Instructions to use liuhaotian/llava-v1.5-13b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use liuhaotian/llava-v1.5-13b with Transformers:

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

pipe = pipeline("image-text-to-text", model="liuhaotian/llava-v1.5-13b")

# Load model directly
from transformers import AutoProcessor, AutoModelForCausalLM

processor = AutoProcessor.from_pretrained("liuhaotian/llava-v1.5-13b")
model = AutoModelForCausalLM.from_pretrained("liuhaotian/llava-v1.5-13b", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use liuhaotian/llava-v1.5-13b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "liuhaotian/llava-v1.5-13b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "liuhaotian/llava-v1.5-13b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/liuhaotian/llava-v1.5-13b

SGLang

How to use liuhaotian/llava-v1.5-13b 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 "liuhaotian/llava-v1.5-13b" \
    --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": "liuhaotian/llava-v1.5-13b",
		"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 "liuhaotian/llava-v1.5-13b" \
        --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": "liuhaotian/llava-v1.5-13b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use liuhaotian/llava-v1.5-13b with Docker Model Runner:
```
docker model run hf.co/liuhaotian/llava-v1.5-13b
```

will first-stage checkpoint of llava-v1.5-13b be released? (trained on caption, but not tuned on instructions)

by Maxlinn - opened Oct 12, 2023

Discussion

Maxlinn

Oct 12, 2023

thanks for your hard work on llava, llava-v1.5 has shown great improvement!

recently i want to do some research on instruction data, but due to tight budget i can not train the first-stage, will it be released in the future?

Maxlinn changed discussion title from will first-stage checkpoint of llava-v1.5-13b release? (trained on caption, but not tuned on instructions) to will first-stage checkpoint of llava-v1.5-13b be released? (trained on caption, but not tuned on instructions) Oct 12, 2023

liuhaotian

Owner Oct 12, 2023

First-stage projectors are already released here:

https://github.com/haotian-liu/LLaVA/blob/main/docs/MODEL_ZOO.md#projector-weights

Maxlinn

Oct 14, 2023

much thanks for timly responding!

Maxlinn changed discussion status to closed Oct 14, 2023

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