update data description docs

.gitignore

@@ -1,12 +1,21 @@
 # temp
 temp*
+
 # run files
 *__pycache__*
 
 # to large to be uploaded
-data/corpus-title.tar.gz
+data/corpus-title.tar.gz                                # raw data = y_preprocess
+
+data/raw/corpus-title.txt                               # y_preprocess for ANN & Transformer model
+data/processed/corpus-title-no-accent.txt               # X_preprocess for ANN & Transformer model
+data/processed/corpus-title-unicode.txt                 # y for ANN model
+data/processed/corpus-title-no-accent-unicode.txt       # X for ANN model
 
-data/raw/corpus-title.txt
-data/processed/corpus-title-no-accent.txt
-data/processed/corpus-title-unicode.txt
-data/processed/corpus-title-no-accent-unicode.txt
+# models
+models/*.h5
+models/*.json
+models/*.joblib
+models/*.pkl
+models/*.pt
+models/*.pth

README.md

@@ -1,6 +1,12 @@
 # How to reconstruct Vietnamese diacritics
 
-### Data Collection, Processing & Model Training
+## Data Preparation
+
+This project utilizes nearly 10 million Vietnamese sentences collected by the research team at [news-corpus](https://github.com/binhvq/news-corpus).
+
+To download the dataset (corpus-title.txt - 578MB), please visit the following link: [Google Drive](https://drive.usercontent.google.com/download?id=1ypvEoGRNWrNLmW246RtBm9iMyKXm_2BP&export=download&authuser=0).
+
+## Data Collection, Processing & Model Training
 
 ```bash
 # Generate text without diacritics from raw text
@@ -15,7 +21,7 @@ nohup python main.py collection > logs/data_collection_0_0_1.log 2>&1 &
 nohup python main.py processing > logs/data_processing_0_0_1.log 2>&1 &
 # Processed 9,487,416 samples. Saved: X -> X_transformer.pt, y -> y_transformer.pt
 
-# Train the model: memory usage stabilized below 30GB, 3 hours and 10 minutes per epoch.
+# Train the model: memory usage stabilized below 30GB.
 # The first epoch took approximately 30 hours.
 nohup python main.py building > logs/model_building_0_0_1.log 2>&1 &
 # Data loaded: X -> torch.Size([9487416, 150]), y -> torch.Size([9487416, 150])
@@ -28,7 +34,7 @@ nohup python main.py building > logs/model_building_0_0_1.log 2>&1 &
 
 ---
 
-### Reconstructing Vietnamese Diacritics
+## Reconstructing Vietnamese Diacritics
 
 Below are the demo results using the model trained for one epoch (approximately 30 hours of training).
 - Training dataset: 8 million sentences.

src/use_app/use_app.py

@@ -1,5 +1,4 @@
 import torch
-import unidecode
 from src.data_collection.data_no_accent import remove_vietnamese_accent
 from src.model_building.model import Transformer