Open
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 pull request includes functionality for loading gpt-oss and other Unsloth models using AtomGPT as well as formatting text into the Alpaca or Harmony template depending on which model is being used. The core functionality is an abstract factory, found in inverse_models/factories.py, and it separates gpt-oss loading from other model loading. This strategy resolved many overt and silent bugs found when loading gpt-oss, and training and inference with gpt-oss successfully executes. Interestingly, the text in this screenshot looks as if ChatGPT is reasoning about a crystal structure, further validating that gpt-oss is working as intended. Before merging this pull request, I need to double check that the harmony template is formatted correctly, and I need to do the README tests.
