ComfyUI_MiniCPM-V

ComfyUI_MiniCPM-V
★ 42

模型集成文本生成ComfyUI节点轻量级大模型
为ComfyUI提供MiniCPM-V模型节点，实现MiniCPM-V在可视化流程中的快速加载与推理，方便测试与集成。
💡 在ComfyUI流程中加载并调用MiniCPM-V进行文本推理与测试。
🍴 6 Forks💻 Python🔄 2024-08-09
📦
网盘下载
复制链接后前往夸克网盘下载
https://pan.quark.cn/s/b6135d9bd930
📦 requirements.txt
einops
gradio==4.15.0
huggingface_hub
matplotlib
numpy>=1.23.4
omegaconf
openai==1.3.5
opencv-python>=4.4.0.46
openpyxl
pandas>=1.5.3
pillow
portalocker
protobuf
pycocoevalcap
python-dotenv
requests
rich
seaborn
sentencepiece
sty
tabulate
tiktoken
timeout-decorator
tqdm
typing_extensions==4.7.1
validators
visual_genome
xlsxwriter
Pillow==10.1.0
sentencepiece==0.1.99
transformers==4.40.0
torch==1.13.1
torchvision
📄 README

ComfyUI MiniCPM-V (Unofficial Support)

Unofficial MiniCPM-V support for ComfyUI.

Important Updates

  • 2024/08/09: Added support for MiniCPM-V 2.6 (16GB+ video memory required)
  • 2024/05/22: Added support for MiniCPM-Llama3-V 2.5 (16GB+ video memory required)
  • 2024/04/18: Added ComfyUI nodes and workflow examples

    • Basic Workflow

    This workflow shows the basic usage on querying an image with Chinese and English.

    Version 2.6 Workflow

    This workflow shows the Verson 2.6’s usage on querying multiple images and a video.

    Installation

  • Install from ComfyUI Manager (search for minicpm)

  • Download or git clone this repository into the ComfyUI/custom_nodes/ directory and run:
    • pip install -r requirements.txt

    Download Checkpoints

    All the models will be downloaded automatically, no worries.