Dataset LLMs & Texto

ibm-research/ScarfBench

Dataset em destaque no Hugging Face — 215 downloads. Scarf (Self-Contained Application Refactoring) is a benchmark suite for evaluating AI agents' ability to migrate enterprise Java applications across J…

Hugging Face · Datasets ·ibm-research · ·↓ 215 ·♥ 16

O dataset ibm-research/ScarfBench está entre os destaques do Hugging Face — dados que alimentam o treinamento e a avaliação dos modelos do momento.

  • 215 downloads
  • 16 curtidas

Sobre o dataset

Scarf (Self-Contained Application Refactoring) is a benchmark suite for evaluating AI agents' ability to migrate enterprise Java applications across Jakarta EE, Quarkus, and Spring while preserving functionality, idiomatic patterns, and architectural integrity.

text-generation code benchmark evaluation java code-translation agentic

Explorar o dataset no Hugging Face →

compartilhar: