Instructions to use HuggingFaceM4/idefics2-8b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use HuggingFaceM4/idefics2-8b with Transformers:

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

pipe = pipeline("image-text-to-text", model="HuggingFaceM4/idefics2-8b")

# Load model directly
from transformers import AutoProcessor, AutoModelForImageTextToText

processor = AutoProcessor.from_pretrained("HuggingFaceM4/idefics2-8b")
model = AutoModelForImageTextToText.from_pretrained("HuggingFaceM4/idefics2-8b")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use HuggingFaceM4/idefics2-8b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "HuggingFaceM4/idefics2-8b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "HuggingFaceM4/idefics2-8b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/HuggingFaceM4/idefics2-8b

SGLang

How to use HuggingFaceM4/idefics2-8b 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 "HuggingFaceM4/idefics2-8b" \
    --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": "HuggingFaceM4/idefics2-8b",
		"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 "HuggingFaceM4/idefics2-8b" \
        --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": "HuggingFaceM4/idefics2-8b",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use HuggingFaceM4/idefics2-8b with Docker Model Runner:
```
docker model run hf.co/HuggingFaceM4/idefics2-8b
```

Is idefics-8b trained with system roles?

#67

by chenzizhao - opened Jun 6, 2024

Discussion

chenzizhao

Jun 6, 2024

In the model card, we have

System prompt used in the playground for `idefics2-8b-chatty`

SYSTEM_PROMPT = "System: The following ..."

Was idefics2-8b trained with such a system prompt as well? Additionally, can I insert snippets of "System: xxx" between conversations of "User: xxx" and "Assistant: xxx"?

Thanks!

VictorSanh

Jun 6, 2024

hi @ch272h
that is a valid question!
we did NOT train with the system prompt, only with User/Assistant turns. the system prompt is this only injected at inference time.

Additionally, can I insert snippets of "System: xxx" between conversations of "User: xxx" and "Assistant: xxx"?
I have never tried it. i would encourage you to test.
what would be the use-case that requires a new system prompt for each turn?

chenzizhao

Jun 6, 2024

I see. Thank you for the clarification.

I am playing a stateful game (like chess) with idefics2, and I'd like to communicate the board state after each turn via system prompts instead of user prompts. Nevertheless, it probably is not one of the use cases Idefics2 was designed for and that's fine! Thanks again.

VictorSanh changed discussion status to closed Jun 7, 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

Is idefics-8b trained with system roles?

System prompt used in the playground for idefics2-8b-chatty

System prompt used in the playground for `idefics2-8b-chatty`