deepscaleR / gdpo_lambda2.0 /logs /step_stats.jsonl
brikdavies's picture
Add gdpo_lambda2.0/logs
8df2912 verified
Raw
History Blame Contribute Delete
159 kB
{"step": 0, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.8160546875, "mean_total_reward": -1.032109375, "std_total_reward": 0.8217654691748278, "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": 0.34345541424176934, "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": "2026-02-15T10:57:55.145475"}
{"step": 1, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.22, "mean_correctness_reward": 0.22, "mean_length_penalty": -0.94798828125, "mean_total_reward": -1.6759765625, "std_total_reward": 0.5692826344580588, "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": 0.38377250099024696, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 28.030738830566406, "avg_step_time_sec": 28.030738830566406, "gpu_allocated_gb": 32.365621248, "gpu_reserved_gb": 79.92246272, "gpu_peak_gb": 73.906566656, "elapsed_sec": 28.036179780960083, "eta_sec": 5578.117027282715, "timestamp": "2026-02-15T10:58:24.772621"}
{"step": 2, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.72, "mean_correctness_reward": 0.72, "mean_length_penalty": -0.795693359375, "mean_total_reward": -0.87138671875, "std_total_reward": 0.757080108042264, "total_tokens": 81479, "mean_completion_length": 1629.58, "median_completion_length": 1716, "min_completion_length": 792, "max_completion_length": 2048, "frac_has_think": 0.76, "frac_has_boxed": 0.76, "mean_reward_std_per_prompt": 0.6070057845931066, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 29.600348472595215, "avg_step_time_sec": 28.81554365158081, "gpu_allocated_gb": 32.365654016, "gpu_reserved_gb": 79.95392, "gpu_peak_gb": 74.127090176, "elapsed_sec": 57.64388966560364, "eta_sec": 5705.4776430130005, "timestamp": "2026-02-15T10:58:52.594611"}
{"step": 3, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.66, "mean_correctness_reward": 0.66, "mean_length_penalty": -0.786689453125, "mean_total_reward": -0.91337890625, "std_total_reward": 0.8435044051294447, "total_tokens": 80557, "mean_completion_length": 1611.14, "median_completion_length": 1812, "min_completion_length": 743, "max_completion_length": 2048, "frac_has_think": 0.72, "frac_has_boxed": 0.72, "mean_reward_std_per_prompt": 0.5089097058786443, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 27.73742985725403, "avg_step_time_sec": 28.456172386805218, "gpu_allocated_gb": 32.365626368, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.127090176, "elapsed_sec": 85.38826036453247, "eta_sec": 5605.865960200628, "timestamp": "2026-02-15T10:59:20.807857"}
{"step": 4, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.24, "mean_correctness_reward": 0.24, "mean_length_penalty": -0.94328125, "mean_total_reward": -1.6465625, "std_total_reward": 0.638796004040282, "total_tokens": 96592, "mean_completion_length": 1931.84, "median_completion_length": 2048, "min_completion_length": 1090, "max_completion_length": 2048, "frac_has_think": 0.26, "frac_has_boxed": 0.26, "mean_reward_std_per_prompt": 0.2922835900204994, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.6, "step_time_sec": 28.32918620109558, "avg_step_time_sec": 28.424425840377808, "gpu_allocated_gb": 32.365638656, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.127090176, "elapsed_sec": 113.72325420379639, "eta_sec": 5571.18746471405, "timestamp": "2026-02-15T10:59:50.497143"}
{"step": 5, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.42, "mean_correctness_reward": 0.42, "mean_length_penalty": -0.824443359375, "mean_total_reward": -1.22888671875, "std_total_reward": 0.9340752650205799, "total_tokens": 84423, "mean_completion_length": 1688.46, "median_completion_length": 2048, "min_completion_length": 791, "max_completion_length": 2048, "frac_has_think": 0.56, "frac_has_boxed": 0.56, "mean_reward_std_per_prompt": 0.4000419471220775, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 29.674192667007446, "avg_step_time_sec": 28.674379205703737, "gpu_allocated_gb": 32.365667328, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.156304384, "elapsed_sec": 143.405291557312, "eta_sec": 5591.503945112228, "timestamp": "2026-02-15T11:00:18.625261"}
{"step": 6, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.808984375, "mean_total_reward": -1.01796875, "std_total_reward": 0.8434698247209818, "total_tokens": 82840, "mean_completion_length": 1656.8, "median_completion_length": 1749, "min_completion_length": 822, "max_completion_length": 2048, "frac_has_think": 0.68, "frac_has_boxed": 0.7, "mean_reward_std_per_prompt": 0.6264407100302983, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 28.013072967529297, "avg_step_time_sec": 28.564161499341328, "gpu_allocated_gb": 32.365632512, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.156304384, "elapsed_sec": 171.42462348937988, "eta_sec": 5541.447330872217, "timestamp": "2026-02-15T11:00:46.537938"}
{"step": 7, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.48, "mean_correctness_reward": 0.48, "mean_length_penalty": -0.858603515625, "mean_total_reward": -1.23720703125, "std_total_reward": 0.8255877222024269, "total_tokens": 87921, "mean_completion_length": 1758.42, "median_completion_length": 2048, "min_completion_length": 829, "max_completion_length": 2048, "frac_has_think": 0.58, "frac_has_boxed": 0.58, "mean_reward_std_per_prompt": 0.3559855415138814, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 28.261310815811157, "avg_step_time_sec": 28.520897115979874, "gpu_allocated_gb": 32.36562432, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.156304384, "elapsed_sec": 199.69243574142456, "eta_sec": 5504.533143384116, "timestamp": "2026-02-15T11:01:15.180738"}
{"step": 8, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.796298828125, "mean_total_reward": -0.99259765625, "std_total_reward": 0.6935917329082132, "total_tokens": 81541, "mean_completion_length": 1630.82, "median_completion_length": 1668, "min_completion_length": 877, "max_completion_length": 2048, "frac_has_think": 0.84, "frac_has_boxed": 0.84, "mean_reward_std_per_prompt": 0.4630827619652521, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 28.180207014083862, "avg_step_time_sec": 28.478310853242874, "gpu_allocated_gb": 32.365616128, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.156304384, "elapsed_sec": 227.87919569015503, "eta_sec": 5467.835683822632, "timestamp": "2026-02-15T11:01:42.824468"}
{"step": 9, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.66, "mean_correctness_reward": 0.66, "mean_length_penalty": -0.733310546875, "mean_total_reward": -0.80662109375, "std_total_reward": 0.8137336518363865, "total_tokens": 75091, "mean_completion_length": 1501.82, "median_completion_length": 1534, "min_completion_length": 653, "max_completion_length": 2048, "frac_has_think": 0.84, "frac_has_boxed": 0.84, "mean_reward_std_per_prompt": 0.5346072052387677, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 28.188136339187622, "avg_step_time_sec": 28.44606924057007, "gpu_allocated_gb": 32.365641728, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.156304384, "elapsed_sec": 256.07338857650757, "eta_sec": 5433.199224948883, "timestamp": "2026-02-15T11:02:10.682863"}
{"step": 10, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.56, "mean_correctness_reward": 0.56, "mean_length_penalty": -0.90291015625, "mean_total_reward": -1.2458203125, "std_total_reward": 0.6970836115728425, "total_tokens": 92458, "mean_completion_length": 1849.16, "median_completion_length": 2025, "min_completion_length": 912, "max_completion_length": 2048, "frac_has_think": 0.66, "frac_has_boxed": 0.66, "mean_reward_std_per_prompt": 0.68436409444089, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 27.768264293670654, "avg_step_time_sec": 28.378288745880127, "gpu_allocated_gb": 32.365712384, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 283.84884333610535, "eta_sec": 5391.874861717224, "timestamp": "2026-02-15T11:02:39.797699"}
{"step": 11, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.58, "mean_correctness_reward": 0.58, "mean_length_penalty": -0.863779296875, "mean_total_reward": -1.14755859375, "std_total_reward": 0.7576697335014029, "total_tokens": 88451, "mean_completion_length": 1769.02, "median_completion_length": 1957, "min_completion_length": 979, "max_completion_length": 2048, "frac_has_think": 0.6, "frac_has_boxed": 0.6, "mean_reward_std_per_prompt": 0.5582519919350974, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 28.763071060180664, "avg_step_time_sec": 28.413268956271086, "gpu_allocated_gb": 32.36563456, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 312.61768078804016, "eta_sec": 5370.107832735235, "timestamp": "2026-02-15T11:03:08.667579"}
{"step": 12, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.52, "mean_correctness_reward": 0.52, "mean_length_penalty": -0.789814453125, "mean_total_reward": -1.05962890625, "std_total_reward": 0.7939293224193148, "total_tokens": 80877, "mean_completion_length": 1617.54, "median_completion_length": 1570, "min_completion_length": 873, "max_completion_length": 2048, "frac_has_think": 0.74, "frac_has_boxed": 0.74, "mean_reward_std_per_prompt": 0.5439528710382309, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 28.871024131774902, "avg_step_time_sec": 28.451415220896404, "gpu_allocated_gb": 32.365643776, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 341.49437832832336, "eta_sec": 5348.866061528524, "timestamp": "2026-02-15T11:03:36.446162"}
{"step": 13, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.48, "mean_correctness_reward": 0.48, "mean_length_penalty": -0.8390625, "mean_total_reward": -1.198125, "std_total_reward": 0.8559249292058475, "total_tokens": 85920, "mean_completion_length": 1718.4, "median_completion_length": 2048, "min_completion_length": 929, "max_completion_length": 2048, "frac_has_think": 0.58, "frac_has_boxed": 0.58, "mean_reward_std_per_prompt": 0.5504662696405266, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 27.769163846969604, "avg_step_time_sec": 28.398934345978958, "gpu_allocated_gb": 32.365641728, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 369.2697916030884, "eta_sec": 5310.6007226980655, "timestamp": "2026-02-15T11:04:05.169831"}
{"step": 14, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.56, "mean_correctness_reward": 0.56, "mean_length_penalty": -0.721796875, "mean_total_reward": -0.88359375, "std_total_reward": 0.7500019582087066, "total_tokens": 73912, "mean_completion_length": 1478.24, "median_completion_length": 1363, "min_completion_length": 954, "max_completion_length": 2048, "frac_has_think": 0.92, "frac_has_boxed": 0.92, "mean_reward_std_per_prompt": 0.2708422906339235, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 28.75686001777649, "avg_step_time_sec": 28.424500465393066, "gpu_allocated_gb": 32.365654016, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 398.03406953811646, "eta_sec": 5286.95708656311, "timestamp": "2026-02-15T11:04:32.375725"}
{"step": 15, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.68, "mean_correctness_reward": 0.68, "mean_length_penalty": -0.760712890625, "mean_total_reward": -0.84142578125, "std_total_reward": 0.8626544734450807, "total_tokens": 77897, "mean_completion_length": 1557.94, "median_completion_length": 1766, "min_completion_length": 644, "max_completion_length": 2048, "frac_has_think": 0.7, "frac_has_boxed": 0.7, "mean_reward_std_per_prompt": 0.4145289895864867, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 27.225886821746826, "avg_step_time_sec": 28.344592889149983, "gpu_allocated_gb": 32.36562944, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 425.26699590682983, "eta_sec": 5243.749684492746, "timestamp": "2026-02-15T11:05:00.574480"}
{"step": 16, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.5, "mean_correctness_reward": 0.5, "mean_length_penalty": -0.860703125, "mean_total_reward": -1.22140625, "std_total_reward": 0.8226196406116968, "total_tokens": 88136, "mean_completion_length": 1762.72, "median_completion_length": 2048, "min_completion_length": 759, "max_completion_length": 2048, "frac_has_think": 0.68, "frac_has_boxed": 0.68, "mean_reward_std_per_prompt": 0.42944772294333167, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 28.10014510154724, "avg_step_time_sec": 28.329314902424812, "gpu_allocated_gb": 32.365674496, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 453.3739538192749, "eta_sec": 5212.5939420461655, "timestamp": "2026-02-15T11:05:28.974724"}
{"step": 17, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.58, "mean_correctness_reward": 0.58, "mean_length_penalty": -0.801240234375, "mean_total_reward": -1.02248046875, "std_total_reward": 0.8492826124952993, "total_tokens": 82047, "mean_completion_length": 1640.94, "median_completion_length": 1840, "min_completion_length": 894, "max_completion_length": 2048, "frac_has_think": 0.72, "frac_has_boxed": 0.74, "mean_reward_std_per_prompt": 0.2067717531426648, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 28.394806623458862, "avg_step_time_sec": 28.333167356603287, "gpu_allocated_gb": 32.365631488, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 481.77581810951233, "eta_sec": 5184.969626258401, "timestamp": "2026-02-15T11:05:57.331247"}
{"step": 18, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.54, "mean_correctness_reward": 0.54, "mean_length_penalty": -0.829599609375, "mean_total_reward": -1.11919921875, "std_total_reward": 0.8585258013536726, "total_tokens": 84951, "mean_completion_length": 1699.02, "median_completion_length": 1883, "min_completion_length": 602, "max_completion_length": 2048, "frac_has_think": 0.62, "frac_has_boxed": 0.62, "mean_reward_std_per_prompt": 0.4732347257524955, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 28.351396083831787, "avg_step_time_sec": 28.334180063671536, "gpu_allocated_gb": 32.36563456, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 510.13418555259705, "eta_sec": 5156.820771588219, "timestamp": "2026-02-15T11:06:25.527165"}
{"step": 19, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.56, "mean_correctness_reward": 0.56, "mean_length_penalty": -0.85595703125, "mean_total_reward": -1.1519140625, "std_total_reward": 0.7826402464159753, "total_tokens": 87650, "mean_completion_length": 1753, "median_completion_length": 1942, "min_completion_length": 1014, "max_completion_length": 2048, "frac_has_think": 0.66, "frac_has_boxed": 0.66, "mean_reward_std_per_prompt": 0.5215769221603203, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 28.249868154525757, "avg_step_time_sec": 28.329742594769126, "gpu_allocated_gb": 32.365650944, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 538.390189409256, "eta_sec": 5127.683409653212, "timestamp": "2026-02-15T11:06:54.345749"}
{"step": 20, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.68, "mean_correctness_reward": 0.68, "mean_length_penalty": -0.782724609375, "mean_total_reward": -0.88544921875, "std_total_reward": 0.7836544359465067, "total_tokens": 80151, "mean_completion_length": 1603.02, "median_completion_length": 1542, "min_completion_length": 917, "max_completion_length": 2048, "frac_has_think": 0.72, "frac_has_boxed": 0.72, "mean_reward_std_per_prompt": 0.4212731820878171, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 28.621649503707886, "avg_step_time_sec": 28.344337940216064, "gpu_allocated_gb": 32.365620224, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 567.0184926986694, "eta_sec": 5101.980829238892, "timestamp": "2026-02-15T11:07:21.930857"}
{"step": 21, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.78, "mean_correctness_reward": 0.78, "mean_length_penalty": -0.699267578125, "mean_total_reward": -0.61853515625, "std_total_reward": 0.807603061819106, "total_tokens": 71605, "mean_completion_length": 1432.1, "median_completion_length": 1301, "min_completion_length": 702, "max_completion_length": 2048, "frac_has_think": 0.86, "frac_has_boxed": 0.86, "mean_reward_std_per_prompt": 0.5139640571078002, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 27.769338130950928, "avg_step_time_sec": 28.316956996917725, "gpu_allocated_gb": 32.365652992, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.261332992, "elapsed_sec": 594.7946524620056, "eta_sec": 5068.735302448273, "timestamp": "2026-02-15T11:07:49.853433"}
{"step": 22, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.46, "mean_correctness_reward": 0.46, "mean_length_penalty": -0.827041015625, "mean_total_reward": -1.19408203125, "std_total_reward": 0.839444168433562, "total_tokens": 84689, "mean_completion_length": 1693.78, "median_completion_length": 2012, "min_completion_length": 782, "max_completion_length": 2048, "frac_has_think": 0.72, "frac_has_boxed": 0.72, "mean_reward_std_per_prompt": 0.38920419112171595, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 28.43941879272461, "avg_step_time_sec": 28.322523442181673, "gpu_allocated_gb": 32.365769728, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.392025088, "elapsed_sec": 623.2401103973389, "eta_sec": 5041.409172708338, "timestamp": "2026-02-15T11:08:18.895726"}
{"step": 23, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.44, "mean_correctness_reward": 0.44, "mean_length_penalty": -0.8037109375, "mean_total_reward": -1.167421875, "std_total_reward": 0.9183013061181822, "total_tokens": 82300, "mean_completion_length": 1646, "median_completion_length": 1958, "min_completion_length": 710, "max_completion_length": 2048, "frac_has_think": 0.6, "frac_has_boxed": 0.6, "mean_reward_std_per_prompt": 0.46641530185584, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 29.388689517974854, "avg_step_time_sec": 28.36887848895529, "gpu_allocated_gb": 32.365832192, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 652.6361367702484, "eta_sec": 5021.291492545086, "timestamp": "2026-02-15T11:08:48.271825"}
{"step": 24, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.8, "mean_correctness_reward": 0.8, "mean_length_penalty": -0.646806640625, "mean_total_reward": -0.49361328125, "std_total_reward": 0.7937859524972447, "total_tokens": 66233, "mean_completion_length": 1324.66, "median_completion_length": 1240, "min_completion_length": 619, "max_completion_length": 2048, "frac_has_think": 0.8, "frac_has_boxed": 0.8, "mean_reward_std_per_prompt": 0.19360674506245223, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 28.35512661933899, "avg_step_time_sec": 28.368305494387943, "gpu_allocated_gb": 32.365621248, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 680.9984676837921, "eta_sec": 4992.821767012278, "timestamp": "2026-02-15T11:09:14.777935"}
{"step": 25, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.591083984375, "mean_total_reward": -0.58216796875, "std_total_reward": 0.46144544683622546, "total_tokens": 60527, "mean_completion_length": 1210.54, "median_completion_length": 1163, "min_completion_length": 735, "max_completion_length": 1891, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.23439236735872684, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 26.542267322540283, "avg_step_time_sec": 28.29526396751404, "gpu_allocated_gb": 32.365627392, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 707.5482177734375, "eta_sec": 4951.671194314957, "timestamp": "2026-02-15T11:09:39.714891"}
{"step": 26, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.7, "mean_correctness_reward": 0.7, "mean_length_penalty": -0.704345703125, "mean_total_reward": -0.70869140625, "std_total_reward": 0.9224601155991601, "total_tokens": 72125, "mean_completion_length": 1442.5, "median_completion_length": 1422, "min_completion_length": 249, "max_completion_length": 2048, "frac_has_think": 0.82, "frac_has_boxed": 0.82, "mean_reward_std_per_prompt": 0.3013751065495344, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 24.347071409225464, "avg_step_time_sec": 28.14341040757986, "gpu_allocated_gb": 32.36541696, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 731.9018037319183, "eta_sec": 4896.953410918896, "timestamp": "2026-02-15T11:10:06.705103"}
{"step": 27, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.8, "mean_correctness_reward": 0.8, "mean_length_penalty": -0.571904296875, "mean_total_reward": -0.34380859375, "std_total_reward": 0.48124803458468896, "total_tokens": 58563, "mean_completion_length": 1171.26, "median_completion_length": 1060, "min_completion_length": 699, "max_completion_length": 2010, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.17562759088840857, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 27.468075275421143, "avg_step_time_sec": 28.118397995277686, "gpu_allocated_gb": 32.3656192, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 759.376149892807, "eta_sec": 4864.48285318304, "timestamp": "2026-02-15T11:10:32.146391"}
{"step": 28, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.779599609375, "mean_total_reward": -0.95919921875, "std_total_reward": 0.9234610935560011, "total_tokens": 79831, "mean_completion_length": 1596.62, "median_completion_length": 1754, "min_completion_length": 576, "max_completion_length": 2048, "frac_has_think": 0.68, "frac_has_boxed": 0.68, "mean_reward_std_per_prompt": 0.45800198063012965, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 25.380982160568237, "avg_step_time_sec": 28.020633144038065, "gpu_allocated_gb": 32.365585408, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 784.7643468379974, "eta_sec": 4819.548900774547, "timestamp": "2026-02-15T11:11:00.379028"}
{"step": 29, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.96, "mean_correctness_reward": 0.96, "mean_length_penalty": -0.57966796875, "mean_total_reward": -0.1993359375, "std_total_reward": 0.48040045087477135, "total_tokens": 59358, "mean_completion_length": 1187.16, "median_completion_length": 1124, "min_completion_length": 709, "max_completion_length": 2048, "frac_has_think": 0.96, "frac_has_boxed": 0.96, "mean_reward_std_per_prompt": 0.36469899513614046, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 28.67519497871399, "avg_step_time_sec": 28.04320424178551, "gpu_allocated_gb": 32.365682688, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 813.4463613033295, "eta_sec": 4795.387925345322, "timestamp": "2026-02-15T11:11:26.527461"}
{"step": 30, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.48, "mean_correctness_reward": 0.48, "mean_length_penalty": -0.704580078125, "mean_total_reward": -0.92916015625, "std_total_reward": 0.5497447794986866, "total_tokens": 72149, "mean_completion_length": 1442.98, "median_completion_length": 1548, "min_completion_length": 540, "max_completion_length": 2048, "frac_has_think": 0.96, "frac_has_boxed": 0.96, "mean_reward_std_per_prompt": 0.3592857215048405, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 25.727460384368896, "avg_step_time_sec": 27.966012779871622, "gpu_allocated_gb": 32.365612032, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 839.1805837154388, "eta_sec": 4754.222172578176, "timestamp": "2026-02-15T11:11:53.519145"}
{"step": 31, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.92, "mean_correctness_reward": 0.92, "mean_length_penalty": -0.488046875, "mean_total_reward": -0.05609375, "std_total_reward": 0.48621426409342433, "total_tokens": 49976, "mean_completion_length": 999.52, "median_completion_length": 999, "min_completion_length": 503, "max_completion_length": 2048, "frac_has_think": 0.98, "frac_has_boxed": 0.98, "mean_reward_std_per_prompt": 0.2802743560101648, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 27.7020206451416, "avg_step_time_sec": 27.95749690455775, "gpu_allocated_gb": 32.365669376, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 866.890026807785, "eta_sec": 4724.81697687026, "timestamp": "2026-02-15T11:12:18.955928"}
{"step": 32, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.56, "mean_correctness_reward": 0.56, "mean_length_penalty": -0.6526953125, "mean_total_reward": -0.745390625, "std_total_reward": 0.8508148335815969, "total_tokens": 66836, "mean_completion_length": 1336.72, "median_completion_length": 1064, "min_completion_length": 648, "max_completion_length": 2048, "frac_has_think": 0.86, "frac_has_boxed": 0.86, "mean_reward_std_per_prompt": 0.2447382922353724, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 24.719195127487183, "avg_step_time_sec": 27.856299974024296, "gpu_allocated_gb": 32.365612032, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 891.6159009933472, "eta_sec": 4679.858395636082, "timestamp": "2026-02-15T11:12:45.564186"}
{"step": 33, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.8, "mean_correctness_reward": 0.8, "mean_length_penalty": -0.684658203125, "mean_total_reward": -0.56931640625, "std_total_reward": 0.7846134504823739, "total_tokens": 70109, "mean_completion_length": 1402.18, "median_completion_length": 1332, "min_completion_length": 251, "max_completion_length": 2048, "frac_has_think": 0.8, "frac_has_boxed": 0.8, "mean_reward_std_per_prompt": 0.2659131619319467, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 26.834537506103516, "avg_step_time_sec": 27.82533747499639, "gpu_allocated_gb": 32.365663232, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 918.4570384025574, "eta_sec": 4646.831358324397, "timestamp": "2026-02-15T11:13:13.196737"}
{"step": 34, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.82, "mean_correctness_reward": 0.82, "mean_length_penalty": -0.64140625, "mean_total_reward": -0.4628125, "std_total_reward": 0.7817670857142506, "total_tokens": 65680, "mean_completion_length": 1313.6, "median_completion_length": 1135, "min_completion_length": 699, "max_completion_length": 2048, "frac_has_think": 0.88, "frac_has_boxed": 0.88, "mean_reward_std_per_prompt": 0.47884287848453133, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 27.833915948867798, "avg_step_time_sec": 27.825589783051434, "gpu_allocated_gb": 32.365696, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 946.2972931861877, "eta_sec": 4619.047903986538, "timestamp": "2026-02-15T11:13:40.034773"}
{"step": 35, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.82, "mean_correctness_reward": 0.82, "mean_length_penalty": -0.50439453125, "mean_total_reward": -0.1887890625, "std_total_reward": 0.8409575971330429, "total_tokens": 51650, "mean_completion_length": 1033, "median_completion_length": 829, "min_completion_length": 533, "max_completion_length": 2048, "frac_has_think": 0.94, "frac_has_boxed": 0.94, "mean_reward_std_per_prompt": 0.1689859705594126, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 26.522064447402954, "avg_step_time_sec": 27.78834620203291, "gpu_allocated_gb": 32.365632512, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 972.8260927200317, "eta_sec": 4585.07712333543, "timestamp": "2026-02-15T11:14:05.459405"}
{"step": 36, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.58, "mean_correctness_reward": 0.58, "mean_length_penalty": -0.593681640625, "mean_total_reward": -0.60736328125, "std_total_reward": 0.8260797176825192, "total_tokens": 60793, "mean_completion_length": 1215.86, "median_completion_length": 1031, "min_completion_length": 666, "max_completion_length": 2048, "frac_has_think": 0.78, "frac_has_boxed": 0.78, "mean_reward_std_per_prompt": 0.20898626297004477, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 25.473427534103394, "avg_step_time_sec": 27.72404290570153, "gpu_allocated_gb": 32.365651968, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 998.3059763908386, "eta_sec": 4546.743036535051, "timestamp": "2026-02-15T11:14:32.288018"}
{"step": 37, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.84, "mean_correctness_reward": 0.84, "mean_length_penalty": -0.637109375, "mean_total_reward": -0.43421875, "std_total_reward": 0.7498222842425912, "total_tokens": 65240, "mean_completion_length": 1304.8, "median_completion_length": 1124, "min_completion_length": 591, "max_completion_length": 2048, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.4083209604073009, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 27.221322774887085, "avg_step_time_sec": 27.71045587513898, "gpu_allocated_gb": 32.36573184, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1025.5335340499878, "eta_sec": 4516.804307647653, "timestamp": "2026-02-15T11:14:59.288435"}
{"step": 38, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.5539453125, "mean_total_reward": -0.507890625, "std_total_reward": 0.9915326956216908, "total_tokens": 56724, "mean_completion_length": 1134.48, "median_completion_length": 915, "min_completion_length": 504, "max_completion_length": 2048, "frac_has_think": 0.84, "frac_has_boxed": 0.84, "mean_reward_std_per_prompt": 0.08977261661675104, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 26.626873016357422, "avg_step_time_sec": 27.68194053674999, "gpu_allocated_gb": 32.365658112, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1052.1666359901428, "eta_sec": 4484.474366953498, "timestamp": "2026-02-15T11:15:25.181543"}
{"step": 39, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.96, "mean_correctness_reward": 0.96, "mean_length_penalty": -0.42453125, "mean_total_reward": 0.1109375, "std_total_reward": 0.4362129681016362, "total_tokens": 43472, "mean_completion_length": 869.44, "median_completion_length": 785, "min_completion_length": 550, "max_completion_length": 2048, "frac_has_think": 0.98, "frac_has_boxed": 0.98, "mean_reward_std_per_prompt": 0.21709050201295246, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 25.67488694190979, "avg_step_time_sec": 27.63047762406178, "gpu_allocated_gb": 32.365620224, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1077.8467524051666, "eta_sec": 4448.506897473946, "timestamp": "2026-02-15T11:15:49.957971"}
{"step": 40, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.56, "mean_correctness_reward": 0.56, "mean_length_penalty": -0.5495703125, "mean_total_reward": -0.539140625, "std_total_reward": 0.6934129006628476, "total_tokens": 56276, "mean_completion_length": 1125.52, "median_completion_length": 1054, "min_completion_length": 588, "max_completion_length": 2048, "frac_has_think": 0.94, "frac_has_boxed": 0.94, "mean_reward_std_per_prompt": 0.2576102663460176, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 24.768519639968872, "avg_step_time_sec": 27.55892867445946, "gpu_allocated_gb": 32.365620224, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1102.6216073036194, "eta_sec": 4409.428587913513, "timestamp": "2026-02-15T11:16:15.638104"}
{"step": 41, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.86, "mean_correctness_reward": 0.86, "mean_length_penalty": -0.529287109375, "mean_total_reward": -0.19857421875, "std_total_reward": 0.8084075591887672, "total_tokens": 54199, "mean_completion_length": 1083.98, "median_completion_length": 942, "min_completion_length": 386, "max_completion_length": 2048, "frac_has_think": 0.9, "frac_has_boxed": 0.9, "mean_reward_std_per_prompt": 0.3645152716384936, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 26.161901712417603, "avg_step_time_sec": 27.524854846116973, "gpu_allocated_gb": 32.365696, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1128.7897243499756, "eta_sec": 4376.451920532599, "timestamp": "2026-02-15T11:16:41.620927"}
{"step": 42, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.94, "mean_correctness_reward": 0.94, "mean_length_penalty": -0.363544921875, "mean_total_reward": 0.21291015625, "std_total_reward": 0.34055928231530874, "total_tokens": 37227, "mean_completion_length": 744.54, "median_completion_length": 670, "min_completion_length": 410, "max_completion_length": 1572, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.22381226374785856, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 25.671804189682007, "avg_step_time_sec": 27.48073459239233, "gpu_allocated_gb": 32.365651968, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1154.4678857326508, "eta_sec": 4341.9560655979885, "timestamp": "2026-02-15T11:17:02.508250"}
{"step": 43, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.96, "mean_correctness_reward": 0.96, "mean_length_penalty": -0.49421875, "mean_total_reward": -0.0284375, "std_total_reward": 0.5428709317800847, "total_tokens": 50608, "mean_completion_length": 1012.16, "median_completion_length": 841, "min_completion_length": 462, "max_completion_length": 2048, "frac_has_think": 0.98, "frac_has_boxed": 0.98, "mean_reward_std_per_prompt": 0.270534792435057, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 19.199666500091553, "avg_step_time_sec": 27.288151613501615, "gpu_allocated_gb": 32.364941824, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1173.6734459400177, "eta_sec": 4284.239803319753, "timestamp": "2026-02-15T11:17:26.154741"}
{"step": 44, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.74, "mean_correctness_reward": 0.74, "mean_length_penalty": -0.508359375, "mean_total_reward": -0.27671875, "std_total_reward": 0.8865599209005379, "total_tokens": 52056, "mean_completion_length": 1041.12, "median_completion_length": 880, "min_completion_length": 346, "max_completion_length": 2048, "frac_has_think": 0.9, "frac_has_boxed": 0.9, "mean_reward_std_per_prompt": 0.25753765484981667, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 25.27553653717041, "avg_step_time_sec": 27.242410361766815, "gpu_allocated_gb": 32.365645824, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1198.9557802677155, "eta_sec": 4249.816016435623, "timestamp": "2026-02-15T11:17:51.607899"}
{"step": 45, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.8, "mean_correctness_reward": 0.8, "mean_length_penalty": -0.50533203125, "mean_total_reward": -0.2106640625, "std_total_reward": 0.9145444367522191, "total_tokens": 51746, "mean_completion_length": 1034.92, "median_completion_length": 844, "min_completion_length": 367, "max_completion_length": 2048, "frac_has_think": 0.8, "frac_has_boxed": 0.8, "mean_reward_std_per_prompt": 0.09431374428194116, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 25.451072454452515, "avg_step_time_sec": 27.202602852715387, "gpu_allocated_gb": 32.365638656, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1224.413857460022, "eta_sec": 4216.403442170885, "timestamp": "2026-02-15T11:18:17.138943"}
{"step": 46, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.94, "mean_correctness_reward": 0.94, "mean_length_penalty": -0.381923828125, "mean_total_reward": 0.17615234375, "std_total_reward": 0.5290273288092735, "total_tokens": 39109, "mean_completion_length": 782.18, "median_completion_length": 660, "min_completion_length": 482, "max_completion_length": 2048, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.19545914159659414, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 26.046528339385986, "avg_step_time_sec": 27.177470798077792, "gpu_allocated_gb": 32.3656704, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1250.4677429199219, "eta_sec": 4185.33050290398, "timestamp": "2026-02-15T11:18:41.931030"}
{"step": 47, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.74, "mean_correctness_reward": 0.74, "mean_length_penalty": -0.489775390625, "mean_total_reward": -0.23955078125, "std_total_reward": 0.7429196442783124, "total_tokens": 50153, "mean_completion_length": 1003.06, "median_completion_length": 901, "min_completion_length": 340, "max_completion_length": 2048, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.2863318373388506, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 24.259246587753296, "avg_step_time_sec": 27.115380921262375, "gpu_allocated_gb": 32.36563456, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1274.7334275245667, "eta_sec": 4148.653280953144, "timestamp": "2026-02-15T11:19:07.203209"}
{"step": 48, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.8, "mean_correctness_reward": 0.8, "mean_length_penalty": -0.476640625, "mean_total_reward": -0.15328125, "std_total_reward": 0.9481783610084487, "total_tokens": 48808, "mean_completion_length": 976.16, "median_completion_length": 765, "min_completion_length": 415, "max_completion_length": 2048, "frac_has_think": 0.8, "frac_has_boxed": 0.8, "mean_reward_std_per_prompt": 0.12604483681034662, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 25.29506516456604, "avg_step_time_sec": 27.077457676331203, "gpu_allocated_gb": 32.3656448, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1300.0352339744568, "eta_sec": 4115.773566802343, "timestamp": "2026-02-15T11:19:32.395668"}
{"step": 49, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.41802734375, "mean_total_reward": -0.2360546875, "std_total_reward": 1.0065597917863192, "total_tokens": 42806, "mean_completion_length": 856.12, "median_completion_length": 691, "min_completion_length": 257, "max_completion_length": 2048, "frac_has_think": 0.8, "frac_has_boxed": 0.8, "mean_reward_std_per_prompt": 0.0346550618538999, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 25.15297222137451, "avg_step_time_sec": 27.038182462964738, "gpu_allocated_gb": 32.365645824, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1325.1931488513947, "eta_sec": 4082.765551907675, "timestamp": "2026-02-15T11:19:57.028381"}
{"step": 50, "num_prompts": 5, "num_rollouts": 50, "accuracy": 1.0, "mean_correctness_reward": 1.0, "mean_length_penalty": -0.355986328125, "mean_total_reward": 0.28802734375, "std_total_reward": 0.2858138293701061, "total_tokens": 36453, "mean_completion_length": 729.06, "median_completion_length": 678, "min_completion_length": 316, "max_completion_length": 1711, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.10634269211766767, "frac_prompts_all_correct": 1.0, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 24.619140625, "avg_step_time_sec": 26.989801626205445, "gpu_allocated_gb": 32.365627392, "gpu_reserved_gb": 79.956017152, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1349.8192420005798, "eta_sec": 4048.4702439308167, "timestamp": "2026-02-15T11:20:18.614619"}
{"step": 51, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.8, "mean_correctness_reward": 0.8, "mean_length_penalty": -0.335302734375, "mean_total_reward": 0.12939453125, "std_total_reward": 0.47792973738569927, "total_tokens": 34335, "mean_completion_length": 686.7, "median_completion_length": 656, "min_completion_length": 375, "max_completion_length": 1750, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.09512862505574209, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 19.48569941520691, "avg_step_time_sec": 26.84266236716626, "gpu_allocated_gb": 32.36515072, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1369.8655779361725, "eta_sec": 3999.5566927077725, "timestamp": "2026-02-15T11:20:38.209798"}
{"step": 52, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.76, "mean_correctness_reward": 0.76, "mean_length_penalty": -0.386591796875, "mean_total_reward": -0.01318359375, "std_total_reward": 0.7348749609744585, "total_tokens": 39587, "mean_completion_length": 791.74, "median_completion_length": 710, "min_completion_length": 304, "max_completion_length": 2048, "frac_has_think": 0.98, "frac_has_boxed": 0.98, "mean_reward_std_per_prompt": 0.3479188873074197, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 19.83918833732605, "avg_step_time_sec": 26.707980174284714, "gpu_allocated_gb": 32.365205504, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1389.7110018730164, "eta_sec": 3952.7810657941377, "timestamp": "2026-02-15T11:21:00.617585"}
{"step": 53, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.84, "mean_correctness_reward": 0.84, "mean_length_penalty": -0.44390625, "mean_total_reward": -0.0478125, "std_total_reward": 0.7453801336090878, "total_tokens": 45456, "mean_completion_length": 909.12, "median_completion_length": 712, "min_completion_length": 404, "max_completion_length": 2048, "frac_has_think": 0.92, "frac_has_boxed": 0.92, "mean_reward_std_per_prompt": 0.4064747391285517, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 24.23162007331848, "avg_step_time_sec": 26.661256398794787, "gpu_allocated_gb": 32.365645824, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1413.9500031471252, "eta_sec": 3919.204690622834, "timestamp": "2026-02-15T11:21:25.249716"}
{"step": 54, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.92, "mean_correctness_reward": 0.92, "mean_length_penalty": -0.47888671875, "mean_total_reward": -0.0377734375, "std_total_reward": 0.7109534746659928, "total_tokens": 49038, "mean_completion_length": 980.76, "median_completion_length": 742, "min_completion_length": 400, "max_completion_length": 2048, "frac_has_think": 0.98, "frac_has_boxed": 0.98, "mean_reward_std_per_prompt": 0.2541618138754811, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 24.1333966255188, "avg_step_time_sec": 26.614444180771155, "gpu_allocated_gb": 32.365657088, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1438.0896172523499, "eta_sec": 3885.708850392589, "timestamp": "2026-02-15T11:21:49.622818"}
{"step": 55, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.412880859375, "mean_total_reward": -0.22576171875, "std_total_reward": 0.9563949911859898, "total_tokens": 42279, "mean_completion_length": 845.58, "median_completion_length": 549, "min_completion_length": 280, "max_completion_length": 2048, "frac_has_think": 0.8, "frac_has_boxed": 0.8, "mean_reward_std_per_prompt": 0.06531152751539154, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 24.314842224121094, "avg_step_time_sec": 26.572633236104792, "gpu_allocated_gb": 32.365641728, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1462.4118480682373, "eta_sec": 3853.0318192351947, "timestamp": "2026-02-15T11:22:13.484393"}
{"step": 56, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.8, "mean_correctness_reward": 0.8, "mean_length_penalty": -0.488662109375, "mean_total_reward": -0.17732421875, "std_total_reward": 1.0036062083925388, "total_tokens": 50039, "mean_completion_length": 1000.78, "median_completion_length": 554, "min_completion_length": 255, "max_completion_length": 2048, "frac_has_think": 0.8, "frac_has_boxed": 0.8, "mean_reward_std_per_prompt": 0.07491196800003273, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 23.85719060897827, "avg_step_time_sec": 26.52414318919182, "gpu_allocated_gb": 32.365648896, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1486.2750549316406, "eta_sec": 3819.476619243622, "timestamp": "2026-02-15T11:22:38.760529"}
{"step": 57, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.74, "mean_correctness_reward": 0.74, "mean_length_penalty": -0.45818359375, "mean_total_reward": -0.1763671875, "std_total_reward": 0.7720105009797774, "total_tokens": 46918, "mean_completion_length": 938.36, "median_completion_length": 1005, "min_completion_length": 367, "max_completion_length": 2048, "frac_has_think": 0.96, "frac_has_boxed": 0.96, "mean_reward_std_per_prompt": 0.3364112545029401, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 25.28809928894043, "avg_step_time_sec": 26.502458208485653, "gpu_allocated_gb": 32.365647872, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1511.570821762085, "eta_sec": 3789.8515238134482, "timestamp": "2026-02-15T11:23:03.404462"}
{"step": 58, "num_prompts": 5, "num_rollouts": 50, "accuracy": 1.0, "mean_correctness_reward": 1.0, "mean_length_penalty": -0.288291015625, "mean_total_reward": 0.42341796875, "std_total_reward": 0.23064653911131708, "total_tokens": 29521, "mean_completion_length": 590.42, "median_completion_length": 575, "min_completion_length": 321, "max_completion_length": 1366, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.09695037666192002, "frac_prompts_all_correct": 1.0, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 25.09900975227356, "avg_step_time_sec": 26.478260821309583, "gpu_allocated_gb": 32.36573696, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1536.6776587963104, "eta_sec": 3759.9130366259606, "timestamp": "2026-02-15T11:23:22.468412"}
{"step": 59, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.34, "mean_correctness_reward": 0.34, "mean_length_penalty": -0.316279296875, "mean_total_reward": -0.29255859375, "std_total_reward": 0.5132161450487468, "total_tokens": 32387, "mean_completion_length": 647.74, "median_completion_length": 722, "min_completion_length": 273, "max_completion_length": 966, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.19423575967630702, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.6, "step_time_sec": 15.729346990585327, "avg_step_time_sec": 26.296075841127816, "gpu_allocated_gb": 32.36467712, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1552.4137892723083, "eta_sec": 3707.746693599022, "timestamp": "2026-02-15T11:23:36.120789"}
{"step": 60, "num_prompts": 5, "num_rollouts": 50, "accuracy": 1.0, "mean_correctness_reward": 1.0, "mean_length_penalty": -0.2498046875, "mean_total_reward": 0.500390625, "std_total_reward": 0.09785448629613898, "total_tokens": 25580, "mean_completion_length": 511.6, "median_completion_length": 490, "min_completion_length": 359, "max_completion_length": 792, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.06622049079580897, "frac_prompts_all_correct": 1.0, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 12.3740873336792, "avg_step_time_sec": 26.064042699337005, "gpu_allocated_gb": 32.364137472, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1564.7949426174164, "eta_sec": 3648.965977907181, "timestamp": "2026-02-15T11:23:47.356509"}
{"step": 61, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.84, "mean_correctness_reward": 0.84, "mean_length_penalty": -0.36912109375, "mean_total_reward": 0.1017578125, "std_total_reward": 0.5528591400707468, "total_tokens": 37798, "mean_completion_length": 755.96, "median_completion_length": 770, "min_completion_length": 234, "max_completion_length": 1501, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.3324185311638072, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 10.333732843399048, "avg_step_time_sec": 25.80616876727245, "gpu_allocated_gb": 32.36386304, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1575.13796043396, "eta_sec": 3587.0574586508706, "timestamp": "2026-02-15T11:24:02.564207"}
{"step": 62, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.8, "mean_correctness_reward": 0.8, "mean_length_penalty": -0.28484375, "mean_total_reward": 0.2303125, "std_total_reward": 0.6900319409924759, "total_tokens": 29168, "mean_completion_length": 583.36, "median_completion_length": 530, "min_completion_length": 204, "max_completion_length": 1402, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.07940076857042026, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 18.026524305343628, "avg_step_time_sec": 25.680690630789726, "gpu_allocated_gb": 32.36489472, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1593.1700766086578, "eta_sec": 3543.9353070489824, "timestamp": "2026-02-15T11:24:19.329466"}
{"step": 63, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.96, "mean_correctness_reward": 0.96, "mean_length_penalty": -0.27767578125, "mean_total_reward": 0.4046484375, "std_total_reward": 0.2427562908767843, "total_tokens": 28434, "mean_completion_length": 568.68, "median_completion_length": 548, "min_completion_length": 342, "max_completion_length": 904, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.1519725220452184, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 16.34808039665222, "avg_step_time_sec": 25.53255396040659, "gpu_allocated_gb": 32.364749824, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1609.5262928009033, "eta_sec": 3497.959892575703, "timestamp": "2026-02-15T11:24:32.400117"}
{"step": 64, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.96, "mean_correctness_reward": 0.96, "mean_length_penalty": -0.308876953125, "mean_total_reward": 0.34224609375, "std_total_reward": 0.38586015848349414, "total_tokens": 31629, "mean_completion_length": 632.58, "median_completion_length": 582, "min_completion_length": 228, "max_completion_length": 1477, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.1670576771502588, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 11.13660192489624, "avg_step_time_sec": 25.307617209851742, "gpu_allocated_gb": 32.364044288, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1620.6702580451965, "eta_sec": 3441.835940539837, "timestamp": "2026-02-15T11:24:47.859057"}
{"step": 65, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.76, "mean_correctness_reward": 0.76, "mean_length_penalty": -0.319404296875, "mean_total_reward": 0.12119140625, "std_total_reward": 0.9639851840530811, "total_tokens": 32707, "mean_completion_length": 654.14, "median_completion_length": 439, "min_completion_length": 214, "max_completion_length": 2048, "frac_has_think": 0.92, "frac_has_boxed": 0.92, "mean_reward_std_per_prompt": 0.19882959388532015, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 17.694401264190674, "avg_step_time_sec": 25.190490810687724, "gpu_allocated_gb": 32.364860928, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1638.3719172477722, "eta_sec": 3400.716259442843, "timestamp": "2026-02-15T11:25:08.790229"}
{"step": 66, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.258818359375, "mean_total_reward": 0.08236328125, "std_total_reward": 0.5784743655764085, "total_tokens": 26503, "mean_completion_length": 530.06, "median_completion_length": 561, "min_completion_length": 172, "max_completion_length": 894, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.06530429067526211, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 23.246143341064453, "avg_step_time_sec": 25.161031000541918, "gpu_allocated_gb": 32.36565504, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1661.6252076625824, "eta_sec": 3371.578154072617, "timestamp": "2026-02-15T11:25:24.318257"}
{"step": 67, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.76, "mean_correctness_reward": 0.76, "mean_length_penalty": -0.3073046875, "mean_total_reward": 0.145390625, "std_total_reward": 0.711888064297237, "total_tokens": 31468, "mean_completion_length": 629.36, "median_completion_length": 480, "min_completion_length": 312, "max_completion_length": 1754, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.20111468014872938, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 10.894620418548584, "avg_step_time_sec": 24.948099499318136, "gpu_allocated_gb": 32.364022784, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1672.5269486904144, "eta_sec": 3318.097233409312, "timestamp": "2026-02-15T11:25:40.654463"}
{"step": 68, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.241416015625, "mean_total_reward": 0.11716796875, "std_total_reward": 0.7645556270003865, "total_tokens": 24721, "mean_completion_length": 494.42, "median_completion_length": 339, "min_completion_length": 242, "max_completion_length": 2048, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.11191386703799691, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 19.67180848121643, "avg_step_time_sec": 24.870506984346054, "gpu_allocated_gb": 32.365238272, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1692.2053356170654, "eta_sec": 3282.9069219336793, "timestamp": "2026-02-15T11:26:01.515364"}
{"step": 69, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.24, "mean_correctness_reward": 0.24, "mean_length_penalty": -0.331787109375, "mean_total_reward": -0.42357421875, "std_total_reward": 0.8136755508142037, "total_tokens": 33975, "mean_completion_length": 679.5, "median_completion_length": 445, "min_completion_length": 172, "max_completion_length": 2048, "frac_has_think": 0.9, "frac_has_boxed": 0.9, "mean_reward_std_per_prompt": 0.29708909640000974, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.6, "step_time_sec": 22.599355936050415, "avg_step_time_sec": 24.83759175176206, "gpu_allocated_gb": 32.365641728, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1714.8118438720703, "eta_sec": 3253.7245194808297, "timestamp": "2026-02-15T11:26:25.295895"}
{"step": 70, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.1506640625, "mean_total_reward": 0.298671875, "std_total_reward": 0.5531642183527603, "total_tokens": 15428, "mean_completion_length": 308.56, "median_completion_length": 271, "min_completion_length": 196, "max_completion_length": 1391, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.0865680567364201, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 23.301621913909912, "avg_step_time_sec": 24.815649325507028, "gpu_allocated_gb": 32.36564992, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1738.1190192699432, "eta_sec": 3226.034412315914, "timestamp": "2026-02-15T11:26:42.415565"}
{"step": 71, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.8, "mean_correctness_reward": 0.8, "mean_length_penalty": -0.166552734375, "mean_total_reward": 0.46689453125, "std_total_reward": 0.5116839820756728, "total_tokens": 17055, "mean_completion_length": 341.1, "median_completion_length": 302, "min_completion_length": 132, "max_completion_length": 649, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.03751466799565311, "frac_prompts_all_correct": 0.8, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 14.276378154754639, "avg_step_time_sec": 24.667208886482346, "gpu_allocated_gb": 32.36470272, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1752.4019222259521, "eta_sec": 3182.0699463562228, "timestamp": "2026-02-15T11:26:52.917549"}
{"step": 72, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.62, "mean_correctness_reward": 0.62, "mean_length_penalty": -0.26359375, "mean_total_reward": 0.0928125, "std_total_reward": 0.7906476999273384, "total_tokens": 26992, "mean_completion_length": 539.84, "median_completion_length": 359, "min_completion_length": 184, "max_completion_length": 2048, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.13078207469576794, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 7.8998863697052, "avg_step_time_sec": 24.434329407082664, "gpu_allocated_gb": 32.363707392, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1760.308081150055, "eta_sec": 3127.594164106581, "timestamp": "2026-02-15T11:27:09.792220"}
{"step": 73, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.34, "mean_correctness_reward": 0.34, "mean_length_penalty": -0.164384765625, "mean_total_reward": 0.01123046875, "std_total_reward": 0.5055303379844583, "total_tokens": 16833, "mean_completion_length": 336.66, "median_completion_length": 320, "min_completion_length": 208, "max_completion_length": 554, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.11303364279873122, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.6, "step_time_sec": 22.299489498138428, "avg_step_time_sec": 24.405085024768358, "gpu_allocated_gb": 32.36562944, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1782.6136484146118, "eta_sec": 3099.4457981455816, "timestamp": "2026-02-15T11:27:22.581480"}
{"step": 74, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.92, "mean_correctness_reward": 0.92, "mean_length_penalty": -0.137353515625, "mean_total_reward": 0.64529296875, "std_total_reward": 0.39171577638122057, "total_tokens": 14065, "mean_completion_length": 281.3, "median_completion_length": 208, "min_completion_length": 78, "max_completion_length": 830, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.16474692948329636, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 7.045936346054077, "avg_step_time_sec": 24.170501934515464, "gpu_allocated_gb": 32.363567616, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1789.6658253669739, "eta_sec": 3045.4832437489486, "timestamp": "2026-02-15T11:27:31.534327"}
{"step": 75, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.72, "mean_correctness_reward": 0.72, "mean_length_penalty": -0.185966796875, "mean_total_reward": 0.34806640625, "std_total_reward": 0.7741458022225237, "total_tokens": 19043, "mean_completion_length": 380.86, "median_completion_length": 223, "min_completion_length": 96, "max_completion_length": 1694, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.2650207243689744, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 9.830961227416992, "avg_step_time_sec": 23.979308058420816, "gpu_allocated_gb": 32.363924992, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1799.5025925636292, "eta_sec": 2997.413507302602, "timestamp": "2026-02-15T11:27:46.457204"}
{"step": 76, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.84, "mean_correctness_reward": 0.84, "mean_length_penalty": -0.110087890625, "mean_total_reward": 0.61982421875, "std_total_reward": 0.4213551235620347, "total_tokens": 11273, "mean_completion_length": 225.46, "median_completion_length": 197, "min_completion_length": 73, "max_completion_length": 660, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.2268039612186037, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 18.40089726448059, "avg_step_time_sec": 23.90590791639529, "gpu_allocated_gb": 32.365162496, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1817.909282207489, "eta_sec": 2964.3325816330157, "timestamp": "2026-02-15T11:27:58.022308"}
{"step": 77, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.11234375, "mean_total_reward": 0.3753125, "std_total_reward": 0.5472985399519134, "total_tokens": 11504, "mean_completion_length": 230.08, "median_completion_length": 213, "min_completion_length": 43, "max_completion_length": 701, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.05456207788510585, "frac_prompts_all_correct": 0.6, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 7.436468601226807, "avg_step_time_sec": 23.692019094120372, "gpu_allocated_gb": 32.363681792, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1825.35209274292, "eta_sec": 2914.1183485768056, "timestamp": "2026-02-15T11:28:05.675060"}
{"step": 78, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.7, "mean_correctness_reward": 0.7, "mean_length_penalty": -0.11037109375, "mean_total_reward": 0.4792578125, "std_total_reward": 0.45647955817991764, "total_tokens": 11302, "mean_completion_length": 226.04, "median_completion_length": 223, "min_completion_length": 121, "max_completion_length": 368, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.19342157998172432, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 7.794365167617798, "avg_step_time_sec": 23.48820301813957, "gpu_allocated_gb": 32.36374016, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1833.1536526679993, "eta_sec": 2865.560768213027, "timestamp": "2026-02-15T11:28:11.662016"}
{"step": 79, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.4, "mean_correctness_reward": 0.4, "mean_length_penalty": -0.07828125, "mean_total_reward": 0.2434375, "std_total_reward": 0.5101898244000822, "total_tokens": 8016, "mean_completion_length": 160.32, "median_completion_length": 163, "min_completion_length": 58, "max_completion_length": 368, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.03516290488770255, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.6, "step_time_sec": 4.953397274017334, "avg_step_time_sec": 23.253585223910175, "gpu_allocated_gb": 32.363296768, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1838.1129553318024, "eta_sec": 2813.683812093131, "timestamp": "2026-02-15T11:28:16.298029"}
{"step": 80, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.56, "mean_correctness_reward": 0.56, "mean_length_penalty": -0.07669921875, "mean_total_reward": 0.4066015625, "std_total_reward": 0.4776664883031377, "total_tokens": 7854, "mean_completion_length": 157.08, "median_completion_length": 148, "min_completion_length": 34, "max_completion_length": 292, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.29277164346154205, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 4.614161252975464, "avg_step_time_sec": 23.02059242427349, "gpu_allocated_gb": 32.36329984, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1842.7344019412994, "eta_sec": 2762.471090912819, "timestamp": "2026-02-15T11:28:20.614806"}
{"step": 81, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.6, "mean_correctness_reward": 0.6, "mean_length_penalty": -0.047568359375, "mean_total_reward": 0.50486328125, "std_total_reward": 0.5091367757527934, "total_tokens": 4871, "mean_completion_length": 97.42, "median_completion_length": 84, "min_completion_length": 40, "max_completion_length": 195, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.25032856022467337, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 4.32066798210144, "avg_step_time_sec": 22.789729159555318, "gpu_allocated_gb": 32.363279872, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1847.0615820884705, "eta_sec": 2711.977769987083, "timestamp": "2026-02-15T11:28:24.035479"}
{"step": 82, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.4, "mean_correctness_reward": 0.4, "mean_length_penalty": -0.029697265625, "mean_total_reward": 0.34060546875, "std_total_reward": 0.4942896842403361, "total_tokens": 3041, "mean_completion_length": 60.82, "median_completion_length": 52, "min_completion_length": 25, "max_completion_length": 146, "frac_has_think": 1.0, "frac_has_boxed": 1.0, "mean_reward_std_per_prompt": 0.2026056168391301, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 2.820793867111206, "avg_step_time_sec": 22.54620555842795, "gpu_allocated_gb": 32.363044352, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1849.8879487514496, "eta_sec": 2660.452255894498, "timestamp": "2026-02-15T11:28:26.424204"}
{"step": 83, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.5, "mean_correctness_reward": 0.5, "mean_length_penalty": -0.050068359375, "mean_total_reward": 0.39986328125, "std_total_reward": 0.544066771743017, "total_tokens": 5127, "mean_completion_length": 102.54, "median_completion_length": 56, "min_completion_length": 26, "max_completion_length": 434, "frac_has_think": 1.0, "frac_has_boxed": 0.98, "mean_reward_std_per_prompt": 0.16198941389617177, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 2.8515894412994385, "avg_step_time_sec": 22.308921026896282, "gpu_allocated_gb": 32.362970112, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1852.7451238632202, "eta_sec": 2610.143760146865, "timestamp": "2026-02-15T11:28:31.135447"}
{"step": 84, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.4, "mean_correctness_reward": 0.4, "mean_length_penalty": -0.039248046875, "mean_total_reward": 0.32150390625, "std_total_reward": 0.5207132821033217, "total_tokens": 4019, "mean_completion_length": 80.38, "median_completion_length": 55, "min_completion_length": 29, "max_completion_length": 402, "frac_has_think": 1.0, "frac_has_boxed": 0.9, "mean_reward_std_per_prompt": 0.03792614492303969, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.6, "step_time_sec": 5.153187990188599, "avg_step_time_sec": 22.104686109792617, "gpu_allocated_gb": 32.36340224, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1857.9044926166534, "eta_sec": 2564.1435887359435, "timestamp": "2026-02-15T11:28:35.927304"}
{"step": 85, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.84, "mean_correctness_reward": 0.84, "mean_length_penalty": -0.02337890625, "mean_total_reward": 0.7932421875, "std_total_reward": 0.3657370392073307, "total_tokens": 2394, "mean_completion_length": 47.88, "median_completion_length": 44, "min_completion_length": 19, "max_completion_length": 81, "frac_has_think": 1.0, "frac_has_boxed": 0.92, "mean_reward_std_per_prompt": 0.24093240374136857, "frac_prompts_all_correct": 0.4, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 4.412422180175781, "avg_step_time_sec": 21.896541828267715, "gpu_allocated_gb": 32.363305984, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1862.3236989974976, "eta_sec": 2518.102310250787, "timestamp": "2026-02-15T11:28:38.442071"}
{"step": 86, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.34, "mean_correctness_reward": 0.34, "mean_length_penalty": -0.027060546875, "mean_total_reward": 0.28587890625, "std_total_reward": 0.48426511677320333, "total_tokens": 2771, "mean_completion_length": 55.42, "median_completion_length": 36, "min_completion_length": 10, "max_completion_length": 202, "frac_has_think": 1.0, "frac_has_boxed": 0.78, "mean_reward_std_per_prompt": 0.259914895309582, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 1.8945462703704834, "avg_step_time_sec": 21.663960484571234, "gpu_allocated_gb": 32.362860544, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1864.223631620407, "eta_sec": 2469.6914952411207, "timestamp": "2026-02-15T11:28:41.057351"}
{"step": 87, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.46, "mean_correctness_reward": 0.46, "mean_length_penalty": -0.018212890625, "mean_total_reward": 0.42357421875, "std_total_reward": 0.4998673381289774, "total_tokens": 1865, "mean_completion_length": 37.3, "median_completion_length": 31, "min_completion_length": 15, "max_completion_length": 119, "frac_has_think": 1.0, "frac_has_boxed": 0.4, "mean_reward_std_per_prompt": 0.3037914526611083, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.0, "step_time_sec": 2.8104519844055176, "avg_step_time_sec": 21.447253490316456, "gpu_allocated_gb": 32.363035136, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1867.0395166873932, "eta_sec": 2423.5396444057596, "timestamp": "2026-02-15T11:28:43.357309"}
{"step": 88, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.04, "mean_correctness_reward": 0.04, "mean_length_penalty": -0.0178125, "mean_total_reward": 0.004375, "std_total_reward": 0.19961408002649567, "total_tokens": 1824, "mean_completion_length": 36.48, "median_completion_length": 29, "min_completion_length": 16, "max_completion_length": 113, "frac_has_think": 1.0, "frac_has_boxed": 0.32, "mean_reward_std_per_prompt": 0.09335958050006506, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.8, "step_time_sec": 2.120818614959717, "avg_step_time_sec": 21.2276349121874, "gpu_allocated_gb": 32.36295424, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1869.1656444072723, "eta_sec": 2377.495110164989, "timestamp": "2026-02-15T11:28:45.461028"}
{"step": 89, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.18, "mean_correctness_reward": 0.18, "mean_length_penalty": -0.021015625, "mean_total_reward": 0.13796875, "std_total_reward": 0.3743774352846242, "total_tokens": 2152, "mean_completion_length": 43.04, "median_completion_length": 23, "min_completion_length": 11, "max_completion_length": 161, "frac_has_think": 1.0, "frac_has_boxed": 0.32, "mean_reward_std_per_prompt": 0.2587201232422357, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 2.080781936645508, "avg_step_time_sec": 21.012501732686932, "gpu_allocated_gb": 32.36296448, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1871.25181722641, "eta_sec": 2332.3876923282496, "timestamp": "2026-02-15T11:28:47.814466"}
{"step": 90, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.28, "mean_correctness_reward": 0.28, "mean_length_penalty": -0.008310546875, "mean_total_reward": 0.26337890625, "std_total_reward": 0.44722166686619197, "total_tokens": 851, "mean_completion_length": 17.02, "median_completion_length": 14, "min_completion_length": 8, "max_completion_length": 48, "frac_has_think": 1.0, "frac_has_boxed": 0.12, "mean_reward_std_per_prompt": 0.14988312547890983, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 2.412895441055298, "avg_step_time_sec": 20.80583944055769, "gpu_allocated_gb": 32.362996224, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1873.6702334880829, "eta_sec": 2288.642338461346, "timestamp": "2026-02-15T11:28:49.694328"}
{"step": 91, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.28, "mean_correctness_reward": 0.28, "mean_length_penalty": -0.009912109375, "mean_total_reward": 0.26017578125, "std_total_reward": 0.4394135437903919, "total_tokens": 1015, "mean_completion_length": 20.3, "median_completion_length": 13, "min_completion_length": 10, "max_completion_length": 53, "frac_has_think": 1.0, "frac_has_boxed": 0.02, "mean_reward_std_per_prompt": 0.24759278780487246, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 1.8584916591644287, "avg_step_time_sec": 20.597626827575347, "gpu_allocated_gb": 32.362992128, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1875.5365719795227, "eta_sec": 2245.141324205713, "timestamp": "2026-02-15T11:28:51.410458"}
{"step": 92, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.28, "mean_correctness_reward": 0.28, "mean_length_penalty": -0.010302734375, "mean_total_reward": 0.25939453125, "std_total_reward": 0.45993285850135085, "total_tokens": 1055, "mean_completion_length": 21.1, "median_completion_length": 10, "min_completion_length": 8, "max_completion_length": 479, "frac_has_think": 1.0, "frac_has_boxed": 0.02, "mean_reward_std_per_prompt": 0.18739697219353232, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.4, "step_time_sec": 1.6376125812530518, "avg_step_time_sec": 20.39153971620228, "gpu_allocated_gb": 32.362834432, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1877.1796307563782, "eta_sec": 2202.2862893498464, "timestamp": "2026-02-15T11:28:55.488823"}
{"step": 93, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.22, "mean_correctness_reward": 0.22, "mean_length_penalty": -0.018544921875, "mean_total_reward": 0.18291015625, "std_total_reward": 0.4295628925323358, "total_tokens": 1899, "mean_completion_length": 37.98, "median_completion_length": 19, "min_completion_length": 8, "max_completion_length": 642, "frac_has_think": 1.0, "frac_has_boxed": 0.16, "mean_reward_std_per_prompt": 0.33044081851348056, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.2, "step_time_sec": 5.527684926986694, "avg_step_time_sec": 20.231713320619317, "gpu_allocated_gb": 32.36358144, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1882.7121787071228, "eta_sec": 2164.793325306267, "timestamp": "2026-02-15T11:29:02.048838"}
{"step": 94, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.04, "mean_correctness_reward": 0.04, "mean_length_penalty": -0.0049609375, "mean_total_reward": 0.030078125, "std_total_reward": 0.19600190827355482, "total_tokens": 508, "mean_completion_length": 10.16, "median_completion_length": 10, "min_completion_length": 8, "max_completion_length": 14, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.08019136638615494, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.8, "step_time_sec": 7.784182071685791, "avg_step_time_sec": 20.099292775417897, "gpu_allocated_gb": 32.363854336, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1890.501636981964, "eta_sec": 2130.525034194297, "timestamp": "2026-02-15T11:29:06.152281"}
{"step": 95, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.36, "mean_correctness_reward": 0.36, "mean_length_penalty": -0.007216796875, "mean_total_reward": 0.34556640625, "std_total_reward": 0.4812457199897188, "total_tokens": 739, "mean_completion_length": 14.78, "median_completion_length": 14, "min_completion_length": 10, "max_completion_length": 26, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.07996881102026236, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.6, "step_time_sec": 1.403536081314087, "avg_step_time_sec": 19.902495336532592, "gpu_allocated_gb": 32.362774528, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1891.9101650714874, "eta_sec": 2089.7620103359222, "timestamp": "2026-02-15T11:29:07.654527"}
{"step": 96, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.16, "mean_correctness_reward": 0.16, "mean_length_penalty": -0.005263671875, "mean_total_reward": 0.14947265625, "std_total_reward": 0.365240245294569, "total_tokens": 539, "mean_completion_length": 10.78, "median_completion_length": 10, "min_completion_length": 8, "max_completion_length": 23, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.08093846136472166, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.8, "step_time_sec": 1.5070323944091797, "avg_step_time_sec": 19.710875930885475, "gpu_allocated_gb": 32.362824192, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1893.4221022129059, "eta_sec": 2049.9310968120894, "timestamp": "2026-02-15T11:29:09.124910"}
{"step": 97, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.2, "mean_correctness_reward": 0.2, "mean_length_penalty": -0.0054296875, "mean_total_reward": 0.189140625, "std_total_reward": 0.39957431267244425, "total_tokens": 556, "mean_completion_length": 11.12, "median_completion_length": 12, "min_completion_length": 8, "max_completion_length": 14, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0002961943630842125, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.8, "step_time_sec": 1.4564878940582275, "avg_step_time_sec": 19.522686363495502, "gpu_allocated_gb": 32.362808832, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1894.884017944336, "eta_sec": 2010.8366954400367, "timestamp": "2026-02-15T11:29:10.553183"}
{"step": 98, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.38, "mean_correctness_reward": 0.38, "mean_length_penalty": -0.006513671875, "mean_total_reward": 0.36697265625, "std_total_reward": 0.48603676678078556, "total_tokens": 667, "mean_completion_length": 13.34, "median_completion_length": 12, "min_completion_length": 8, "max_completion_length": 31, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.06097508261564494, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.6, "step_time_sec": 1.421922206878662, "avg_step_time_sec": 19.337984688427984, "gpu_allocated_gb": 32.362779648, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1896.3127255439758, "eta_sec": 1972.4744382196543, "timestamp": "2026-02-15T11:29:12.072338"}
{"step": 99, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.2, "mean_correctness_reward": 0.2, "mean_length_penalty": -0.0051171875, "mean_total_reward": 0.189765625, "std_total_reward": 0.4002372724616963, "total_tokens": 524, "mean_completion_length": 10.48, "median_completion_length": 10, "min_completion_length": 8, "max_completion_length": 14, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.000234375, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.8, "step_time_sec": 1.5367252826690674, "avg_step_time_sec": 19.15817398735971, "gpu_allocated_gb": 32.362820096, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1897.8566048145294, "eta_sec": 1934.9755727233307, "timestamp": "2026-02-15T11:29:13.503009"}
{"step": 100, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.16, "mean_correctness_reward": 0.16, "mean_length_penalty": -0.00525390625, "mean_total_reward": 0.1494921875, "std_total_reward": 0.36437465870548447, "total_tokens": 538, "mean_completion_length": 10.76, "median_completion_length": 10, "min_completion_length": 8, "max_completion_length": 16, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.07955400686308421, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.8, "step_time_sec": 1.404435396194458, "avg_step_time_sec": 18.980636601448058, "gpu_allocated_gb": 32.362769408, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1899.2669832706451, "eta_sec": 1898.063660144806, "timestamp": "2026-02-15T11:29:15.467209"}
{"step": 101, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.2, "mean_correctness_reward": 0.2, "mean_length_penalty": -0.00609375, "mean_total_reward": 0.1878125, "std_total_reward": 0.3982868434532689, "total_tokens": 624, "mean_completion_length": 12.48, "median_completion_length": 14, "min_completion_length": 8, "max_completion_length": 16, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.00038273277230987154, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.8, "step_time_sec": 1.4188649654388428, "avg_step_time_sec": 18.80675767435886, "gpu_allocated_gb": 32.362825216, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1901.2254354953766, "eta_sec": 1861.8690097615272, "timestamp": "2026-02-15T11:29:16.895452"}
{"step": 102, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00541015625, "mean_total_reward": -0.0108203125, "std_total_reward": 0.0017989967093233522, "total_tokens": 554, "mean_completion_length": 11.08, "median_completion_length": 10, "min_completion_length": 8, "max_completion_length": 16, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0006310740008360746, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.418421983718872, "avg_step_time_sec": 18.63628379503886, "gpu_allocated_gb": 32.362816, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1902.6495316028595, "eta_sec": 1826.3558119138083, "timestamp": "2026-02-15T11:29:18.419403"}
{"step": 103, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.005029296875, "mean_total_reward": -0.01005859375, "std_total_reward": 0.0027010384152225254, "total_tokens": 515, "mean_completion_length": 10.3, "median_completion_length": 9, "min_completion_length": 8, "max_completion_length": 19, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0011629590581398595, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.5332667827606201, "avg_step_time_sec": 18.470235086181788, "gpu_allocated_gb": 32.362890752, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1904.1931371688843, "eta_sec": 1791.6128033596333, "timestamp": "2026-02-15T11:29:19.875993"}
{"step": 104, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00560546875, "mean_total_reward": -0.0112109375, "std_total_reward": 0.00278441927924949, "total_tokens": 574, "mean_completion_length": 11.48, "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": 0.001038903044619743, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.441885232925415, "avg_step_time_sec": 18.3065009529774, "gpu_allocated_gb": 32.362816512, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1905.6406700611115, "eta_sec": 1757.4240914858306, "timestamp": "2026-02-15T11:29:21.302077"}
{"step": 105, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.06, "mean_correctness_reward": 0.06, "mean_length_penalty": -0.00609375, "mean_total_reward": 0.0478125, "std_total_reward": 0.2365481091333649, "total_tokens": 624, "mean_completion_length": 12.48, "median_completion_length": 12, "min_completion_length": 10, "max_completion_length": 22, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.13988034183012563, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.6, "step_time_sec": 1.38527512550354, "avg_step_time_sec": 18.145346421287172, "gpu_allocated_gb": 32.362781184, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1907.0330593585968, "eta_sec": 1723.8079100222812, "timestamp": "2026-02-15T11:29:22.754681"}
{"step": 106, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00458984375, "mean_total_reward": -0.0091796875, "std_total_reward": 0.002734375, "total_tokens": 470, "mean_completion_length": 9.4, "median_completion_length": 8, "min_completion_length": 8, "max_completion_length": 15, "frac_has_think": 0.98, "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": 1.4746975898742676, "avg_step_time_sec": 17.988076149292713, "gpu_allocated_gb": 32.362796544, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1908.5151987075806, "eta_sec": 1690.879158033515, "timestamp": "2026-02-15T11:29:24.224540"}
{"step": 107, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.005546875, "mean_total_reward": -0.01109375, "std_total_reward": 0.0034541058420333257, "total_tokens": 568, "mean_completion_length": 11.36, "median_completion_length": 10, "min_completion_length": 8, "max_completion_length": 19, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0006746130830520426, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.4547982215881348, "avg_step_time_sec": 17.83355953314594, "gpu_allocated_gb": 32.362866688, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1909.976764202118, "eta_sec": 1658.5210365825726, "timestamp": "2026-02-15T11:29:25.667617"}
{"step": 108, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.02, "mean_correctness_reward": 0.02, "mean_length_penalty": -0.0046875, "mean_total_reward": 0.010625, "std_total_reward": 0.13968280280832937, "total_tokens": 480, "mean_completion_length": 9.6, "median_completion_length": 8, "min_completion_length": 8, "max_completion_length": 14, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.06, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.8, "step_time_sec": 1.9514427185058594, "avg_step_time_sec": 17.686502895973348, "gpu_allocated_gb": 32.362815488, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1911.9340341091156, "eta_sec": 1627.158266429548, "timestamp": "2026-02-15T11:29:27.588326"}
{"step": 109, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.004296875, "mean_total_reward": -0.00859375, "std_total_reward": 0.0015625, "total_tokens": 440, "mean_completion_length": 8.8, "median_completion_length": 8, "min_completion_length": 8, "max_completion_length": 12, "frac_has_think": 1.0, "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": 1.3943426609039307, "avg_step_time_sec": 17.53703353601858, "gpu_allocated_gb": 32.36277248, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1913.3353638648987, "eta_sec": 1595.8700517776908, "timestamp": "2026-02-15T11:29:28.961935"}
{"step": 110, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00390625, "mean_total_reward": -0.0078125, "std_total_reward": 0.0, "total_tokens": 400, "mean_completion_length": 8, "median_completion_length": 8, "min_completion_length": 8, "max_completion_length": 8, "frac_has_think": 1.0, "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": 1.3761911392211914, "avg_step_time_sec": 17.39011678695679, "gpu_allocated_gb": 32.362778112, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1914.7170028686523, "eta_sec": 1565.110510826111, "timestamp": "2026-02-15T11:29:30.329388"}
{"step": 111, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.2, "mean_correctness_reward": 0.2, "mean_length_penalty": -0.0046875, "mean_total_reward": 0.190625, "std_total_reward": 0.396875, "total_tokens": 480, "mean_completion_length": 9.6, "median_completion_length": 8, "min_completion_length": 8, "max_completion_length": 16, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.8, "step_time_sec": 1.3617980480194092, "avg_step_time_sec": 17.245717519038433, "gpu_allocated_gb": 32.362791936, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1916.0842852592468, "eta_sec": 1534.8688591944206, "timestamp": "2026-02-15T11:29:31.782830"}
{"step": 112, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00572265625, "mean_total_reward": -0.0114453125, "std_total_reward": 0.012699878986412873, "total_tokens": 586, "mean_completion_length": 11.72, "median_completion_length": 8, "min_completion_length": 8, "max_completion_length": 81, "frac_has_think": 1.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.004658274779978233, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.4339251518249512, "avg_step_time_sec": 17.104540801474027, "gpu_allocated_gb": 32.362802688, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1917.5243735313416, "eta_sec": 1505.1995905297144, "timestamp": "2026-02-15T11:29:33.562928"}
{"step": 113, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00384765625, "mean_total_reward": -0.0076953125, "std_total_reward": 0.0003173452501810922, "total_tokens": 394, "mean_completion_length": 7.88, "median_completion_length": 8, "min_completion_length": 7, "max_completion_length": 8, "frac_has_think": 0.88, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 9.568319307746788e-05, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.8133935928344727, "avg_step_time_sec": 16.969220914671908, "gpu_allocated_gb": 32.36291072, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1919.3445463180542, "eta_sec": 1476.322219576456, "timestamp": "2026-02-15T11:29:34.967910"}
{"step": 114, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.0036328125, "mean_total_reward": -0.007265625, "std_total_reward": 0.0005226206312601427, "total_tokens": 372, "mean_completion_length": 7.44, "median_completion_length": 7, "min_completion_length": 7, "max_completion_length": 9, "frac_has_think": 0.4, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.00018365477026236032, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.358365535736084, "avg_step_time_sec": 16.832283586786506, "gpu_allocated_gb": 32.362780672, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1920.7105588912964, "eta_sec": 1447.5763884636394, "timestamp": "2026-02-15T11:29:36.320782"}
{"step": 115, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.003486328125, "mean_total_reward": -0.00697265625, "std_total_reward": 0.0005527723319564416, "total_tokens": 357, "mean_completion_length": 7.14, "median_completion_length": 7, "min_completion_length": 6, "max_completion_length": 9, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.00021484375000000003, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.3512029647827148, "avg_step_time_sec": 16.697665494421255, "gpu_allocated_gb": 32.362786816, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1922.0674850940704, "eta_sec": 1419.3015670258067, "timestamp": "2026-02-15T11:29:37.686866"}
{"step": 116, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.2, "mean_correctness_reward": 0.2, "mean_length_penalty": -0.003447265625, "mean_total_reward": 0.19310546875, "std_total_reward": 0.3998828625825711, "total_tokens": 353, "mean_completion_length": 7.06, "median_completion_length": 7, "min_completion_length": 7, "max_completion_length": 8, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 8.950343154210625e-05, "frac_prompts_all_correct": 0.2, "frac_prompts_all_incorrect": 0.8, "step_time_sec": 1.361145257949829, "avg_step_time_sec": 16.565454113072363, "gpu_allocated_gb": 32.362778624, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1923.4355716705322, "eta_sec": 1391.4981454980784, "timestamp": "2026-02-15T11:29:39.060747"}
{"step": 117, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.003427734375, "mean_total_reward": -0.00685546875, "std_total_reward": 0.0007931226759471466, "total_tokens": 351, "mean_completion_length": 7.02, "median_completion_length": 7, "min_completion_length": 6, "max_completion_length": 11, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0003542647880315831, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.3673059940338135, "avg_step_time_sec": 16.435555411200237, "gpu_allocated_gb": 32.362766336, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1924.8109962940216, "eta_sec": 1364.1510991296198, "timestamp": "2026-02-15T11:29:40.498998"}
{"step": 118, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.02, "mean_correctness_reward": 0.02, "mean_length_penalty": -0.10458984375, "mean_total_reward": -0.1891796875, "std_total_reward": 0.6179474850165516, "total_tokens": 10710, "mean_completion_length": 214.2, "median_completion_length": 7, "min_completion_length": 5, "max_completion_length": 2048, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.22301478763391863, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 0.8, "step_time_sec": 1.41752028465271, "avg_step_time_sec": 16.308283927076953, "gpu_allocated_gb": 32.362875392, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1926.2364070415497, "eta_sec": 1337.2792820203101, "timestamp": "2026-02-15T11:29:54.827622"}
{"step": 119, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.002822265625, "mean_total_reward": -0.00564453125, "std_total_reward": 0.0009425725400010667, "total_tokens": 289, "mean_completion_length": 5.78, "median_completion_length": 5, "min_completion_length": 5, "max_completion_length": 7, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0006613323578013919, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 21.64835262298584, "avg_step_time_sec": 16.353158453933332, "gpu_allocated_gb": 32.365689856, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1947.891706943512, "eta_sec": 1324.6058347685998, "timestamp": "2026-02-15T11:30:03.493261"}
{"step": 120, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00283203125, "mean_total_reward": -0.0056640625, "std_total_reward": 0.0016105881350068987, "total_tokens": 290, "mean_completion_length": 5.8, "median_completion_length": 5, "min_completion_length": 5, "max_completion_length": 11, "frac_has_think": 0.0, "frac_has_boxed": 0.16, "mean_reward_std_per_prompt": 0.00017469281074217107, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.3440101146697998, "avg_step_time_sec": 16.2280822177728, "gpu_allocated_gb": 32.362784256, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1949.2425932884216, "eta_sec": 1298.246577421824, "timestamp": "2026-02-15T11:30:04.883648"}
{"step": 121, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.0024609375, "mean_total_reward": -0.004921875, "std_total_reward": 0.0002734375, "total_tokens": 252, "mean_completion_length": 5.04, "median_completion_length": 5, "min_completion_length": 5, "max_completion_length": 7, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.0001171875, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.8650195598602295, "avg_step_time_sec": 16.10937922059997, "gpu_allocated_gb": 32.36283136, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1951.1146869659424, "eta_sec": 1272.6409584273977, "timestamp": "2026-02-15T11:30:06.720587"}
{"step": 122, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00267578125, "mean_total_reward": -0.0053515625, "std_total_reward": 0.0010004881620988826, "total_tokens": 274, "mean_completion_length": 5.48, "median_completion_length": 5, "min_completion_length": 5, "max_completion_length": 9, "frac_has_think": 0.0, "frac_has_boxed": 0.04, "mean_reward_std_per_prompt": 0.00046875000000000004, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.349156141281128, "avg_step_time_sec": 15.988393785523586, "gpu_allocated_gb": 32.362811904, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1952.4710326194763, "eta_sec": 1247.0947152708397, "timestamp": "2026-02-15T11:30:08.101339"}
{"step": 123, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00255859375, "mean_total_reward": -0.0051171875, "std_total_reward": 0.0008414320011147663, "total_tokens": 262, "mean_completion_length": 5.24, "median_completion_length": 5, "min_completion_length": 5, "max_completion_length": 9, "frac_has_think": 0.0, "frac_has_boxed": 0.04, "mean_reward_std_per_prompt": 0.0003125, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.3631820678710938, "avg_step_time_sec": 15.869489625217469, "gpu_allocated_gb": 32.362789888, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1953.8419511318207, "eta_sec": 1221.9507011417452, "timestamp": "2026-02-15T11:30:09.439631"}
{"step": 124, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.2, "mean_correctness_reward": 0.2, "mean_length_penalty": -0.00244140625, "mean_total_reward": 0.1951171875, "std_total_reward": 0.4, "total_tokens": 250, "mean_completion_length": 5, "median_completion_length": 5, "min_completion_length": 5, "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": 0.2, "frac_prompts_all_incorrect": 0.8, "step_time_sec": 1.3344495296478271, "avg_step_time_sec": 15.752271559930616, "gpu_allocated_gb": 32.362771456, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1955.1812212467194, "eta_sec": 1197.1726385547267, "timestamp": "2026-02-15T11:30:10.782717"}
{"step": 125, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.022646484375, "mean_total_reward": -0.04529296875, "std_total_reward": 0.2792664763013395, "total_tokens": 2319, "mean_completion_length": 46.38, "median_completion_length": 5, "min_completion_length": 5, "max_completion_length": 2048, "frac_has_think": 0.0, "frac_has_boxed": 0.02, "mean_reward_std_per_prompt": 0.1195473478360511, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.3411774635314941, "avg_step_time_sec": 15.636982807159423, "gpu_allocated_gb": 32.362771456, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1956.5291571617126, "eta_sec": 1172.7737105369567, "timestamp": "2026-02-15T11:30:23.445955"}
{"step": 126, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00244140625, "mean_total_reward": -0.0048828125, "std_total_reward": 0.0, "total_tokens": 250, "mean_completion_length": 5, "median_completion_length": 5, "min_completion_length": 5, "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": 19.9596688747406, "avg_step_time_sec": 15.671289839441814, "gpu_allocated_gb": 32.365696, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1976.4958600997925, "eta_sec": 1159.6754481186942, "timestamp": "2026-02-15T11:30:32.076631"}
{"step": 127, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00244140625, "mean_total_reward": -0.0048828125, "std_total_reward": 0.0, "total_tokens": 250, "mean_completion_length": 5, "median_completion_length": 5, "min_completion_length": 5, "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": 1.3167991638183594, "avg_step_time_sec": 15.558262353806985, "gpu_allocated_gb": 32.3627776, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1977.8187992572784, "eta_sec": 1135.75315182791, "timestamp": "2026-02-15T11:30:33.398423"}
{"step": 128, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00244140625, "mean_total_reward": -0.0048828125, "std_total_reward": 0.0, "total_tokens": 250, "mean_completion_length": 5, "median_completion_length": 5, "min_completion_length": 5, "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": 1.325087308883667, "avg_step_time_sec": 15.44706567376852, "gpu_allocated_gb": 32.362763264, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1979.1498687267303, "eta_sec": 1112.1887285113335, "timestamp": "2026-02-15T11:30:34.741450"}
{"step": 129, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00244140625, "mean_total_reward": -0.0048828125, "std_total_reward": 0.0, "total_tokens": 250, "mean_completion_length": 5, "median_completion_length": 5, "min_completion_length": 5, "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": 1.3127224445343018, "avg_step_time_sec": 15.337497121603914, "gpu_allocated_gb": 32.362785792, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1980.4705052375793, "eta_sec": 1088.9622956338778, "timestamp": "2026-02-15T11:30:36.067878"}
{"step": 130, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00244140625, "mean_total_reward": -0.0048828125, "std_total_reward": 0.0, "total_tokens": 250, "mean_completion_length": 5, "median_completion_length": 5, "min_completion_length": 5, "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": 1.3262536525726318, "avg_step_time_sec": 15.229718325688289, "gpu_allocated_gb": 32.362780672, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1981.8048284053802, "eta_sec": 1066.0802827981802, "timestamp": "2026-02-15T11:30:37.382432"}
{"step": 131, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00244140625, "mean_total_reward": -0.0048828125, "std_total_reward": 0.0, "total_tokens": 250, "mean_completion_length": 5, "median_completion_length": 5, "min_completion_length": 5, "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": 1.3006246089935303, "avg_step_time_sec": 15.123389366018863, "gpu_allocated_gb": 32.36278272, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1983.1112418174744, "eta_sec": 1043.5138662553015, "timestamp": "2026-02-15T11:30:38.690862"}
{"step": 132, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00244140625, "mean_total_reward": -0.0048828125, "std_total_reward": 0.0, "total_tokens": 250, "mean_completion_length": 5, "median_completion_length": 5, "min_completion_length": 5, "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": 1.3125677108764648, "avg_step_time_sec": 15.018761929237481, "gpu_allocated_gb": 32.362745856, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1984.4291656017303, "eta_sec": 1021.2758111881487, "timestamp": "2026-02-15T11:30:39.998475"}
{"step": 133, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00216796875, "mean_total_reward": -0.0043359375, "std_total_reward": 0.001355418091632615, "total_tokens": 222, "mean_completion_length": 4.44, "median_completion_length": 5, "min_completion_length": 1, "max_completion_length": 5, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.000703125, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.7351791858673096, "avg_step_time_sec": 14.918885367257255, "gpu_allocated_gb": 32.362788864, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1986.1696305274963, "eta_sec": 999.5653196062361, "timestamp": "2026-02-15T11:30:41.738980"}
{"step": 134, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00103515625, "mean_total_reward": -0.0020703125, "std_total_reward": 0.001753901900050285, "total_tokens": 106, "mean_completion_length": 2.12, "median_completion_length": 1, "min_completion_length": 1, "max_completion_length": 5, "frac_has_think": 0.0, "frac_has_boxed": 0.0, "mean_reward_std_per_prompt": 0.00125, "frac_prompts_all_correct": 0.0, "frac_prompts_all_incorrect": 1.0, "step_time_sec": 1.3078320026397705, "avg_step_time_sec": 14.817310342148168, "gpu_allocated_gb": 32.362781696, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1987.4830050468445, "eta_sec": 977.9424825817791, "timestamp": "2026-02-15T11:30:43.117850"}
{"step": 135, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.3746416568756104, "avg_step_time_sec": 14.717735018553558, "gpu_allocated_gb": 32.362797056, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1988.8645071983337, "eta_sec": 956.6527762059812, "timestamp": "2026-02-15T11:30:44.436000"}
{"step": 136, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2841994762420654, "avg_step_time_sec": 14.618959021918913, "gpu_allocated_gb": 32.362750464, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1990.1542475223541, "eta_sec": 935.6133774028104, "timestamp": "2026-02-15T11:30:45.745816"}
{"step": 137, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.298048973083496, "avg_step_time_sec": 14.521726101854421, "gpu_allocated_gb": 32.36276992, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1991.4608833789825, "eta_sec": 914.8687444168286, "timestamp": "2026-02-15T11:30:47.009104"}
{"step": 138, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2507143020629883, "avg_step_time_sec": 14.42555934968202, "gpu_allocated_gb": 32.362781184, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1992.7178037166595, "eta_sec": 894.3846796802852, "timestamp": "2026-02-15T11:30:48.288323"}
{"step": 139, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2763924598693848, "avg_step_time_sec": 14.330961026733728, "gpu_allocated_gb": 32.362786304, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1994.0042402744293, "eta_sec": 874.1886226307574, "timestamp": "2026-02-15T11:30:49.579043"}
{"step": 140, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2835321426391602, "avg_step_time_sec": 14.237765106133052, "gpu_allocated_gb": 32.362813952, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1995.2984194755554, "eta_sec": 854.2659063679831, "timestamp": "2026-02-15T11:30:50.849230"}
{"step": 141, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2472591400146484, "avg_step_time_sec": 14.145633858146397, "gpu_allocated_gb": 32.362779136, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1996.552953004837, "eta_sec": 834.5923976306374, "timestamp": "2026-02-15T11:30:52.110946"}
{"step": 142, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.260667085647583, "avg_step_time_sec": 14.054894655523166, "gpu_allocated_gb": 32.362772992, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1997.8204548358917, "eta_sec": 815.1838900203436, "timestamp": "2026-02-15T11:30:53.374299"}
{"step": 143, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2778935432434082, "avg_step_time_sec": 13.965544997395336, "gpu_allocated_gb": 32.362758656, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 1999.1047356128693, "eta_sec": 796.0360648515341, "timestamp": "2026-02-15T11:30:54.670650"}
{"step": 144, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2749786376953125, "avg_step_time_sec": 13.877416064341864, "gpu_allocated_gb": 32.362783232, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2000.387713432312, "eta_sec": 777.1352996031444, "timestamp": "2026-02-15T11:30:55.950612"}
{"step": 145, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.701082706451416, "avg_step_time_sec": 13.793441351528825, "gpu_allocated_gb": 32.3627648, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2002.0962529182434, "eta_sec": 758.6392743340854, "timestamp": "2026-02-15T11:30:57.681506"}
{"step": 146, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.3109326362609863, "avg_step_time_sec": 13.707944716492744, "gpu_allocated_gb": 32.362792448, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2003.4161179065704, "eta_sec": 740.2290146906082, "timestamp": "2026-02-15T11:30:58.977444"}
{"step": 147, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2682061195373535, "avg_step_time_sec": 13.623320644404613, "gpu_allocated_gb": 32.362774016, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2004.6916499137878, "eta_sec": 722.0359941534444, "timestamp": "2026-02-15T11:31:00.236796"}
{"step": 148, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2561204433441162, "avg_step_time_sec": 13.539758480883933, "gpu_allocated_gb": 32.36278016, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2005.9536764621735, "eta_sec": 704.0674410059645, "timestamp": "2026-02-15T11:31:01.503448"}
{"step": 149, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2630243301391602, "avg_step_time_sec": 13.45736429195276, "gpu_allocated_gb": 32.362792448, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2007.2226145267487, "eta_sec": 686.3255788895908, "timestamp": "2026-02-15T11:31:02.785068"}
{"step": 150, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.26320481300354, "avg_step_time_sec": 13.376069895426433, "gpu_allocated_gb": 32.3627392, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2008.4942209720612, "eta_sec": 668.8034947713217, "timestamp": "2026-02-15T11:31:04.609132"}
{"step": 151, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2618045806884766, "avg_step_time_sec": 13.295842972812274, "gpu_allocated_gb": 32.362767872, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2010.3275353908539, "eta_sec": 651.4963056678014, "timestamp": "2026-02-15T11:31:05.893867"}
