深度網絡學習-PyTorch_應用模型進行推斷-單張圖片的預測

說明

沒有聯網，先把模型下載下來
先學習怎么推斷，
  然后再看怎么進行Dataset Dataloader transform
 接着看怎么訓練和評價

軟件和硬件

cuda
  查看cuda 版本
  whereis nvcc
  /usr/local/cuda-10.0/bin/nvcc -V
  cat /usr/local/cuda/version.txt   
libcudnn.so最終鏈接的文件名，文件名中包含版本號
GPU查看
  lspci | grep -i nvidia 
  nvidia-sm
  watch -n 1 nvidia-sm

示例代碼

import torch
import torch.cuda
import torch.nn
import torchvision.models as models
import torchvision.transforms as transforms
import numpy as np
import cv2

def get_model():
    # 加載模型 model_ft = torchvision.models.vgg16(pretrained=False)
    model_ft = models.resnet101(pretrained=False)
    #model_path ="./models/vgg16-397923af.pth"
    model_path ="./models/resnet101-5d3b4d8f.pth"
    pre = torch.load(model_path)
    model_ft.load_state_dict(pre)
    model_ft.cuda()
    return model_ft

    # # 查看模型結構
    # print(model_ft)
    # # 查看網絡參數
    # for name, parameters in model_ft.named_parameters():
    #     print(name, ':', parameters.size())
    # # 網絡模型的卷積方式以及權重數值
    # print("#############-parameters")
    # for child in model_ft.children():
    #     print(child) 
    #     # for param in child.parameters():
    #     #     print(param)

def deal_img(img_path):
    """Transforming images on GPU"""
    image = cv2.imread(img_path) 
    image_new =  cv2.resize(image, (224,224))
    my_transforms= transforms.Compose(
        [ 
        transforms.ToTensor(),
        transforms.Normalize(mean=[0.485, 0.456, 0.406],std=[0.229,0.224,0.225]) 
        ]
        )
    my_tensor = my_transforms(image_new)
    my_tensor = my_tensor.resize_(1,3,224,224)
    my_tensor= my_tensor.cuda()
    return my_tensor

def cls_inference(cls_model,imgpth):
    input_tensor = deal_img(imgpth)
    cls_model.eval()
    result = cls_model(input_tensor)
    result_npy = result.data.cpu().numpy()
    max_index = np.argmax(result_npy[0])
    return max_index

def feature_extract(cls_model,imgpth):
    cls_model.fc = torch.nn.LeakyReLU(0.1)
    cls_model.eval()
    input_tensor = deal_img(imgpth)
    result = cls_model(input_tensor)
    result_npy = result.data.cpu().numpy()
    return result_npy[0]


if __name__ == "__main__":
    image_path="./pytorch/data/train/cat/08.jpg"
    model = get_model()
    cls_label = cls_inference(model,image_path)
    print(cls_label)
    feature = feature_extract(model,image_path)
    print(feature)

參考

使用pytorch預訓練模型分類與特征提取 https://blog.csdn.net/u010165147/article/details/72829969?spm=1001.2014.3001.5502