torch-brain 0.1.1

A package for deep learning models for neuroscience

Documentation | Join our Discord community

torch_brain is a Python library for various deep learning models designed for neuroscience.

Features

• Multi-recording training
• Optimized data loading with with on-demand data access -- only loads data when needed
• Advanced samplers that enable arbitrary slicing of data on the fly
• Advanced data collation strategies including chaining and padding
• Support for arbitrary neural and behavioral modalities
• Collection of useful nn.Modules like stitchers, multi-output readouts, infinite vocab embeddings, etc.
• Collection of neural and behavioral transforms and augmentation strategies
• Implementations of various deep learning models for neuroscience

List of implemented models

• POYO: A Unified, Scalable Framework for Neural Population Decoding (Azabou et al. 2023)
• More coming soon...

Installation

torch_brain is available for Python >= 3.10 and can be installed via pip.

We recommend using a virtual environment to manage dependencies, and Python 3.10.

To create and activate a new virtual environment with venv, run:

python3 -m venv venv
source venv/bin/activate  # On Windows, use: .venv\Scripts\activate
pip install torch_brain

Contributing

If you are planning to contribute to the package, you can install the package in development mode by running the following command:

pip install -e ".[dev]"

Install pre-commit hooks:

pre-commit install

Unit tests are located under test/. Run the entire test suite with

pytest

or test individual files via, e.g., pytest test/test_binning.py

Cite

Please cite our paper if you use this code in your own work:

@inproceedings{
    azabou2023unified,
    title={A Unified, Scalable Framework for Neural Population Decoding},
    author={Mehdi Azabou and Vinam Arora and Venkataramana Ganesh and Ximeng Mao and Santosh Nachimuthu and Michael Mendelson and Blake Richards and Matthew Perich and Guillaume Lajoie and Eva L. Dyer},
    booktitle={Thirty-seventh Conference on Neural Information Processing Systems},
    year={2023},
}