Thanks to visit codestin.com
Credit goes to github.com

Skip to content
This repository was archived by the owner on Jun 30, 2025. It is now read-only.

Conversation

@mart-r
Copy link
Collaborator

@mart-r mart-r commented May 21, 2025

Adds RelCAT conversion.

Tests are done within:

tests.components.addons.relation_extraction.test_rel_cat_in_model_pack

They involve downloading a RelCAT model from S3 (it's 250MB or so), converting it, adding to model pack, and saving model pack with it. And subsequently loading both the RelCAT model separately as well as loading the whole model pack and ensuring it has a RelCAT addon.

@mart-r mart-r merged commit 3513f11 into main May 22, 2025
4 checks passed
alhendrickson pushed a commit to CogStack/cogstack-nlp that referenced this pull request Jun 24, 2025
* CU-86994t45r: Update RelCAT implementation.

Only use Base_tokenizer where needed / applicable

* CU-86994t45r: Redo init of data paths for RelCAT when loading off disk - with correct tokenizer

* CU-86994t45r: Add comment regarding data paths / tokenizer

* CU-86994t45r: Add method to load base RelExtr component using pre-existing config

* CU-86994t45r: Add legacy converter for RelCAT

* CU-86994t45r: Add RelCAT conversion to overall converter

* CU-86994t45r: Allow RelCAT to be loaded from disk as part of model pack

* CU-86994t45r: Add test for RelCAT in model pack along with legacy conversion

* CU-86994t45r: Add note regarding rel cat conversion tests

* CU-86994t45r: Fix argument name
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants