MaziyarPanahi/Llama-Nemotron-Post-Training-Dataset-v1-Smol-ShareGPT
收藏Hugging Face2025-03-19 更新2025-04-12 收录
下载链接：
https://hf-mirror.com/datasets/MaziyarPanahi/Llama-Nemotron-Post-Training-Dataset-v1-Smol-ShareGPT
下载链接
链接失效反馈官方服务：
资源简介：
Llama-Nemotron-Post-训练-Dataset-v1-Smol-ShareGPT是一个较小的数据集，基于NVIDIA的Llama-Nemotron-Post-训练-Dataset-v1，转换为ShareGPT格式，并合并为一个数据集。它包含原始的所有列，并增加了`messages`字段和`original_split`来追踪来源。数据集的采样方式根据分割的大小来确定，小于1M例子的分割包含100%数据，大于1M例子的分割包含20%数据。
Llama-Nemotron-Post-Training-Dataset-v1-Smol-ShareGPT is a smaller version of NVIDIAs Llama-Nemotron-Post-Training-Dataset-v1, converted to ShareGPT format, and merged into a single dataset. It includes all original columns and adds a `messages` field and `original_split` to track the source. The datasets sampling method is determined by the size of the split, with splits smaller than 1M examples containing 100% of the data, and splits larger than 1M examples containing 20% of the data.
提供机构：
MaziyarPanahi
