Instructions to use Efficient-Large-Model/NVILA-Lite-8B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Efficient-Large-Model/NVILA-Lite-8B with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="Efficient-Large-Model/NVILA-Lite-8B")

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("Efficient-Large-Model/NVILA-Lite-8B", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Efficient-Large-Model/NVILA-Lite-8B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Efficient-Large-Model/NVILA-Lite-8B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Efficient-Large-Model/NVILA-Lite-8B",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/Efficient-Large-Model/NVILA-Lite-8B

SGLang

How to use Efficient-Large-Model/NVILA-Lite-8B 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 "Efficient-Large-Model/NVILA-Lite-8B" \
    --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": "Efficient-Large-Model/NVILA-Lite-8B",
		"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 "Efficient-Large-Model/NVILA-Lite-8B" \
        --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": "Efficient-Large-Model/NVILA-Lite-8B",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use Efficient-Large-Model/NVILA-Lite-8B with Docker Model Runner:
```
docker model run hf.co/Efficient-Large-Model/NVILA-Lite-8B
```

quantizing into int8 precision

by tanvij - opened Mar 31, 2025

Discussion

tanvij

Mar 31, 2025

•

edited Mar 31, 2025

I'm looking into quantizing this model to int8 precision and I'm wondering if I should manually quantize the weights or use an automated technique like AWQ or bitsandbytes. Any recommendations on which method works best for this model? Thanks! @klldmofashi

tanvij changed discussion title from quantizing to quantizing into int8 precision Mar 31, 2025

Louym

Efficient-Large-Model org Mar 31, 2025

This comment has been hidden (marked as Off-Topic)

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment