Bangla Sentence Correction and Autocompletion with Sequence-to-Sequence Model
Implementation
Step 1
Extract dataset.zip located in the data folder, and make sure linesb.txt file is in the data folder.
Step 2
Run process-data.py
Step 3
Run make-dataset.py, located in data folder.
Step 4
Run train.py, it will train the model and it would take some time.
Step 5 (Optional)
If you want to see the graphs of training time, loss graph and accuraccy graph. Then,
Run make-graph.py
It will make graph and save it in graph folder.
Step 6 (Optional)
If you want to interfere with the model by inputing invalid sentence, and want to see the corrected sentence. Then,
Run the interference.py