Simply added test output every epoch as the model trains. You can change the number of characters on lines 131 of train.py. After training, you can resume with --resume. It will stop after 100, so you can use the sample.py script to view output.
Multi-layer recurrent neural networks for training and sampling from texts, inspired by Andrej Karpathy's article and the original torch source code karpathy/char-rnn.
This code is written in Python 3, and it requires the Keras deep learning library.
All input data should be placed in the data/ directory. Sample training texts are provided.
To train the model with default settings:
$ python train.py --input tiny-shakespeare.txtTo sample the model at epoch 100:
$ python sample.py 100Training loss/accuracy is stored in logs/training_log.csv. Model results, including intermediate model weights during training, are stored in the model directory. These are also used by sample.py for sampling.