Some basic ML models

Under/ image-transfer we can find a basic model that transfers style from one image to another.

• Results :
  

rnn-text/lstm-passangers.py

• Results :
• Blue is actual data, red is our training set, green is the test set.

lstm-nlp/imdb-reviews.py

• LSTM model that predicts if a review is positive or negative.
• Trained on the infamous IMDB dataset.
• Best Results :

•  Test loss: 0.279
   Test accuracy: 88.455%
  

transformsers/train.py

• Creating similar text to the one that its been trained on(in this case- tiny shakespeare dataset)
• Model overfitting after certain amount of training.(10k iterations ~40 minutes)
• Generated text does look similar to the dataset, but not really english.
•