Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Doc2Vec defaults shouldn't override Word2Vec defaults without good reason #795

Closed
gojomo opened this issue Jul 17, 2016 · 0 comments
Closed
Labels
difficulty easy Easy issue: required small fix wishlist Feature request

Comments

@gojomo
Copy link
Collaborator

gojomo commented Jul 17, 2016

Following #534, Word2Vec matches the word2vec.c defaults (as of the time Google stopped updating/maintaining that code). However, where our Doc2Vec class takes its own parameters – hs, negative, sample, window, and dm_mean – it's still using the older defaults. There's not really any case that these older defaults are better – and in some cases they're probably worse.

Unless there's a reason for different defaults, Doc2Vec should use the same values as Word2Vec... ideally by not specifying them at all (to avoid duplication). This probably means removing those params from the __init__() signature – but still documenting them in the doc-comment – so that the extra keywords just pass through to Word2Vec.__init__() if/when present.

@tmylk tmylk added wishlist Feature request difficulty easy Easy issue: required small fix labels Oct 5, 2016
markroxor added a commit to markroxor/gensim that referenced this issue Oct 9, 2016
markroxor added a commit to markroxor/gensim that referenced this issue Oct 9, 2016
more parameters excluded
markroxor added a commit to markroxor/gensim that referenced this issue Nov 8, 2016
markroxor added a commit to markroxor/gensim that referenced this issue Nov 8, 2016
more parameters excluded
@tmylk tmylk closed this as completed in 14f12f4 Nov 22, 2016
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
difficulty easy Easy issue: required small fix wishlist Feature request
Projects
None yet
Development

No branches or pull requests

2 participants