Downloads · 30 days
0
native-fb-vla-libero-10k-s1000
kunhsiang/native-fb-vla-libero-10k-s1000
native-fb-vla-libero-10k-s1000 is a machine learning model from kunhsiang. Use it for the machine learning 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.
備份自 ID Lab lerobot-vla-fbagent 專案，含完整中間 checkpoint。
Access
Public
Updated Jun 20, 2026
Repo size
2 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors2 GB · 100%
At a glance
- Library
- lerobot
- License
- mit
- Access
- Public
- Created
- Jun 19, 2026
- Updated
- Jun 20, 2026
- SHA
- 31f0ebf4
Model card
From the Hugging Face model README
native-fb-vla-libero-10k-s1000
備份自 ID Lab lerobot-vla-fbagent 專案，含完整中間 checkpoint。
訓練資訊
| 項目 | 值 |
|---|---|
| Policy Type | native_fb_vla |
| Dataset | HuggingFaceVLA/libero |
| Max Steps | 10,000 |
| Checkpoints | 10 個 |
| Seed | ? |
| Batch Size | ? |
| Learning Rate | ? |
| Num Workers | ? |
| Chunk Size | 100 |
| Git Commit | 4839863 |
| Source Dir | native_fb_vla_libero_10k_s1000 |
復現指令
git clone --recurse-submodules [email protected]:KunHsiang/lerobot-vla-fbagent.git
cd lerobot-vla-fbagent
git checkout 4839863
uv sync
HF_HUB_ENABLE_HF_TRANSFER=1 .venv/bin/python scripts/train_with_fb.py \
--policy.type=native_fb_vla \
--policy.repo_id=kunhsiang/native-fb-vla-libero-10k-s1000 \
--dataset.repo_id=HuggingFaceVLA/libero \
--training.seed=? \
--training.batch_size=? \
--training.learning_rate=? \
--training.num_workers=?
Checkpoint 結構
checkpoints/
001000/ # step 1,000
002000/ # step 2,000
003000/ # step 3,000
004000/ # step 4,000
005000/ # step 5,000
...
010000/ # final (10,000 steps)
每個 step 目錄包含:
pretrained_model/model.safetensors— 模型權重pretrained_model/config.json— 模型設定pretrained_model/train_config.json— 完整訓練設定training_state/optimizer_state.safetensors— Optimizer 狀態（可 resume）training_state/rng_state.safetensors— RNG 狀態（可 resume）
Resume 訓練
.venv/bin/python scripts/train_with_fb.py \
--policy.type=native_fb_vla \
--resume=true \
--policy.pretrained_path=checkpoints/010000/pretrained_model
- Library
- lerobot
- License
- mit
- Created
- Jun 19, 2026
- Updated
- Jun 20, 2026