Tweak ESM tokenizer for Nucleotide Transformer #22770
Merged
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Outdated suggestions cannot be applied.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.
Nucleotide Transformer is a model that takes DNA inputs. It uses the same model architecture as the protein model ESM, but in addition to a different vocabulary it tokenizes inputs without a
<sep>
or<eos>
token at the end. This PR makes a small tweak to the tokenization code for ESM, so that it doesn't try to addself.eos_token_id
to sequences when the tokenizer does not have aneos_token
set. With this change, we can fully support Nucleotide Transformer as an ESM checkpoint.