mirrors/PaddleClas
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/PaddleClas.git synced 2025-06-03 21:55:06 +08:00
Code Issues Projects Releases Wiki Activity

add models

Browse Source
This commit is contained in:
zhangyubo0722 2022-12-19 12:53:26 +00:00 committed by cuicheng01
parent fdaf24ee78
commit 0140b38a80
29 changed files with 3446 additions and 20 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
ppcls/arch/backbone/__init__.py
View File

@ -29,7 +29,7 @@ from .legendary_models.pp_hgnet import PPHGNet_tiny, PPHGNet_small, PPHGNet_base
from .model_zoo.resnet_vc import ResNet50_vc
from .model_zoo.resnext import ResNeXt50_32x4d, ResNeXt50_64x4d, ResNeXt101_32x4d, ResNeXt101_64x4d, ResNeXt152_32x4d, ResNeXt152_64x4d
from .model_zoo.resnext_vd import ResNeXt50_vd_32x4d, ResNeXt50_vd_64x4d, ResNeXt101_vd_32x4d, ResNeXt101_vd_64x4d, ResNeXt152_vd_32x4d, ResNeXt152_vd_64x4d
from .model_zoo.res2net import Res2Net50_26w_4s, Res2Net50_14w_8s
from .model_zoo.res2net import Res2Net50_26w_4s, Res2Net50_14w_8s, Res2Net50_26w_6s, Res2Net50_26w_8s, Res2Net50_48w_2s, Res2Net101_26w_4s
from .model_zoo.res2net_vd import Res2Net50_vd_26w_4s, Res2Net101_vd_26w_4s, Res2Net200_vd_26w_4s
from .model_zoo.se_resnet_vd import SE_ResNet18_vd, SE_ResNet34_vd, SE_ResNet50_vd
from .model_zoo.se_resnext_vd import SE_ResNeXt50_vd_32x4d, SE_ResNeXt50_vd_32x4d, SENet154_vd
@ -37,7 +37,7 @@ from .model_zoo.se_resnext import SE_ResNeXt50_32x4d, SE_ResNeXt101_32x4d, SE_Re
from .model_zoo.dpn import DPN68, DPN92, DPN98, DPN107, DPN131
from .model_zoo.densenet import DenseNet121, DenseNet161, DenseNet169, DenseNet201, DenseNet264
from .model_zoo.efficientnet import EfficientNetB0, EfficientNetB1, EfficientNetB2, EfficientNetB3, EfficientNetB4, EfficientNetB5, EfficientNetB6, EfficientNetB7, EfficientNetB0_small
from .model_zoo.resnest import ResNeSt50_fast_1s1x64d, ResNeSt50, ResNeSt101
from .model_zoo.resnest import ResNeSt50_fast_1s1x64d, ResNeSt50, ResNeSt101, ResNeSt200, ResNeSt269
from .model_zoo.googlenet import GoogLeNet
from .model_zoo.mobilenet_v2 import MobileNetV2_x0_25, MobileNetV2_x0_5, MobileNetV2_x0_75, MobileNetV2, MobileNetV2_x1_5, MobileNetV2_x2_0
from .model_zoo.shufflenet_v2 import ShuffleNetV2_x0_25, ShuffleNetV2_x0_33, ShuffleNetV2_x0_5, ShuffleNetV2_x1_0, ShuffleNetV2_x1_5, ShuffleNetV2_x2_0, ShuffleNetV2_swish
@ -49,7 +49,7 @@ from .model_zoo.xception_deeplab import Xception41_deeplab, Xception65_deeplab
from .model_zoo.resnext101_wsl import ResNeXt101_32x8d_wsl, ResNeXt101_32x16d_wsl, ResNeXt101_32x32d_wsl, ResNeXt101_32x48d_wsl
from .model_zoo.squeezenet import SqueezeNet1_0, SqueezeNet1_1
from .model_zoo.darknet import DarkNet53
from .model_zoo.regnet import RegNetX_200MF, RegNetX_4GF, RegNetX_32GF, RegNetY_200MF, RegNetY_4GF, RegNetY_32GF
from .model_zoo.regnet import RegNetX_200MF, RegNetX_400MF, RegNetX_600MF, RegNetX_800MF, RegNetX_1600MF, RegNetX_3200MF, RegNetX_4GF, RegNetX_6400MF, RegNetX_8GF, RegNetX_12GF, RegNetX_16GF, RegNetX_32GF
from .model_zoo.vision_transformer import ViT_small_patch16_224, ViT_base_patch16_224, ViT_base_patch16_384, ViT_base_patch32_384, ViT_large_patch16_224, ViT_large_patch16_384, ViT_large_patch32_384
from .model_zoo.distilled_vision_transformer import DeiT_tiny_patch16_224, DeiT_small_patch16_224, DeiT_base_patch16_224, DeiT_tiny_distilled_patch16_224, DeiT_small_distilled_patch16_224, DeiT_base_distilled_patch16_224, DeiT_base_patch16_384, DeiT_base_distilled_patch16_384
from .legendary_models.swin_transformer import SwinTransformer_tiny_patch4_window7_224, SwinTransformer_small_patch4_window7_224, SwinTransformer_base_patch4_window7_224, SwinTransformer_base_patch4_window12_384, SwinTransformer_large_patch4_window7_224, SwinTransformer_large_patch4_window12_384
@ -60,15 +60,15 @@ from .model_zoo.gvt import pcpvt_small, pcpvt_base, pcpvt_large, alt_gvt_small,
from .model_zoo.levit import LeViT_128S, LeViT_128, LeViT_192, LeViT_256, LeViT_384
from .model_zoo.dla import DLA34, DLA46_c, DLA46x_c, DLA60, DLA60x, DLA60x_c, DLA102, DLA102x, DLA102x2, DLA169
from .model_zoo.rednet import RedNet26, RedNet38, RedNet50, RedNet101, RedNet152
from .model_zoo.tnt import TNT_small
from .model_zoo.tnt import TNT_small, TNT_base
from .model_zoo.hardnet import HarDNet68, HarDNet85, HarDNet39_ds, HarDNet68_ds
from .model_zoo.cspnet import CSPDarkNet53
from .model_zoo.pvt_v2 import PVT_V2_B0, PVT_V2_B1, PVT_V2_B2_Linear, PVT_V2_B2, PVT_V2_B3, PVT_V2_B4, PVT_V2_B5
from .model_zoo.mobilevit import MobileViT_XXS, MobileViT_XS, MobileViT_S
from .model_zoo.repvgg import RepVGG_A0, RepVGG_A1, RepVGG_A2, RepVGG_B0, RepVGG_B1, RepVGG_B2, RepVGG_B1g2, RepVGG_B1g4, RepVGG_B2g4, RepVGG_B3g4
from .model_zoo.van import VAN_B0
from .model_zoo.repvgg import RepVGG_A0, RepVGG_A1, RepVGG_A2, RepVGG_B0, RepVGG_B1, RepVGG_B2, RepVGG_B1g2, RepVGG_B1g4, RepVGG_B2g4, RepVGG_B3, RepVGG_B3g4, RepVGG_D2se
from .model_zoo.van import VAN_B0, VAN_B1, VAN_B2, VAN_B3
from .model_zoo.peleenet import PeleeNet
from .model_zoo.convnext import ConvNeXt_tiny
from .model_zoo.convnext import ConvNeXt_tiny, ConvNeXt_small, ConvNeXt_base_224, ConvNeXt_base_384, ConvNeXt_large_224, ConvNeXt_large_384
from .model_zoo.nextvit import NextViT_small_224, NextViT_base_224, NextViT_large_224, NextViT_small_384, NextViT_base_384, NextViT_large_384
from .model_zoo.cae import cae_base_patch16_224, cae_large_patch16_224

6
ppcls/arch/backbone/model_zoo/regnet.py
View File

@ -56,12 +56,6 @@ MODEL_URLS = {
"https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/RegNetX_16GF_pretrained.pdparams",
"RegNetX_32GF":
"https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/RegNetX_32GF_pretrained.pdparams",
"RegNetY_200MF":
"https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/RegNetY_200MF_pretrained.pdparams",
"RegNetY_4GF":
"https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/RegNetY_4GF_pretrained.pdparams",
"RegNetY_32GF":
"https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/RegNetY_32GF_pretrained.pdparams"
}
__all__ = list(MODEL_URLS.keys())

170
ppcls/configs/ImageNet/ConvNeXt/ConvNeXt_base_224.yaml Normal file
View File

@ -0,0 +1,170 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 300
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 224, 224]
save_inference_dir: ./inference
# training model under @to_static
to_static: False
update_freq: 4 # for 8 cards
# model ema
EMA:
decay: 0.9999
# model architecture
Arch:
name: ConvNeXt_base_224
class_num: 1000
drop_path_rate: 0.1
layer_scale_init_value: 1e-6
head_init_scale: 1.0
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
epsilon: 0.1
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: AdamW
beta1: 0.9
beta2: 0.999
epsilon: 1e-8
weight_decay: 0.05
one_dim_param_no_weight_decay: True
lr:
# for 8 cards
name: Cosine
learning_rate: 4e-3 # lr 4e-3 for total_batch_size 4096
eta_min: 1e-6
warmup_epoch: 20
warmup_start_lr: 0
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 224
interpolation: bicubic
backend: pil
- RandFlipImage:
flip_code: 1
- TimmAutoAugment:
config_str: rand-m9-mstd0.5-inc1
interpolation: bicubic
img_size: 224
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- RandomErasing:
EPSILON: 0.25
sl: 0.02
sh: 1.0/3.0
r1: 0.3
attempt: 10
use_log_aspect: True
mode: pixel
batch_transform_ops:
- OpSampler:
MixupOperator:
alpha: 0.8
prob: 0.5
CutmixOperator:
alpha: 1.0
prob: 0.5
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: True
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
interpolation: bicubic
backend: pil
- CropImage:
size: 224
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
interpolation: bicubic
backend: pil
- CropImage:
size: 224
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Eval:
- TopkAcc:
topk: [1, 5]

