Instructions to use coisini9293/lingbot_pot14 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use coisini9293/lingbot_pot14 with Transformers:
# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("coisini9293/lingbot_pot14", device_map="auto") - Notebooks
- Google Colab
- Kaggle
|
Download README.md from coisini9293/lingbot_pot14: direct link, hf CLI and curl.
- Browser
- Download file 1.26 kB
-
https://huggingface.co/coisini9293/lingbot_pot14/resolve/main/README.md
- Command line
-
hf download hf://coisini9293/lingbot_pot14/README.md
-
curl -L -o README.md https://huggingface.co/coisini9293/lingbot_pot14/resolve/main/README.md
1.26 kB
metadata
license: apache-2.0
tags:
- robotics
- vla
- lingbot-vla
- pot14
library_name: transformers
lingbot_pot14
LingBot-VLA 在 pot14 自采右臂数据上微调到 global_step_1024 的部署权重。
- 基座: robbyant/lingbot-vla-4b
- 任务: 杯子等(训练文本形如
完成任务:杯子) - 动作: 7 维单活动臂(6 关节 + 1 effector)
- 相机: top / left / right(由拼接画面裁剪)
仓库内容
| 路径 | 说明 |
|---|---|
model-*.safetensors + config.json 等 |
HF 格式权重(hf_ckpt) |
configs/pot14_right_arm.json |
归一化统计量 |
configs/robot_pot14.yaml |
机器人特征映射 |
configs/vla_pot14.yaml |
训练配置参考 |
下载
export HF_ENDPOINT=https://hf-mirror.com # 国内可选
huggingface-cli download coisini9293/lingbot_pot14 \
--local-dir models/finetuned/lingbot_pot14
推理还需要 Qwen 处理器:
huggingface-cli download Qwen/Qwen2.5-VL-3B-Instruct \
--local-dir models/pretrained/Qwen2.5-VL-3B-Instruct
完整训练 / 云端推理 / 本地串口说明见项目文档docs/pot14_full_guide.md。