You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Tell us about your request. Provide a summary of the request and all versions that are affected.
The cosine similarity metric for k-NN is computed as (a . b)/(|a|*|b|) where "." means dotproduct and || means the magnitude of the vector. If a or b is all zero's, this would lead to a divide by 0 exception. So, in opensearch-project/k-NN#1501, we are rejecting requests for indices that use cosine similarity if the vector is 0.
For this, we need to update the documentation to tell users that we will reject these requests. I think the best place to do this would be here:
What do you want to do?
Tell us about your request. Provide a summary of the request and all versions that are affected.
The cosine similarity metric for k-NN is computed as
(a . b)/(|a|*|b|)
where "." means dotproduct and || means the magnitude of the vector. If a or b is all zero's, this would lead to a divide by 0 exception. So, in opensearch-project/k-NN#1501, we are rejecting requests for indices that use cosine similarity if the vector is 0.For this, we need to update the documentation to tell users that we will reject these requests. I think the best place to do this would be here:
What other resources are available? Provide links to related issues, POCs, steps for testing, etc.
Related to this PR: opensearch-project/k-NN#1501.
The text was updated successfully, but these errors were encountered: