| |
| """Build a compact source-of-truth artifact index for the research project. |
| |
| The index is intentionally selective. It lists the files behind the public |
| project readouts, not every prediction array or checkpoint in the repository. |
| """ |
|
|
| from __future__ import annotations |
|
|
| import hashlib |
| import json |
| from datetime import datetime, timezone |
| from pathlib import Path |
|
|
|
|
| ROOT = Path(__file__).resolve().parents[1] |
| OUTPUT = ROOT / "docs/data/artifact_index.json" |
| QWEN3_FUTURE_TASK_PROBE_RUN_ID = "xperience10m_qwen3_omni_v6_future_task_probes_a100_20260616T143608Z" |
| COSMOS3_SUPER_INTERACTION_TEXT_TASK_PROBE_RUN_ID = ( |
| "xperience10m_cosmos3_super_interaction_text_task15_textonly_v1_20260620T1558Z" |
| ) |
|
|
| ARTIFACTS = [ |
| { |
| "id": "project_brief", |
| "title": "Project brief", |
| "path": "PROJECT_BRIEF.md", |
| "kind": "project_path", |
| "surface": "repo_hf", |
| "shows": "Gives first-pass readers a concise project shape before the detailed artifact trail.", |
| }, |
| { |
| "id": "project_brief_json", |
| "title": "Project brief JSON", |
| "path": "docs/data/project_brief.json", |
| "kind": "project_path", |
| "surface": "website_hf", |
| "shows": "Machine-readable first-reader project brief for the website and Hugging Face mirrors.", |
| }, |
| { |
| "id": "project_status", |
| "title": "Project status", |
| "path": "PROJECT_STATUS.md", |
| "kind": "project_path", |
| "surface": "repo_hf", |
| "shows": "Gives a compact current-state table for first-pass readers.", |
| }, |
| { |
| "id": "project_status_json", |
| "title": "Project status JSON", |
| "path": "docs/data/project_status.json", |
| "kind": "project_path", |
| "surface": "website_hf", |
| "shows": "Machine-readable copy of the current project status for website and HF mirrors.", |
| }, |
| { |
| "id": "glossary", |
| "title": "Glossary", |
| "path": "GLOSSARY.md", |
| "kind": "project_path", |
| "surface": "repo_hf", |
| "shows": "Defines terminology that can be confused across data scope, task metrics, model branches, and public mirrors.", |
| }, |
| { |
| "id": "glossary_json", |
| "title": "Glossary JSON", |
| "path": "docs/data/glossary.json", |
| "kind": "project_path", |
| "surface": "website_hf", |
| "shows": "Machine-readable terminology layer for the website, artifact dataset, model mirror, and public QA checks.", |
| }, |
| { |
| "id": "research_roadmap", |
| "title": "Research roadmap", |
| "path": "RESEARCH_ROADMAP.md", |
| "kind": "project_path", |
| "surface": "repo_hf", |
| "shows": "Defines the path from public-sample task development to multi-episode held-out evaluation and larger omni-model extensions.", |
| }, |
| { |
| "id": "research_roadmap_json", |
| "title": "Research roadmap JSON", |
| "path": "docs/data/research_roadmap.json", |
| "kind": "project_path", |
| "surface": "website_hf", |
| "shows": "Machine-readable research roadmap for the website and Hugging Face mirrors.", |
| }, |
| { |
| "id": "foundation_model_plan", |
| "title": "Foundation model plan", |
| "path": "FOUNDATION_MODEL_PLAN.md", |
| "kind": "project_path", |
| "surface": "repo_hf", |
| "shows": "Defines the post-data-gate backbone choices: Qwen3-Omni first, Cosmos 3 for world modeling, and VLA/policy models after action-target conversion.", |
| }, |
| { |
| "id": "foundation_model_plan_json", |
| "title": "Foundation model plan JSON", |
| "path": "docs/data/foundation_model_plan.json", |
| "kind": "project_path", |
| "surface": "website_hf", |
| "shows": "Machine-readable foundation-model selection matrix with source links, entry conditions, and evaluation additions.", |
| }, |
| { |
| "id": "three_foundation_pipelines", |
| "title": "Three foundation pipeline tracks", |
| "path": "THREE_FOUNDATION_PIPELINES.md", |
| "kind": "project_path", |
| "surface": "repo_hf", |
| "shows": "Frames spatial intelligence, human-video world modeling, and vision-language-action as three pipeline tracks with explicit inputs, outputs, maturity, and next evidence gates.", |
| }, |
| { |
| "id": "three_foundation_pipelines_json", |
| "title": "Three foundation pipeline tracks JSON", |
| "path": "docs/data/three_foundation_pipelines.json", |
| "kind": "project_path", |
| "surface": "website_hf", |
| "shows": "Machine-readable pipeline-track contract for the website and Hugging Face mirrors.", |
| }, |
| { |
| "id": "spatial_intelligence_slide_diagram", |
| "title": "Spatial intelligence slide diagram", |
| "path": "docs/assets/foundation-pipelines/spatial-intelligence-pipeline.png", |
| "kind": "visual_asset", |
| "surface": "website_hf", |
| "shows": "High-resolution slide diagram for the spatial intelligence model training pipeline direction.", |
| }, |
| { |
| "id": "human_video_world_model_slide_diagram", |
| "title": "Human-video world model slide diagram", |
| "path": "docs/assets/foundation-pipelines/human-video-world-model-pipeline.png", |
| "kind": "visual_asset", |
| "surface": "website_hf", |
| "shows": "High-resolution slide diagram for the human-video world-model training pipeline direction.", |
| }, |
| { |
| "id": "vision_language_action_slide_diagram", |
| "title": "Vision-language-action slide diagram", |
| "path": "docs/assets/foundation-pipelines/vision-language-action-pipeline.png", |
| "kind": "visual_asset", |
| "surface": "website_hf", |
| "shows": "High-resolution slide diagram for the vision-language-action training pipeline direction.", |
| }, |
| { |
| "id": "spatial_intelligence_source_slide", |
| "title": "Spatial intelligence source slide", |
| "path": "docs/assets/foundation-pipelines/source-slides/spatial-intelligence-slide.png", |
| "kind": "visual_asset_source", |
| "surface": "repo_hf", |
| "shows": "Clean source slide PNG supplied for the spatial intelligence public direction figure.", |
| }, |
| { |
| "id": "human_video_world_model_source_slide", |
| "title": "Human-video world model source slide", |
| "path": "docs/assets/foundation-pipelines/source-slides/human-video-world-model-slide.png", |
| "kind": "visual_asset_source", |
| "surface": "repo_hf", |
| "shows": "Clean source slide PNG supplied for the human-video world-model public direction figure.", |
| }, |
| { |
| "id": "vision_language_action_source_slide", |
| "title": "Vision-language-action source slide", |
| "path": "docs/assets/foundation-pipelines/source-slides/vision-language-action-slide.png", |
| "kind": "visual_asset_source", |
| "surface": "repo_hf", |
| "shows": "Clean source slide PNG supplied for the vision-language-action public direction figure.", |
| }, |
| { |
| "id": "omni_model_extension_contract", |
| "title": "Omni model extension contract", |
| "path": "OMNI_MODEL_EXTENSION_CONTRACT.md", |
| "kind": "scaleup_contract", |
| "surface": "repo_hf", |
| "shows": "Defines the shared manifest, episode split, held-out evaluation, packaging, and public-safety rules for Qwen3-Omni, Cosmos3, and VLA/policy model tracks.", |
| }, |
| { |
| "id": "omni_backbone_registry_configs", |
| "title": "Omni backbone registry configs", |
| "path": "configs/omni_backbones", |
| "kind": "scaleup_contract", |
| "surface": "repo_hf", |
| "shows": "Stores the implemented Qwen3-Omni LoRA contract and planned Cosmos-style world-model and VLA/policy branch contracts.", |
| }, |
| { |
| "id": "omni_backbone_registry_validator", |
| "title": "Omni backbone registry validator", |
| "path": "scripts/omni/backbone_registry.py", |
| "kind": "scaleup_contract", |
| "surface": "repo_hf", |
| "shows": "Validates backbone ids, split defaults, leakage guards, required metrics, required files, and forbidden public package categories.", |
| }, |
| { |
| "id": "omni_model_neutral_window_index_exporter", |
| "title": "Model-neutral window index exporter", |
| "path": "scripts/omni/export_model_neutral_window_index.py", |
| "kind": "scaleup_contract", |
| "surface": "repo_hf", |
| "shows": "Converts Qwen JSONL records into a model-neutral window index with Qwen, Cosmos-style, and policy/VLA adapter views.", |
| }, |
| { |
| "id": "omni_backbone_scaffolder", |
| "title": "Omni backbone scaffolder", |
| "path": "scripts/omni/scaffold_omni_backbone.py", |
| "kind": "scaleup_contract", |
| "surface": "repo_hf", |
| "shows": "Creates a validated planned-backbone config from an existing contract template so new model families inherit the shared rules.", |
| }, |
| { |
| "id": "omni_backbone_packaging_smoke", |
| "title": "Omni backbone packaging smoke test", |
| "path": "scripts/omni/smoke_test_backbone_packaging.py", |
| "kind": "scaleup_contract", |
| "surface": "repo_hf", |
| "shows": "Builds synthetic verified packages for every configured backbone and audits them against the public-safe package contract.", |
| }, |
| { |
| "id": "qwen3_omni_error_analysis_script", |
| "title": "Qwen3-Omni held-out error-analysis script", |
| "path": "scripts/omni/analyze_qwen3_omni_errors.py", |
| "kind": "scaleup_contract", |
| "surface": "repo_hf", |
| "shows": "Computes public-safe held-out error-analysis tables by episode, action family, train-seen status, required-modality state, and object category.", |
| }, |
| { |
| "id": "multi_episode_128_baseline_script", |
| "title": "128-episode aligned baseline runner", |
| "path": "scripts/omni/run_128_task_baselines.py", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Runs simple metadata and neural MLP baselines on the same selected 96/16/16 episode split used by the Qwen3-Omni diagnostic pilot.", |
| }, |
| { |
| "id": "task_suite_enhancement_128", |
| "title": "128-episode task-suite enhancement pack", |
| "path": "TASK_SUITE_ENHANCEMENT_128.md", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Records the no-new-episode dense-window, hierarchical-target, bottleneck, and experiment-backlog plan for pushing the current 128-episode suite harder without overwriting prior results.", |
| }, |
| { |
| "id": "task_suite_enhancement_128_json", |
| "title": "128-episode task-suite enhancement JSON", |
| "path": "docs/data/task_suite_enhancement_128.json", |
| "kind": "scaleup_status", |
| "surface": "website_hf", |
| "shows": "Machine-readable enhancement pack for the website and Hugging Face mirrors.", |
| }, |
| { |
| "id": "task_suite_enhancement_128_result", |
| "title": "128-episode task-suite enhancement result package", |
| "path": "results/omni_finetune/task_suite_enhancement_128_v1_20260608/enhancement_plan.json", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Versioned result directory with dense-window estimates, hierarchical target contract, task bottlenecks, Qwen action-family error summary, and experiment cards.", |
| }, |
| { |
| "id": "task_suite_enhancement_128_builder", |
| "title": "128-episode task-suite enhancement builder", |
| "path": "scripts/omni/build_task_suite_enhancement_128.py", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Regenerates the enhancement pack from committed 128-episode windows, baseline summaries, verified Qwen predictions, and Cosmos reference metrics.", |
| }, |
| { |
| "id": "xperience10m_128_episode_feature_index", |
| "title": "Xperience-10M 128-episode source and feature index", |
| "path": "XPERIENCE10M_128_EPISODE_FEATURE_INDEX.md", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Links each selected official session UUID and episode id to the gated Xperience-10M source tree plus the public-safe processed feature artifacts derived from the selected split.", |
| }, |
| { |
| "id": "xperience10m_128_episode_feature_index_json", |
| "title": "Xperience-10M 128-episode source and feature index JSON", |
| "path": "docs/data/xperience10m_128_episode_feature_index.json", |
| "kind": "scaleup_status", |
| "surface": "website_hf", |
| "shows": "Machine-readable 128-episode source-to-feature map for GitHub Pages, HF Space, artifact dataset, and baseline-model mirrors.", |
| }, |
| { |
| "id": "xperience10m_128_episode_feature_index_builder", |
| "title": "128-episode source and feature index builder", |
| "path": "scripts/omni/build_128_episode_feature_index.py", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Regenerates the public-safe source/feature index from the selected episode manifest, dense multiscale export, metadata matrices, and raw20 baseline summary.", |
| }, |
| { |
| "id": "xperience10m_128_dense_multiscale_windows", |
| "title": "128-episode dense multiscale public-safe windows", |
| "path": "results/omni_finetune/xperience10m_128ep_dense_multiscale_hierarchical_v1_20260608/dense_multiscale_windows.jsonl", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Public-safe dense/medium/long window rows over the exportable selected episodes, linked back to official source episode ids without redistributing raw gated files.", |
| }, |
| { |
| "id": "xperience10m_128_metadata_matrix_v2", |
| "title": "128-episode metadata feature matrix v2", |
| "path": "results/omni_finetune/a100_128_metadata_task_baselines_20260616_v2/metadata_feature_matrix.npz", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Public-safe 34,269 x 394 metadata/text feature matrix used by the aligned 128-episode metadata baseline layer.", |
| }, |
| { |
| "id": "qwen3_full_parameter_gates", |
| "title": "Qwen3-Omni full-parameter feasibility gates", |
| "path": "results/omni_finetune/QWEN3_FULL_PARAMETER_GATES_20260609.md", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Summarizes the 2026-06-09 full-parameter FSDP feasibility gates: 1/8/32/64-step guarded runs passed, the 128-step opportunistic pilot was preempted for Qwen v5 handoff, and no full checkpoints or weights are published.", |
| }, |
| { |
| "id": "qwen3_full_parameter_gates_json", |
| "title": "Qwen3-Omni full-parameter feasibility gates JSON", |
| "path": "docs/data/qwen3_full_parameter_gates.json", |
| "kind": "scaleup_status", |
| "surface": "website_hf", |
| "shows": "Machine-readable summary of full-parameter feasibility evidence and publication policy for website and Hugging Face mirrors.", |
| }, |
| { |
| "id": "qwen3_v5_v6_comparison", |
| "title": "Qwen3-Omni v5/v6 comparison", |
| "path": "results/omni_finetune/QWEN3_V5_V6_COMPARISON_20260614.md", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Reader-facing comparison of the verified Qwen3 v5 release row and the latest verified v6 row, including metric deltas and release-tag policy.", |
| }, |
| { |
| "id": "qwen3_v5_v6_comparison_json", |
| "title": "Qwen3-Omni v5/v6 comparison JSON", |
| "path": "docs/data/qwen3_v5_v6_comparison.json", |
| "kind": "scaleup_status", |
| "surface": "website_hf", |
| "shows": "Machine-readable v5/v6 metric deltas and publication recommendation for website and Hugging Face mirrors.", |
| }, |
| { |
| "id": "qwen3_full_parameter_gates_builder", |
| "title": "Qwen3-Omni full-parameter gate summary builder", |
| "path": "scripts/omni/build_qwen3_full_parameter_gate_summary.py", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Regenerates the full-parameter feasibility-gate Markdown and JSON summaries from the run-local evidence files.", |
| }, |
| { |
| "id": "qwen3_full_parameter_post_verified_deferrer", |
| "title": "Qwen3-Omni post-verified full-parameter deferrer", |
| "path": "scripts/omni/defer_qwen3_fullparam_after_verified_qwen.sh", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Waits for a verified Qwen held-out package, then launches a bounded 128-step full-parameter feasibility pilot on the same multiscale v5 dataset with no checkpoints or weights saved.", |
| }, |
| { |
| "id": "qwen3_lora_hf_package_builder", |
| "title": "Qwen3 LoRA HF package builder", |
| "path": "scripts/omni/prepare_qwen3_lora_hf_package.py", |
| "kind": "publication_workflow", |
| "surface": "repo_hf", |
| "shows": "Builds the upload-ready Hugging Face adapter folder from a verified Qwen3 LoRA result summary and adapter directory.", |
| }, |
| { |
| "id": "qwen3_private_gpu_repro_smoke", |
| "title": "Qwen3 private staged-GPU reproduction smoke", |
| "path": "scripts/omni/run_private_gpu_qwen3_v6_repro_smoke.sh", |
| "kind": "reproducibility", |
| "surface": "repo_hf", |
| "shows": "Runs the owner-side Qwen3-Omni v6 one-sample reproduction smoke from a private staged model, adapter, JSONL, and exported media cache.", |
| }, |
| { |
| "id": "qwen3_video_feature_compat_patch", |
| "title": "Qwen3 video-feature compatibility patch checker", |
| "path": "scripts/omni/patch_qwen3_omni_video_features.py", |
| "kind": "reproducibility", |
| "surface": "repo_hf", |
| "shows": "Checks and narrowly repairs the installed Qwen3-Omni video-feature branch so private staged-GPU reproduction uses the verified source-compatible behavior.", |
| }, |
| { |
| "id": "additional_development_directions", |
| "title": "Additional development directions", |
| "path": "ADDITIONAL_DEVELOPMENT_DIRECTIONS.md", |
| "kind": "project_path", |
| "surface": "repo_hf", |
| "shows": "Records concrete non-backbone Xperience-10M development tracks: taxonomy, benchmark protocol, representation learning, skill graphs, affordances, 3D/4D memory, QA, and policy transfer.", |
| }, |
| { |
| "id": "additional_development_directions_json", |
| "title": "Additional development directions JSON", |
| "path": "docs/data/additional_development_directions.json", |
| "kind": "project_path", |
| "surface": "website_hf", |
| "shows": "Machine-readable additional development directions for the website and Hugging Face mirrors.", |
| }, |
| { |
| "id": "xperience_embodied_foundation_pretraining", |
| "title": "Xperience Embodied Foundation Model pretraining goal", |
| "path": "XPERIENCE_EMBODIED_FOUNDATION_MODEL_PRETRAINING.md", |
| "kind": "project_path", |
| "surface": "repo_hf", |
| "shows": "Describes the future full-corpus Xperience-native pretraining goal, target modules, objectives, staged scale-up, hardware ranges, and evaluation protocol.", |
| }, |
| { |
| "id": "evidence_contract", |
| "title": "Evidence contract", |
| "path": "EVIDENCE_CONTRACT.md", |
| "kind": "project_scope", |
| "surface": "repo", |
| "shows": "Defines the implemented scope, setup-stage items, and multi-episode prerequisites.", |
| }, |
| { |
| "id": "project_packet", |
| "title": "Project packet", |
| "path": "docs/data/project_packet.json", |
| "kind": "project_path", |
| "surface": "website_hf", |
| "shows": "Gives a short project path with scope status and public surfaces.", |
| }, |
| { |
| "id": "artifact_guide", |
| "title": "Artifact guide", |
| "path": "ARTIFACT_GUIDE.md", |
| "kind": "project_path", |
| "surface": "repo_hf", |
| "shows": "Gives the human-readable map from project scope to data, tasks, platform mirrors, and scale-up status.", |
| }, |
| { |
| "id": "official_dataset_card_alignment", |
| "title": "Official Xperience-10M dataset-card alignment", |
| "path": "XPERIENCE10M_DATASET_CARD_ALIGNMENT.md", |
| "kind": "source_alignment", |
| "surface": "repo_hf", |
| "shows": "Aligns public dataset wording with the official gated Xperience-10M card, public sample card, HF API metadata, and current project coverage.", |
| }, |
| { |
| "id": "official_dataset_card_alignment_json", |
| "title": "Official Xperience-10M dataset-card alignment JSON", |
| "path": "docs/data/xperience10m_dataset_card_alignment.json", |
| "kind": "source_alignment", |
| "surface": "website_hf", |
| "shows": "Machine-readable upstream dataset-card, sample-card, and HF API alignment facts for website and HF mirrors.", |
| }, |
| { |
| "id": "source_alignment", |
| "title": "Source alignment", |
| "path": "SOURCE_ALIGNMENT_AUDIT.md", |
| "kind": "source_alignment", |
| "surface": "repo_hf", |
| "shows": "Summarizes the pass/fail check for full-dataset facts, sample-card facts, API-listing notes, and project coverage.", |
| }, |
| { |
| "id": "source_alignment_json", |
| "title": "Source alignment JSON", |
| "path": "docs/data/source_alignment_audit.json", |
| "kind": "source_alignment", |
| "surface": "website_hf", |
| "shows": "Machine-readable source-alignment pass/fail check for repo, website, and HF surfaces.", |
| }, |
| { |
| "id": "source_alignment_validator", |
| "title": "Source alignment validator", |
| "path": "scripts/validate_source_alignment.py", |
| "kind": "source_alignment", |
| "surface": "repo_hf", |
| "shows": "Regenerates the source-alignment report from committed facts and public card text.", |
| }, |
| { |
| "id": "hf_publisher", |
| "title": "Hugging Face publisher", |
| "path": "scripts/publish_hf_bundles.py", |
| "kind": "publication_workflow", |
| "surface": "repo_hf", |
| "shows": "Publishes prepared Space, artifact dataset, and model bundles, including an explicit model-binary upload batch.", |
| }, |
| { |
| "id": "github_package_dockerfile", |
| "title": "GitHub package Dockerfile", |
| "path": "Dockerfile", |
| "kind": "publication_workflow", |
| "surface": "repo", |
| "shows": "Builds the static-dashboard container package for GitHub Container Registry.", |
| }, |
| { |
| "id": "github_package_workflow", |
| "title": "GitHub package workflow", |
| "path": ".github/workflows/publish-ghcr.yml", |
| "kind": "publication_workflow", |
| "surface": "repo", |
| "shows": "Publishes the static-dashboard image to GitHub Container Registry on main or manual dispatch.", |
| }, |
| { |
| "id": "evaluation_protocol", |
| "title": "Evaluation protocol", |
| "path": "EVALUATION_PROTOCOL.md", |
| "kind": "evaluation_protocol", |
| "surface": "repo_hf", |
| "shows": "Defines the window unit, chronological split, task metrics, leakage controls, and current limitations.", |
| }, |
| { |
| "id": "evaluation_protocol_json", |
| "title": "Evaluation protocol JSON", |
| "path": "docs/data/evaluation_protocol.json", |
| "kind": "evaluation_protocol", |
| "surface": "website_hf", |
| "shows": "Machine-readable protocol generated from committed task metrics for website and HF mirrors.", |
| }, |
| { |
| "id": "evaluation_protocol_builder", |
| "title": "Evaluation protocol builder", |
| "path": "scripts/build_evaluation_protocol.py", |
| "kind": "evaluation_protocol", |
| "surface": "repo_hf", |
| "shows": "Regenerates the protocol from committed summary metrics and task artifacts.", |
| }, |
| { |
| "id": "task_suite_20", |
| "title": "Unified 20-task suite", |
| "path": "TASK_SUITE_20.md", |
| "kind": "evaluation_protocol", |
| "surface": "repo_hf", |
| "shows": "Reader-facing table for the single unified public-sample task suite under the same window, split, feature, and baseline contract.", |
| }, |
| { |
| "id": "task_suite_20_json", |
| "title": "Unified 20-task suite JSON", |
| "path": "docs/data/task_suite_20.json", |
| "kind": "website_data", |
| "surface": "website_hf", |
| "shows": "Machine-readable unified 20-task index for the website, Hugging Face mirrors, and live verification.", |
| }, |
| { |
| "id": "task_suite_20_builder", |
| "title": "Unified 20-task suite builder", |
| "path": "scripts/build_unified_task_suite.py", |
| "kind": "evaluation_protocol", |
| "surface": "repo_hf", |
| "shows": "Regenerates the unified 20-task JSON and Markdown from the public-sample metrics plus the historical provenance result bundle.", |
| }, |
| { |
| "id": "unified_task_model_radar_json", |
| "title": "Unified 20-task model radar JSON", |
| "path": "docs/data/unified_task_model_radar.json", |
| "kind": "website_data", |
| "surface": "website_hf", |
| "shows": "Stores normalized 20-axis radar values, raw task metrics, grouped chart-design metadata, Qwen3-Omni/Cosmos3 source mappings, method-card caveats, proxy flags, and source artifacts.", |
| }, |
| { |
| "id": "single_episode_task_model_radar_json", |
| "title": "Single-episode 20-task model radar JSON", |
