Add config for rotary 350M oscar #5
Merged
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
thomasw21
force-pushed
the
rotary-debug
branch
from
bf51bb7 to
a836be9
Compare
thomasw21
changed the title
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Outdated suggestions cannot be applied.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.
We suspect that the implementation for rotary is wrong, causing some of the following observations:
Or responsible for the implementation: bigscience-workshop/Megatron-DeepSpeed#7
Tensorboard for 1B3 run: https://huggingface.co/bigscience/tr4-1B3-rotary-tensorboard
Default implementation for rotary: https://github.com/EleutherAI/gpt-neox/blob/main/megatron/model/positional_embeddings.py
We run 350M on oscar using similar config as tr3h.