Skip to content

P
PaddleClas

PaddlePaddle / PaddleClas
大约 2 年 前同步成功

通知 118
Star 4999
Fork 1114
P
PaddleClas
未验证 提交 b9786424 编写于 作者: C cuicheng01 提交者： GitHub
浏览文件
操作

Merge pull request #819 from cuicheng01/develop_reg 

Update configs
上级 4afc61a3 d187ce20
隐藏空白更改
内联 并排
Showing with 1935 addition and 1933 deletion
ppcls/configs/ImageNet/HRNet/HRNet_W18_C.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "HRNet_W18_C" name: HRNet_W18_C
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -46,80 +46,80 @@ Optimizer: ...@@ -46,80 +46,80 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/HRNet/HRNet_W30_C.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "HRNet_W30_C" name: HRNet_W30_C
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -46,80 +46,80 @@ Optimizer: ...@@ -46,80 +46,80 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/HRNet/HRNet_W32_C.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "HRNet_W32_C" name: HRNet_W32_C
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -46,80 +46,80 @@ Optimizer: ...@@ -46,80 +46,80 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/HRNet/HRNet_W40_C.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "HRNet_W40_C" name: HRNet_W40_C
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -46,80 +46,80 @@ Optimizer: ...@@ -46,80 +46,80 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/HRNet/HRNet_W44_C.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "HRNet_W44_C" name: HRNet_W44_C
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -46,78 +46,80 @@ Optimizer: ...@@ -46,78 +46,80 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
size: 224 resize_short: 256
- NormalizeImage: - CropImage:
scale: 0.00392157 size: 224
mean: [0.485, 0.456, 0.406] - NormalizeImage:
std: [0.229, 0.224, 0.225] scale: 1.0/255.0
order: '' mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225]
order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/HRNet/HRNet_W48_C.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "HRNet_W48_C" name: HRNet_W48_C
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -46,80 +46,80 @@ Optimizer: ...@@ -46,80 +46,80 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/HRNet/HRNet_W64_C.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "HRNet_W64_C" name: HRNet_W64_C
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -46,80 +46,80 @@ Optimizer: ...@@ -46,80 +46,80 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/Ineption/InceptionV3.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -12,18 +12,19 @@ Global: ...@@ -12,18 +12,19 @@ Global:
print_batch_step: 10 print_batch_step: 10
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 299, 299]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "InceptionV3" name: InceptionV3
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
Train: Train:
- CELoss: - CELoss:
weight: 1.0 weight: 1.0
epsilon: 0.1
Eval: Eval:
- CELoss: - CELoss:
weight: 1.0 weight: 1.0
...@@ -35,8 +36,6 @@ Optimizer: ...@@ -35,8 +36,6 @@ Optimizer:
lr: lr:
name: Cosine name: Cosine
learning_rate: 0.045 learning_rate: 0.045
decay_epochs: [30, 60, 90]
values: [0.1, 0.01, 0.001, 0.0001]
regularizer: regularizer:
name: 'L2' name: 'L2'
coeff: 0.0001 coeff: 0.0001
...@@ -46,80 +45,80 @@ Optimizer: ...@@ -46,80 +45,80 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 299 size: 299
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 320 resize_short: 320
- CropImage: - CropImage:
size: 299 size: 299
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 320 resize_short: 320
- CropImage: - CropImage:
size: 299 size: 299
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/MobileNetV1/MobileNetV1.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "MobileNetV1" name: MobileNetV1
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -39,87 +39,87 @@ Optimizer: ...@@ -39,87 +39,87 @@ Optimizer:
values: [0.1, 0.01, 0.001, 0.0001] values: [0.1, 0.01, 0.001, 0.0001]
regularizer: regularizer:
name: 'L2' name: 'L2'
coeff: 0.00003 coeff: 0.0003
# data loader for train and eval # data loader for train and eval
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/MobileNetV1/MobileNetV1_x0_25.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "MobileNetV1_x0_25" name: MobileNetV1_x0_25
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -39,87 +39,87 @@ Optimizer: ...@@ -39,87 +39,87 @@ Optimizer:
values: [0.1, 0.01, 0.001, 0.0001] values: [0.1, 0.01, 0.001, 0.0001]
regularizer: regularizer:
name: 'L2' name: 'L2'
coeff: 0.00003 coeff: 0.0003
# data loader for train and eval # data loader for train and eval
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/MobileNetV1/MobileNetV1_x0_5.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "MobileNetV1_x0_5" name: MobileNetV1_x0_5
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -39,87 +39,87 @@ Optimizer: ...@@ -39,87 +39,87 @@ Optimizer:
values: [0.1, 0.01, 0.001, 0.0001] values: [0.1, 0.01, 0.001, 0.0001]
regularizer: regularizer:
name: 'L2' name: 'L2'
coeff: 0.00003 coeff: 0.0003
# data loader for train and eval # data loader for train and eval
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/MobileNetV1/MobileNetV1_x0_75.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "MobileNetV1_x0_75" name: MobileNetV1_x0_75
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -39,87 +39,87 @@ Optimizer: ...@@ -39,87 +39,87 @@ Optimizer:
values: [0.1, 0.01, 0.001, 0.0001] values: [0.1, 0.01, 0.001, 0.0001]
regularizer: regularizer:
name: 'L2' name: 'L2'
coeff: 0.00003 coeff: 0.0003
# data loader for train and eval # data loader for train and eval
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/MobileNetV3/MobileNetV3_large_x0_35.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "MobileNetV3_large_x0_35" name: MobileNetV3_large_x0_35
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -45,80 +45,80 @@ Optimizer: ...@@ -45,80 +45,80 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 512 batch_size: 512
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/MobileNetV3/MobileNetV3_large_x0_5.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "MobileNetV3_large_x0_5" name: MobileNetV3_large_x0_5
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -45,80 +45,80 @@ Optimizer: ...@@ -45,80 +45,80 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 512 batch_size: 512
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/MobileNetV3/MobileNetV3_large_x0_75.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "MobileNetV3_large_x0_75" name: MobileNetV3_large_x0_75
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -45,80 +45,80 @@ Optimizer: ...@@ -45,80 +45,80 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 512 batch_size: 512
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/MobileNetV3/MobileNetV3_large_x1_0.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "MobileNetV3_large_x1_0" name: MobileNetV3_large_x1_0
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -45,81 +45,81 @@ Optimizer: ...@@ -45,81 +45,81 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- AutoAugment: - AutoAugment:
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 512 batch_size: 512
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/MobileNetV3/MobileNetV3_large_x1_25.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "MobileNetV3_large_x1_25" name: MobileNetV3_large_x1_25
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -38,87 +38,87 @@ Optimizer: ...@@ -38,87 +38,87 @@ Optimizer:
learning_rate: 1.3 learning_rate: 1.3
regularizer: regularizer:
name: 'L2' name: 'L2'
coeff: 0.00002 coeff: 0.00004
# data loader for train and eval # data loader for train and eval
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 512 batch_size: 512
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/MobileNetV3/MobileNetV3_small_x0_35.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "MobileNetV3_small_x0_35" name: MobileNetV3_small_x0_35
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -38,87 +38,87 @@ Optimizer: ...@@ -38,87 +38,87 @@ Optimizer:
learning_rate: 1.3 learning_rate: 1.3
regularizer: regularizer:
name: 'L2' name: 'L2'
coeff: 0.00002 coeff: 0.00001
# data loader for train and eval # data loader for train and eval
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 512 batch_size: 512
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/MobileNetV3/MobileNetV3_small_x0_5.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "MobileNetV3_small_x0_5" name: MobileNetV3_small_x0_5
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -38,87 +38,87 @@ Optimizer: ...@@ -38,87 +38,87 @@ Optimizer:
learning_rate: 1.3 learning_rate: 1.3
regularizer: regularizer:
name: 'L2' name: 'L2'
coeff: 0.00002 coeff: 0.00001
