Added prompting along with e5 instruct #888
Merged
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
All encode calls now include a prompt_name (following the variable name in sentence transformer's encode), which provided the task name. The model can then use a custom prompt pr. task if they wish (or encode based on e.g. task type). Additionally to test it out I have also added the two e5 instruct models.
KennethEnevoldsen
requested review from
imenelydiaker,
Muennighoff and
x-tabdeveloping
13 tasks
Co-authored-by: Niklas Muennighoff <n.muennighoff@gmail.com>
Muennighoff
reviewed
Jun 13, 2024
There was a problem hiding this comment.
Sorry for the delay! It looks great tho would love if @orionw could have a quick look whether or not this aligns with Retrieval w/ Instructions 🙌
|
Looks great @KennethEnevoldsen! Using the same as Retrieval for InstructionRetrieval is great and it will be nice for all the many models using it these days. For many tasks in InstructionRetrieval the models did better with the prefix and the instruction so I added them both manually, glad to have it be part of this. Thanks for the ping @Muennighoff |
Co-authored-by: Niklas Muennighoff <n.muennighoff@gmail.com>
10 tasks
This was referenced Jun 26, 2024
Closed
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Outdated suggestions cannot be applied.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.
All encode calls now include a prompt_name (following the variable name in sentence transformer's encode), which provided the task name. The model can then use a custom prompt pr. task if they wish (or encode based on e.g. task type).
Additionally to test it out I have also added the two e5 instruct models.
Checklist
make test.make lint.Adding a model checklist
mteb.get_model(model_name, revision_id)andmteb.get_model_meta(model_name, revision_id)