Add custom prepare_for_training logic to ECD model for LLM encoder adapter initialization
#3874
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
jeffkinnison
requested review from
w4nderlust,
tgaddair,
justinxzhao,
arnavgarg1,
geoffreyangus,
Infernaught and
alexsherstinsky
as code owners
justinxzhao
approved these changes
Jan 10, 2024
arnavgarg1
approved these changes
Jan 11, 2024
jeffkinnison
changed the title
prepare_for_trianing logic to ECD model for LLM encoder adapter initializationprepare_for_training logic to ECD model for LLM encoder adapter initialization
vijayi1
pushed a commit
to vijayi1/ludwig
that referenced
this pull request
Jan 23, 2024
…adapter initialization (ludwig-ai#3874)
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Outdated suggestions cannot be applied.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.
The LLM model type initializes the adapter weights and quantization at training time using
LLM.prepare_for_training. WhenLLMEncoderwas added, ECD model did not have a correspondingprepare_for_trainingmethod, so adapter initialization occurred at encoder initialization. This PR addsECD.prepare_for_training, which bringsLLMEncoderadapter initialization to parity with LLM models.