相关数据集
GPT-3 Training Traces
该数据集是在训练各种GPT-3模型变体时收集的追踪数据,旨在评估模型的表现和执行行为。这些追踪数据是使用PyTorch Kineto工具收集的,包含了在不同并行策略下的详细执行分解。该数据集涵盖了从150亿到1750亿参数规模的模型训练,其任务是对大规模语言模型训练的性能建模与估计。
arXiv270
The Importance of Being Constrained - Dataset
This repository contains all of the raw datafiles used in the creation of the paper 'The Importance of Being Constrained' File explanation: Because of the large amounts of data, it has been turned int
Mendeley Data2024-05-10 更新60
open-llm-leaderboard-old/details_allknowingroger__mergekit-slerp-zplzqvn
--- pretty_name: Evaluation run of allknowingroger/mergekit-slerp-zplzqvn dataset_summary: "Dataset automatically created during the evaluation run of model\ \ [allknowingroger/mergekit-slerp-zplzqv
Hugging Face2024-04-10 更新180
Results on paired t-test for statistical significance.
“H” represents the hypothesis that there is no difference in performance between S2S and convi6.
Figshare2021-03-01 更新40



