Vikas Yadav

icon indicating a website link https://sites.google.com/email.arizona.edu/vikasy95/home icon indicating an email [email protected]

icon company University of Arizona icon location Tucson icon location PhD student in Information and Computer Sciences department, University of Arizona, Tucson

Results 6 comments of Vikas Yadav

Retrieving objects for a set or list of URL's in parallel

@wumpus - thanks for the response :D I am trying to retrieve meta data for nearly 10k webpages. I am feeding URL of each webpage one by one to the...

AttLayer has no attribute init

@koelscha - could you solve this problem?? Even I am getting the same error. @richliao - Can you help us by pointing out at a specific function (or line number)...

When will you release this code?

Hi Jin, Thanks, the furnished codes will take few days/weeks as I have deadlines+exams in this month which could cause a delay. I am planning to dump the experimental codes...

Unable to reproduce the result

Hi, Sorry, I dint see this issue until now. We were getting higher performance specifically with FastText Dutch embeddings. The hyperparameters for Dutch were different from other experiments. If required,...

Unable to reproduce the result

Hi, Did you run build_data.py first to generate all the required dictionaries from Spanish CoNLL data? We had used FastText embeddings for Spanish too. It would be helpful if you...

100- dimension Glove embeddings for CoNLL 2003 English data

Hi, You can download the 100D English GloVe word embeddings from here - https://nlp.stanford.edu/projects/glove/ Particularly, this line - (6B tokens, 400K vocab, uncased, 50d, 100d, 200d, & 300d vectors, 822...