![]() + cp deepmind-qa/* $DATAPATH/deepmind-qa/ The training questions are expected to be in `$DATAPATH/deepmind-qa/cnn/questions/training`. ![]() +Set the environment variable `DATAPATH` to the folder containing the DeepMind QA dataset. +The corresponding dataset is provided by () but if the script does not work (or you are tired of waiting) you can check () by (). + pip install git+git:///mila-udem/blocks-extras.git +Anaconda also includes a Bokeh server, but you still need to install `blocks-extras` if you want to have the plot: + pip install git+git:///mila-udem/blocks.git -r + pip install git+git:///mila-udem/fuel.git + pip install git+git:///Theano/Theano.git +We recommend using () and installing them with the following commands (where `pip` refers to the `pip` command from Anaconda): +* Blocks Extras and a Bokeh server for the plot +Here is an example of attention weights used by the attentive reader model on an example: +We trained the three models during 2 to 4 days on a Titan Black GPU. +- `deep_bidir_lstm_2x128` implements a two-layer bidirectional LSTM reader +- `deepmind_attentive_reader` reproduces the experimental settings of the DeepMind paper for the Attentive reader +- `deepmind_deep_lstm` reproduces the experimental settings of the DeepMind paper for the LSTM reader The corresponding dataset is provided by () but if the script does not work you can check () by (). +The three models implemented in this repository are: This repository also contains an implementation of a Deep Bidirectional LSTM. ![]() This repository contains an implementation of the two models (the Deep LSTM and the Attentive Reader) described in *Teaching Machines to Read and Comprehend* by Karl Moritz Hermann and al., NIPS, 2015. ![]() Update README.md with more precise installation instructions. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |