https://github.com/Erlemar/pytorch_tempest
My repo for training neural nets using pytorch-lightning and hydra
https://github.com/Erlemar/pytorch_tempest
deep-learning hacktoberfest hydra pytorch-lightning training-pipeline
Last synced: 10 days ago
JSON representation
My repo for training neural nets using pytorch-lightning and hydra
- Host: GitHub
- URL: https://github.com/Erlemar/pytorch_tempest
- Owner: Erlemar
- License: mit
- Created: 2020-08-30T10:46:59.000Z (over 4 years ago)
- Default Branch: master
- Last Pushed: 2024-08-03T07:27:14.000Z (10 months ago)
- Last Synced: 2024-08-04T03:12:43.901Z (10 months ago)
- Topics: deep-learning, hacktoberfest, hydra, pytorch-lightning, training-pipeline
- Language: Python
- Homepage: https://pytorch-tempest.readthedocs.io/en/latest/
- Size: 114 MB
- Stars: 208
- Watchers: 8
- Forks: 21
- Open Issues: 5
-
Metadata Files:
- Readme: README.md
- Contributing: contributing.md
- License: LICENSE
Awesome Lists containing this project
README
# tempest
[](https://deepsource.io/gh/Erlemar/pytorch_tempest/?ref=repository-badge)
This repository has my pipeline for training neural nets.
Main frameworks used:
* [hydra](https://github.com/facebookresearch/hydra)
* [pytorch-lightning](https://github.com/PyTorchLightning/pytorch-lightning)
The main ideas of the pipeline:
* all parameters and modules are defined in configs;
* prepare configs beforehand for different optimizers/schedulers and so on, so it is easy to switch between them;
* have templates for different deep learning tasks. Currently, image classification and named entity recognition are supported;
Examples of running the pipeline:
This will run training on MNIST (data will be downloaded):
```shell
>>> python train.py --config-name mnist_config model.encoder.params.to_one_channel=True
```
Running on MPS (M1 macbook)
```shell
python train.py --config-name mnist_config model.encoder.params.to_one_channel=True trainer.accelerator=mps +trainer.devices=1 optimizer=adan training.lr=0.001
```
Running on MPS (M1 macbook) with schedule free optimizer https://github.com/facebookresearch/schedule_free/tree/main
```shell
python train.py --config-name mnist_config model.encoder.params.to_one_channel=True trainer.accelerator=mps trainer.devices=1 optimizer=adamwschedulefree training.lr=0.001 scheduler.params.patience=100
```
The default run:
```shell
>>> python train.py
```
The default version of the pipeline is run on imagenette dataset. To do it, download the data from this repository:
https://github.com/fastai/imagenette
unzip it and define the path to it in conf/datamodule/image_classification.yaml path