Instructions to use lamm-mit/Graph-PRefLexOR-4B-Inpainting with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use lamm-mit/Graph-PRefLexOR-4B-Inpainting with Transformers:

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

pipe = pipeline("text-generation", model="lamm-mit/Graph-PRefLexOR-4B-Inpainting")
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("lamm-mit/Graph-PRefLexOR-4B-Inpainting")
model = AutoModelForMultimodalLM.from_pretrained("lamm-mit/Graph-PRefLexOR-4B-Inpainting", 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 lamm-mit/Graph-PRefLexOR-4B-Inpainting with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "lamm-mit/Graph-PRefLexOR-4B-Inpainting"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "lamm-mit/Graph-PRefLexOR-4B-Inpainting",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/lamm-mit/Graph-PRefLexOR-4B-Inpainting

SGLang

How to use lamm-mit/Graph-PRefLexOR-4B-Inpainting 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 "lamm-mit/Graph-PRefLexOR-4B-Inpainting" \
    --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": "lamm-mit/Graph-PRefLexOR-4B-Inpainting",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "lamm-mit/Graph-PRefLexOR-4B-Inpainting" \
        --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": "lamm-mit/Graph-PRefLexOR-4B-Inpainting",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use lamm-mit/Graph-PRefLexOR-4B-Inpainting with Docker Model Runner:
```
docker model run hf.co/lamm-mit/Graph-PRefLexOR-4B-Inpainting
```

mjbuehler commited on Jul 18

Commit

35feffc

verified ·

1 Parent(s): 2936a5f

Update README.md

Browse files

Files changed (1) hide show

README.md +3 -0

README.md CHANGED Viewed

@@ -48,6 +48,9 @@ The training data contains six graph-inpainting and repair modes:
 For additive completion, supplied objects should normally be marked fixed.
 For repair or removal, the supplied graph must remain editable.
 ## Output contract
 The model may use Gemma's native thinking channel. The scored final response is

 For additive completion, supplied objects should normally be marked fixed.
 For repair or removal, the supplied graph must remain editable.
+![image](https://cdn-uploads.huggingface.co/production/uploads/623ce1c6b66fedf374859fe7/MxCwVjW4v6kILliJv47Xm.png)
 ## Output contract
 The model may use Gemma's native thinking channel. The scored final response is