Instructions to use Salesforce/xgen-mm-phi3-mini-instruct-r-v1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Salesforce/xgen-mm-phi3-mini-instruct-r-v1 with Transformers:

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

pipe = pipeline("image-text-to-text", model="Salesforce/xgen-mm-phi3-mini-instruct-r-v1", trust_remote_code=True)
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 AutoModelForVision2Seq
model = AutoModelForVision2Seq.from_pretrained("Salesforce/xgen-mm-phi3-mini-instruct-r-v1", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Salesforce/xgen-mm-phi3-mini-instruct-r-v1 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Salesforce/xgen-mm-phi3-mini-instruct-r-v1"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Salesforce/xgen-mm-phi3-mini-instruct-r-v1",
		"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/Salesforce/xgen-mm-phi3-mini-instruct-r-v1

SGLang

How to use Salesforce/xgen-mm-phi3-mini-instruct-r-v1 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 "Salesforce/xgen-mm-phi3-mini-instruct-r-v1" \
    --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": "Salesforce/xgen-mm-phi3-mini-instruct-r-v1",
		"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 "Salesforce/xgen-mm-phi3-mini-instruct-r-v1" \
        --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": "Salesforce/xgen-mm-phi3-mini-instruct-r-v1",
		"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 Salesforce/xgen-mm-phi3-mini-instruct-r-v1 with Docker Model Runner:
```
docker model run hf.co/Salesforce/xgen-mm-phi3-mini-instruct-r-v1
```

The model does not currently support batch size more than 1

by Napron - opened May 19, 2024

Discussion

Napron

May 19, 2024

I am trying to run inference with more than 1 image and it throws the error, could you please fix this.

Thanks in advance.

Napron

May 19, 2024

tokenizer.padding_side = "left"

#Adding one more img in the list
inputs = image_processor([img, img], return_tensors="pt", image_aspect_ratio='anyres') # image_aspect_ratio='anyres')

prompt = apply_prompt_template(sample['question'])
language_inputs = tokenizer([prompt], return_tensors="pt")
inputs.update(language_inputs)

#Batch size becomes 2
torch.Size([1, 2, 5, 3, 378, 378])

RuntimeError: Expected 3D (unbatched) or 4D (batched) input to conv2d, but got input of size: [2, 5, 3, 378, 378]

UncleFish

Salesforce AI Research org May 24, 2024

@Napron Thanks for your interest in our work. We will work on it and update you when the batch inference feature is ready.

Manli

Salesforce AI Research org May 28, 2024

Hi @Napron , thank you for being patient. Check out our latest notebook for batch inference. Let us know if you have any questions.

Napron

May 29, 2024

Thank you, appreciate it!

Manli changed discussion status to closed May 29, 2024

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