Up to higher level directory | |||
Name | Date | Size | |
---|---|---|---|
BUILD | 21-Aug-2018 | 659 | |
README.md | 21-Aug-2018 | 965 | |
rnn_colorbot.py | 21-Aug-2018 | 12.7K | |
rnn_colorbot_test.py | 21-Aug-2018 | 2.2K |
1 RNN Colorbot: An RNN that predicts colors using eager execution. 2 3 To train and generate colors, run: 4 5 ``` 6 python rnn_colorbot.py 7 ``` 8 9 This example shows how to: 10 1. read, process, (one-hot) encode, and pad text data via the 11 Datasets API; 12 2. build a trainable model; 13 3. implement a multi-layer RNN using Python control flow 14 constructs (e.g., a for loop); 15 4. train a model using an iterative gradient-based method; and 16 5. log training and evaluation loss for consumption by TensorBoard 17 (to view summaries, use: tensorboard --log_dir=<dir>/summaries). 18 19 The data used in this example is licensed under the Creative Commons 20 Attribution-ShareAlike License and is available at 21 https://en.wikipedia.org/wiki/List_of_colors:_A-F 22 https://en.wikipedia.org/wiki/List_of_colors:_G-M 23 https://en.wikipedia.org/wiki/List_of_colors:_N-Z 24 25 This example was adapted from 26 https://github.com/random-forests/tensorflow-workshop/tree/master/extras/colorbot 27