Team name: Natural Language Processors
Kindly View the report for this project here
Steps to execude the code -
- Clone the repository
- First Run the preprocessing folder to generate the correct dataset files in csv format. This will clean the data, both train and validation.
- Then start running by milestone 2, upload each .pynb file in colab and run each cell.
- Then start running for milestone 3, upload each file in colab and keep running each cell to see results.
- Please contact us incase of issues.
https://huggingface.co/sentence-transformers/multi-qa-MiniLM-L6-cos-v1 https://discuss.huggingface.co/t/trainer-only-doing-3-epochs-no-matter-the-trainingarguments/19347/5 https://huggingface.co/docs/transformers/tasks/question_answering#preprocess https://towardsdatascience.com/fine-tune-transformer-models-for-question-answering-on-custom-data-513eaac37a80 https://huggingface.co/transformers/v3.3.1/custom_datasets.html#qa-squad https://towardsdatascience.com/question-answering-with-pretrained-transformers-using-pytorch-c3e7a44b4012 https://scikit-learn.org/stable/modules/generated/sklearn.feature_extraction.text.TfidfVectorizer.html https://zenodo.org/record/6362726#.YsbdSTVBzrk