diff --git a/deploy/configs/build_inshop.yaml b/deploy/configs/build_inshop.yaml
index c736d067a965a04e0a064b4532845c37e2eeac26..c2638e979ec5cac2e2b40fddcdcf6782f8922a36 100644
--- a/deploy/configs/build_inshop.yaml
+++ b/deploy/configs/build_inshop.yaml
@@ -1,5 +1,5 @@
 Global:
-  rec_inference_model_dir: "./models/product_ResNet50_vd_Inshop_v1.0_infer/"
+  rec_inference_model_dir: "./models/product_ResNet50_vd_Inshop_v1.0_infer"
   batch_size: 1
   use_gpu: True
   enable_mkldnn: True
@@ -26,7 +26,7 @@ RecPostProcess: null
 
 # indexing engine config
 IndexProcess:
-  index_path: "./dataset/product_demo_data_v1.0/query/"
+  index_path: "./dataset/product_demo_data_v1.0/index"
   image_root: "./dataset/product_demo_data_v1.0"
   data_file:  "./dataset/product_demo_data_v1.0/data_file.txt"
   delimiter: " "
diff --git a/deploy/configs/inference_inshop.yaml b/deploy/configs/inference_inshop.yaml
index f9446efe1e71fd21a170dc1e5dc99e55af455c0f..1e7db144d1223f0c5b969d0d27f9597665f505ad 100644
--- a/deploy/configs/inference_inshop.yaml
+++ b/deploy/configs/inference_inshop.yaml
@@ -1,11 +1,11 @@
 Global:
-  infer_imgs: "./dataset/product_demo_data_v1.0/query/"
-  det_inference_model_dir: "./models/ppyolov2_r50vd_dcn_mainbody_v1.0_infer/"
-  rec_inference_model_dir: "./models/product_ResNet50_vd_Inshop_v1.0_infer/"
+  infer_imgs: "./dataset/product_demo_data_v1.0/query"
+  det_inference_model_dir: "./models/ppyolov2_r50vd_dcn_mainbody_v1.0_infer"
+  rec_inference_model_dir: "./models/product_ResNet50_vd_Inshop_v1.0_infer"
   batch_size: 1
   image_shape: [3, 640, 640]
   threshold: 0.0
-  max_det_results: 3
+  max_det_results: 1
   labe_list:
   - foreground
 
@@ -48,7 +48,7 @@ RecPostProcess: null
 
 # indexing engine config
 IndexProcess:
-  index_path: "./dataset/product_demo_data_v1.0/index/"
+  index_path: "./dataset/product_demo_data_v1.0/index"
   search_budget: 100
   return_k: 5
   dist_type: "IP"
diff --git a/ppcls/configs/ImageNet/ResNet/ResNet101.yaml b/ppcls/configs/ImageNet/ResNet/ResNet101.yaml
index f3ecafe6682128404a6a6ed6cc22bba44bde6606..8b9e8dd9ef1800ffcf70f3b01ea68144cc335d67 100644
--- a/ppcls/configs/ImageNet/ResNet/ResNet101.yaml
+++ b/ppcls/configs/ImageNet/ResNet/ResNet101.yaml
@@ -30,10 +30,10 @@ Loss:
 
 
 Optimizer:
-  name: Momentum
+  name: "Momentum"
   momentum: 0.9
   lr:
-    name: Piecewise
+    name: "Piecewise"
     learning_rate: 0.1
     decay_epochs: [30, 60, 90]
     values: [0.1, 0.01, 0.001, 0.0001]
@@ -46,80 +46,80 @@ Optimizer:
 DataLoader:
   Train:
     dataset:
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
-        transform_ops:
-          - RandCropImage:
-              size: 224
-          - RandFlipImage:
-              flip_code: 1
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
+      transform_ops:
+        - RandCropImage:
+            size: 224
+        - RandFlipImage:
+            flip_code: 1
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
 
     sampler:
-        name: DistributedBatchSampler
-        batch_size: 64
-        drop_last: False
-        shuffle: True
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: True
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
   Eval:
     # TOTO: modify to the latest trainer
     dataset: 
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
-        transform_ops:
-          - ResizeImage:
-              resize_short: 256
-          - CropImage:
-              size: 224
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
+      transform_ops:
+        - 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
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: False
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
 Infer:
   infer_imgs: "docs/images/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:
+    - 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
+    name: "Topk"
     topk: 5
     class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt"
 
 Metric:
-    Train:
+  Train:
     - TopkAcc:
         topk: [1, 5]
-    Eval:
+  Eval:
     - TopkAcc:
         topk: [1, 5]
diff --git a/ppcls/configs/ImageNet/ResNet/ResNet101_vd.yaml b/ppcls/configs/ImageNet/ResNet/ResNet101_vd.yaml
index 37bdb4080afa98e065d5584c0799646e12fcb9fb..a446b71ec31f4e42e84db7f8b86c9e0566492bfa 100644
--- a/ppcls/configs/ImageNet/ResNet/ResNet101_vd.yaml
+++ b/ppcls/configs/ImageNet/ResNet/ResNet101_vd.yaml
@@ -24,16 +24,17 @@ Loss:
   Train:
     - CELoss:
         weight: 1.0
+        epsilon: 0.1
   Eval:
     - CELoss:
         weight: 1.0
 
 
 Optimizer:
-  name: Momentum
+  name: "Momentum"
   momentum: 0.9
   lr:
-    name: Cosine
+    name: "Cosine"
     learning_rate: 0.1
   regularizer:
     name: 'L2'
@@ -44,80 +45,81 @@ Optimizer:
 DataLoader:
   Train:
     dataset:
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
-        transform_ops:
-          - RandCropImage:
-              size: 224
-          - RandFlipImage:
-              flip_code: 1
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
+      transform_ops:
+        - RandCropImage:
+            size: 224
+        - RandFlipImage:
+            flip_code: 1
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+      batch_transform_ops:
+        - MixupOperator:
+            alpha: 0.2
 
     sampler:
-        name: DistributedBatchSampler
-        batch_size: 64
-        drop_last: False
-        shuffle: True
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: True
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
   Eval:
     # TOTO: modify to the latest trainer
     dataset: 
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
-        transform_ops:
-          - ResizeImage:
-              resize_short: 256
-          - CropImage:
-              size: 224
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
+      transform_ops:
+        - 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
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: False
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
 Infer:
   infer_imgs: "docs/images/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:
+    - 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
+    name: "Topk"
     topk: 5
     class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt"
 
 Metric:
-    Train:
-    - TopkAcc:
-        topk: [1, 5]
-    Eval:
+  Train:
+  Eval:
     - TopkAcc:
         topk: [1, 5]
diff --git a/ppcls/configs/ImageNet/ResNet/ResNet152.yaml b/ppcls/configs/ImageNet/ResNet/ResNet152.yaml
index da05e570deabd94bc7dcdb24bb5429ff686baa49..1d1883901d96a7258ca53fc811b2de8a0b1631d2 100644
--- a/ppcls/configs/ImageNet/ResNet/ResNet152.yaml
+++ b/ppcls/configs/ImageNet/ResNet/ResNet152.yaml
@@ -30,10 +30,10 @@ Loss:
 
 
 Optimizer:
-  name: Momentum
+  name: "Momentum"
   momentum: 0.9
   lr:
-    name: Piecewise
+    name: "Piecewise"
     learning_rate: 0.1
     decay_epochs: [30, 60, 90]
     values: [0.1, 0.01, 0.001, 0.0001]
@@ -46,80 +46,80 @@ Optimizer:
 DataLoader:
   Train:
     dataset:
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
-        transform_ops:
-          - RandCropImage:
-              size: 224
-          - RandFlipImage:
-              flip_code: 1
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
+      transform_ops:
+        - RandCropImage:
+            size: 224
+        - RandFlipImage:
+            flip_code: 1
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
 
     sampler:
-        name: DistributedBatchSampler
-        batch_size: 64
-        drop_last: False
-        shuffle: True
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: True
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
   Eval:
     # TOTO: modify to the latest trainer
     dataset: 
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
-        transform_ops:
-          - ResizeImage:
-              resize_short: 256
-          - CropImage:
-              size: 224
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
+      transform_ops:
+        - 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
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: False
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
 Infer:
   infer_imgs: "docs/images/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:
+    - 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
+    name: "Topk"
     topk: 5
     class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt"
 