| "path": "docs/data/single_episode_task_model_radar.json", |
| "kind": "website_data", |
| "surface": "website_hf", |
| "shows": "Machine-readable split radar for the one-episode Minimal and Neural MLP baselines, both scored on all 20 task contracts.", |
| }, |
| { |
| "id": "episode128_task_model_radar_json", |
| "title": "128-episode 20-task model radar JSON", |
| "path": "docs/data/episode128_task_model_radar.json", |
| "kind": "website_data", |
| "surface": "website_hf", |
| "shows": "Machine-readable split radar for selected 128-episode metadata/raw baselines, Qwen3-Omni v6, Cosmos3-Super, and Cosmos3-Nano, now complete at 140/140 scored rows with proxy notes retained.", |
| }, |
| { |
| "id": "task_method_20_result_matrix_json", |
| "title": "Task-method 20-result matrix JSON", |
| "path": "docs/data/task_method_20_result_matrix.json", |
| "kind": "website_data", |
| "surface": "website_hf", |
| "shows": "Machine-readable 9-method by 20-task matrix where every method has 20 records and the current release is complete at 180/180 scored rows.", |
| }, |
| { |
| "id": "task_method_20_result_matrix", |
| "title": "Task-method 20-result matrix", |
| "path": "TASK_METHOD_20_RESULT_MATRIX.md", |
| "kind": "evaluation_protocol", |
| "surface": "repo_hf", |
| "shows": "Reader-facing table that separates 20 records per method, direct numeric scores, documented compact-proxy scores, and source artifacts.", |
| }, |
| { |
| "id": "task_method_20_gap_audit_json", |
| "title": "Task-method 20-result gap audit JSON", |
| "path": "docs/data/task_method_20_gap_audit.json", |
| "kind": "website_data", |
| "surface": "website_hf", |
| "shows": "Machine-readable 180-record completion ledger with numeric scores, proxy flags, explicit status reasons, and source artifacts.", |
| }, |
| { |
| "id": "task_method_20_gap_audit", |
| "title": "Task-method 20-result gap audit", |
| "path": "TASK_METHOD_20_GAP_AUDIT.md", |
| "kind": "evaluation_protocol", |
| "surface": "repo_hf", |
| "shows": "Reader-facing ledger confirming 180/180 scored method-task cells and listing the six compact-proxy records separately.", |
| }, |
| { |
| "id": "task_method_20_source_audit_json", |
| "title": "Task-method 20-result source audit JSON", |
| "path": "docs/data/task_method_20_source_audit.json", |
| "kind": "website_data", |
| "surface": "website_hf", |
| "shows": "Machine-readable check that scored JSON-backed matrix cells match their declared metric source values.", |
| }, |
| { |
| "id": "task_method_20_source_audit", |
| "title": "Task-method 20-result source audit", |
| "path": "TASK_METHOD_20_SOURCE_AUDIT.md", |
| "kind": "evaluation_protocol", |
| "surface": "repo_hf", |
| "shows": "Reader-facing source-value audit for the 180-result matrix.", |
| }, |
| { |
| "id": "two_evidence_line_map_chart", |
| "title": "Two evidence-line map", |
| "path": "docs/assets/charts/two_evidence_line_map.svg", |
| "kind": "generated_figure", |
| "surface": "website_hf", |
| "shows": "Explains the public result organization: one sample-episode task-lab line, one selected-128 comparison line, and the combined 180/180 scored method-task ledger.", |
| }, |
| { |
| "id": "unified_task_model_radar_chart", |
| "title": "Unified 20-task model radar", |
| "path": "docs/assets/charts/unified_task_model_radar.svg", |
| "kind": "generated_figure", |
| "surface": "website_hf", |
| "shows": "Groups all nine methods into small-multiple 20-task radar panels so single-episode, 128-episode metadata/text, 128-episode raw-feature, and foundation-model rows remain readable.", |
| }, |
| { |
| "id": "single_episode_task_model_radar_chart", |
| "title": "Single-episode 20-task model radar", |
| "path": "docs/assets/charts/single_episode_task_model_radar.svg", |
| "kind": "generated_figure", |
| "surface": "website_hf", |
| "shows": "Shows the one-episode Minimal and Neural MLP 20/20 scored baselines in one enlarged radar panel with local legend and task key.", |
| }, |
| { |
| "id": "episode128_task_model_radar_chart", |
| "title": "128-episode 20-task model radar", |
| "path": "docs/assets/charts/episode128_task_model_radar.svg", |
| "kind": "generated_figure", |
| "surface": "website_hf", |
| "shows": "Separates selected 128-episode methods into metadata/text, raw-feature, and foundation-model radar panels with all 140 result rows scored and proxy notes retained.", |
| }, |
| { |
| "id": "unified_task_model_radar_builder", |
| "title": "Unified 20-task model radar builder", |
| "path": "scripts/build_unified_task_model_radar.py", |
| "kind": "visualization_builder", |
| "surface": "repo_hf", |
| "shows": "Regenerates grouped 20-task radar charts plus machine-readable metric, source, chart-design, and proxy metadata.", |
| }, |
| { |
| "id": "task_method_20_gap_audit_builder", |
| "title": "Task-method gap-audit builder", |
| "path": "scripts/build_task_method_20_gap_audit.py", |
| "kind": "publication_workflow", |
| "surface": "repo_hf", |
| "shows": "Regenerates the public completion/proxy audit from the 9-method by 20-task matrix without inventing unsupported scores.", |
| }, |
| { |
| "id": "task_method_20_source_audit_validator", |
| "title": "Task-method source-audit validator", |
| "path": "scripts/validate_task_method_matrix_sources.py", |
| "kind": "publication_workflow", |
| "surface": "repo_hf", |
| "shows": "Fails release checks if a scored matrix row disagrees with its JSON metric source.", |
| }, |
| { |
| "id": "all_task_model_scoring_waiter", |
| "title": "All-task model scoring guarded waiter", |
| "path": "scripts/omni/launch_all_task_model_scoring_when_free.sh", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Launches a user-provided all-task model scoring command only after enough private GPU capacity is idle, writing status logs under results/omni_finetune/deferred_launchers.", |
| }, |
| { |
| "id": "model_output_probe_readiness", |
| "title": "Model-output probe readiness", |
| "path": "results/omni_finetune/model_output_probe_readiness/model_output_probe_readiness.json", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Checks whether Qwen3-Omni and Cosmos3 runs have train, validation, and test prediction files before extending model overlays to all 20 task contracts.", |
| }, |
| { |
| "id": "model_output_probe_script", |
| "title": "Model-output probe readiness script", |
| "path": "scripts/omni/score_model_output_probes.py", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Audits model-output split availability and writes a readiness report without assigning new numeric task scores.", |
| }, |
| { |
| "id": "existing_model_output_task_probe", |
| "title": "Existing model-output task probe package", |
| "path": "results/omni_finetune/model_output_task_probes_20260616/summary.json", |
| "kind": "model_result", |
| "surface": "repo_hf", |
| "shows": "Scores task-specific Qwen3-Omni and Cosmos3 overlays only where verified held-out prediction JSON or compact target maps already contain the required targets.", |
| }, |
| { |
| "id": "existing_model_output_task_probe_script", |
| "title": "Existing model-output task probe scorer", |
| "path": "scripts/omni/score_existing_model_output_task_probes.py", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Derives task-specific scores from committed verified model outputs without running new inference or backfilling absent targets.", |
| }, |
| { |
| "id": "a100_128_metadata_task_baselines", |
| "title": "128-episode metadata task baselines", |
| "path": "results/omni_finetune/a100_128_metadata_task_baselines_20260616_v2/summary_report.json", |
| "kind": "model_result", |
| "surface": "repo_hf", |
| "shows": "Rerun of JSONL metadata/text simple and neural baselines over the selected 128-episode multiscale dataset; supports radar overlays on JSONL-supported task axes.", |
| }, |
| { |
| "id": "a100_128_raw20_task_baselines", |
| "title": "128-episode raw-feature 20-task baselines", |
| "path": "results/omni_finetune/a100_128_raw20_task_baselines_complete20_proxy_20260616T091500Z/run_summary_all.json", |
| "kind": "model_result", |
| "surface": "repo_hf", |
| "shows": "Rerun of simple and neural baselines over 34,269 windows and staged 4430-dimensional sensor NPZ features; covers 20 of 20 task axes, with interaction text and camera-view sync marked as compact-proxy completions because the 128 export lacks raw interaction strings and paired video-view embeddings.", |
| }, |
| { |
| "id": "research_takeaways", |
| "title": "Research takeaways", |
| "path": "RESEARCH_TAKEAWAYS.md", |
| "kind": "result_interpretation", |
| "surface": "repo_hf", |
| "shows": "Summarizes the main research lessons from committed metrics and identifies which experiments need held-out episodes.", |
| }, |
| { |
| "id": "research_takeaways_json", |
| "title": "Research takeaways JSON", |
| "path": "docs/data/research_takeaways.json", |
| "kind": "result_interpretation", |
| "surface": "website_hf", |
| "shows": "Machine-readable result interpretation for the website, HF cards, and mirror checks.", |
| }, |
| { |
| "id": "research_takeaways_builder", |
| "title": "Research takeaways builder", |
| "path": "scripts/build_research_takeaways.py", |
| "kind": "result_interpretation", |
| "surface": "repo_hf", |
| "shows": "Regenerates the research takeaways from committed summary metrics and task result artifacts.", |
| }, |
| { |
| "id": "audio_ablation_script", |
| "title": "Audio contribution script", |
| "path": "scripts/audio_ablation_and_raw_upgrade.py", |
| "kind": "result_interpretation", |
| "surface": "repo_hf", |
| "shows": "Measures audio contribution variants across the walkthrough-backed task contracts.", |
| }, |
| { |
| "id": "audio_ablation_summary", |
| "title": "Audio ablation summary", |
| "path": "results/audio_ablation/audio_ablation_summary.json", |
| "kind": "metrics_source", |
| "surface": "repo_hf", |
| "shows": "Stores per-task audio deltas for all current features, no-audio, audio-only, alternate-audio-only, replacement, and all-plus-alternate variants.", |
| }, |
| { |
| "id": "audio_ablation_summary_md", |
| "title": "Audio ablation summary report", |
| "path": "results/audio_ablation/AUDIO_ABLATION_SUMMARY.md", |
| "kind": "result_interpretation", |
| "surface": "repo_hf", |
| "shows": "Human-readable table showing the measured audio contribution and alternate-representation delta for every task.", |
| }, |
| { |
| "id": "audio_ablation_website_json", |
| "title": "Audio ablation website JSON", |
| "path": "docs/data/audio_ablation_summary.json", |
| "kind": "website_data", |
| "surface": "website_hf", |
| "shows": "Machine-readable audio ablation summary mirrored into the static website and Hugging Face bundles.", |
| }, |
| { |
| "id": "audio_ablation_delta_chart", |
| "title": "Audio ablation delta chart", |
| "path": "docs/assets/charts/audio_ablation_delta.svg", |
| "kind": "visual_evidence", |
| "surface": "website_hf", |
| "shows": "Bar chart of measured current-audio primary-metric deltas across the walkthrough-backed tasks.", |
| }, |
| { |
| "id": "figure_index", |
| "title": "Figure index", |
| "path": "FIGURE_INDEX.md", |
| "kind": "visual_evidence", |
| "surface": "repo_hf", |
| "shows": "Catalogs public figures, charts, modality thumbnails, dimensions, hashes, roles, and source scripts.", |
| }, |
| { |
| "id": "figure_index_json", |
| "title": "Figure index JSON", |
| "path": "docs/data/figure_index.json", |
| "kind": "visual_evidence", |
| "surface": "website_hf", |
| "shows": "Machine-readable visual asset index for website and Hugging Face mirrors.", |
| }, |
| { |
| "id": "figure_index_builder", |
| "title": "Figure index builder", |
| "path": "scripts/build_figure_index.py", |
| "kind": "visual_evidence", |
| "surface": "repo_hf", |
| "shows": "Regenerates visual-asset hashes, dimensions, and source-script provenance.", |
| }, |
| { |
| "id": "brand_assets_json", |
| "title": "Brand assets manifest", |
| "path": "docs/data/brand_assets.json", |
| "kind": "visual_evidence", |
| "surface": "website_hf", |
| "shows": "Machine-readable manifest for the generated logo system, favicon, social card, dimensions, hashes, and usage roles.", |
| }, |
| { |
| "id": "brand_logo_social_card", |
| "title": "Brand logo social card", |
| "path": "docs/assets/brand/xperience10m-logo-social-card.png", |
| "kind": "visual_evidence", |
| "surface": "website_hf", |
| "shows": "Provides the project logo card used in README, Hugging Face cards, and social previews.", |
| }, |
| { |
| "id": "brand_asset_builder", |
| "title": "Brand asset builder", |
| "path": "scripts/build_brand_assets.py", |
| "kind": "visual_evidence", |
| "surface": "repo_hf", |
| "shows": "Regenerates logo derivatives, favicon variants, app icons, and the Open Graph social card from the generated logo mark.", |
| }, |
| { |
| "id": "raw_sample_files_manifest", |
| "title": "Raw public sample file manifest", |
| "path": "docs/data/raw_sample_files.json", |
| "kind": "dataset_context", |
| "surface": "website_hf", |
| "shows": "Lists the official public sample HDF5, MP4, and RRD files, derived browser-preview clips, playback/download URLs, file sizes, browser behavior, and HDF5 group organization.", |
| }, |
| { |
| "id": "quality_gates", |
| "title": "Release checks", |
| "path": "QUALITY_GATES.md", |
| "kind": "quality_gate", |
| "surface": "repo_hf", |
| "shows": "Lists the automated and post-publish checks used to keep the release current.", |
| }, |
| { |
| "id": "quality_gate_manifest", |
| "title": "Release-check manifest", |
| "path": "docs/data/quality_gates.json", |
| "kind": "quality_gate", |
| "surface": "website_hf", |
| "shows": "Machine-readable release-check summary for validators, mirrors, and public project surfaces.", |
| }, |
| { |
| "id": "public_surface_qa", |
| "title": "Public project surface", |
| "path": "PUBLIC_SURFACE_QA.md", |
| "kind": "quality_gate", |
| "surface": "repo_hf", |
| "shows": "Keeps the repo, website, and Hugging Face cards aligned as one cohesive research project surface.", |
| }, |
| { |
| "id": "public_reader_map", |
| "title": "Public reader map", |
| "path": "PUBLIC_READER_MAP.md", |
| "kind": "project_path", |
| "surface": "repo_hf", |
| "shows": "Provides the first-pass navigation layer for GitHub, GitHub Pages, Hugging Face mirrors, Qwen3-Omni/Cosmos3 repos, evidence lines, and result-reading lanes.", |
| }, |
| { |
| "id": "public_reader_map_json", |
| "title": "Public reader map JSON", |
| "path": "docs/data/public_reader_map.json", |
| "kind": "project_path", |
| "surface": "website_hf", |
| "shows": "Machine-readable public reader map used by the website and Hugging Face mirrors to keep entry points and surface responsibilities explicit.", |
| }, |
| { |
| "id": "public_surface_qa_json", |
| "title": "Public project surface JSON", |
| "path": "docs/data/public_surface_qa.json", |
| "kind": "quality_gate", |
| "surface": "website_hf", |
| "volatile": True, |
| "shows": "Machine-readable report for SEO/social metadata, accessible tab semantics, public links, project links, and clear project presentation.", |
| }, |
| { |
| "id": "public_surface_qa_builder", |
| "title": "Public project surface builder", |
| "path": "scripts/build_public_surface_qa.py", |
| "kind": "quality_gate", |
| "surface": "repo_hf", |
| "shows": "Regenerates the public presentation report before release.", |
| }, |
| { |
| "id": "task_surface_integrity", |
| "title": "Task-surface integrity report", |
| "path": "docs/data/task_surface_integrity.json", |
| "kind": "quality_gate", |
| "surface": "website_hf", |
| "volatile": True, |
| "shows": "Confirms the public original-task cards use human-readable research names, representative modality thumbnails, and the interactive walkthrough/player JSON contract.", |