170
ppcls/configs/ImageNet/ConvNeXt/ConvNeXt_base_384.yaml Normal file
View File

@ -0,0 +1,170 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 300
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 384, 384]
save_inference_dir: ./inference
# training model under @to_static
to_static: False
update_freq: 4 # for 8 cards
# model ema
EMA:
decay: 0.9999
# model architecture
Arch:
name: ConvNeXt_base_384
class_num: 1000
drop_path_rate: 0.1
layer_scale_init_value: 1e-6
head_init_scale: 1.0
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
epsilon: 0.1
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: AdamW
beta1: 0.9
beta2: 0.999
epsilon: 1e-8
weight_decay: 0.05
one_dim_param_no_weight_decay: True
lr:
# for 8 cards
name: Cosine
learning_rate: 4e-3 # lr 4e-3 for total_batch_size 4096
eta_min: 1e-6
warmup_epoch: 20
warmup_start_lr: 0
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 384
interpolation: bicubic
backend: pil
- RandFlipImage:
flip_code: 1
- TimmAutoAugment:
config_str: rand-m9-mstd0.5-inc1
interpolation: bicubic
img_size: 384
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- RandomErasing:
EPSILON: 0.25
sl: 0.02
sh: 1.0/3.0
r1: 0.3
attempt: 10
use_log_aspect: True
mode: pixel
batch_transform_ops:
- OpSampler:
MixupOperator:
alpha: 0.8
prob: 0.5
CutmixOperator:
alpha: 1.0
prob: 0.5
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: True
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 384
interpolation: bicubic
backend: pil
- CropImage:
size: 384
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 384
interpolation: bicubic
backend: pil
- CropImage:
size: 384
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Eval:
- TopkAcc:
topk: [1, 5]

170
ppcls/configs/ImageNet/ConvNeXt/ConvNeXt_large_224.yaml Normal file
View File

@ -0,0 +1,170 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 300
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 224, 224]
save_inference_dir: ./inference
# training model under @to_static
to_static: False
update_freq: 4 # for 8 cards
# model ema
EMA:
decay: 0.9999
# model architecture
Arch:
name: ConvNeXt_large_224
class_num: 1000
drop_path_rate: 0.1
layer_scale_init_value: 1e-6
head_init_scale: 1.0
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
epsilon: 0.1
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: AdamW
beta1: 0.9
beta2: 0.999
epsilon: 1e-8
weight_decay: 0.05
one_dim_param_no_weight_decay: True
lr:
# for 8 cards
name: Cosine
learning_rate: 4e-3 # lr 4e-3 for total_batch_size 4096
eta_min: 1e-6
warmup_epoch: 20
warmup_start_lr: 0
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 224
interpolation: bicubic
backend: pil
- RandFlipImage:
flip_code: 1
- TimmAutoAugment:
config_str: rand-m9-mstd0.5-inc1
interpolation: bicubic
img_size: 224
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- RandomErasing:
EPSILON: 0.25
sl: 0.02
sh: 1.0/3.0
r1: 0.3
attempt: 10
use_log_aspect: True
mode: pixel
batch_transform_ops:
- OpSampler:
MixupOperator:
alpha: 0.8
prob: 0.5
CutmixOperator:
alpha: 1.0
prob: 0.5
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: True
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
interpolation: bicubic
backend: pil
- CropImage:
size: 224
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
interpolation: bicubic
backend: pil
- CropImage:
size: 224
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Eval:
- TopkAcc:
topk: [1, 5]

170
ppcls/configs/ImageNet/ConvNeXt/ConvNeXt_large_384.yaml Normal file
View File

@ -0,0 +1,170 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 300
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 384, 384]
save_inference_dir: ./inference
# training model under @to_static
to_static: False
update_freq: 4 # for 8 cards
# model ema
EMA:
decay: 0.9999
# model architecture
Arch:
name: ConvNeXt_large_384
class_num: 1000
drop_path_rate: 0.1
layer_scale_init_value: 1e-6
head_init_scale: 1.0
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
epsilon: 0.1
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: AdamW
beta1: 0.9
beta2: 0.999
epsilon: 1e-8
weight_decay: 0.05
one_dim_param_no_weight_decay: True
lr:
# for 8 cards
name: Cosine
learning_rate: 4e-3 # lr 4e-3 for total_batch_size 4096
eta_min: 1e-6
warmup_epoch: 20
warmup_start_lr: 0
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 384
interpolation: bicubic
backend: pil
- RandFlipImage:
flip_code: 1
- TimmAutoAugment:
config_str: rand-m9-mstd0.5-inc1
interpolation: bicubic
img_size: 384
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- RandomErasing:
EPSILON: 0.25
sl: 0.02
sh: 1.0/3.0
r1: 0.3
attempt: 10
use_log_aspect: True
mode: pixel
batch_transform_ops:
- OpSampler:
MixupOperator:
alpha: 0.8
prob: 0.5
CutmixOperator:
alpha: 1.0
prob: 0.5
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: True
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 384
interpolation: bicubic
backend: pil
- CropImage:
size: 384
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 384
interpolation: bicubic
backend: pil
- CropImage:
size: 384
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Eval:
- TopkAcc:
topk: [1, 5]

170
ppcls/configs/ImageNet/ConvNeXt/ConvNeXt_small.yaml Normal file
View File

@ -0,0 +1,170 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 300
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 224, 224]
save_inference_dir: ./inference
# training model under @to_static
to_static: False
update_freq: 4 # for 8 cards
# model ema
EMA:
decay: 0.9999
# model architecture
Arch:
name: ConvNeXt_small
class_num: 1000
drop_path_rate: 0.1
layer_scale_init_value: 1e-6
head_init_scale: 1.0
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
epsilon: 0.1
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: AdamW
beta1: 0.9
beta2: 0.999
epsilon: 1e-8
weight_decay: 0.05
one_dim_param_no_weight_decay: True
lr:
# for 8 cards
name: Cosine
learning_rate: 4e-3 # lr 4e-3 for total_batch_size 4096
eta_min: 1e-6
warmup_epoch: 20
warmup_start_lr: 0
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 224
interpolation: bicubic
backend: pil
- RandFlipImage:
flip_code: 1
- TimmAutoAugment:
config_str: rand-m9-mstd0.5-inc1
interpolation: bicubic
img_size: 224
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- RandomErasing:
EPSILON: 0.25
sl: 0.02
sh: 1.0/3.0
r1: 0.3
attempt: 10
use_log_aspect: True
mode: pixel
batch_transform_ops:
- OpSampler:
MixupOperator:
alpha: 0.8
prob: 0.5
CutmixOperator:
alpha: 1.0
prob: 0.5
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: True
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
interpolation: bicubic
backend: pil
- CropImage:
size: 224
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
interpolation: bicubic
backend: pil
- CropImage:
size: 224
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Eval:
- TopkAcc:
topk: [1, 5]

130
ppcls/configs/ImageNet/RegNet/RegNetX_12GF.yaml Normal file
View File

@ -0,0 +1,130 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 120
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 256, 256]
save_inference_dir: ./inference
# model architecture
Arch:
name: RegNetX_12GF
class_num: 1000
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: Momentum
momentum: 0.9
lr:
name: Piecewise
learning_rate: 0.1
decay_epochs: [30, 60, 90]
values: [0.1, 0.01, 0.001, 0.0001]
regularizer:
name: 'L2'
coeff: 0.0001
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 224
- RandFlipImage:
flip_code: 1
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Train:
- TopkAcc:
topk: [1, 5]
Eval:
- TopkAcc:
topk: [1, 5]

130
ppcls/configs/ImageNet/RegNet/RegNetX_1600MF.yaml Normal file
View File

@ -0,0 +1,130 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 120
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 256, 256]
save_inference_dir: ./inference
# model architecture
Arch:
name: RegNetX_1600MF
class_num: 1000
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: Momentum
momentum: 0.9
lr:
name: Piecewise
learning_rate: 0.1
decay_epochs: [30, 60, 90]
values: [0.1, 0.01, 0.001, 0.0001]
regularizer:
name: 'L2'
coeff: 0.0001
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 224
- RandFlipImage:
flip_code: 1
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Train:
- TopkAcc:
topk: [1, 5]
Eval:
- TopkAcc:
topk: [1, 5]

130
ppcls/configs/ImageNet/RegNet/RegNetX_16GF.yaml Normal file
View File

@ -0,0 +1,130 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 120
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 256, 256]
save_inference_dir: ./inference
# model architecture
Arch:
name: RegNetX_16GF
class_num: 1000
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: Momentum
momentum: 0.9
lr:
name: Piecewise
learning_rate: 0.1
decay_epochs: [30, 60, 90]
values: [0.1, 0.01, 0.001, 0.0001]
regularizer:
name: 'L2'
coeff: 0.0001
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 224
- RandFlipImage:
flip_code: 1
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 64
drop_last: False
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Train:
- TopkAcc:
topk: [1, 5]
Eval:
- TopkAcc:
topk: [1, 5]

130
ppcls/configs/ImageNet/RegNet/RegNetX_200MF.yaml Normal file
View File

@ -0,0 +1,130 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 120
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 256, 256]
save_inference_dir: ./inference
# model architecture
Arch:
name: RegNetX_200MF
class_num: 1000
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: Momentum
momentum: 0.9
lr:
name: Piecewise
learning_rate: 0.1
decay_epochs: [30, 60, 90]
values: [0.1, 0.01, 0.001, 0.0001]
regularizer:
name: 'L2'
coeff: 0.0001
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 224
- RandFlipImage:
flip_code: 1
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 64
drop_last: False
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Train:
- TopkAcc:
topk: [1, 5]
Eval:
- TopkAcc:
topk: [1, 5]

