Enable HF PretrainedModel loading for speculative model training #122
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
…te from fms directly to extract hidden states; removed unnecessary classes
|
Nice, this does make more sense once models are partitioned into headless/head components! |
|
Couldn't follow the reset logic. Rest everything looks good! |
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Outdated suggestions cannot be applied.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 enables HF PretrainedModel loading. To use this feature, simply set the architecture to "hf_pretrained", and the variant to a huggingface variant (model_id). This was enabled my removing the need to create special adapters, by wrapping a model in a HiddenStatesExtractor (extracts hidden states from base model). With this new wrapper, the adapters and overridden model classes that used include_embeds were not required, as well as generate could be used from fms main