creating_models.md

Working With Models

In each model folder, there are 3 sub folders:

• content/
  • Information about model and performance metrics.
• docs/
  • Documentation for the model.
• src/
  • Code for the model.

In the src/ folder, we organize the code into multiple .py files, each serving a specific purpose:

• methods.py
  • Containes methods that handle the core logic, such as training and evaluation routines.
• architecture.py
  • Defines the model architecture.
• dataset.py
  • Handles the dataset and data augmentation definitions.

To run the code and execute these methods, we work from Jupyter notebooks located in the src/notebooks/ folder. Each notebook is tailored for a specific environment, such as Windows, macOS, Google Colab, and more. This allows you to choose the environment that best suits your needs while leveraging the methods defined in methods.py. A tutorial on how to work with cloud environments.

For example, while methods.py includes the methods to train a model, the actual training process is initiated and managed from a Jupyter notebook specific to your environment in the src/notebooks/ folder. This setup keeps our code organized and makes it easier to experiment and iterate across different platforms.

Creating Models

1. Navigate to src/scripts.
2. Right click on either create_model.py or create_copy_of_model.py
   • create_model.py creates a new model from the template
   • create_copy_of_model.py creates a copy of a model using its model id
3. Click "Run Python File in Termainl".
4. Answer the prompts in the terminal.