Usage

• image_sm: Single-modality Image Tasks
  • Image Classification: script & model
• video_sm: Single-modality Video Tasks
  • Short-term Video Understanding: script & model
  • Long-term Video Understanding: script & model
  • Masked Modeling: script, model
• video_mm: Multi-modality Video Tasks
  • Video-Text Retrieval: script & model

Installation

• Clone this repo:

  git clone https://github.com/OpenGVLab/VideoMamba
  cd VideoMamba

• Create Conda environments

  conda create -n mamba python=3.10
  conda activate mamba

• Install PyTorch 2.1.1+cu118

  pip install torch==2.1.1 torchvision==0.16.1 torchaudio==2.1.1 --index-url https://download.pytorch.org/whl/cu118

• Install causal_conv1d and mamba

  pip install -r requirements.txt
  pip install -e causal-conv1d
  pip install -e mamba