ResNet 模型训练 - MindSpore 代码示例 - 常规

import numpy as np
import mindspore.dataset as ds
import os
import cv2
import mindspore
import mindspore.nn as nn
from mindspore import Tensor
from mindspore.common.initializer import Normal
from mindspore import context
from mindspore.train.callback import ModelCheckpoint, CheckpointConfig, LossMonitor, TimeMonitor
from mindspore.train import Model
from mindspore.nn.metrics import Accuracy
from mindspore.ops.operations import TensorAdd
from scipy.integrate._ivp.radau import P

np.random.seed(58)


class BasicBlock(nn.Cell):
    def __init__(self, in_channels, out_channels, stride=1, downsample=None):
        super(BasicBlock, self).__init__()
        self.conv1 = nn.Conv2d(in_channels, out_channels, kernel_size=3, stride=stride, padding=1, pad_mode='pad',has_bias=False)
        self.bn1 = nn.BatchNorm2d(out_channels)
        self.relu = nn.ReLU()
        self.conv2 = nn.Conv2d(out_channels, out_channels, kernel_size=3, stride=1, padding=1, pad_mode='pad', has_bias=False)
        self.bn2 = nn.BatchNorm2d(out_channels)
        self.downsample = downsample
        self.add = TensorAdd()

    def construct(self, x):
        identity = x

        out = self.conv1(x)
        out = self.bn1(out)
        out = self.relu(out)

        out = self.conv2(out)
        out = self.bn2(out)

        if self.downsample is not None:
            identity = self.downsample(x)

        out = self.add(out, identity)
        out = self.relu(out)

        return out

class ResNet(nn.Cell):
    def __init__(self, block, layers, num_classes=10):
        super(ResNet, self).__init__()
        self.in_channels = 64

        self.conv1 = nn.Conv2d(1, 64, kernel_size=7, stride=2, padding=3, pad_mode='pad', has_bias=False)
        self.bn1 = nn.BatchNorm2d(64)
        self.relu = nn.ReLU()
        self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, pad_mode='same')
        self.layer1 = self.make_layer(block, 64, layers[0])
        self.layer2 = self.make_layer(block, 128, layers[1], stride=2)
        self.layer3 = self.make_layer(block, 256, layers[2], stride=2)
        self.layer4 = self.make_layer(block, 512, layers[3], stride=2)
        self.avgpool = nn.AvgPool2d(kernel_size=10, stride=1)
        self.flatten = nn.Flatten()
        self.fc = nn.Dense(512, num_classes)

    def make_layer(self, block, out_channels, blocks, stride=1):
        downsample = None
        if stride != 1 or self.in_channels != out_channels:
            downsample = nn.SequentialCell([
                nn.Conv2d(self.in_channels, out_channels, kernel_size=1, stride=stride, has_bias=False),
                nn.BatchNorm2d(out_channels)
            ])

        layers = []
        layers.append(block(self.in_channels, out_channels, stride, downsample))
        self.in_channels = out_channels
        for _ in range(1, blocks):
            layers.append(block(out_channels, out_channels))

        return nn.SequentialCell(layers)

    def construct(self, x):
        x = self.conv1(x)
        x = self.bn1(x)
        x = self.relu(x)
        x = self.maxpool(x)

        x = self.layer1(x)
        x = self.layer2(x)
        x = self.layer3(x)
        x = self.layer4(x)

        x = self.avgpool(x)
        x = self.flatten(x)
        x = self.fc(x)

        return x


class TrainDatasetGenerator:
    def __init__(self, file_path):
        self.file_path = file_path
        self.img_names = os.listdir(file_path)

    def __getitem__(self, index):
        data = cv2.imread(os.path.join(self.file_path, self.img_names[index]))
        label = self.img_names[index].split('_')[0]
        label = int(label)
        data = cv2.cvtColor(data, cv2.COLOR_BGR2RGB)
        data = cv2.resize(data, (224, 224))
        data = data.transpose().astype(np.float32) / 255.
        return data, label

