Conchylicultor/DeepQA

Can't change the corpus

Worrah opened this issue · 2 comments

Good day to you.
In some way, i can't change corpus from default. What am i doing wrong?

D:\Bots\DeepQA>python main.py --corpus lightweight
Welcome to DeepQA v0.1 !

TensorFlow detected: v1.4.0

Warning: Restoring parameters:
globStep: 1978
watsonMode: False
autoEncode: False
corpus: cornell
datasetTag:
maxLength: 10
filterVocab: 1
skipLines: False
vocabularySize: 40000
hiddenSize: 512
numLayers: 2
softmaxSamples: 0
initEmbeddings: False
embeddingSize: 64
embeddingSource: GoogleNews-vectors-negative300.bin

Training samples not found. Creating dataset...

use the switch --modelTag

Check readme for more options. It's mentioned under the heading "Running".

firstly,i Download english corpus directly from here: http://opus.lingfil.uu.se/download.php?f=OpenSubtitles/en.tar.gz
and i unpack the archive in\DeepQA-master\data\opensubs,
last, i run main.py --corpus opensubs, but it show that "Loaded opensubs: 4 words, 0 QA"
what is wrong with me?