tensor合numpy转换

在PyTorch中，将Tensor转换为NumPy数组有几种方法。一种常用的方法是使用`.numpy()`方法。这个方法将Tensor对象转换为NumPy数组，但是需要注意的是，只有当Tensor对象在CPU上时才能进行转换。因此，如果你的Tensor对象在GPU上，需要先将其转移到CPU上，然后再使用`.numpy()`方法进行转换。具体代码如下所示：

import torch
import numpy as np

# 创建一个Tensor对象
a = torch.tensor([1, 2, 3])

# 将Tensor对象转换为NumPy数组
b = a.cpu().numpy()

# 打印转换后的结果
print(type(a))
print(type(b))
print(a)
print(b)

输出结果为：

<class 'torch.Tensor'>
<class 'numpy.ndarray'>
tensor([1, 2, 3])
<span class="em">1</span><span class="em">2</span><span class="em">3</span>
#### 引用[.reference_title]
-  *1* [Tensor和Numpy互相转换](https://blog.csdn.net/m0_56676881/article/details/126912457)[target="_blank" data-report-click={"spm":"1018.2226.3001.9630","extra":{"utm_source":"vip_chatgpt_common_search_pc_result","utm_medium":"distribute.pc_search_result.none-task-cask-2~all~insert_cask~default-1-null.142^v93^chatsearchT3_1"}}] [.reference_item style="max-width: 33.333333333333336%"]
-  *2* [tensor与numpy转换资源合集](https://download.csdn.net/download/Rocky006/87842415)[target="_blank" data-report-click={"spm":"1018.2226.3001.9630","extra":{"utm_source":"vip_chatgpt_common_search_pc_result","utm_medium":"distribute.pc_search_result.none-task-cask-2~all~insert_cask~default-1-null.142^v93^chatsearchT3_1"}}] [.reference_item style="max-width: 33.333333333333336%"]
-  *3* [pytorch 实现tensor与numpy数组转换](https://download.csdn.net/download/weixin_38668225/13762306)[target="_blank" data-report-click={"spm":"1018.2226.3001.9630","extra":{"utm_source":"vip_chatgpt_common_search_pc_result","utm_medium":"distribute.pc_search_result.none-task-cask-2~all~insert_cask~default-1-null.142^v93^chatsearchT3_1"}}] [.reference_item style="max-width: 33.333333333333336%"]
[ .reference_list ]