facebookresearch / facebookresearch/blt

Why do you require that first patch should have one non-boe token?

Open
#129 0 comments 0 reactions 0 assignees View on GitHub
Dominant language
Python
Stars
2.1k
Forks
196
PR merge metrics
No merged PRs in 30d

Description

Why do you impose the following requirement: "First patch should have one non-boe token"?

See the assert statement here:
https://github.com/facebookresearch/blt/blob/main/bytelatent/model/blt.py#L126

Thank you for the paper and code!

Contributor guide

Open the contributing guide

Research direction

Start at bytelatent/model/blt.py line 126 and read the surrounding assertion. Compare that requirement with the paper and the issue's question; done means documenting why the first patch must contain one non-boe token, or identifying what needs clarification.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning
Issue type
Documentation
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.