Tools for training PyTorch models
Project description
torchninja
torchninja is a Python package that provides tools for training PyTorch models. It offers a flexible and easy-to-use training framework with built-in functionality for logging, metrics tracking, and model checkpointing.
Features
- Simplifies the process of training PyTorch models.
- Flexible and customizable trainer class.
- Automatic logging of training progress and metrics.
- Support for model checkpointing to save and load model states.
- Easy integration with TensorBoard for visualizing training metrics.
- Provides utility functions for common tasks in PyTorch model training.
Installation
You can install torchninja using pip:
pip install torchninja
torchninja has the following dependencies:
- torch
- torchvision
- tensorboard
- tqdm
- numpy
- matplotlib
Make sure these dependencies are installed in your environment before using torchninja.
Getting Started
To get started with torchninja, you can refer to the examples
directory in the project repository. It contains example scripts that demonstrate how to use the torchninja trainer class to train a PyTorch model. You can modify these examples according to your specific use case.
Documentation
The documentation for torchninja is available on the project's GitHub repository:
The documentation provides detailed information on the usage of the torchninja package, including the trainer class, available functionality, and usage examples.
Contributing
Contributions to torchninja are welcome! If you find any issues, have suggestions for improvements, or would like to add new features, please open an issue or submit a pull request on the project's GitHub repository.
License
torchninja is released under the MIT License. See the LICENSE file for details.
Contact
For any questions or inquiries, you can reach out to the project author:
- Author: Silvan Ferreira
- Email: silvanfj@gmail.com
Feel free to contact the author for any assistance or collaboration opportunities related to torchninja.
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file torchninja-0.1.1.tar.gz
.
File metadata
- Download URL: torchninja-0.1.1.tar.gz
- Upload date:
- Size: 8.4 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.10.9
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | d4bc50185f6b878aacf633d7f7072d4d4eef6e176d22b93d25c2cdda5a4a0f08 |
|
MD5 | e421397a21d410dbc344bc88079e481f |
|
BLAKE2b-256 | f270b5f7c55fa18522a473688d321460291d6e7b2b5ed80bd5d8a4c3f2ea7fbd |
File details
Details for the file torchninja-0.1.1-py3-none-any.whl
.
File metadata
- Download URL: torchninja-0.1.1-py3-none-any.whl
- Upload date:
- Size: 8.8 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.10.9
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 01f5c3235c46393a45fcf99dbfe1b97dae463f65c6d1895fb56b7cbb4e2d8e83 |
|
MD5 | 7d1111d4ad91457c764b457fe23f0b50 |
|
BLAKE2b-256 | 6ad191a566da18ade9362c4cef107285c958f2f3c21501fa4b627a4b4cb7bc57 |