.
├── data
│    ├── shakespeare  
│    │   ├── alllines.txt.txt                         # All Shakespeare Lines 
│    │   ├── clean_modern.txt                         # Modern Shakespeare lines before tokenizing
│    │   ├── clean_original.txt                       # Original Shakespeare lines before tokenizing
│    │   └── processed
│    │        ├── all_proc.txt                        # All Lines after cleaning and tokenizing
│    │        ├── modern_proc.txt                     # Modern Shakespeare lines after tokenizing
│    │        └── original_proc.txt                   # Original Shakespeare lines after tokenizing
│    └── trump
│    │        ├── speeches
│    │        │    ├── cln_speech.txt
│    │        │    └── speech.txt
│    │        └── tweets 
│    │             ├── cln_tweets.txt 
│    │             ├── tweets.json.tx
│    │             └── tweets_01-08-2021.csv.txt
└──src                    
     ├── baseline 
     │    └── baseline.py                              # Creates baseline model
     ├── cleaning   
     │    ├── shake_cleaning.py                        # Cleans and tokenizes Shakespeare lines
     │    └── trump_cleaning.py                        # Cleans and tokenizes Trump tweets and speeches
     └── model_hw2
         ├── data
         │    ├── s_to_t
         │    │    ├── shake_fake_1                     # Pseudo Shakespeare data produced by IBT
         │    │    ├── shake_fake_2                     # Pseudo Shakespeare data produced by IBT
         │    │    └── shake_fake_3                     # Pseudo Shakespeare data produced by IBT
         │    ├── t_to_s
         │    │    ├── trump_fake_1                     # Pseudo Shakespeare data produced by IBT
         │    │    ├── trump_fake_2                     # Pseudo Shakespeare data produced by IBT
         │    │    └── trump_fake_3                     # Psuedo Trump data produced by IBT 
         │    ├── temp_combine.txt                      # Temporary total combined data of (source, target) (100%)
         │    ├── temp_dev.txt                          # Temporary combined dev data of (source, target) for training (80%)
         │    ├── temp_train.txt                        # Temporary total combined data of (source, target) (20%)
         │    └── truth
         │        ├── all_proc.txt                     # Copied from data/shakespeare/processed/
         │        ├── modern_proc.txt                  # Copied from data/shakespeare/processed/
         │        ├── original_proc.txt                # Copied from data/shakespeare/processed/
         │        └── trump.txt                        # Combined Trump tweets and speeches
         ├── driver.py                                 # Main file to run IBT process
         ├── layers.py                                 # HW2 Layers for NMT
         ├── output.txt                                # Trainer output for some of the epochs
         ├── transformer.py                            # Modified HW2 Transformer for GPU
         └── saved_models   
             ├── model_ST_1                            # IBT Process Shakespeare to Trump Model --> first
             ├── model_ST_2                            # IBT Process Shakespeare to Trump Model --> best
             ├── model_ST_3                            # IBT Process Shakespeare to Trump Model --> diverge
             ├── model_TS_1                            # IBT Process Trump to Shakespeare Model --> seed
             ├── model_TS_2                            # IBT Process Trump to Shakespeare Model --> best
             └── model_TS_3                            # IBT Process Trump to Shakespeare Model --> seed