Please download the following data from the given links, EmpatheticDialogues, GoEmotions, Affect in Tweet and NRC VAD Lexicon
Note: For EmpatheticDialogue dataset, please use ed_data_extract.py to transform the data before preprocessing.
For data preprocessing, run the following command
python preprocess.py
For training the model, go to config.py or config_multilabel.py to set the required parameters.
The training for this work was done entirely in Google Colab due to resource requirements. Use kea_singlelabel_colab_notebook for single label setting and kea_multilabel_colab notebook for multilabel settings.
Follow the below instructions to use the python scripts
python train.py ## for single-label settings
python train_multilabel.py ## for multi-label settings
Install the required packages mentioned in requirements.txt using pip.
pip install -r requirements.txt
This application uses Open Source components. You can find the source code of their open source projects along with license information below. We acknowledge and are grateful to these developers for their contributions to open source.
-
Project: Text-Classification-Pytorch https://github.com/prakashpandey9/Text-Classification-Pytorch
License https://github.com/prakashpandey9/Text-Classification-Pytorch/blob/master/LICENSE.txt -
Project:EmpatheticDialogues https://github.com/facebookresearch/EmpatheticDialogues License https://github.com/facebookresearch/EmpatheticDialogues/blob/master/LICENSE
-
Project:GoEmotions https://github.com/google-research/google-research/tree/master/goemotions
-
Dataset:Affect in Tweets https://competitions.codalab.org/competitions/17751#learn_the_details-datasets
-
Project:MoEL https://github.com/HLTCHKUST/MoEL
License https://github.com/HLTCHKUST/MoEL/blob/master/LICENSE