Commit History

  1. removed debugging messages now that search is working

    4c7c106
  2. making sure all the correct arguments are there in the function calls

    67412fc
  3. pass IDF DTM to the function factory

    84c4c3d
  4. the factory has to actually return the function

    35bf0ca
  5. corrected the name of the fasttext model binary

    4842291
  6. send row names to be included in the results

    2f234b8

  1. switched to the new search functions using sentence-transformers

    68fd999
  2. added search URL to the README

    1aa0d22
  3. renamed modules that do the search

    c1d8ce6

  1. changes TF-IDF search to cosine similarity from dot product

    ab4ff40
  2. changes TF-IDF search to cosine similarity from dot product

    1310186

  1. switch tfidf search to use file list saved by joblib

    3facea3
  2. reshape with shape, not size

    1cf6271
  3. reduce files list to 1-d

    7b687d4
  4. testing separate query methods

    c70ddc5
  5. use block_embeddings_df from the compressed serialized parquet file

    68239c7
  6. copied encode function directly into search_embeddings.py

    6b6def4
  7. need to specify path to encode because this is a hacky prototype

    88bbcb9
  8. use a pre-serialized dtm_svd

    d503cc1
  9. allow pickle for deserializing data

    4e5f6d2
  10. load the DTM and file list from serialized versions

    21ca93f
  11. don't need to install glob

    17f8024
  12. added scripts for testing inference

    c795cd4