search

sherjilozair / char-rnn-tensorflow

Multi-layer Recurrent Neural Networks (LSTM, RNN) for character-level language models in Python using Tensorflow
MIT License
2.64k stars 959 forks source link

Continue train #96

Closed woto closed 7 years ago

woto commented 7 years ago

Hi, sorry for lame question. How i should run subsequent (continue from last point) trains? And which files should i save? is it enough 'data' and 'save' directories? Can't test by myself, due to space limit in aws.

Jaspi10 commented 7 years ago

From python3 train.py --help:

--init_from INIT_FROM
                        continue training from saved model at this path. Path
                        must contain files saved by previous training process:
                        'config.pkl' : configuration; 
                        'chars_vocab.pkl' : vocabulary definitions; 
                        'checkpoint' : paths to model
                        file(s) (created by tf). Note: this file contains
                        absolute paths, be careful when moving files around;
                        'model.ckpt-*' : file(s) with model definition
                        (created by tf) (default: None)

All of these are located in your save directory. Which means apart from your data you only need the files in there.

woto commented 7 years ago

@Jaspi10 oh, sorry for the innatantion, thank you!