793445312, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 26.196780681610107, "avg_step_time_sec": 26.722718052430587, "gpu_allocated_gb": 32.365641728, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1470.688624382019, "eta_sec": 3874.794117602435, "timestamp": "2026-02-15T11:57:39.436864"} {"step": 56, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.76, "mean_correctness_reward": 0.76, "mean_length_penalty": -0.536845703125, "mean_total_reward": 0.223154296875, "std_total_reward": 0.7085589617314204, "total_tokens": 54973, "mean_completion_length": 1099.46, "median_completion_length": 718, "min_completion_length": 347, "max_completion_length": 2048, "frac_has_think": 0.8, "frac_has_boxed": 0.78, "mean_reward_std_per_prompt": 0.11963611181451377, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 24.96063780784607, "avg_step_time_sec": 26.691252333777292, "gpu_allocated_gb": 32.365648896, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1495.6553137302399, "eta_sec": 3843.54033606393, "timestamp": "2026-02-15T11:58:05.121322"} {"step": 57, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.602470703125, "mean_total_reward": -0.002470703125, "std_total_reward": 0.7270903569258446, "total_tokens": 61693, "mean_completion_length": 1233.86, "median_completion_length": 1275, "min_completion_length": 428, "max_completion_length": 2048, "frac_has_think": 0.76, "frac_has_boxed": 0.76, "mean_reward_std_per_prompt": 0.3126164645276175, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 25.694030046463013, "avg_step_time_sec": 26.67375720592967, "gpu_allocated_gb": 32.365647872, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1521.3552141189575, "eta_sec": 3814.347280447943, "timestamp": "2026-02-15T11:58:32.017672"} {"step": 58, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.88, "mean_correctness_reward": 0.88, "mean_length_penalty": -0.39087890625, "mean_total_reward": 0.48912109375, "std_total_reward": 0.4748762274653714, "total_tokens": 40026, "mean_completion_length": 800.52, "median_completion_length": 758, "min_completion_length": 359, "max_completion_length": 2048, "frac_has_think": 0.98, "frac_has_boxed": 0.98, "mean_reward_std_per_prompt": 0.24564742347953247, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 27.34681463241577, "avg_step_time_sec": 26.685361644317364, "gpu_allocated_gb": 32.36573696, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1548.7078304290771, "eta_sec": 3789.3213534930655, "timestamp": "2026-02-15T11:58:56.793357"} {"step": 59, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.38, "mean_correctness_reward": 0.38, "mean_length_penalty": -0.37716796875, "mean_total_reward": 0.00283203125, "std_total_reward": 0.5124249193973779, "total_tokens": 38622, "mean_completion_length": 772.44, "median_completion_length": 770, "min_completion_length": 392, "max_completion_length": 1233, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.10351032586251642, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.6, "step_time_sec": 24.230899333953857, "avg_step_time_sec": 26.643760588209506, "gpu_allocated_gb": 32.365630464, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1572.9448411464691, "eta_sec": 3756.7702429375404, "timestamp": "2026-02-15T11:59:15.645463"} {"step": 60, "num_prompts": 5, "num_rollouts": 50, "accuracy": 1.0, "mean_correctness_reward": 1.0, "mean_length_penalty": -0.315419921875, "mean_total_reward": 0.684580078125, "std_total_reward": 0.05898643560789105, "total_tokens": 32299, "mean_completion_length": 645.98, "median_completion_length": 625, "min_completion_length": 400, "max_completion_length": 911, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.03905128274063249, "frac_prompts_all_correct": 1.0, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 15.729955911636353, "avg_step_time_sec": 26.461863843599954, "gpu_allocated_gb": 32.364511232, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1588.6803641319275, "eta_sec": 3704.6609381039934, "timestamp": "2026-02-15T11:59:28.773712"} {"step": 61, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.8, "mean_correctness_reward": 0.8, "mean_length_penalty": -0.459130859375, "mean_total_reward": 0.340869140625, "std_total_reward": 0.4935659161986017, "total_tokens": 47015, "mean_completion_length": 940.3, "median_completion_length": 947, "min_completion_length": 289, "max_completion_length": 2048, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.3238812687583328, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 11.724182605743408, "avg_step_time_sec": 26.220262511831816, "gpu_allocated_gb": 32.364028928, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1600.4111433029175, "eta_sec": 3644.6164891446224, "timestamp": "2026-02-15T11:59:49.591401"} {"step": 62, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.8, "mean_correctness_reward": 0.8, "mean_length_penalty": -0.356953125, "mean_total_reward": 0.443046875, "std_total_reward": 0.5431766403810455, "total_tokens": 36552, "mean_completion_length": 731.04, "median_completion_length": 731, "min_completion_length": 244, "max_completion_length": 1615, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.04502425774615019, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 25.452081441879272, "avg_step_time_sec": 26.207872494574517, "gpu_allocated_gb": 32.365659136, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1625.870953798294, "eta_sec": 3616.6864042512834, "timestamp": "2026-02-15T12:00:10.694753"} {"step": 63, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.88, "mean_correctness_reward": 0.88, "mean_length_penalty": -0.3666015625, "mean_total_reward": 0.5133984375, "std_total_reward": 0.3850260133748651, "total_tokens": 37540, "mean_completion_length": 750.8, "median_completion_length": 701, "min_completion_length": 334, "max_completion_length": 2048, "frac_has_think": 0.98, "frac_has_boxed": 0.98, "mean_reward_std_per_prompt": 0.34063865320744424, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 19.324589729309082, "avg_step_time_sec": 26.098614037983, "gpu_allocated_gb": 32.365047296, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1645.201611995697, "eta_sec": 3575.5101232036714, "timestamp": "2026-02-15T12:00:31.861392"} {"step": 64, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.92, "mean_correctness_reward": 0.92, "mean_length_penalty": -0.401181640625, "mean_total_reward": 0.518818359375, "std_total_reward": 0.43023302118557, "total_tokens": 41081, "mean_completion_length": 821.62, "median_completion_length": 769, "min_completion_length": 378, "max_completion_length": 2048, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.1651860611395597, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 22.854740619659424, "avg_step_time_sec": 26.047928515821695, "gpu_allocated_gb": 32.365643776, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1668.0630550384521, "eta_sec": 3542.5182781517506, "timestamp": "2026-02-15T12:00:55.327371"} {"step": 65, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.76, "mean_correctness_reward": 0.76, "mean_length_penalty": -0.38638671875, "mean_total_reward": 0.37361328125, "std_total_reward": 0.6861395352645315, "total_tokens": 39566, "mean_completion_length": 791.32, "median_completion_length": 553, "min_completion_length": 295, "max_completion_length": 2048, "frac_has_think": 0.92, "frac_has_boxed": 0.92, "mean_reward_std_per_prompt": 0.12164736391209385, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 23.519651889801025, "avg_step_time_sec": 26.009031952344454, "gpu_allocated_gb": 32.36566016, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1691.5895102024078, "eta_sec": 3511.219313566501, "timestamp": "2026-02-15T12:01:19.375555"} {"step": 66, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.357138671875, "mean_total_reward": 0.242861328125, "std_total_reward": 0.5445039445333112, "total_tokens": 36571, "mean_completion_length": 731.42, "median_completion_length": 748, "min_completion_length": 213, "max_completion_length": 1453, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.04301517212182919, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 24.0369975566864, "avg_step_time_sec": 25.97915264331933, "gpu_allocated_gb": 32.36565504, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1715.6332619190216, "eta_sec": 3481.2064542047906, "timestamp": "2026-02-15T12:01:38.871263"} {"step": 67, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.8, "mean_correctness_reward": 0.8, "mean_length_penalty": -0.355712890625, "mean_total_reward": 0.444287109375, "std_total_reward": 0.5176118084005709, "total_tokens": 36425, "mean_completion_length": 728.5, "median_completion_length": 653, "min_completion_length": 434, "max_completion_length": 1519, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.030599404202508396, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 16.920116186141968, "avg_step_time_sec": 25.843943143958477, "gpu_allocated_gb": 32.36480512, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1732.560350894928, "eta_sec": 3437.2444381464775, "timestamp": "2026-02-15T12:01:56.332987"} {"step": 68, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.343056640625, "mean_total_reward": 0.256943359375, "std_total_reward": 0.6759258934149693, "total_tokens": 35129, "mean_completion_length": 702.58, "median_completion_length": 478, "min_completion_length": 326, "max_completion_length": 2048, "frac_has_think": 0.98, "frac_has_boxed": 0.98, "mean_reward_std_per_prompt": 0.059026060445571696, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 17.796121835708618, "avg_step_time_sec": 25.72559283060186, "gpu_allocated_gb": 32.364908544, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1750.3639442920685, "eta_sec": 