Instructions to use mistral-experimental/pixtral-12b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use mistral-experimental/pixtral-12b with Transformers:

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

pipe = pipeline("image-text-to-text", model="mistral-experimental/pixtral-12b")
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, AutoModelForImageTextToText

processor = AutoProcessor.from_pretrained("mistral-experimental/pixtral-12b")
model = AutoModelForImageTextToText.from_pretrained("mistral-experimental/pixtral-12b")
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 mistral-experimental/pixtral-12b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "mistral-experimental/pixtral-12b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "mistral-experimental/pixtral-12b",
		"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/mistral-experimental/pixtral-12b

SGLang

How to use mistral-experimental/pixtral-12b 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 "mistral-experimental/pixtral-12b" \
    --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": "mistral-experimental/pixtral-12b",
		"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 "mistral-experimental/pixtral-12b" \
        --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": "mistral-experimental/pixtral-12b",
		"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 mistral-experimental/pixtral-12b with Docker Model Runner:
```
docker model run hf.co/mistral-experimental/pixtral-12b
```

Support of flash attention 2?

#29

by LuciusLan - opened Feb 10, 2025

Discussion

LuciusLan

Feb 10, 2025

I notice that there is no FlashAttention class in the vision tower part (which is fine)
But seems that even when I overwrite the config for the LM part (by config = AutoConfig.from_pretrained(this_repo) and config.text_config._attn_implementation="flash_attention_2"), the attention implementation still becomes SDPA.
(And because there is no flash attention for vision tower, overwritting config for whole modelconfig._attn_implementation="flash_attention_2" will of course throw error)
Is there any way to switch the LM to use flash attention 2?

RaushanTurganbay

Feb 10, 2025

Can you try indicating attention while loading with model = LlavaForConditionalGeneration.from_pretrained(model_id, attn_implementation={"text_config": "flash_attention_2", "vision_config": "sdpa"}, torch_dtype=torch.float16)?

LMK if that works, prob the attention is not picked up from sub-config. I am planning to fix all attention related issues in VLMs soon

LuciusLan

Feb 10, 2025

Thank you for the suggestion! Not sure if it is because my transformers version (4.46.1), the vision tower also don't have sdpa, but changing it to eager works!

LuciusLan changed discussion status to closed Feb 10, 2025

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