Skip to content

Commit c723f1e

Browse files
committed
remove original proposal
1 parent dbbe1cc commit c723f1e

File tree

1 file changed

+0
-5
lines changed

1 file changed

+0
-5
lines changed

README.md

Lines changed: 0 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -7,8 +7,3 @@ My MSc thesis on Embeddings Monitoring
77
![](https://raw.githubusercontent.com/theofpa/embeddings-store-thesis/main/images/compacte-logo.jpg)
88

99
A Dissertation submitted in partial fulfillment of the requirements for the degree of Master of Science.
10-
11-
Abstract
12-
==========
13-
14-
Embeddings are taking the AI world by storm. Word embeddings enabled Google’s BERT model, embeddings lie behind Google Translate and other ad- vances in NLP, while graph embeddings have enabled breakthroughs in fighting financial crime. Embeddings are challenging to operationalize. If you change how they are computed, you need a new version of them. In a feature store, this means you may need to retrain all training datasets that use that embedding as a feature. Embeddings may also be published to an embeddings store for similarity search (find me the closest items to ‘X’). In this thesis, I will work on adding MLOps support for computing embeddings, and adding orchestration support in Hopsworks for automating the re-computation of derived features and training datasets when embeddings are re-computed.

0 commit comments

Comments
 (0)