3395.7782536394457, "timestamp": "2026-02-15T12:02:17.402981"} {"step": 69, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.2, "mean_correctness_reward": 0.2, "mean_length_penalty": -0.4664453125, "mean_total_reward": -0.2664453125, "std_total_reward": 0.5851082628281166, "total_tokens": 47764, "mean_completion_length": 955.28, "median_completion_length": 626, "min_completion_length": 285, "max_completion_length": 2048, "frac_has_think": 0.78, "frac_has_boxed": 0.78, "mean_reward_std_per_prompt": 0.05371580896134758, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.8, "step_time_sec": 23.27153468132019, "avg_step_time_sec": 25.690026770467345, "gpu_allocated_gb": 32.365641728, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1773.6417968273163, "eta_sec": 3365.393506931222, "timestamp": "2026-02-15T12:02:42.283308"} {"step": 70, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.54, "mean_correctness_reward": 0.54, "mean_length_penalty": -0.24333984375, "mean_total_reward": 0.29666015625, "std_total_reward": 0.5276944016481147, "total_tokens": 24918, "mean_completion_length": 498.36, "median_completion_length": 465, "min_completion_length": 277, "max_completion_length": 2048, "frac_has_think": 0.98, "frac_has_boxed": 0.98, "mean_reward_std_per_prompt": 0.19605155153843273, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 24.892953157424927, "avg_step_time_sec": 25.678640004566738, "gpu_allocated_gb": 32.36564992, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1798.541380405426, "eta_sec": 3338.2232005936758, "timestamp": "2026-02-15T12:03:03.984606"} {"step": 71, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.74, "mean_correctness_reward": 0.74, "mean_length_penalty": -0.25529296875, "mean_total_reward": 0.48470703125, "std_total_reward": 0.5019509347593266, "total_tokens": 26142, "mean_completion_length": 522.84, "median_completion_length": 469, "min_completion_length": 281, "max_completion_length": 896, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.11500330641560212, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 21.512423276901245, "avg_step_time_sec": 25.61996089572638, "gpu_allocated_gb": 32.365621248, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1820.0603475570679, "eta_sec": 3304.974955548703, "timestamp": "2026-02-15T12:03:19.417710"} {"step": 72, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.62, "mean_correctness_reward": 0.62, "mean_length_penalty": -0.381025390625, "mean_total_reward": 0.238974609375, "std_total_reward": 0.6765155033413047, "total_tokens": 39017, "mean_completion_length": 780.34, "median_completion_length": 515, "min_completion_length": 251, "max_completion_length": 2048, "frac_has_think": 0.9, "frac_has_boxed": 0.9, "mean_reward_std_per_prompt": 0.10782974165968642, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 11.064551830291748, "avg_step_time_sec": 25.41780243648423, "gpu_allocated_gb": 32.364051456, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1831.1312742233276, "eta_sec": 3253.4787118699815, "timestamp": "2026-02-15T12:03:38.476716"} {"step": 73, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.4, "mean_correctness_reward": 0.4, "mean_length_penalty": -0.270107421875, "mean_total_reward": 0.129892578125, "std_total_reward": 0.547162700637928, "total_tokens": 27659, "mean_completion_length": 553.18, "median_completion_length": 481, "min_completion_length": 263, "max_completion_length": 1442, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.06607526502039808, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.6, "step_time_sec": 23.586691856384277, "avg_step_time_sec": 25.392718729907518, "gpu_allocated_gb": 32.36562944, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1854.7245230674744, "eta_sec": 3224.875278698255, "timestamp": "2026-02-15T12:03:57.877326"} {"step": 74, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.9, "mean_correctness_reward": 0.9, "mean_length_penalty": -0.225498046875, "mean_total_reward": 0.674501953125, "std_total_reward": 0.3214094909000148, "total_tokens": 23091, "mean_completion_length": 461.82, "median_completion_length": 365, "min_completion_length": 151, "max_completion_length": 1158, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.17861237818238057, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 16.936902284622192, "avg_step_time_sec": 25.278450940106367, "gpu_allocated_gb": 32.364811264, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1871.667988061905, "eta_sec": 3185.0848184534025, "timestamp": "2026-02-15T12:04:12.232356"} {"step": 75, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.78, "mean_correctness_reward": 0.78, "mean_length_penalty": -0.283330078125, "mean_total_reward": 0.496669921875, "std_total_reward": 0.6460580407898693, "total_tokens": 29013, "mean_completion_length": 580.26, "median_completion_length": 370, "min_completion_length": 157, "max_completion_length": 2048, "frac_has_think": 0.94, "frac_has_boxed": 0.94, "mean_reward_std_per_prompt": 0.12066046759519608, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 13.150975704193115, "avg_step_time_sec": 25.116751270294188, "gpu_allocated_gb": 32.364383744, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1884.825438261032, "eta_sec": 3139.5939087867737, "timestamp": "2026-02-15T12:04:31.391658"} {"step": 76, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.84, "mean_correctness_reward": 0.84, "mean_length_penalty": -0.250224609375, "mean_total_reward": 0.589775390625, "std_total_reward": 0.5161040330186365, "total_tokens": 25623, "mean_completion_length": 512.46, "median_completion_length": 357, "min_completion_length": 122, "max_completion_length": 2048, "frac_has_think": 0.98, "frac_has_boxed": 0.98, "mean_reward_std_per_prompt": 0.16088789698613076, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 22.822327375411987, "avg_step_time_sec": 25.086561482203635, "gpu_allocated_gb": 32.365657088, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1907.653977394104, "eta_sec": 3110.733623793251, "timestamp": "2026-02-15T12:04:53.636185"} {"step": 77, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.242626953125, "mean_total_reward": 0.357373046875, "std_total_reward": 0.5789677534495806, "total_tokens": 24845, "mean_completion_length": 496.9, "median_completion_length": 418, "min_completion_length": 233, "max_completion_length": 1316, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.04209073744788806, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 22.139278173446655, "avg_step_time_sec": 25.0482850755964, "gpu_allocated_gb": 32.36562432, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1929.7995879650116, "eta_sec": 3080.939064298357, "timestamp": "2026-02-15T12:05:11.417254"} {"step": 78, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.78, "mean_correctness_reward": 0.78, "mean_length_penalty": -0.248447265625, "mean_total_reward": 0.531552734375, "std_total_reward": 0.4126015034260358, "total_tokens": 25441, "mean_completion_length": 508.82, "median_completion_length": 516, "min_completion_length": 300, "max_completion_length": 791, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.08640171789963916, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 15.2257661819458, "avg_step_time_sec": 24.922355346190624, "gpu_allocated_gb": 32.36460288, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1945.0310986042023, "eta_sec": 3040.5273522352563, "timestamp": "2026-02-15T12:05:23.588146"} {"step": 79, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.4, "mean_correctness_reward": 0.4, "mean_length_penalty": -0.2121484375, "mean_total_reward": 0.1878515625, "std_total_reward": 0.5336831508825144, "total_tokens": 21724, "mean_completion_length": 434.48, "median_completion_length": 450, "min_completion_length": 186, "max_completion_length": 745, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.0326196007851527, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.6, "step_time_sec": 9.832759618759155, "avg_step_time_sec": 24.731347805337062, "gpu_allocated_gb": 32.363888128, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1954.8702638149261, "eta_sec": 2992.4930844457845, "timestamp": "2026-02-15T12:05:32.851770"} {"step": 80, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.76, "mean_correctness_reward": 0.76, "mean_length_penalty": -0.2226953125, "mean_total_reward": 0.5373046875, "std_total_reward": 0.39056229965539346, "total_tokens": 22804, "mean_completion_length": 456.08, "median_completion_length": 401, "min_completion_length": 215, "max_completion_length": 904, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.14363714136092373, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 9.140454292297363, "avg_step_time_sec": 24.536461636424065, "gpu_allocated_gb": 32.363829248, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1964.0164303779602, "eta_sec": 2944.3753963708878, "timestamp": "2026-02-15T12:05:43.145441"} {"step": 81, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.66, "mean_correctness_reward": 0.66, "mean_length_penalty": -0.152197265625, "mean_total_reward": 0.507802734375, "std_total_reward": 0.5074648890070934, "total_tokens": 15585, "mean_completion_length": 311.7, "median_completion_length": 281, "min_completion_length": 145, "max_completion_length": 592, "frac_has_think": 0.98, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.10373407000164664, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 11.149046421051025, "avg_step_time_sec": 24.371184905370075, "gpu_allocated_gb": 32.364136448, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1975.171305656433, "eta_sec": 2900.171003739039, "timestamp": "2026-02-15T12:05:51.836827"} {"step": 82, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.58, "mean_correctness_reward": 0.58, "mean_length_penalty": -0.12244140625, "mean_total_reward": 0.45755859375, "std_total_reward": 0.5046988267129465, "total_tokens": 12538, "mean_completion_length": 250.76, "median_completion_length": 244, "min_completion_length": 120, "max_completion_length": 484, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.07470575935353967, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 7.229802370071411, "avg_step_time_sec": 24.162143654939605, "gpu_allocated_gb": 32.363599872, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1982.4081795215607, "eta_sec": 2851.1329512828734, "timestamp": "2026-02-15T12:05:58.135184"} {"step": 83, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.76, "mean_correctness_reward": 0.76, "mean_length_penalty": -0.200654296875, "mean_total_reward": 0.559345703125, "std_total_reward": 0.42237440386817865, "total_tokens": 20547, "mean_completion_length": 410.94, "median_completion_length": 218, "min_completion_length": 94, "max_completion_length": 1553, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.13182299336429257, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 5.891402006149292, "avg_step_time_sec": 23.9420142374843, "gpu_allocated_gb": 32.3634432, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1988.3058326244354, "eta_sec": 2801.215665785663, "timestamp": "2026-02-15T12:06:11.145637"} {"step": 84, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.62, "mean_correctness_reward": 0.62, "mean_length_penalty": -0.183232421875, "mean_total_reward": 0.436767578125, "std_total_reward": 0.5166475110451011, "total_tokens": 18763, "mean_completion_length": 375.26, "median_completion_length": 265, "min_completion_length": 139, "max_completion_length": 1174, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.14858813307138047, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 17.48445439338684, "avg_step_time_sec": 23.86513852505457, "gpu_allocated_gb": 32.364968448, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2005.7965848445892, "eta_sec": 2768.35606890633, "timestamp": "2026-02-15T12:06:26.115432"} {"step": 85, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.92, "mean_correctness_reward": 0.92, "mean_length_penalty": -0.09146484375, "mean_total_reward": 0.82853515625, "std_total_reward": 0.2733088836744714, "total_tokens": 9366, "mean_completion_length": 187.32, "median_completion_length": 195, "min_completion_length": 57, "max_completion_length": 319, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.1099429752087724, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 12.903345108032227, "avg_step_time_sec": 23.736176249560188, "gpu_allocated_gb": 32.364385792, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2018.7058906555176, "eta_sec": 2729.6602686994215, "timestamp": "2026-02-15T12:06:33.272665"} {"step": 86, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.56, "mean_correctness_reward": 0.56, "mean_length_penalty": -0.13470703125, "mean_total_reward": 0.42529296875, "std_total_reward": 0.5217734049513275, "total_tokens": 13794, "mean_completion_length": 275.88, "median_completion_length": 240, "min_completion_length": 84, "max_completion_length": 702, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.10333928338887041, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 4.123839616775513, "avg_step_time_sec": 23.508125823597574, "gpu_allocated_gb": 32.363193856, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2022.8358500003815, "eta_sec": 2679.9263438901235, "timestamp": "2026-02-15T12:06:39.929721"} {"step": 87, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.9, "mean_correctness_reward": 0.9, "mean_length_penalty": -0.08982421875, "mean_total_reward": 0.81017578125, "std_total_reward": 0.30680082275399667, "total_tokens": 9198, "mean_completion_length": 183.96, "median_completion_length": 154, "min_completion_length": 58, "max_completion_length": 450, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.22198254049662386, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 8.086495876312256, "avg_step_time_sec": 23.330865709260962, "gpu_allocated_gb": 32.363736576, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2030.928050994873, "eta_sec": 2636.387825146489, "timestamp": "2026-02-15T12:06:46.268803"} {"step": 88, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.36, "mean_correctness_reward": 0.36, "mean_length_penalty": -0.111796875, "mean_total_reward": 0.248203125, "std_total_reward": 0.5122904878685249, "total_tokens": 11448, "mean_completion_length": 228.96, "median_completion_length": 209, "min_completion_length": 64, "max_completion_length": 769, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.27763560622809413, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 5.476812839508057, "avg_step_time_sec": 23.127978744831953, "gpu_allocated_gb": 32.3634176, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2036.4108202457428, "eta_sec": 2590.333619421179, "timestamp": "2026-02-15T12:06:53.755250"} {"step": 89, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.68, "mean_correctness_reward": 0.68, "mean_length_penalty": -0.1563671875, "mean_total_reward": 0.5236328125, "std_total_reward": 0.48833853179629694, "total_tokens": 16012, "mean_completion_length": 320.24, "median_completion_length": 202, "min_completion_length": 101, "max_completion_length": 1294, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.318328712293294, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 8.60616135597229, "avg_step_time_sec": 22.964812257316677, "gpu_allocated_gb": 32.363881984, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2045.0235376358032, "eta_sec": 2549.094160562151, "timestamp": "2026-02-15T12:07:05.840532"} {"step": 90, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.78, "mean_correctness_reward": 0.78, "mean_length_penalty": -0.089765625, "mean_total_reward": 0.690234375, "std_total_reward": 0.43199932953111403, "total_tokens": 9192, "mean_completion_length": 183.84, "median_completion_length": 138, "min_completion_length": 40, "max_completion_length": 801, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.19222918757072013, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 13.908793687820435, "avg_step_time_sec": 22.864189828766715, "gpu_allocated_gb": 32.364582912, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2058.9386823177338, "eta_sec": 2515.0608811643388, "timestamp": "2026-02-15T12:07:16.494454"} {"step": 91, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.067880859375, "mean_total_reward": 0.532119140625, "std_total_reward": 0.49559979493750395, "total_tokens": 6951, "mean_completion_length": 139.02, "median_completion_length": 126, "min_completion_length": 59, "max_completion_length": 283, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.1263970917375661, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 9.8855881690979, "avg_step_time_sec": 22.72156783250662, "gpu_allocated_gb": 32.364045824, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2068.83114027977, "eta_sec": 2476.6508937432213, "timestamp": "2026-02-15T12:07:22.988860"} {"step": 92, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.5, "mean_correctness_reward": 0.5, "mean_length_penalty": -0.0603515625, "mean_total_reward": 0.4396484375, "std_total_reward": 0.5134438824619708, "total_tokens": 6180, "mean_completion_length": 123.6, "median_completion_length": 117, "min_completion_length": 36, "max_completion_length": 514, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.2476633280485788, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 3.7109248638153076, "avg_step_time_sec": 22.51493040893389, "gpu_allocated_gb": 32.363156992, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2072.5488319396973, "eta_sec": 2431.6124841648602, "timestamp": "2026-02-15T12:07:28.100864"} {"step": 93, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.42, "mean_correctness_reward": 0.42, "mean_length_penalty": -0.0793359375, "mean_total_reward": 0.3406640625, "std_total_reward": 0.47492945739546744, "total_tokens": 8124, "mean_completion_length": 162.48, "median_completion_length": 122, "min_completion_length": 50, "max_completion_length": 448, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.25500383623111, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 6.296195030212402, "avg_step_time_sec": 22.340535404861615, "gpu_allocated_gb": 32.363629056, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2078.852345228195, "eta_sec": 2390.4372883201927, "timestamp": "2026-02-15T12:07:34.306942"} {"step": 94, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.36, "mean_correctness_reward": 0.36, "mean_length_penalty": -0.038466796875, "mean_total_reward": 0.321533203125, "std_total_reward": 0.4767020624641501, "total_tokens": 3939, "mean_completion_length": 78.78, "median_completion_length": 73, "min_completion_length": 31, "max_completion_length": 144, "frac_has_think": 1.0, "frac_has_boxed": 0.98, "mean_reward_std_per_prompt": 0.3731958566582783, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 6.190309286117554, "avg_step_time_sec": 22.16872448870476, "gpu_allocated_gb": 32.363581952, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2085.04900097847, "eta_sec": 2349.8847958027045, "timestamp": "2026-02-15T12:07:38.221111"} {"step": 95, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.032666015625, "mean_total_reward": 0.567333984375, "std_total_reward": 0.49427078643680694, "total_tokens": 3345, "mean_completion_length": 66.9, "median_completion_length": 63, "min_completion_length": 28, "max_completion_length": 144, "frac_has_think": 1.0, "frac_has_boxed": 0.98, "mean_reward_std_per_prompt": 0.2540133642171448, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 2.391721725463867, "avg_step_time_sec": 21.960545512249595, "gpu_allocated_gb": 32.362957312, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2087.446749687195, "eta_sec": 2305.8572787862076, "timestamp": "2026-02-15T12:07:40.616039"} {"step": 96, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.4, "mean_correctness_reward": 0.4, "mean_length_penalty": -0.03341796875, "mean_total_reward": 0.36658203125, "std_total_reward": 0.49012746243003014, "total_tokens": 3422, "mean_completion_length": 68.44, "median_completion_length": 65, "min_completion_length": 35, "max_completion_length": 163, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.26898371686285305, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 2.4477717876434326, "avg_step_time_sec": 21.757287452618282, "gpu_allocated_gb": 32.362989056, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2089.902505159378, "eta_sec": 2262.7578950723014, "timestamp": "2026-02-15T12:07:43.140084"} {"step": 97, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.34, "mean_correctness_reward": 0.34, "mean_length_penalty": -0.023330078125, "mean_total_reward": 0.316669921875, "std_total_reward": 0.4756774409253905, "total_tokens": 2389, "mean_completion_length": 47.78, "median_completion_length": 47, "min_completion_length": 20, "max_completion_length": 113, "frac_has_think": 1.0, "frac_has_boxed": 0.98, "mean_reward_std_per_prompt": 0.09704032365828769, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.6, "step_time_sec": 2.455653190612793, "avg_step_time_sec": 21.55830153239142, "gpu_allocated_gb": 32.363005952, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2092.363744735718, "eta_sec": 2220.5050578363166, "timestamp": "2026-02-15T12:07:45.242398"} {"step": 98, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.62, "mean_correctness_reward": 0.62, "mean_length_penalty": -0.024326171875, "mean_total_reward": 0.595673828125, "std_total_reward": 0.48648901951111795, "total_tokens": 2491, "mean_completion_length": 49.82, "median_completion_length": 51, "min_completion_length": 23, "max_completion_length": 79, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.2516091343317751, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 2.095844268798828, "avg_step_time_sec": 21.3597050297017, "gpu_allocated_gb": 32.362919424, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2094.465828895569, "eta_sec": 2178.6899130295733, "timestamp": "2026-02-15T12:07:47.163658"} {"step": 99, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.62, "mean_correctness_reward": 0.62, "mean_length_penalty": -0.019111328125, "mean_total_reward": 0.600888671875, "std_total_reward": 0.4846247848773958, "total_tokens": 1957, "mean_completion_length": 39.14, "median_completion_length": 40, "min_completion_length": 19, "max_completion_length": 65, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.3686976252099313, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 1.8559808731079102, "avg_step_time_sec": 21.16269771498863, "gpu_allocated_gb": 32.362885632, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2096.326985359192, "eta_sec": 2137.4324692138516, "timestamp": "2026-02-15T12:07:48.915324"} {"step": 100, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.46, "mean_correctness_reward": 0.46, "mean_length_penalty": -0.020126953125, "mean_total_reward": 0.439873046875, "std_total_reward": 0.49851282211924225, "total_tokens": 2061, "mean_completion_length": 41.22, "median_completion_length": 37, "min_completion_length": 21, "max_completion_length": 71, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.14199403594389695, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 1.7908530235290527, "avg_step_time_sec": 20.968979268074037, "gpu_allocated_gb": 32.362841088, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2098.1244933605194, "eta_sec": 2096.8979268074036, "timestamp": "2026-02-15T12:07:52.664261"} {"step": 101, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.46, "mean_correctness_reward": 0.46, "mean_length_penalty": -0.013779296875, "mean_total_reward": 0.446220703125, "std_total_reward": 0.4961664390836806, "total_tokens": 1411, "mean_completion_length": 28.22, "median_completion_length": 27, "min_completion_length": 17, "max_completion_length": 74, "frac_has_think": 1.0, "frac_has_boxed": 0.92, "mean_reward_std_per_prompt": 0.0914379867470349, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 1.8271172046661377, "avg_step_time_sec": 20.77945588130762, "gpu_allocated_gb": 32.362901504, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2101.847535610199, "eta_sec": 2057.166132249454, "timestamp": "2026-02-15T12:07:54.465283"} {"step": 102, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.2, "mean_correctness_reward": 0.2, "mean_length_penalty": -0.022646484375, "mean_total_reward": 0.177353515625, "std_total_reward": 0.4017642012900887, "total_tokens": 2319, "mean_completion_length": 46.38, "median_completion_length": 28, "min_completion_length": 8, "max_completion_length": 302, "frac_has_think": 1.0, "frac_has_boxed": 0.88, "mean_reward_std_per_prompt": 0.2674311902966641, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 2.271826982498169, "avg_step_time_sec": 20.59800853916243, "gpu_allocated_gb": 32.362895872, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2104.126271009445, "eta_sec": 2018.604836837918, "timestamp": "2026-02-15T12:07:58.254715"} {"step": 103, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.02, "mean_correctness_reward": 0.02, "mean_length_penalty": -0.012470703125, "mean_total_reward": 0.007529296875, "std_total_reward": 0.1402926468769148, "total_tokens": 1277, "mean_completion_length": 25.54, "median_completion_length": 23, "min_completion_length": 12, "max_completion_length": 108, "frac_has_think": 1.0, "frac_has_boxed": 0.9, "mean_reward_std_per_prompt": 0.062850860325171, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.8, "step_time_sec": 4.02325439453125, "avg_step_time_sec": 20.43708859601067, "gpu_allocated_gb": 32.3632896, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2108.158737897873, "eta_sec": 1982.3975938130352, "timestamp": "2026-02-15T12:08:00.933455"} {"step": 104, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.28, "mean_correctness_reward": 0.28, "mean_length_penalty": -0.00912109375, "mean_total_reward": 0.27087890625, "std_total_reward": 0.44861338881684354, "total_tokens": 934, "mean_completion_length": 18.68, "median_completion_length": 17, "min_completion_length": 12, "max_completion_length": 45, "frac_has_think": 1.0, "frac_has_boxed": 0.34, "mean_reward_std_per_prompt": 0.09816172852776857, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.6, "step_time_sec": 1.96044921875, "avg_step_time_sec": 20.25942860199855, "gpu_allocated_gb": 32.362941952, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2110.1251163482666, "eta_sec": 1944.9051457918608, "timestamp": "2026-02-15T12:08:02.539623"} {"step": 105, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.22, "mean_correctness_reward": 0.22, "mean_length_penalty": -0.00671875, "mean_total_reward": 0.21328125, "std_total_reward": 0.41452035601320764, "total_tokens": 688, "mean_completion_length": 13.76, "median_completion_length": 12, "min_completion_length": 8, "max_completion_length": 24, "frac_has_think": 1.0, "frac_has_boxed": 0.16, "mean_reward_std_per_prompt": 0.14070116133973698, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.6, "step_time_sec": 1.5592093467712402, "avg_step_time_sec": 20.081331275758288, "gpu_allocated_gb": 32.362821632, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2111.694237470627, "eta_sec": 1907.7264711970374, "timestamp": "2026-02-15T12:08:03.964999"} {"step": 106, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.005576171875, "mean_total_reward": -0.005576171875, "std_total_reward": 0.0021037840038867503, "total_tokens": 571, "mean_completion_length": 11.42, "median_completion_length": 10, "min_completion_length": 8, "max_completion_length": 22, "frac_has_think": 1.0, "frac_has_boxed": 0.12, "mean_reward_std_per_prompt": 0.0008225655298736011, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.4174046516418457, "avg_step_time_sec": 19.905256496285492, "gpu_allocated_gb": 32.362799104, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2113.1172926425934, "eta_sec": 1871.0941106508362, "timestamp": "2026-02-15T12:08:05.432823"} {"step": 107, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.24, "mean_correctness_reward": 0.24, "mean_length_penalty": -0.006083984375, "mean_total_reward": 0.233916015625, "std_total_reward": 0.42766918125191333, "total_tokens": 623, "mean_completion_length": 12.46, "median_completion_length": 14, "min_completion_length": 8, "max_completion_length": 19, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.08029441521484727, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.6, "step_time_sec": 1.4590966701507568, "avg_step_time_sec": 19.732862479218813, "gpu_allocated_gb": 32.36287744, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2114.5824682712555, "eta_sec": 1835.1562105673497, "timestamp": "2026-02-15T12:08:06.836948"} {"step": 108, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.18, "mean_correctness_reward": 0.18, "mean_length_penalty": -0.00564453125, "mean_total_reward": 0.17435546875, "std_total_reward": 0.38408863397738074, "total_tokens": 578, "mean_completion_length": 11.56, "median_completion_length": 12, "min_completion_length": 8, "max_completion_length": 16, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.06007615194307747, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.8, "step_time_sec": 1.3827319145202637, "avg_step_time_sec": 19.562953862879013, "gpu_allocated_gb": 32.362815488, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2115.971464395523, "eta_sec": 1799.7917553848692, "timestamp": "2026-02-15T12:08:08.215093"} {"step": 109, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.005244140625, "mean_total_reward": -0.005244140625, "std_total_reward": 0.001512285151078932, "total_tokens": 537, "mean_completion_length": 10.74, "median_completion_length": 10, "min_completion_length": 8, "max_completion_length": 25, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0005620182922284557, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.3696107864379883, "avg_step_time_sec": 19.396042458507992, "gpu_allocated_gb": 32.362776064, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2117.3494749069214, "eta_sec": 1765.0398637242272, "timestamp": "2026-02-15T12:08:09.630108"} {"step": 110, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.38, "mean_correctness_reward": 0.38, "mean_length_penalty": -0.00470703125, "mean_total_reward": 0.37529296875, "std_total_reward": 0.48488704073835687, "total_tokens": 482, "mean_completion_length": 9.64, "median_completion_length": 10, "min_completion_length": 8, "max_completion_length": 12, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.06033203125, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.6, "step_time_sec": 1.4270122051239014, "avg_step_time_sec": 19.232687638022682, "gpu_allocated_gb": 32.362796032, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2118.782965183258, "eta_sec": 1730.9418874220414, "timestamp": "2026-02-15T12:08:10.985621"} {"step": 111, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.2, "mean_correctness_reward": 0.2, "mean_length_penalty": -0.006640625, "mean_total_reward": 0.193359375, "std_total_reward": 0.39941621140925593, "total_tokens": 680, "mean_completion_length": 13.6, "median_completion_length": 14, "min_completion_length": 10, "max_completion_length": 24, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.000390625, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.8, "step_time_sec": 1.346412181854248, "avg_step_time_sec": 19.071550021300446, "gpu_allocated_gb": 32.362796544, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2120.135471343994, "eta_sec": 1697.3679518957397, "timestamp": "2026-02-15T12:08:12.405864"} {"step": 112, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.14, "mean_correctness_reward": 0.14, "mean_length_penalty": -0.006962890625, "mean_total_reward": 0.133037109375, "std_total_reward": 0.346627584800616, "total_tokens": 713, "mean_completion_length": 14.26, "median_completion_length": 13, "min_completion_length": 8, "max_completion_length": 26, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.09238020788514563, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.8, "step_time_sec": 1.4056477546691895, "avg_step_time_sec": 18.913818751062667, "gpu_allocated_gb": 32.36281344, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2121.546905517578, "eta_sec": 1664.4160500935147, "timestamp": "2026-02-15T12:08:13.844629"} {"step": 113, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.18, "mean_correctness_reward": 0.18, "mean_length_penalty": -0.00490234375, "mean_total_reward": 0.17509765625, "std_total_reward": 0.38373977150472305, "total_tokens": 502, "mean_completion_length": 10.04, "median_completion_length": 10, "min_completion_length": 8, "max_completion_length": 12, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.14005859375000002, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.6, "step_time_sec": 1.4411442279815674, "avg_step_time_sec": 18.759193312805312, "gpu_allocated_gb": 32.362834432, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2122.993424654007, "eta_sec": 1632.049818214062, "timestamp": "2026-02-15T12:08:15.194343"} {"step": 114, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.06, "mean_correctness_reward": 0.06, "mean_length_penalty": -0.0052734375, "mean_total_reward": 0.0547265625, "std_total_reward": 0.23758878233017597, "total_tokens": 540, "mean_completion_length": 10.8, "median_completion_length": 10, "min_completion_length": 6, "max_completion_length": 14, "frac_has_think": 0.72, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.09187176629312065, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.8, "step_time_sec": 1.3469908237457275, "avg_step_time_sec": 18.606454694480227, "gpu_allocated_gb": 32.36278528, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2124.3459990024567, "eta_sec": 1600.1551037252996, "timestamp": "2026-02-15T12:08:16.576066"} {"step": 115, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00435546875, "mean_total_reward": -0.00435546875, "std_total_reward": 0.0007098742908354003, "total_tokens": 446, "mean_completion_length": 8.92, "median_completion_length": 8, "min_completion_length": 8, "max_completion_length": 12, "frac_has_think": 0.82, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0002579219404000865, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.8049235343933105, "avg_step_time_sec": 18.46035442352295, "gpu_allocated_gb": 32.36279296, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2126.156998872757, "eta_sec": 1569.130125999451, "timestamp": "2026-02-15T12:08:18.362462"} {"step": 116, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.04, "mean_correctness_reward": 0.04, "mean_length_penalty": -0.004697265625, "mean_total_reward": 0.035302734375, "std_total_reward": 0.1961324402836571, "total_tokens": 481, "mean_completion_length": 9.62, "median_completion_length": 8, "min_completion_length": 6, "max_completion_length": 31, "frac_has_think": 0.54, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.080579719324506, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.8, "step_time_sec": 1.3344924449920654, "avg_step_time_sec": 18.312717682328717, "gpu_allocated_gb": 32.362782208, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2127.4975061416626, "eta_sec": 1538.2682853156123, "timestamp": "2026-02-15T12:08:19.784968"} {"step": 117, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00451171875, "mean_total_reward": -0.00451171875, "std_total_reward": 0.0013302774378013925, "total_tokens": 462, "mean_completion_length": 9.24, "median_completion_length": 9, "min_completion_length": 5, "max_completion_length": 14, "frac_has_think": 0.5, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.00037856437845668034, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.4098403453826904, "avg_step_time_sec": 18.168248645260803, "gpu_allocated_gb": 32.362798592, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2128.913149356842, "eta_sec": 1507.9646375566467, "timestamp": "2026-02-15T12:08:21.187009"} {"step": 118, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.008447265625, "mean_total_reward": -0.008447265625, "std_total_reward": 0.008901864025920993, "total_tokens": 865, "mean_completion_length": 17.3, "median_completion_length": 9, "min_completion_length": 6, "max_completion_length": 70, "frac_has_think": 0.34, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0026016493468486515, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.4193172454833984, "avg_step_time_sec": 18.02630854865252, "gpu_allocated_gb": 32.362878976, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2130.339476108551, "eta_sec": 1478.1573009895067, "timestamp": "2026-02-15T12:08:22.917059"} {"step": 119, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.003251953125, "mean_total_reward": -0.003251953125, "std_total_reward": 0.0008263934764859386, "total_tokens": 333, "mean_completion_length": 6.66, "median_completion_length": 6, "min_completion_length": 4, "max_completion_length": 10, "frac_has_think": 0.02, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.00011880030654210625, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.7062625885009766, "avg_step_time_sec": 17.8891653052899, "gpu_allocated_gb": 32.362921984, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2132.051345348358, "eta_sec": 1449.0223897284818, "timestamp": "2026-02-15T12:08:24.227515"} {"step": 120, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.0029296875, "mean_total_reward": -0.0029296875, "std_total_reward": 0.0, "total_tokens": 300, "mean_completion_length": 6, "median_completion_length": 6, "min_completion_length": 6, "max_completion_length": 6, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2879548072814941, "avg_step_time_sec": 17.750821884473165, "gpu_allocated_gb": 32.36278784, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2133.3446428775787, "eta_sec": 1420.0657507578533, "timestamp": "2026-02-15T12:08:25.516275"} {"step": 121, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00298828125, "mean_total_reward": -0.00298828125, "std_total_reward": 0.000591122888688342, "total_tokens": 306, "mean_completion_length": 6.12, "median_completion_length": 6, "min_completion_length": 4, "max_completion_length": 8, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0002128706930774679, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2935028076171875, "avg_step_time_sec": 17.614810983011544, "gpu_allocated_gb": 32.362824192, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2134.6434729099274, "eta_sec": 1391.570067657912, "timestamp": "2026-02-15T12:08:26.847803"} {"step": 122, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.002822265625, "mean_total_reward": -0.002822265625, "std_total_reward": 0.000719548321085494, "total_tokens": 289, "mean_completion_length": 5.78, "median_completion_length": 6, "min_completion_length": 4, "max_completion_length": 9, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.00023191139731315938, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.3177905082702637, "avg_step_time_sec": 17.481228847972684, "gpu_allocated_gb": 32.362814464, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2135.9665610790253, "eta_sec": 1363.5358501418693, "timestamp": "2026-02-15T12:08:28.134148"} {"step": 123, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00205078125, "mean_total_reward": -0.00205078125, "std_total_reward": 0.0001953125, "total_tokens": 210, "mean_completion_length": 4.2, "median_completion_length": 4, "min_completion_length": 4, "max_completion_length": 5, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2778823375701904, "avg_step_time_sec": 17.349494323497865, "gpu_allocated_gb": 32.362789888, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2137.249755859375, "eta_sec": 1335.9110629093357, "timestamp": "2026-02-15T12:08:29.390088"} {"step": 124, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.004111328125, "mean_total_reward": -0.004111328125, "std_total_reward": 0.0035842504584469195, "total_tokens": 421, "mean_completion_length": 8.42, "median_completion_length": 6, "min_completion_length": 4, "max_completion_length": 26, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.00046273702205139024, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2501251697540283, "avg_step_time_sec": 17.219660701290255, "gpu_allocated_gb": 32.362764288, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2138.5055997371674, "eta_sec": 1308.6942132980594, "timestamp": "2026-02-15T12:08:30.773399"} {"step": 125, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00234375, "mean_total_reward": -0.00234375, "std_total_reward": 0.00047841596538733945, "total_tokens": 240, "mean_completion_length": 4.8, "median_completion_length": 4, "min_completion_length": 4, "max_completion_length": 6, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.3810288906097412, "avg_step_time_sec": 17.09295164680481, "gpu_allocated_gb": 32.362802688, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2139.893019914627, "eta_sec": 1281.9713735103608, "timestamp": "2026-02-15T12:08:32.097246"} {"step": 126, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.001953125, "mean_total_reward": -0.001953125, "std_total_reward": 0.0, "total_tokens": 200, "mean_completion_length": 4, "median_completion_length": 4, "min_completion_length": 4, "max_completion_length": 4, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.728348731994629, "avg_step_time_sec": 16.971010353830124, "gpu_allocated_gb": 32.362838528, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2141.6269714832306, "eta_sec": 1255.854766183429, "timestamp": "2026-02-15T12:08:33.762014"} {"step": 127, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.001953125, "mean_total_reward": -0.001953125, "std_total_reward": 0.0, "total_tokens": 200, "mean_completion_length": 4, "median_completion_length": 4, "min_completion_length": 4, "max_completion_length": 4, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2489385604858398, "avg_step_time_sec": 16.84721451293765, "gpu_allocated_gb": 32.3627776, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2142.8821470737457, "eta_sec": 1229.8466594444485, "timestamp": "2026-02-15T12:08:35.045630"} {"step": 128, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.0021484375, "mean_total_reward": -0.0021484375, "std_total_reward": 0.000390625, "total_tokens": 220, "mean_completion_length": 4.4, "median_completion_length": 4, "min_completion_length": 4, "max_completion_length": 6, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2618129253387451, "avg_step_time_sec": 16.725453563034534, "gpu_allocated_gb": 32.362763264, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2144.150094270706, "eta_sec": 1204.2326565384865, "timestamp": "2026-02-15T12:08:36.305225"} {"step": 129, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.001953125, "mean_total_reward": -0.001953125, "std_total_reward": 0.0, "total_tokens": 200, "mean_completion_length": 4, "median_completion_length": 4, "min_completion_length": 4, "max_completion_length": 4, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2740583419799805, "avg_step_time_sec": 16.60567530550698, "gpu_allocated_gb": 32.362788352, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2145.4303183555603, "eta_sec": 1179.0029466909957, "timestamp": "2026-02-15T12:08:37.574412"} {"step": 130, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.001953125, "mean_total_reward": -0.001953125, "std_total_reward": 0.0, "total_tokens": 200, "mean_completion_length": 4, "median_completion_length": 4, "min_completion_length": 4, "max_completion_length": 4, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2435400485992432, "avg_step_time_sec": 16.487505034299996, "gpu_allocated_gb": 32.362780672, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2146.681410074234, "eta_sec": 1154.1253524009996, "timestamp": "2026-02-15T12:08:38.820107"} {"step": 131, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.001953125, "mean_total_reward": -0.001953125, "std_total_reward": 0.0, "total_tokens": 200, "mean_completion_length": 4, "median_completion_length": 4, "min_completion_length": 4, "max_completion_length": 4, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2443439960479736, "avg_step_time_sec": 16.37114502637441, "gpu_allocated_gb": 32.36278272, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2147.93137550354, "eta_sec": 1129.6090068198343, "timestamp": "2026-02-15T12:08:40.071808"} {"step": 132, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.0019140625, "mean_total_reward": -0.0019140625, "std_total_reward": 0.0001913663861549358, "total_tokens": 196, "mean_completion_length": 3.92, "median_completion_length": 4, "min_completion_length": 2, "max_completion_length": 4, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 7.8125e-05, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2437665462493896, "avg_step_time_sec": 16.25654367425225, "gpu_allocated_gb": 32.362745856, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2149.1812064647675, "eta_sec": 1105.444969849153, "timestamp": "2026-02-15T12:08:41.319009"} {"step": 133, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.0013671875, "mean_total_reward": -0.0013671875, "std_total_reward": 0.00047841596538733945, "total_tokens": 140, "mean_completion_length": 2.8, "median_completion_length": 2, "min_completion_length": 2, "max_completion_length": 4, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.00015625, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.238485336303711, "avg_step_time_sec": 16.143625942387974, "gpu_allocated_gb": 32.362788864, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2150.42759680748, "eta_sec": 1081.6229381399942, "timestamp": "2026-02-15T12:08:42.564548"} {"step": 134, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.0009765625, "mean_total_reward": -0.0009765625, "std_total_reward": 0.0, "total_tokens": 100, "mean_completion_length": 2, "median_completion_length": 2, "min_completion_length": 2, "max_completion_length": 2, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2301604747772217, "avg_step_time_sec": 16.03233142397297, "gpu_allocated_gb": 32.362781696, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2151.668218612671, "eta_sec": 1058.133873982216, "timestamp": "2026-02-15T12:08:43.837228"} {"step": 135, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.0009765625, "mean_total_reward": -0.0009765625, "std_total_reward": 0.0, "total_tokens": 100, "mean_completion_length": 2, "median_completion_length": 2, "min_completion_length": 2, "max_completion_length": 2, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.285768747329712, "avg_step_time_sec": 15.923097626368206, "gpu_allocated_gb": 32.362793472, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2152.9610295295715, "eta_sec": 1035.0013457139335, "timestamp": "2026-02-15T12:08:45.084867"} {"step": 136, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.0009765625, "mean_total_reward": -0.0009765625, "std_total_reward": 0.0, "total_tokens": 100, "mean_completion_length": 2, "median_completion_length": 2, "min_completion_length": 2, "max_completion_length": 2, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2287302017211914, "avg_step_time_sec": 15.81505080706933, "gpu_allocated_gb": 32.362751488, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2154.1960504055023, "eta_sec": 1012.1632516524371, "timestamp": "2026-02-15T12:08:46.327600"} {"step": 137, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.000908203125, "mean_total_reward": -0.000908203125, "std_total_reward": 0.00016942726145407688, "total_tokens": 93, "mean_completion_length": 1.86, "median_completion_length": 2, "min_completion_length": 1, "max_completion_length": 2, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 4.4751715771053124e-05, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2371587753295898, "avg_step_time_sec": 15.708642836034732, "gpu_allocated_gb": 32.362770944, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2155.440062046051, "eta_sec": 989.6444986701881, "timestamp": "2026-02-15T12:08:47.562550"} {"step": 138, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00078125, "mean_total_reward": -0.00078125, "std_total_reward": 0.00023920798269366973, "total_tokens": 80, "mean_completion_length": 1.6, "median_completion_length": 2, "min_completion_length": 1, "max_completion_length": 2, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 8.950343154210625e-05, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.6040847301483154, "avg_step_time_sec": 15.606435893238455, "gpu_allocated_gb": 32.362782208, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2157.0507888793945, "eta_sec": 967.5990253807843, "timestamp": "2026-02-15T12:08:49.183232"} {"step": 139, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.000791015625, "mean_total_reward": -0.000791015625, "std_total_reward": 0.00023700509959983588, "total_tokens": 81, "mean_completion_length": 1.62, "median_completion_length": 2, "min_completion_length": 1, "max_completion_length": 2, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 2.9296875e-05, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2346410751342773, "avg_step_time_sec": 15.503041685913963, "gpu_allocated_gb": 32.362787328, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2158.293071746826, "eta_sec": 945.6855428407517, "timestamp": "2026-02-15T12:08:50.430521"} {"step": 140, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.000546875, "mean_total_reward": -0.000546875, "std_total_reward": 0.0001586726250905461, "total_tokens": 56, "mean_completion_length": 1.12, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 2, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 7.8125e-05, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2452375888824463, "avg_step_time_sec": 15.401200228078025, "gpu_allocated_gb": 32.362814976, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2159.5449693202972, "eta_sec": 924.0720136846815, "timestamp": "2026-02-15T12:08:51.649028"} {"step": 141, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.0005859375, "mean_total_reward": -0.0005859375, "std_total_reward": 0.0001953125, "total_tokens": 60, "mean_completion_length": 1.2, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 2, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2030017375946045, "avg_step_time_sec": 15.300503784883107, "gpu_allocated_gb": 32.36278016, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2160.7535824775696, "eta_sec": 902.7297233081033, "timestamp": "2026-02-15T12:08:52.863474"} {"step": 142, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2150263786315918, "avg_step_time_sec": 15.201310282022181, "gpu_allocated_gb": 32.362774016, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2161.973851442337, "eta_sec": 881.6759963572865, "timestamp": "2026-02-15T12:08:54.070867"} {"step": 143, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2010300159454346, "avg_step_time_sec": 15.103406224217448, "gpu_allocated_gb": 32.362758656, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2163.180398464203, "eta_sec": 860.8941547803945, "timestamp": "2026-02-15T12:08:55.298264"} {"step": 144, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2154450416564941, "avg_step_time_sec": 15.006962049338552, "gpu_allocated_gb": 32.362783232, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2164.4060683250427, "eta_sec": 840.3898747629589, "timestamp": "2026-02-15T12:08:56.526237"} {"step": 145, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2176260948181152, "avg_step_time_sec": 14.911863180686687, "gpu_allocated_gb": 32.3627648, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2165.6310317516327, "eta_sec": 820.1524749377678, "timestamp": "2026-02-15T12:08:57.748185"} {"step": 146, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2205302715301514, "avg_step_time_sec": 14.818086927884245, "gpu_allocated_gb": 32.362792448, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2166.8581671714783, "eta_sec": 800.1766941057492, "timestamp": "2026-02-15T12:08:58.975086"} {"step": 147, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2071657180786133, "avg_step_time_sec": 14.725495627137269, "gpu_allocated_gb": 32.362774016, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2168.073805809021, "eta_sec": 780.4512682382752, "timestamp": "2026-02-15T12:09:00.169080"} {"step": 148, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.1956079006195068, "avg_step_time_sec": 14.634077466822959, "gpu_allocated_gb": 32.36278016, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2169.2748215198517, "eta_sec": 760.9720282747938, "timestamp": "2026-02-15T12:09:01.399460"} {"step": 149, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2347736358642578, "avg_step_time_sec": 14.544149253192364, "gpu_allocated_gb": 32.362792448, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2170.517758846283, "eta_sec": 741.7516119128105, "timestamp": "2026-02-15T12:09:02.623038"} {"step": 150, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.639561653137207, "avg_step_time_sec": 14.458118669191997, "gpu_allocated_gb": 32.3627392, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2172.1638202667236, "eta_sec": 722.9059334595999, "timestamp": "2026-02-15T12:09:06.191449"} {"step": 151, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.25315260887146, "avg_step_time_sec": 14.370668562832257, "gpu_allocated_gb": 32.362767872, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2175.3397767543793, "eta_sec": 704.1627595787807, "timestamp": "2026-02-15T12:09:07.496950"} {"step": 152, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2911884784698486, "avg_step_time_sec": 14.284619351750926, "gpu_allocated_gb": 32.36283136, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2176.640824317932, "eta_sec": 685.6617288840445, "timestamp": "2026-02-15T12:09:08.762484"} {"step": 153, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2264790534973145, "avg_step_time_sec": 14.199272029540118, "gpu_allocated_gb": 32.362781184, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2177.874758720398, "eta_sec": 667.3657853883856, "timestamp": "2026-02-15T12:09:10.009276"} {"step": 154, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2546424865722656, "avg_step_time_sec": 14.11521599354682, "gpu_allocated_gb": 32.362824192, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2179.1363565921783, "eta_sec": 649.2999357031538, "timestamp": "2026-02-15T12:09:11.246778"} {"step": 155, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2387356758117676, "avg_step_time_sec": 14.032141926980788, "gpu_allocated_gb": 32.362782208, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2180.380944252014, "eta_sec": 631.4463867141354, "timestamp": "2026-02-15T12:09:12.499884"} {"step": 156, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2463412284851074, "avg_step_time_sec": 13.950181666092995, "gpu_allocated_gb": 32.362803712, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2181.633028268814, "eta_sec": 613.8079933080918, "timestamp": "2026-02-15T12:09:13.764601"} {"step": 157, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2351174354553223, "avg_step_time_sec": 13.869193995834156, "gpu_allocated_gb": 32.362784256, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2182.876441717148, "eta_sec": 596.3753418208687, "timestamp": "2026-02-15T12:09:14.994668"} {"step": 158, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2423274517059326, "avg_step_time_sec": 13.789277118972585, "gpu_allocated_gb": 32.362816, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2184.124587535858, "eta_sec": 579.1496389968486, "timestamp": "2026-02-15T12:09:16.237627"} {"step": 159, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.195594310760498, "avg_step_time_sec": 13.710071566719678, "gpu_allocated_gb": 32.362809856, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2185.326675891876, "eta_sec": 562.1129342355068, "timestamp": "2026-02-15T12:09:17.434959"} {"step": 160, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.23575758934021, "avg_step_time_sec": 13.632107104361058, "gpu_allocated_gb": 32.362781184, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2186.5683188438416, "eta_sec": 545.2842841744423, "timestamp": "2026-02-15T12:09:18.698205"} {"step": 161, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.7238974571228027, "avg_step_time_sec": 13.558143069285043, "gpu_allocated_gb": 32.362819072, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2188.298748254776, "eta_sec": 528.7675797021167, "timestamp": "2026-02-15T12:09:20.417855"} {"step": 162, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2471086978912354, "avg_step_time_sec": 13.482149029955451, "gpu_allocated_gb": 32.362753536, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2189.551507949829, "eta_sec": 512.3216631383071, "timestamp": "2026-02-15T12:09:21.662790"} {"step": 163, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.227229118347168, "avg_step_time_sec": 13.406965472215523, "gpu_allocated_gb": 32.36279552, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2190.7848172187805, "eta_sec": 496.05772247197433, "timestamp": "2026-02-15T12:09:22.901501"} {"step": 164, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2121195793151855, "avg_step_time_sec": 13.3326066557954, "gpu_allocated_gb": 32.362760704, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2192.005999326706, "eta_sec": 479.9738396086344, "timestamp": "2026-02-15T12:09:24.140531"} {"step": 165, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2512588500976562, "avg_step_time_sec": 13.259386366063898, "gpu_allocated_gb": 32.362817024, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2193.264920949936, "eta_sec": 464.07852281223643, "timestamp": "2026-02-15T12:09:25.366173"} {"step": 166, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2024240493774414, "avg_step_time_sec": 13.18675406295133, "gpu_allocated_gb": 32.362758656, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2194.474839925766, "eta_sec": 448.34963814034523, "timestamp": "2026-02-15T12:09:26.611026"} {"step": 167, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2433958053588867, "avg_step_time_sec": 13.115236947636404, "gpu_allocated_gb": 32.362817024, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2195.727208137512, "eta_sec": 432.80281927200133, "timestamp": "2026-02-15T12:09:27.849037"} {"step": 168, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.223698377609253, "avg_step_time_sec": 13.044453979957671, "gpu_allocated_gb": 