agu18dec/longhealth-100k-t5gemma-2b-inference
收藏Hugging Face2025-10-17 更新2025-10-25 收录
下载链接：
https://hf-mirror.com/datasets/agu18dec/longhealth-100k-t5gemma-2b-inference
下载链接
链接失效反馈官方服务：
资源简介：
Long Health 100k是一个包含10万医疗/健康样本的数据集，这些样本具有长文本上下文，用于T5-Gemma 2B模型推断。每个样本包含长上下文输入令牌、输出序列令牌、推理掩码、教师模型预测令牌、教师模型的对数几率以及教师令牌的位置索引等信息。
Long Health 100k is a dataset containing 100,000 medical/health samples with long context for T5-Gemma 2B inference. Each sample includes long context input tokens, output sequence tokens, reasoning mask, teacher models predicted tokens, teacher models logits, and position indices for teacher tokens.
提供机构：
agu18dec
