|
Download README.md from sadjava/latent-dynamics-libero90-s1000-shuffle: direct link, hf CLI and curl.
- Browser
- Download file 857 Bytes
-
https://huggingface.co/sadjava/latent-dynamics-libero90-s1000-shuffle/resolve/main/README.md
- Command line
-
hf download hf://sadjava/latent-dynamics-libero90-s1000-shuffle/README.md
-
curl -L -o README.md https://huggingface.co/sadjava/latent-dynamics-libero90-s1000-shuffle/resolve/main/README.md
857 Bytes
metadata
tags:
- libero
- latent-dynamics
- action-free
- robotics
- shuffled-future
- control
Latent dynamics (IDM/FDM) · libero_90 · seed 1000 · shuffled-future control
Action-free visual dynamics pretrain: frozen SmolVLA vision features → IDM → FDM.
Trained on RGB from nvidia/libero_90 only (no action / state / reward labels).
Control run: trained with shuffle_future=True — the future RGB frame is sampled from a different episode, so temporal dynamics are broken. Same architecture / steps / data volume as the real IDM; use as a negative control for video-latent fine-tuning (VIDEO_CKPT=.../idm_fdm.pt from this repo).
Files:
idm_fdm.pt— model weights + config dictidm_fdm.json— training hyperparameters
Used as frozen teacher for video-latent few-shot SmolVLA fine-tuning
(--video.checkpoint).