File size: 5,316 Bytes
d0f68bc |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 |
from torch.utils.data import Dataset, DataLoader, Subset
from robust_detection.data_utils.rcnn_data_utils import *
import pytorch_lightning as pl
import robust_detection.transforms as T
DATA_FOLDER = os.path.join(os.path.dirname(__file__))
def get_transform():
transforms = []
transforms.append(T.ToTensor())
return T.Compose(transforms)
class Objects_Smiles(pl.LightningDataModule):
def __init__(self, data_path, **kwargs):
super().__init__()
self.batch_size = 1
self.num_workers = 4
self.data_path = data_path
self.transforms = get_transform()
self.base_class = Objects_Detection_Predictor_Dataset
def prepare_data(self):
dataset = self.base_class(os.path.join(DATA_FOLDER, self.data_path), self.transforms)
self.train = dataset
self.test = dataset
self.val = dataset
self.test_ood = dataset
def train_dataloader(self):
return DataLoader(
self.train,
batch_size=self.batch_size,
shuffle=True,
num_workers=self.num_workers,
drop_last=False,
pin_memory=True,
collate_fn=collate_tuple
)
def val_dataloader(self):
return DataLoader(
self.val,
batch_size=self.batch_size,
shuffle=False,
num_workers=self.num_workers,
drop_last=False,
pin_memory=True,
collate_fn=collate_tuple
)
def test_dataloader(self):
return DataLoader(
self.test,
batch_size=self.batch_size,
shuffle=False,
num_workers=self.num_workers,
drop_last=False,
pin_memory=True,
collate_fn=collate_tuple
)
def test_ood_dataloader(self, shuffle=False):
return DataLoader(
self.test_ood,
batch_size=self.batch_size,
shuffle=shuffle,
num_workers=self.num_workers,
drop_last=False,
pin_memory=True,
collate_fn=collate_tuple
)
@classmethod
def add_dataset_specific_args(cls, parent):
import argparse
parser = argparse.ArgumentParser(parents=[parent], add_help=False)
parser.add_argument('--data_path', type=str,
default="mnist/alldigits/")
return parser
class Objects_fold_Smiles(pl.LightningDataModule):
def __init__(self, data_path, fold, **kwargs):
super().__init__()
self.batch_size = 1
self.num_workers = 4
self.data_path = data_path
self.fold = fold
self.transforms = get_transform()
# self.base_class = Objects_Detection_Predictor_Dataset
self.base_class = Objects_Detection_Dataset
def prepare_data(self):
dataset = self.base_class(os.path.join(DATA_FOLDER, self.data_path), self.transforms)
if self.fold > -1:
train_idx = np.load(os.path.join(DATA_FOLDER, f"{self.data_path}", "../folds", str(self.fold), "train_idx.npy"))
self.train = Subset(dataset, train_idx)
val_idx = np.load(os.path.join(DATA_FOLDER, f"{self.data_path}", "../folds", str(self.fold), "val_idx.npy"))
self.val = Subset(dataset, val_idx)
else:
self.train = dataset
self.val = dataset
self.test = self.val
self.test_ood = self.test
def train_dataloader(self):
return DataLoader(
self.train,
batch_size=self.batch_size,
shuffle=True,
num_workers=self.num_workers,
drop_last=False,
pin_memory=True,
collate_fn=collate_tuple
)
def val_dataloader(self):
return DataLoader(
self.val,
batch_size=self.batch_size,
shuffle=False,
num_workers=self.num_workers,
drop_last=False,
pin_memory=True,
collate_fn=collate_tuple
)
def test_dataloader(self):
return DataLoader(
self.test,
batch_size=self.batch_size,
shuffle=False,
num_workers=self.num_workers,
drop_last=False,
pin_memory=True,
collate_fn=collate_tuple
)
def test_ood_dataloader(self, shuffle=False):
return DataLoader(
self.test_ood,
batch_size=self.batch_size,
shuffle=shuffle,
num_workers=self.num_workers,
drop_last=False,
pin_memory=True,
collate_fn=collate_tuple
)
@classmethod
def add_dataset_specific_args(cls, parent):
import argparse
parser = argparse.ArgumentParser(parents=[parent], add_help=False)
parser.add_argument('--data_path', type=str,
default="mnist/alldigits/")
parser.add_argument('--fold', type=int,
default=0)
return parser
|