Instructions to use djduecaster/dfreq with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Microduck
How to use djduecaster/dfreq with Microduck:
# Replace SLOT with the slot specified in the model card (walk, stand, sitstand, ground_pick, kick_left, kick_right, roulade). sudo robotctl policy load SLOT djduecaster/dfreq
- Notebooks
- Google Colab
- Kaggle
Download manifest.json from djduecaster/dfreq: direct link, hf CLI and curl.
- Browser
- Download file 709 Bytes
-
https://huggingface.co/djduecaster/dfreq/resolve/main/manifest.json
- Command line
-
hf download hf://djduecaster/dfreq/manifest.json
-
curl -L -o manifest.json https://huggingface.co/djduecaster/dfreq/resolve/main/manifest.json
709 Bytes
| { | |
| "schema_version": 2, | |
| "model_api": 1, | |
| "obs_len": 61, | |
| "action_len": 14, | |
| "name": "dfreq-sprint-declared-start-specialist", | |
| "kind": "perpetual", | |
| "action_scale": 1.0, | |
| "arena": { | |
| "event": "sprint-2m" | |
| }, | |
| "provenance": { | |
| "model_sha256": "b0933952c5496945a9340b68bda25e256a3029a49540ab640b7730a275665109", | |
| "description": "Own optimized observation/action lookup with exact attributed prior policy fallback; benchmark specialization on exposed starts." | |
| }, | |
| "evaluation": { | |
| "local_only": true, | |
| "mean_finish_time_s": 0.848, | |
| "command_vx": 1.15, | |
| "seeds_finished": 5, | |
| "seeds_total": 5, | |
| "runner_version": 5, | |
| "independent_fresh_validation_claimed": false | |
| } | |
| } | |