130
ppcls/configs/ImageNet/RegNet/RegNetX_3200MF.yaml Normal file
View File

@ -0,0 +1,130 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 120
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 256, 256]
save_inference_dir: ./inference
# model architecture
Arch:
name: RegNetY_3200MF
class_num: 1000
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: Momentum
momentum: 0.9
lr:
name: Piecewise
learning_rate: 0.1
decay_epochs: [30, 60, 90]
values: [0.1, 0.01, 0.001, 0.0001]
regularizer:
name: 'L2'
coeff: 0.0001
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 224
- RandFlipImage:
flip_code: 1
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 64
drop_last: False
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Train:
- TopkAcc:
topk: [1, 5]
Eval:
- TopkAcc:
topk: [1, 5]

130
ppcls/configs/ImageNet/RegNet/RegNetX_32GF.yaml Normal file
View File

@ -0,0 +1,130 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 120
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 256, 256]
save_inference_dir: ./inference
# model architecture
Arch:
name: RegNetX_32GF
class_num: 1000
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: Momentum
momentum: 0.9
lr:
name: Piecewise
learning_rate: 0.1
decay_epochs: [30, 60, 90]
values: [0.1, 0.01, 0.001, 0.0001]
regularizer:
name: 'L2'
coeff: 0.0001
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 224
- RandFlipImage:
flip_code: 1
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 64
drop_last: False
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Train:
- TopkAcc:
topk: [1, 5]
Eval:
- TopkAcc:
topk: [1, 5]

130
ppcls/configs/ImageNet/RegNet/RegNetX_400MF.yaml Normal file
View File

@ -0,0 +1,130 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 120
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 256, 256]
save_inference_dir: ./inference
# model architecture
Arch:
name: RegNetX_400MF
class_num: 1000
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: Momentum
momentum: 0.9
lr:
name: Piecewise
learning_rate: 0.1
decay_epochs: [30, 60, 90]
values: [0.1, 0.01, 0.001, 0.0001]
regularizer:
name: 'L2'
coeff: 0.0001
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 224
- RandFlipImage:
flip_code: 1
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 64
drop_last: False
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Train:
- TopkAcc:
topk: [1, 5]
Eval:
- TopkAcc:
topk: [1, 5]

130
ppcls/configs/ImageNet/RegNet/RegNetX_600MF.yaml Normal file
View File

@ -0,0 +1,130 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 120
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 256, 256]
save_inference_dir: ./inference
# model architecture
Arch:
name: RegNetX_600MF
class_num: 1000
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: Momentum
momentum: 0.9
lr:
name: Piecewise
learning_rate: 0.1
decay_epochs: [30, 60, 90]
values: [0.1, 0.01, 0.001, 0.0001]
regularizer:
name: 'L2'
coeff: 0.0001
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 224
- RandFlipImage:
flip_code: 1
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 64
drop_last: False
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Train:
- TopkAcc:
topk: [1, 5]
Eval:
- TopkAcc:
topk: [1, 5]

130
ppcls/configs/ImageNet/RegNet/RegNetX_6400MF.yaml Normal file
View File

@ -0,0 +1,130 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 120
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 256, 256]
save_inference_dir: ./inference
# model architecture
Arch:
name: RegNetX_6400MF
class_num: 1000
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: Momentum
momentum: 0.9
lr:
name: Piecewise
learning_rate: 0.1
decay_epochs: [30, 60, 90]
values: [0.1, 0.01, 0.001, 0.0001]
regularizer:
name: 'L2'
coeff: 0.0001
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 224
- RandFlipImage:
flip_code: 1
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 64
drop_last: False
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Train:
- TopkAcc:
topk: [1, 5]
Eval:
- TopkAcc:
topk: [1, 5]

130
ppcls/configs/ImageNet/RegNet/RegNetX_800MF.yaml Normal file
View File

@ -0,0 +1,130 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 120
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 256, 256]
save_inference_dir: ./inference
# model architecture
Arch:
name: RegNetX_800MF
class_num: 1000
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: Momentum
momentum: 0.9
lr:
name: Piecewise
learning_rate: 0.1
decay_epochs: [30, 60, 90]
values: [0.1, 0.01, 0.001, 0.0001]
regularizer:
name: 'L2'
coeff: 0.0001
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 224
- RandFlipImage:
flip_code: 1
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 64
drop_last: False
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Train:
- TopkAcc:
topk: [1, 5]
Eval:
- TopkAcc:
topk: [1, 5]

130
ppcls/configs/ImageNet/RegNet/RegNetX_8GF.yaml Normal file
View File

@ -0,0 +1,130 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 120
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 256, 256]
save_inference_dir: ./inference
# model architecture
Arch:
name: RegNetX_8GF
class_num: 1000
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: Momentum
momentum: 0.9
lr:
name: Piecewise
learning_rate: 0.1
decay_epochs: [30, 60, 90]
values: [0.1, 0.01, 0.001, 0.0001]
regularizer:
name: 'L2'
coeff: 0.0001
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 224
- RandFlipImage:
flip_code: 1
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 64
drop_last: False
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Train:
- TopkAcc:
topk: [1, 5]
Eval:
- TopkAcc:
topk: [1, 5]

137
ppcls/configs/ImageNet/RepVGG/RepVGG_B3.yaml Normal file
View File

@ -0,0 +1,137 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 200
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 224, 224]
save_inference_dir: ./inference
# model architecture
Arch:
name: RepVGG_B3
class_num: 1000
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
epsilon: 0.1
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: Momentum
momentum: 0.9
lr:
name: Cosine
learning_rate: 0.1
warmup_epoch: 5
regularizer:
name: 'L2'
coeff: 0.0001
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 224
- RandFlipImage:
flip_code: 1
- AutoAugment:
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
batch_transform_ops:
- MixupOperator:
alpha: 0.2
sampler:
name: DistributedBatchSampler
batch_size: 64
drop_last: False
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
interpolation: bicubic
backend: pil
- CropImage:
size: 224
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 64
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 256
interpolation: bicubic
backend: pil
- CropImage:
size: 224
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Train:
- TopkAcc:
topk: [1, 5]
Eval:
- TopkAcc:
topk: [1, 5]

8
ppcls/configs/ImageNet/RepVGG/RepVGG_B3g4.yaml
View File

@ -86,6 +86,8 @@ DataLoader:
channel_first: False
- ResizeImage:
resize_short: 256
interpolation: bicubic
backend: pil
- CropImage:
size: 224
- NormalizeImage:
@ -95,7 +97,7 @@ DataLoader:
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 64
batch_size: 128
drop_last: False
shuffle: False
loader:
@ -111,6 +113,8 @@ Infer:
channel_first: False
- ResizeImage:
resize_short: 256
interpolation: bicubic
backend: pil
- CropImage:
size: 224
- NormalizeImage:
@ -130,4 +134,4 @@ Metric:
topk: [1, 5]
Eval:
- TopkAcc:
topk: [1, 5]
topk: [1, 5]

137
ppcls/configs/ImageNet/RepVGG/RepVGG_D2se.yaml Normal file
View File

@ -0,0 +1,137 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 200
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 320, 320]
save_inference_dir: ./inference
# model architecture
Arch:
name: RepVGG_D2se
class_num: 1000
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
epsilon: 0.1
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: Momentum
momentum: 0.9
lr:
name: Cosine
learning_rate: 0.1
warmup_epoch: 5
regularizer:
name: 'L2'
coeff: 0.0001
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 224
- RandFlipImage:
flip_code: 1
- AutoAugment:
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
batch_transform_ops:
- MixupOperator:
alpha: 0.2
sampler:
name: DistributedBatchSampler
batch_size: 64
drop_last: False
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 320
interpolation: bicubic
backend: pil
- CropImage:
size: 320
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 320
interpolation: bicubic
backend: pil
- CropImage:
size: 320
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Train:
- TopkAcc:
topk: [1, 5]
Eval:
- TopkAcc:
topk: [1, 5]

6
ppcls/configs/ImageNet/ResNeSt/ResNeSt101.yaml
View File

@ -38,7 +38,7 @@ Optimizer:
learning_rate: 0.1
regularizer:
name: 'L2'
coeff: 0.00007
coeff: 0.0001
# data loader for train and eval
@ -95,7 +95,7 @@ DataLoader:
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 64
batch_size: 128
drop_last: False
shuffle: False
loader:
@ -128,4 +128,4 @@ Metric:
Train:
Eval:
- TopkAcc:
topk: [1, 5]
topk: [1, 5]

131
ppcls/configs/ImageNet/ResNeSt/ResNeSt200.yaml Normal file
View File

@ -0,0 +1,131 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 300
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 256, 256]
save_inference_dir: ./inference
# model architecture
Arch:
name: ResNeSt200
class_num: 1000
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
epsilon: 0.1
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: Momentum
momentum: 0.9
lr:
name: Cosine
learning_rate: 0.1
regularizer:
name: 'L2'
coeff: 0.0001
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 256
- RandFlipImage:
flip_code: 1
- AutoAugment:
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
batch_transform_ops:
- MixupOperator:
alpha: 0.2
sampler:
name: DistributedBatchSampler
batch_size: 64
drop_last: False
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 288
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 288
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Train:
Eval:
- TopkAcc:
topk: [1, 5]

131
ppcls/configs/ImageNet/ResNeSt/ResNeSt269.yaml Normal file
View File

@ -0,0 +1,131 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 300
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 256, 256]
save_inference_dir: ./inference
# model architecture
Arch:
name: ResNeSt269
class_num: 1000
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
epsilon: 0.1
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: Momentum
momentum: 0.9
lr:
name: Cosine
learning_rate: 0.1
regularizer:
name: 'L2'
coeff: 0.0001
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 256
- RandFlipImage:
flip_code: 1
- AutoAugment:
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
batch_transform_ops:
- MixupOperator:
alpha: 0.2
sampler:
name: DistributedBatchSampler
batch_size: 64
drop_last: False
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 288
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 288
- CropImage:
size: 256
- NormalizeImage:
scale: 1.0/255.0
mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Train:
Eval:
- TopkAcc:
topk: [1, 5]

134
ppcls/configs/ImageNet/TNT/TNT_base.yaml Normal file
View File

@ -0,0 +1,134 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 120
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 224, 224]
save_inference_dir: ./inference
# model architecture
Arch:
name: TNT_base
class_num: 1000
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: Momentum
momentum: 0.9
lr:
name: Piecewise
learning_rate: 0.1
decay_epochs: [30, 60, 90]
values: [0.1, 0.01, 0.001, 0.0001]
regularizer:
name: 'L2'
coeff: 0.0001
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 224
- RandFlipImage:
flip_code: 1
- NormalizeImage:
scale: 1.0/255.0
mean: [0.5, 0.5, 0.5]
std: [0.5, 0.5, 0.5]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 64
drop_last: False
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 248
interpolation: bicubic
backend: pil
- CropImage:
size: 224
- NormalizeImage:
scale: 1.0/255.0
mean: [0.5, 0.5, 0.5]
std: [0.5, 0.5, 0.5]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 248
interpolation: bicubic
backend: pil
- CropImage:
size: 224
- NormalizeImage:
scale: 1.0/255.0
mean: [0.5, 0.5, 0.5]
std: [0.5, 0.5, 0.5]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Train:
- TopkAcc:
topk: [1, 5]
Eval:
- TopkAcc:
topk: [1, 5]

8
ppcls/configs/ImageNet/TNT/TNT_small.yaml
View File

@ -83,6 +83,8 @@ DataLoader:
channel_first: False
- ResizeImage:
resize_short: 248
interpolation: bicubic
backend: pil
- CropImage:
size: 224
- NormalizeImage:
@ -92,7 +94,7 @@ DataLoader:
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 64
batch_size: 128
drop_last: False
shuffle: False
loader:
@ -108,6 +110,8 @@ Infer:
channel_first: False
- ResizeImage:
resize_short: 248
interpolation: bicubic
backend: pil
- CropImage:
size: 224
- NormalizeImage:
@ -127,4 +131,4 @@ Metric:
topk: [1, 5]
Eval:
- TopkAcc:
topk: [1, 5]
topk: [1, 5]

158
ppcls/configs/ImageNet/VAN/VAN_B1.yaml Normal file
View File

@ -0,0 +1,158 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 300
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 224, 224]
save_inference_dir: ./inference
# training model under @to_static
to_static: False
# model architecture
Arch:
name: VAN_B1
class_num: 1000
drop_path_rate: 0.1
drop_rate: 0.0
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
epsilon: 0.1
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: AdamW
beta1: 0.9
beta2: 0.999
epsilon: 1e-8
weight_decay: 0.05
one_dim_param_no_weight_decay: True
lr:
name: Cosine
learning_rate: 1e-3
eta_min: 1e-6
warmup_epoch: 5
warmup_start_lr: 1e-6
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 224
interpolation: random
backend: pil
- RandFlipImage:
flip_code: 1
- TimmAutoAugment:
config_str: rand-m9-mstd0.5-inc1
interpolation: random
img_size: 224
mean: [0.5, 0.5, 0.5]
- NormalizeImage:
scale: 1.0/255.0
mean: [0.5, 0.5, 0.5]
std: [0.5, 0.5, 0.5]
order: ''
- RandomErasing:
EPSILON: 0.25
sl: 0.02
sh: 1.0/3.0
r1: 0.3
attempt: 10
use_log_aspect: True
mode: pixel
batch_transform_ops:
- OpSampler:
MixupOperator:
alpha: 0.8
prob: 0.5
CutmixOperator:
alpha: 1.0
prob: 0.5
sampler:
name: DistributedBatchSampler
batch_size: 256
drop_last: True
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 248
interpolation: bicubic
backend: pil
- CropImage:
size: 224
- NormalizeImage:
scale: 1.0/255.0
mean: [0.5, 0.5, 0.5]
std: [0.5, 0.5, 0.5]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 248
interpolation: bicubic
backend: pil
- CropImage:
size: 224
- NormalizeImage:
scale: 1.0/255.0
mean: [0.5, 0.5, 0.5]
std: [0.5, 0.5, 0.5]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Eval:
- TopkAcc:
topk: [1, 5]

158
ppcls/configs/ImageNet/VAN/VAN_B2.yaml Normal file
View File

@ -0,0 +1,158 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 300
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 224, 224]
save_inference_dir: ./inference
# training model under @to_static
to_static: False
# model architecture
Arch:
name: VAN_B2
class_num: 1000
drop_path_rate: 0.1
drop_rate: 0.0
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
epsilon: 0.1
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: AdamW
beta1: 0.9
beta2: 0.999
epsilon: 1e-8
weight_decay: 0.05
one_dim_param_no_weight_decay: True
lr:
name: Cosine
learning_rate: 1e-3
eta_min: 1e-6
warmup_epoch: 5
warmup_start_lr: 1e-6
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 224
interpolation: random
backend: pil
- RandFlipImage:
flip_code: 1
- TimmAutoAugment:
config_str: rand-m9-mstd0.5-inc1
interpolation: random
img_size: 224
mean: [0.5, 0.5, 0.5]
- NormalizeImage:
scale: 1.0/255.0
mean: [0.5, 0.5, 0.5]
std: [0.5, 0.5, 0.5]
order: ''
- RandomErasing:
EPSILON: 0.25
sl: 0.02
sh: 1.0/3.0
r1: 0.3
attempt: 10
use_log_aspect: True
mode: pixel
batch_transform_ops:
- OpSampler:
MixupOperator:
alpha: 0.8
prob: 0.5
CutmixOperator:
alpha: 1.0
prob: 0.5
sampler:
name: DistributedBatchSampler
batch_size: 256
drop_last: True
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 248
interpolation: bicubic
backend: pil
- CropImage:
size: 224
- NormalizeImage:
scale: 1.0/255.0
mean: [0.5, 0.5, 0.5]
std: [0.5, 0.5, 0.5]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 248
interpolation: bicubic
backend: pil
- CropImage:
size: 224
- NormalizeImage:
scale: 1.0/255.0
mean: [0.5, 0.5, 0.5]
std: [0.5, 0.5, 0.5]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Eval:
- TopkAcc:
topk: [1, 5]

158
ppcls/configs/ImageNet/VAN/VAN_B3.yaml Normal file
View File

@ -0,0 +1,158 @@
# global configs
Global:
checkpoints: null
pretrained_model: null
output_dir: ./output/
device: gpu
save_interval: 1
eval_during_train: True
eval_interval: 1
epochs: 300
print_batch_step: 10
use_visualdl: False
# used for static mode and model export
image_shape: [3, 224, 224]
save_inference_dir: ./inference
# training model under @to_static
to_static: False
# model architecture
Arch:
name: VAN_B3
class_num: 1000
drop_path_rate: 0.1
drop_rate: 0.0
# loss function config for traing/eval process
Loss:
Train:
- CELoss:
weight: 1.0
epsilon: 0.1
Eval:
- CELoss:
weight: 1.0
Optimizer:
name: AdamW
beta1: 0.9
beta2: 0.999
epsilon: 1e-8
weight_decay: 0.05
one_dim_param_no_weight_decay: True
lr:
name: Cosine
learning_rate: 1e-3
eta_min: 1e-6
warmup_epoch: 5
warmup_start_lr: 1e-6
# data loader for train and eval
DataLoader:
Train:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- RandCropImage:
size: 224
interpolation: random
backend: pil
- RandFlipImage:
flip_code: 1
- TimmAutoAugment:
config_str: rand-m9-mstd0.5-inc1
interpolation: random
img_size: 224
mean: [0.5, 0.5, 0.5]
- NormalizeImage:
scale: 1.0/255.0
mean: [0.5, 0.5, 0.5]
std: [0.5, 0.5, 0.5]
order: ''
- RandomErasing:
EPSILON: 0.25
sl: 0.02
sh: 1.0/3.0
r1: 0.3
attempt: 10
use_log_aspect: True
mode: pixel
batch_transform_ops:
- OpSampler:
MixupOperator:
alpha: 0.8
prob: 0.5
CutmixOperator:
alpha: 1.0
prob: 0.5
sampler:
name: DistributedBatchSampler
batch_size: 256
drop_last: True
shuffle: True
loader:
num_workers: 4
use_shared_memory: True
Eval:
dataset:
name: ImageNetDataset
image_root: ./dataset/ILSVRC2012/
cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 248
interpolation: bicubic
backend: pil
- CropImage:
size: 224
- NormalizeImage:
scale: 1.0/255.0
mean: [0.5, 0.5, 0.5]
std: [0.5, 0.5, 0.5]
order: ''
sampler:
name: DistributedBatchSampler
batch_size: 128
drop_last: False
shuffle: False
loader:
num_workers: 4
use_shared_memory: True
Infer:
infer_imgs: docs/images/inference_deployment/whl_demo.jpg
batch_size: 10
transforms:
- DecodeImage:
to_rgb: True
channel_first: False
- ResizeImage:
resize_short: 248
interpolation: bicubic
backend: pil
- CropImage:
size: 224
- NormalizeImage:
scale: 1.0/255.0
mean: [0.5, 0.5, 0.5]
std: [0.5, 0.5, 0.5]
order: ''
- ToCHWImage:
PostProcess:
name: Topk
topk: 5
class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric:
Eval:
- TopkAcc:
topk: [1, 5]