Skip to content

Latest commit

 

History

History
44 lines (29 loc) · 1.91 KB

README.md

File metadata and controls

44 lines (29 loc) · 1.91 KB
Raw

Mac port

  • in the makefile:

    • replace 'gcc' with 'clang'
    • replace '-Ofast' with '-O2'
    • replace '-Who-unused-result' with '-Wunused-result'

  • where needed in the *.c files, replace '#include <malloc.h>' with '#include <stdlib.h>'

  • intall 'wget' (I use Homebrew, brew install wget)

  • change references to gzip to unzip

I changed the bash files to simply point to sample_data folder and zipped up the text8 sample data file, rather than have it try to download them everytime.

Unzip the text8.zip file in /sample_data before running any of the bash files.

Then run bash demo-word.sh

I also included convertvec (ported to Mac) in the directory /convertvectotext https://github.com/marekrei/convertvec

Tools for computing distributed representation of words

We provide an implementation of the Continuous Bag-of-Words (CBOW) and the Skip-gram model (SG), as well as several demo scripts.

Given a text corpus, the word2vec tool learns a vector for every word in the vocabulary using the Continuous Bag-of-Words or the Skip-Gram neural network architectures. The user should to specify the following:

  • desired vector dimensionality
  • the size of the context window for either the Skip-Gram or the Continuous Bag-of-Words model
  • training algorithm: hierarchical softmax and / or negative sampling
  • threshold for downsampling the frequent words
  • number of threads to use
  • the format of the output word vector file (text or binary)

Usually, the other hyper-parameters such as the learning rate do not need to be tuned for different training sets.

The script demo-word.sh downloads a small (100MB) text corpus from the web, and trains a small word vector model. After the training is finished, the user can interactively explore the similarity of the words.

More information about the scripts is provided at https://code.google.com/p/word2vec/