feature: Embed Service #182
Conversation
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #182 +/- ##
=======================================
Coverage 93.61% 93.61%
=======================================
Files 9 9
Lines 407 407
=======================================
Hits 381 381
Misses 26 26 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
JNygaard-Skylight
requested review from
BradySkylight,
bamader,
m-goggins and
robertandremitchell
as code owners
|
|
||
| def embed(input_text: str) -> Tensor: | ||
| """Embed text.""" | ||
| return model.encode(input_text) |
There was a problem hiding this comment.
This is awesome and simple! Should we update the doc string to be a bit more descriptive? This may only be relevant for maybe our actual APIs though....but just checking.
There was a problem hiding this comment.
Definitely want to add more to the doc string. It's fairly standard to include the parameters and expected output.
|
|
||
| def embed(input_text: str) -> Tensor: | ||
| """Embed text.""" | ||
| return model.encode(input_text) |
There was a problem hiding this comment.
Definitely want to add more to the doc string. It's fairly standard to include the parameters and expected output.
|
|
||
| from dibbs_text_to_code.configs import MODEL_NAME | ||
|
|
||
| model = SentenceTransformer(MODEL_NAME) |
There was a problem hiding this comment.
We likely want to refactor this into a lazy load to make the API and lambda cold start a little faster. Up to you if you want to do that in this PR or another.
There was a problem hiding this comment.
I can tackle this in a different PR with an update to the doc strings (see comment below).
| @@ -0,0 +1,40 @@ | |||
| from dibbs_text_to_code import configs | |||
|
|
|||
There was a problem hiding this comment.
I do think you should add at least one test to show we've returned a tensor object that is the same dimensions as the model expects (768 for Qwen). I could be wrong, but I don't think you'll have to mock anything as long as you load this in the test file from embeddings import embed
There was a problem hiding this comment.
I like this idea as well!
…s docstrings to be on a single continuous line with no returns
Description
Creates a simple "embed" service with a single function.
Related Issues
Closes #157