Here I used Bert for a sentiment analysis on twitter dataset. Bert is a large-scale transformer-based Language Model that can be finetuned for a variety of tasks. After importing the dataset I found there are some angry, sad, happy post that are collected from twitter comments. I used a pretrained model from Bert. The tqdm is imported from a progress bar.
Some addition details on this project is added in the image folder, where I also added one pdf which consists of bert mechanism for pretraining and fine tuning.
Original paper can be found here
Hugging Face documentation can be found here
Bert Documentation can be found here