Skip to content

search: add BERT cosine similarity#7

Merged
mweiden merged 3 commits intomainfrom
codex/update-search-index-to-use-cosine-similarities
Aug 10, 2025
Merged

search: add BERT cosine similarity#7
mweiden merged 3 commits intomainfrom
codex/update-search-index-to-use-cosine-similarities

Conversation

@mweiden
Copy link
Copy Markdown
Owner

@mweiden mweiden commented Aug 10, 2025

Summary

  • replace TF-IDF ranking with cosine similarity over BERT embeddings
  • inject embedding model into InvertedIndex and expose top_k
  • adapt Flask search route and tests

Testing

  • make lint
  • make test (fails: selenium.common.exceptions.NoSuchDriverException: Unable to obtain driver for chrome)
  • pytest src/search/tests -q

https://chatgpt.com/codex/tasks/task_e_68981d0ce1888324ad8411443be76a4d

@mweiden mweiden merged commit 34c796e into main Aug 10, 2025
2 checks passed
@mweiden mweiden deleted the codex/update-search-index-to-use-cosine-similarities branch August 10, 2025 05:19
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant