A Supervised ULMFiT model to Twitter US Airlines Sentiment.
This is a repository created purely with my understandings on ULMFIT by executing it on Twitter Airlines Sentiment hosted at kaggle.
Note: I have not included weights.You can just run and it just takes few minutes to train and predict. I recommend you to tweak the hyparameters to get even more results.
- This Repository is primarily inspired from this github repository.
- "Universal Language Model Fine-tuning for Text Classification"
- "A Comparison of Pre-processing Techniques for Twitter Sentiment Analysis"
- "A disciplined approach to neural network hyper-parameters: Part 1 -- learning rate, batch size, momentum, and weight decay"