keras-language-modeling icon indicating copy to clipboard operation
keras-language-modeling copied to clipboard

Published 20 hours ago • verified publisher icon codekansas

models

Open heri opened this issue 8 years ago • 3 comments

When running insurance_qa_eval.py, I have :

  File "/Users/heri/anaconda/lib/python2.7/site-packages/numpy/lib/npyio.py", line 362, in load
    fid = open(file, "rb")
IOError: [Errno 2] No such file or directory: 'models/word2vec_100_dim.h5'

heri avatar May 04 '16 14:05 heri

I'm still trying things out. In insurance_qa_eval.py I loaded some pre-trained embeddings, but I haven't put the embeddings on Github yet. To generate them, I trained Gensim's Word2Vec model to augment the weights after training an EmbeddingModel. When I feel like I've got good embeddings I'll upload them.

codekansas avatar May 04 '16 15:05 codekansas

Thanks

I'm following the blog post and interested in integrating this in a bot I'm working on.

Currently I'm using a retrieval-based system (with rules and using Yelp API to get responses), but I'm logging all the positive answers (with respective user questions), so if I get enough data (25,000 answers?) I might be able to use keras

heri avatar May 04 '16 16:05 heri

Neat! Let me know how it turns out. Originally I was hoping to use this framework for Trec. I think the deep learning part would probably work better if there was more data.

codekansas avatar May 04 '16 16:05 codekansas