scientific-guide-notebooks

scientific-guide-notebooks is a collection of machine learning and deep learning notebooks.

Archive structure

.
├── Cycle_GAN                 - Converting Horses into Zebras and vice versa, 
implementing CycleGAN and its residual blocks.
├── Evaluating_Unbiasing_GANs - Exploring two GAN Evaluation metrics and investigating
the possible biases that a GAN might have.
├── MNIST_GANs                - Generating hand-written images of digits, exploring 
different kind of GANs and GAN techniques on MNIST dataset.
├── Pix2Pix                  - Converting aerial satellite imagery into map routes, 
exploring Pix2Pix GAN architecture and its generator U-Net.
├── chatbot                  - Two Chatbots Dialogue, Reformer: The Efficient 
Transformer Model, trained on MultiWoz dataset
├── machine-translation      - English-to-German Translator, LSTM Model with attention, 
trained from scratch on opus/medical dataset
├── question-answering       - Context-Based Question Answering, pretrained T5 Model, 
fine-tuned on Stanford Question Answering Dataset (SQuAD)
├── text-summarizer          - Abstractive Text Summarizer Model, Transformer Language 
Model, trained from scratch on CNN/DailyMail non-anonymized summarization dataset
└── 

Libraries Used

• Trax — Deep Learning with Clear Code and Speed library, maintained by Google Brain team.
• PyTorch - Tensors and Dynamic neural networks in Python with strong GPU acceleration

Note

• If you are unable to load one of the notebook in this repository, please copy then paste its link in nbviewer.
• To clone this repository, you may need to use (git-lfs)[https://git-lfs.github.com/] in order to retrieve files that exceed 100MB or download this files directly. In general, for large files, we provide a smaller version.

License

This project is licensed under the terms of the MIT license. See the LICENSE file.