Instructions to use fancyfeast/llama-joycaption-beta-one-hf-llava with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use fancyfeast/llama-joycaption-beta-one-hf-llava with Transformers:

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

pipe = pipeline("image-text-to-text", model="fancyfeast/llama-joycaption-beta-one-hf-llava")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("fancyfeast/llama-joycaption-beta-one-hf-llava")
model = AutoModelForMultimodalLM.from_pretrained("fancyfeast/llama-joycaption-beta-one-hf-llava", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use fancyfeast/llama-joycaption-beta-one-hf-llava with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "fancyfeast/llama-joycaption-beta-one-hf-llava"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "fancyfeast/llama-joycaption-beta-one-hf-llava",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/fancyfeast/llama-joycaption-beta-one-hf-llava

SGLang

How to use fancyfeast/llama-joycaption-beta-one-hf-llava 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 "fancyfeast/llama-joycaption-beta-one-hf-llava" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "fancyfeast/llama-joycaption-beta-one-hf-llava",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

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 "fancyfeast/llama-joycaption-beta-one-hf-llava" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "fancyfeast/llama-joycaption-beta-one-hf-llava",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use fancyfeast/llama-joycaption-beta-one-hf-llava with Docker Model Runner:
```
docker model run hf.co/fancyfeast/llama-joycaption-beta-one-hf-llava
```

Anyone have a script to run this with vLLM?

#15

by Ataylorm - opened Jan 16

Discussion

Ataylorm

Jan 16

I have been running this on vLLM for months without any issue. Last night RunPod sent me an email that there was an issue with my pod. Which turns out to be they literally wiped it out. Now I am trying to get it setup again, but my scripts I had to run it, are no longer working, with a variety of errors. I am guessing something has changed with vLLM that is causing these issues. Wondering if anyone has a working setup script they could share?

Here is my latest error:

Here is my run script:

Here is my setup script:

fancyfeast

Owner Jan 17

I know that the docker image for vLLM v0.10.2 works, so I would try to install specifically v0.10.2 of vLLM, or if possible use that docker image as the base for your pod (I'm not familiar with what runpod requires).

If that fails let me know what versions of vLLM, torch, and transformers your pod is running with.

Hubbyboyluvboobs

Jan 22

•

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