Image-Text-to-Text
Transformers
ONNX
Safetensors
English
vision-language-action
edge-deployment
tensorRT
qwen
Instructions to use xintaozhen/MiniVLA with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use xintaozhen/MiniVLA with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-text-to-text", model="xintaozhen/MiniVLA")# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("xintaozhen/MiniVLA", dtype="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps
- vLLM
How to use xintaozhen/MiniVLA with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "xintaozhen/MiniVLA" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "xintaozhen/MiniVLA", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/xintaozhen/MiniVLA
- SGLang
How to use xintaozhen/MiniVLA with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "xintaozhen/MiniVLA" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "xintaozhen/MiniVLA", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "xintaozhen/MiniVLA" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "xintaozhen/MiniVLA", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use xintaozhen/MiniVLA with Docker Model Runner:
docker model run hf.co/xintaozhen/MiniVLA
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
# MiniVLA
|
| 2 |
|
| 3 |
This repository hosts **MiniVLA** โ a modular and deployment-friendly Vision-Language-Action (VLA) model designed for **edge hardware** (e.g., Jetson Orin Nano).
|
|
@@ -20,7 +36,7 @@ By exporting the **vision encoder** into ONNX and TensorRT engines, we significa
|
|
| 20 |
The MiniVLA deployment is designed with modular microservices:
|
| 21 |
|
| 22 |
<p align="center">
|
| 23 |
-
<img src="System_Architecture.svg" width="
|
| 24 |
</p>
|
| 25 |
|
| 26 |
|
|
@@ -33,7 +49,7 @@ The MiniVLA deployment is designed with modular microservices:
|
|
| 33 |
### Hybrid Acceleration
|
| 34 |
|
| 35 |
<p align="center">
|
| 36 |
-
<img src="MiniVLA_Architecture.svg" width="
|
| 37 |
</p>
|
| 38 |
|
| 39 |
|
|
@@ -136,4 +152,4 @@ We also acknowledge and thank the authors of [Stanford-ILIAD/minivla-vq-libero90
|
|
| 136 |
## ๐ Related Project
|
| 137 |
|
| 138 |
For full implementation and code, please visit the companion GitHub repository:
|
| 139 |
-
๐ [https://github.com/Zhenxintao/MiniVLA](https://github.com/Zhenxintao/MiniVLA)
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
+
tags:
|
| 6 |
+
- vision-language-action
|
| 7 |
+
- edge-deployment
|
| 8 |
+
- tensorRT
|
| 9 |
+
- qwen
|
| 10 |
+
base_model: Stanford-ILIAD/minivla-vq-libero90-prismatic
|
| 11 |
+
library_name: transformers
|
| 12 |
+
datasets:
|
| 13 |
+
- LIBERO
|
| 14 |
+
pipeline_tag: text-to-action
|
| 15 |
+
---
|
| 16 |
+
|
| 17 |
# MiniVLA
|
| 18 |
|
| 19 |
This repository hosts **MiniVLA** โ a modular and deployment-friendly Vision-Language-Action (VLA) model designed for **edge hardware** (e.g., Jetson Orin Nano).
|
|
|
|
| 36 |
The MiniVLA deployment is designed with modular microservices:
|
| 37 |
|
| 38 |
<p align="center">
|
| 39 |
+
<img src="System_Architecture.svg" width="100%">
|
| 40 |
</p>
|
| 41 |
|
| 42 |
|
|
|
|
| 49 |
### Hybrid Acceleration
|
| 50 |
|
| 51 |
<p align="center">
|
| 52 |
+
<img src="MiniVLA_Architecture.svg" width="100%">
|
| 53 |
</p>
|
| 54 |
|
| 55 |
|
|
|
|
| 152 |
## ๐ Related Project
|
| 153 |
|
| 154 |
For full implementation and code, please visit the companion GitHub repository:
|
| 155 |
+
๐ [https://github.com/Zhenxintao/MiniVLA](https://github.com/Zhenxintao/MiniVLA)
|