Skip to content

Commit

Permalink
Merge 712df80 into 9eac968
Browse files Browse the repository at this point in the history
  • Loading branch information
de-code committed Jun 24, 2019
2 parents 9eac968 + 712df80 commit 0d03a16
Show file tree
Hide file tree
Showing 2 changed files with 53 additions and 0 deletions.
37 changes: 37 additions & 0 deletions docker/Dockerfile.grobid-with-delft
@@ -0,0 +1,37 @@
ARG GROBID_VERSION
ARG GROBID_REPO=lfoppiano/grobid
FROM ${GROBID_REPO}:${GROBID_VERSION}

RUN apt-get update \
&& apt-get install --yes \
python3-minimal python3-venv python3-pip \
openjdk-8-jdk-headless \
curl \
&& JAVA_HOME=/usr/lib/jvm/java-8-openjdk-amd64 pip3 install jep==3.8.2 \
&& apt-get remove --yes openjdk-8-jdk-headless \
&& rm -rf /var/lib/apt/lists/*

# create virtual env (see also https://bugs.python.org/issue24875)
ENV VENV=/opt/venv
RUN python3 -m venv ${VENV} \
&& python3 -m venv --system-site-packages ${VENV}
ENV VIRTUAL_ENV=${VENV} PYTHONUSERBASE=${VENV} PATH=${VENV}/bin:$PATH

# install python build dependencies
RUN pip install setuptools==41.0.1 wheel==0.33.4

# download and install delft
ARG delft_repo=kermitt2/delft
ARG delft_tag=master
RUN curl --progress-bar --location \
"https://github.com/${delft_repo}/archive/${delft_tag}.tar.gz" \
--output "/tmp/${delft_tag}.tar.gz" \
&& tar -C "/opt" -xvf "/tmp/${delft_tag}.tar.gz" \
&& rm "/tmp/${delft_tag}.tar.gz" \
&& ln -s "/opt/delft-${delft_tag}" "/opt/delft"

COPY requirements.delft.minimal.txt /opt/delft/
RUN pip install -r /opt/delft/requirements.delft.minimal.txt

# enable delft
RUN sed -i 's/^grobid.crf.engine.*/grobid.crf.engine=delft/' /opt/grobid/grobid-home/config/grobid.properties
16 changes: 16 additions & 0 deletions docker/requirements.delft.minimal.txt
@@ -0,0 +1,16 @@
keras==2.2.4
numpy==1.16.1
pandas==0.22.0
bleach>=2.1.0
regex==2018.2.21
scikit-learn==0.19.1
tqdm==4.21
gensim==3.4.0
langdetect==1.0.7
textblob==0.15.1
h5py==2.7.1
unidecode==1.0.22
pydot==1.2.4
lmdb==0.94
keras-bert==0.39.0
tensorflow==1.12.0

0 comments on commit 0d03a16

Please sign in to comment.