 Metric:
-    Train:
+  Train:
     - TopkAcc:
         topk: [1, 5]
-    Eval:
+  Eval:
     - TopkAcc:
         topk: [1, 5]
diff --git a/ppcls/configs/ImageNet/ResNet/ResNet152_vd.yaml b/ppcls/configs/ImageNet/ResNet/ResNet152_vd.yaml
index 9bfcfc0620ea56a31d07315bdbe930d3483f1b49..1676b267999cca9bebe4c4526ee9482e69ae3a65 100644
--- a/ppcls/configs/ImageNet/ResNet/ResNet152_vd.yaml
+++ b/ppcls/configs/ImageNet/ResNet/ResNet152_vd.yaml
@@ -24,16 +24,17 @@ Loss:
   Train:
     - CELoss:
         weight: 1.0
+        epsilon: 0.1
   Eval:
     - CELoss:
         weight: 1.0
 
 
 Optimizer:
-  name: Momentum
+  name: "Momentum"
   momentum: 0.9
   lr:
-    name: Cosine
+    name: "Cosine"
     learning_rate: 0.1
   regularizer:
     name: 'L2'
@@ -44,80 +45,81 @@ Optimizer:
 DataLoader:
   Train:
     dataset:
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
-        transform_ops:
-          - RandCropImage:
-              size: 224
-          - RandFlipImage:
-              flip_code: 1
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
+      transform_ops:
+        - RandCropImage:
+            size: 224
+        - RandFlipImage:
+            flip_code: 1
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+      batch_transform_ops:
+        - MixupOperator:
+            alpha: 0.2
 
     sampler:
-        name: DistributedBatchSampler
-        batch_size: 64
-        drop_last: False
-        shuffle: True
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: True
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
   Eval:
     # TOTO: modify to the latest trainer
     dataset: 
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
-        transform_ops:
-          - ResizeImage:
-              resize_short: 256
-          - CropImage:
-              size: 224
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
+      transform_ops:
+        - 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
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: False
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
 Infer:
   infer_imgs: "docs/images/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:
+    - 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
+    name: "Topk"
     topk: 5
     class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt"
 
 Metric:
-    Train:
-    - TopkAcc:
-        topk: [1, 5]
-    Eval:
+  Train:
+  Eval:
     - TopkAcc:
         topk: [1, 5]
diff --git a/ppcls/configs/ImageNet/ResNet/ResNet18.yaml b/ppcls/configs/ImageNet/ResNet/ResNet18.yaml
index 12a38ccba5d6c14f884069f49623fdfa94dcbf2f..d9c22f29c563984474bb18f2d797f5276c12653d 100644
--- a/ppcls/configs/ImageNet/ResNet/ResNet18.yaml
+++ b/ppcls/configs/ImageNet/ResNet/ResNet18.yaml
@@ -30,10 +30,10 @@ Loss:
 
 
 Optimizer:
-  name: Momentum
+  name: "Momentum"
   momentum: 0.9
   lr:
-    name: Piecewise
+    name: "Piecewise"
     learning_rate: 0.1
     decay_epochs: [30, 60, 90]
     values: [0.1, 0.01, 0.001, 0.0001]
@@ -46,80 +46,80 @@ Optimizer:
 DataLoader:
   Train:
     dataset:
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
-        transform_ops:
-          - RandCropImage:
-              size: 224
-          - RandFlipImage:
-              flip_code: 1
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
+      transform_ops:
+        - RandCropImage:
+            size: 224
+        - RandFlipImage:
+            flip_code: 1
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
 
     sampler:
-        name: DistributedBatchSampler
-        batch_size: 64
-        drop_last: False
-        shuffle: True
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: True
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
   Eval:
     # TOTO: modify to the latest trainer
     dataset: 
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
-        transform_ops:
-          - ResizeImage:
-              resize_short: 256
-          - CropImage:
-              size: 224
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
+      transform_ops:
+        - 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
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: False
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
 Infer:
   infer_imgs: "docs/images/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:
+    - 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
+    name: "Topk"
     topk: 5
     class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt"
 
