NCE-CNN-Tensorflow
Getting Started
1.
Please install the tensorflow library
2.
Checkout our repo:
git clone https://github.com/whaleloops/NCE-CNN-Tensorflow.git
3.
Copy and paste codes into this repo: https://github.com/castorini/NCE-CNN-Torch
4.
Using following script to download and preprocess the Glove word embedding:
$ sh fetch_and_preprocess.sh
Please make sure your python version >= 2.7, otherwise you will encounter an exception when unzip the downloaded embedding file.
5.
TODO: add more
Running
To evaluate on the TrecQA raw dataset, with MAX sampling and number of negative pairs as 8, run:
$ python2 PairwiseTrainQA.py --dataset TrecQA --version raw --neg_mode 2 --num_pairs 8
To evaluate on the TrecQA clean dataset, simply change -version to clean. Similarly, if you want to evaluate on the WikiQA dataset, change -dataset to WikiQA (don't need to set the -version). You can also change the -neg_mode and -num_pairs to select different sampling strategies or negative pairs.