Downloads · 30 days
0
smolvla-grpo-libero-object
MorpheusTzz/smolvla-grpo-libero-object
smolvla-grpo-libero-object is a robotics model from MorpheusTzz. Use it for the robotics task on the model card, and read the license before you ship it in a product. It is set up for lerobot. The card lists the license as mit.
Checkpoints from FlowVLA-RL: critic-free GRPO on SmolVLA's flow-matching action expert, made possible by a marginal-preserving ODE-to-SDE retrofit that gives every denoising step a closed-form log-probability.
Access
Public
Updated Aug 5, 2026
Repo size
2.2 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors1.8 GB · 100%
At a glance
- Task
- Robotics
- Library
- lerobot
- License
- mit
- Access
- Public
- Created
- Aug 5, 2026
- Updated
- Aug 5, 2026
- SHA
- a120aa9b
Base models
Model card
From the Hugging Face model README
SmolVLA + online GRPO on LIBERO-Object
Checkpoints from FlowVLA-RL: critic-free GRPO on SmolVLA's flow-matching action expert, made possible by a marginal-preserving ODE-to-SDE retrofit that gives every denoising step a closed-form log-probability.
| folder | what | LIBERO-Object success (n=200, deterministic ODE) |
|---|---|---|
sft-100pct-baseline/ | SFT on the full LIBERO dataset (25k steps, bs 64) | 58.5% |
grpo-seed11-update300/ | the above + 300 GRPO updates (~2000 sim episodes) | 68.5% |
68.5% is this seed's peak. Across the seed family the fixed-budget mean is 64.8% (+6.3 pp, 4/4 seeds improve); protocol, statistics and caveats are in the repo's REPORT.md.
Usage
Checkpoints are in lerobot v0.6.0 SmolVLA format (3 camera slots; LIBERO provides 2, the third is an empty camera).
hf download MorpheusTzz/smolvla-grpo-libero-object --local-dir ckpts
# evaluate with the FlowVLA-RL harness
git clone https://github.com/BlackMirean/FlowVLA-RL
python scripts/run_baseline_eval.py --checkpoint ckpts/grpo-seed11-update300 \
--suite libero_object --episodes 20 --batch-size 5 --seed 1000
One thing to know before evaluating: lerobot v0.6.0 computes _max_episode_steps for LIBERO but
never applies it, so the caller must enforce the step caps (280/280/300 per suite) or success
rates come out inflated. The harness above does this; details in the repo's
eval protocol.
MIT license, covering the checkpoints and this card. Base model: lerobot/smolvla_base.
- Task
- Robotics
- Library
- lerobot
- License
- mit
- Created
- Aug 5, 2026
- Updated
- Aug 5, 2026