-
Notifications
You must be signed in to change notification settings - Fork 0
/
tensor_dataloader.py
47 lines (42 loc) · 2.29 KB
/
tensor_dataloader.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
44
45
46
47
from typing import *
import math
from torch.utils.data import Sampler, TensorDataset, DataLoader
import torch
######################################
# Custom Sampler
######################################
class TensorBatchSampler(Sampler):
def __init__(self, data_source, batch_size, shuffle=False, drop_last=False):
if not isinstance(data_source, TensorDataset):
raise ValueError(f"data_source should be an instance of torch.utils.data.TensorDataset, but got data_source={data_source}")
if not isinstance(batch_size, int) or isinstance(batch_size, bool) or \
batch_size <= 0:
raise ValueError(f"batch_size should be a positive integer value, but got batch_size={batch_size}")
if not isinstance(shuffle, bool):
raise ValueError(f"shuffle should be a boolean value, but got shuffle={shuffle}")
if not isinstance(drop_last, bool):
raise ValueError(f"drop_last should be a boolean value, but got drop_last={drop_last}")
self.data_source = data_source
self.batch_size = batch_size
self.shuffle = shuffle
self.drop_last = drop_last
def __iter__(self):
if self.shuffle:
idxs = torch.randperm(len(self.data_source)).split(self.batch_size)
else:
idxs = torch.arange(len(self.data_source)).split(self.batch_size)
if self.drop_last and len(self.data_source) % self.batch_size != 0:
idxs = idxs[:-1]
return iter(idxs)
def __len__(self):
return (math.floor if self.drop_last else math.ceil)(len(self.data_source) / self.batch_size)
def unwrap_collate_fn(batch):
return batch[0]
class TensorDataLoader(DataLoader):
def __init__(self, dataset, batch_size=1, shuffle=False, sampler=None,
batch_sampler=None, *, drop_last=False, collate_fn=None, **kwargs):
if sampler is not None or batch_sampler is not None or collate_fn is not None:
raise ValueError("TensorDataLoader does not support alternate samplers, batch samplers, or collate functions.")
sampler = TensorBatchSampler(dataset, batch_size, shuffle, drop_last)
super().__init__(dataset, batch_size=1, shuffle=False, sampler=sampler,
drop_last=False, collate_fn = unwrap_collate_fn, **kwargs)