fix github link

Author: nan2088

docs/installation.mdx

@@ -144,13 +144,13 @@ For more examples, see [Showcase](./showcase/showcase.mdx).
 
 ## Customizing Dockerfile {#selfdocker}
 
-Refer to the [example Dockerfile](https://github.com/torchpipe/torchpipe/-/blob/master/docker/torchpipe.base). After downloading TensorRT and OpenCV in advance, you can compile the corresponding base image.
+Refer to the [example Dockerfile](https://github.com/torchpipe/torchpipe/blob/main/docker/trt9.1.base). After downloading [TensorRT](https://github.com/NVIDIA/TensorRT/tree/release/9.1#optional---if-not-using-tensorrt-container-specify-the-tensorrt-ga-release-build-path) in advance, you can compile the corresponding base image.
 ```
-# put TensorRT-8.6.1.6.Linux.x86_64-gnu.cuda-11.8.tar.gz into thirdparty/
-wget https://codeload.github.com/opencv/opencv/zip/refs/tags/4.5.4 -O thirdparty/opencv-4.5.4.zip
-# docker build --network=host -f docker/torchpipe.base -t hub.c.163.com/neteaseis/ai/torchpipe:base_trt-8.6 .
+# put TensorRT-9.1.0.4.Linux.x86_64-gnu.cuda-11.8.tar.gz into thirdparty/
 
-# docker run --rm   --network=host --gpus=all --ulimit memlock=-1 --ulimit stack=67108864 --privileged=true  -v `pwd`:/workspace -it hub.c.163.com/neteaseis/ai/torchpipe:base_trt-8.6  /bin/bash 
+# docker build --network=host -f docker/trt9.1.base -t torchpipe:base_trt-9.1 .
+
+# docker run --rm   --network=host --gpus=all --ulimit memlock=-1 --ulimit stack=67108864 --privileged=true  -v `pwd`:/workspace -it torchpipe:base_trt-9.1  /bin/bash 
 
 ```
 Base images compiled in this way have smaller sizes than NGC PyTorch images. Please note that `_GLIBCXX_USE_CXX11_ABI==0`.

docs/python/test.mdx

@@ -392,7 +392,7 @@ def test_all_files(file_dir:str, num_clients=10, batch_size = 1,
 
 
 ### Clients with Different Batch Sizes
-In the example provided [here](https://github.com/torchpipe/torchpipe/-/blob/master/examples/yolox/yolox_multithreads_test.py), we use ten clients, each requesting different amounts of data per request, ranging from 1 to 10. We validate the consistency of the results in this case.
+In the example provided [here](https://github.com/torchpipe/torchpipe/blob/main/examples/yolox/yolox_multithreads_test.py), we use ten clients, each requesting different amounts of data per request, ranging from 1 to 10. We validate the consistency of the results in this case.
 
 Typically, users can iterate through all the data in a directory and repeatedly send requests to verify the stability and consistency of the results.
 

i18n/zh/docusaurus-plugin-content-docs/current/installation.mdx

@@ -133,13 +133,13 @@ print(input["result"].shape)  # 失败则此键值一定不存在，即使输入
 
 ## 自定义dockerfile {#selfdocker}
 
-参考[示例dockerfile](https://github.com/torchpipe/torchpipe/-/blob/master/docker/torchpipe.base)，预先下载tensorrt和opencv后可编译相关基础镜像。
-```
-# put TensorRT-8.6.1.6.Linux.x86_64-gnu.cuda-11.8.tar.gz into thirdparty/
-wget https://codeload.github.com/opencv/opencv/zip/refs/tags/4.5.4 -O thirdparty/opencv-4.5.4.zip
-# docker build --network=host -f docker/torchpipe.base -t hub.c.163.com/neteaseis/ai/torchpipe:base_trt-8.6 .
+参考[示例dockerfile](https://github.com/torchpipe/torchpipe/blob/main/docker/trt9.1.base)，预先下载[TensorRT](https://github.com/NVIDIA/TensorRT/tree/release/9.1#optional---if-not-using-tensorrt-container-specify-the-tensorrt-ga-release-build-path)后可编译相关基础环境镜像。
+```bash
+# put TensorRT-9.1.0.4.Linux.x86_64-gnu.cuda-11.8.tar.gz into thirdparty/
+
+# docker build --network=host -f docker/trt9.1.base -t torchpipe:base_trt-9.1 .
 
-# docker run --rm   --network=host --gpus=all --ulimit memlock=-1 --ulimit stack=67108864 --privileged=true  -v `pwd`:/workspace -it hub.c.163.com/neteaseis/ai/torchpipe:base_trt-8.6  /bin/bash 
+# docker run --rm   --network=host --gpus=all --ulimit memlock=-1 --ulimit stack=67108864 --privileged=true  -v `pwd`:/workspace -it torchpipe:base_trt-9.1  /bin/bash 
 
 ```
 这种方式编译出的基础镜像比NGC pytorch镜像体积更小. 需要注意，其`_GLIBCXX_USE_CXX11_ABI==0`

i18n/zh/docusaurus-plugin-content-docs/current/python/test.mdx

@@ -389,7 +389,7 @@ def test_all_files(file_dir:str, num_clients=10, batch_size = 1,
 ```
 
 ### 不同batchsize的客户端
-在[此处例子](https://github.com/torchpipe/torchpipe/-/blob/master/examples/yolox/yolox_multithreads_test.py)中，我们使用十个客户端，客户端每次请求的数据量分别为1，2，3，...10。我们验证了在这种情况下结果的一致性。
+在[此处例子](https://github.com/torchpipe/torchpipe/blob/main/examples/yolox/yolox_multithreads_test.py)中，我们使用十个客户端，客户端每次请求的数据量分别为1，2，3，...10。我们验证了在这种情况下结果的一致性。
 通常来讲，用户可遍历一个目录中的所有数据，并多次重复发送，以验证结果的稳定性和一致性。
 :::warning tensorrt模型推理结果不一致
 如果模型本身batchsize>1,tensorrt推理结果可能会有少量差异（输入数据的batchsize不同时，模型选取的优化方法可能不同）；如果batchsize==1，这种差异不应该存在，然而重新生成模型后，推理结果仍然有可能变化。