dogtooth/uf_Meta-Llama-3.1-8B-Instruct_3_Skywork-Reward-Gemma-2-27B-v0.2
收藏Hugging Face2024-12-14 更新2024-12-21 收录
下载链接：
https://hf-mirror.com/datasets/dogtooth/uf_Meta-Llama-3.1-8B-Instruct_3_Skywork-Reward-Gemma-2-27B-v0.2
下载链接
链接失效反馈官方服务：
资源简介：
该数据集是allenai/open_instruct项目中的Rejection Sampling Dataset，主要用于通过拒绝采样算法对模型生成的文本进行筛选和评分。数据集涉及使用Skywork/Skywork-Reward-Gemma-2-27B-v0.2模型进行文本生成和评分，配置参数包括最大前向批量大小、生成文本的数量、使用的GPU数量等。
This dataset is part of the allenai/open_instruct project, specifically the Rejection Sampling Dataset, which is used to filter and score model-generated text using the rejection sampling algorithm. The dataset involves using the Skywork/Skywork-Reward-Gemma-2-27B-v0.2 model for text generation and scoring, with configuration parameters including the maximum forward batch size, the number of generated texts, the number of GPUs used, etc.
提供机构：
dogtooth
