zhaodefeng

any ideas? besides, i can't reproduce methods like ewc, PI and RWalk and i noticed their weight importance seemed to be large in the argparser.py

> Hey, > > Sorry I cannot respond ealier, I'm a busy right now with the writing of my thesis manuscript. > > 1. I'm not sure about your results...

> With --verbose falg, the following message appears that I am connecting remotely via SSH Please help to check it! > > (base) feadre@e2680v2:~sudo sudo (which coolgpus) --speed 99 99...

 

i ask this question mainly because i found some distortions in the output, when i was testing.

后序遍历，自己控制想要返回的东西 ```py class Solution: def reverseKGroup(self, head: Optional[ListNode], k: int) -> Optional[ListNode]: next_iter_head = head new_head = head last_sub_tail = None i = 1 while True: sub_head, sub_tail, next_iter_head =...

我看官方题解下面也有人在讨论DNA这个题，如本书作者所说(下图1)，O(L)的时间复杂度来自于生成子串，可是加入了hash后还是存在生成子串的过程吧(下图2)？所以时间复杂度不是应该还是要考虑L？  

提个小意见，我觉得大家对random选择target那一步有疑问，是因为上面这个图有误导性。比如w[0]=1，其实是{1}这个点被分配了对应给w[0]，再比如w[1]=3，其实是{2,3,4}分配给了w[1]。 而作者的图中，“对区间进行了着色”，再加上例子中刚刚好w[0]=1，让人误解了好像是“[0,1]这个区间对应w[0]”，并且w[0]=1和random中为了取到右边界的而+1的1意义不同，但是数值上都是1，更导致大家不容易理解。 所以是否图中不对区间着色，或者用别的方式指示weight对应的“集合范围”更好呢？

> Hey! > > I think we would need a new kind of taskset, and a new scenario. I'm not super familiar with continual object detection, but I assume it's...

Though it is not an issue, but I like it. Cause it's hard to find a customed theme on the internet lol