get docker build working, update documentation in README

.dockerignore

@@ -0,0 +1,4 @@
+.git
+venv/
+__pycache__/
+.output/

.gitignore

@@ -4,3 +4,4 @@ dist/
 venv/
 __pycache__/
 .output/
+.data/

Dockerfile

@@ -0,0 +1,61 @@
+FROM nvidia/cuda:11.7.1-devel-ubuntu22.04
+# FROM python:3.10.13-bookworm
+
+ENV LC_ALL C.UTF-8
+ENV LANG C.UTF-8
+ARG UID=1000
+ARG UNAME=ubuntu
+ARG DEBIAN_FRONTEND=noninteractive
+ENV PATH="/home/$UNAME/.local/bin:$PATH"
+
+# setup user
+RUN useradd -m -u $UID $UNAME
+WORKDIR /home/$UNAME
+RUN chown -R $UNAME:$UNAME /home/$UNAME
+
+# install Python
+RUN apt-get update -y && \
+    apt-get install -y software-properties-common build-essential git gcc curl make vim xorg-dev libxcb-shm0 libglu1-mesa-dev clang libc++-dev libc++abi-dev libsdl2-dev ninja-build libxi-dev libtbb-dev libosmesa6-dev libudev-dev autoconf libtool && \
+    add-apt-repository -y ppa:deadsnakes/ppa && \
+    apt-get update -y && \
+    apt-get install -y python3.10 python3.10-dev python3.10-distutils
+RUN curl -sS https://bootstrap.pypa.io/get-pip.py | python3.10
+
+USER $UNAME
+
+# install python packages
+COPY --chown=$UNAME requirements.txt /tmp/requirements.txt
+RUN python3.10 -m pip install --no-cache-dir --upgrade setuptools distlib pip && \
+    python3.10 -m pip install --no-cache-dir -r /tmp/requirements.txt && \
+    rm /tmp/requirements.txt
+RUN python3.10 -m pip install git+https://github.com/facebookresearch/detectron2.git@fc9c33b1f6e5d4c37bbb46dde19af41afc1ddb2a
+RUN git clone https://github.com/isl-org/Open3D.git && \
+    cd Open3D/ && \
+    mkdir build && \
+    cd build && \
+    cmake -DENABLE_HEADLESS_RENDERING=ON \
+          -DBUILD_GUI=OFF \
+          -DBUILD_WEBRTC=OFF \
+          -DUSE_SYSTEM_GLEW=OFF \
+          -DUSE_SYSTEM_GLFW=OFF \
+          .. && \
+    make && \
+    make install-pip-package
+
+# copy files
+COPY --chown=$UNAME . /home/$UNAME/opdmulti-demo
+
+ENV TORCH_CUDA_ARCH_LIST='8.0'
+ENV FORCE_CUDA=1
+# TODO: just install into /tmp instead of creating a user folder
+RUN mkdir tmp && \
+    cd opdmulti-demo/mask2former/modeling/pixel_decoder/ops && \
+    python3.10 setup.py build install --prefix /home/$UNAME/tmp
+USER root
+RUN mv /home/$UNAME/tmp/local/lib/python3.10/dist-packages/MultiScaleDeformableAttention-1.0-py3.10-linux-x86_64.egg /usr/local/lib/python3.10/dist-packages/ && rm -rf /home/$UNAME/tmp
+USER $UNAME
+ENV PYTHONPATH="/usr/local/lib/python3.10/dist-packages/MultiScaleDeformableAttention-1.0-py3.10-linux-x86_64.egg"
+
+WORKDIR /home/$UNAME/opdmulti-demo
+
+CMD ["python3.10", "app.py"]

README.md

@@ -3,26 +3,47 @@ title: Opdmulti Demo
 emoji: 🌍
 colorFrom: gray
 colorTo: red
-sdk: gradio
-sdk_version: 3.44.3
-app_file: app.py
+sdk: docker
+app_port: 7860
 pinned: false
 license: mit
 ---
 
 Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
 
-# Installation
+## Installation
+
+## Requirements
+
+For the docker build, you will just need docker in order to build and run the container, else you will need
+
+* python 3.10
+* git
+* libosmesa6-dev (for open3d headless rendering)
+
+### Docker (preferred)
+
+To build the docker container, run
+```
+docker build -f Dockerfile -t opdmulti-demo .
+```
+
+### Local
 
 To setup the environment, run the following (recommended in a virtual environment):
 ```
 # install base requirements
-pip install -r pre-requirements.txt
-pip install -r requirements.txt
+python3.10 -m pip install -r requirements.txt
+
+# install detectron2 (must be done after some of the libraries in requirements.txt)
+python3.10 -m pip install git+https://github.com/facebookresearch/detectron2.git@fc9c33b1f6e5d4c37bbb46dde19af41afc1ddb2a
 
+# build library for model
 cd mask2former/modeling/pixel_decoder/ops
 python setup.py build install
 
+# INSTALL OPEN3D
+# --------------
 # Option A: running locally only
 pip install open3d==0.17.0
 
@@ -39,10 +60,21 @@ cd ../examples/python/visualization/
 python headless_rendering.py
 ```
 
-The setup with pre-requirements.txt and requirements.txt resolves the issue that certain packages need to be installed
-prior to others. By default, most additional packages should be added to requirements.txt.
-
 ## Usage
+
+### Docker (preferred)
+
+To run the docker container, execute
+```
+docker run -d --network host -t opdmulti-demo
+```
+If you want to see the output of the container or interact with it,
+
+* use `-it` to run in interactive mode, and remove the `-d` option
+* add `bash` to the end to open into a console rather than running the app directly
+
+### Local
+
 To startup the application locally, run
 ```
 gradio app.py
@@ -50,4 +82,4 @@ gradio app.py
 
 You can view the app on the specified port (usually 7860). To run over an ssh connection, setup port forwarding using
 `-L 7860:localhost:7860` when you create your ssh connection. Note that you will need to install Open3D in headless 
-rendering for this to work.
+rendering for this to work, as described above.

app.py

@@ -19,7 +19,7 @@ SCORE_THRESHOLD = 0.8
 MAX_PARTS = 5
 ARGS = SimpleNamespace(
     config_file="configs/coco/instance-segmentation/swin/opd_v1_real.yaml",
-    model="../data/models/motion_state_pred_opdformerp_rgb.pth",
+    model=".data/models/motion_state_pred_opdformerp_rgb.pth",
     input_format="RGB",
     output=".output",
     cpu=True,
@@ -54,6 +54,7 @@ def predict(rgb_image: str, depth_image: str, intrinsics: np.ndarray, num_sample
         return images
 
     # clear old predictions
+    os.makedirs(ARGS.output, exist_ok=True)
     for path in os.listdir(ARGS.output):
         full_path = os.path.join(ARGS.output, path)
         if os.path.isdir(full_path):

pre-requirements.txt

@@ -1,6 +0,0 @@
-numpy==1.25.2
-Pillow==10.0.1
-torch==2.0.1
-torchaudio==2.0.2
-torchvision==0.15.2
-urllib3==1.26.16

requirements.txt

@@ -1,3 +1,9 @@
+numpy==1.25.2
+Pillow==10.0.1
+torch==2.0.1
+torchaudio==2.0.2
+torchvision==0.15.2
+urllib3==1.26.16
 h5py==3.9.0
 imageio==2.31.3
 opencv-python==4.8.0.76
@@ -10,5 +16,3 @@ timm==0.9.7
 black==23.9.1
 gradio==3.44.3
 huggingface-hub==0.17.2
-detectron2 @ git+https://github.com/facebookresearch/detectron2.git@fc9c33b1f6e5d4c37bbb46dde19af41afc1ddb2a
--e mask2former/modeling/pixel_decoder/ops/