Downloads · 30 days
0
retarget-legacy-checkpoints
myx160/retarget-legacy-checkpoints
retarget-legacy-checkpoints is a reinforcement learning model from myx160. Use it for the reinforcement learning task on the model card, and read the license before you ship it in a product.
Original contents of the project's logs/latesttrain archive, uploaded on 2026-09-16 without converting the files. The latesttrain/ directory preserves historical run names, .pt checkpoints, configuration JSON files, a…
Access
Public
Updated Sep 16, 2026
Repo size
392 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.0294 MB · 75%
At a glance
- Task
- Reinforcement Learning
- Access
- Public
- Created
- Sep 16, 2026
- Updated
- Sep 16, 2026
- SHA
- ceaad711
Model card
From the Hugging Face model README
DA legacy training checkpoints
Original contents of the project's logs/latest_train archive, uploaded on 2026-09-16 without converting the files. The latest_train/ directory preserves historical run names, .pt checkpoints, configuration JSON files, and TensorBoard events.
This is the legacy training archive, not the A800 runs currently being resumed in September 2026. Legacy checkpoints and metrics must not be assumed to use the same data version or evaluation protocol as newer runs. Consult each run's saved configuration before loading it. These custom PyTorch policies require the project's DA/legged_gym environment and are not Transformers models. Only load pickle-based checkpoints from trusted sources.
Related data stages: https://huggingface.co/datasets/myx160/retarget-da-dataset
The uploader authorized this public release; no additional blanket license over third-party material is asserted here.
- Task
- Reinforcement Learning
- Created
- Sep 16, 2026
- Updated
- Sep 16, 2026