facebookresearch/metaseq

Unify tokenizers

Open

#308 建立於 2022年8月20日

在 GitHub 查看
 (1 留言) (1 反應) (1 負責人)Python (6,195 star) (701 fork)batch import
better-engenhancementgood first issue

描述

🚀 Feature Request

With #305, we now have two ways to specify a tokenizer: with the GPT2 tokenizer (provided as two files), and with the universal HF format (specified as one file). These are in two separate code paths, but they don't need to be: we could (manually) merge the two GPT2 files into the universal HF format and switch to only that, and we should.

The resulting code would be cleaner, but catching all the other places the old method is used (e.g. API and old sweeps) needs thorough review.

貢獻者指南

技術棧
python
領域
backendapi
議題類型
refactor
難度面向新貢獻者的預計實作難度，1 表示很小改動，5 表示專家級工作。
3
預計時間有經驗貢獻者完成調查、實作、測試並準備 pull request 的粗略時間範圍。
1-2 days
活動狀態議題目前的可參與程度：新鮮、活躍、陳舊、阻塞或等待維護者輸入。
stale
清晰度議題是否清楚說明預期改動、驗收標準和下一步。
clear
前置要求
Knowledge of tokenizer formatsFamiliarity with HuggingFace tokenizer
新手友善度1-100 的估計分數，表示該議題對首次貢獻者的友善程度。
35
研究方向
Examine PR #305 that introduced the HuggingFace tokenizer format. Locate the two tokenizer code paths in the repository, likely in separate files or modules. Review how tokenization is used in the API and old sweeps to ensure all occurrences are updated to the unified format.
Unify tokenizers · facebookresearch/metaseq#308 | Good First Issue