5、深入剖析PyTorch DataLoader源码

🕗 发布于 2024-11-21 15:19 线性代数 numpy 矩阵 人工智能 深度学习

文章目录

1. 重要类
2. DataSet
3. DataLoader
4. Python实例

参考大神B站，记录学习笔记
5、深入剖析PyTorch DataLoader源码
其他大神笔记： pytorch数据操作—dataset，dataloader，transform
在这里插入图片描述

1. 重要类

Data Loader
Dataset
Sample
Random Sampler
Sequential Sampler
Batch Sample
Shuffle
torch.randperm
yield from
next,iter
torch.Generator
collate_fn
multi_processing
SingleProcessDataLoaderIter
get_iterator
index_sampler
BaseDataLoaderIter
index_sampler

2. DataSet

DataSet：主要是生成特征和标签对，自定义Dataset需要继承自抽象类dataset,需要实例化dataset的三种方法：__init__,__len__,__getitem__
init: 主要是定义特征features,标签labels的来源，有的特征features是图片，有的是csv格式文件，有的需要对图片进行一些变化，保证最后得到的特征是张量
len: 表示的是整个dataset数据集中的大小
getitem:是最重要的部分，需要形成(特征，标签)对，这样方便后续训练和识别，有训练数据集中的特征，标签的前处理，主要是能够根据dataset[i]返回第i个特征标签对,
$dataset[i]=(feature_i,label_i)$
注：属于训练数据集的预处理阶段，dataset根据服务器的大小进行分块处理，以便能够进行下去

3. DataLoader

DataLoader主要是为了将样本图片和标签一起按照指定的batchsize打包起来，形成一捆一捆的，这样能加速训练，打包中涉及到顺序打包，随机打包，涉及到采样的概念，为了增加程序的鲁棒性，一般会打乱打包；
比如我们的dataset有500个数据，DataLoader中的batchsize=5,那么一个DataLoader[i]中包含5个dataset[i],一共有100个dataloader[i]

4. Python实例

python 代码

#!/usr/bin/env python
# -*- coding:utf-8 -*-
# @FileName  :my_test_label.py
# @Time      :2024/11/19 20:15
# @Author    :Jason Zhang
import numpy as np
import torch
from torch.utils.data import Dataset, DataLoader

np.set_printoptions(suppress=True, precision=2)
np.random.seed(43434)


class MyDataset(Dataset):
    def __init__(self, in_data, in_label):
        self.data = in_data
        self.label = in_label

    def __len__(self):
        return len(self.label)

    def __getitem__(self, item):
        index_data = torch.tensor(self.data[item])
        index_label = torch.tensor(self.label[item])
        return index_data, index_label


if __name__ == "__main__":
    run_code = 0
    my_data = np.linspace(1, 100, 500, dtype=np.float64)
    print(f"my_data=\n{my_data}")
    my_label = np.sin(my_data)
    print(f"my_label=\n{my_label}")
    my_dataset = MyDataset(my_data, my_label)
    my_data_loader = DataLoader(my_dataset, batch_size=5, shuffle=True, drop_last=True)
    i = 0
    for loader in my_data_loader:
        print(f"loader[{i}]={loader}")
        i += 1

