tw/gpu/train_fail2.py

#!/usr/bin/python
#coding=utf-8    
'''
If there are Chinese comments in the code，please add at the beginning：
#!/usr/bin/python
#coding=utf-8    

Due to the adaptability of a100, before using the training environment, please use the recommended image of the 
platform with cuda 11.Then adjust the code and submit the image.
The image of this example is: dockerhub.pcl.ac.cn:5000/user-images/openi:cuda111_python37_pytorch191
In the training environment, the uploaded dataset will be automatically placed in the /dataset directory. 
If it is a single dataset: 
if MnistDataset_torch.zip is selected,Then the dataset directory is /dataset/train, /dataset/test;
If it is a multiple dataset: 
If MnistDataset_torch.zip and checkpoint_epoch1_0.73.zip are selected, 
the dataset directory is /dataset/MnistDataset_torch/train, /dataset/MnistDataset_torch/test
and /dataset/checkpoint_epoch1_0.73/mnist_epoch1_0.73.pkl

The model download path is under /model by default. Please specify the model output location to /model, 
and the Qizhi platform will provide file downloads under the /model directory.
'''


from model import Model
import numpy as np
import torch
from torchvision.datasets import mnist
from torch.nn import CrossEntropyLoss
from torch.optim import SGD
from torch.utils.data import DataLoader
from torchvision.transforms import ToTensor
import argparse
import datetime

# Training settings
parser = argparse.ArgumentParser(description='PyTorch MNIST Example')
#The dataset location is placed under /dataset
parser.add_argument('--traindata', default="/dataset/train" ,help='path to train dataset')
parser.add_argument('--testdata', default="/dataset/test" ,help='path to test dataset')
parser.add_argument('--epoch_size', type=int, default=1, help='how much epoch to train')
parser.add_argument('--batch_size', type=int, default=256, help='how much batch_size in epoch')

def gettime():
    timestr = datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')
    return timestr

if __name__ == '__main__':
    args, unknown = parser.parse_known_args()
    #log output
    print(gettime(), 'cuda is available:{}'.format(torch.cuda.is_available()))  
    device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
    batch_size = args.batch_size
    train_dataset = mnist.MNIST(root=args.traindata, train=True, transform=ToTensor(),download=False)
    test_dataset = mnist.MNIST(root=args.testdata, train=False, transform=ToTensor(),download=False)
    train_loader = DataLoader(train_dataset, batch_size=batch_size)
    test_loader = DataLoader(test_dataset, batch_size=batch_size)
    model = Model().to(device)
    sgd = SGD(model.parameters(), lr=1e-1)
    cost = CrossEntropyLoss()
    epoch = args.epoch_size
    print(gettime(), 'epoch_size is:{}'.format(epoch))
    for _epoch in range(epoch):
        print(gettime(), 'the {} epoch_size begin'.format(_epoch + 1))
        model.train()
        for idx, (train_x, train_label) in enumerate(train_loader):
            train_x = train_x.to(device)
            train_label = train_label.to(device)
            label_np = np.zeros((train_label.shape[0], 10))
            sgd.zero_grad()
            predict_y = model(train_x.float())
            loss = cost(predict_y, train_label.long())
            if idx % 10 == 0:
                print(gettime(), 'idx: {}, loss: {}'.format(idx, loss.sum().item()))
            loss.backward()
            sgd.step()

        correct = 0
        _sum = 0
        model.eval()
        for idx, (test_x, test_label) in enumerate(test_loader):
            test_x = test_x
            test_label = test_label
            predict_y = model(test_x.to(device).float()).detach()
            predict_ys = np.argmax(predict_y.cpu(), axis=-1)
            label_np = test_label.numpy()
            _ = predict_ys == test_label
            correct += np.sum(_.numpy(), axis=-1)
            _sum += _.shape[0]
        print(gettime(), 'accuracy: {:.2f}'.format(correct / _sum))
        #The model output location is placed under /model
        torch.save(model, '/model/mnist_epoch{}_{:.2f}.pkl'.format(_epoch+1, correct / _sum))
        print("----------this is the end--------")
        print("abc"