config.label_smooth_factor = 0.0
loss = CrossEntropy(smooth_factor=config.label_smooth_factor,
num_classes=config.class_num)
if args_opt.do_train:
dataset = create_dataset(dataset_path=args_opt.dataset_path,
do_train=True,
repeat_num=epoch_size,
batch_size=config.batch_size)
step_size = dataset.get_dataset_size()
loss_scale = FixedLossScaleManager(config.loss_scale,
drop_overflow_update=False)
# learning rate strategy with cosine
lr = Tensor(
warmup_cosine_annealing_lr(config.lr, step_size,
config.warmup_epochs,
config.epoch_size))
opt = Momentum(filter(lambda x: x.requires_grad,
net.get_parameters()), lr, config.momentum,
config.weight_decay, config.loss_scale)
model = Model(net,
loss_fn=loss,
optimizer=opt,
amp_level='O2',
keep_batchnorm_fp32=False,
loss_scale_manager=loss_scale,
metrics={'acc'})
time_cb = TimeMonitor(data_size=step_size)
loss_cb = LossMonitor()
cb = [time_cb, loss_cb]
if config.save_checkpoint:
if not config.label_smooth:
config.label_smooth_factor = 0.0
loss = CrossEntropy(smooth_factor=config.label_smooth_factor,
num_classes=config.class_num)
if args_opt.do_train:
dataset = create_dataset(dataset_path=args_opt.dataset_path,
do_train=True,
repeat_num=epoch_size,
batch_size=config.batch_size)
step_size = dataset.get_dataset_size()
loss_scale = FixedLossScaleManager(config.loss_scale,
drop_overflow_update=False)
lr = Tensor(
warmup_cosine_annealing_lr(0.035, step_size, config.warmup_epochs,
50, config.T_max, config.eta_min))
opt = THOR(
filter(lambda x: x.requires_grad, net.get_parameters()), lr,
config.momentum, damping, config.frequency,
filter(lambda x: 'matrix_A' in x.name, net.get_parameters()),
filter(lambda x: 'matrix_G' in x.name, net.get_parameters()),
filter(lambda x: 'spatial_norm' in x.name, net.get_parameters()),
config.weight_decay, config.loss_scale)
model = Model(net,
loss_fn=loss,
optimizer=opt,
amp_level='O2',
loss_scale_manager=loss_scale,
keep_batchnorm_fp32=False,
metrics={'acc'},
if args_opt.do_train:
dataset = create_dataset(dataset_path=args_opt.dataset_path,
do_train=True,
repeat_num=epoch_size,
batch_size=config.batch_size)
step_size = dataset.get_dataset_size()
loss_scale = FixedLossScaleManager(config.loss_scale,
drop_overflow_update=False)
if args_opt.pre_trained:
param_dict = load_checkpoint(args_opt.pre_trained)
load_param_into_net(net, param_dict)
# learning rate strategy with cosine
lr = Tensor(
warmup_cosine_annealing_lr(config.lr, step_size,
config.warmup_epochs, 120,
config.pretrain_epoch_size * step_size))
opt = Momentum(filter(lambda x: x.requires_grad,
net.get_parameters()), lr, config.momentum,
config.weight_decay, config.loss_scale)
model = Model(net,
loss_fn=loss,
optimizer=opt,
amp_level='O2',
keep_batchnorm_fp32=False,
loss_scale_manager=loss_scale,
metrics={'acc'})
time_cb = TimeMonitor(data_size=step_size)
loss_cb = LossMonitor()
cb = [time_cb, loss_cb]
if config.save_checkpoint: