File size: 12,003 Bytes
747451d | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 | # /*---------------------------------------------------------------------------------------------
# * Copyright (c) 2025 STMicroelectronics.
# * All rights reserved.
# *
# * This software is licensed under terms that can be found in the LICENSE file in
# * the root directory of this software component.
# * If no LICENSE file comes with this software, it is provided AS-IS.
# * Loader Factory, Fast Collate, CUDA Prefetcher
# * Prefetcher and Fast Collate inspired by NVIDIA APEX example at
# * https://github.com/NVIDIA/apex/commit/d5e2bb4bdeedd27b1dfaf5bb2b24d6c000dee9be#diff-cf86c282ff7fba81fad27a559379d5bf
# * Copyright 2019, Ross Wightman
# *--------------------------------------------------------------------------------------------*/
from functools import partial
import torch
from timm.data.dataset import IterableImageDataset
from timm.data.distributed_sampler import (OrderedDistributedSampler,
RepeatAugSampler)
from timm.data.loader import (MultiEpochsDataLoader, PrefetchLoader,
_worker_init, fast_collate)
from image_classification.pt.src.datasets.augmentations.augs import (
IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD)
from torch.utils.data import Dataset
from torchvision import transforms
from PIL import Image
import os
class PredictionDataset(Dataset):
def __init__(self, folder_path, transform=None):
self.folder_path = folder_path
self.image_paths = [
os.path.join(folder_path, f)
for f in os.listdir(folder_path)
if f.lower().endswith((".png", ".jpg", ".jpeg"))
]
self.transform = transform or transforms.Compose([
transforms.Resize((224, 224)),
transforms.ToTensor(),
transforms.Normalize(mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]),
])
def __getitem__(self, idx):
img_path = self.image_paths[idx]
img = Image.open(img_path).convert("RGB")
img = self.transform(img)
return img, img_path
def __len__(self):
return len(self.image_paths)
def create_loader(
dataset,
input_size,
batch_size,
is_training=False,
use_prefetcher=True,
no_aug=False,
re_prob=0.,
re_mode='const',
re_count=1,
re_num_splits=0,
num_aug_repeats=0,
mean=IMAGENET_DEFAULT_MEAN,
std=IMAGENET_DEFAULT_STD,
num_workers=1,
distributed=False,
collate_fn=None,
pin_memory=False,
fp16=False, # deprecated, use img_dtype
img_dtype=torch.float32,
device=torch.device('cuda'),
use_multi_epochs_loader=False,
persistent_workers=True,
worker_seeding='all',
):
if isinstance(input_size, int):
input_size = (3, input_size, input_size)
if isinstance(dataset, IterableImageDataset):
# give Iterable datasets early knowledge of num_workers so that sample estimates
# are correct before worker processes are launched
dataset.set_loader_cfg(num_workers=num_workers)
sampler = None
if distributed and not isinstance(dataset, torch.utils.data.IterableDataset):
if is_training:
if num_aug_repeats:
sampler = RepeatAugSampler(dataset, num_repeats=num_aug_repeats)
else:
sampler = torch.utils.data.distributed.DistributedSampler(dataset)
else:
# This will add extra duplicate entries to result in equal num
# of samples per-process, will slightly alter validation results
sampler = OrderedDistributedSampler(dataset)
else:
assert num_aug_repeats == 0, "RepeatAugment not currently supported in non-distributed or IterableDataset use"
if collate_fn is None:
collate_fn = fast_collate if use_prefetcher else torch.utils.data.dataloader.default_collate
loader_class = torch.utils.data.DataLoader
if use_multi_epochs_loader:
loader_class = MultiEpochsDataLoader
loader_args = {
'batch_size': batch_size,
'shuffle': not isinstance(dataset, torch.utils.data.IterableDataset) and sampler is None and is_training,
'num_workers': num_workers,
'sampler': sampler,
'collate_fn': collate_fn,
'pin_memory': pin_memory,
'drop_last': is_training,
'worker_init_fn': partial(_worker_init, worker_seeding=worker_seeding),
'persistent_workers': persistent_workers
}
try:
loader = loader_class(dataset, **loader_args)
except TypeError:
loader_args.pop('persistent_workers') # only in Pytorch 1.7+
loader = loader_class(dataset, **loader_args)
if use_prefetcher:
prefetch_re_prob = re_prob if is_training and not no_aug else 0.
loader = PrefetchLoader( # pylint: disable=unexpected-keyword-arg
loader,
mean=mean,
std=std,
channels=input_size[0],
device=device,
fp16=fp16, # deprecated, use img_dtype
img_dtype=img_dtype,
re_prob=prefetch_re_prob,
re_mode=re_mode,
re_count=re_count,
re_num_splits=re_num_splits
)
return loader
from torch.utils.data.dataloader import default_collate
from torch.utils.data.distributed import DistributedSampler as DS
def get_dataloader(
dataset, batch_size=32, num_workers=4, fp16=False, distributed=False, shuffle=False,
collate_fn=None, device="cuda"
):
if collate_fn is None:
collate_fn = default_collate
def half_precision(x):
x = collate_fn(x)
x = [_x.half() if isinstance(_x, torch.FloatTensor) else _x for _x in x]
return x
dataloader = torch.utils.data.DataLoader(
dataset,
batch_size=batch_size,
shuffle=shuffle,
num_workers=num_workers,
collate_fn= half_precision if fp16 else collate_fn,
sampler=DS(dataset) if distributed else None,
)
return dataloader
from timm import utils
from timm.data import FastCollateMixup
def prepare_kwargs_for_dataloader(cfg):
