huggingface/lighteval

[BUG] Optimize tokenization

Aberta

#732 aberto em 15 de mai. de 2025

 (1 comentário) (0 reação) (1 responsável)Python (514 forks)auto 404
buggood first issuehelp wantedscience-team

Métricas do repositório

Stars
 (2.496 estrelas)
Métricas de merge de PR
 (Métricas PR pendentes)

Description

Describe the bug

For now tokenization is bing made in a for loop, making the whole process very expensive for large benchmarks.

To Reproduce

run any big benchmarks

Expected behavior

Use batch tokenize to speed it up.

Version info

Please provide your operating system, lighteval version or commit if you installed from main, and pip/conda environment if your problem concerns dependencies.

Guia do colaborador