InsultedByMathematics/llama3-ultrafeedback-armo-test-evaluation-rewards-logprobs
收藏数据链接：
官方服务：
资源简介：
该数据集包含多个与微调响应相关的字段，如微调响应、提示、奖励值、选择的响应、拒绝的响应等。数据集主要用于测试偏好，包含1563个示例，总大小为124402784字节。
This dataset contains multiple fields related to fine-tuned responses, such as fine-tuned responses, prompts, reward values, chosen responses, rejected responses, etc. The dataset is primarily used for testing preferences, containing 1563 examples with a total size of 124402784 bytes.
