Author: Shuvendu Roy Bikash Dept of CSE, KUET
Predicting next Bangla char or word from the previous text written using LSTM in tensorflow
Multi-layer Recurrent Neural Networks (LSTM, RNN) for character-level language models in Python using Tensorflow.
Inspired by Sherjil Ozair's char-nn which is a tensorflow implementations of the project of from Andrej Karpathy's char-rnn.
- Tensorflow 1.0 or greater
To train with default parameters on the tinyshakespeare corpus, run python train.py. To access all the parameters use python train.py --help.
you might want to specify the data directory using --data_dir "directory"
To sample from a checkpointed model, python sample.py --prime "input_text".
To get one word instead of a sentence python word_predict.py "input_text"
You can use any plain text file as input. But is must be it utf-8 format to represent bangla with out breakdown
Then start train from the top level directory using python train.py --data_dir=./data/sherlock/
please be aware that depending on your console settings you may not see bangla in desired font. Even it may be in machine language. In that case you have to show it in some other GUI window or some other python editor may show you in perfect front. Just be aware the the prediction is in uni-code (utf-8) format. so the word is there, how it will show will depent on the font renderer of your system
To visualize training progress, model graphs, and internal state histograms: fire up Tensorboard and point it at your log_dir. E.g.:
$ tensorboard --logdir=./logs/Then open a browser to http://localhost:6006 or the correct IP/Port specified.