-
Notifications
You must be signed in to change notification settings - Fork 0
/
dataset.py
43 lines (37 loc) · 1.38 KB
/
dataset.py
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
import numpy as np
import pytorch_lightning as pl
from typing import Any
import torch
from torch.utils.data import DataLoader
import torchvision
import torchvision.transforms as transforms
class ToTensor:
def __call__(self, arr):
if isinstance(arr, torch.Tensor):
return arr
arr = np.asarray(arr)
return torch.from_numpy(arr)
class CIFAR10(pl.LightningDataModule):
def __init__(self, train_batch=64, val_batch=32):
super().__init__()
self._train_batch = train_batch
self._val_batch = val_batch
self._train_xform = transforms.Compose([
transforms.RandomHorizontalFlip(),
ToTensor(),
])
self._val_xform = transforms.Compose([
ToTensor(),
])
def train_dataloader(self) -> DataLoader:
trainset = torchvision.datasets.CIFAR10(
root='./data', train=True, download=True, transform=self._train_xform)
trainloader = torch.utils.data.DataLoader(
trainset, batch_size=self._train_batch, shuffle=True)
return trainloader
def val_dataloader(self) -> DataLoader:
testset = torchvision.datasets.CIFAR10(
root='./data', train=False, download=True, transform=self._val_xform)
testloader = torch.utils.data.DataLoader(
testset, batch_size=self._val_batch, shuffle=False)
return testloader