Sentiment analysis on twitter data using BERT from google (https://arxiv.org/abs/1810.04805) using the pytorch implementation with the help of the libary tranformers. (https://huggingface.co/transformers/)
Use the Train class implementation (refer to .\deeplearning\README.md) to train your own model for sentiment classification using BERT.
Data should be in csv in the folder .\data with the required columns 'sentiment' and 'content' which has your emotions and text sentences respectively.
Navigate to the folder Sentiment-Analysis-with-BERT and run the command below:
pip install -r requirements.txt
Note that pytorch will have some issues of installing. Make sure you install pytorch separately if you have any issues.
If you find any issues/bugs with the code. I am available on email at venkyr91193@gmail.com