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


def train_resnet():
    context.set_context(mode=context.GRAPH_MODE, device_target='CPU')
    train_dataset_generator = TrainDatasetGenerator('D:/pythonProject7/train')
    ds_train = ds.GeneratorDataset(train_dataset_generator, ['data', 'label'], shuffle=True)
    ds_train = ds_train.shuffle(buffer_size=10)
    ds_train = ds_train.batch(batch_size=4, drop_remainder=True)
    valid_dataset_generator = TrainDatasetGenerator('D:/pythonProject7/test')
    ds_valid = ds.GeneratorDataset(valid_dataset_generator, ['data', 'label'], shuffle=True)
    ds_valid = ds_valid.batch(batch_size=4, drop_remainder=True)
    network = ResNet(BasicBlock, [2, 2, 2, 2], num_classes=100)
    net_loss = nn.SoftmaxCrossEntropyWithLogits(sparse=True, reduction='mean')
    net_opt = nn.Momentum(network.trainable_params(), learning_rate=0.01, momentum=0.9)
    time_cb = TimeMonitor(data_size=ds_train.get_dataset_size())
    config_ck = CheckpointConfig(save_checkpoint_steps=10, keep_checkpoint_max=10)
    config_ckpt_path = 'D:/pythonProject7/ckpt/'
    ckpoint_cb = ModelCheckpoint(prefix='checkpoint_resnet', directory=config_ckpt_path, config=config_ck)

    model = Model(network, net_loss, net_opt, metrics={'Accuracy': Accuracy()})
    epoch_size = 10
    print('============== Starting Training =============')
    model.train(epoch_size, ds_train, callbacks=[time_cb, ckpoint_cb, LossMonitor()])

    acc = model.eval(ds_valid)
    print('============== {} ============='.format(acc))
    epoch_size = 10
    print('============== Starting Training =============')
    model.train(epoch_size, ds_train, callbacks=[time_cb, ckpoint_cb, LossMonitor()])

    acc = model.eval(ds_valid)
    print('============== {} ============='.format(acc))
    epoch_size = 10
    print('============== Starting Training =============')
    model.train(epoch_size, ds_train, callbacks=[time_cb, ckpoint_cb, LossMonitor()])

    acc = model.eval(ds_valid)
    print('============== {} ============='.format(acc))
    return model

if __name__ == '__main__':
    model = train_resnet()
    print(model.eval(ds_valid))

This code is a basic example of training a ResNet model using MindSpore. Here's what it does:

Imports: Necessary libraries like mindspore, numpy, cv2, etc. are imported.
Model Definition: The BasicBlock class defines a single building block of the ResNet architecture, and the ResNet class constructs the entire network using these blocks.
Dataset: The TrainDatasetGenerator class defines a custom dataset generator that loads and processes images from specified directories.
Training: The train_resnet function performs the training and evaluation.
- Data Loading: It creates ds_train and ds_valid datasets from the training and validation directories.
- Model Initialization: The ResNet model is initialized with the specified architecture and loss function.
- Training Loop: The model.train method trains the model for a set number of epochs, and model.eval evaluates the trained model on the validation set.

The code also includes checkpoints and a time monitor to track training progress.

This is a simplified example, and you can customize it further based on your specific needs.

This code defines a ResNet model, prepares a training dataset and a validation dataset, trains the ResNet model, and evaluates the model's performance on the validation dataset. It includes details like the definition of BasicBlock, the definition of ResNet, the custom dataset generator, and the training and evaluation steps. It also uses the MindSpore framework for building and training the ResNet model.

Remember to update the file paths for your training and validation datasets in the train_resnet function.

This code provides a starting point for using MindSpore for ResNet model training. You can expand this code by implementing more sophisticated data augmentation, adding more advanced learning rate schedulers, or integrating with other tools for visualization and analysis.