train_cifar10.py

import argparse, time, logging
import os
import numpy as np
import mxnet as mx
from tqdm import tqdm
from mxnet import gluon
from mxnet import autograd as ag
from mxnet.gluon.data.vision import transforms

import gluoncv as gcv
gcv.utils.check_version('0.6.0')
from gluoncv.model_zoo import get_model
from gluoncv.utils import makedirs, TrainingHistory
from gluoncv.data import transforms as gcv_transforms
from mxboard import SummaryWriter


# CLI
def parse_args():
    parser = argparse.ArgumentParser(description='Train a model for image classification.')
    parser.add_argument('--batch-size', type=int, default=512,
                        help='training batch size per device (CPU/GPU).')
    parser.add_argument('--num-gpus', type=int, default=0,
                        help='number of gpus to use.')
    parser.add_argument('--model', type=str, default='cifar_resnet20_v2',
                        help='model to use. options are resnet and wrn. default is resnet.')
    parser.add_argument('-j', '--num-data-workers', dest='num_workers', default=4, type=int,
                        help='number of preprocessing workers')
    parser.add_argument('--num-epochs', type=int, default=40,
                        help='number of training epochs.')
    parser.add_argument('--lr', type=float, default=0.1,
                        help='learning rate. default is 0.1.')
    parser.add_argument('--momentum', type=float, default=0.9,
                        help='momentum value for optimizer, default is 0.9.')
    parser.add_argument('--wd', type=float, default=0.0001,
                        help='weight decay rate. default is 0.0001.')
    parser.add_argument('--lr-decay', type=float, default=0.1,
                        help='decay rate of learning rate. default is 0.1.')
    parser.add_argument('--lr-decay-period', type=int, default=0,
                        help='period in epoch for learning rate decays. default is 0 (has no effect).')
    parser.add_argument('--lr-decay-epoch', type=str, default='20,30',
                        help='epochs at which learning rate decays. default is 40,60.')
    parser.add_argument('--drop-rate', type=float, default=0.0,
                        help='dropout rate for wide resnet. default is 0.')
    parser.add_argument('--mode', type=str,
                        help='mode in which to train the model. options are imperative, hybrid')
    parser.add_argument('--save-period', type=int, default=5,
                        help='period in epoch of model saving.')
    parser.add_argument('--save-dir', type=str, default='snapshots',
                        help='directory of saved models')
    parser.add_argument('--resume-from', type=str,
                        help='resume training from the model')
    opt = parser.parse_args()
    return opt


def main():
    opt = parse_args()
    batch_size = opt.batch_size
    classes = 10

    log_dir = os.path.join(opt.save_dir, "logs")
    model_dir = os.path.join(opt.save_dir, "params")
    if not os.path.exists(model_dir):
        os.makedirs(model_dir)

    # Init dataloader
    jitter_param = 0.4
    transform_train = transforms.Compose([
        gcv_transforms.RandomCrop(32, pad=4),
        transforms.RandomFlipLeftRight(),
        transforms.RandomBrightness(jitter_param),
        transforms.RandomColorJitter(jitter_param),
        transforms.RandomContrast(jitter_param),
        transforms.RandomSaturation(jitter_param),
        transforms.ToTensor(),
        transforms.Normalize([0.4914, 0.4822, 0.4465], [0.2023, 0.1994, 0.2010])
    ])

    transform_test = transforms.Compose([
        transforms.ToTensor(),
        transforms.Normalize([0.4914, 0.4822, 0.4465], [0.2023, 0.1994, 0.2010])
    ])

    train_data = gluon.data.DataLoader(
        gluon.data.vision.CIFAR10(train=True).transform_first(transform_train),
        batch_size=batch_size, shuffle=True, last_batch='discard', num_workers=opt.num_workers)

    val_data = gluon.data.DataLoader(
        gluon.data.vision.CIFAR10(train=False).transform_first(transform_test),
        batch_size=batch_size, shuffle=False, num_workers=opt.num_workers)

    num_gpus = opt.num_gpus
    batch_size *= max(1, num_gpus)
    context = [mx.gpu(i) for i in range(num_gpus)] if num_gpus > 0 else [mx.cpu()]

    lr_decay = opt.lr_decay
    lr_decay_epoch = [int(i) for i in opt.lr_decay_epoch.split(',')] + [np.inf]

    model_name = opt.model
    model_name = opt.model
    if model_name.startswith('cifar_wideresnet'):
        kwargs = {'classes': classes,
                'drop_rate': opt.drop_rate}
    else:
        kwargs = {'classes': classes}
    net = get_model(model_name, **kwargs)

    if opt.resume_from:
        net.load_parameters(opt.resume_from, ctx=context)
    optimizer = 'nag'

    save_period = opt.save_period
    if opt.save_dir and save_period:
        save_dir = opt.save_dir
        makedirs(save_dir)
    else:
        save_dir = ''
        save_period = 0

    def test(ctx, val_loader):
        metric = mx.metric.Accuracy()
        for i, batch in enumerate(val_loader):
            data = gluon.utils.split_and_load(batch[0], ctx_list=ctx, batch_axis=0)
            label = gluon.utils.split_and_load(batch[1], ctx_list=ctx, batch_axis=0)
            outputs = [net(X) for X in data]
            metric.update(label, outputs)
        return metric.get()

    def train(train_data, val_data, epochs, ctx):
        if isinstance(ctx, mx.Context):
            ctx = [ctx]

        net.hybridize()
        net.initialize(mx.init.Xavier(), ctx=ctx)
        net.forward(mx.nd.ones((1, 3, 30, 30), ctx=ctx[0]))
        with SummaryWriter(logdir=log_dir, verbose=False) as sw:
            sw.add_graph(net)

        trainer = gluon.Trainer(net.collect_params(), optimizer,
                                {'learning_rate': opt.lr, 'wd': opt.wd, 'momentum': opt.momentum})
        metric = mx.metric.Accuracy()
        train_metric = mx.metric.Accuracy()
        loss_fn = gluon.loss.SoftmaxCrossEntropyLoss()

        iteration = 0
        lr_decay_count = 0

        best_val_score = 0
        global_step = 0

        for epoch in range(epochs):
            tic = time.time()
            train_metric.reset()
            metric.reset()
            train_loss = 0
            num_batch = len(train_data)
            alpha = 1

            if epoch == lr_decay_epoch[lr_decay_count]:
                trainer.set_learning_rate(trainer.learning_rate*lr_decay)
                lr_decay_count += 1

            tbar = tqdm(train_data)

            for i, batch in enumerate(tbar):
                data = gluon.utils.split_and_load(batch[0], ctx_list=ctx, batch_axis=0)
                label = gluon.utils.split_and_load(batch[1], ctx_list=ctx, batch_axis=0)

                with ag.record():
                    output = [net(X) for X in data]
                    loss = [loss_fn(yhat, y) for yhat, y in zip(output, label)]
                for l in loss:
                    l.backward()
                trainer.step(batch_size)
                train_loss += sum([l.sum().asscalar() for l in loss])

                train_metric.update(label, output)
                name, acc = train_metric.get()
                iteration += 1
                global_step += len(loss)

            train_loss /= batch_size * num_batch
            name, acc = train_metric.get()
            name, val_acc = test(ctx, val_data)

            if val_acc > best_val_score:
                best_val_score = val_acc
                net.save_parameters('{}/{}-{}-{:04.3f}-best.params'.format(model_dir, model_name, epoch, best_val_score))

            with SummaryWriter(logdir=log_dir, verbose=False) as sw:
                sw.add_scalar(tag="TrainLos", value=train_loss, global_step=global_step)
                sw.add_scalar(tag="TrainAcc", value=acc, global_step=global_step)
                sw.add_scalar(tag="ValAcc", value=val_acc, global_step=global_step)
                sw.add_graph(net)

            logging.info('[Epoch %d] train=%f val=%f loss=%f time: %f' %
                         (epoch, acc, val_acc, train_loss, time.time() - tic))

            if save_period and save_dir and (epoch + 1) % save_period == 0:
                net.save_parameters('{}/{}-{}.params'.format(save_dir, model_name, epoch))

        if save_period and save_dir:
            net.save_parameters('{}/{}-{}.params'.format(save_dir, model_name, epochs-1))

    if opt.mode == 'hybrid':
        net.hybridize()
    train(train_data, val_data, opt.num_epochs, context)


if __name__ == '__main__':
    main()