python 矩阵之上11 OnMatrix1-tensor的gpu和cpu转换

一个热身实验,熟悉一下pytorch的tensor。

普通的计算都是发生在cpu里的，如果数据在cpu中，那么可以通过 x.numpy()方法将tensor转为numpy
要调用gpu计算要 x.cuda(), 如果还要回到内存要先.cpu
除chartensor外所有tensor都可以转换为numpy

import torch as tor
import numpy as np 

a = tor.ones(5)
a
tensor([1., 1., 1., 1., 1.])


a.device
device(type='cpu')

a1 = a.cuda()
a1
tensor([1., 1., 1., 1., 1.], device='cuda:0')

b = a.numpy()
b
array([1., 1., 1., 1., 1.], dtype=float32)

# 要先cpu()才行
b1 = a1.numpy()
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
 in 
----> 1 b1 = a1.numpy()

TypeError: can't convert cuda:0 device type tensor to numpy. Use Tensor.cpu() to copy the tensor to host memory first.

查看显存（这就花了830M? 应该是有有些固定开销。)

!nvidia-smi
  
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 460.73.01    Driver Version: 460.73.01    CUDA Version: 11.2     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  GeForce RTX 306...  Off  | 00000000:01:00.0 Off |                  N/A |
|  0%   45C    P8    14W / 200W |    830MiB /  7979MiB |      0%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
+-----------------------------------------------------------------------------+