 Metric:
-    Train:
+  Train:
     - TopkAcc:
         topk: [1, 5]
-    Eval:
+  Eval:
     - TopkAcc:
         topk: [1, 5]
diff --git a/ppcls/configs/ImageNet/ResNet/ResNet18_vd.yaml b/ppcls/configs/ImageNet/ResNet/ResNet18_vd.yaml
index 6d207d65016f92536d9c6b13057542f4d1577dbe..b7ae6f1fc7902fac2f99cb706a6e157fdf77440f 100644
--- a/ppcls/configs/ImageNet/ResNet/ResNet18_vd.yaml
+++ b/ppcls/configs/ImageNet/ResNet/ResNet18_vd.yaml
@@ -24,16 +24,17 @@ Loss:
   Train:
     - CELoss:
         weight: 1.0
+        epsilon: 0.1
   Eval:
     - CELoss:
         weight: 1.0
 
 
 Optimizer:
-  name: Momentum
+  name: "Momentum"
   momentum: 0.9
   lr:
-    name: Cosine
+    name: "Cosine"
     learning_rate: 0.1
   regularizer:
     name: 'L2'
@@ -44,80 +45,81 @@ Optimizer:
 DataLoader:
   Train:
     dataset:
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
-        transform_ops:
-          - RandCropImage:
-              size: 224
-          - RandFlipImage:
-              flip_code: 1
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
+      transform_ops:
+        - RandCropImage:
+            size: 224
+        - RandFlipImage:
+            flip_code: 1
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+      batch_transform_ops:
+        - MixupOperator:
+            alpha: 0.2
 
     sampler:
-        name: DistributedBatchSampler
-        batch_size: 64
-        drop_last: False
-        shuffle: True
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: True
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
   Eval:
     # TOTO: modify to the latest trainer
     dataset: 
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
-        transform_ops:
-          - ResizeImage:
-              resize_short: 256
-          - CropImage:
-              size: 224
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
+      transform_ops:
+        - 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
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: False
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
 Infer:
   infer_imgs: "docs/images/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:
+    - 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
+    name: "Topk"
     topk: 5
     class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt"
 
 Metric:
-    Train:
-    - TopkAcc:
-        topk: [1, 5]
-    Eval:
+  Train:
+  Eval:
     - TopkAcc:
         topk: [1, 5]
diff --git a/ppcls/configs/ImageNet/ResNet/ResNet200_vd.yaml b/ppcls/configs/ImageNet/ResNet/ResNet200_vd.yaml
index d20aeb12fb510a304ab23f8cb1bfd18fca6e3109..1436de4e5f25c38a5df55078a5f7244e01ff7ac8 100644
--- a/ppcls/configs/ImageNet/ResNet/ResNet200_vd.yaml
+++ b/ppcls/configs/ImageNet/ResNet/ResNet200_vd.yaml
@@ -24,16 +24,17 @@ Loss:
   Train:
     - CELoss:
         weight: 1.0
+        epsilon: 0.1
   Eval:
     - CELoss:
         weight: 1.0
 
 
 Optimizer:
-  name: Momentum
+  name: "Momentum"
   momentum: 0.9
   lr:
-    name: Cosine
+    name: "Cosine"
     learning_rate: 0.1
   regularizer:
     name: 'L2'
@@ -44,80 +45,81 @@ Optimizer:
 DataLoader:
   Train:
     dataset:
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
-        transform_ops:
-          - RandCropImage:
-              size: 224
-          - RandFlipImage:
-              flip_code: 1
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
+      transform_ops:
+        - RandCropImage:
+            size: 224
+        - RandFlipImage:
+            flip_code: 1
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+      batch_transform_ops:
+        - MixupOperator:
+            alpha: 0.2
 
     sampler:
-        name: DistributedBatchSampler
-        batch_size: 64
-        drop_last: False
-        shuffle: True
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: True
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
   Eval:
     # TOTO: modify to the latest trainer
     dataset: 
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
-        transform_ops:
-          - ResizeImage:
-              resize_short: 256
-          - CropImage:
-              size: 224
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
+      transform_ops:
+        - 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
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: False
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
 Infer:
   infer_imgs: "docs/images/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:
+    - 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
+    name: "Topk"
     topk: 5
     class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt"
 
