Merge pull request #1971 from cuicheng01/Add_PULC_demo

[WIP]Add pulc demo

ppcls/arch/backbone/legendary_models/pp_lcnet.py

@@ -189,6 +189,7 @@ class PPLCNet(TheseusLayer):
                  dropout_prob=0.2,
                  class_expand=1280,
                  lr_mult_list=[1.0, 1.0, 1.0, 1.0, 1.0, 1.0],
+                 stride_list=[2, 2, 2, 2, 2],
                  use_last_conv=True,
                  return_patterns=None,
                  return_stages=None,
@@ -198,6 +199,8 @@ class PPLCNet(TheseusLayer):
         self.class_expand = class_expand
         self.lr_mult_list = lr_mult_list
         self.use_last_conv = use_last_conv
+        self.stride_list = stride_list
+        self.net_config = NET_CONFIG
         if isinstance(self.lr_mult_list, str):
             self.lr_mult_list = eval(self.lr_mult_list)
 
@@ -206,17 +209,27 @@ class PPLCNet(TheseusLayer):
         )), "lr_mult_list should be in (list, tuple) but got {}".format(
             type(self.lr_mult_list))
         assert len(self.lr_mult_list
-                   ) == 6, "lr_mult_list length should be 5 but got {}".format(
+                   ) == 6, "lr_mult_list length should be 6 but got {}".format(
                        len(self.lr_mult_list))
 
+        assert isinstance(self.stride_list, (
+            list, tuple
+        )), "stride_list should be in (list, tuple) but got {}".format(
+            type(self.stride_list))
+        assert len(self.stride_list
+                   ) == 5, "stride_list length should be 5 but got {}".format(
+                       len(self.stride_list))
+
+        for i, stride in enumerate(stride_list[1:]):
+            self.net_config["blocks{}".format(i + 3)][0][3] = stride
         self.conv1 = ConvBNLayer(
             num_channels=3,
             filter_size=3,
             num_filters=make_divisible(16 * scale),
-            stride=2,
+            stride=stride_list[0],
             lr_mult=self.lr_mult_list[0])
 
-        self.blocks2 = nn.Sequential(* [
+        self.blocks2 = nn.Sequential(*[
             DepthwiseSeparable(
                 num_channels=make_divisible(in_c * scale),
                 num_filters=make_divisible(out_c * scale),
@@ -224,10 +237,11 @@ class PPLCNet(TheseusLayer):
                 stride=s,
                 use_se=se,
                 lr_mult=self.lr_mult_list[1])
-            for i, (k, in_c, out_c, s, se) in enumerate(NET_CONFIG["blocks2"])
+            for i, (k, in_c, out_c, s, se
+                    ) in enumerate(self.net_config["blocks2"])
         ])
 
-        self.blocks3 = nn.Sequential(* [
+        self.blocks3 = nn.Sequential(*[
             DepthwiseSeparable(
                 num_channels=make_divisible(in_c * scale),
                 num_filters=make_divisible(out_c * scale),
@@ -235,10 +249,11 @@ class PPLCNet(TheseusLayer):
                 stride=s,
                 use_se=se,
                 lr_mult=self.lr_mult_list[2])
-            for i, (k, in_c, out_c, s, se) in enumerate(NET_CONFIG["blocks3"])
+            for i, (k, in_c, out_c, s, se
+                    ) in enumerate(self.net_config["blocks3"])
         ])
 
-        self.blocks4 = nn.Sequential(* [
+        self.blocks4 = nn.Sequential(*[
             DepthwiseSeparable(
                 num_channels=make_divisible(in_c * scale),
                 num_filters=make_divisible(out_c * scale),
@@ -246,10 +261,11 @@ class PPLCNet(TheseusLayer):
                 stride=s,
                 use_se=se,
                 lr_mult=self.lr_mult_list[3])
-            for i, (k, in_c, out_c, s, se) in enumerate(NET_CONFIG["blocks4"])
+            for i, (k, in_c, out_c, s, se
+                    ) in enumerate(self.net_config["blocks4"])
         ])
 
-        self.blocks5 = nn.Sequential(* [
+        self.blocks5 = nn.Sequential(*[
             DepthwiseSeparable(
                 num_channels=make_divisible(in_c * scale),
                 num_filters=make_divisible(out_c * scale),
@@ -257,10 +273,11 @@ class PPLCNet(TheseusLayer):
                 stride=s,
                 use_se=se,
                 lr_mult=self.lr_mult_list[4])
-            for i, (k, in_c, out_c, s, se) in enumerate(NET_CONFIG["blocks5"])
+            for i, (k, in_c, out_c, s, se
+                    ) in enumerate(self.net_config["blocks5"])
         ])
 
-        self.blocks6 = nn.Sequential(* [
+        self.blocks6 = nn.Sequential(*[
             DepthwiseSeparable(
                 num_channels=make_divisible(in_c * scale),
                 num_filters=make_divisible(out_c * scale),
@@ -268,13 +285,14 @@ class PPLCNet(TheseusLayer):
                 stride=s,
                 use_se=se,
                 lr_mult=self.lr_mult_list[5])
-            for i, (k, in_c, out_c, s, se) in enumerate(NET_CONFIG["blocks6"])
+            for i, (k, in_c, out_c, s, se
+                    ) in enumerate(self.net_config["blocks6"])
         ])
 
         self.avg_pool = AdaptiveAvgPool2D(1)
         if self.use_last_conv:
             self.last_conv = Conv2D(
-                in_channels=make_divisible(NET_CONFIG["blocks6"][-1][2] *
+                in_channels=make_divisible(self.net_config["blocks6"][-1][2] *
                                            scale),
                 out_channels=self.class_expand,
                 kernel_size=1,
@@ -286,8 +304,9 @@ class PPLCNet(TheseusLayer):
         else:
             self.last_conv = None
         self.flatten = nn.Flatten(start_axis=1, stop_axis=-1)
-        self.fc = Linear(self.class_expand if self.use_last_conv else
-                         NET_CONFIG["blocks6"][-1][2], class_num)
+        self.fc = Linear(
+            self.class_expand if self.use_last_conv else
+            make_divisible(self.net_config["blocks6"][-1][2]), class_num)
 
         super().init_res(
             stages_pattern,

ppcls/arch/backbone/legendary_models/resnet.py

@@ -122,21 +122,18 @@ class ConvBNLayer(TheseusLayer):
         self.is_vd_mode = is_vd_mode
         self.act = act
         self.avg_pool = AvgPool2D(
-            kernel_size=2, stride=2, padding=0, ceil_mode=True)
+            kernel_size=2, stride=stride, padding="SAME", ceil_mode=True)
         self.conv = Conv2D(
             in_channels=num_channels,
             out_channels=num_filters,
             kernel_size=filter_size,
-            stride=stride,
+            stride=1 if is_vd_mode else stride,
             padding=(filter_size - 1) // 2,
             groups=groups,
             weight_attr=ParamAttr(learning_rate=lr_mult),
             bias_attr=False,
             data_format=data_format)
 
-        weight_attr = ParamAttr(learning_rate=lr_mult, trainable=True)
-        bias_attr = ParamAttr(learning_rate=lr_mult, trainable=True)
-
         self.bn = BatchNorm(
             num_filters,
             param_attr=ParamAttr(learning_rate=lr_mult),
@@ -164,7 +161,6 @@ class BottleneckBlock(TheseusLayer):
                  lr_mult=1.0,
                  data_format="NCHW"):
         super().__init__()
-
         self.conv0 = ConvBNLayer(
             num_channels=num_channels,
             num_filters=num_filters,
@@ -193,7 +189,7 @@ class BottleneckBlock(TheseusLayer):
                 num_channels=num_channels,
                 num_filters=num_filters * 4,
                 filter_size=1,
-                stride=stride if if_first else 1,
+                stride=stride,
                 is_vd_mode=False if if_first else True,
                 lr_mult=lr_mult,
                 data_format=data_format)
@@ -248,7 +244,7 @@ class BasicBlock(TheseusLayer):
                 num_channels=num_channels,
                 num_filters=num_filters,
                 filter_size=1,
-                stride=stride if if_first else 1,
+                stride=stride,
                 is_vd_mode=False if if_first else True,
                 lr_mult=lr_mult,
                 data_format=data_format)
@@ -287,6 +283,7 @@ class ResNet(TheseusLayer):
                  stem_act="relu",
                  class_num=1000,
                  lr_mult_list=[1.0, 1.0, 1.0, 1.0, 1.0],
+                 stride_list=[2, 2, 2, 2, 2],
                  data_format="NCHW",
                  input_image_channel=3,
                  return_patterns=None,
@@ -296,6 +293,7 @@ class ResNet(TheseusLayer):
 
         self.cfg = config
         self.lr_mult_list = lr_mult_list
+        self.stride_list = stride_list
         self.is_vd_mode = version == "vd"
         self.class_num = class_num
         self.num_filters = [64, 128, 256, 512]
@@ -312,11 +310,19 @@ class ResNet(TheseusLayer):
                    ) == 5, "lr_mult_list length should be 5 but got {}".format(
                        len(self.lr_mult_list))
 
+        assert isinstance(self.stride_list, (
+            list, tuple
+        )), "stride_list should be in (list, tuple) but got {}".format(
+            type(self.stride_list))
+        assert len(self.stride_list
+                   ) == 5, "stride_list length should be 5 but got {}".format(
+                       len(self.stride_list))
+
         self.stem_cfg = {
             #num_channels, num_filters, filter_size, stride
-            "vb": [[input_image_channel, 64, 7, 2]],
-            "vd":
-            [[input_image_channel, 32, 3, 2], [32, 32, 3, 1], [32, 64, 3, 1]]
+            "vb": [[input_image_channel, 64, 7, self.stride_list[0]]],
+            "vd": [[input_image_channel, 32, 3, self.stride_list[0]],
+                   [32, 32, 3, 1], [32, 64, 3, 1]]
         }
 
         self.stem = nn.Sequential(*[
@@ -332,7 +338,10 @@ class ResNet(TheseusLayer):
         ])
 
         self.max_pool = MaxPool2D(
-            kernel_size=3, stride=2, padding=1, data_format=data_format)
+            kernel_size=3,
+            stride=stride_list[1],
+            padding=1,
+            data_format=data_format)
         block_list = []
         for block_idx in range(len(self.block_depth)):
             shortcut = False
@@ -341,7 +350,8 @@ class ResNet(TheseusLayer):
                     num_channels=self.num_channels[block_idx] if i == 0 else
                     self.num_filters[block_idx] * self.channels_mult,
                     num_filters=self.num_filters[block_idx],
-                    stride=2 if i == 0 and block_idx != 0 else 1,
+                    stride=self.stride_list[block_idx + 1]
+                    if i == 0 and block_idx != 0 else 1,
                     shortcut=shortcut,
                     if_first=block_idx == i == 0 if version == "vd" else True,
                     lr_mult=self.lr_mult_list[block_idx + 1],

ppcls/configs/PULC/person_attribute/MobileNetV3_large_x1_0.yaml

+# global configs
+Global:
+  checkpoints: null
+  pretrained_model: null
+  output_dir: "./output/"
+  device: "gpu"
+  save_interval: 5
+  eval_during_train: True
+  eval_interval: 1
+  epochs: 20
+  print_batch_step: 20
+  use_visualdl: False
+  # used for static mode and model export
+  image_shape: [3, 256, 192]
+  save_inference_dir: "./inference"
+  use_multilabel: True
+
+# model architecture
+Arch:
+  name: "MobileNetV3_large_x1_0"
+  pretrained: True
+  class_num: 26
+
+# loss function config for traing/eval process
+Loss:
+  Train:
+    - MultiLabelLoss:
+        weight: 1.0
+        weight_ratio: True
+        size_sum: True
+  Eval:
+    - MultiLabelLoss:
+        weight: 1.0
+        weight_ratio: True
+        size_sum: True
+
+Optimizer:
+  name: Momentum
+  momentum: 0.9
+  lr:
+    name: Cosine
+    learning_rate: 0.01
+    warmup_epoch: 5
+  regularizer:
+    name: 'L2'
+    coeff: 0.0005
+  #clip_norm: 10
+
+# data loader for train and eval
+DataLoader:
+  Train:
+    dataset:
+      name: MultiLabelDataset
+      image_root: "dataset/pa100k/"
+      cls_label_path: "dataset/pa100k/train_val_list.txt"
+      label_ratio: True
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            size: [192, 256]
+        - Padv2:
+            size: [212, 276]
+            pad_mode: 1
+            fill_value: 0
+        - RandomCropImage:
+            size: [192, 256]
+        - RandFlipImage:
+            flip_code: 1
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 64
+      drop_last: True
+      shuffle: True
+    loader:
+      num_workers: 4
+      use_shared_memory: True
+  Eval:
+    dataset:
+      name: MultiLabelDataset
+      image_root: "dataset/pa100k/"
+      cls_label_path: "dataset/pa100k/test_list.txt"
+      label_ratio: True
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            size: [192, 256]
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 64
+      drop_last: False
+      shuffle: False
+    loader:
+      num_workers: 4
+      use_shared_memory: True
+
+
+Metric:
+  Eval:
+    - ATTRMetric:
+
+

ppcls/configs/PULC/person_attribute/PPLCNet_x1_0.yaml

+# global configs
+Global:
+  checkpoints: null
+  pretrained_model: null
+  output_dir: "./output/"
+  device: "gpu"
+  save_interval: 1
+  eval_during_train: True
+  eval_interval: 1
+  epochs: 20
+  print_batch_step: 10
+  use_visualdl: False
+  # used for static mode and model export
+  image_shape: [3, 256, 192]
+  save_inference_dir: "./inference"
+  use_multilabel: True
+
+# model architecture
+Arch:
+  name: "PPLCNet_x1_0"
+  pretrained: True
+  use_ssld: True
+  class_num: 26
+  
+
+# loss function config for traing/eval process
+Loss:
+  Train:
+    - MultiLabelLoss:
+        weight: 1.0
+        weight_ratio: True
+        size_sum: True
+  Eval:
+    - MultiLabelLoss:
+        weight: 1.0
+        weight_ratio: True
+        size_sum: True
+
+Optimizer:
+  name: Momentum
+  momentum: 0.9
+  lr:
+    name: Cosine
+    learning_rate: 0.01
+    warmup_epoch: 5
+  regularizer:
+    name: 'L2'
+    coeff: 0.0005
+
+# data loader for train and eval
+DataLoader:
+  Train:
+    dataset:
+      name: MultiLabelDataset
+      image_root: "dataset/pa100k/"
+      cls_label_path: "dataset/pa100k/train_val_list.txt"
+      label_ratio: True
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            size: [192, 256]
+        - TimmAutoAugment:
+            prob: 0.8
+            config_str: rand-m9-mstd0.5-inc1
+            interpolation: bicubic
+            img_size: [192, 256]
+        - Padv2:
+            size: [212, 276]
+            pad_mode: 1
+            fill_value: 0
+        - RandomCropImage:
+            size: [192, 256]
+        - RandFlipImage:
+            flip_code: 1
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+        - RandomErasing:
+            EPSILON: 0.4
+            sl: 0.02
+            sh: 1.0/3.0
+            r1: 0.3
+            attempt: 10
+            use_log_aspect: True
+            mode: pixel
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 64
+      drop_last: True
+      shuffle: True
+    loader:
+      num_workers: 4
+      use_shared_memory: True
+  Eval:
+    dataset:
+      name: MultiLabelDataset
+      image_root: "dataset/pa100k/"
+      cls_label_path: "dataset/pa100k/test_list.txt"
+      label_ratio: True
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            size: [192, 256]
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 64
+      drop_last: False
+      shuffle: False
+    loader:
+      num_workers: 4
+      use_shared_memory: True
+
+
+Metric:
+  Eval:
+    - ATTRMetric:
+
+

ppcls/configs/PULC/person_attribute/PPLCNet_x1_0_Distillation.yaml

+# global configs
+Global:
+  checkpoints: null
+  pretrained_model: null
+  output_dir: ./output
+  device: gpu
+  save_interval: 1
+  eval_during_train: True
+  start_eval_epoch: 1
+  eval_interval: 1
+  epochs: 20
+  print_batch_step: 10
+  use_visualdl: False
+  # used for static mode and model export
+  image_shape: [3, 256, 192]
+  save_inference_dir: ./inference
+  # training model under @to_static
+  to_static: False
+  use_dali: False
+  use_multilabel: True
+
+# model architecture
+Arch:
+  name: "DistillationModel"
+  class_num: &class_num 26
+  # if not null, its lengths should be same as models
+  pretrained_list:
+  # if not null, its lengths should be same as models
+  freeze_params_list:
+  - True
+  - False
+  use_sync_bn: True
+  models:
+    - Teacher:
+        name: ResNet101_vd
+        class_num: *class_num
+    - Student:
+        name: PPLCNet_x1_0
+        class_num: *class_num
+        pretrained: True
+        use_ssld: True
+
+  infer_model_name: "Student"
+ 
+# loss function config for traing/eval process
+Loss:
+  Train:
+    - DistillationDMLLoss:
+        weight: 1.0
+        model_name_pairs:
+        - ["Student", "Teacher"]
+    - DistillationMultiLabelLoss:
+        weight: 1.0
+        weight_ratio: True
+        model_names: ["Student"]
+        size_sum: True
+  Eval:
+    - MultiLabelLoss:
+        weight: 1.0
+        weight_ratio: True
+        size_sum: True
+
+Optimizer:
+  name: Momentum
+  momentum: 0.9
+  lr:
+    name: Cosine
+    learning_rate: 0.01
+    warmup_epoch: 5
+  regularizer:
+    name: 'L2'
+    coeff: 0.0005
+
+
+# data loader for train and eval
+# data loader for train and eval
+DataLoader:
+  Train:
+    dataset:
+      name: MultiLabelDataset
+      image_root: "dataset/pa100k/"
+      cls_label_path: "dataset/pa100k/train_val_list.txt"
+      label_ratio: True
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            size: [192, 256]
+        - TimmAutoAugment:
+            prob: 0.8
+            config_str: rand-m9-mstd0.5-inc1
+            interpolation: bicubic
+            img_size: [192, 256]
+        - Padv2:
+            size: [212, 276]
+            pad_mode: 1
+            fill_value: 0
+        - RandomCropImage:
+            size: [192, 256]
+        - RandFlipImage:
+            flip_code: 1
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+        - RandomErasing:
+            EPSILON: 0.4
+            sl: 0.02
+            sh: 1.0/3.0
+            r1: 0.3
+            attempt: 10
+            use_log_aspect: True
+            mode: pixel
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 64
+      drop_last: True
+      shuffle: True
+    loader:
+      num_workers: 4
+      use_shared_memory: True
+  Eval:
+    dataset:
+      name: MultiLabelDataset
+      image_root: "dataset/pa100k/"
+      cls_label_path: "dataset/pa100k/test_list.txt"
+      label_ratio: True
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            size: [192, 256]
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 64
+      drop_last: False
+      shuffle: False
+    loader:
+      num_workers: 4
+      use_shared_memory: True
+
+
+
+Metric:
+  Eval:
+    - ATTRMetric:

ppcls/configs/PULC/person_attribute/PPLCNet_x1_0_search.yaml

+# global configs
+Global:
+  checkpoints: null
+  pretrained_model: null
+  output_dir: "./output/"
+  device: "gpu"
+  save_interval: 1
+  eval_during_train: True
+  eval_interval: 1
+  epochs: 20
+  print_batch_step: 10
+  use_visualdl: False
+  # used for static mode and model export
+  image_shape: [3, 256, 192]
+  save_inference_dir: "./inference"
+  use_multilabel: True
+
+# model architecture
+Arch:
+  name: "PPLCNet_x1_0"
+  pretrained: True
+  use_ssld: True
+  class_num: 26
+  
+
+# loss function config for traing/eval process
+Loss:
+  Train:
+    - MultiLabelLoss:
+        weight: 1.0
+        weight_ratio: True
+        size_sum: True
+  Eval:
+    - MultiLabelLoss:
+        weight: 1.0
+        weight_ratio: True
+        size_sum: True
+
+Optimizer:
+  name: Momentum
+  momentum: 0.9
+  lr:
+    name: Cosine
+    learning_rate: 0.01
+    warmup_epoch: 5
+  regularizer:
+    name: 'L2'
+    coeff: 0.0005
+
+# data loader for train and eval
+DataLoader:
+  Train:
+    dataset:
+      name: MultiLabelDataset
+      image_root: "dataset/pa100k/"
+      cls_label_path: "dataset/pa100k/train_val_list.txt"
+      label_ratio: True
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            size: [192, 256]
+        - TimmAutoAugment:
+            prob: 0.0
+            config_str: rand-m9-mstd0.5-inc1
+            interpolation: bicubic
+            img_size: [192, 256]
+        - Padv2:
+            size: [212, 276]
+            pad_mode: 1
+            fill_value: 0
+        - RandomCropImage:
+            size: [192, 256]
+        - RandFlipImage:
+            flip_code: 1
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+        - RandomErasing:
+            EPSILON: 0.0
+            sl: 0.02
+            sh: 1.0/3.0
+            r1: 0.3
+            attempt: 10
+            use_log_aspect: True
+            mode: pixel
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 64
+      drop_last: True
+      shuffle: True
+    loader:
+      num_workers: 4
+      use_shared_memory: True
+  Eval:
+    dataset:
+      name: MultiLabelDataset
+      image_root: "dataset/pa100k"
+      cls_label_path: "dataset/pa100k/test_list.txt"
+      label_ratio: True
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            size: [192, 256]
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 64
+      drop_last: False
+      shuffle: False
+    loader:
+      num_workers: 4
+      use_shared_memory: True
+
+
+Metric:
+  Eval:
+    - ATTRMetric:
+
+

ppcls/configs/PULC/person_attribute/Res2Net200_vd_26w_4s.yaml

+# global configs
+Global:
+  checkpoints: null
+  pretrained_model: null
+  output_dir: "./output/"
+  device: "gpu"
+  save_interval: 5
+  eval_during_train: True
+  eval_interval: 1
+  epochs: 20
+  print_batch_step: 20
+  use_visualdl: False
+  # used for static mode and model export
+  image_shape: [3, 256, 192]
+  save_inference_dir: "./inference"
+  use_multilabel: True
+
+# model architecture
+Arch:
+  name: "Res2Net200_vd_26w_4s"
+  pretrained: True
+  class_num: 26
+
+# loss function config for traing/eval process
+Loss:
+  Train:
+    - MultiLabelLoss:
+        weight: 1.0
+        weight_ratio: True
+        size_sum: True
+  Eval:
+    - MultiLabelLoss:
+        weight: 1.0
+        weight_ratio: True
+        size_sum: True
+
+Optimizer:
+  name: Momentum
+  momentum: 0.9
+  lr:
+    name: Cosine
+    learning_rate: 0.01
+    warmup_epoch: 5
+  regularizer:
+    name: 'L2'
+    coeff: 0.0005
+  #clip_norm: 10
+
+# data loader for train and eval
+DataLoader:
+  Train:
+    dataset:
+      name: MultiLabelDataset
+      image_root: "dataset/pa100k/"
+      cls_label_path: "dataset/pa100k/train_val_list.txt"
+      label_ratio: True
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            size: [192, 256]
+        - Padv2:
+            size: [212, 276]
+            pad_mode: 1
+            fill_value: 0
+        - RandomCropImage:
+            size: [192, 256]
+        - RandFlipImage:
+            flip_code: 1
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 64
+      drop_last: True
+      shuffle: True
+    loader:
+      num_workers: 4
+      use_shared_memory: True
+  Eval:
+    dataset:
+      name: MultiLabelDataset
+      image_root: "dataset/pa100k/"
+      cls_label_path: "dataset/pa100k/test_list.txt"
+      label_ratio: True
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            size: [192, 256]
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 64
+      drop_last: False
+      shuffle: False
+    loader:
+      num_workers: 4
+      use_shared_memory: True
+
+
+Metric:
+  Eval:
+    - ATTRMetric:
+
+

ppcls/configs/PULC/person_attribute/search.yaml

+base_config_file: ppcls/configs/PULC/person_attribute/PPLCNet_x1_0_search.yaml
+distill_config_file: ppcls/configs/PULC/person_attribute/PPLCNet_x1_0_Distillation.yaml
+
+gpus: 0,1,2,3
+output_dir: output/search_attr
+search_times: 1
+search_dict:
+  - search_key: lrs
+    replace_config:
+      - Optimizer.lr.learning_rate
+    search_values: [0.0001, 0.005, 0.01, 0.02, 0.05]
+  - search_key: resolutions
+    replace_config:
+      - DataLoader.Train.dataset.transform_ops.1.ResizeImage.size
+      - DataLoader.Train.dataset.transform_ops.4.RandomCropImage.size
+      - DataLoader.Train.dataset.transform_ops.2.TimmAutoAugment.img_size
+    search_values: [[192, 256]]
+  - search_key: ra_probs
+    replace_config:
+      - DataLoader.Train.dataset.transform_ops.2.TimmAutoAugment.prob
+    search_values: [0.0, 0.2, 0.4, 0.6, 0.8, 1.0]
+  - search_key: re_probs
+    replace_config:
+      - DataLoader.Train.dataset.transform_ops.7.RandomErasing.EPSILON
+    search_values: [0.0, 0.2, 0.4, 0.6, 0.8, 1.0]
+  - search_key: lr_mult_list
+    replace_config:
+      - Arch.lr_mult_list
+    search_values:
+      - [0.0, 0.2, 0.4, 0.6, 0.8, 1.0]
+      - [0.0, 0.4, 0.4, 0.8, 0.8, 1.0]
+      - [1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
+teacher:
+  rm_keys:
+    - Arch.lr_mult_list
+  search_values:
+    - ResNet101_vd
+    - ResNet50_vd
+final_replace:
+  Arch.lr_mult_list: Arch.models.1.Student.lr_mult_list
+

ppcls/configs/PULC/text_direction/MobileNetV3_large_x1_0.yaml

+# global configs
+Global:
+  checkpoints: null
+  pretrained_model: null
+  output_dir: ./output/
+  device: gpu
+  save_interval: 1
+  eval_during_train: True
+  eval_interval: 1
+  start_eval_epoch: 18
+  epochs: 20
+  print_batch_step: 10
+  use_visualdl: False
+  # used for static mode and model export
+  image_shape: [3, 224, 224]
+  save_inference_dir: ./inference
+  # training model under @to_static
+  to_static: False
+  use_dali: False
+
+# model architecture
+Arch:
+  name: MobileNetV3_large_x1_0
+  class_num: 2
+  pretrained: True
+  use_sync_bn: True
+ 
+# loss function config for traing/eval process
+Loss:
+  Train:
+    - CELoss:
+        weight: 1.0
+        epsilon: 0.1
+  Eval:
+    - CELoss:
+        weight: 1.0
+
+
+Optimizer:
+  name: Momentum
+  momentum: 0.9
+  lr:
+    name: Cosine
+    learning_rate: 0.13
+    warmup_epoch: 5
+  regularizer:
+    name: 'L2'
+    coeff: 0.00002
+
+
+# data loader for train and eval
+DataLoader:
+  Train:
+    dataset:
+      name: ImageNetDataset
+      image_root: ./dataset/text_direction/
+      cls_label_path: ./dataset/text_direction/train_list.txt
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - RandCropImage:
+            size: 224
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 512
+      drop_last: False
+      shuffle: True
+    loader:
+      num_workers: 8
+      use_shared_memory: True
+
+  Eval:
+    dataset: 
+      name: ImageNetDataset
+      image_root: ./dataset/text_direction/
+      cls_label_path: ./dataset/text_direction/val_list.txt
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            resize_short: 256
+        - CropImage:
+            size: 224
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 64
+      drop_last: False
+      shuffle: False
+    loader:
+      num_workers: 4
+      use_shared_memory: True
+
+Infer:
+  infer_imgs: docs/images/inference_deployment/whl_demo.jpg
+  batch_size: 10
+  transforms:
+    - DecodeImage:
+        to_rgb: True
+        channel_first: False
+    - ResizeImage:
+        resize_short: 256
+    - CropImage:
+        size: 224
+    - NormalizeImage:
+        scale: 1.0/255.0
+        mean: [0.485, 0.456, 0.406]
+        std: [0.229, 0.224, 0.225]
+        order: ''
+    - ToCHWImage:
+  PostProcess:
+    name: Topk
+    topk: 5
+    class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
+
+Metric:
+  Train:
+    - TopkAcc:
+        topk: [1, 2]
+  Eval:
+    - TopkAcc:
+        topk: [1, 2]

ppcls/configs/PULC/text_direction/PPLCNet_x1_0.yaml

+# global configs
+Global:
+  checkpoints: null
+  pretrained_model: null
+  output_dir: ./output/
+  device: gpu
+  save_interval: 1
+  eval_during_train: True
+  start_eval_epoch: 18
+  eval_interval: 1
+  epochs: 20
+  print_batch_step: 10
+  use_visualdl: False
+  # used for static mode and model export
+  image_shape: [3, 80, 160]
+  save_inference_dir: ./inference
+  # training model under @to_static
+  to_static: False
+  use_dali: False
+
+# model architecture
+Arch:
+  name: PPLCNet_x1_0
+  class_num: 2
+  pretrained: True
+  use_ssld: True
+  stride_list: [2, [2, 1], [2, 1], [2, 1], [2, 1]]
+ 
+# loss function config for traing/eval process
+Loss:
+  Train:
+    - CELoss:
+        weight: 1.0
+  Eval:
+    - CELoss:
+        weight: 1.0
+
+
+Optimizer:
+  name: Momentum
+  momentum: 0.9
+  lr:
+    name: Cosine
+    learning_rate: 0.8
+    warmup_epoch: 5
+  regularizer:
+    name: 'L2'
+    coeff: 0.00004
+
+
+# data loader for train and eval
+DataLoader:
+  Train:
+    dataset:
+      name: ImageNetDataset
+      image_root: ./dataset/text_direction/
+      cls_label_path: ./dataset/text_direction/train_list.txt
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            size: [160, 80]
+        - TimmAutoAugment:
+            prob: 1.0
+            config_str: rand-m9-mstd0.5-inc1
+            interpolation: bicubic
+            img_size: [160, 80]
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+        - RandomErasing:
+            EPSILON: 0.0
+            sl: 0.02
+            sh: 1.0/3.0
+            r1: 0.3
+            attempt: 10
+            use_log_aspect: True
+            mode: pixel
+
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 256
+      drop_last: False
+      shuffle: True
+    loader:
+      num_workers: 16
+      use_shared_memory: True
+
+  Eval:
+    dataset: 
+      name: ImageNetDataset
+      image_root: ./dataset/text_direction/
+      cls_label_path: ./dataset/text_direction/val_list.txt
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            size: [160, 80]
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 128
+      drop_last: False
+      shuffle: False
+    loader:
+      num_workers: 8
+      use_shared_memory: True
+
+Infer:
+  infer_imgs: docs/images/inference_deployment/whl_demo.jpg
+  batch_size: 10
+  transforms:
+    - DecodeImage:
+        to_rgb: True
+        channel_first: False
+    - ResizeImage:
+        size: [160, 80]
+    - NormalizeImage:
+        scale: 1.0/255.0
+        mean: [0.485, 0.456, 0.406]
+        std: [0.229, 0.224, 0.225]
+        order: ''
+    - ToCHWImage:
+  PostProcess:
+    name: Topk
+    topk: 5
+    class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
+
+Metric:
+  Train:
+    - TopkAcc:
+        topk: [1, 2]
+  Eval:
+    - TopkAcc:
+        topk: [1, 2]

ppcls/configs/PULC/text_direction/PPLCNet_x1_0_distillation.yaml

+# global configs
+Global:
+  checkpoints: null
+  pretrained_model: null
+  output_dir: ./output/
+  device: gpu
+  save_interval: 1
+  eval_during_train: True
+  start_eval_epoch: 18
+  eval_interval: 1
+  epochs: 20
+  print_batch_step: 10
+  use_visualdl: False
+  # used for static mode and model export
+  image_shape: [3, 80, 160]
+  save_inference_dir: ./inference
+  # training model under @to_static
+  to_static: False
+  use_dali: False
+
+# model architecture
+Arch:
+  name: "DistillationModel"
+  class_num: &class_num 2
+  # if not null, its lengths should be same as models
+  pretrained_list:
+  # if not null, its lengths should be same as models
+  freeze_params_list:
+  - True
+  - False
+  use_sync_bn: True
+  models:
+    - Teacher:
+        name: ResNet101_vd
+        class_num: *class_num
+        stride_list: [2, [2, 1], [2, 1], [2, 1], [2, 1]]
+    - Student:
+        name: PPLCNet_x1_0
+        class_num: *class_num
+        stride_list: [2, [2, 1], [2, 1], [2, 1], [2, 1]]
+        pretrained: True
+        use_ssld: True
+
+  infer_model_name: "Student"
+ 
+# loss function config for traing/eval process
+Loss:
+  Train:
+    - DistillationDMLLoss:
+        weight: 1.0
+        model_name_pairs:
+        - ["Student", "Teacher"]
+  Eval:
+    - CELoss:
+        weight: 1.0
+
+
+Optimizer:
+  name: Momentum
+  momentum: 0.9
+  lr:
+    name: Cosine
+    learning_rate: 0.8
+    warmup_epoch: 5
+  regularizer:
+    name: 'L2'
+    coeff: 0.00004
+
+
+# data loader for train and eval
+DataLoader:
+  Train:
+    dataset:
+      name: ImageNetDataset
+      image_root: ./dataset/text_direction/
+      cls_label_path: ./dataset/text_direction/train_list.txt
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            size: [160, 80]
+        - TimmAutoAugment:
+            prob: 1.0
+            config_str: rand-m9-mstd0.5-inc1
+            interpolation: bicubic
+            img_size: [160, 80]
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+        - RandomErasing:
+            EPSILON: 0.0
+            sl: 0.02
+            sh: 1.0/3.0
+            r1: 0.3
+            attempt: 10
+            use_log_aspect: True
+            mode: pixel
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 256
+      drop_last: False
+      shuffle: True
+    loader:
+      num_workers: 16
+      use_shared_memory: True
+
+  Eval:
+    dataset: 
+      name: ImageNetDataset
+      image_root: ./dataset/text_direction/
+      cls_label_path: ./dataset/text_direction/val_list.txt
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            size: [160, 80]
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 128
+      drop_last: False
+      shuffle: False
+    loader:
+      num_workers: 8
+      use_shared_memory: True
+
+Infer:
+  infer_imgs: docs/images/inference_deployment/whl_demo.jpg
+  batch_size: 10
+  transforms:
+    - DecodeImage:
+        to_rgb: True
+        channel_first: False
+    - ResizeImage:
+        size: [160, 80]
+    - NormalizeImage:
+        scale: 1.0/255.0
+        mean: [0.485, 0.456, 0.406]
+        std: [0.229, 0.224, 0.225]
+        order: ''
+    - ToCHWImage:
+  PostProcess:
+    name: Topk
+    topk: 5
+    class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
+
+Metric:
+    Train:
+    - DistillationTopkAcc:
+        model_key: "Student"
+        topk: [1, 2]
+    Eval:
+    - TopkAcc:
+        topk: [1, 2]

ppcls/configs/PULC/text_direction/PPLCNet_x1_0_search.yaml

+# global configs
+Global:
+  checkpoints: null
+  pretrained_model: null
+  output_dir: ./output/
+  device: gpu
+  save_interval: 1
+  eval_during_train: True
+  start_eval_epoch: 18
+  eval_interval: 1
+  epochs: 20
+  print_batch_step: 10
+  use_visualdl: False
+  # used for static mode and model export
+  image_shape: [3, 48, 192]
+  save_inference_dir: ./inference
+  # training model under @to_static
+  to_static: False
+  use_dali: False
+
+# model architecture
+Arch:
+  name: PPLCNet_x1_0
+  class_num: 2
+  pretrained: True
+  use_ssld: True
+  stride_list: [2, [2, 1], [2, 1], [2, 1], [2, 1]]
+  
+ 
+# loss function config for traing/eval process
+Loss:
+  Train:
+    - CELoss:
+        weight: 1.0
+  Eval:
+    - CELoss:
+        weight: 1.0
+
+
+Optimizer:
+  name: Momentum
+  momentum: 0.9
+  lr:
+    name: Cosine
+    learning_rate: 0.5
+    warmup_epoch: 5
+  regularizer:
+    name: 'L2'
+    coeff: 0.00004
+
+
+# data loader for train and eval
+DataLoader:
+  Train:
+    dataset:
+      name: ImageNetDataset
+      image_root: ./dataset/text_direction/
+      cls_label_path: ./dataset/text_direction/train_list.txt
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            size: [192, 48]
+        - TimmAutoAugment:
+            prob: 0.0
+            config_str: rand-m9-mstd0.5-inc1
+            interpolation: bicubic
+            img_size: [192, 48]
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+        - RandomErasing:
+            EPSILON: 0.0
+            sl: 0.02
+            sh: 1.0/3.0
+            r1: 0.3
+            attempt: 10
+            use_log_aspect: True
+            mode: pixel
+
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 256
+      drop_last: False
+      shuffle: True
+    loader:
+      num_workers: 16
+      use_shared_memory: True
+
+  Eval:
+    dataset: 
+      name: ImageNetDataset
+      image_root: ./dataset/text_direction/
+      cls_label_path: ./dataset/text_direction/val_list.txt
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            size: [192, 48]
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 128
+      drop_last: False
+      shuffle: False
+    loader:
+      num_workers: 8
+      use_shared_memory: True
+
+Infer:
+  infer_imgs: docs/images/inference_deployment/whl_demo.jpg
+  batch_size: 10
+  transforms:
+    - DecodeImage:
+        to_rgb: True
+        channel_first: False
+    - ResizeImage:
+        size: [192, 48]
+    - NormalizeImage:
+        scale: 1.0/255.0
+        mean: [0.485, 0.456, 0.406]
+        std: [0.229, 0.224, 0.225]
+        order: ''
+    - ToCHWImage:
+  PostProcess:
+    name: Topk
+    topk: 5
+    class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
+
+Metric:
+  Train:
+    - TopkAcc:
+        topk: [1, 2]
+  Eval:
+    - TopkAcc:
+        topk: [1, 2]

