A template to implement Machine Learning projects faster.
You need to define the following based on your project:
- Your model in a file named "models.py", and the class name which can be anything. I set it to TheModel.
- Your dataset loader in a file named "datasets.py" and a function called data_loader
- analyze function in which you can save model's ouput and make some plots and csv files.
- compute_metrics function in order to compute metrics you want including f1 score and accuracy
Run the following command:
bash run_ML.sh develop 200 100 attention train-test random 42 tanh