Support for Seq2Seq Models (T5, T5Gemma, etc.) #3153
Draft
+27
−8
Conversation
This file contains hidden or 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
|
Hey @maxzuo thanks for the contribution |
|
@Datta0 sure I'm actively working on it, actually why I converted this to a draft. Will let you know! |
|
@maxzuo did u work on it? |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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Outdated suggestions cannot be applied.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.
PR Description
Adds support for Seq2Seq models:
AutoModelForSeq2SeqLM.Why
Seq2Seq models are not directly supported, despite support for all model architectures. This is because
FastModel.from_pretrainedsets theauto_modelparameter to eitherAutoModelForCausalLMorAutoModelForVision2Seq/AutoModelForImageTextToText.Further, since models like T5 have class names ending in
ForConditionalGeneration, unsloth registers this as a VLM and tries to load it as such.I use
AutoModelForSeq2SeqLM._model_mappingto check if a model config is registered as a Seq2Seq model. This logic can be extended to other auto models (e.g.,AutoModelForSequenceClassification) if desired.Links
Support for T5 has some community interest: