Add configs for MNIST, CIFAR10 and ImageNet
yanglei
yl-1993
parent
563fa63d06
commit
5c88675a71
|
|
@ -0,0 +1,30 @@
|
|||
# dataset settings
|
||||
dataset_type = 'CIFAR10'
|
||||
img_norm_cfg = dict(
|
||||
mean=[125.307, 122.961, 113.8575],
|
||||
std=[51.5865, 50.847, 51.255],
|
||||
to_rgb=True)
|
||||
train_pipeline = [
|
||||
dict(type='RandomCrop', size=32, padding=4),
|
||||
dict(type='RandomFlip', flip_prob=0.5, direction='horizontal'),
|
||||
dict(type='Normalize', **img_norm_cfg),
|
||||
dict(type='ImageToTensor', keys=['img']),
|
||||
dict(type='ToTensor', keys=['gt_label']),
|
||||
dict(type='Collect', keys=['img', 'gt_label'])
|
||||
]
|
||||
test_pipeline = [
|
||||
dict(type='Normalize', **img_norm_cfg),
|
||||
dict(type='ImageToTensor', keys=['img']),
|
||||
dict(type='ToTensor', keys=['gt_label']),
|
||||
dict(type='Collect', keys=['img', 'gt_label'])
|
||||
]
|
||||
data = dict(
|
||||
samples_per_gpu=128,
|
||||
workers_per_gpu=2,
|
||||
train=dict(
|
||||
type=dataset_type, data_prefix='data/cifar10',
|
||||
pipeline=train_pipeline),
|
||||
val=dict(
|
||||
type=dataset_type, data_prefix='data/cifar10', pipeline=test_pipeline),
|
||||
test=dict(
|
||||
type=dataset_type, data_prefix='data/cifar10', pipeline=test_pipeline))
|
||||
|
|
@ -0,0 +1,41 @@
|
|||
# dataset settings
|
||||
dataset_type = 'ImageNet'
|
||||
img_norm_cfg = dict(
|
||||
mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True)
|
||||
train_pipeline = [
|
||||
dict(type='LoadImageFromFile'),
|
||||
dict(type='RandomResizedCrop', size=224),
|
||||
dict(type='RandomFlip', flip_prob=0.5, direction='horizontal'),
|
||||
dict(type='Normalize', **img_norm_cfg),
|
||||
dict(type='ImageToTensor', keys=['img']),
|
||||
dict(type='ToTensor', keys=['gt_label']),
|
||||
dict(type='Collect', keys=['img', 'gt_label'])
|
||||
]
|
||||
test_pipeline = [
|
||||
dict(type='LoadImageFromFile'),
|
||||
dict(type='Resize', size=256),
|
||||
dict(type='CenterCrop', crop_size=224),
|
||||
dict(type='Normalize', **img_norm_cfg),
|
||||
dict(type='ImageToTensor', keys=['img']),
|
||||
dict(type='ToTensor', keys=['gt_label']),
|
||||
dict(type='Collect', keys=['img', 'gt_label'])
|
||||
]
|
||||
data = dict(
|
||||
samples_per_gpu=32,
|
||||
workers_per_gpu=2,
|
||||
train=dict(
|
||||
type=dataset_type,
|
||||
data_prefix='data/imagenet/train',
|
||||
pipeline=train_pipeline),
|
||||
val=dict(
|
||||
type=dataset_type,
|
||||
data_prefix='data/imagenet/val',
|
||||
ann_file='data/imagenet/meta/val.txt',
|
||||
pipeline=test_pipeline),
|
||||
test=dict(
|
||||
# replace `data/val` with `data/test` for standard test
|
||||
type=dataset_type,
|
||||
data_prefix='data/imagenet/val',
|
||||
ann_file='data/imagenet/meta/val.txt',
|
||||
pipeline=test_pipeline))
|
||||
evaluation = dict(interval=1, metric='accuracy')
|
||||
|
|
@ -0,0 +1,41 @@
|
|||
# dataset settings
|
||||
dataset_type = 'ImageNet'
|
||||
img_norm_cfg = dict(
|
||||
mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True)
|
||||
train_pipeline = [
|
||||
dict(type='LoadImageFromFile'),
|
||||
dict(type='RandomResizedCrop', size=224),
|
||||
dict(type='RandomFlip', flip_prob=0.5, direction='horizontal'),
|
||||
dict(type='Normalize', **img_norm_cfg),
|
||||
dict(type='ImageToTensor', keys=['img']),
|
||||
dict(type='ToTensor', keys=['gt_label']),
|
||||
dict(type='Collect', keys=['img', 'gt_label'])
|
||||
]
|
||||
test_pipeline = [
|
||||
dict(type='LoadImageFromFile'),
|
||||
dict(type='Resize', size=256),
|
||||
dict(type='CenterCrop', crop_size=224),
|
||||
dict(type='Normalize', **img_norm_cfg),
|
||||
dict(type='ImageToTensor', keys=['img']),
|
||||
dict(type='ToTensor', keys=['gt_label']),
|
||||
dict(type='Collect', keys=['img', 'gt_label'])
|
||||
]
|
||||
data = dict(
|
||||
samples_per_gpu=64,
|
||||
workers_per_gpu=2,
|
||||
train=dict(
|
||||
type=dataset_type,
|
||||
data_prefix='data/imagenet/train',
|
||||
pipeline=train_pipeline),
|
||||
val=dict(
|
||||
type=dataset_type,
|
||||
data_prefix='data/imagenet/val',
|
||||
ann_file='data/imagenet/meta/val.txt',
|
||||
pipeline=test_pipeline),
|
||||
test=dict(
|
||||
# replace `data/val` with `data/test` for standard test
|
||||
type=dataset_type,
|
||||
data_prefix='data/imagenet/val',
|
||||
ann_file='data/imagenet/meta/val.txt',
|
||||
pipeline=test_pipeline))
|
||||
evaluation = dict(interval=1, metric='accuracy')
|
||||
|
|
@ -0,0 +1,15 @@
|
|||
# checkpoint saving
|
||||
checkpoint_config = dict(interval=1)
|
||||
# yapf:disable
|
||||
log_config = dict(
|
||||
interval=100,
|
||||
hooks=[
|
||||
dict(type='TextLoggerHook'),
|
||||
# dict(type='TensorboardLoggerHook')
|
||||
])
|
||||
# yapf:enable
|
||||
dist_params = dict(backend='nccl')
|
||||
log_level = 'INFO'
|
||||
load_from = None
|
||||
resume_from = None
|
||||
workflow = [('train', 1)]
|
||||
|
|
@ -0,0 +1,12 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(type='MobileNetV2', widen_factor=1.0),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=1000,
|
||||
in_channels=1280,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
topk=(1, 5),
|
||||
))
|
||||
|
|
@ -0,0 +1,17 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(
|
||||
type='ResNet',
|
||||
depth=101,
|
||||
num_stages=4,
|
||||
out_indices=(3, ),
|
||||
style='pytorch'),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=1000,
|
||||
in_channels=2048,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
topk=(1, 5),
|
||||
))
|
||||
|
|
@ -0,0 +1,17 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(
|
||||
type='ResNet',
|
||||
depth=152,
|
||||
num_stages=4,
|
||||
out_indices=(3, ),
|
||||
style='pytorch'),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=1000,
|
||||
in_channels=2048,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
topk=(1, 5),
|
||||
))
|
||||
|
|
@ -0,0 +1,17 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(
|
||||
type='ResNet',
|
||||
depth=18,
|
||||
num_stages=4,
|
||||
out_indices=(3, ),
|
||||
style='pytorch'),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=1000,
|
||||
in_channels=512,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
topk=(1, 5),
|
||||
))
|
||||
|
|
@ -0,0 +1,16 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(
|
||||
type='ResNet_CIFAR',
|
||||
depth=18,
|
||||
num_stages=4,
|
||||
out_indices=(3, ),
|
||||
style='pytorch'),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=10,
|
||||
in_channels=512,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
))
|
||||
|
|
@ -0,0 +1,17 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(
|
||||
type='ResNet',
|
||||
depth=34,
|
||||
num_stages=4,
|
||||
out_indices=(3, ),
|
||||
style='pytorch'),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=1000,
|
||||
in_channels=512,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
topk=(1, 5),
|
||||
))
|
||||
|
|
@ -0,0 +1,17 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(
|
||||
type='ResNet',
|
||||
depth=50,
|
||||
num_stages=4,
|
||||
out_indices=(3, ),
|
||||
style='pytorch'),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=1000,
|
||||
in_channels=2048,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
topk=(1, 5),
|
||||
))
|
||||
|
|
@ -0,0 +1,16 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(
|
||||
type='ResNet_CIFAR',
|
||||
depth=50,
|
||||
num_stages=4,
|
||||
out_indices=(3, ),
|
||||
style='pytorch'),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=10,
|
||||
in_channels=2048,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
))
|
||||
|
|
@ -0,0 +1,17 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(
|
||||
type='ResNetV1d',
|
||||
depth=101,
|
||||
num_stages=4,
|
||||
out_indices=(3, ),
|
||||
style='pytorch'),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=1000,
|
||||
in_channels=2048,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
topk=(1, 5),
|
||||
))
|
||||
|
|
@ -0,0 +1,17 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(
|
||||
type='ResNetV1d',
|
||||
depth=152,
|
||||
num_stages=4,
|
||||
out_indices=(3, ),
|
||||
style='pytorch'),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=1000,
|
||||
in_channels=2048,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
topk=(1, 5),
|
||||
))
|
||||
|
|
@ -0,0 +1,17 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(
|
||||
type='ResNetV1d',
|
||||
depth=50,
|
||||
num_stages=4,
|
||||
out_indices=(3, ),
|
||||
style='pytorch'),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=1000,
|
||||
in_channels=2048,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
topk=(1, 5),
|
||||
))
|
||||
|
|
@ -0,0 +1,19 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(
|
||||
type='ResNeXt',
|
||||
depth=101,
|
||||
num_stages=4,
|
||||
out_indices=(3, ),
|
||||
groups=32,
|
||||
width_per_group=4,
|
||||
style='pytorch'),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=1000,
|
||||
in_channels=2048,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
topk=(1, 5),
|
||||
))
|
||||
|
|
@ -0,0 +1,19 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(
|
||||
type='ResNeXt',
|
||||
depth=101,
|
||||
num_stages=4,
|
||||
out_indices=(3, ),
|
||||
groups=32,
|
||||
width_per_group=8,
|
||||
style='pytorch'),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=1000,
|
||||
in_channels=2048,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
topk=(1, 5),
|
||||
))
|
||||
|
|
@ -0,0 +1,19 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(
|
||||
type='ResNeXt',
|
||||
depth=152,
|
||||
num_stages=4,
|
||||
out_indices=(3, ),
|
||||
groups=32,
|
||||
width_per_group=4,
|
||||
style='pytorch'),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=1000,
|
||||
in_channels=2048,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
topk=(1, 5),
|
||||
))
|
||||
|
|
@ -0,0 +1,19 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(
|
||||
type='ResNeXt',
|
||||
depth=50,
|
||||
num_stages=4,
|
||||
out_indices=(3, ),
|
||||
groups=32,
|
||||
width_per_group=4,
|
||||
style='pytorch'),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=1000,
|
||||
in_channels=2048,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
topk=(1, 5),
|
||||
))
|
||||
|
|
@ -0,0 +1,17 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(
|
||||
type='SEResNet',
|
||||
depth=101,
|
||||
num_stages=4,
|
||||
out_indices=(3, ),
|
||||
style='pytorch'),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=1000,
|
||||
in_channels=2048,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
topk=(1, 5),
|
||||
))
|
||||
|
|
@ -0,0 +1,17 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(
|
||||
type='SEResNet',
|
||||
depth=50,
|
||||
num_stages=4,
|
||||
out_indices=(3, ),
|
||||
style='pytorch'),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=1000,
|
||||
in_channels=2048,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
topk=(1, 5),
|
||||
))
|
||||
|
|
@ -0,0 +1,20 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(
|
||||
type='SEResNeXt',
|
||||
depth=101,
|
||||
num_stages=4,
|
||||
out_indices=(3, ),
|
||||
groups=32,
|
||||
width_per_group=4,
|
||||
se_ratio=16,
|
||||
style='pytorch'),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=1000,
|
||||
in_channels=2048,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
topk=(1, 5),
|
||||
))
|
||||
|
|
@ -0,0 +1,20 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(
|
||||
type='SEResNeXt',
|
||||
depth=50,
|
||||
num_stages=4,
|
||||
out_indices=(3, ),
|
||||
groups=32,
|
||||
width_per_group=4,
|
||||
se_ratio=16,
|
||||
style='pytorch'),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=1000,
|
||||
in_channels=2048,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
topk=(1, 5),
|
||||
))
|
||||
|
|
@ -0,0 +1,11 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(type='ShuffleNetV1', groups=3),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=1000,
|
||||
in_channels=960,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
))
|
||||
|
|
@ -0,0 +1,10 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(type='ShuffleNetV2', widen_factor=1.0),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearClsHead',
|
||||
num_classes=1000,
|
||||
in_channels=1024,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0)))
|
||||
|
|
@ -0,0 +1,6 @@
|
|||
# optimizer
|
||||
optimizer = dict(type='SGD', lr=0.1, momentum=0.9, weight_decay=0.0001)
|
||||
optimizer_config = dict(grad_clip=None)
|
||||
# learning policy
|
||||
lr_config = dict(policy='step', step=[100, 150])
|
||||
total_epochs = 200
|
||||
|
|
@ -0,0 +1,12 @@
|
|||
# optimizer
|
||||
optimizer = dict(type='SGD', lr=0.5, momentum=0.9, weight_decay=0.00004)
|
||||
optimizer_config = dict(grad_clip=None)
|
||||
# learning policy
|
||||
lr_config = dict(
|
||||
policy='poly',
|
||||
min_lr=0,
|
||||
by_epoch=False,
|
||||
warmup='constant',
|
||||
warmup_iters=5000,
|
||||
)
|
||||
total_epochs = 360
|
||||
|
|
@ -0,0 +1,12 @@
|
|||
# optimizer
|
||||
optimizer = dict(
|
||||
type='SGD', lr=0.8, momentum=0.9, weight_decay=0.0001, nesterov=True)
|
||||
optimizer_config = dict(grad_clip=None)
|
||||
# learning policy
|
||||
lr_config = dict(
|
||||
policy='step',
|
||||
warmup='linear',
|
||||
warmup_iters=2500,
|
||||
warmup_ratio=0.25,
|
||||
step=[30, 60, 90])
|
||||
total_epochs = 100
|
||||
|
|
@ -0,0 +1,12 @@
|
|||
# optimizer
|
||||
optimizer = dict(
|
||||
type='SGD', lr=0.8, momentum=0.9, weight_decay=0.0001, nesterov=True)
|
||||
optimizer_config = dict(grad_clip=None)
|
||||
# learning policy
|
||||
lr_config = dict(
|
||||
policy='CosineAnealing',
|
||||
min_lr=0,
|
||||
warmup='linear',
|
||||
warmup_iters=2500,
|
||||
warmup_ratio=0.25)
|
||||
total_epochs = 100
|
||||
|
|
@ -0,0 +1,12 @@
|
|||
# optimizer
|
||||
optimizer = dict(type='SGD', lr=0.5, momentum=0.9, weight_decay=0.00004)
|
||||
optimizer_config = dict(grad_clip=None)
|
||||
# learning policy
|
||||
lr_config = dict(
|
||||
policy='poly',
|
||||
min_lr=0,
|
||||
by_epoch=False,
|
||||
warmup='constant',
|
||||
warmup_iters=5000,
|
||||
)
|
||||
total_epochs = 300
|
||||
|
|
@ -0,0 +1,6 @@
|
|||
# optimizer
|
||||
optimizer = dict(type='SGD', lr=0.1, momentum=0.9, weight_decay=0.0001)
|
||||
optimizer_config = dict(grad_clip=None)
|
||||
# learning policy
|
||||
lr_config = dict(policy='step', step=[30, 60, 90])
|
||||
total_epochs = 100
|
||||
|
|
@ -0,0 +1,6 @@
|
|||
# optimizer
|
||||
optimizer = dict(type='SGD', lr=0.1, momentum=0.9, weight_decay=0.0001)
|
||||
optimizer_config = dict(grad_clip=None)
|
||||
# learning policy
|
||||
lr_config = dict(policy='step', step=[40, 80, 120])
|
||||
total_epochs = 140
|
||||
|
|
@ -0,0 +1,6 @@
|
|||
# optimizer
|
||||
optimizer = dict(type='SGD', lr=0.1, momentum=0.9, weight_decay=0.0001)
|
||||
optimizer_config = dict(grad_clip=None)
|
||||
# learning policy
|
||||
lr_config = dict(policy='CosineAnealing', min_lr=0)
|
||||
total_epochs = 100
|
||||
|
|
@ -0,0 +1,6 @@
|
|||
# optimizer
|
||||
optimizer = dict(type='SGD', lr=0.045, momentum=0.9, weight_decay=0.00004)
|
||||
optimizer_config = dict(grad_clip=None)
|
||||
# learning policy
|
||||
lr_config = dict(policy='step', gamma=0.98, step=1)
|
||||
total_epochs = 300
|
||||
|
|
@ -0,0 +1,4 @@
|
|||
_base_ = [
|
||||
'../_base_/models/resnet18_cifar.py', '../_base_/datasets/cifar10.py',
|
||||
'../_base_/schedules/cifar10.py', '../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,4 @@
|
|||
_base_ = [
|
||||
'../_base_/models/resnet50_cifar.py', '../_base_/datasets/cifar10.py',
|
||||
'../_base_/schedules/cifar10.py', '../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -1,59 +0,0 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='xxx',
|
||||
pretrained='torchvision://resnet50',
|
||||
backbone=dict(
|
||||
type='ResNet',
|
||||
depth=50,
|
||||
num_stages=4,
|
||||
out_indices=(0, 1, 2, 3),
|
||||
frozen_stages=1,
|
||||
style='pytorch'))
|
||||
# dataset settings
|
||||
dataset_type = 'XXXDataset'
|
||||
img_norm_cfg = dict(
|
||||
mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True)
|
||||
train_pipeline = []
|
||||
test_pipeline = []
|
||||
data = dict(
|
||||
imgs_per_gpu=2,
|
||||
workers_per_gpu=2,
|
||||
train=dict(
|
||||
type=dataset_type,
|
||||
ann_file='',
|
||||
data_prefix='',
|
||||
pipeline=train_pipeline),
|
||||
val=dict(
|
||||
type=dataset_type, ann_file='', data_prefix='',
|
||||
pipeline=test_pipeline),
|
||||
test=dict(
|
||||
type=dataset_type, ann_file='', data_prefix='',
|
||||
pipeline=test_pipeline))
|
||||
# optimizer
|
||||
optimizer = dict(type='SGD', lr=0.02, momentum=0.9, weight_decay=0.0001)
|
||||
optimizer_config = dict(grad_clip=None)
|
||||
# learning policy
|
||||
lr_config = dict(
|
||||
policy='step',
|
||||
warmup='linear',
|
||||
warmup_iters=500,
|
||||
warmup_ratio=1.0 / 3,
|
||||
step=[8, 11])
|
||||
# checkpoint saving
|
||||
checkpoint_config = dict(interval=1)
|
||||
# yapf:disable
|
||||
log_config = dict(
|
||||
interval=50,
|
||||
hooks=[
|
||||
dict(type='TextLoggerHook'),
|
||||
# dict(type='TensorboardLoggerHook')
|
||||
])
|
||||
# yapf:enable
|
||||
# runtime settings
|
||||
total_epochs = 12
|
||||
dist_params = dict(backend='nccl')
|
||||
log_level = 'INFO'
|
||||
work_dir = './work_dirs/xxx'
|
||||
load_from = None
|
||||
resume_from = None
|
||||
workflow = [('train', 1)]
|
||||
|
|
@ -0,0 +1,6 @@
|
|||
_base_ = [
|
||||
'../_base_/models/mobilenet_v2_1x.py',
|
||||
'../_base_/datasets/imagenet_bs32.py',
|
||||
'../_base_/schedules/imagenet_bs256_epochstep.py',
|
||||
'../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,4 @@
|
|||
_base_ = [
|
||||
'../_base_/models/resnet101.py', '../_base_/datasets/imagenet_bs32.py',
|
||||
'../_base_/schedules/imagenet_bs256.py', '../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,4 @@
|
|||
_base_ = [
|
||||
'../_base_/models/resnet152.py', '../_base_/datasets/imagenet_bs32.py',
|
||||
'../_base_/schedules/imagenet_bs256.py', '../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,4 @@
|
|||
_base_ = [
|
||||
'../_base_/models/resnet18.py', '../_base_/datasets/imagenet_bs32.py',
|
||||
'../_base_/schedules/imagenet_bs256.py', '../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,4 @@
|
|||
_base_ = [
|
||||
'../_base_/models/resnet34.py', '../_base_/datasets/imagenet_bs32.py',
|
||||
'../_base_/schedules/imagenet_bs256.py', '../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,4 @@
|
|||
_base_ = [
|
||||
'../_base_/models/resnet50.py', '../_base_/datasets/imagenet_bs64.py',
|
||||
'../_base_/schedules/imagenet_bs2048.py', '../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,5 @@
|
|||
_base_ = [
|
||||
'../_base_/models/resnet50.py', '../_base_/datasets/imagenet_bs64.py',
|
||||
'../_base_/schedules/imagenet_bs2048_coslr.py',
|
||||
'../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,12 @@
|
|||
_base_ = ['./resnet50_batch2048_warmup.py']
|
||||
model = dict(
|
||||
head=dict(
|
||||
type='LinearHead',
|
||||
num_classes=1000,
|
||||
in_channels=2048,
|
||||
loss=dict(
|
||||
type='LabelSmoothLoss',
|
||||
loss_weight=1.0,
|
||||
label_smooth_val=0.1,
|
||||
num_classes=1000),
|
||||
))
|
||||
|
|
@ -0,0 +1,4 @@
|
|||
_base_ = [
|
||||
'../_base_/models/resnet50.py', '../_base_/datasets/imagenet_bs32.py',
|
||||
'../_base_/schedules/imagenet_bs256.py', '../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,5 @@
|
|||
_base_ = [
|
||||
'../_base_/models/resnet50.py', '../_base_/datasets/imagenet_bs32.py',
|
||||
'../_base_/schedules/imagenet_bs256_coslr.py',
|
||||
'../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,12 @@
|
|||
_base_ = ['./resnet50_imagenet_bs256.py']
|
||||
model = dict(
|
||||
head=dict(
|
||||
type='LinearHead',
|
||||
num_classes=1000,
|
||||
in_channels=2048,
|
||||
loss=dict(
|
||||
type='LabelSmoothLoss',
|
||||
loss_weight=1.0,
|
||||
label_smooth_val=0.1,
|
||||
num_classes=1000),
|
||||
))
|
||||
|
|
@ -0,0 +1,98 @@
|
|||
import os.path as osp
|
||||
|
||||
# run with 8 GPUs
|
||||
|
||||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(
|
||||
type='ResNet',
|
||||
depth=50,
|
||||
num_stages=4,
|
||||
out_indices=(3, ),
|
||||
style='pytorch'),
|
||||
neck=dict(type='GlobalAveragePooling'),
|
||||
head=dict(
|
||||
type='LinearHead',
|
||||
num_classes=1000,
|
||||
in_channels=2048,
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
))
|
||||
# dataset settings
|
||||
dataset_type = 'ImageNet'
|
||||
img_norm_cfg = dict(
|
||||
mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True)
|
||||
memcached_root = '/mnt/lustre/share/memcached_client/'
|
||||
train_pipeline = [
|
||||
dict(
|
||||
type='LoadImageFromFile',
|
||||
file_client_args=dict(
|
||||
backend='memcached',
|
||||
server_list_cfg=osp.join(memcached_root, 'server_list.conf'),
|
||||
client_cfg=osp.join(memcached_root, 'client.conf'))),
|
||||
dict(type='ToPIL'),
|
||||
dict(type='RandomResizedCrop', size=224),
|
||||
dict(type='RandomHorizontalFlip', p=0.5),
|
||||
dict(type='ToNumpy'),
|
||||
dict(type='Normalize', **img_norm_cfg),
|
||||
dict(type='ImageToTensor', keys=['img']),
|
||||
dict(type='ToTensor', keys=['gt_labels']),
|
||||
dict(type='Collect', keys=['img', 'gt_labels'])
|
||||
]
|
||||
test_pipeline = [
|
||||
dict(
|
||||
type='LoadImageFromFile',
|
||||
file_client_args=dict(
|
||||
backend='memcached',
|
||||
server_list_cfg=osp.join(memcached_root, 'server_list.conf'),
|
||||
client_cfg=osp.join(memcached_root, 'client.conf'))),
|
||||
dict(type='ToPIL'),
|
||||
dict(type='Resize', size=256),
|
||||
dict(type='CenterCrop', size=224),
|
||||
dict(type='ToNumpy'),
|
||||
dict(type='Normalize', **img_norm_cfg),
|
||||
dict(type='ImageToTensor', keys=['img']),
|
||||
dict(type='ToTensor', keys=['gt_labels']),
|
||||
dict(type='Collect', keys=['img', 'gt_labels'])
|
||||
]
|
||||
data = dict(
|
||||
samples_per_gpu=64,
|
||||
workers_per_gpu=2,
|
||||
train=dict(
|
||||
type=dataset_type,
|
||||
data_prefix='data/imagenet/train',
|
||||
pipeline=train_pipeline),
|
||||
val=dict(
|
||||
type=dataset_type,
|
||||
data_prefix='data/imagenet/val',
|
||||
ann_file='data/imagenet/meta/val.txt',
|
||||
pipeline=test_pipeline),
|
||||
test=dict(
|
||||
# replace `data/val` with `data/test` for standard test
|
||||
type=dataset_type,
|
||||
data_prefix='data/imagenet/val',
|
||||
ann_file='data/imagenet/meta/val.txt',
|
||||
pipeline=test_pipeline))
|
||||
# optimizer
|
||||
optimizer = dict(type='SGD', lr=0.1, momentum=0.9, weight_decay=0.0001)
|
||||
optimizer_config = dict(grad_clip=None)
|
||||
# learning policy
|
||||
lr_config = dict(policy='step', step=[30, 60, 90])
|
||||
# checkpoint saving
|
||||
checkpoint_config = dict(interval=1)
|
||||
# yapf:disable
|
||||
log_config = dict(
|
||||
interval=100,
|
||||
hooks=[
|
||||
dict(type='TextLoggerHook'),
|
||||
# dict(type='TensorboardLoggerHook')
|
||||
])
|
||||
# yapf:enable
|
||||
# runtime settings
|
||||
total_epochs = 100
|
||||
dist_params = dict(backend='nccl')
|
||||
log_level = 'INFO'
|
||||
work_dir = './work_dirs/imagenet_resnet50_batch256'
|
||||
load_from = None
|
||||
resume_from = None
|
||||
workflow = [('train', 1)]
|
||||
|
|
@ -0,0 +1,4 @@
|
|||
_base_ = [
|
||||
'../_base_/models/resnetv1d101.py', '../_base_/datasets/imagenet_bs32.py',
|
||||
'../_base_/schedules/imagenet_bs256.py', '../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,4 @@
|
|||
_base_ = [
|
||||
'../_base_/models/resnetv1d152.py', '../_base_/datasets/imagenet_bs32.py',
|
||||
'../_base_/schedules/imagenet_bs256.py', '../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,4 @@
|
|||
_base_ = [
|
||||
'../_base_/models/resnetv1d50.py', '../_base_/datasets/imagenet_bs32.py',
|
||||
'../_base_/schedules/imagenet_bs256.py', '../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,5 @@
|
|||
_base_ = [
|
||||
'../_base_/models/resnext101_32x4d.py',
|
||||
'../_base_/datasets/imagenet_bs32.py',
|
||||
'../_base_/schedules/imagenet_bs256.py', '../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,5 @@
|
|||
_base_ = [
|
||||
'../_base_/models/resnext101_32x8d.py',
|
||||
'../_base_/datasets/imagenet_bs32.py',
|
||||
'../_base_/schedules/imagenet_bs256.py', '../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,5 @@
|
|||
_base_ = [
|
||||
'../_base_/models/resnext152_32x4d.py',
|
||||
'../_base_/datasets/imagenet_bs32.py',
|
||||
'../_base_/schedules/imagenet_bs256.py', '../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,5 @@
|
|||
_base_ = [
|
||||
'../_base_/models/resnext50_32x4d.py',
|
||||
'../_base_/datasets/imagenet_bs32.py',
|
||||
'../_base_/schedules/imagenet_bs256.py', '../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,4 @@
|
|||
_base_ = [
|
||||
'../_base_/models/seresnet101.py', '../_base_/datasets/imagenet_bs32.py',
|
||||
'../_base_/schedules/imagenet_bs256.py', '../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,5 @@
|
|||
_base_ = [
|
||||
'../_base_/models/seresnet50.py', '../_base_/datasets/imagenet_bs32.py',
|
||||
'../_base_/schedules/imagenet_bs256_140e.py',
|
||||
'../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,5 @@
|
|||
_base_ = [
|
||||
'../_base_/models/seresnext101_32x4d.py',
|
||||
'../_base_/datasets/imagenet_bs32.py',
|
||||
'../_base_/schedules/imagenet_bs256.py', '../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,5 @@
|
|||
_base_ = [
|
||||
'../_base_/models/seresnext50_32x4d.py',
|
||||
'../_base_/datasets/imagenet_bs32.py',
|
||||
'../_base_/schedules/imagenet_bs256.py', '../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,6 @@
|
|||
_base_ = [
|
||||
'../_base_/models/shufflenet_v1_1x.py',
|
||||
'../_base_/datasets/imagenet_bs64.py',
|
||||
'../_base_/schedules/imagenet_bs1024_lineardecay.py',
|
||||
'../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,6 @@
|
|||
_base_ = [
|
||||
'../_base_/models/shufflenet_v2_1x.py',
|
||||
'../_base_/datasets/imagenet_bs64.py',
|
||||
'../_base_/schedules/imagenet_bs2048_lineardecay.py',
|
||||
'../_base_/default_runtime.py'
|
||||
]
|
||||
|
|
@ -0,0 +1,56 @@
|
|||
# model settings
|
||||
model = dict(
|
||||
type='ImageClassifier',
|
||||
backbone=dict(type='LeNet5', num_classes=10),
|
||||
neck=None,
|
||||
head=dict(
|
||||
type='ClsHead',
|
||||
loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
|
||||
))
|
||||
# dataset settings
|
||||
dataset_type = 'MNIST'
|
||||
img_norm_cfg = dict(mean=[33.46], std=[78.87], to_rgb=True)
|
||||
train_pipeline = [
|
||||
dict(type='Resize', size=32),
|
||||
dict(type='Normalize', **img_norm_cfg),
|
||||
dict(type='ImageToTensor', keys=['img']),
|
||||
dict(type='ToTensor', keys=['gt_label']),
|
||||
]
|
||||
test_pipeline = [
|
||||
dict(type='Resize', size=32),
|
||||
dict(type='Normalize', **img_norm_cfg),
|
||||
dict(type='ImageToTensor', keys=['img']),
|
||||
dict(type='ToTensor', keys=['gt_label']),
|
||||
]
|
||||
data = dict(
|
||||
samples_per_gpu=128,
|
||||
workers_per_gpu=2,
|
||||
train=dict(
|
||||
type=dataset_type, data_prefix='data/mnist', pipeline=train_pipeline),
|
||||
val=dict(
|
||||
type=dataset_type, data_prefix='data/mnist', pipeline=test_pipeline),
|
||||
test=dict(
|
||||
type=dataset_type, data_prefix='data/mnist', pipeline=test_pipeline))
|
||||
# optimizer
|
||||
optimizer = dict(type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0001)
|
||||
optimizer_config = dict(grad_clip=None)
|
||||
# learning policy
|
||||
lr_config = dict(policy='step', step=[15])
|
||||
# checkpoint saving
|
||||
checkpoint_config = dict(interval=1)
|
||||
# yapf:disable
|
||||
log_config = dict(
|
||||
interval=100,
|
||||
hooks=[
|
||||
dict(type='TextLoggerHook'),
|
||||
# dict(type='TensorboardLoggerHook')
|
||||
])
|
||||
# yapf:enable
|
||||
# runtime settings
|
||||
total_epochs = 20
|
||||
dist_params = dict(backend='nccl')
|
||||
log_level = 'INFO'
|
||||
work_dir = './work_dirs/mnist/'
|
||||
load_from = None
|
||||
resume_from = None
|
||||
workflow = [('train', 1)]
|
||||
|
|
@ -4,7 +4,7 @@
|
|||
|
||||
Following typical conventions, we use `Dataset` and `DataLoader` for data loading
|
||||
with multiple workers. `Dataset` returns a dict of data items corresponding to
|
||||
the arguments of models' forward method.
|
||||
the arguments of models forward method.
|
||||
|
||||
The data preparation pipeline and the dataset is decomposed. Usually a dataset
|
||||
defines how to process the annotations and a data pipeline defines all the steps to prepare a data dict.
|
||||
|
|
@ -36,6 +36,22 @@ test_pipeline = [
|
|||
]
|
||||
```
|
||||
|
||||
By fault, `LoadImageFromFile` loads images from disk but it may lead to IO bottleneck for efficient small models.
|
||||
Various backends are supported by mmcv to accelerate this process. For example, if the training machines have setup
|
||||
[memcached](https://memcached.org/), we can revise the config as follows.
|
||||
```
|
||||
memcached_root = '/mnt/xxx/memcached_client/'
|
||||
train_pipeline = [
|
||||
dict(
|
||||
type='LoadImageFromFile',
|
||||
file_client_args=dict(
|
||||
backend='memcached',
|
||||
server_list_cfg=osp.join(memcached_root, 'server_list.conf'),
|
||||
client_cfg=osp.join(memcached_root, 'client.conf'))),
|
||||
]
|
||||
```
|
||||
More supported backends can be found in [mmcv.fileio.FileClient](https://github.com/open-mmlab/mmcv/blob/master/mmcv/fileio/file_client.py).
|
||||
|
||||
For each operation, we list the related dict fields that are added/updated/removed.
|
||||
At the end of the pipeline, we use `Collect` to only retain the necessary items for forward computation.
|
||||
|
||||
|
|
|
|||
Loading…
Reference in New Issue