Small fixes when resuming training #245
Merged
+77
−19
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
Member
NouamaneTazi
commented
Nov 19, 2024
•
NouamaneTazi
commented
- adapted config to be able to resume training with nanotron
- made sure optimizer states reshard correctly when going from TP=1 to TP>1
- fixed optimizer states dtype when loading from checkpoint (load directly in fp32 instead of loading in bf16 then recasting to fp32)
- fix small bug where LR is set to initial_lr only in 1st step instead of resuming from last lr (never trust pytorch too much)
- small fix in case of resuming training with DP>1 and TP>1
- fix loading non sharded optim states
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Outdated suggestions cannot be applied.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.