Taywon/alpaca_Llama-3.1-8B-Instruct_octopus_paraphrased_greedy
收藏数据链接：
官方服务：
资源简介：
该数据集包含了一系列的文本对，每一对由原始提示（prompt）和对应的改写响应（paraphrased_response）组成。此外，每个样本还包含了模型的参数配置信息，如backend、batch_size等，以及指令（instruction）、输入（input）、原始响应（original_response）和评分（score）。数据集分为训练集，共有约2万个示例。
The dataset consists of a series of text pairs, each pair includes an original prompt and its corresponding paraphrased response. Additionally, each sample contains information about the models parameter configurations, such as backend, batch_size, etc., as well as instruction, input, original response, and score. The dataset is split into a training set with approximately 20,000 examples.
提供机构：
Taywon
