add some

Author: LeetJoe

.gitignore

@@ -19,6 +19,7 @@ esearch/csvmappings.py
 
 # C extensions
 *.so
+*.pth
 
 # Distribution / packaging
 .DS_Store

pyTorch/BookPyTorch2/chapter2/Softmax.py

@@ -0,0 +1,15 @@
+import math
+
+import numpy
+def softmax(inMatrix):
+    m,n = numpy.shape(inMatrix)
+    outMatrix = numpy.mat(numpy.zeros((m,n)))
+    soft_sum = 0
+    for idx in range(0,n):
+        outMatrix[0,idx] = math.exp(inMatrix[0,idx])
+        soft_sum += outMatrix[0,idx]
+    for idx in range(0,n):
+        outMatrix[0,idx] = outMatrix[0,idx] / soft_sum
+    return outMatrix
+a = numpy.array([[1, 2, 1, 2, 1, 1, 3]])
+print(softmax(a))

pyTorch/BookPyTorch2/chapter2/hello_pytorch.py

@@ -0,0 +1,3 @@
+import torch
+result = torch.tensor(1) + torch.tensor(2.0)
+print(result)

pyTorch/BookPyTorch2/chapter2/img/img.jpg

@@ -0,0 +1,26 @@
+import numpy as np
+import matplotlib.pyplot as plt
+import torch
+
+grid_size = 5
+x_train = np.load("../dataset/mnist/x_train.npy")
+#image = torch.tensor(x_train[0]).to("cuda")
+image = torch.tensor(x_train[0]).to("cpu")
+
+image = image
+print(image.shape)
+image = image.cpu().numpy()
+plt.imshow(image)
+plt.savefig("./img/img.jpg")
+plt.show()
+
+
+
+
+# image = torch.tensor(x_train[0]).to("cuda")
+# image = torch.squeeze(image,dim=0)
+# image = image.cpu().numpy()
+# plt.imshow(image)
+# plt.show()
+
+

pyTorch/BookPyTorch2/chapter2/img/img_0.jpg

@@ -0,0 +1,49 @@
+import os
+os.environ['CUDA_VISIBLE_DEVICES'] = '0' #指定GPU编
+import torch
+import numpy as np
+import unet
+import matplotlib.pyplot as plt
+from tqdm import tqdm
+
+batch_size = 320                        #设定每次训练的批次数
+epochs = 1024                           #设定训练次数
+
+#device = "cpu"                         #Pytorch的特性，需要指定计算的硬件，如果没有GPU的存在，就使用CPU进行计算
+device = "cuda"                         #在这里读者默认使用GPU，如果读者出现运行问题可以将其改成cpu模式
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+device = "cpu"
+
+model = unet.Unet()                     #导入Unet模型
+model = model.to(device)                #将计算模型传入GPU硬件等待计算
+#model = torch.compile(model)            #Pytorch2.0的特性，加速计算速度
+optimizer = torch.optim.Adam(model.parameters(), lr=2e-5)   #设定优化函数
+
+#载入数据
+x_train = np.load("../dataset/mnist/x_train.npy")
+y_train_label = np.load("../dataset/mnist/y_train_label.npy")
+
+x_train_batch = []
+for i in range(len(y_train_label)):
+    if y_train_label[i] <= 10:                    #为了加速演示作者只对数据集中的小于2的数字，也就是0和1进行运行，读者可以自行增加训练个数
+        x_train_batch.append(x_train[i])
+
+x_train = np.reshape(x_train_batch, [-1, 1, 28, 28])  #修正数据输入维度：([30596, 28, 28])
+x_train /= 512.
+
+image = x_train[np.random.randint(28)]                    #随机挑选一条数据进行计算
+image = np.reshape(image,[28,28])                                   #修正数据维度
+plt.imshow(image)
+plt.show()
+
+state_dict = torch.load("./saver/unet.pth")
+model.load_state_dict(state_dict)
+image = torch.reshape(torch.tensor(image),[1,1,28,28])
+img = model(image)
+img = torch.reshape(img, shape=[28,28])                             #修正模型输出结果
+img = img.detach().cpu().numpy()
+plt.imshow(img)
+plt.show()
+
+
+

pyTorch/BookPyTorch2/chapter2/img/img_1.jpg

@@ -0,0 +1,10 @@
+import torch
+
+print(torch.__version__)
+print(torch.cuda.is_available())
+
+print("是否可用：", torch.cuda.is_available())        # 查看GPU是否可用
+print("GPU数量：", torch.cuda.device_count())        # 查看GPU数量
+print("torch方法查看CUDA版本：", torch.version.cuda)  # torch方法查看CUDA版本
+print("GPU索引号：", torch.cuda.current_device())    # 查看GPU索引号
+print("GPU名称：", torch.cuda.get_device_name(0))    # 根据索引号得到GPU名称