结果：

my_data=
[  1.     1.2    1.4    1.6    1.79   1.99   2.19   2.39   2.59   2.79
   2.98   3.18   3.38   3.58   3.78   3.98   4.17   4.37   4.57   4.77
   4.97   5.17   5.36   5.56   5.76   5.96   6.16   6.36   6.56   6.75
   6.95   7.15   7.35   7.55   7.75   7.94   8.14   8.34   8.54   8.74
   8.94   9.13   9.33   9.53   9.73   9.93  10.13  10.32  10.52  10.72
  10.92  11.12  11.32  11.52  11.71  11.91  12.11  12.31  12.51  12.71
  12.9   13.1   13.3   13.5   13.7   13.9   14.09  14.29  14.49  14.69
  14.89  15.09  15.28  15.48  15.68  15.88  16.08  16.28  16.47  16.67
  16.87  17.07  17.27  17.47  17.67  17.86  18.06  18.26  18.46  18.66
  18.86  19.05  19.25  19.45  19.65  19.85  20.05  20.24  20.44  20.64
  20.84  21.04  21.24  21.43  21.63  21.83  22.03  22.23  22.43  22.63
  22.82  23.02  23.22  23.42  23.62  23.82  24.01  24.21  24.41  24.61
  24.81  25.01  25.2   25.4   25.6   25.8   26.    26.2   26.39  26.59
  26.79  26.99  27.19  27.39  27.59  27.78  27.98  28.18  28.38  28.58
  28.78  28.97  29.17  29.37  29.57  29.77  29.97  30.16  30.36  30.56
  30.76  30.96  31.16  31.35  31.55  31.75  31.95  32.15  32.35  32.55
  32.74  32.94  33.14  33.34  33.54  33.74  33.93  34.13  34.33  34.53
  34.73  34.93  35.12  35.32  35.52  35.72  35.92  36.12  36.31  36.51
  36.71  36.91  37.11  37.31  37.51  37.7   37.9   38.1   38.3   38.5
  38.7   38.89  39.09  39.29  39.49  39.69  39.89  40.08  40.28  40.48
  40.68  40.88  41.08  41.27  41.47  41.67  41.87  42.07  42.27  42.46
  42.66  42.86  43.06  43.26  43.46  43.66  43.85  44.05  44.25  44.45
  44.65  44.85  45.04  45.24  45.44  45.64  45.84  46.04  46.23  46.43
  46.63  46.83  47.03  47.23  47.42  47.62  47.82  48.02  48.22  48.42
  48.62  48.81  49.01  49.21  49.41  49.61  49.81  50.    50.2   50.4
  50.6   50.8   51.    51.19  51.39  51.59  51.79  51.99  52.19  52.38
  52.58  52.78  52.98  53.18  53.38  53.58  53.77  53.97  54.17  54.37
  54.57  54.77  54.96  55.16  55.36  55.56  55.76  55.96  56.15  56.35
  56.55  56.75  56.95  57.15  57.34  57.54  57.74  57.94  58.14  58.34
  58.54  58.73  58.93  59.13  59.33  59.53  59.73  59.92  60.12  60.32
  60.52  60.72  60.92  61.11  61.31  61.51  61.71  61.91  62.11  62.3
  62.5   62.7   62.9   63.1   63.3   63.49  63.69  63.89  64.09  64.29
  64.49  64.69  64.88  65.08  65.28  65.48  65.68  65.88  66.07  66.27
  66.47  66.67  66.87  67.07  67.26  67.46  67.66  67.86  68.06  68.26
  68.45  68.65  68.85  69.05  69.25  69.45  69.65  69.84  70.04  70.24
  70.44  70.64  70.84  71.03  71.23  71.43  71.63  71.83  72.03  72.22
  72.42  72.62  72.82  73.02  73.22  73.41  73.61  73.81  74.01  74.21
  74.41  74.61  74.8   75.    75.2   75.4   75.6   75.8   75.99  76.19
  76.39  76.59  76.79  76.99  77.18  77.38  77.58  77.78  77.98  78.18
  78.37  78.57  78.77  78.97  79.17  79.37  79.57  79.76  79.96  80.16
  80.36  80.56  80.76  80.95  81.15  81.35  81.55  81.75  81.95  82.14
  82.34  82.54  82.74  82.94  83.14  83.33  83.53  83.73  83.93  84.13
  84.33  84.53  84.72  84.92  85.12  85.32  85.52  85.72  85.91  86.11
  86.31  86.51  86.71  86.91  87.1   87.3   87.5   87.7   87.9   88.1
  88.29  88.49  88.69  88.89  89.09  89.29  89.48  89.68  89.88  90.08
  90.28  90.48  90.68  90.87  91.07  91.27  91.47  91.67  91.87  92.06
  92.26  92.46  92.66  92.86  93.06  93.25  93.45  93.65  93.85  94.05
  94.25  94.44  94.64  94.84  95.04  95.24  95.44  95.64  95.83  96.03
  96.23  96.43  96.63  96.83  97.02  97.22  97.42  97.62  97.82  98.02
  98.21  98.41  98.61  98.81  99.01  99.21  99.4   99.6   99.8  100.  ]
my_label=
[ 0.84  0.93  0.98  1.    0.98  0.91  0.81  0.68  0.53  0.35  0.16 -0.04
 -0.24 -0.42 -0.59 -0.74 -0.86 -0.94 -0.99 -1.   -0.97 -0.9  -0.79 -0.66
 -0.5  -0.32 -0.12  0.07  0.27  0.45  0.62  0.76  0.88  0.95  0.99  1.
  0.96  0.88  0.77  0.63  0.47  0.29  0.09 -0.11 -0.3  -0.48 -0.65 -0.78
