昇思25天学习打卡营第02天|快速入门

import time
import mindspore
from mindspore import nn
from mindspore.dataset import vision, transforms
from mindspore.dataset import MnistDataset
from download import download

url = "https://mindspore-website.obs.cn-north-4.myhuaweicloud.com/" \
      "notebook/datasets/MNIST_Data.zip"
path = download(url, "./", kind="zip", replace=True)
train_dataset = MnistDataset('MNIST_Data/train')
test_dataset = MnistDataset('MNIST_Data/test')
print(train_dataset.get_col_names())

def datapipe(dataset, batch_size):
    image_transforms = [
        vision.Rescale(1.0 / 255.0, 0),
        vision.Normalize(mean=(0.1307,), std=(0.3081,)),
        vision.HWC2CHW()
    ]
    label_transform = transforms.TypeCast(mindspore.int32)

    dataset = dataset.map(image_transforms, 'image')
    dataset = dataset.map(label_transform, 'label')
    dataset = dataset.batch(batch_size)
    return dataset

# Map vision transforms and batch dataset
train_dataset = datapipe(train_dataset, 64)
test_dataset = datapipe(test_dataset, 64)

for image, label in test_dataset.create_tuple_iterator():
    print(f"Shape of image [N, C, H, W]: {image.shape} {image.dtype}")
    print(f"Shape of label: {label.shape} {label.dtype}")
    break
for data in test_dataset.create_dict_iterator():
    print(f"Shape of image [N, C, H, W]: {data['image'].shape} {data['image'].dtype}")
    print(f"Shape of label: {data['label'].shape} {data['label'].dtype}")
    break

    # Define model
class Network(nn.Cell):
    def __init__(self):
        super().__init__()
        self.flatten = nn.Flatten()
        self.dense_relu_sequential = nn.SequentialCell(
            nn.Dense(28*28, 512),
            nn.ReLU(),
            nn.Dense(512, 512),
            nn.ReLU(),
            nn.Dense(512, 10)
        )

    def construct(self, x):
        x = self.flatten(x)
        logits = self.dense_relu_sequential(x)
        return logits

model = Network()
print(model)

# Instantiate loss function and optimizer
loss_fn = nn.CrossEntropyLoss()
optimizer = nn.SGD(model.trainable_params(), 1e-2)

# 1. Define forward function
def forward_fn(data, label):
    logits = model(data)
    loss = loss_fn(logits, label)
    return loss, logits

# 2. Get gradient function
grad_fn = mindspore.value_and_grad(forward_fn, None, optimizer.parameters, has_aux=True)

# 3. Define function of one-step training
def train_step(data, label):
    (loss, _), grads = grad_fn(data, label)
    optimizer(grads)
    return loss

def train(model, dataset):
    size = dataset.get_dataset_size()
    model.set_train()
    for batch, (data, label) in enumerate(dataset.create_tuple_iterator()):
        loss = train_step(data, label)

        if batch % 100 == 0:
            loss, current = loss.asnumpy(), batch
            print(f"loss: {loss:>7f}  [{current:>3d}/{size:>3d}]")
    
    
def test(model, dataset, loss_fn):
    num_batches = dataset.get_dataset_size()
    model.set_train(False)
    total, test_loss, correct = 0, 0, 0
    for data, label in dataset.create_tuple_iterator():
        pred = model(data)
        total += len(data)
        test_loss += loss_fn(pred, label).asnumpy()
        correct += (pred.argmax(1) == label).asnumpy().sum()
    test_loss /= num_batches
    correct /= total
    print(f"Test: \n Accuracy: {(100*correct):>0.1f}%, Avg loss: {test_loss:>8f} \n")

epochs = 10
for t in range(epochs):
    print(f"Epoch {t+1}\n-------------------------------")
    train(model, train_dataset)
    test(model, test_dataset, loss_fn)
print("Done!")
# Save checkpoint
mindspore.save_checkpoint(model, "model.ckpt")
print("Saved Model to model.ckpt")

# Instantiate a random initialized model
model = Network()
# Load checkpoint and load parameter to model
param_dict = mindspore.load_checkpoint("model.ckpt")
param_not_load, _ = mindspore.load_param_into_net(model, param_dict)
print(param_not_load)


model.set_train(False)
for data, label in test_dataset:
    pred = model(data)
    predicted = pred.argmax(1)
    print(f'Predicted: "{predicted[:10]}", Actual: "{label[:10]}"')
    break

print(time.strftime("%Y-%m-%d %H:%M:%S", time.localtime()),'skywp')

Downloading data from https://mindspore-website.obs.cn-north-4.myhuaweicloud.com/notebook/datasets/MNIST_Data.zip (10.3 MB)

file_sizes: 100%|███████████████████████████| 10.8M/10.8M [00:00<00:00, 161MB/s]
Extracting zip file...
Successfully downloaded / unzipped to ./
['image', 'label']
Shape of image [N, C, H, W]: (64, 1, 28, 28) Float32
Shape of label: (64,) Int32
Shape of image [N, C, H, W]: (64, 1, 28, 28) Float32
Shape of label: (64,) Int32
Network<
  (flatten): Flatten<>
  (dense_relu_sequential): SequentialCell<
    (0): Dense<input_channels=784, output_channels=512, has_bias=True>
    (1): ReLU<>
    (2): Dense<input_channels=512, output_channels=512, has_bias=True>
    (3): ReLU<>
    (4): Dense<input_channels=512, output_channels=10, has_bias=True>
    >
  >
Epoch 1
-------------------------------
loss: 2.279789  [  0/938]
loss: 1.712688  [100/938]
loss: 0.872912  [200/938]
loss: 0.499927  [300/938]
loss: 0.528868  [400/938]
loss: 0.501849  [500/938]
loss: 0.398775  [600/938]
loss: 0.447360  [700/938]
loss: 0.286581  [800/938]
loss: 0.332299  [900/938]
Test: 
 Accuracy: 90.8%, Avg loss: 0.323961 

Epoch 2
-------------------------------
loss: 0.233778  [  0/938]
loss: 0.255780  [100/938]
loss: 0.150093  [200/938]
loss: 0.279230  [300/938]
loss: 0.365830  [400/938]
loss: 0.210970  [500/938]
loss: 0.168809  [600/938]
loss: 0.235510  [700/938]
loss: 0.294178  [800/938]
loss: 0.176371  [900/938]
Test: 
 Accuracy: 92.9%, Avg loss: 0.247941 

Epoch 3
-------------------------------
loss: 0.275178  [  0/938]
loss: 0.174475  [100/938]
loss: 0.232013  [200/938]
loss: 0.145759  [300/938]
loss: 0.285019  [400/938]
loss: 0.306831  [500/938]
loss: 0.347860  [600/938]
loss: 0.131523  [700/938]
loss: 0.173969  [800/938]
loss: 0.240118  [900/938]
Test: 
 Accuracy: 94.0%, Avg loss: 0.209006 

Epoch 4
-------------------------------
loss: 0.270546  [  0/938]
loss: 0.136092  [100/938]
loss: 0.149328  [200/938]
loss: 0.177321  [300/938]
loss: 0.129258  [400/938]
loss: 0.248159  [500/938]
loss: 0.129819  [600/938]
loss: 0.149704  [700/938]
loss: 0.122814  [800/938]
loss: 0.192139  [900/938]
Test: 
 Accuracy: 94.6%, Avg loss: 0.182635 

Epoch 5
-------------------------------
loss: 0.286487  [  0/938]
loss: 0.242977  [100/938]
loss: 0.093043  [200/938]
loss: 0.092097  [300/938]
loss: 0.199909  [400/938]
loss: 0.107868  [500/938]
loss: 0.186808  [600/938]
loss: 0.175348  [700/938]
loss: 0.135079  [800/938]
loss: 0.134168  [900/938]
Test: 
 Accuracy: 95.3%, Avg loss: 0.160946 

Epoch 6
-------------------------------
loss: 0.140025  [  0/938]
loss: 0.091131  [100/938]
loss: 0.199530  [200/938]
loss: 0.117356  [300/938]
loss: 0.171679  [400/938]
loss: 0.186101  [500/938]
loss: 0.119611  [600/938]
loss: 0.143975  [700/938]
loss: 0.273533  [800/938]
loss: 0.036094  [900/938]
Test: 
 Accuracy: 95.8%, Avg loss: 0.143220 

Epoch 7
-------------------------------
loss: 0.239285  [  0/938]
loss: 0.075213  [100/938]
loss: 0.026216  [200/938]
loss: 0.115669  [300/938]
loss: 0.132865  [400/938]
loss: 0.114236  [500/938]
loss: 0.139998  [600/938]
loss: 0.204587  [700/938]
loss: 0.098077  [800/938]
loss: 0.157028  [900/938]
Test: 
 Accuracy: 96.2%, Avg loss: 0.129179 

Epoch 8
-------------------------------
loss: 0.108100  [  0/938]
loss: 0.080826  [100/938]
loss: 0.139983  [200/938]
loss: 0.159044  [300/938]
loss: 0.279953  [400/938]
loss: 0.056621  [500/938]
loss: 0.137742  [600/938]
loss: 0.133608  [700/938]
loss: 0.111276  [800/938]
loss: 0.090705  [900/938]
Test: 
 Accuracy: 96.5%, Avg loss: 0.118926 

Epoch 9
-------------------------------
loss: 0.115459  [  0/938]
loss: 0.094403  [100/938]
loss: 0.131634  [200/938]
loss: 0.071918  [300/938]
loss: 0.215124  [400/938]
loss: 0.090745  [500/938]
loss: 0.101857  [600/938]
loss: 0.257950  [700/938]
loss: 0.084232  [800/938]
loss: 0.174996  [900/938]
Test: 
 Accuracy: 96.8%, Avg loss: 0.108092 

Epoch 10
-------------------------------
loss: 0.055152  [  0/938]
loss: 0.023567  [100/938]
loss: 0.121823  [200/938]
loss: 0.061770  [300/938]
loss: 0.106167  [400/938]
loss: 0.106240  [500/938]
loss: 0.133518  [600/938]
loss: 0.060433  [700/938]
loss: 0.029802  [800/938]
loss: 0.066283  [900/938]
Test: 
 Accuracy: 96.8%, Avg loss: 0.105468 

Done!
Saved Model to model.ckpt
[]
Predicted: "[9 7 0 1 3 7 0 5 8 2]", Actual: "[9 2 0 1 3 7 0 5 8 2]"
2024-06-26 13:53:48 skywp

ps:怎么说呢，今天的学习真的是一头雾水的赶脚，勉强把代码都跑了下，嗯，任重而道远。