With a functional dockerfile

Dockerfile

@@ -0,0 +1,18 @@
+FROM pytorch/pytorch:latest
+
+# prevent apt from hanging
+ARG DEBIAN_FRONTEND=noninteractive
+
+ENV HOME /workspace
+
+WORKDIR $HOME
+
+# dependency: lang-segment-anything
+RUN apt update
+RUN apt install -y git
+RUN apt install libgl1-mesa-glx -y
+RUN apt install libglib2.0-0 -y
+RUN git clone https://github.com/kauevestena/lang-segment-anything.git
+WORKDIR $HOME/lang-segment-anything
+RUN pip install -e .
+RUN python running_test.py

README.md

@@ -42,6 +42,21 @@ conda env create -f environment.yml
 conda activate lsa
 ```
 
+#### Docker Installation
+
+first buid the image:
+
+```
+docker build --tag 'lang-segment-anything' .
+```
+
+then run it (on windows use powershell):
+
+```
+docker run --name running-langsam --mount type=bind,source=D:\data,target=/workspace/data --gpus all -it 'lang-segment-anything' 
+```
+
+replace `D:\data` with the desired path for mounting a shared folder. If you wanna more than one running container, you can remove `--name running_deep_pavements`. Include `--detach` to run in background and `--rm` to remove on exit
 
 ### Usage