#####################################
# Hydra config to flat config with last keys
#####################################
from types import SimpleNamespace
from common.utils import flatten_config
# exceptional handling of scale variable. as config has two variables of same name.
# and when u flatten config they both becomes same
scale = None
if "data_augmentation" in cfg and "scale" in cfg.data_augmentation:
scale = cfg.data_augmentation.scale
if not scale:
scale = [0.08, 1.0]
# convert hydra heirarchical config to flat config
args_raw = flatten_config(cfg)
# convert dict based config to argument parser type
if isinstance(args_raw, dict):
args = SimpleNamespace(**args_raw)
else:
args = args_raw
#####################################
args.prefetcher = not getattr(args, 'no_prefetcher', False)
if not hasattr(args, "batch_size") or args.batch_size is None:
args.batch_size = 128
if not hasattr(args, "validation_batch_size") or args.validation_batch_size is None:
args.validation_batch_size = args.batch_size
num_aug_splits = 0
collate_fn = None
args.mixup = getattr(args, "mixup", 0)
args.cutmix = getattr(args, "cutmix", 0)
args.cutmix_minmax = getattr(args, "cutmix_minmax", None)
mixup_active = (args.mixup > 0) or (args.cutmix > 0) or (args.cutmix_minmax is not None)
# with current config following "if" is false
if mixup_active:
mixup_args = dict(
mixup_alpha=args.mixup,
cutmix_alpha=args.cutmix,
cutmix_minmax=args.cutmix_minmax,
prob=getattr(args, "mixup_prob", 1.0),
switch_prob=getattr(args, "mixup_switch_prob", 0.5),
mode=getattr(args, "mixup_mode", 'batch'),
label_smoothing=getattr(args, "smoothing", 0.1),
num_classes=args.num_classes
)
if args.prefetcher:
assert not num_aug_splits # collate conflict (need to support deinterleaving in collate mixup)
collate_fn = FastCollateMixup(**mixup_args)
device = args.device # already set in the main py
if getattr(args, 'aug_splits', 0) > 0:
assert args.aug_splits > 1, 'A split of 1 makes no sense'
num_aug_splits = args.aug_splits
extra_loader_kwargs = {}
if getattr(args, 'dataset_name', '') in ['imagenet','custom']:
extra_loader_kwargs = {
'train_split': getattr(args, 'train_split', 'train'),
'val_split': getattr(args, 'val_split', 'val'),
'class_map': getattr(args, 'class_map', None),
'seed': getattr(args, 'seed', None),
'repeats': getattr(args, 'epoch_repeats', None),
}
# create data loaders w/ augmentation pipeline
test_interpolation = 'bicubic'
train_interpolation = getattr(args, 'train_interpolation', "random")
args.no_aug = getattr(args, 'no_aug', False)
if args.no_aug or not train_interpolation:
train_interpolation = test_interpolation
args.mean= getattr(args, 'mean', [0.485, 0.456, 0.406])
args.std= getattr(args, 'std', [0.229, 0.224, 0.225])
batch_size = getattr(args, 'batch_size', 128)
validation_batch_size = getattr(args, 'validation_batch_size', batch_size)
loader_kwargs = dict(
img_size=tuple(args.input_shape),
batch_size=batch_size,
test_batch_size=validation_batch_size,
download=getattr(args, 'data_download', False),
distributed=getattr(args, 'distributed', False), # created same way as before Nikhil
use_prefetcher=args.prefetcher, # created same way as before Nikhil
no_aug=args.no_aug,
re_prob=getattr(args, 'reprob', 0),
re_mode=getattr(args, 'remode', 'pixel'),
re_count=getattr(args, 'recount', 1),
re_split=getattr(args, 'resplit', False),
scale=scale,
ratio=getattr(args, 'ratio', [0.75, 1.33]),
hflip=getattr(args, 'hflip', 0.5),
vflip=getattr(args, 'vflip', 0.0),
color_jitter=getattr(args, 'color_jitter', 0.4),
auto_augment=getattr(args, 'aa', None),
num_aug_repeats=getattr(args, 'aug_repeats', 0),
num_aug_splits=num_aug_splits,
train_interpolation=train_interpolation,
test_interpolation=test_interpolation, # the way it was done in
mean=args.mean, # from config directly Nikhil
std=args.std, # from config directly Nikhil
num_workers=getattr(args, 'workers', 4),
collate_fn=collate_fn, # created same way as before Nikhil
pin_memory=getattr(args, 'pin_mem', False),
device=device, # TODO multi gpu
use_multi_epochs_loader=getattr(args, 'use_multi_epochs_loader', False),
worker_seeding=getattr(args, 'worker_seeding', False),
test_path=getattr(args, 'test_path', None),
prediction_path=getattr(args, 'prediction_path', None),
quantization_path=getattr(args, 'quantization_path', None),
quantization_split = getattr(args, 'quantization_split', 0.1),
num_classes = args.num_classes, # has to be there
**extra_loader_kwargs)
return loader_kwargs
def get_from_config(cfg, key_path, default=None):
"""
Safely get a nested value from cfg using dot-separated keys.
Works for OmegaConf DictConfig, dict, or objects with attributes.
Example:
get_from_config(cfg, "a.b.c", default=42)
"""
keys = key_path.split(".")
current = cfg
for key in keys:
if isinstance(current, dict):
if key not in current:
return default
current = current[key]
elif hasattr(current, key):
current = get_from_config(current, key)
else:
return default
return current
|