Flashvl 2B Dynamic ISS

🚀 FlashVL-2B-Dynamic-ISS

We're thrilled to present FlashVL, a groundbreaking method for optimizing Vision-Language Models (VLMs) in real - time applications. It aims for ultra - low latency and high throughput without compromising accuracy. By leveraging advanced architecture enhancements and efficient computational strategies, Flash - VL 2B maximizes throughput by cutting processing time while maintaining competitive performance across multiple vision - language benchmarks.

Key Information

[📜 FlashVL]

🚀 Quick Start

Environment Setup

pip install torch==2.1.2
pip install transformers==4.50.0.dev0

How to use it?

import torch
from PIL import Image
import requests
from io import BytesIO
from transformers import AutoModel, AutoTokenizer, CLIPImageProcessor

model_path = "FlashVL/FlashVL-2B-Dynamic-ISS"
model = AutoModel.from_pretrained(model_path, torch_dtype=torch.bfloat16,trust_remote_code=True,device_map='cuda')
model.tokenizer = AutoTokenizer.from_pretrained(model_path,device_map='cuda')
model.im_trans = CLIPImageProcessor.from_pretrained(model_path)

# single-image single-round conversation (单图单轮对话)
image_url ="https://s3plus.meituan.net/automl-datasets/mlm/0516.png"
response = requests.get(image_url)
image_data = BytesIO(response.content)
pil_image = Image.open(image_data).convert('RGB')   
messages = [{'role': 'user', 'content': "生成图中菜品的菜谱"}] # answer: EXTRA
answer = model.chat(pil_image, messages, do_sample=False, max_new_tokens=256)
print(answer)

# single-image multi-round conversation (单图多轮对话)
messages = [
    {'role': 'user', 'content': '这是什么'},
    {"role": "assistant", "content": '这是一道看起来像是银耳莲子汤的甜品。\
     银耳是一种常见的食材，通常用于制作甜品和汤品，具有软糯的口感和清润的口感。莲 \
     子是莲子的干燥部分，常用于中医和食疗中，具有补脾止泻的功效。图片中还可以看到 \
     一些枸杞和核桃，枸杞富含维生素和抗氧化物质，核桃则提供丰富的蛋白质和健康脂肪。 \
     整体来看，这道甜品不仅美味，还具有一定的营养价值。'},
    {'role': 'user', 'content': '对图中菜品卡路里分析'}
    ]
answer = model.chat(pil_image, messages, do_sample=False, max_new_tokens=256)
print(answer)

# pure-text single-round conversation (纯文本对话）
messages = [{'role': 'user', 'content': "who are you"}]
answer = model.chat(None, messages, do_sample=False, max_new_tokens=256)
print(answer)

✨ Features

We are excited to introduce FlashVL, a novel approach to optimizing Vision - Language Models (VLMs) for real - time applications. It targets ultra - low latency and high throughput without sacrificing accuracy. Leveraging advanced architectural enhancements and efficient computational strategies, Flash - VL 2B is designed to maximize throughput by reducing processing time while maintaining competitive performance across multiple vision - language benchmarks. Our approach includes tailored architectural choices, token compression mechanisms, data curation, training schemes, and a novel image processing technique called implicit semantic stitching that effectively balances computational load and model performance.

📚 Documentation

Evaluation

We use [VLMEvalKit](https://github.com/open - compass/VLMEvalKit) to evaluate FlashVL - 2B - Static.

📄 License

This project is under the Apache - 2.0 license.

📜 Citation

If you find this project useful in your research, please consider citing:

@misc{zhang2025flashvl2boptimizingvisionlanguage,
      title={Flash-VL 2B: Optimizing Vision-Language Model Performance for Ultra-Low Latency and High Throughput}, 
      author={Bo Zhang and Shuo Li and Runhe Tian and Yang Yang and Jixin Tang and Jinhao Zhou and Lin Ma},
      year={2025},
      eprint={2505.09498},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2505.09498}, 
}