diff --git a/.dockerignore b/.dockerignore
new file mode 100644
index 000000000..5d36f76ab
--- /dev/null
+++ b/.dockerignore
@@ -0,0 +1,2 @@
+datasets/*
+logs/*
diff --git a/.gitignore b/.gitignore
index 8be82c3e7..a2fc1b6a6 100644
--- a/.gitignore
+++ b/.gitignore
@@ -1,4 +1,4 @@
-
+datasets/*
 logs
 
 # compilation and distribution
diff --git a/docker/Dockerfile b/Dockerfile_CUDA10
similarity index 55%
rename from docker/Dockerfile
rename to Dockerfile_CUDA10
index ff5682d17..c3a67946b 100644
--- a/docker/Dockerfile
+++ b/Dockerfile_CUDA10
@@ -9,20 +9,17 @@ RUN apt-get update && apt-get install -y \
 RUN ln -sv /usr/bin/python3 /usr/bin/python
 
 # create a non-root user
-ARG USER_ID=1000
-RUN useradd -m --no-log-init --system  --uid ${USER_ID} appuser -g sudo
-RUN echo '%sudo ALL=(ALL) NOPASSWD:ALL' >> /etc/sudoers
-USER appuser
+COPY . /home/appuser
 WORKDIR /home/appuser
 
 # https://github.com/facebookresearch/detectron2/issues/3933
 ENV PATH="/home/appuser/.local/bin:${PATH}"
 RUN wget https://bootstrap.pypa.io/pip/3.6/get-pip.py && \
-	python3 get-pip.py --user && \
+	python3 get-pip.py && \
 	rm get-pip.py
 
 # install dependencies
 # See https://pytorch.org/ for other options if you use a different version of CUDA
-RUN pip install --user tensorboard cmake   # cmake from apt-get is too old
-RUN pip install --user torch==1.6.0+cu101 torchvision==0.7.0+cu101 -f https://download.pytorch.org/whl/cu101/torch_stable.html
-RUN pip install --user -i https://pypi.tuna.tsinghua.edu.cn/simple tensorboard opencv-python cython yacs termcolor scikit-learn tabulate gdown gpustat faiss-gpu ipdb h5py
+RUN pip install tensorboard cmake   # cmake from apt-get is too old
+RUN pip install torch==1.6.0+cu101 torchvision==0.7.0+cu101 -f https://download.pytorch.org/whl/cu101/torch_stable.html
+RUN pip install -i https://pypi.tuna.tsinghua.edu.cn/simple tensorboard opencv-python cython yacs termcolor scikit-learn tabulate gdown gpustat faiss-gpu ipdb h5py
diff --git a/Dockerfile_CUDA11 b/Dockerfile_CUDA11
new file mode 100644
index 000000000..6f3424311
--- /dev/null
+++ b/Dockerfile_CUDA11
@@ -0,0 +1,28 @@
+# FROM nvidia/cuda:10.1-cudnn7-devel
+FROM nvidia/cuda:11.1.1-cudnn8-devel-ubuntu18.04
+
+# https://github.com/NVIDIA/nvidia-docker/issues/1632
+#RUN rm /etc/apt/sources.list.d/cuda.list
+#RUN rm /etc/apt/sources.list.d/nvidia-ml.list
+ENV DEBIAN_FRONTEND noninteractive
+RUN apt-get update && apt-get install -y \
+	python3-opencv ca-certificates python3-dev git wget sudo ninja-build
+RUN ln -sv /usr/bin/python3 /usr/bin/python
+
+# create a non-root user
+COPY . /home/appuser
+WORKDIR /home/appuser
+
+# https://github.com/facebookresearch/detectron2/issues/3933
+ENV PATH="/home/appuser/.local/bin:${PATH}"
+RUN wget https://bootstrap.pypa.io/pip/3.6/get-pip.py && \
+	python3 get-pip.py && \
+	rm get-pip.py
+
+# install dependencies
+# See https://pytorch.org/ for other options if you use a different version of CUDA
+RUN pip install tensorboard cmake   # cmake from apt-get is too old
+RUN pip install torch==1.10 torchvision==0.11.1 -f https://download.pytorch.org/whl/cu111/torch_stable.html
+# RUN pip install torch==1.6.0+cu101 torchvision==0.7.0+cu101 -f https://download.pytorch.org/whl/cu101/torch_stable.html
+# RUN pip install -i https://pypi.tuna.tsinghua.edu.cn/simple tensorboard opencv-python cython yacs termcolor scikit-learn tabulate gdown gpustat faiss-gpu ipdb h5py
+RUN pip install  tensorboard opencv-python cython yacs termcolor scikit-learn tabulate gdown gpustat faiss-gpu ipdb h5py
diff --git a/docker/README.md b/docker/README.md
index 351b30d09..a120ab995 100644
--- a/docker/README.md
+++ b/docker/README.md
@@ -1,22 +1,29 @@
 # Use the container
 
+### Build Container
+
 ```shell script
 cd docker/
-# Build:
-docker build -t=fastreid:v0 .
-# Launch (requires GPUs)
-nvidia-docker run -v server_path:docker_path --name=fastreid --net=host --ipc=host -it fastreid:v0 /bin/sh
+# Build with the corresponding CUDA version
+
+# CUDA 10
+docker build -t=fastreid:v0 -f Dockerfile_CUDA10 .
+# CUDA 11
+docker build -t=fastreid:v0 -f Dockerfile_CUDA11 .
 ```
 
-## Install new dependencies
+### Run Container
 
-Add the following to `Dockerfile` to make persist changes.
-```shell script
-RUN sudo apt-get update && sudo apt-get install -y vim
 ```
+# Launch (requires GPUs)
+nvidia-docker run -v ${PWD}:/home/appuser --name=fastreid --net=host --ipc=host -it fastreid:v0
+```
+
+### Run Training
+
+Next, follow the [Get Started Doc](https://github.com/JDAI-CV/fast-reid/blob/master/GETTING_STARTED.md#compile-with-cython-to-accelerate-evalution).
 
-Or run them in the container to make temporary changes.
 
 ## A more complete docker container
 
-If you want to use a complete docker container which contains many useful tools, you can check my development environment [Dockerfile](https://github.com/L1aoXingyu/fastreid_docker)
\ No newline at end of file
+If you want to use a complete docker container which contains many useful tools, you can check my development environment [Dockerfile](https://github.com/L1aoXingyu/fastreid_docker)