qubvel / segmentation_models

Segmentation models with pretrained backbones. Keras and TensorFlow Keras.
MIT License
4.77k stars 1.03k forks source link

dockerfile #516

Open elliestath opened 2 years ago

elliestath commented 2 years ago

Hi, does anyone have a dockerfile for this?

I tried to compile my own. Everything seems to compile ok but I had no luck running the jupyter notebook

FROM nvidia/cuda:10.2-cudnn7-runtime-ubuntu18.04

ARG NUM_THREADS=1

ENV TZ=Europe/Rome
RUN ln -snf /usr/share/zoneinfo/$TZ /etc/localtime && echo $TZ > /etc/timezone

#ENV PATH=/usr/local/cuda-10.2/bin:$PATH

# Get dependencies
RUN apt-get update -qq && apt-get install -qq

RUN apt-get install -y wget build-essential checkinstall
RUN apt-get install -y libreadline-gplv2-dev libncursesw5-dev libssl-dev \
    libsqlite3-dev tk-dev libgdbm-dev libc6-dev libbz2-dev libffi-dev zlib1g-dev libjpeg62 libjpeg62-dev

RUN wget https://www.python.org/ftp/python/3.7.9/Python-3.7.9.tgz
RUN tar xzf Python-3.7.9.tgz

WORKDIR Python-3.7.9
RUN ./configure --enable-optimizations
RUN make altinstall

RUN apt-get install -y python3-pip
RUN pip3 install --upgrade pip

WORKDIR ../
RUN apt-get install -y git

##pip needes to be upgrades to include tensorflow 2.1

RUN pip3 install jupyter
RUN git clone https://github.com/qubvel/segmentation_models

# install dependencies
COPY requirements.txt /tmp

RUN pip3 install -r /tmp/requirements.txt

I am not that familiar with docker syntax but when I try to run my notebook like

docker run -it -p 8888:8888 image:version
jupyter notebook --ip 0.0.0.0 --no-browser --allow-root
localhost:8888/tree‌

I get "file not found"

Shubham-Gupta3197 commented 2 years ago

@elliestath hey, I can help you in this. Please assign this issue to me.

elliestath commented 2 years ago

hey @Shubham-Gupta3197 thanks! I guess I don't have permission to assign an issue in a public repo