[models] - chore: improve seq2seq - #85
Closed
JulesBelveze wants to merge 2 commits into
Closed
Conversation
…seq2seq models - Enable beam search with early stopping during the generation process in seq2seq model configurations - Introduce source and target prefixes in data module configurations to support structured input and output for seq2seq tasks - Update the parameter grouping for optimizer to include additional no-decay layer norms and biases across various model modules
…prefix configuration - Changed the retrieval of `source_prefix` and `target_prefix` from using a `get` method to direct attribute access to avoid errors related to missing configuration - Removed fallback to empty strings for `source_prefix` and `target_prefix` to enforce explicit configuration values
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Description
Improves seq2seq model configuration and training by enabling beam search generation with early stopping for better output quality, adding support for source/target prefixes in data modules for structured input/output, and expanding the no-decay parameter list for optimizers to properly handle layer normalization variants across different model architectures (including
layer_norm,layernorm, andln_f).Risk
Low. Changes are configuration improvements that don't alter core model behavior.
Tests
Manual testing with updated configurations.