32.362812928, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2196.957038640976, "eta_sec": 417.4225273586455, "timestamp": "2026-02-15T12:09:29.077624"} {"step": 169, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2277288436889648, "avg_step_time_sec": 12.974532529447206, "gpu_allocated_gb": 32.362766848, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2198.1915578842163, "eta_sec": 402.21050841286336, "timestamp": "2026-02-15T12:09:30.296855"} {"step": 170, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2025823593139648, "avg_step_time_sec": 12.90528576374054, "gpu_allocated_gb": 32.362753536, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2199.4013011455536, "eta_sec": 387.1585729122162, "timestamp": "2026-02-15T12:09:31.521118"} {"step": 171, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.227935552597046, "avg_step_time_sec": 12.836997166014553, "gpu_allocated_gb": 32.362767872, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2200.6374900341034, "eta_sec": 372.27291781442204, "timestamp": "2026-02-15T12:09:32.766590"} {"step": 172, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2302980422973633, "avg_step_time_sec": 12.769516357155734, "gpu_allocated_gb": 32.362835456, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2201.8740339279175, "eta_sec": 357.54645800036053, "timestamp": "2026-02-15T12:09:33.978918"} {"step": 173, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.6335458755493164, "avg_step_time_sec": 12.705146585585753, "gpu_allocated_gb": 32.36277504, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2203.513543844223, "eta_sec": 343.03895781081536, "timestamp": "2026-02-15T12:09:35.618296"} {"step": 174, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2143619060516357, "avg_step_time_sec": 12.63910759317464, "gpu_allocated_gb": 32.362755584, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2204.733687400818, "eta_sec": 328.6167974225406, "timestamp": "2026-02-15T12:09:36.835838"} {"step": 175, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2050817012786865, "avg_step_time_sec": 12.573770302363805, "gpu_allocated_gb": 32.362770944, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2205.9443969726562, "eta_sec": 314.3442575590951, "timestamp": "2026-02-15T12:09:38.057952"} {"step": 176, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2155334949493408, "avg_step_time_sec": 12.50923486595804, "gpu_allocated_gb": 32.362774016, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2207.1663377285004, "eta_sec": 300.22163678299296, "timestamp": "2026-02-15T12:09:39.274224"} {"step": 177, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2106842994689941, "avg_step_time_sec": 12.445401246938328, "gpu_allocated_gb": 32.362750464, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2208.383103132248, "eta_sec": 286.24422867958157, "timestamp": "2026-02-15T12:09:40.494066"} {"step": 178, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2125294208526611, "avg_step_time_sec": 12.382295225443466, "gpu_allocated_gb": 32.362778112, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2209.602066755295, "eta_sec": 272.41049495975625, "timestamp": "2026-02-15T12:09:41.703857"} {"step": 179, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.21065354347229, "avg_step_time_sec": 12.319883819398934, "gpu_allocated_gb": 32.362781184, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2210.8181540966034, "eta_sec": 258.71756020737763, "timestamp": "2026-02-15T12:09:42.974278"} {"step": 180, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2667021751403809, "avg_step_time_sec": 12.258477254708607, "gpu_allocated_gb": 32.362833408, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2212.090565443039, "eta_sec": 245.16954509417215, "timestamp": "2026-02-15T12:09:44.205775"} {"step": 181, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.21866774559021, "avg_step_time_sec": 12.197483832006297, "gpu_allocated_gb": 32.362789376, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2213.3154876232147, "eta_sec": 231.75219280811964, "timestamp": "2026-02-15T12:09:45.431632"} {"step": 182, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.219571828842163, "avg_step_time_sec": 12.137165634186713, "gpu_allocated_gb": 32.362765824, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2214.5417902469635, "eta_sec": 218.46898141536084, "timestamp": "2026-02-15T12:09:46.661767"} {"step": 183, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2220215797424316, "avg_step_time_sec": 12.077520038260788, "gpu_allocated_gb": 32.36275456, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2215.771108150482, "eta_sec": 205.3178406504334, "timestamp": "2026-02-15T12:09:47.883023"} {"step": 184, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2124650478363037, "avg_step_time_sec": 12.018470826356307, "gpu_allocated_gb": 32.362770944, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2216.9911913871765, "eta_sec": 192.2955332217009, "timestamp": "2026-02-15T12:09:49.102240"} {"step": 185, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.6322405338287354, "avg_step_time_sec": 11.962329040991293, "gpu_allocated_gb": 32.362799616, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2218.628843307495, "eta_sec": 179.4349356148694, "timestamp": "2026-02-15T12:09:50.759040"} {"step": 186, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2320613861083984, "avg_step_time_sec": 11.904639429943536, "gpu_allocated_gb": 32.362830336, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2219.8668444156647, "eta_sec": 166.66495201920952, "timestamp": "2026-02-15T12:09:52.000629"} {"step": 187, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2334060668945312, "avg_step_time_sec": 11.847574010889797, "gpu_allocated_gb": 32.362788352, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2221.109279155731, "eta_sec": 154.01846214156737, "timestamp": "2026-02-15T12:09:53.274838"} {"step": 188, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.274733543395996, "avg_step_time_sec": 11.79133549776483, "gpu_allocated_gb": 32.3628416, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2222.3912992477417, "eta_sec": 141.49602597317795, "timestamp": "2026-02-15T12:09:54.507931"} {"step": 189, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.220282793045044, "avg_step_time_sec": 11.735404001972663, "gpu_allocated_gb": 32.362792448, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2223.618245124817, "eta_sec": 129.08944402169928, "timestamp": "2026-02-15T12:09:55.712039"} {"step": 190, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.1913795471191406, "avg_step_time_sec": 11.679909136420802, "gpu_allocated_gb": 32.362753536, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2224.815311193466, "eta_sec": 116.79909136420802, "timestamp": "2026-02-15T12:09:56.926033"} {"step": 191, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2118394374847412, "avg_step_time_sec": 11.625102488782394, "gpu_allocated_gb": 32.362774016, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2226.0334129333496, "eta_sec": 104.62592239904154, "timestamp": "2026-02-15T12:09:58.242264"} {"step": 192, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.3090934753417969, "avg_step_time_sec": 11.571373275170723, "gpu_allocated_gb": 32.362886656, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2227.349367618561, "eta_sec": 92.57098620136578, "timestamp": "2026-02-15T12:09:59.444111"} {"step": 193, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.193842887878418, "avg_step_time_sec": 11.5176036876718, "gpu_allocated_gb": 32.362758656, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2228.5485892295837, "eta_sec": 80.6232258137026, "timestamp": "2026-02-15T12:10:00.703410"} {"step": 194, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.258042335510254, "avg_step_time_sec": 11.464719350804987, "gpu_allocated_gb": 32.362847744, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2229.8144133090973, "eta_sec": 68.78831610482993, "timestamp": "2026-02-15T12:10:01.926456"} {"step": 195, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2118966579437256, "avg_step_time_sec": 11.41214077289288, "gpu_allocated_gb": 32.36278016, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2231.033007144928, "eta_sec": 57.0607038644644, "timestamp": "2026-02-15T12:10:03.138256"} {"step": 196, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.204622507095337, "avg_step_time_sec": 11.36006159806738, "gpu_allocated_gb": 32.362762752, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2232.244201183319, "eta_sec": 45.44024639226952, "timestamp": "2026-02-15T12:10:04.351382"} {"step": 197, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.606395959854126, "avg_step_time_sec": 11.310550604980005, "gpu_allocated_gb": 32.362781184, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2233.8562495708466, "eta_sec": 33.93165181494001, "timestamp": "2026-02-15T12:10:06.142743"} {"step": 198, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.4744317531585693, "avg_step_time_sec": 11.260873237041512, "gpu_allocated_gb": 32.362960384, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2235.3393704891205, "eta_sec": 22.521746474083024, "timestamp": "2026-02-15T12:10:07.438326"} {"step": 199, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.00048828125, "std_total_reward": 0.0, "total_tokens": 50, "mean_completion_length": 1, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 1, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.2025697231292725, "avg_step_time_sec": 11.210328998278134, "gpu_allocated_gb": 32.362765824, "gpu_reserved_gb": 79.960211456, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2236.548176765442, "eta_sec": 11.210328998278134, "timestamp": "2026-02-15T12:10:08.696192"}