| }, |
| { |
| "id": "rendered_site_check", |
| "title": "Rendered website check", |
| "path": "RENDERED_SITE_CHECK.md", |
| "kind": "quality_gate", |
| "surface": "repo_hf", |
| "volatile": True, |
| "shows": "Records the latest browser-level load, tab, walkthrough deep-link, control-click, and console-health check.", |
| }, |
| { |
| "id": "rendered_site_check_json", |
| "title": "Rendered website check JSON", |
| "path": "docs/data/rendered_site_check.json", |
| "kind": "quality_gate", |
| "surface": "website_hf", |
| "volatile": True, |
| "shows": "Machine-readable browser-level website check for the public static site.", |
| }, |
| { |
| "id": "rendered_site_check_builder", |
| "title": "Rendered website check builder", |
| "path": "scripts/build_rendered_site_check.py", |
| "kind": "quality_gate", |
| "surface": "repo_hf", |
| "shows": "Builds the rendered website check from browser observations.", |
| }, |
| { |
| "id": "task_surface_validator", |
| "title": "Task-surface integrity validator", |
| "path": "scripts/validate_task_surface.py", |
| "kind": "quality_gate", |
| "surface": "repo_hf", |
| "shows": "Regenerates the task-surface integrity report and fails if task cards expose raw artifact ids or lose the interactive player wiring.", |
| }, |
| { |
| "id": "live_publication_status", |
| "title": "Live publication status", |
| "path": "docs/data/live_publication_status.json", |
| "kind": "quality_gate", |
| "surface": "website_hf", |
| "volatile": True, |
| "shows": "Records the last live GitHub/HF URL verification after upload.", |
| }, |
| { |
| "id": "live_publication_verifier", |
| "title": "Live publication verifier", |
| "path": "scripts/verify_live_publication.py", |
| "kind": "quality_gate", |
| "surface": "repo", |
| "shows": "Fetches the published GitHub/HF URLs and compares live hashes and public-card markers against the release assets.", |
| }, |
| { |
| "id": "reproducibility_contract", |
| "title": "Reproducibility contract", |
| "path": "REPRODUCIBILITY.md", |
| "kind": "reproducibility", |
| "surface": "repo_hf", |
| "shows": "Defines public reproduction commands, expected outputs, and non-reproducible scale-up boundaries.", |
| }, |
| { |
| "id": "reproducibility_matrix", |
| "title": "Reproducibility matrix", |
| "path": "docs/data/reproducibility_matrix.json", |
| "kind": "reproducibility", |
| "surface": "website_hf", |
| "shows": "Machine-readable reproduction steps with expected artifacts and public boundaries.", |
| }, |
| { |
| "id": "artifact_index_builder", |
| "title": "Artifact index builder", |
| "path": "scripts/build_artifact_index.py", |
| "kind": "project_path", |
| "surface": "repo_hf", |
| "shows": "Generates the selective artifact catalog from local files.", |
| }, |
| { |
| "id": "publication_audit", |
| "title": "Public bundle contents", |
| "path": "docs/data/publication_audit.json", |
| "kind": "publication_package_check", |
| "surface": "website_hf", |
| "volatile": True, |
| "shows": "Confirms public bundles exclude raw data, caches, heavy archives, and credential text.", |
| }, |
| { |
| "id": "scale_up_status_check", |
| "title": "Multi-episode pilot status", |
| "path": "docs/data/scope_claims_audit.json", |
| "kind": "scale_up_status", |
| "surface": "website_hf", |
| "volatile": True, |
| "shows": "Separates setup paths from completed held-out-episode results.", |
| }, |
| { |
| "id": "mirror_parity", |
| "title": "Prepared mirror parity report", |
| "path": "docs/data/mirror_parity.json", |
| "kind": "mirror_parity", |
| "surface": "website_hf", |
| "volatile": True, |
| "shows": "Confirms prepared GitHub/HF Space/artifact/model mirrors share the same critical data, figure, website HTML, and validator files.", |
| }, |
| { |
| "id": "website_integrity", |
| "title": "Website integrity report", |
| "path": "docs/data/website_integrity.json", |
| "kind": "integrity_report", |
| "surface": "website_hf", |
| "volatile": True, |
| "shows": "Confirms local website links, anchors, JSON data files, and referenced images resolve.", |
| }, |
| { |
| "id": "project_manifest", |
| "title": "Project manifest", |
| "path": "docs/data/project_manifest.json", |
| "kind": "metadata", |
| "surface": "website_hf", |
| "shows": "Lists public URLs, upstream sources, and machine-readable project metadata.", |
| }, |
| { |
| "id": "task_summary", |
| "title": "Original task summary report", |
| "path": "results/episode_task_suite/summary_report.json", |
| "kind": "metrics_source", |
| "surface": "repo_hf", |
| "shows": "Stores the task definitions, splits, feature dimension, and minimal/neural metrics.", |
| }, |
| { |
| "id": "website_metrics_bundle", |
| "title": "Website metrics bundle", |
| "path": "docs/data/summary_metrics.json", |
| "kind": "website_data", |
| "surface": "website_hf", |
| "shows": "Mirrors task metrics for the static dashboard.", |
| }, |
| { |
| "id": "feature_manifest", |
| "title": "Feature manifest", |
| "path": "results/episode_task_suite/feature_manifest.json", |
| "kind": "data_contract", |
| "surface": "repo_hf", |
| "shows": "Maps the current window vector back to source feature blocks.", |
| }, |
| { |
| "id": "available_modalities", |
| "title": "Available modalities", |
| "path": "results/episode_task_suite/available_modalities.json", |
| "kind": "data_contract", |
| "surface": "repo_hf", |
| "shows": "Documents which sample modalities entered the current extracted feature contract.", |
| }, |
| { |
| "id": "windows_table", |
| "title": "Aligned windows table", |
| "path": "results/episode_task_suite/windows.csv", |
| "kind": "data_contract", |
| "surface": "repo_hf", |
| "shows": "Lists the 1,161 aligned windows and their frame/action/subtask labels.", |
| }, |
| { |
| "id": "neural_mlp_directory", |
| "title": "Neural MLP task-head results", |
| "path": "results/episode_task_suite/neural_mlp", |
| "kind": "result_directory", |
| "surface": "repo_hf_model", |
| "shows": "Stores matching PyTorch MLP results for the walkthrough-backed task contracts.", |
| }, |
| { |
| "id": "research_direction_taxonomy", |
| "title": "Research direction taxonomy", |
| "path": "results/episode_task_suite/research_directions/research_direction_taxonomy.json", |
| "kind": "taxonomy", |
| "surface": "repo_hf", |
| "shows": "Maps the walkthrough-backed tasks to the four Ropedia research directions as direct/proxy/diagnostic.", |
| }, |
| { |
| "id": "research_direction_extensions", |
| "title": "Research direction extension probes", |
| "path": "results/episode_task_suite/research_direction_extensions/research_direction_extension_results.json", |
| "kind": "metrics_source", |
| "surface": "repo_hf", |
| "shows": "Stores one coded extension probe per research direction with minimal and neural metrics.", |
| }, |
| { |
| "id": "tier2_task_suite", |
| "title": "Unified 20-task provenance bundle", |
| "path": "results/episode_task_suite/tier2_task_suite/tier2_task_suite_results.json", |
| "kind": "metrics_source", |
| "surface": "repo_hf", |
| "shows": "Stores the historical result bundle for provenance rows with minimal and neural baselines aligned to the same 20-task window/split setup.", |
| }, |
| { |
| "id": "tier2_task_suite_json", |
| "title": "Unified 20-task provenance JSON", |
| "path": "docs/data/tier2_task_suite.json", |
| "kind": "website_data", |
| "surface": "website_hf", |
| "shows": "Machine-readable provenance definitions, setup alignment, metrics, and public source paths; the file name is historical.", |
| }, |
| { |
| "id": "tier2_task_suite_chart", |
| "title": "Unified 20-task provenance chart", |
| "path": "docs/assets/charts/tier2_task_suite.svg", |
| "kind": "generated_figure", |
| "surface": "website_hf", |
| "shows": "Visual summary of the historical provenance baseline metrics inside the unified 20-task suite.", |
