A simple LSTM based Statement of Purpose Generator for grad school. :)
- Pytorch 0.4
- Python 2.x / 3.x
Generate samples from the trained model.
python generate.py --data ./data/sop/ --seed 6 # play with seed to get different text
python main.py --cuda --epochs 500 --data data/sop/ --tied # Train a tied LSTM on SOP dataset with CUDA for 500 epochs
The model uses the
nn.RNN module (and its sister modules
which will automatically use the cuDNN backend if run on CUDA with cuDNN installed.
During training, if a keyboard interrupt (Ctrl-C) is received, training is stopped and the current model is evaluated against the test dataset.