I tried to realise better cross entropy loss by Triton, but failed, still trying to find the reason…
Update to 3.2, replace self-hosted translation server with Google Gemini 2.0 experimental
I met three great table tennis coaches recently in table tennis room. They have different styles to teach people, one of them are strict but another of them always say good things to their student. last one is patient and really nice, he can taught everyone with different levels with enough patience.
When I have a idea, do it as quick as possible.
Weekly blog is becoming shorter and shorter, not a good trend.
This post is licensed under CC BY 4.0 by the author.
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。