diff --git a/configs/ViTDet/ViTDet-ViTAE-Small-100e.py b/configs/ViTDet/ViTDet-ViTAE-Small-100e.py
index 632dedd..539b912 100644
--- a/configs/ViTDet/ViTDet-ViTAE-Small-100e.py
+++ b/configs/ViTDet/ViTDet-ViTAE-Small-100e.py
@@ -35,6 +35,7 @@
         NC_group=[1, 1, 96], 
         RC_group=[1, 1, 1],
         mlp_ratio=3.,
+        window=[2, 5, 8, 13]
         ),
     neck=dict(
         type='FPN',
diff --git a/mmdet/models/backbones/vitae_s.py b/mmdet/models/backbones/vitae_s.py
index 7f44fd6..4e89b91 100644
--- a/mmdet/models/backbones/vitae_s.py
+++ b/mmdet/models/backbones/vitae_s.py
@@ -40,7 +40,7 @@ def forward(self, x, size):
 class BasicLayer(nn.Module):
     def __init__(self, img_size=224, in_chans=3, embed_dims=64, token_dims=64, downsample_ratios=4, kernel_size=7, RC_heads=1, NC_heads=6, dilations=[1, 2, 3, 4],
                 RC_tokens_type='performer', NC_tokens_type='transformer', RC_group=1, NC_group=64, NC_depth=2, dpr=0.1, mlp_ratio=4., qkv_bias=True, 
-                qk_scale=None, drop=0, attn_drop=0., norm_layer=nn.LayerNorm, window_size=(14, 14), use_checkpoint=False, interval=3):
+                qk_scale=None, drop=0, attn_drop=0., norm_layer=nn.LayerNorm, window_size=(14, 14), use_checkpoint=False, window=[2, 5, 8, 13]):
         super().__init__()
         self.img_size = img_size
         self.in_chans = in_chans
@@ -72,8 +72,8 @@ def __init__(self, img_size=224, in_chans=3, embed_dims=64, token_dims=64, downs
             NormalCell(token_dims, NC_heads, mlp_ratio=mlp_ratio, qkv_bias=qkv_bias, qk_scale=qk_scale, drop=drop, attn_drop=attn_drop,
                        drop_path=dpr[i] if isinstance(dpr, list) else dpr, norm_layer=norm_layer, 
                        group=NC_group, tokens_type=NC_tokens_type,
-                       img_size=img_size // downsample_ratios, window_size=window_size if (i + 1) % interval != 0 else full_image_size, 
-                       window=True if (i + 1) % interval != 0 else False)
+                       img_size=img_size // downsample_ratios, window_size=window_size if i in window else full_image_size, 
+                       window=True if i in window else False)
                        for i in range(NC_depth)])
 
     def forward(self, x, size):
@@ -129,7 +129,7 @@ def __init__(self,
                 frozen_stages=-1,
                 use_checkpoint=False,
                 load_ema=True,
-                interval=3,
+                window=[2, 5, 8, 13],
                 pretrained=None):
         super().__init__()
 
@@ -169,7 +169,7 @@ def __init__(self,
                 self.kernel_size[i], self.RC_heads[i], self.NC_heads[i], self.dilaions[i],
                 self.RC_tokens_type[i], self.NC_tokens_type[i], self.RC_group[i], self.NC_group[i], self.NC_depth[i], dpr[startDpr:self.NC_depth[i]+startDpr],
                 mlp_ratio=self.mlp_ratio[i], qkv_bias=self.qkv_bias[i], qk_scale=self.qk_scale[i], drop=self.drop[i], attn_drop=self.attn_drop[i],
-                norm_layer=self.norm_layer[i], window_size=window_size, use_checkpoint=use_checkpoint, interval=interval)
+                norm_layer=self.norm_layer[i], window_size=window_size, use_checkpoint=use_checkpoint, window=window)
             )
             img_size = img_size // self.downsample_ratios[i]
             in_chans = self.tokens_dims[i]