update vitdet model links (#6602)

* update model pretrained links

configs/vitdet/README.md

-# Vision transformer Detection
+# Vision Transformer Detection
 
 ## Introduction
 
@@ -15,8 +15,8 @@ non-trivial when new architectures, such as Vision Transformer (ViT) models, arr
 
 | Backbone | Pretrained | Model | Scheduler | Images/GPU  | Box AP | Config | Download |
 |:------:|:--------:|:--------------:|:--------------:|:--------------:|:------:|:------:|:--------:|
-| ViT-base | CAE | Cascade RCNN  | 1x | 1 | -- | [config](./cascade_rcnn_vit_base_hrfpn_cae_1x_coco.yml) | [coming soon]() |
-| ViT-large | CAE | Cascade RCNN  | 1x | 1 | -- | [config](./cascade_rcnn_vit_large_hrfpn_cae_1x_coco.yml) | [coming soon]() |
+| ViT-base | CAE | Cascade RCNN  | 1x | 1 | 52.7 | [config](./cascade_rcnn_vit_base_hrfpn_cae_1x_coco.yml) | [model](https://bj.bcebos.com/v1/paddledet/models/cascade_rcnn_vit_base_hrfpn_cae_1x_coco.pdparams) |
+| ViT-large | CAE | Cascade RCNN  | 1x | 1 | 55.7 | [config](./cascade_rcnn_vit_large_hrfpn_cae_1x_coco.yml) | [model](https://bj.bcebos.com/v1/paddledet/models/cascade_rcnn_vit_large_hrfpn_cae_1x_coco.pdparams) |
 
 **Notes:**
 - Model is trained on COCO train2017 dataset and evaluated on val2017 results of `mAP(IoU=0.5:0.95)`.

configs/vitdet/cascade_rcnn_vit_base_hrfpn_cae_1x_coco.yml

@@ -52,7 +52,7 @@ VisionTransformer:
   epsilon: 0.000001 # 1e-6
   out_indices: [3, 5, 7, 11]
   with_fpn: True
-  pretrained: ~
+  pretrained: https://bj.bcebos.com/v1/paddledet/models/pretrained/vit_base_cae_pretrained.pdparams
 
 HRFPN:
   out_channel: 256

configs/vitdet/cascade_rcnn_vit_large_hrfpn_cae_1x_coco.yml

@@ -15,7 +15,7 @@ VisionTransformer:
   num_heads: 16
   drop_path_rate: 0.25
   out_indices: [7, 11, 15, 23]
-  pretrained: ~
+  pretrained: https://bj.bcebos.com/v1/paddledet/models/pretrained/vit_large_cae_pretrained.pdparams
 
 HRFPN:
   in_channels: [*dim, *dim, *dim, *dim]