
boss
@crypto-boss
225 Following
157 Followers
0 reply
4 recasts
54 reactions
0 reply
14 recasts
32 reactions
0 reply
14 recasts
32 reactions
8 replies
12 recasts
190 reactions
45 replies
188 recasts
922 reactions
0 reply
2 recasts
34 reactions
0 reply
6 recasts
36 reactions
0 reply
11 recasts
55 reactions
3 replies
2 recasts
25 reactions
0 reply
9 recasts
52 reactions
0 reply
7 recasts
47 reactions
0 reply
2 recasts
52 reactions
0 reply
3 recasts
46 reactions
0 reply
6 recasts
49 reactions
0 reply
1 recast
34 reactions
0 reply
22 recasts
234 reactions
The differences between the APIs of numpy, cupy and torch are so fascinating....
```
>>> import torch as np
>>> np.arange(10)
tensor([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
>>> a = np.zeros(20)
>>> a[19:-1:-2] = np.arange(10)
```
Torch doesn't let you have ranges that go backwards ๐คฃ
Would love more consistency 19 replies
206 recasts
934 reactions
1 reply
2 recasts
11 reactions
0 reply
4 recasts
29 reactions
0 reply
5 recasts
20 reactions