Instructions to use FlashVL/FlashVL-2B-Dynamic with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use FlashVL/FlashVL-2B-Dynamic with Transformers:

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

pipe = pipeline("image-text-to-text", model="FlashVL/FlashVL-2B-Dynamic", 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 AutoModel
model = AutoModel.from_pretrained("FlashVL/FlashVL-2B-Dynamic", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use FlashVL/FlashVL-2B-Dynamic with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "FlashVL/FlashVL-2B-Dynamic"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "FlashVL/FlashVL-2B-Dynamic",
		"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/FlashVL/FlashVL-2B-Dynamic

SGLang

How to use FlashVL/FlashVL-2B-Dynamic 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 "FlashVL/FlashVL-2B-Dynamic" \
    --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": "FlashVL/FlashVL-2B-Dynamic",
		"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 "FlashVL/FlashVL-2B-Dynamic" \
        --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": "FlashVL/FlashVL-2B-Dynamic",
		"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 FlashVL/FlashVL-2B-Dynamic with Docker Model Runner:
```
docker model run hf.co/FlashVL/FlashVL-2B-Dynamic
```

FlashVL-2B-Dynamic

File size: 732 Bytes

import copy
import transformers
from transformers import PretrainedConfig, Qwen2Config
from .configuration_aimv2 import AIMv2Config

class FlashVLDynamicConfig(PretrainedConfig):
    model_type = 'FlashVLDynamicConfig'
    is_composition = True

    def __init__(
            self,
            vision_config,
            llm_config,
            **kwargs
        ):
        super().__init__(**kwargs)
        self.vision_config = AIMv2Config(**vision_config)
        self.llm_config = Qwen2Config(**llm_config)

    def to_dict(self):
        
        output = copy.deepcopy(self.__dict__)
        output['vision_config'] = self.vision_config.to_dict()
        output['llm_config'] = self.llm_config.to_dict()

        return output