Model Caching #5
Draft
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
CalebCourier
commented
Apr 22, 2024
Comment on lines
+324
to
+328
| def st_embed_function(sources: list[str]): | ||
| print("Running st_embed_function...") | ||
| return DefaultEncodingModel().encode(sources) | ||
|
|
||
| embed_function = st_embed_function |
There was a problem hiding this comment.
We probably don't need st_embed_function anymore as a wrapper. We should be able to just pass DefaultEncodingModel().encode
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR demonstrates how we can properly cache local models during post-install for quick recall in a containerized environment (i.e. docker).
Currently it saves to a relative directory in the root of wherever the validator is installed which isn't ideal. We will need to figure out what directories we can use both locally and in a docker environment that will not require additional permissions (the reason we're not writing to something like
.cache).Beside caching the model weights for quick retrieval, we also introduce a singleton pattern for initializing the instance of the model.
In order to test and demonstrate the above, we switch the default encoding function back to the sentence-transformer model.