# data loader for train and eval # data loader for train and eval
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 512 batch_size: 512
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/MobileNetV3/MobileNetV3_small_x0_75.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "MobileNetV3_small_x0_75" name: MobileNetV3_small_x0_75
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -45,80 +45,80 @@ Optimizer: ...@@ -45,80 +45,80 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 512 batch_size: 512
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/MobileNetV3/MobileNetV3_small_x1_0.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "MobileNetV3_small_x1_0" name: MobileNetV3_small_x1_0
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -45,80 +45,80 @@ Optimizer: ...@@ -45,80 +45,80 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 512 batch_size: 512
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/MobileNetV3/MobileNetV3_small_x1_25.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "MobileNetV3_small_x1_25" name: MobileNetV3_small_x1_25
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -45,80 +45,80 @@ Optimizer: ...@@ -45,80 +45,80 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 512 batch_size: 512
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/ResNet/ResNet101.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "ResNet101" name: ResNet101
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -30,10 +30,10 @@ Loss: ...@@ -30,10 +30,10 @@ Loss:
Optimizer: Optimizer:
name: "Momentum" name: Momentum
momentum: 0.9 momentum: 0.9
lr: lr:
name: "Piecewise" name: Piecewise
learning_rate: 0.1 learning_rate: 0.1
decay_epochs: [30, 60, 90] decay_epochs: [30, 60, 90]
values: [0.1, 0.01, 0.001, 0.0001] values: [0.1, 0.01, 0.001, 0.0001]
...@@ -46,9 +46,9 @@ Optimizer: ...@@ -46,9 +46,9 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
...@@ -61,20 +61,20 @@ DataLoader: ...@@ -61,20 +61,20 @@ DataLoader:
order: '' order: ''
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
...@@ -86,16 +86,16 @@ DataLoader: ...@@ -86,16 +86,16 @@ DataLoader:
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
...@@ -112,9 +112,9 @@ Infer: ...@@ -112,9 +112,9 @@ Infer:
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: "Topk" name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
......
ppcls/configs/ImageNet/ResNet/ResNet101_vd.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "ResNet101_vd" name: ResNet101_vd
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -31,10 +31,10 @@ Loss: ...@@ -31,10 +31,10 @@ Loss:
Optimizer: Optimizer:
name: "Momentum" name: Momentum
momentum: 0.9 momentum: 0.9
lr: lr:
name: "Cosine" name: Cosine
learning_rate: 0.1 learning_rate: 0.1
regularizer: regularizer:
name: 'L2' name: 'L2'
...@@ -45,9 +45,9 @@ Optimizer: ...@@ -45,9 +45,9 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
...@@ -63,20 +63,20 @@ DataLoader: ...@@ -63,20 +63,20 @@ DataLoader:
alpha: 0.2 alpha: 0.2
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
...@@ -88,16 +88,16 @@ DataLoader: ...@@ -88,16 +88,16 @@ DataLoader:
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
...@@ -114,9 +114,9 @@ Infer: ...@@ -114,9 +114,9 @@ Infer:
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: "Topk" name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
......
ppcls/configs/ImageNet/ResNet/ResNet152.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "ResNet152" name: ResNet152
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -30,10 +30,10 @@ Loss: ...@@ -30,10 +30,10 @@ Loss:
Optimizer: Optimizer:
name: "Momentum" name: Momentum
momentum: 0.9 momentum: 0.9
lr: lr:
name: "Piecewise" name: Piecewise
learning_rate: 0.1 learning_rate: 0.1
decay_epochs: [30, 60, 90] decay_epochs: [30, 60, 90]
values: [0.1, 0.01, 0.001, 0.0001] values: [0.1, 0.01, 0.001, 0.0001]
...@@ -46,9 +46,9 @@ Optimizer: ...@@ -46,9 +46,9 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
...@@ -61,20 +61,20 @@ DataLoader: ...@@ -61,20 +61,20 @@ DataLoader:
order: '' order: ''
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
...@@ -86,16 +86,16 @@ DataLoader: ...@@ -86,16 +86,16 @@ DataLoader:
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
...@@ -112,9 +112,9 @@ Infer: ...@@ -112,9 +112,9 @@ Infer:
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: "Topk" name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
......
ppcls/configs/ImageNet/ResNet/ResNet152_vd.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "ResNet152_vd" name: ResNet152_vd
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -31,10 +31,10 @@ Loss: ...@@ -31,10 +31,10 @@ Loss:
Optimizer: Optimizer:
name: "Momentum" name: Momentum
momentum: 0.9 momentum: 0.9
lr: lr:
name: "Cosine" name: Cosine
learning_rate: 0.1 learning_rate: 0.1
regularizer: regularizer:
name: 'L2' name: 'L2'
...@@ -45,9 +45,9 @@ Optimizer: ...@@ -45,9 +45,9 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
...@@ -63,20 +63,20 @@ DataLoader: ...@@ -63,20 +63,20 @@ DataLoader:
alpha: 0.2 alpha: 0.2
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
...@@ -88,16 +88,16 @@ DataLoader: ...@@ -88,16 +88,16 @@ DataLoader:
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
...@@ -114,9 +114,9 @@ Infer: ...@@ -114,9 +114,9 @@ Infer:
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: "Topk" name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
......
ppcls/configs/ImageNet/ResNet/ResNet18.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "ResNet18" name: ResNet18
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -30,10 +30,10 @@ Loss: ...@@ -30,10 +30,10 @@ Loss:
Optimizer: Optimizer:
name: "Momentum" name: Momentum
momentum: 0.9 momentum: 0.9
lr: lr:
name: "Piecewise" name: Piecewise
learning_rate: 0.1 learning_rate: 0.1
decay_epochs: [30, 60, 90] decay_epochs: [30, 60, 90]
values: [0.1, 0.01, 0.001, 0.0001] values: [0.1, 0.01, 0.001, 0.0001]
...@@ -46,9 +46,9 @@ Optimizer: ...@@ -46,9 +46,9 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
...@@ -61,20 +61,20 @@ DataLoader: ...@@ -61,20 +61,20 @@ DataLoader:
order: '' order: ''
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
...@@ -86,16 +86,16 @@ DataLoader: ...@@ -86,16 +86,16 @@ DataLoader:
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
...@@ -112,9 +112,9 @@ Infer: ...@@ -112,9 +112,9 @@ Infer:
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: "Topk" name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
......
ppcls/configs/ImageNet/ResNet/ResNet18_vd.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "ResNet18_vd" name: ResNet18_vd
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -31,10 +31,10 @@ Loss: ...@@ -31,10 +31,10 @@ Loss:
Optimizer: Optimizer:
name: "Momentum" name: Momentum
momentum: 0.9 momentum: 0.9
lr: lr:
name: "Cosine" name: Cosine
learning_rate: 0.1 learning_rate: 0.1
regularizer: regularizer:
name: 'L2' name: 'L2'
...@@ -45,9 +45,9 @@ Optimizer: ...@@ -45,9 +45,9 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
...@@ -63,20 +63,20 @@ DataLoader: ...@@ -63,20 +63,20 @@ DataLoader:
alpha: 0.2 alpha: 0.2
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
...@@ -88,16 +88,16 @@ DataLoader: ...@@ -88,16 +88,16 @@ DataLoader:
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
...@@ -114,9 +114,9 @@ Infer: ...@@ -114,9 +114,9 @@ Infer:
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: "Topk" name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
......
ppcls/configs/ImageNet/ResNet/ResNet200_vd.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "ResNet200_vd" name: ResNet200_vd
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -31,10 +31,10 @@ Loss: ...@@ -31,10 +31,10 @@ Loss:
Optimizer: Optimizer:
name: "Momentum" name: Momentum
momentum: 0.9 momentum: 0.9
lr: lr:
name: "Cosine" name: Cosine
learning_rate: 0.1 learning_rate: 0.1
regularizer: regularizer:
name: 'L2' name: 'L2'
...@@ -45,9 +45,9 @@ Optimizer: ...@@ -45,9 +45,9 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
...@@ -63,20 +63,20 @@ DataLoader: ...@@ -63,20 +63,20 @@ DataLoader:
alpha: 0.2 alpha: 0.2
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
...@@ -88,16 +88,16 @@ DataLoader: ...@@ -88,16 +88,16 @@ DataLoader:
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
...@@ -114,9 +114,9 @@ Infer: ...@@ -114,9 +114,9 @@ Infer:
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: "Topk" name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
......
ppcls/configs/ImageNet/ResNet/ResNet34.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "ResNet34" name: ResNet34
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -30,10 +30,10 @@ Loss: ...@@ -30,10 +30,10 @@ Loss:
Optimizer: Optimizer:
name: "Momentum" name: Momentum
momentum: 0.9 momentum: 0.9
lr: lr:
name: "Piecewise" name: Piecewise
learning_rate: 0.1 learning_rate: 0.1
decay_epochs: [30, 60, 90] decay_epochs: [30, 60, 90]
values: [0.1, 0.01, 0.001, 0.0001] values: [0.1, 0.01, 0.001, 0.0001]
...@@ -46,9 +46,9 @@ Optimizer: ...@@ -46,9 +46,9 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
...@@ -61,20 +61,20 @@ DataLoader: ...@@ -61,20 +61,20 @@ DataLoader:
order: '' order: ''
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
...@@ -86,16 +86,16 @@ DataLoader: ...@@ -86,16 +86,16 @@ DataLoader:
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
...@@ -112,9 +112,9 @@ Infer: ...@@ -112,9 +112,9 @@ Infer:
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: "Topk" name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
......
ppcls/configs/ImageNet/ResNet/ResNet34_vd.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "ResNet34_vd" name: ResNet34_vd
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -31,10 +31,10 @@ Loss: ...@@ -31,10 +31,10 @@ Loss:
Optimizer: Optimizer:
name: "Momentum" name: Momentum
momentum: 0.9 momentum: 0.9
lr: lr:
name: "Cosine" name: Cosine
learning_rate: 0.1 learning_rate: 0.1
regularizer: regularizer:
name: 'L2' name: 'L2'
...@@ -45,9 +45,9 @@ Optimizer: ...@@ -45,9 +45,9 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
...@@ -63,20 +63,20 @@ DataLoader: ...@@ -63,20 +63,20 @@ DataLoader:
alpha: 0.2 alpha: 0.2
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
...@@ -88,16 +88,16 @@ DataLoader: ...@@ -88,16 +88,16 @@ DataLoader:
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
...@@ -114,9 +114,9 @@ Infer: ...@@ -114,9 +114,9 @@ Infer:
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: "Topk" name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
......
ppcls/configs/ImageNet/ResNet/ResNet50.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "ResNet50" name: ResNet50
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -30,10 +30,10 @@ Loss: ...@@ -30,10 +30,10 @@ Loss:
Optimizer: Optimizer:
name: "Momentum" name: Momentum
momentum: 0.9 momentum: 0.9
lr: lr:
name: "Piecewise" name: Piecewise
learning_rate: 0.1 learning_rate: 0.1
decay_epochs: [30, 60, 90] decay_epochs: [30, 60, 90]
values: [0.1, 0.01, 0.001, 0.0001] values: [0.1, 0.01, 0.001, 0.0001]
...@@ -46,9 +46,9 @@ Optimizer: ...@@ -46,9 +46,9 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
...@@ -61,20 +61,20 @@ DataLoader: ...@@ -61,20 +61,20 @@ DataLoader:
order: '' order: ''
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
...@@ -86,16 +86,16 @@ DataLoader: ...@@ -86,16 +86,16 @@ DataLoader:
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
...@@ -112,9 +112,9 @@ Infer: ...@@ -112,9 +112,9 @@ Infer:
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: "Topk" name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
......
ppcls/configs/ImageNet/ResNet/ResNet50_vd.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "ResNet50_vd" name: ResNet50_vd
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -31,10 +31,10 @@ Loss: ...@@ -31,10 +31,10 @@ Loss:
Optimizer: Optimizer:
name: "Momentum" name: Momentum
momentum: 0.9 momentum: 0.9
lr: lr:
name: "Cosine" name: Cosine
learning_rate: 0.1 learning_rate: 0.1
regularizer: regularizer:
name: 'L2' name: 'L2'
...@@ -45,9 +45,9 @@ Optimizer: ...@@ -45,9 +45,9 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
...@@ -63,20 +63,20 @@ DataLoader: ...@@ -63,20 +63,20 @@ DataLoader:
alpha: 0.2 alpha: 0.2
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
...@@ -88,16 +88,16 @@ DataLoader: ...@@ -88,16 +88,16 @@ DataLoader:
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: "DistributedBatchSampler" name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
...@@ -114,9 +114,9 @@ Infer: ...@@ -114,9 +114,9 @@ Infer:
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: "Topk" name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
......
ppcls/configs/ImageNet/VGG/VGG11.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "VGG11" name: VGG11
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -44,80 +44,80 @@ Optimizer: ...@@ -44,80 +44,80 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 128
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/VGG/VGG13.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "VGG13" name: VGG13
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -44,80 +44,80 @@ Optimizer: ...@@ -44,80 +44,80 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/VGG/VGG16.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "VGG16" name: VGG16
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -44,80 +44,80 @@ Optimizer: ...@@ -44,80 +44,80 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/ImageNet/VGG/VGG19.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 1000 class_num: 1000
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,11 +13,11 @@ Global: ...@@ -13,11 +13,11 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
# model architecture # model architecture
Arch: Arch:
name: "VGG19" name: VGG19
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -44,80 +44,80 @@ Optimizer: ...@@ -44,80 +44,80 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/train_list.txt" cls_label_path: ./dataset/ILSVRC2012/train_list.txt
transform_ops: transform_ops:
- RandCropImage: - RandCropImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: ImageNetDataset name: ImageNetDataset
image_root: "./dataset/ILSVRC2012/" image_root: ./dataset/ILSVRC2012/
cls_label_path: "./dataset/ILSVRC2012/val_list.txt" cls_label_path: ./dataset/ILSVRC2012/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
PostProcess: PostProcess:
name: Topk name: Topk
topk: 5 topk: 5
class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt" class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
Eval: Eval:
- TopkAcc: - TopkAcc:
topk: [1, 5] topk: [1, 5]
ppcls/configs/Products/ResNet50_vd_Aliproduct.yaml
浏览文件 @ b9786424
...@@ -2,8 +2,8 @@ ...@@ -2,8 +2,8 @@
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null pretrained_model: null
output_dir: "./output/" output_dir: ./output/
device: "gpu" device: gpu
class_num: 50030 class_num: 50030
save_interval: 1 save_interval: 1
eval_during_train: True eval_during_train: True
...@@ -13,23 +13,23 @@ Global: ...@@ -13,23 +13,23 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
eval_mode: "classification" eval_mode: classification
# model architecture # model architecture
Arch: Arch:
name: "RecModel" name: RecModel
Backbone: Backbone:
name: "ResNet50_vd" name: ResNet50_vd
pretrained: False pretrained: True
BackboneStopLayer: BackboneStopLayer:
name: "flatten_0" name: flatten_0
Neck: Neck:
name: "FC" name: FC
embedding_size: 2048 embedding_size: 2048
class_num: 512 class_num: 512
Head: Head:
name: "FC" name: FC
embedding_size: 512 embedding_size: 512
class_num: 50030 class_num: 50030
...@@ -56,52 +56,52 @@ Optimizer: ...@@ -56,52 +56,52 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/Aliproduct/" image_root: ./dataset/Aliproduct/
cls_label_path: "./dataset/Aliproduct/train_list.txt" cls_label_path: ./dataset/Aliproduct/train_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 0.00392157
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
# TOTO: modify to the latest trainer # TOTO: modify to the latest trainer
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/Aliproduct/" image_root: ./dataset/Aliproduct/
cls_label_path: "./dataset/Aliproduct/val_list.txt" cls_label_path: ./dataset/Aliproduct/val_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
resize_short: 256 resize_short: 256
- CropImage: - CropImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 0.00392157
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
sampler: sampler:
name: DistributedBatchSampler name: DistributedBatchSampler
batch_size: 64 batch_size: 64
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Metric: Metric:
Train: Train:
- TopkAcc: - TopkAcc:
...@@ -111,17 +111,17 @@ Metric: ...@@ -111,17 +111,17 @@ Metric:
topk: [1, 5] topk: [1, 5]
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 224 resize_short: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
ppcls/configs/Products/ResNet50_vd_Inshop.yaml
浏览文件 @ b9786424
# global configs # global configs
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null # please download pretrained model via this link:
output_dir: "./output/" # https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/rec/models/pretrain/product_ResNet50_vd_Aliproduct_v1.0_pretrained.pdparams
device: "gpu" pretrained_model: product_ResNet50_vd_Aliproduct_v1.0_pretrained
output_dir: ./output/
device: gpu
class_num: 3997 class_num: 3997
save_interval: 10 save_interval: 10
eval_during_train: True eval_during_train: True
...@@ -13,29 +15,30 @@ Global: ...@@ -13,29 +15,30 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
eval_mode: "retrieval" eval_mode: retrieval
# model architecture # model architecture
Arch: Arch:
name: "RecModel" name: RecModel
infer_output_key: features
infer_add_softmax: False
Backbone: Backbone:
name: "ResNet50_vd" name: ResNet50_vd
pretrained: False pretrained: False
BackboneStopLayer: BackboneStopLayer:
name: "flatten_0" name: flatten_0
Neck: Neck:
name: "FC" name: FC
embedding_size: 2048 embedding_size: 2048
class_num: 512 class_num: 512
Head: Head:
name: "ArcMargin" name: ArcMargin
embedding_size: 512 embedding_size: 512
class_num: 3997 class_num: 3997
margin: 0.15 margin: 0.15
scale: 30 scale: 30
infer_output_key: "features"
infer_add_softmax: False
# loss function config for traing/eval process # loss function config for traing/eval process
Loss: Loss:
...@@ -67,42 +70,41 @@ Optimizer: ...@@ -67,42 +70,41 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/Inshop/" image_root: ./dataset/Inshop/
cls_label_path: "./dataset/Inshop/train_list.txt" cls_label_path: ./dataset/Inshop/train_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 0.00392157
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- RandomErasing: - RandomErasing:
EPSILON: 0.5 EPSILON: 0.5
sl: 0.02 sl: 0.02
sh: 0.4 sh: 0.4
r1: 0.3 r1: 0.3
mean: [0., 0., 0.] mean: [0., 0., 0.]
sampler: sampler:
name: DistributedRandomIdentitySampler name: DistributedRandomIdentitySampler
batch_size: 64 batch_size: 64
num_instances: 2 num_instances: 2
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Eval: Eval:
Query: Query:
# TOTO: modify to the latest trainer
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/Inshop/" image_root: ./dataset/Inshop/
cls_label_path: "./dataset/Inshop/query_list.txt" cls_label_path: ./dataset/Inshop/query_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
size: 224 size: 224
...@@ -117,20 +119,19 @@ DataLoader: ...@@ -117,20 +119,19 @@ DataLoader:
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Gallery: Gallery:
# TOTO: modify to the latest trainer
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/Inshop/" image_root: ./dataset/Inshop/
cls_label_path: "./dataset/Inshop/gallery_list.txt" cls_label_path: ./dataset/Inshop/gallery_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
size: 224 size: 224
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
...@@ -140,7 +141,7 @@ DataLoader: ...@@ -140,7 +141,7 @@ DataLoader:
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Metric: Metric:
...@@ -149,17 +150,17 @@ Metric: ...@@ -149,17 +150,17 @@ Metric:
topk: [1, 5] topk: [1, 5]
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 224 resize_short: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
ppcls/configs/Products/ResNet50_vd_SOP.yaml
浏览文件 @ b9786424
# global configs # global configs
Global: Global:
checkpoints: null checkpoints: null
pretrained_model: null # please download pretrained model via this link:
output_dir: "./output/" # https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/rec/models/pretrain/product_ResNet50_vd_Aliproduct_v1.0_pretrained.pdparams
device: "gpu" pretrained_model: product_ResNet50_vd_Aliproduct_v1.0_pretrained
output_dir: ./output/
device: gpu
class_num: 11319 class_num: 11319
save_interval: 10 save_interval: 10
eval_during_train: True eval_during_train: True
...@@ -13,28 +15,28 @@ Global: ...@@ -13,28 +15,28 @@ Global:
use_visualdl: False use_visualdl: False
# used for static mode and model export # used for static mode and model export
image_shape: [3, 224, 224] image_shape: [3, 224, 224]
save_inference_dir: "./inference" save_inference_dir: ./inference
eval_mode: "retrieval" eval_mode: retrieval
# model architecture # model architecture
Arch: Arch:
name: "RecModel" name: RecModel
Backbone: Backbone:
name: "ResNet50_vd" name: ResNet50_vd
pretrained: False pretrained: False
BackboneStopLayer: BackboneStopLayer:
name: "flatten_0" name: flatten_0
Neck: Neck:
name: "FC" name: FC
embedding_size: 2048 embedding_size: 2048
class_num: 512 class_num: 512
Head: Head:
name: "ArcMargin" name: ArcMargin
embedding_size: 512 embedding_size: 512
class_num: 11319 class_num: 11319
margin: 0.15 margin: 0.15
scale: 30 scale: 30
infer_output_key: "features" infer_output_key: features
infer_add_softmax: False infer_add_softmax: False
# loss function config for traing/eval process # loss function config for traing/eval process
...@@ -67,42 +69,41 @@ Optimizer: ...@@ -67,42 +69,41 @@ Optimizer:
DataLoader: DataLoader:
Train: Train:
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/Stanford_Online_Products/" image_root: ./dataset/Stanford_Online_Products/
cls_label_path: "./dataset/Stanford_Online_Products/train_list.txt" cls_label_path: ./dataset/Stanford_Online_Products/train_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
size: 224 size: 224
- RandFlipImage: - RandFlipImage:
flip_code: 1 flip_code: 1
- NormalizeImage: - NormalizeImage:
scale: 0.00392157 scale: 0.00392157
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- RandomErasing: - RandomErasing:
EPSILON: 0.5 EPSILON: 0.5
sl: 0.02 sl: 0.02
sh: 0.4 sh: 0.4
r1: 0.3 r1: 0.3
mean: [0., 0., 0.] mean: [0., 0., 0.]
sampler: sampler:
name: DistributedRandomIdentitySampler name: DistributedRandomIdentitySampler
batch_size: 64 batch_size: 64
num_instances: 2 num_instances: 2
drop_last: False drop_last: False
shuffle: True shuffle: True
loader: loader:
num_workers: 6 num_workers: 6
use_shared_memory: True use_shared_memory: True
Eval: Eval:
Query: Query:
# TOTO: modify to the latest trainer
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/Stanford_Online_Products/" image_root: ./dataset/Stanford_Online_Products/
cls_label_path: "./dataset/Stanford_Online_Products/test_list.txt" cls_label_path: ./dataset/Stanford_Online_Products/test_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
size: 224 size: 224
...@@ -117,15 +118,14 @@ DataLoader: ...@@ -117,15 +118,14 @@ DataLoader:
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Gallery: Gallery:
# TOTO: modify to the latest trainer
dataset: dataset:
name: "ImageNetDataset" name: ImageNetDataset
image_root: "./dataset/Stanford_Online_Products/" image_root: ./dataset/Stanford_Online_Products/
cls_label_path: "./dataset/Stanford_Online_Products/test_list.txt" cls_label_path: ./dataset/Stanford_Online_Products/test_list.txt
transform_ops: transform_ops:
- ResizeImage: - ResizeImage:
size: 224 size: 224
...@@ -140,7 +140,7 @@ DataLoader: ...@@ -140,7 +140,7 @@ DataLoader:
drop_last: False drop_last: False
shuffle: False shuffle: False
loader: loader:
num_workers: 6 num_workers: 4
use_shared_memory: True use_shared_memory: True
Metric: Metric:
...@@ -149,17 +149,17 @@ Metric: ...@@ -149,17 +149,17 @@ Metric:
topk: [1, 5] topk: [1, 5]
Infer: Infer:
infer_imgs: "docs/images/whl/demo.jpg" infer_imgs: docs/images/whl/demo.jpg
batch_size: 10 batch_size: 10
transforms: transforms:
- DecodeImage: - DecodeImage:
to_rgb: True to_rgb: True
channel_first: False channel_first: False
- ResizeImage: - ResizeImage:
resize_short: 224 resize_short: 224
- NormalizeImage: - NormalizeImage:
scale: 1.0/255.0 scale: 1.0/255.0
mean: [0.485, 0.456, 0.406] mean: [0.485, 0.456, 0.406]
std: [0.229, 0.224, 0.225] std: [0.229, 0.224, 0.225]
order: '' order: ''
- ToCHWImage: - ToCHWImage:
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或