{"step": 152, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2908596992492676, "avg_step_time_sec": 13.216862819696727, "gpu_allocated_gb": 32.36283136, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2011.624356508255, "eta_sec": 634.4094153454429, "timestamp": "2026-02-15T11:31:07.169066"}
{"step": 153, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.259418249130249, "avg_step_time_sec": 13.138709587209364, "gpu_allocated_gb": 32.362781184, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2012.8891706466675, "eta_sec": 617.5193505988401, "timestamp": "2026-02-15T11:31:08.461674"}
{"step": 154, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.281118631362915, "avg_step_time_sec": 13.061712243340232, "gpu_allocated_gb": 32.362824192, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2014.1787309646606, "eta_sec": 600.8387631936506, "timestamp": "2026-02-15T11:31:09.732490"}
{"step": 155, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.261413335800171, "avg_step_time_sec": 12.985581282646425, "gpu_allocated_gb": 32.362782208, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2015.4468059539795, "eta_sec": 584.3511577190891, "timestamp": "2026-02-15T11:31:11.006839"}
{"step": 156, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.7007079124450684, "avg_step_time_sec": 12.91324235078616, "gpu_allocated_gb": 32.362803712, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2017.1534123420715, "eta_sec": 568.1826634345911, "timestamp": "2026-02-15T11:31:12.701822"}
{"step": 157, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2444872856140137, "avg_step_time_sec": 12.838919070116274, "gpu_allocated_gb": 32.362784256, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2018.4034163951874, "eta_sec": 552.0735200149998, "timestamp": "2026-02-15T11:31:13.957534"}
{"step": 158, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2610688209533691, "avg_step_time_sec": 12.765641536893725, "gpu_allocated_gb": 32.362816, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2019.6714389324188, "eta_sec": 536.1569445495364, "timestamp": "2026-02-15T11:31:15.246918"}
{"step": 159, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2769277095794678, "avg_step_time_sec": 12.693385475086716, "gpu_allocated_gb": 32.362809856, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2020.956267118454, "eta_sec": 520.4288044785553, "timestamp": "2026-02-15T11:31:16.500967"}
{"step": 160, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2409563064575195, "avg_step_time_sec": 12.621807792782784, "gpu_allocated_gb": 32.362781184, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2022.2026104927063, "eta_sec": 504.87231171131134, "timestamp": "2026-02-15T11:31:17.751826"}
{"step": 161, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.262016773223877, "avg_step_time_sec": 12.551250084586766, "gpu_allocated_gb": 32.362819072, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2023.4703485965729, "eta_sec": 489.49875329888386, "timestamp": "2026-02-15T11:31:19.019240"}
{"step": 162, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.271580457687378, "avg_step_time_sec": 12.481622494297262, "gpu_allocated_gb": 32.362753536, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2024.7474768161774, "eta_sec": 474.301654783296, "timestamp": "2026-02-15T11:31:20.298192"}
{"step": 163, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2816450595855713, "avg_step_time_sec": 12.412910976292897, "gpu_allocated_gb": 32.36279552, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2026.0351085662842, "eta_sec": 459.2777061228372, "timestamp": "2026-02-15T11:31:21.591427"}
{"step": 164, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2587954998016357, "avg_step_time_sec": 12.344898077045999, "gpu_allocated_gb": 32.362760704, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2027.3006219863892, "eta_sec": 444.41633077365594, "timestamp": "2026-02-15T11:31:22.864610"}
{"step": 165, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2811989784240723, "avg_step_time_sec": 12.277845355236169, "gpu_allocated_gb": 32.362817024, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2028.5876870155334, "eta_sec": 429.7245874332659, "timestamp": "2026-02-15T11:31:24.133373"}
{"step": 166, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2439935207366943, "avg_step_time_sec": 12.211376368281353, "gpu_allocated_gb": 32.362758656, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2029.8378298282623, "eta_sec": 415.186796521566, "timestamp": "2026-02-15T11:31:25.396478"}
{"step": 167, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2624363899230957, "avg_step_time_sec": 12.145813853440885, "gpu_allocated_gb": 32.362817024, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2031.1071724891663, "eta_sec": 400.8118571635492, "timestamp": "2026-02-15T11:31:26.656306"}
{"step": 168, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.6284615993499756, "avg_step_time_sec": 12.083210566214152, "gpu_allocated_gb": 32.362812928, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2032.7408781051636, "eta_sec": 386.6627381188529, "timestamp": "2026-02-15T11:31:28.298607"}
{"step": 169, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.257678747177124, "avg_step_time_sec": 12.019154164918076, "gpu_allocated_gb": 32.362766848, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2034.006563425064, "eta_sec": 372.59377911246037, "timestamp": "2026-02-15T11:31:29.552155"}
{"step": 170, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2428176403045654, "avg_step_time_sec": 11.955763950067407, "gpu_allocated_gb": 32.362753536, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2035.2560863494873, "eta_sec": 358.6729185020222, "timestamp": "2026-02-15T11:31:30.809832"}
{"step": 171, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2577621936798096, "avg_step_time_sec": 11.893202536287363, "gpu_allocated_gb": 32.362767872, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2036.5217232704163, "eta_sec": 344.9028735523335, "timestamp": "2026-02-15T11:31:32.100224"}
{"step": 172, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2842693328857422, "avg_step_time_sec": 11.83152269208154, "gpu_allocated_gb": 32.362835456, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2037.8122725486755, "eta_sec": 331.28263537828315, "timestamp": "2026-02-15T11:31:33.377060"}
{"step": 173, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2817974090576172, "avg_step_time_sec": 11.770541621081401, "gpu_allocated_gb": 32.36277504, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2039.1027765274048, "eta_sec": 317.8046237691978, "timestamp": "2026-02-15T11:31:34.648819"}
{"step": 174, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2543928623199463, "avg_step_time_sec": 11.710103984536795, "gpu_allocated_gb": 32.362755584, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2040.363031387329, "eta_sec": 304.4627035979567, "timestamp": "2026-02-15T11:31:35.916519"}
{"step": 175, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2508423328399658, "avg_step_time_sec": 11.650336775098529, "gpu_allocated_gb": 32.362770944, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2041.6203918457031, "eta_sec": 291.2584193774632, "timestamp": "2026-02-15T11:31:37.186386"}
{"step": 176, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2716376781463623, "avg_step_time_sec": 11.591366893865846, "gpu_allocated_gb": 32.362774016, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2042.9000334739685, "eta_sec": 278.1928054527803, "timestamp": "2026-02-15T11:31:38.445288"}
{"step": 177, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2557125091552734, "avg_step_time_sec": 11.53297336626861, "gpu_allocated_gb": 32.362750464, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2044.1630549430847, "eta_sec": 265.258387424178, "timestamp": "2026-02-15T11:31:39.727763"}
{"step": 178, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2679352760314941, "avg_step_time_sec": 11.475304612952671, "gpu_allocated_gb": 32.362778112, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2045.4383232593536, "eta_sec": 252.45670148495876, "timestamp": "2026-02-15T11:31:40.998933"}
{"step": 179, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2707409858703613, "avg_step_time_sec": 11.418295877605843, "gpu_allocated_gb": 32.362781184, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2046.7174754142761, "eta_sec": 239.78421342972268, "timestamp": "2026-02-15T11:31:42.337796"}
{"step": 180, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.7793333530426025, "avg_step_time_sec": 11.364746085802715, "gpu_allocated_gb": 32.362833408, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2048.507253885269, "eta_sec": 227.2949217160543, "timestamp": "2026-02-15T11:31:44.087311"}
{"step": 181, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2912631034851074, "avg_step_time_sec": 11.309091483690462, "gpu_allocated_gb": 32.362789376, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2049.8063945770264, "eta_sec": 214.87273819011878, "timestamp": "2026-02-15T11:31:45.353743"}
{"step": 182, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2509503364562988, "avg_step_time_sec": 11.253826971892472, "gpu_allocated_gb": 32.362765824, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2051.063127756119, "eta_sec": 202.5688854940645, "timestamp": "2026-02-15T11:31:46.617283"}
{"step": 183, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2563300132751465, "avg_step_time_sec": 11.199195840971067, "gpu_allocated_gb": 32.36275456, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2052.327018737793, "eta_sec": 190.38632929650814, "timestamp": "2026-02-15T11:31:47.878039"}
{"step": 184, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2593047618865967, "avg_step_time_sec": 11.14517469380213, "gpu_allocated_gb": 32.362770944, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2053.593661546707, "eta_sec": 178.32279510083407, "timestamp": "2026-02-15T11:31:49.151698"}
{"step": 185, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2638721466064453, "avg_step_time_sec": 11.091762247601071, "gpu_allocated_gb": 32.362799616, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2054.8628718852997, "eta_sec": 166.37643371401606, "timestamp": "2026-02-15T11:31:50.426977"}
{"step": 186, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2751753330230713, "avg_step_time_sec": 11.038984898597963, "gpu_allocated_gb": 32.362830336, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2056.143615245819, "eta_sec": 154.5457885803715, "timestamp": "2026-02-15T11:31:51.696683"}
{"step": 187, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.25425386428833, "avg_step_time_sec": 10.986660133708607, "gpu_allocated_gb": 32.362788352, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2057.4041526317596, "eta_sec": 142.8265817382119, "timestamp": "2026-02-15T11:31:53.015711"}
{"step": 188, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.328725814819336, "avg_step_time_sec": 10.935288142650686, "gpu_allocated_gb": 32.3628416, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2058.741499185562, "eta_sec": 131.22345771180824, "timestamp": "2026-02-15T11:31:54.297401"}
{"step": 189, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2614855766296387, "avg_step_time_sec": 10.884104002089728, "gpu_allocated_gb": 32.362792448, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2060.0095632076263, "eta_sec": 119.72514402298701, "timestamp": "2026-02-15T11:31:55.595773"}
{"step": 190, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2840819358825684, "avg_step_time_sec": 10.833577570162321, "gpu_allocated_gb": 32.362753536, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2061.3029561042786, "eta_sec": 108.3357757016232, "timestamp": "2026-02-15T11:31:56.855782"}
{"step": 191, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2589771747589111, "avg_step_time_sec": 10.783448772280629, "gpu_allocated_gb": 32.362774016, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2062.5687971115112, "eta_sec": 97.05103895052565, "timestamp": "2026-02-15T11:31:58.218152"}
{"step": 192, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.817572832107544, "avg_step_time_sec": 10.736751501758894, "gpu_allocated_gb": 32.362886656, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2064.3936352729797, "eta_sec": 85.89401201407115, "timestamp": "2026-02-15T11:31:59.957737"}
{"step": 193, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2817373275756836, "avg_step_time_sec": 10.68776179101183, "gpu_allocated_gb": 32.362758656, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2065.684767484665, "eta_sec": 74.8143325370828, "timestamp": "2026-02-15T11:32:01.278138"}
{"step": 194, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2948980331420898, "avg_step_time_sec": 10.639344967517657, "gpu_allocated_gb": 32.362847744, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2066.988857984543, "eta_sec": 63.836069805105936, "timestamp": "2026-02-15T11:32:02.524223"}
{"step": 195, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2322392463684082, "avg_step_time_sec": 10.591103399716891, "gpu_allocated_gb": 32.36278016, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2068.2269191741943, "eta_sec": 52.95551699858446, "timestamp": "2026-02-15T11:32:03.771553"}
{"step": 196, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2444000244140625, "avg_step_time_sec": 10.543416137598, "gpu_allocated_gb": 32.362762752, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2069.4769468307495, "eta_sec": 42.173664550392, "timestamp": "2026-02-15T11:32:05.025761"}
{"step": 197, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2607753276824951, "avg_step_time_sec": 10.496296133486753, "gpu_allocated_gb": 32.362781184, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2070.7428119182587, "eta_sec": 31.48888840046026, "timestamp": "2026-02-15T11:32:06.454516"}
{"step": 198, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.4861667156219482, "avg_step_time_sec": 10.450790429356122, "gpu_allocated_gb": 32.362960384, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2072.2367153167725, "eta_sec": 20.901580858712244, "timestamp": "2026-02-15T11:32:07.769242"}
{"step": 199, "num_prompts": 5, "num_rollouts": 50, "accuracy": 0.0, "mean_correctness_reward": 0.0, "mean_length_penalty": -0.00048828125, "mean_total_reward": -0.0009765625, "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": 1.2387380599975586, "avg_step_time_sec": 10.404498708907084, "gpu_allocated_gb": 32.362765824, "gpu_reserved_gb": 79.958114304, "gpu_peak_gb": 74.53288704, "elapsed_sec": 2073.4804158210754, "eta_sec": 10.404498708907084, "timestamp": "2026-02-15T11:32:09.090856"}