Conversion script bugfixes #1218
Merged
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
|
I think it would be more desirable to preserve the PP=1 -> PipelineModule with one pipeline stage behavior and have pinged the PR creator on Discord to discuss. |
|
should #1214 get closed and pulled into this? |
|
@dmahan93 yep! I agree with Stella that going back to the behavior of pp=1 operating as a PipelineModule is desirable though. |
Quentin-Anthony
approved these changes
Jun 7, 2024
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Outdated suggestions cannot be applied.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.
Updates the NeoX-to-HF conversion utilities to fix the following problems:
is_pipe_parallelbehavior s.t. PP=1 models no longer are trained using PipelineModules, since MoE is not PP-compatible. This affects the save format of such checkpoints, so bringing the conversion script is_pipe_parallel detection logic in line with this.KeyError: no key 'final_linear.weight' could be foundwhen trying to convert a model with tied input/output embeddings. So adjusted the script to now save input embeds into the LM head state when converting a tied embedding model.