 Metric:
-    Train:
-    - TopkAcc:
-        topk: [1, 5]
-    Eval:
+  Train:
+  Eval:
     - TopkAcc:
         topk: [1, 5]
diff --git a/ppcls/configs/ImageNet/ResNet/ResNet34.yaml b/ppcls/configs/ImageNet/ResNet/ResNet34.yaml
index c05228cacd1779686df6a4ee4d99d8d3e3455eb2..56cc95125105f7bf3abdd4ed32ebc7bbc1e1559f 100644
--- a/ppcls/configs/ImageNet/ResNet/ResNet34.yaml
+++ b/ppcls/configs/ImageNet/ResNet/ResNet34.yaml
@@ -30,10 +30,10 @@ Loss:
 
 
 Optimizer:
-  name: Momentum
+  name: "Momentum"
   momentum: 0.9
   lr:
-    name: Piecewise
+    name: "Piecewise"
     learning_rate: 0.1
     decay_epochs: [30, 60, 90]
     values: [0.1, 0.01, 0.001, 0.0001]
@@ -46,80 +46,80 @@ Optimizer:
 DataLoader:
   Train:
     dataset:
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
-        transform_ops:
-          - RandCropImage:
-              size: 224
-          - RandFlipImage:
-              flip_code: 1
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
+      transform_ops:
+        - RandCropImage:
+            size: 224
+        - RandFlipImage:
+            flip_code: 1
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
 
     sampler:
-        name: DistributedBatchSampler
-        batch_size: 64
-        drop_last: False
-        shuffle: True
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: True
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
   Eval:
     # TOTO: modify to the latest trainer
     dataset: 
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
-        transform_ops:
-          - ResizeImage:
-              resize_short: 256
-          - CropImage:
-              size: 224
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
+      transform_ops:
+        - 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
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: False
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
 Infer:
   infer_imgs: "docs/images/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:
+    - 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
+    name: "Topk"
     topk: 5
     class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt"
 
 Metric:
-    Train:
+  Train:
     - TopkAcc:
         topk: [1, 5]
-    Eval:
+  Eval:
     - TopkAcc:
         topk: [1, 5]
diff --git a/ppcls/configs/ImageNet/ResNet/ResNet34_vd.yaml b/ppcls/configs/ImageNet/ResNet/ResNet34_vd.yaml
index ea5be38a9ee7b89d8d23654a166e01d85ec3773f..482e53a202ce6cc7bc08a473851f3b52da18d8a1 100644
--- a/ppcls/configs/ImageNet/ResNet/ResNet34_vd.yaml
+++ b/ppcls/configs/ImageNet/ResNet/ResNet34_vd.yaml
@@ -24,16 +24,17 @@ Loss:
   Train:
     - CELoss:
         weight: 1.0
+        epsilon: 0.1
   Eval:
     - CELoss:
         weight: 1.0
 
 
 Optimizer:
-  name: Momentum
+  name: "Momentum"
   momentum: 0.9
   lr:
-    name: Cosine
+    name: "Cosine"
     learning_rate: 0.1
   regularizer:
     name: 'L2'
@@ -44,80 +45,81 @@ Optimizer:
 DataLoader:
   Train:
     dataset:
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
-        transform_ops:
-          - RandCropImage:
-              size: 224
-          - RandFlipImage:
-              flip_code: 1
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
+      transform_ops:
+        - RandCropImage:
+            size: 224
+        - RandFlipImage:
+            flip_code: 1
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+      batch_transform_ops:
+        - MixupOperator:
+            alpha: 0.2
 
     sampler:
-        name: DistributedBatchSampler
-        batch_size: 64
-        drop_last: False
-        shuffle: True
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: True
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
   Eval:
     # TOTO: modify to the latest trainer
     dataset: 
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
-        transform_ops:
-          - ResizeImage:
-              resize_short: 256
-          - CropImage:
-              size: 224
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
+      transform_ops:
+        - 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
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: False
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
 Infer:
   infer_imgs: "docs/images/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:
+    - 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
+    name: "Topk"
     topk: 5
     class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt"
 
 Metric:
-    Train:
-    - TopkAcc:
-        topk: [1, 5]
-    Eval:
+  Train:
+  Eval:
     - TopkAcc:
         topk: [1, 5]
diff --git a/ppcls/configs/ImageNet/ResNet/ResNet50.yaml b/ppcls/configs/ImageNet/ResNet/ResNet50.yaml
index a19f78c691a2c10595e6fe2f7aa4913aad4f9e0c..c63c9822d8fbacbaaaee735068dd674eb0c79cf1 100644
--- a/ppcls/configs/ImageNet/ResNet/ResNet50.yaml
+++ b/ppcls/configs/ImageNet/ResNet/ResNet50.yaml
@@ -30,10 +30,10 @@ Loss:
 
 
 Optimizer:
-  name: Momentum
+  name: "Momentum"
   momentum: 0.9
   lr:
-    name: Piecewise
+    name: "Piecewise"
     learning_rate: 0.1
     decay_epochs: [30, 60, 90]
     values: [0.1, 0.01, 0.001, 0.0001]
@@ -46,80 +46,80 @@ Optimizer:
 DataLoader:
   Train:
     dataset:
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
-        transform_ops:
-          - RandCropImage:
-              size: 224
-          - RandFlipImage:
-              flip_code: 1
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
+      transform_ops:
+        - RandCropImage:
+            size: 224
+        - RandFlipImage:
+            flip_code: 1
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
 
     sampler:
-        name: DistributedBatchSampler
-        batch_size: 64
-        drop_last: False
-        shuffle: True
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: True
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
   Eval:
     # TOTO: modify to the latest trainer
     dataset: 
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
-        transform_ops:
-          - ResizeImage:
-              resize_short: 256
-          - CropImage:
-              size: 224
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
+      transform_ops:
+        - 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
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: False
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
 Infer:
   infer_imgs: "docs/images/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:
+    - 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
+    name: "Topk"
     topk: 5
     class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt"
 
 Metric:
-    Train:
+  Train:
     - TopkAcc:
         topk: [1, 5]
-    Eval:
+  Eval:
     - TopkAcc:
         topk: [1, 5]
diff --git a/ppcls/configs/ImageNet/ResNet/ResNet50_retrieval.yml b/ppcls/configs/ImageNet/ResNet/ResNet50_retrieval.yml
deleted file mode 100644
index bf143c861d6b9e774c20c291867d5501763cb7f0..0000000000000000000000000000000000000000
--- a/ppcls/configs/ImageNet/ResNet/ResNet50_retrieval.yml
+++ /dev/null
@@ -1,109 +0,0 @@
-# global configs
-Global:
-  checkpoints: null
-  pretrained_model: null
-  output_dir: "./output/"
-  device: "gpu"
-  class_num: 1000
-  save_interval: 1
-  eval_during_train: True
-  eval_interval: 1
-  epochs: 120
-  print_batch_step: 10
-  use_visualdl: False
-  image_shape: [3, 224, 224]
-  infer_imgs:
-
-# model architecture
-Arch:  
-  name: "RecModel"
-  Backbone:
-    name: "ResNet50"
-  Stoplayer: 
-    name: "flatten_0"
-    output_dim: 2048
-    embedding_size: 512
-  Head:
-    name: "ArcMargin"
-    margin: 0.5
-    scale:  80
- 
-# loss function config for traing/eval process
-Loss:
-  Train:
-    - CELoss:
-        weight: 1.0
-  Eval:
-    - CELoss:
-        weight: 1.0
-
-
-Optimizer:
-  name: Momentum
-  momentum: 0.9
-  lr:
-    name: Piecewise
-    learning_rate: 0.1
-    decay_epochs: [30, 60, 90]
-    values: [0.1, 0.01, 0.001, 0.0001]
-  regularizer:
-    name: 'L2'
-    coeff: 0.0001
-
-# data loader for train and eval
-DataLoader:
-  Train:
-    # Dataset:
-    # Sampler:
-    # Loader:
-    batch_size: 256
-    num_workers: 4
-    file_list: "./dataset/ILSVRC2012/train_list.txt"
-    data_dir: "./dataset/ILSVRC2012/"
-    shuffle_seed: 0
-    transforms:
-        - DecodeImage:
-            to_rgb: True
-            channel_first: False
-        - RandCropImage:
-            size: 224
-        - RandFlipImage:
-            flip_code: 1
-        - NormalizeImage:
-            scale: 1./255.
-            mean: [0.485, 0.456, 0.406]
-            std: [0.229, 0.224, 0.225]
-            order: ''
-        - ToCHWImage:
-  Eval:
-    # TOTO: modify to the latest trainer
-    # Dataset:
-    # Sampler:
-    # Loader:
-    batch_size: 128
-    num_workers: 4
-    file_list: "./dataset/ILSVRC2012/val_list.txt"
-    data_dir: "./dataset/ILSVRC2012/"
-    shuffle_seed: 0
-    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:
-
-Metric:
-    Train:
-    - Topk:
-        k: [1, 5]
-    Eval:
-    - Topk:
-        k: [1, 5]
diff --git a/ppcls/configs/ImageNet/ResNet/ResNet50_vd.yaml b/ppcls/configs/ImageNet/ResNet/ResNet50_vd.yaml
index 15a02e7c29696e551bc3653524f44ba4b427fe8e..f05f8857cfb14ea6d47cc7c68b7b2c9afeb5d791 100644
--- a/ppcls/configs/ImageNet/ResNet/ResNet50_vd.yaml
+++ b/ppcls/configs/ImageNet/ResNet/ResNet50_vd.yaml
@@ -24,16 +24,17 @@ Loss:
   Train:
     - CELoss:
         weight: 1.0
+        epsilon: 0.1
   Eval:
     - CELoss:
         weight: 1.0
 
