-
Notifications
You must be signed in to change notification settings - Fork 3
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Miscellaneous fixes to the x-transformers implementation #79
Open
Waino
wants to merge
8
commits into
main
Choose a base branch
from
feat/model_blowout
base: main
Could not load branches
Branch not found: {{ refName }}
Loading
Could not load tags
Nothing to show
Loading
Are you sure you want to change the base?
Some commits from the old base branch may be removed from the timeline,
and old review comments may become outdated.
Commits on Oct 7, 2024
-
Skip backward if loss is NaN. Stop training if enough batches are skipped.
Configuration menu - View commit details
-
Copy full SHA for 5a73b4a - Browse repository at this point
Copy the full SHA 5a73b4aView commit details -
Configuration menu - View commit details
-
Copy full SHA for 17b6ced - Browse repository at this point
Copy the full SHA 17b6cedView commit details -
Configuration menu - View commit details
-
Copy full SHA for 4f6620c - Browse repository at this point
Copy the full SHA 4f6620cView commit details -
Configuration menu - View commit details
-
Copy full SHA for 7229141 - Browse repository at this point
Copy the full SHA 7229141View commit details -
Bugfix: Statisics inherits n_correct from previous instance
The default value must be either zero or None, depending on whether accuracy is reported or not.
Configuration menu - View commit details
-
Copy full SHA for 203d4d5 - Browse repository at this point
Copy the full SHA 203d4d5View commit details -
Configuration menu - View commit details
-
Copy full SHA for 83c8c26 - Browse repository at this point
Copy the full SHA 83c8c26View commit details
Commits on Oct 14, 2024
-
Distributed component for TransformerWrapper
Parameters in the TransformerWrapper, e.g. to_logits, need their own distributed component and optimizer.
Configuration menu - View commit details
-
Copy full SHA for 97bc2d9 - Browse repository at this point
Copy the full SHA 97bc2d9View commit details -
The adapter injection code was causing parameter duplication. Another issue: to normalize or not to normalize? We compute a normalization based on either tokens or sents, but never apply it. The effect can be compensated for using the learning rate, as long as batches are approximately the same size. Too high learning rates lead to gradient clipping, which is extra detrimental because each component is individually clipped. Clipping deterministically requires one of the following: - access to gradients for all parameters of the entire model (infeasible) - component local clipping (current approach) - communicating a clipping factor across devices (maybe we should do this?)
Configuration menu - View commit details
-
Copy full SHA for 3a16fa0 - Browse repository at this point
Copy the full SHA 3a16fa0View commit details
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.