返回全部动态
base-zone-grpo：基于GRPO微调的新模型发布
原标题：sullivan1502/base-zone-grpo
AI 摘要
Hugging Face 上发布了新模型 base-zone-grpo，由用户 sullivan1502 基于 transformers 库微调，使用 TRL 框架和 GRPO 方法训练。该模型可用于文本生成任务，示例展示了如何通过 pipeline 调用。训练基于 DeepSeekMath 论文中提出的 GRPO 算法，相关框架版本已列出。
以上摘要由 AI 生成，可能存在误差。事实请以原文为准。
正文节选
--- library_name: transformers model_name: base-zone-grpo tags: - generated_from_trainer - grpo - trl licence: license --- # Model Card for base-zone-grpo This model is a fine-tuned version of [None](https://huggingface.co/None). It has been trained using [TRL](https://github.com/huggingface/trl). ## Quick start ```python from transformers import pipeline question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?
发布时间：2026-08-13 22:14
抓取时间：2026-08-12 08:07
来源机构：Hugging Face