| }, |
| { |
| "id": "tier2_task_suite_builder", |
| "title": "Unified 20-task provenance builder", |
| "path": "scripts/tier2_task_suite.py", |
| "kind": "evaluation_protocol", |
| "surface": "repo_hf", |
| "shows": "Regenerates the historical provenance rows from shared windows plus the local public-sample annotation HDF5; the script name is historical.", |
| }, |
| { |
| "id": "task_walkthroughs", |
| "title": "Task walkthroughs", |
| "path": "results/episode_task_suite/task_walkthroughs/TASK_WALKTHROUGHS.md", |
| "kind": "onboarding_doc", |
| "surface": "repo_hf", |
| "shows": "Explains every task with case study, input, process modules, output, and limitation.", |
| }, |
| { |
| "id": "task_suite_infographic", |
| "title": "Original task-suite infographic", |
| "path": "docs/assets/task_suite_infographic.png", |
| "kind": "generated_figure", |
| "surface": "website_hf", |
| "shows": "Presents the task suite and sample modality thumbnails with metrics generated from committed files.", |
| }, |
| { |
| "id": "modality_atlas", |
| "title": "Responsive modality atlas", |
| "path": "docs/data/modality_atlas.json", |
| "kind": "website_data", |
| "surface": "website_hf", |
| "shows": "Documents the seven public-sample modality cards and their derived thumbnail assets.", |
| }, |
| { |
| "id": "modality_thumbnails", |
| "title": "Standalone modality thumbnails", |
| "path": "docs/assets/modalities", |
| "kind": "generated_figure_assets", |
| "surface": "website_hf", |
| "shows": "Stores small derived thumbnails for readable website modality cards without raw data redistribution.", |
| }, |
| { |
| "id": "pipeline_figure", |
| "title": "Pipeline figure", |
| "path": "docs/assets/pipeline_diagram.png", |
| "kind": "generated_figure", |
| "surface": "website_hf", |
| "shows": "Shows the raw-episode to artifact pipeline with verified labels.", |
| }, |
| { |
| "id": "architecture_figure", |
| "title": "Architecture figure", |
| "path": "docs/assets/task_architectures.png", |
| "kind": "generated_figure", |
| "surface": "website_hf", |
| "shows": "Shows the shared feature pipeline and minimal/neural head families.", |
| }, |
| { |
| "id": "qwen_data_access_status", |
| "title": "Qwen3-Omni data access status", |
| "path": "results/omni_finetune/DATA_ACCESS_STATUS.md", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Summarizes the data-readiness checks required before a held-out Qwen3-Omni pilot can report metrics.", |
| }, |
| { |
| "id": "qwen3_lora_hf_upload_note", |
| "title": "Qwen3 LoRA HF upload note", |
| "path": "results/omni_finetune/HF_UPLOAD.md", |
| "kind": "publication_workflow", |
| "surface": "repo_hf", |
| "shows": "Documents the final 128-episode LoRA adapter upload path, target model repo, package builder, and forbidden files.", |
| }, |
| { |
| "id": "multi_episode_access_status", |
| "title": "Multi-episode access status", |
| "path": "results/omni_finetune/MULTI_EPISODE_ACCESS_STATUS.md", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Documents the public multi-episode access status and 32-episode pilot selection.", |
| }, |
| { |
| "id": "qwen3_omni_error_analysis_report", |
| "title": "Qwen3-Omni held-out error-analysis report", |
| "path": "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_96train_16val_16test_valmon_20260605_eval/analysis/ERROR_ANALYSIS.md", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Summarizes the earlier validation-aware Qwen3-Omni held-out failures by episode, action family, train-seen status, required-modality state, and object category.", |
| }, |
| { |
| "id": "qwen3_omni_error_analysis_json", |
| "title": "Qwen3-Omni held-out error-analysis JSON", |
| "path": "results/omni_finetune/verified_public/xperience10m_qwen3_omni_128ep_96train_16val_16test_valmon_20260605_eval/analysis/error_analysis_summary.json", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Machine-readable Qwen3-Omni held-out error analysis with grouped metrics and sanitized failure examples.", |
| }, |
| { |
| "id": "multi_episode_128_baseline_report", |
| "title": "128-episode aligned baseline report", |
| "path": "results/omni_finetune/multi_episode_128_task_baselines/BASELINE_ALIGNMENT_REPORT.md", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Summarizes same-split simple and neural metadata baselines for the walkthrough-backed task ids, with unsupported markers for tasks that need missing raw 128 feature blocks.", |
| }, |
| { |
| "id": "multi_episode_128_baseline_summary", |
| "title": "128-episode aligned baseline summary", |
| "path": "results/omni_finetune/multi_episode_128_task_baselines/summary_report.json", |
| "kind": "metrics_source", |
| "surface": "repo_hf", |
| "shows": "Machine-readable 96/16/16 split counts, run configuration, per-task simple metrics, neural metrics, and raw-feature unsupported statuses.", |
| }, |
| { |
| "id": "omni_model_comparison_report", |
| "title": "Omni model comparison report", |
| "path": "results/omni_finetune/OMNI_MODEL_COMPARISON.md", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Reader-facing comparison of the single-episode task suite, 128-episode aligned baselines, Qwen3-Omni packages, and Cosmos3 future-window branch.", |
| }, |
| { |
| "id": "omni_model_comparison_json", |
| "title": "Omni model comparison JSON", |
| "path": "docs/data/omni_model_comparison.json", |
| "kind": "metrics_source", |
| "surface": "repo_hf", |
| "shows": "Machine-readable comparison of the current result versions, per-task aligned baselines, verified Qwen3 packages, and Cosmos3 package.", |
| }, |
| { |
| "id": "cosmos3_nano_verified_summary", |
| "title": "Cosmos3-Nano verified package summary", |
| "path": "results/omni_finetune/verified_public/xperience10m_cosmos3_nano_128ep_future_window_h5_compat_adapter_eval_test_full/verified_result_summary.json", |
| "kind": "metrics_source", |
| "surface": "repo_hf", |
| "shows": "Machine-readable verified public summary for the Cosmos3-Nano future-window compatibility package.", |
| }, |
| { |
| "id": "cosmos3_nano_run_report", |
| "title": "Cosmos3-Nano future-window run report", |
| "path": "results/omni_finetune/verified_public/xperience10m_cosmos3_nano_128ep_future_window_h5_compat_adapter_eval_test_full/eval/RUN_REPORT.md", |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Reader-facing held-out metrics and interpretation for the Cosmos3-Nano future-window compatibility branch.", |
| }, |
| { |
| "id": "citation", |
| "title": "Citation metadata", |
| "path": "CITATION.cff", |
| "kind": "citation", |
| "surface": "repo_hf", |
| "shows": "Makes the project externally citable.", |
| }, |
| { |
| "id": "license", |
| "title": "License and data terms", |
| "path": "LICENSE", |
| "kind": "license", |
| "surface": "repo_hf", |
| "shows": "Separates MIT-scoped code from original Xperience-10M data terms.", |
| }, |
| ] |
|
|
|
|
| def sha256(path: Path) -> str: |
| digest = hashlib.sha256() |
| with path.open("rb") as handle: |
| for chunk in iter(lambda: handle.read(1024 * 1024), b""): |
| digest.update(chunk) |
| return digest.hexdigest() |
|
|
|
|
| def directory_stats(path: Path) -> dict: |
| files = [item for item in path.rglob("*") if item.is_file()] |
| return { |
| "file_count": len(files), |
| "bytes": sum(item.stat().st_size for item in files), |
| } |
|
|
|
|
| def verified_public_package_artifacts() -> list[dict]: |
| verified_root = ROOT / "results/omni_finetune/verified_public" |
| if not verified_root.exists(): |
| return [] |
|
|
| artifacts: list[dict] = [] |
| for summary_path in sorted(verified_root.glob("*/verified_result_summary.json")): |
| package_dir = summary_path.parent |
| slug = package_dir.name |
| payload = json.loads(summary_path.read_text(encoding="utf-8")) |
| title = payload.get("backbone_display_name") or payload.get("eval_run_id") or slug |
