meta-pytorch/tnt
A lightweight library for PyTorch training tools and utilities
What it solves
TNT provides a collection of tools and utilities specifically designed to streamline and enhance the process of training PyTorch models.
How it works
It functions as a library of utilities that integrates with PyTorch to provide developers with helper functions and tools to manage the training workflow.
Who it’s for
Developers and researchers using PyTorch for machine learning model training.
Highlights
- PyTorch-native training utilities
- Available via pip and conda
- Support for nightly builds for early access to new features
Related
- Project
neuralhydrology/neuralhydrologyA PyTorch-based Python library designed for deep learning research in hydrology, allowing users to train neural networks via configuration files.
- Project
Xilinx/brevitasA PyTorch library for neural network quantization that supports both post-training quantization and quantization-aware training to optimize model efficiency.
- Project
sktime/pytorch-forecastingA PyTorch-based library for time series forecasting that provides a high-level API and state-of-the-art deep learning architectures like Temporal Fusion Transformers and N-BEATS.
- Project
Talmaj/onnx2pytorchA library that transforms ONNX models into PyTorch models, allowing users to utilize PyTorch's backend and neural network manipulation tools.
- Project
Lightning-AI/pytorch-lightningA deep learning framework that abstracts away the engineering boilerplate of PyTorch, enabling researchers to pretrain and finetune AI models at scale across various hardware.