Directory structure:

1. Part (a) - Task 1

cbow.ipynb and plot.ipynb were used for this task. cbow.ipynb involves training a neural network model and it's weights were saved in model_weights.h5.

2. Part (a) - Task 2

glove.ipynb was used in this part. The GloVe model was trained in the GloVe directory and it's embeddings were stored in vector.txt

3. Part (b) - Task 1 and Task 2

phrase.ipynb and sentence.ipynb were used for Phrase and Sentence similarity. Note that Sentence similarity model takes time to train.

4. Part (c) - Task 1

transformer.ipynb was used for this. BERT transformer was loaded in this from the internet.

5. Report - Report summarizing the entire task

Dependencies

1. Numpy
2. Matplotlib
3. Datasets
4. Nltk
5. Sklearn
6. Tensorflow
7. Sentence-Transformers
8. Pandas
9. Re
10. Gensim