| backbone = payload.get("backbone", "unknown_backbone") |
| status = payload.get("status", "unknown") |
| eval_run_id = payload.get("eval_run_id", slug) |
| artifacts.append( |
| { |
| "id": f"verified_public_package_{slug}", |
| "title": f"Verified public package: {title}", |
| "path": package_dir.relative_to(ROOT).as_posix(), |
| "kind": "verified_public_package", |
| "surface": "repo_hf", |
| "shows": ( |
| f"Public-safe verified package for {eval_run_id} " |
| f"({backbone}, status={status})." |
| ), |
| } |
| ) |
| artifacts.append( |
| { |
| "id": f"verified_public_summary_{slug}", |
| "title": f"Verified summary: {title}", |
| "path": summary_path.relative_to(ROOT).as_posix(), |
| "kind": "metrics_source", |
| "surface": "repo_hf", |
| "shows": f"Machine-readable verified summary for {eval_run_id}.", |
| } |
| ) |
| for relative_path, kind, label in [ |
| ("PUBLIC_RESULT_SUMMARY.md", "scaleup_status", "public result summary"), |
| ("eval/RUN_REPORT.md", "scaleup_status", "run report"), |
| ("eval/metrics.json", "metrics_source", "metrics JSON"), |
| ("package_audit.json", "publication_audit", "package audit"), |
| ]: |
| path = package_dir / relative_path |
| if not path.exists(): |
| continue |
| safe_label = label.replace(" ", "_") |
| artifacts.append( |
| { |
| "id": f"verified_public_{safe_label}_{slug}", |
| "title": f"Verified {label}: {title}", |
| "path": path.relative_to(ROOT).as_posix(), |
| "kind": kind, |
| "surface": "repo_hf", |
| "shows": f"{label.capitalize()} for {eval_run_id}.", |
| } |
| ) |
| return artifacts |
|
|
|
|
| def qwen3_future_task_probe_artifacts() -> list[dict]: |
| run_dir = ROOT / "results/omni_finetune" / QWEN3_FUTURE_TASK_PROBE_RUN_ID |
| if not run_dir.exists(): |
| return [] |
|
|
| artifacts: list[dict] = [ |
| { |
| "id": "qwen3_future_task_probe_package", |
| "title": "Qwen3 v6 future-task probe package", |
| "path": run_dir.relative_to(ROOT).as_posix(), |
| "kind": "model_result", |
| "surface": "repo_hf", |
| "shows": ( |
| "Two-shard Qwen3-Omni v6 inference probe for tasks 13, 14, " |
| "and 17, with public-safe metrics, predictions, progress logs, " |
| "and merge report." |
| ), |
| } |
| ] |
| for relative_path, kind, label in [ |
| ("summary.json", "metrics_source", "merged summary"), |
| ("collection_validation.json", "publication_audit", "collection validation"), |
| ("RUN_REPORT.md", "scaleup_status", "run report"), |
| ("long_horizon_next_action/metrics.json", "metrics_source", "task 13 metrics"), |
| ("next_subtask_forecast/metrics.json", "metrics_source", "task 14 metrics"), |
| ("object_set_forecast/metrics.json", "metrics_source", "task 17 metrics"), |
| ]: |
| path = run_dir / relative_path |
| if path.exists(): |
| artifacts.append( |
| { |
| "id": f"qwen3_future_task_probe_{relative_path.replace('/', '_').replace('.', '_')}", |
| "title": f"Qwen3 future-task probe {label}", |
| "path": path.relative_to(ROOT).as_posix(), |
| "kind": kind, |
| "surface": "repo_hf", |
| "shows": f"Public-safe {label} for {QWEN3_FUTURE_TASK_PROBE_RUN_ID}.", |
| } |
| ) |
|
|
| launcher_log = ( |
| ROOT |
| / "results/omni_finetune/deferred_launchers" |
| / f"{QWEN3_FUTURE_TASK_PROBE_RUN_ID}.launcher.log" |
| ) |
| if launcher_log.exists(): |
| artifacts.append( |
| { |
| "id": "qwen3_future_task_probe_launcher_log", |
| "title": "Qwen3 future-task probe launcher log", |
| "path": launcher_log.relative_to(ROOT).as_posix(), |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Launch and merge log for the two-shard future-task probe.", |
| } |
| ) |
| return artifacts |
|
|
|
|
| def cosmos3_super_interaction_text_probe_artifacts() -> list[dict]: |
| run_dir = ROOT / "results/omni_finetune" / COSMOS3_SUPER_INTERACTION_TEXT_TASK_PROBE_RUN_ID |
| if not run_dir.exists(): |
| return [] |
|
|
| artifacts: list[dict] = [ |
| { |
| "id": "cosmos3_super_interaction_text_probe_package", |
| "title": "Cosmos3-Super interaction-text task-15 probe package", |
| "path": run_dir.relative_to(ROOT).as_posix(), |
| "kind": "model_result", |
| "surface": "repo_hf", |
| "shows": ( |
| "Four-shard Cosmos3-Super text-only inference probe for task 15 " |
| "over raw annotation.hdf5 interaction text labels." |
| ), |
| } |
| ] |
| for relative_path, kind, label in [ |
| ("summary.json", "metrics_source", "merged summary"), |
| ("launch_env.txt", "scaleup_status", "launch environment"), |
| ("interaction_text_prediction/RUN_REPORT.md", "scaleup_status", "task 15 run report"), |
| ("interaction_text_prediction/metrics.json", "metrics_source", "task 15 metrics"), |
| ("interaction_text_prediction/per_class_metrics.csv", "metrics_source", "task 15 per-class metrics"), |
| ("interaction_text_prediction/confusion_matrix.csv", "metrics_source", "task 15 confusion matrix"), |
| ]: |
| path = run_dir / relative_path |
| if path.exists(): |
| artifacts.append( |
| { |
| "id": f"cosmos3_super_interaction_text_{relative_path.replace('/', '_').replace('.', '_')}", |
| "title": f"Cosmos3-Super interaction-text probe {label}", |
| "path": path.relative_to(ROOT).as_posix(), |
| "kind": kind, |
| "surface": "repo_hf", |
| "shows": f"Public-safe {label} for {COSMOS3_SUPER_INTERACTION_TEXT_TASK_PROBE_RUN_ID}.", |
| } |
| ) |
|
|
| launcher_log = ( |
| ROOT |
| / "results/omni_finetune/deferred_launchers" |
| / f"{COSMOS3_SUPER_INTERACTION_TEXT_TASK_PROBE_RUN_ID}.launcher.log" |
| ) |
| if launcher_log.exists(): |
| artifacts.append( |
| { |
| "id": "cosmos3_super_interaction_text_probe_launcher_log", |
| "title": "Cosmos3-Super interaction-text probe launcher log", |
| "path": launcher_log.relative_to(ROOT).as_posix(), |
| "kind": "scaleup_status", |
| "surface": "repo_hf", |
| "shows": "Launch and merge log for the four-shard task-15 probe.", |
| } |
| ) |
| return artifacts |
|
|
|
|
| def artifact_entry(item: dict) -> dict: |
| path = ROOT / item["path"] |
| entry = { |
| **item, |
| "exists": path.exists(), |
| } |
| if path.is_file(): |
| entry["bytes"] = path.stat().st_size |
| if item.get("volatile"): |
| entry["hash_policy"] = "existence_and_size_only" |
| else: |
| entry["sha256"] = sha256(path) |
| elif path.is_dir(): |
| entry.update(directory_stats(path)) |
| else: |
| entry.update({"bytes": 0}) |
| return entry |
|
|
|
|
| def main() -> int: |
| artifacts = [dict(item) for item in ARTIFACTS] |
| artifacts.extend(verified_public_package_artifacts()) |
| artifacts.extend(qwen3_future_task_probe_artifacts()) |
| artifacts.extend(cosmos3_super_interaction_text_probe_artifacts()) |
| summary_path = ROOT / "results/episode_task_suite/summary_report.json" |
| if summary_path.exists(): |
| summary = json.loads(summary_path.read_text(encoding="utf-8")) |
| feature_dim = int(summary.get("feature_dim", 0)) |
| for item in artifacts: |
| if item["id"] == "feature_manifest" and feature_dim: |
| item["shows"] = f"Maps the {feature_dim:,}-dimensional window vector back to source feature blocks." |
| entries = [artifact_entry(item) for item in artifacts] |
| missing = [entry["path"] for entry in entries if not entry["exists"]] |
| by_kind: dict[str, int] = {} |
| for entry in entries: |
| by_kind[entry["kind"]] = by_kind.get(entry["kind"], 0) + 1 |
|
|
| report = { |
| "title": "Ropedia Xperience-10M Task Suite Artifact Index", |
| "generated_at_utc": datetime.now(timezone.utc).isoformat(timespec="seconds"), |
| "status": "pass" if not missing else "fail", |
| "artifact_count": len(entries), |
| "missing": missing, |
| "by_kind": by_kind, |
| "artifacts": entries, |
| } |
| OUTPUT.parent.mkdir(parents=True, exist_ok=True) |
| OUTPUT.write_text(json.dumps(report, indent=2) + "\n", encoding="utf-8") |
| print(f"{report['status'].upper()}: wrote {OUTPUT}") |
| if missing: |
| for path in missing: |
| print(f"- missing: {path}") |
| return 1 |
| return 0 |
|
|
|
|
| if __name__ == "__main__": |
| raise SystemExit(main()) |
|
|
