diff --git a/yolov3-spp/README.md b/yolov3-spp/README.md
index 1bd914f..ad34994 100644
--- a/yolov3-spp/README.md
+++ b/yolov3-spp/README.md
@@ -2,7 +2,7 @@
 
 Currently this is supporting dynamic input shape, if you want to use non-dynamic version, please checkout commit [659fd2b](https://github.com/wang-xinyu/tensorrtx/commit/659fd2b23482197b19dccf746a5a3dbff1611381).
 
-The Pytorch implementation is [ultralytics/yolov3](https://github.com/ultralytics/yolov3). It provides two trained weights of yolov3-spp, `yolov3-spp.pt` and `yolov3-spp-ultralytics.pt`(originally named `ultralytics68.pt`).
+The Pytorch implementation is [ultralytics/yolov3 archive branch](https://github.com/ultralytics/yolov3/tree/archive). It provides two trained weights of yolov3-spp, `yolov3-spp.pt` and `yolov3-spp-ultralytics.pt`(originally named `ultralytics68.pt`).
 
 ## Config
 
diff --git a/yolov3-tiny/README.md b/yolov3-tiny/README.md
index 502175d..5fa7d5d 100644
--- a/yolov3-tiny/README.md
+++ b/yolov3-tiny/README.md
@@ -1,6 +1,6 @@
 # yolov3-tiny
 
-The Pytorch implementation is [ultralytics/yolov3](https://github.com/ultralytics/yolov3).
+The Pytorch implementation is [ultralytics/yolov3 archive branch](https://github.com/ultralytics/yolov3/tree/archive).
 
 ## Excute:
 
diff --git a/yolov3/README.md b/yolov3/README.md
index 8830c58..8bf00d0 100644
--- a/yolov3/README.md
+++ b/yolov3/README.md
@@ -1,6 +1,6 @@
 # yolov3
 
-The Pytorch implementation is [ultralytics/yolov3](https://github.com/ultralytics/yolov3). It provides two trained weights of yolov3, `yolov3.weights` and `yolov3.pt`
+The Pytorch implementation is [ultralytics/yolov3 archive branch](https://github.com/ultralytics/yolov3/tree/archive). It provides two trained weights of yolov3, `yolov3.weights` and `yolov3.pt`
 
 This branch is using tensorrt7 API, there is also a yolov3 implementation using tensorrt4 API, go to [branch trt4/yolov3](https://github.com/wang-xinyu/tensorrtx/tree/trt4/yolov3), which is using [ayooshkathuria/pytorch-yolo-v3](https://github.com/ayooshkathuria/pytorch-yolo-v3).
 
diff --git a/yolov4/README.md b/yolov4/README.md
index 6ef32c5..e2ae169 100644
--- a/yolov4/README.md
+++ b/yolov4/README.md
@@ -1,6 +1,6 @@
 # yolov4
 
-The Pytorch implementation is from [ultralytics/yolov3](https://github.com/ultralytics/yolov3). It can load yolov4.cfg and yolov4.weights(from AlexeyAB/darknet).
+The Pytorch implementation is from [ultralytics/yolov3 archive branch](https://github.com/ultralytics/yolov3/tree/archive). It can load yolov4.cfg and yolov4.weights(from AlexeyAB/darknet).
 
 ## Config
 
diff --git a/yolov5/README.md b/yolov5/README.md
index 4b0cc43..50b00cc 100644
--- a/yolov5/README.md
+++ b/yolov5/README.md
@@ -41,11 +41,11 @@ python gen_wts.py yolov5s.pt
 2. build tensorrtx/yolov5 and run
 
 ```
-cp {ultralytics}/yolov5/yolov5s.wts {tensorrtx}/yolov5/
 cd {tensorrtx}/yolov5/
 // update CLASS_NUM in yololayer.h if your model is trained on custom dataset
 mkdir build
 cd build
+cp {ultralytics}/yolov5/yolov5s.wts {tensorrtx}/yolov5/build
 cmake ..
 make
 sudo ./yolov5 -s [.wts] [.engine] [s/m/l/x/s6/m6/l6/x6 or c/c6 gd gw]  // serialize model to plan file