KNU-Mamba is a PyTorch-based implementation of the Mamba State Space Model (SSM) architecture, developed and maintained by the Kyungpook National University (KNU).
This repository serves as a codebase for exploring efficient sequence modeling using Mamba, capable of handling long-range dependencies with linear computational complexity.
Note: This project builds upon the original Mamba architecture proposed by Gu & Dao (2023).
- Efficient Implementation: optimized for fast training and inference on GPUs.
- Modular Design: Easy integration into existing deep learning pipelines (e.g., for Computer Vision, NLP, or Time-Series analysis).
- Research Ready: Structured for experimentation with different state-space configurations.
- Linux
- NVIDIA GPU with CUDA 11.6+
- Python 3.8+
- PyTorch 1.12+
We recommend using Conda to manage the environment, as described below:
conda create -n knu-mamba python=3.10
conda activate knu-mamba
📄 License
This project is licensed under the MIT License. See the LICENSE file for details.