update test.py in README and default params in main.py

README.md

@@ -136,14 +136,15 @@ bash ./train_multi.sh
 
 如没有预训练模型, 可从 参考资料 中提供的模型下载, 选择模型对应的config后进行效果验证。
 
-如果想要查看预测效果（预测结果输出到图片上）, 请执行:
+如果想要查看预测效果（预测结果输出到图片上）, 请执行(其余参数与训练config参数需一致, 详情请参考代码里面的参数配置):
+
+以 r50_hybrid_branch_lambda1_group6_t1500_dp0_mqs_lft_deformable_detr_plus_iterative_bbox_refinement_plus_plus_two_stage_12eps 为例：
+python test.py --pre_trained_model /path/r50_hybrid_branch_lambda1_group6_t1500_dp0_mqs_lft_deformable_detr_plus_iterative_bbox_refinement_plus_plus_two_stage_12eps.pth --coco_path /path/coco/data --two_stage --num_queries_one2many 1500 --mixed_selection --look_forward_twice
 
 ```
 python test.py --pre_trained_model <checkpoint path> --coco_path <coco path>
 ```
 
-其余对应参数与训练模型参数需一致, 详情请参考代码里面的参数配置:
-
 #### 多卡推理
 
 ```

main.py

@@ -204,12 +204,12 @@ def get_args_parser():
 
     # dataset parameters
     parser.add_argument("--dataset_file", default="coco")
-    parser.add_argument("--coco_path", default="/public/DL_DATA/COCO2017", type=str)
+    parser.add_argument("--coco_path", default="/home/datasets/COCO2017", type=str)
     parser.add_argument("--coco_panoptic_path", type=str)
     parser.add_argument("--remove_difficult", action="store_true")
 
     parser.add_argument(
-        "--output_dir", default="", help="path where to save, empty for no saving"
+        "--output_dir", default="./results", help="path where to save, empty for no saving"
     )
     parser.add_argument(
         "--device", default="cuda", help="device to use for training / testing"

test.py

@@ -74,7 +74,7 @@ def get_parser():
     parser.add_argument("--two_stage", default=False, action="store_true")
     parser.add_argument("--dataset_file", default="coco")
     parser.add_argument("--coco_path", default="/home/datasets/COCO2017", type=str)
-    parser.add_argument("--save_path", default="./result", type=str)
+    parser.add_argument("--save_path", default="./result_img", type=str)
     parser.add_argument(
         "--cache_mode",
         default=False,

train.sh

@@ -8,6 +8,6 @@ echo "Training start ..."
 # coco_path是训练数据集地址，数据是coco format
 sh <config path> --coco_path <coco path>
 # example
-# sh ./configs/two_stage/deformable-detr-hybrid-branch/12eps/r50_hybrid_branch_lambda1_group6_t1500_dp0_mqs_lft_deformable_detr_plus_iterative_bbox_refinement_plus_plus_two_stage.sh \
+# sh ./configs/two_stage/deformable-detr-hybrid-branch/12eps/r50_hybrid_branch_lambda1_group6_t1500_dp0_mqs_lft_deformable_detr_plus_iterative_bbox_refinement_plus_plus_two_stage.sh
 
 echo "Training finished."