Instructions to use YanweiLi/llama-vid-7b-full-336 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use YanweiLi/llama-vid-7b-full-336 with Transformers:

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

pipe = pipeline("text-generation", model="YanweiLi/llama-vid-7b-full-336")

# Load model directly
from transformers import AutoProcessor, AutoModelForCausalLM

processor = AutoProcessor.from_pretrained("YanweiLi/llama-vid-7b-full-336")
model = AutoModelForCausalLM.from_pretrained("YanweiLi/llama-vid-7b-full-336")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use YanweiLi/llama-vid-7b-full-336 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "YanweiLi/llama-vid-7b-full-336"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "YanweiLi/llama-vid-7b-full-336",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/YanweiLi/llama-vid-7b-full-336

SGLang

How to use YanweiLi/llama-vid-7b-full-336 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 "YanweiLi/llama-vid-7b-full-336" \
    --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": "YanweiLi/llama-vid-7b-full-336",
		"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 "YanweiLi/llama-vid-7b-full-336" \
        --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": "YanweiLi/llama-vid-7b-full-336",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use YanweiLi/llama-vid-7b-full-336 with Docker Model Runner:
```
docker model run hf.co/YanweiLi/llama-vid-7b-full-336
```

LLaMA-VID Model Card

Model details

LLaMA-VID empowers existing frameworks to support hour-long videos and pushes their upper limit with an extra context token.

Model type: LLaMA-VID is an open-source chatbot trained by fine-tuning LLaMA/Vicuna on GPT-generated multimodal instruction-following data. LLaMA-VID empowers existing frameworks to support hour-long videos and pushes their upper limit with an extra context token. We build this repo based on LLaVA.

Model date: llama-vid-7b-full-336 was trained on 11/2023.

License

Where to send questions or comments about the model: https://github.com/dvlab-research/LLaMA-VID/issues

Intended use

Primary intended uses: The primary use of LLaMA-VID is research on large multimodal models and chatbots.

Primary intended users: The primary intended users of the model are researchers and hobbyists in computer vision, natural language processing, machine learning, and artificial intelligence.

Training data

This model is trained based on LLaVA-1.5 dataset, including

558K filtered image-text pairs from LAION/CC/SBU, captioned by BLIP.
158K GPT-generated multimodal instruction-following data.
450K academic-task-oriented VQA data mixture.
40K ShareGPT data.

Downloads last month: 11

Collection including YanweiLi/llama-vid-7b-full-336

LLaMA-VID

Collection

LLaMA-VID checkpoints. Please refer to project page for more detail: https://llama-vid.github.io/ • 11 items • Updated Dec 3, 2023 • 4

Paper for YanweiLi/llama-vid-7b-full-336

LLaMA-VID: An Image is Worth 2 Tokens in Large Language Models

Paper • 2311.17043 • Published Nov 28, 2023