 -0.89 -0.96 -1.   -0.99 -0.95 -0.87 -0.75 -0.61 -0.44 -0.25 -0.06  0.14
  0.33  0.51  0.67  0.8   0.9   0.97  1.    0.99  0.94  0.85  0.73  0.58
  0.41  0.22  0.03 -0.17 -0.36 -0.54 -0.69 -0.82 -0.92 -0.98 -1.   -0.98
 -0.93 -0.83 -0.71 -0.56 -0.38 -0.19  0.01  0.2   0.39  0.57  0.72  0.84
  0.93  0.98  1.    0.98  0.91  0.82  0.69  0.53  0.35  0.16 -0.04 -0.24
 -0.42 -0.59 -0.74 -0.86 -0.94 -0.99 -1.   -0.97 -0.9  -0.8  -0.66 -0.5
 -0.32 -0.13  0.07  0.27  0.45  0.62  0.76  0.87  0.95  0.99  1.    0.96
  0.88  0.78  0.64  0.47  0.29  0.09 -0.1  -0.3  -0.48 -0.64 -0.78 -0.89
 -0.96 -1.   -0.99 -0.95 -0.87 -0.75 -0.61 -0.44 -0.26 -0.06  0.14  0.33
  0.51  0.67  0.8   0.9   0.97  1.    0.99  0.94  0.85  0.73  0.58  0.41
  0.22  0.03 -0.17 -0.36 -0.54 -0.69 -0.82 -0.92 -0.98 -1.   -0.98 -0.93
 -0.83 -0.71 -0.56 -0.38 -0.19  0.    0.2   0.39  0.56  0.72  0.84  0.93
  0.98  1.    0.98  0.91  0.82  0.69  0.53  0.35  0.16 -0.04 -0.23 -0.42
 -0.59 -0.74 -0.86 -0.94 -0.99 -1.   -0.97 -0.9  -0.8  -0.66 -0.5  -0.32
 -0.13  0.07  0.26  0.45  0.62  0.76  0.87  0.95  0.99  1.    0.96  0.89
  0.78  0.64  0.47  0.29  0.1  -0.1  -0.3  -0.48 -0.64 -0.78 -0.89 -0.96
 -1.   -0.99 -0.95 -0.87 -0.76 -0.61 -0.44 -0.26 -0.06  0.13  0.33  0.51
  0.67  0.8   0.9   0.97  1.    0.99  0.94  0.85  0.73  0.59  0.41  0.23
  0.03 -0.17 -0.36 -0.54 -0.69 -0.82 -0.92 -0.98 -1.   -0.98 -0.93 -0.84
 -0.71 -0.56 -0.38 -0.19  0.    0.2   0.39  0.56  0.71  0.84  0.93  0.98
  1.    0.98  0.91  0.82  0.69  0.53  0.35  0.16 -0.04 -0.23 -0.42 -0.59
 -0.74 -0.86 -0.94 -0.99 -1.   -0.97 -0.9  -0.8  -0.66 -0.5  -0.32 -0.13
  0.07  0.26  0.45  0.62  0.76  0.87  0.95  0.99  1.    0.96  0.89  0.78
  0.64  0.47  0.29  0.1  -0.1  -0.29 -0.48 -0.64 -0.78 -0.89 -0.96 -1.
 -0.99 -0.95 -0.87 -0.76 -0.61 -0.45 -0.26 -0.06  0.13  0.33  0.51  0.67
  0.8   0.9   0.97  1.    0.99  0.94  0.85  0.74  0.59  0.42  0.23  0.03
 -0.17 -0.36 -0.53 -0.69 -0.82 -0.92 -0.98 -1.   -0.98 -0.93 -0.84 -0.71
 -0.56 -0.39 -0.2   0.    0.2   0.39  0.56  0.71  0.84  0.93  0.98  1.
  0.98  0.92  0.82  0.69  0.53  0.36  0.16 -0.03 -0.23 -0.42 -0.59 -0.74
 -0.85 -0.94 -0.99 -1.   -0.97 -0.9  -0.8  -0.67 -0.5  -0.32 -0.13  0.07
  0.26  0.45  0.61  0.76  0.87  0.95  0.99  1.    0.96  0.89  0.78  0.64
  0.48  0.29  0.1  -0.1  -0.29 -0.48 -0.64 -0.78 -0.89 -0.96 -1.   -0.99
 -0.95 -0.87 -0.76 -0.61 -0.45 -0.26 -0.07  0.13  0.32  0.5   0.66  0.8
  0.9   0.97  1.    0.99  0.94  0.86  0.74  0.59  0.42  0.23  0.03 -0.16
 -0.35 -0.53 -0.69 -0.82 -0.92 -0.98 -1.   -0.98 -0.93 -0.84 -0.71 -0.56
 -0.39 -0.2  -0.    0.2   0.39  0.56  0.71  0.84  0.93  0.98  1.    0.98
  0.92  0.82  0.69  0.53  0.36  0.17 -0.03 -0.23 -0.42 -0.59 -0.73 -0.85
 -0.94 -0.99 -1.   -0.97 -0.9  -0.8  -0.67 -0.51]
loader[0]=[tensor([84.7234, 50.5992, 31.9499, 72.4228, 30.7595], dtype=torch.float64), tensor([ 0.0994,  0.3276,  0.5090, -0.1655, -0.6103], dtype=torch.float64)]
loader[1]=[tensor([93.4529, 27.9820, 60.5190, 38.2986, 27.5852], dtype=torch.float64), tensor([-0.7138,  0.2882, -0.7371,  0.5642,  0.6359], dtype=torch.float64)]
loader[2]=[tensor([86.9058, 34.9259, 56.5511, 83.5331, 24.2124], dtype=torch.float64), tensor([-0.8718, -0.3601,  0.0024,  0.9608, -0.7958], dtype=torch.float64)]
loader[3]=[tensor([43.8537, 10.7214, 36.1162, 85.7154, 11.9118], dtype=torch.float64), tensor([-0.1282, -0.9627, -0.9999, -0.7786, -0.6088], dtype=torch.float64)]
loader[4]=[tensor([35.7194, 83.3347, 90.0802, 57.1463, 75.2004], dtype=torch.float64), tensor([-0.9176,  0.9966,  0.8552,  0.5627, -0.1965], dtype=torch.float64)]
loader[5]=[tensor([ 3.7776, 36.5130, 52.1864, 57.9399, 58.5351], dtype=torch.float64), tensor([-0.5940, -0.9269,  0.9393,  0.9839,  0.9149], dtype=torch.float64)]
loader[6]=[tensor([19.2525, 78.3747, 87.3026, 83.9299, 13.8958], dtype=torch.float64), tensor([ 0.3921,  0.1643, -0.6147,  0.7790,  0.9710], dtype=torch.float64)]