ppcls/configs/PULC/text_direction/SwinTransformer_tiny_patch4_window7_224.yaml

+# global configs
+Global:
+  checkpoints: null
+  pretrained_model: null
+  output_dir: ./output/
+  device: gpu
+  save_interval: 1
+  eval_during_train: True
+  eval_interval: 1
+  start_eval_epoch: 10
+  epochs: 20
+  print_batch_step: 10
+  use_visualdl: False
+  # used for static mode and model export
+  image_shape: [3, 224, 224]
+  save_inference_dir: ./inference
+  # training model under @to_static
+  to_static: False
+  use_dali: False
+
+# mixed precision training
+AMP:
+  scale_loss: 128.0
+  use_dynamic_loss_scaling: True
+  # O1: mixed fp16
+  level: O1
+
+# model architecture
+Arch:
+  name: SwinTransformer_tiny_patch4_window7_224
+  class_num: 2
+  pretrained: True
+ 
+# loss function config for traing/eval process
+Loss:
+  Train:
+    - CELoss:
+        weight: 1.0
+        epsilon: 0.1
+  Eval:
+    - CELoss:
+        weight: 1.0
+
+Optimizer:
+  name: AdamW
+  beta1: 0.9
+  beta2: 0.999
+  epsilon: 1e-8
+  weight_decay: 0.05
+  no_weight_decay_name: absolute_pos_embed relative_position_bias_table .bias norm
+  one_dim_param_no_weight_decay: True
+  lr:
+    name: Cosine
+    learning_rate: 1e-4
+    eta_min: 2e-6
+    warmup_epoch: 5
+    warmup_start_lr: 2e-7
+
+
+# data loader for train and eval
+DataLoader:
+  Train:
+    dataset:
+      name: ImageNetDataset
+      image_root: ./dataset/text_direction/
+      cls_label_path: ./dataset/text_direction/train_list.txt
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - RandCropImage:
+            size: 224
+            interpolation: bicubic
+            backend: pil
+        - TimmAutoAugment:
+            config_str: rand-m9-mstd0.5-inc1
+            interpolation: bicubic
+            img_size: 224
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+        - RandomErasing:
+            EPSILON: 0.25
+            sl: 0.02
+            sh: 1.0/3.0
+            r1: 0.3
+            attempt: 10
+            use_log_aspect: True
+            mode: pixel
+      batch_transform_ops:
+        - OpSampler:
+            MixupOperator:
+              alpha: 0.8
+              prob: 0.5
+            CutmixOperator:
+              alpha: 1.0
+              prob: 0.5
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 128
+      drop_last: False
+      shuffle: True
+    loader:
+      num_workers: 8
+      use_shared_memory: True
+
+  Eval:
+    dataset: 
+      name: ImageNetDataset
+      image_root: ./dataset/text_direction/
+      cls_label_path: ./dataset/text_direction/val_list.txt
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            resize_short: 256
+        - CropImage:
+            size: 224
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 64
+      drop_last: False
+      shuffle: False
+    loader:
+      num_workers: 8
+      use_shared_memory: True
+
+Infer:
+  infer_imgs: docs/images/inference_deployment/whl_demo.jpg
+  batch_size: 10
+  transforms:
+    - DecodeImage:
+        to_rgb: True
+        channel_first: False
+    - ResizeImage:
+        resize_short: 256
+    - CropImage:
+        size: 224
+    - NormalizeImage:
+        scale: 1.0/255.0
+        mean: [0.485, 0.456, 0.406]
+        std: [0.229, 0.224, 0.225]
+        order: ''
+    - ToCHWImage:
+  PostProcess:
+    name: Topk
+    topk: 5
+    class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
+
+Metric:
+  Train:
+    - TopkAcc:
+        topk: [1, 2]
+  Eval:
+    - TopkAcc:
+        topk: [1, 2]

ppcls/configs/PULC/text_direction/search.yaml

+base_config_file: ppcls/configs/PULC/text_direction/PPLCNet_x1_0.yaml
+distill_config_file: ppcls/configs/PULC/text_direction/PPLCNet_x1_0_distillation.yaml
+
+gpus: 0,1,2,3
+output_dir: output/search_text
+search_times: 1
+search_dict:
+  - search_key: lrs
+    replace_config:
+      - Optimizer.lr.learning_rate
+    search_values: [0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8]
+  - search_key: resolutions
+    replace_config:
+      - DataLoader.Train.dataset.transform_ops.1.ResizeImage.size
+      - DataLoader.Train.dataset.transform_ops.2.TimmAutoAugment.img_size
+      - DataLoader.Eval.dataset.transform_ops.1.ResizeImage.size
+    search_values: [[192, 48], [180, 60], [160, 80]]
+  - search_key: ra_probs
+    replace_config:
+      - DataLoader.Train.dataset.transform_ops.2.TimmAutoAugment.prob
+    search_values: [0.0, 0.2, 0.4, 0.6, 0.8, 1.0]
+  - search_key: re_probs
+    replace_config:
+      - DataLoader.Train.dataset.transform_ops.4.RandomErasing.EPSILON
+    search_values: [0.0, 0.2, 0.4, 0.6, 0.8, 1.0]
+  - search_key: lr_mult_list
+    replace_config:
+      - Arch.lr_mult_list
+    search_values:
+      - [0.0, 0.2, 0.4, 0.6, 0.8, 1.0]
+      - [0.0, 0.4, 0.4, 0.8, 0.8, 1.0]
+      - [1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
+teacher:
+  rm_keys:
+    - Arch.lr_mult_list
+  search_values:
+    - ResNet101_vd
+    - ResNet50_vd
+final_replace:
+  Arch.lr_mult_list: Arch.models.1.Student.lr_mult_list
+