def parse_args():
parser = argparse.ArgumentParser(
description="Train classification models on ImageNet",
formatter_class=argparse.ArgumentDefaultsHelpFormatter)
models.add_model_args(parser)
fit.add_fit_args(parser)
data.add_data_args(parser)
dali.add_dali_args(parser)
data.add_data_aug_args(parser)
return parser.parse_args()
default='sgd',
help='the optimizer type')
parser.add_argument('--model-prefix',
type=str,
default='models/',
help='model prefix')
parser.add_argument('--disp-batches',
type=int,
default=20,
help='show progress for every n batches')
parser.add_argument('--predict',
action='store_true',
default=False,
help='run prediction instead of training')
data.add_data_args(parser)
parser.set_defaults(
# config
model_prefix='models/',
disp_batches=20,
# data
data_config='data_ak8_pfcand_reduced_cloud',
data_train=train_val_fname,
train_val_split=0.8,
data_test=test_fname,
data_example=example_fname,
data_names=None,
num_examples=-1,
# train
batch_size=1024,
num_epochs=200,
import argparse
from data import get_data, add_data_args
import torchvision.utils as vutils
import matplotlib.pyplot as plt
###########
## Setup ##
###########
parser = argparse.ArgumentParser()
add_data_args(parser)
parser.add_argument('--num_cols', type=int, default=8)
args = parser.parse_args()
###############
## Load data ##
###############
print('Loading data...')
train_loader = get_data(args)[0]
images = next(iter(train_loader))
###############
## Plot data ##
###############
print('images.shape = {}'.format(images.shape))
print('Plotting data...')