loader[7]=[tensor([70.4389, 39.6874, 68.6533, 88.6914, 94.6433], dtype=torch.float64), tensor([ 0.9697,  0.9141, -0.4455,  0.6645,  0.3853], dtype=torch.float64)]
loader[8]=[tensor([74.8036, 97.8176, 75.0020, 77.5812, 81.9459], dtype=torch.float64), tensor([-0.5602, -0.4153, -0.3859,  0.8184,  0.2614], dtype=torch.float64)]
loader[9]=[tensor([68.8517, 26.3948,  8.1423, 87.1042, 35.1242], dtype=torch.float64), tensor([-0.2603,  0.9527,  0.9587, -0.7581, -0.5369], dtype=torch.float64)]
loader[10]=[tensor([25.7996, 69.0501, 32.9419, 84.9218, 85.5170], dtype=torch.float64), tensor([ 0.6185, -0.0649,  0.9990, -0.0987, -0.6396], dtype=torch.float64)]
loader[11]=[tensor([42.8617, 54.3687, 33.1403, 35.9178, 40.4810], dtype=torch.float64), tensor([-0.9004, -0.8201,  0.9882, -0.9779,  0.3520], dtype=torch.float64)]
loader[12]=[tensor([66.0741, 78.9699, 89.4850, 28.9739, 38.1002], dtype=torch.float64), tensor([-0.1005, -0.4170,  0.9987, -0.6439,  0.3904], dtype=torch.float64)]
loader[13]=[tensor([16.0782, 54.5671, 36.7114, 74.6052,  2.9840], dtype=torch.float64), tensor([-0.3618, -0.9168, -0.8348, -0.7125,  0.1570], dtype=torch.float64)]
loader[14]=[tensor([92.2625,  3.5792, 14.6894, 47.6232, 89.0882], dtype=torch.float64), tensor([-0.9153, -0.4237,  0.8514, -0.4789,  0.9017], dtype=torch.float64)]
loader[15]=[tensor([90.4770, 66.2725, 11.5150, 11.1182, 64.4870], dtype=torch.float64), tensor([ 0.5885, -0.2947, -0.8681, -0.9925,  0.9964], dtype=torch.float64)]
loader[16]=[tensor([76.7876, 68.0581, 32.1483, 13.4990, 77.1844], dtype=torch.float64), tensor([ 0.9836, -0.8708,  0.6686,  0.8032,  0.9769], dtype=torch.float64)]
loader[17]=[tensor([18.4589, 64.6854, 73.0180, 84.3267, 19.6493], dtype=torch.float64), tensor([-0.3808,  0.9603, -0.6899,  0.4762,  0.7172], dtype=torch.float64)]
loader[18]=[tensor([73.6132, 25.6012,  1.9920, 29.9659, 75.7956], dtype=torch.float64), tensor([-0.9771,  0.4515,  0.9126, -0.9927,  0.3870], dtype=torch.float64)]
loader[19]=[tensor([53.1784, 40.0842, 84.5251, 50.4008, 31.5531], dtype=torch.float64), tensor([0.2267, 0.6864, 0.2936, 0.1349, 0.1367], dtype=torch.float64)]
loader[20]=[tensor([24.6092,  6.1583, 80.9539, 81.7475, 21.8317], dtype=torch.float64), tensor([-0.4999, -0.1245, -0.6650,  0.0660,  0.1588], dtype=torch.float64)]
loader[21]=[tensor([69.8437, 95.6353, 35.3226, 17.0701, 77.3828], dtype=torch.float64), tensor([ 0.6659,  0.9832, -0.6926, -0.9783,  0.9156], dtype=torch.float64)]
loader[22]=[tensor([41.8697, 12.1102, 91.6673, 30.9579,  3.3808], dtype=torch.float64), tensor([-0.8568, -0.4405, -0.5322, -0.4422, -0.2369], dtype=torch.float64)]
loader[23]=[tensor([76.3908, 69.2485, 66.6693, 96.8257, 90.6754], dtype=torch.float64), tensor([ 0.8374,  0.1331, -0.6411,  0.5343,  0.4176], dtype=torch.float64)]
loader[24]=[tensor([67.8597, 34.7275, 20.4429, 29.3707, 71.8277], dtype=torch.float64), tensor([-0.9506, -0.1691,  0.9997, -0.8896,  0.4159], dtype=torch.float64)]
loader[25]=[tensor([98.0160, 99.0080, 13.1022, 68.4549, 26.7916], dtype=torch.float64), tensor([-0.5864, -0.9989,  0.5106, -0.6132,  0.9961], dtype=torch.float64)]
loader[26]=[tensor([61.9078, 30.1643, 81.3507, 54.7655, 73.2164], dtype=torch.float64), tensor([-0.7980, -0.9495, -0.3247, -0.9775, -0.8191], dtype=torch.float64)]
loader[27]=[tensor([98.8096, 23.2204, 92.8577, 46.8297, 53.7735], dtype=torch.float64), tensor([-0.9887, -0.9423, -0.9837,  0.2900, -0.3583], dtype=torch.float64)]
loader[28]=[tensor([74.0100, 56.3527, 17.4669, 37.9018,  4.5711], dtype=torch.float64), tensor([-0.9834, -0.1947, -0.9823,  0.2013, -0.9900], dtype=torch.float64)]
loader[29]=[tensor([61.5110, 58.3367, 48.2184, 12.5070, 13.6974], dtype=torch.float64), tensor([-0.9689,  0.9765, -0.8887, -0.0593,  0.9048], dtype=torch.float64)]
loader[30]=[tensor([40.6794, 90.2786, 22.4269, 54.1703, 10.5230], dtype=torch.float64), tensor([ 0.1606,  0.7363, -0.4220, -0.6913, -0.8904], dtype=torch.float64)]
loader[31]=[tensor([30.5611, 80.5571, 90.8737,  9.5311, 39.0922], dtype=torch.float64), tensor([-0.7544, -0.9020,  0.2304, -0.1061,  0.9842], dtype=torch.float64)]
loader[32]=[tensor([47.0281, 95.0401, 64.2886,  1.5952, 21.6333], dtype=torch.float64), tensor([0.0957, 0.7120, 0.9935, 0.9997, 0.3503], dtype=torch.float64)]
loader[33]=[tensor([ 1.3968, 91.4689, 32.3467, 55.9559, 32.7435], dtype=torch.float64), tensor([ 0.9849, -0.3548,  0.8021, -0.5586,  0.9706], dtype=torch.float64)]
loader[34]=[tensor([20.2445, 45.0441, 65.4790,  5.5631, 44.2505], dtype=torch.float64), tensor([ 0.9846,  0.8732,  0.4746, -0.6594,  0.2650], dtype=torch.float64)]
loader[35]=[tensor([41.2745, 29.7675, 39.2906,  1.1984, 39.4890], dtype=torch.float64), tensor([-0.4204, -0.9970,  0.9998,  0.9315,  0.9761], dtype=torch.float64)]
loader[36]=[tensor([67.2645, 85.1202, 62.8998, 47.8216,  6.7535], dtype=torch.float64), tensor([-0.9611, -0.2929,  0.0679, -0.6425,  0.4532], dtype=torch.float64)]
loader[37]=[tensor([97.4208,  2.3888, 88.8898, 65.0822, 59.3287], dtype=torch.float64), tensor([-0.0315,  0.6837,  0.7987,  0.7779,  0.3538], dtype=torch.float64)]
loader[38]=[tensor([11.7134, 72.0261, 86.3106, 28.3788, 88.0962], dtype=torch.float64), tensor([-0.7532,  0.2285, -0.9965, -0.1042,  0.1312], dtype=torch.float64)]
loader[39]=[tensor([34.3307, 36.3146, 26.5932, 46.4329, 42.2665], dtype=torch.float64), tensor([ 0.2249, -0.9827,  0.9939,  0.6373, -0.9895], dtype=torch.float64)]
loader[40]=[tensor([99.4048, 76.5892, 23.0220, 24.0140,  2.1904], dtype=torch.float64), tensor([-0.9028,  0.9287, -0.8578, -0.8995,  0.8141], dtype=torch.float64)]
loader[41]=[tensor([51.1944, 92.4609, 38.6954, 51.7896, 59.7255], dtype=torch.float64), tensor([ 0.8010, -0.9767,  0.8395,  0.9989, -0.0352], dtype=torch.float64)]
loader[42]=[tensor([75.5972, 49.8056, 83.1363, 12.3086,  1.7936], dtype=torch.float64), tensor([ 0.1977, -0.4438,  0.9933, -0.2549,  0.9753], dtype=torch.float64)]
loader[43]=[tensor([65.2806, 82.3427, 14.0942,  4.9679, 49.0120], dtype=torch.float64), tensor([ 0.6388,  0.6141,  0.9991, -0.9675, -0.9501], dtype=torch.float64)]
loader[44]=[tensor([29.1723, 75.9940, 57.3447, 71.2325, 63.8918], dtype=torch.float64), tensor([-0.7821,  0.5611,  0.7146,  0.8543,  0.8723], dtype=torch.float64)]
loader[45]=[tensor([82.5411, 23.8156, 67.4629, 18.0621, 34.5291], dtype=torch.float64), tensor([ 0.7576, -0.9680, -0.9967, -0.7085,  0.0285], dtype=torch.float64)]
loader[46]=[tensor([ 94.8417, 100.0000,  51.3928,  64.0902,  30.3627], dtype=torch.float64), tensor([ 0.5596, -0.5064,  0.9033,  0.9516, -0.8690], dtype=torch.float64)]
loader[47]=[tensor([45.2425, 66.4709, 80.3587, 75.3988, 22.2285], dtype=torch.float64), tensor([ 9.5215e-01, -4.7723e-01, -9.6938e-01,  5.7391e-04, -2.3509e-01],
       dtype=torch.float64)]
loader[48]=[tensor([ 6.3567, 76.9860, 86.7074, 57.7415, 61.1142], dtype=torch.float64), tensor([ 0.0735,  0.9999, -0.9512,  0.9294, -0.9892], dtype=torch.float64)]
loader[49]=[tensor([ 5.1663, 17.8637, 50.2024, 15.4830, 79.3667], dtype=torch.float64), tensor([-0.8987, -0.8337, -0.0630,  0.2231, -0.7358], dtype=torch.float64)]
loader[50]=[tensor([96.4289,  5.9599, 33.3387, 31.3547, 52.3848], dtype=torch.float64), tensor([ 0.8195, -0.3177,  0.9387, -0.0612,  0.8533], dtype=torch.float64)]
loader[51]=[tensor([47.2265, 94.4449, 97.6192, 22.8236, 62.1062], dtype=torch.float64), tensor([-0.1024,  0.1958, -0.2278, -0.7396, -0.6636], dtype=torch.float64)]
loader[52]=[tensor([31.7515,  6.5551, 63.0982, 63.6934, 41.6713], dtype=torch.float64), tensor([ 0.3293,  0.2686,  0.2632,  0.7588, -0.7384], dtype=torch.float64)]
loader[53]=[tensor([ 9.9279, 70.2405, 56.7495, 61.3126, 50.7976], dtype=torch.float64), tensor([-0.4821,  0.9025,  0.1995, -0.9987,  0.5074], dtype=torch.float64)]
loader[54]=[tensor([52.9800, 81.1523, 98.4128, 93.0561,  2.5872], dtype=torch.float64), tensor([ 0.4142, -0.5048, -0.8539, -0.9290,  0.5264], dtype=torch.float64)]
loader[55]=[tensor([60.3206, 72.8196, 19.0541, 79.5651, 91.8657], dtype=torch.float64), tensor([-0.5895, -0.5337,  0.2031, -0.8549, -0.6886], dtype=torch.float64)]
loader[56]=[tensor([ 7.5471, 19.8477, 87.8978, 97.2224, 18.8557], dtype=torch.float64), tensor([ 0.9533,  0.8405, -0.0667,  0.1662,  0.0062], dtype=torch.float64)]
loader[57]=[tensor([26.9900, 25.9980, 27.3868, 67.0661, 73.4148], dtype=torch.float64), tensor([ 0.9593,  0.7613,  0.7755, -0.8879, -0.9161], dtype=torch.float64)]
loader[58]=[tensor([63.2966, 82.1443, 22.0301, 85.3186, 89.6834], dtype=torch.float64), tensor([ 0.4482,  0.4465, -0.0389, -0.4756,  0.9891], dtype=torch.float64)]
loader[59]=[tensor([69.4469, 61.7094, 58.1383, 59.5271, 96.2305], dtype=torch.float64), tensor([ 0.3258, -0.9012,  0.9998,  0.1625,  0.9164], dtype=torch.float64)]
loader[60]=[tensor([95.8337, 64.8838, 40.2826, 50.9960, 77.9780], dtype=torch.float64), tensor([0.9999, 0.8865, 0.5296, 0.6672, 0.5328], dtype=torch.float64)]
loader[61]=[tensor([48.0200, 78.5731, 96.0321, 37.7034, 63.4950], dtype=torch.float64), tensor([-0.7809, -0.0333,  0.9773,  0.0043,  0.6156], dtype=torch.float64)]
loader[62]=[tensor([71.6293, 20.0461, 58.7335, 57.5431, 28.5772], dtype=torch.float64), tensor([ 0.5870,  0.9308,  0.8173,  0.8384, -0.2982], dtype=torch.float64)]
loader[63]=[tensor([86.5090, 46.6313, 88.4930,  7.3487, 98.6112], dtype=torch.float64), tensor([-0.9934,  0.4729,  0.5041,  0.8750, -0.9397], dtype=torch.float64)]
loader[64]=[tensor([14.2926, 62.3046,  8.5391, 44.0521, 23.4188], dtype=torch.float64), tensor([ 0.9879, -0.5032,  0.7744,  0.0698, -0.9898], dtype=torch.float64)]
loader[65]=[tensor([18.6573, 99.6032, 32.5451, 89.2866, 55.5591], dtype=torch.float64), tensor([-0.1911, -0.8003,  0.9041,  0.9692, -0.8358], dtype=torch.float64)]
loader[66]=[tensor([40.8778,  8.3407, 25.0060, 38.4970, 43.2585], dtype=torch.float64), tensor([-0.0370,  0.8839, -0.1264,  0.7159, -0.6622], dtype=torch.float64)]
loader[67]=[tensor([79.1683, 99.8016, 29.5691,  7.9439, 65.8758], dtype=torch.float64), tensor([-0.5879, -0.6664, -0.9622,  0.9960,  0.0975], dtype=torch.float64)]
loader[68]=[tensor([15.6814, 79.9619, 52.5832,  9.7295, 43.0601], dtype=torch.float64), tensor([ 0.0266, -0.9890,  0.7338, -0.3000, -0.7969], dtype=torch.float64)]
loader[69]=[tensor([ 3.1824, 59.1303, 39.8858, 33.5371, 45.6393], dtype=torch.float64), tensor([-0.0408,  0.5312,  0.8163,  0.8523,  0.9963], dtype=torch.float64)]
loader[70]=[tensor([71.4309,  2.7856, 25.4028,  7.1503, 17.6653], dtype=torch.float64), tensor([ 0.7351,  0.3485,  0.2668,  0.7625, -0.9262], dtype=torch.float64)]
loader[71]=[tensor([25.2044, 74.4068, 16.2766, 98.2144, 12.9038], dtype=torch.float64), tensor([ 0.0716, -0.8368, -0.5384, -0.7346,  0.3311], dtype=torch.float64)]
loader[72]=[tensor([78.1764, 16.8717, 58.9319, 78.7715, 67.6613], dtype=torch.float64), tensor([ 0.3555, -0.9183,  0.6878, -0.2297, -0.9932], dtype=torch.float64)]
loader[73]=[tensor([55.7575, 41.0762, 15.2846, 47.4248, 42.6633], dtype=torch.float64), tensor([-0.7112, -0.2333,  0.4109, -0.2964, -0.9685], dtype=torch.float64)]
loader[74]=[tensor([ 4.7695, 93.2545, 94.0481, 21.2365, 48.4168], dtype=torch.float64), tensor([-0.9984, -0.8378, -0.1984,  0.6851, -0.9616], dtype=torch.float64)]
loader[75]=[tensor([69.6453, 60.7174, 13.3006, 70.0421, 93.6513], dtype=torch.float64), tensor([ 0.5058, -0.8558,  0.6700,  0.7999, -0.5617], dtype=torch.float64)]
loader[76]=[tensor([70.6373, 44.6473, 35.5210, 27.1884, 87.5010], dtype=torch.float64), tensor([ 0.9988,  0.6171, -0.8212,  0.8847, -0.4472], dtype=torch.float64)]
loader[77]=[tensor([91.2705, 43.4569, 18.2605, 44.8457, 33.7355], dtype=torch.float64), tensor([-0.1636, -0.5015, -0.5556,  0.7601,  0.7325], dtype=torch.float64)]
loader[78]=[tensor([45.8377, 46.0361, 77.7796, 53.5752, 99.2064], dtype=torch.float64), tensor([ 0.9598,  0.8856,  0.6891, -0.1673, -0.9698], dtype=torch.float64)]
loader[79]=[tensor([ 9.1343, 37.5050, 87.6994, 20.8397, 95.4369], dtype=torch.float64), tensor([ 0.2864, -0.1929, -0.2621,  0.9134,  0.9280], dtype=torch.float64)]
loader[80]=[tensor([ 1.0000, 86.1122, 71.0341, 41.4729, 31.1563], dtype=torch.float64), tensor([ 0.8415, -0.9606,  0.9400, -0.5910, -0.2567], dtype=torch.float64)]
loader[81]=[tensor([50.0040, 55.1623, 21.0381, 27.7836, 66.8677], dtype=torch.float64), tensor([-0.2585, -0.9830,  0.8152,  0.4713, -0.7798], dtype=torch.float64)]
loader[82]=[tensor([84.1283, 89.8818, 56.9479, 62.7014, 49.2104], dtype=torch.float64), tensor([ 0.6402,  0.9406,  0.3887, -0.1301, -0.8699], dtype=torch.float64)]
loader[83]=[tensor([81.5491, 53.9719, 11.3166, 28.1804, 38.8938], dtype=torch.float64), tensor([-0.1319, -0.5353, -0.9489,  0.0938,  0.9301], dtype=torch.float64)]
loader[84]=[tensor([44.4489, 91.0721, 15.0862, 96.6273, 82.7395], dtype=torch.float64), tensor([0.4499, 0.0340, 0.5825, 0.6905, 0.8714], dtype=torch.float64)]
loader[85]=[tensor([ 8.9359, 93.8497, 33.9339, 37.1082, 16.6733], dtype=torch.float64), tensor([ 0.4697, -0.3876,  0.5840, -0.5571, -0.8223], dtype=torch.float64)]
loader[86]=[tensor([52.7816, 59.9238,  4.3727, 54.9639, 60.9158], dtype=torch.float64), tensor([ 0.5855, -0.2315, -0.9429, -0.9999, -0.9410], dtype=torch.float64)]
loader[87]=[tensor([37.3066, 74.2084, 22.6253, 68.2565, 34.1323], dtype=torch.float64), tensor([-0.3825, -0.9283, -0.5925, -0.7569,  0.4126], dtype=torch.float64)]
loader[88]=[tensor([76.1924, 94.2465, 28.7756,  7.7455, 14.8878], dtype=torch.float64), tensor([ 0.7133, -0.0013, -0.4805,  0.9941,  0.7313], dtype=torch.float64)]
loader[89]=[tensor([51.5912, 95.2385, 80.1603, 42.4649,  9.3327], dtype=torch.float64), tensor([ 0.9701,  0.8364, -0.9988, -0.9986,  0.0920], dtype=torch.float64)]
loader[90]=[tensor([19.4509, 36.9098,  4.1743, 92.6593, 24.8076], dtype=torch.float64), tensor([ 0.5658, -0.7099, -0.8587, -0.9998, -0.3194], dtype=torch.float64)]
loader[91]=[tensor([72.6212, 83.7315, 21.4349, 23.6172, 79.7635], dtype=torch.float64), tensor([-0.3566,  0.8873,  0.5280, -0.9985, -0.9404], dtype=torch.float64)]
loader[92]=[tensor([49.4088, 17.2685, 62.5030, 12.7054, 82.9379], dtype=torch.float64), tensor([-0.7557, -0.9999, -0.3230,  0.1386,  0.9510], dtype=torch.float64)]
loader[93]=[tensor([10.1263,  3.9760, 65.6774, 10.9198, 45.4409], dtype=torch.float64), tensor([-0.6453, -0.7409,  0.2918, -0.9971,  0.9937], dtype=torch.float64)]
loader[94]=[tensor([49.6072, 46.2345,  8.7375, 15.8798, 73.8116], dtype=torch.float64), tensor([-0.6117,  0.7767,  0.6345, -0.1710, -0.9999], dtype=torch.float64)]
loader[95]=[tensor([55.3607,  5.3647, 80.7555, 48.6152, 48.8136], dtype=torch.float64), tensor([-0.9276, -0.7947, -0.7992, -0.9968, -0.9929], dtype=torch.float64)]
loader[96]=[tensor([97.0240, 10.3246, 43.6553, 42.0681, 85.9138], dtype=torch.float64), tensor([ 0.3573, -0.7832, -0.3212, -0.9416, -0.8870], dtype=torch.float64)]
loader[97]=[tensor([ 6.9519, 14.4910, 24.4108,  5.7615, 26.1964], dtype=torch.float64), tensor([ 0.6200,  0.9381, -0.6608, -0.4983,  0.8741], dtype=torch.float64)]
loader[98]=[tensor([53.3768, 20.6413, 56.1543, 16.4749, 70.8357], dtype=torch.float64), tensor([ 0.0303,  0.9757, -0.3842, -0.6940,  0.9888], dtype=torch.float64)]
loader[99]=[tensor([51.9880, 60.1222, 92.0641, 72.2244, 88.2946], dtype=torch.float64), tensor([ 0.9885, -0.4187, -0.8180,  0.0322,  0.3240], dtype=torch.float64)]

Process finished with exit code 0

原文地址：https://blog.csdn.net/scar2016/article/details/143866506

免责声明：本站文章内容转载自网络资源，如本站内容侵犯了原著者的合法权益，可联系本站删除。更多内容请关注自学内容网（zxcms.com）！

上一篇：基于Java Springboot汽车4s店管理系统
下一篇：# TCP、UDP、HTTP、Socket

论文阅读——Intrusion detection systems using longshort‑term memory (LSTM)
作者提出的 LSTM 模型能够有效区分正常网络流量和攻击流量。除此之外，模型结合主成分分析（PCA）和互信息作为降维方法。实验结果表明，基于 PCA 的模型（特别是使用2个主成分）在二分类和多分类任务
阅读更多2024-11-21
用源码编译虚幻引擎，并打包到安卓平台
本文详细介绍了如何用源码编译虚幻引擎，并将其打包到安卓平台。
阅读更多2024-11-21
Vue项目开发 element-UI 前端实现 1到10排列选择的按钮
在 Element UI 中，你可以通过来实现按钮的排列选择，例如让用户选择 1 到 10 之间的数字。为了实现这一功能，我们可以使用来动态生成 1 到 10 的按钮，并通过按钮点击事件来更新
阅读更多2024-11-21
Java EE 【知识改变命运】01计算机的一些知识点
计算机一些基础知识
阅读更多2024-11-21
04 —— Webpack打包CSS代码
加载器style-loader：把解析后的css代码插入到DOM。加载器css-loader ：解析css代码。直接引用，不用变量接收。
阅读更多2024-11-21
万能程序补丁工具 C# 源代码详解
万能程序补丁工具程序目的：搜索二进制可执行 EXE 或 DLL 文件分析的特征代码，替换特征代码，达到调试修正目标程序的功能。
阅读更多2024-11-21
【MySQL数据库】C#实现MySQL数据库最简单的查询和执行函数
C#和MySQL数据库是常见的数据交互，标准的查询和执行方法如下，做个记录。
阅读更多2024-11-21
单条推理转批量推理prompt
在每个线程中设置环境变量 CUDA_VISIBLE_DEVICES，以确保每个线程只使用指定的GPU。使用 concurrent.futures.ThreadPoolExecutor 来管理多线程任务
阅读更多2024-11-21
【AIGC】ChatGPT提示词Prompt解析：情感分析，分手后还可以做朋友吗？
【AIGC】在情感博弈中，最重要的是保持清醒的认知和优雅的态度。识别控制话术不是为了对抗，而是为了更好地保护自己的情感自由，实现真正的成长。
阅读更多2024-11-21
pycharm中配置pyqt5
PyQt和wxPython则提供了更多的控件和更强大的功能，适合于需要复杂用户界面的应用程序。pyQt生成的应用程序，引用图片通常是将资源文件装换为 python 文件，然后引用资源文件，而不能直接加
阅读更多2024-11-21