The current state-of-the-art for single-card training of models is at what level? 目前给出的单卡训练模型方案大概是什么水平? #3210
ryan-utopia
started this conversation in
Community | General
Replies: 1 comment
-
Hi @JacketChenlll There is no standard answer to this as there are many influencing factors such as hardware configuration, model size, parameter configuration, dataset size, batchsize, etc. |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
For example, how long would it take to train a model (GPT-2 or PaLM) using a single A100 card? 比如使用一张A100训练的话大概要训练多久呢?
Beta Was this translation helpful? Give feedback.
All reactions