DeepLearningTransformers

Structure

Files:

• data_downloader.ipnyb - notebook for loading data from kaggle
• Audiodataset.py - code for processing audio files
• data_sampler.ipynb - notebook for sampling data for train dataset
• [architecture_name].ipynb - notebook for running experiments
• plots.ipynb - notebook for plots generation
• requirements.txt - venv requirements for running the project

Run the code

Prerequisites

Before running the code, ensure you have the necessary libraries installed:

pip install -r requirements.txt

1. Experiments

Initialize and train the model by running architecture_name.ipynb file, for example cnn_advanced.ipnyb. The output of expewriments will be saved i hyper_results and loss_results folder. All folders will be created automatically.

2. Plots

Generate plots via plots.ipnyb file, the output of the code will be saved in plots folder.