pytorch-starter-code

A generic starter code for PyTorch projects. A config driven architecture which takes care of the boilerplate code for any project -

• Ability to run different experiments with just config changes.
• Saving logs, plots and models as we run the experiment with an ability to resume experiments.
• A standard training method which can be tweaked based on the project.
• Single command execution.

Usage

• Define the configuration for your experiment in ./config directory. See default.json to see the structure and available options.
• Implement factories to return project specific models, datasets and transforms based on config. Add more flags as per requirement in the config.
• Tweak experiment.py based on the project requirements.
• After defining the configuration (say my_exp.json) - simply run python3 main.py my_exp to start the experiment
• The logs, stats, plots and saved models would be stored in ../experiment_data/my_exp dir. This can be configured in contants.py
• To resume an ongoing experiment, simply run the same command again. It will load the latest stats and models and resume training.

Files

• main.py: Main driver class
• experiment.py: Main experiment class. Initialized based on config - takes care of training, saving stats and plots, logging and resuming experiments.
• dataset_factory: Factory to build datasets based on config
• transforms_factory.py: Factory to build image transforms based on config
• model_factory.py: Factory to build models based on config
• constants.py: constants used across the project
• file_utils.py: utility functions for handling files