Emotion classification on MSP-Podcast(Temporarily)

Introduction

The repository is for the emotion classification task on MSP-Podcast dataset. The dataset is not public yet, so you have to prepare corpus file first. The code is written in PyTorch.

File Structure

.
├── config.toml.example
├── display.py
├── net
├── normal.py
├── PerformanceLog
├── README.md
├── test.ipynb
├── train.py
├── todo.todo
└── utils.py

Files

config.toml.example : Example configuration file. Copy this file to config.toml and modify it to your needs.

train.py : Main train file.

utils.py : Utility functions. Store function like PyTorch dataset, mixup, train report, etc.

display.py : Display functions, for terminal display.

test.ipynb : Jupyter notebook for testing/writing scratchy code.

todo.todo : My working log.

Folders

net : Store all the network architecture files here.

PerformanceLog : Store all the performance logs here.