LLamaCausalLM add support for tokenizer.json #9967
Open
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
sed -i '' 's/\r$//' .gitignore (apparently I had a couple in there?)
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Outdated suggestions cannot be applied.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.
SentencePiece by default uses BPE, which by default also uses a tokenizer.json. This does not have to be customized with tokens you cannot read from tokenizer.model, but as the BSC-LT/salamanda-7b and related models show, it can be. modified the LlamaModel class to augment the vocabulatry from the json file if it is present.
related issue: #9899
note: this was recreated because I made the mistake of force pushing the other day to pull the fork and update my PR branch, leading to all changes in master to look like they were being reapplied... this cost a like by LIN72H on the PR :( I'll not do it again.