 
 Optimizer:
-  name: Momentum
+  name: "Momentum"
   momentum: 0.9
   lr:
-    name: Cosine
+    name: "Cosine"
     learning_rate: 0.1
   regularizer:
     name: 'L2'
@@ -44,81 +45,81 @@ Optimizer:
 DataLoader:
   Train:
     dataset:
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
-        transform_ops:
-          - RandCropImage:
-              size: 224
-          - RandFlipImage:
-              flip_code: 1
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/train_list.txt"
+      transform_ops:
+        - RandCropImage:
+            size: 224
+        - RandFlipImage:
+            flip_code: 1
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.485, 0.456, 0.406]
+            std: [0.229, 0.224, 0.225]
+            order: ''
+      batch_transform_ops:
+        - MixupOperator:
+            alpha: 0.2
 
     sampler:
-        name: DistributedBatchSampler
-        batch_size: 64
-        drop_last: False
-        shuffle: True
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: True
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
   Eval:
     # TOTO: modify to the latest trainer
     dataset: 
-        name: ImageNetDataset
-        image_root: "./dataset/ILSVRC2012/"
-        cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
-        transform_ops:
-          - ResizeImage:
-              resize_short: 256
-          - CropImage:
-              size: 224
-          - NormalizeImage:
-              scale: 0.00392157
-              mean: [0.485, 0.456, 0.406]
-              std: [0.229, 0.224, 0.225]
-              order: ''
+      name: "ImageNetDataset"
+      image_root: "./dataset/ILSVRC2012/"
+      cls_label_path: "./dataset/ILSVRC2012/val_list.txt"
+      transform_ops:
+        - 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
+      name: "DistributedBatchSampler"
+      batch_size: 64
+      drop_last: False
+      shuffle: False
     loader:
-        num_workers: 6
-        use_shared_memory: True
+      num_workers: 6
+      use_shared_memory: True
 
 Infer:
   infer_imgs: "docs/images/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:
+    - 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
+    name: "Topk"
     topk: 5
     class_id_map_file: "ppcls/utils/imagenet1k_label_list.txt"
 
 Metric:
-    Train:
-    - TopkAcc:
-        topk: [1, 5]
-    Eval:
+  Train:
+  Eval:
     - TopkAcc:
         topk: [1, 5]
-