| { |
| "host": "eai-Raider-16-HX-AI-A2XWIG", |
| "created": "2026-09-20T11:50:15.379559", |
| "task": "turning_on_radio", |
| "task_id": 0, |
| "split": "2026 public_test", |
| "instance_ids": [ |
| 301, |
| 302, |
| 303, |
| 304, |
| 305, |
| 306, |
| 307, |
| 308, |
| 309, |
| 310, |
| 311, |
| 312, |
| 313, |
| 314, |
| 315, |
| 316, |
| 317, |
| 318, |
| 319, |
| 320 |
| ], |
| "rollouts_per_instance": 1, |
| "variants": [ |
| "human", |
| "codegen_success", |
| "mix_h50_c50" |
| ], |
| "seed": 0, |
| "seed_reset": "Each simulator reset and each policy reset", |
| "max_steps": 3224, |
| "timeout": "Official 1.5 x mean human-demo length", |
| "simulator_repository": "/home/eai/code/b1k", |
| "simulator_commit": "b1979916ec1549b10a4e65e630bc6504a9af1b00", |
| "simulator_python": "/home/eai/miniconda3/envs/behavior-392dev/bin/python", |
| "policy_python": "/home/eai/code/training_models/evaluation/.policy-venv/bin/python", |
| "upstream_policy_commit": "ca556f74a455cef7987a2be4537b5ac85cc56dd7", |
| "openpi_commit": "01177e0242a1c7e8fad2547caa0e987def614cda", |
| "archived_model_sources": [ |
| { |
| "path": "/home/eai/Documents/Codex/2026-09-18/b1k-chat-part1-b1k-gui-jsonl/work/source/src/b1k/models/pi_behavior.py", |
| "sha256": "68ff7c1409b6c6dae97a7ffcc8a14aec412650797204cd763bf4433d9f276198", |
| "matches_training_manifest": true |
| }, |
| { |
| "path": "/home/eai/Documents/Codex/2026-09-18/b1k-chat-part1-b1k-gui-jsonl/work/source/src/b1k/models/pi_behavior_config.py", |
| "sha256": "95139c8172d7cd300a82df7dec3a467b2e0fe047edf1896e347441610a77a878", |
| "matches_training_manifest": true |
| } |
| ], |
| "normalization": "Each variant checkpoint-14999/assets/IliaLarchenko/behavior_224_rgb/norm_stats.json", |
| "inference": { |
| "parameter_dtype": "bfloat16 (upstream server default)", |
| "num_flow_inference_steps": 20, |
| "action_horizon": 30, |
| "actions_to_execute": 26, |
| "actions_to_keep": 4, |
| "execute_in_n_steps": 20, |
| "stage_history": 3, |
| "stage_votes": 2, |
| "apply_eval_tricks": true, |
| "task_embedding_id": 0, |
| "no_da3": true, |
| "no_skill_conditioning": true |
| }, |
| "compatibility_notes": [ |
| "Full training source was not found locally. Archived pi_behavior.py and pi_behavior_config.py are used with upstream OpenPI and supporting modules.", |
| "Disabled DA3 keyword arguments and imports are removed; the adapter rejects enabled DA3/skill branches. Upstream Gemma is used for the RGB-only path; full training Gemma source was unavailable.", |
| "61D proprioception indices copied from the actual 2026 evaluator.", |
| "Full parameter tree and shapes checked with remove_extra_params=False.", |
| "Fixed correlation-aware inpainting linear system solved on CPU in float32 and residual verified <1e-4; GPU cuSolver fails on this eai setup." |
| ], |
| "metric_source": "Official evaluator success and q_score.final; simulator/infrastructure errors are not counted as task failures", |
| "simulator_source_hashes": { |
| "/home/eai/code/b1k/OmniGibson/omnigibson/eval/evaluator.py": "8aa93be2902a784d2f1b711965ac4abc59f3a51eb9b1d1b9669291d4f4bcd45d", |
| "/home/eai/code/b1k/OmniGibson/omnigibson/eval/eval.py": "829922a257f084ebff2c6675251ab09783f789dc74aa77c18fb44983e5e34d82", |
| "/home/eai/code/b1k/OmniGibson/omnigibson/eval/r1pro.yaml": "a98fa8a81472adfecfb642778d4d7a01e20131be7f70824e0ae095d665cdbb93", |
| "/home/eai/code/b1k/OmniGibson/omnigibson/eval/utils/eval_utils.py": "659aceb8e8f404cabb00b3697f474b068254e1245c544ffab73c0f64947334d3", |
| "/home/eai/code/b1k/datasets/2026-challenge-task-instances/metadata/task.jsonl": "7dbb5f7f00a2e06c693dfbb3045e5c28e627b6179e775f1ad93ca4b61f08e6dc" |
| }, |
| "runner_source_hashes": { |
| "/home/eai/code/training_models/evaluation/serve_checkpoint.py": "3e4f9f6a803895d893771f5b0675c13707227dfb030405ece0c2bcbf1b7c6713", |
| "/home/eai/code/training_models/evaluation/run_comparison.py": "cd338e542238e358bd46fce568b893849cece994c707a7b3d73935bc1fc80dc8", |
| "/home/eai/code/training_models/evaluation/summarize.py": "11ed9c2e94287f357b4210d22cbd3c09fb6dc6755c1f90820627835dfee5264c", |
| "/home/eai/code/training_models/evaluation/run_simulator.py": "017558d9affd2e9acd1418e78203cf684183e3c8f91f93968f4e23347c275d0f", |
| "/home/eai/code/training_models/evaluation/prepare_runtime.py": "d2304c4d8b622a2037814461b06f758b92fdbdd90f736868704895a35706b846", |
| "/home/eai/code/training_models/evaluation/b1k_eval_contract.py": "cfb4367f6d2b3c994b2e0c79812b73f831b20d79b70344605a45e9bb79aecada", |
| "/home/eai/code/training_models/evaluation/audit_results.py": "a8b3c17a9b1cb7d774dc883147decd517c3592eac1e162ee9c517df301287789", |
| "/home/eai/code/training_models/evaluation/progress.py": "0db83004c290a768a2f7fbe29b49e8166fc048b06cbf9c93ac9dca51e9d70764" |
| }, |
| "timeout_note": "Official max_steps=3224 produces 3225 recorded simulator steps for timeout rollouts due to the evaluator termination check.", |
| "hardware": { |
| "gpu": "NVIDIA GeForce RTX 5080 Laptop GPU", |
| "gpu_memory_mib": 16303, |
| "driver_version": "580.173.02" |
| }, |
| "completion": { |
| "status": "complete", |
| "rollouts": 60, |
| "completed_at": "2026-09-20T14:53:24.163149", |
| "audit": "audit.json", |
| "raw_result_checksums": "RESULTS_SHA256SUMS", |
| "gpu_compute_processes_after_run": 0 |
| } |
| } |
|
|
