Practice jupyter notebooks for RNN, encoder-decoder system, attention, LSTM staff implemented in pyTorch.

• 1. PyTorch Basics
• 2. Logistic Regression (sklearn/pytorch)
• 3. Deep Learning Training Workflow
• 4. Building a Bag-of-Words Model
• 5. Document Classification with FastText
• 6. RNN/CNN based Language Classification
• 7. Intrinsic Evaluation of Word Vectors (GloVe/FastText)
• 8. Language Modeling with KenLM
• 9. Language Modeling with Byte Pair Encoding

• Bag-of-Ngram and Document Classification
• RNN/CNN based Natural Language Inference

(for MacOS)

1. Download and install conda (Python3.6)

   • XCode is assumed
   • run bash [installer_file_that_ends_with.sh]
   • run conda list to confirm that installation succeeded

2. Setup conda environment and install jupyter notebook

   $ conda create -n learn_nlp python=3.6
   $ conda activate learn_nlp
   $ conda install jupyter notebook matplotlib scikit-learn
   $ conda install -c conda-forge jupyterlab  # very helpful
   

3. Install pyTorch

   conda install pytorch torchvision -c pytorch