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

Libraries

How to use ys-zong/llava-v1.5-13b-Clean with Transformers:

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

pipe = pipeline("text-generation", model="ys-zong/llava-v1.5-13b-Clean")

# Load model directly
from transformers import AutoProcessor, AutoModelForCausalLM

processor = AutoProcessor.from_pretrained("ys-zong/llava-v1.5-13b-Clean")
model = AutoModelForCausalLM.from_pretrained("ys-zong/llava-v1.5-13b-Clean")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use ys-zong/llava-v1.5-13b-Clean with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/ys-zong/llava-v1.5-13b-Clean

SGLang

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

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

VLGuard

[Website] [Paper] [Code]

Safety Fine-Tuning at (Almost) No Cost: A Baseline for Vision Large Language Models. (ICML 2024)

Weights

This is the model weight for LLaVA-v1.5-13B re-trained after removing harmful samples in the training data. You can use them in exactly the same way as the original LLaVA.

Usage

Please refer to Github for detailed usage.

Downloads last month: 6

Collection including ys-zong/llava-v1.5-13b-Clean

VLGuard

Collection

Data and Model weights for VLGuard: https://ys-zong.github.io/VLGuard/ • 13 items • Updated Jun 18, 2024 • 1

Paper for ys-zong/llava-v1.5-13b-Clean

Safety Fine-Tuning at (Almost) No Cost: A Baseline for Vision Large Language Models

Paper • 2402.02207 • Published Feb 3, 2024 • 2