"pytorch fundamentals" video 13

[nikhilww3]

hello
The tensor ndim is [1,3,3] and in the book shows the 1 dimension is first [ ], but for the first dimension it should be the second [ ]

[Haji-Suleman]

A shape [1, 3, 3] in PyTorch means

1 → first dimension (batch or outermost)

3 → second dimension (rows)

3 → third dimension (columns)

The books might display [rows x columns] without the batch dimension, which appears "swapped" because of this.

You can access properties such as:

X [0 : 1 : 2]

0th batch, 1st row,

So PyTorch is always counting the dimensions from outermost to innermost.

[sergo8online]

@nikhilww3
i'm not sure i understand your quetion, but i'll try to give an answer
in python list [a, b, c] we have elements --> a, b and c
and also we have indexes from 0 to lenght-1
in my example list is [a, b, c] so here is indexes and appropriate elements
0 index ---> a
1 index ---> b
2 index ---> c
so in python if index=0 so it's first element's index, if index=1 it's second's and so on...

P.S.
maybe my english is not so good