optimize the dockerfile

.github/workflows/ci-build.yml

@@ -36,19 +36,19 @@ jobs:
     runs-on: ubuntu-latest
 
     steps:
-      # - name: Maximize build disk space
-      #   uses: easimon/maximize-build-space@v10
-      #   with:
-      #     root-reserve-mb: 512
-      #     swap-size-mb: 1024
-      #     remove-dotnet: 'true'
-      - name: Create more disk space
-        run: |
-          sudo rm -rf /usr/share/dotnet
-          sudo rm -rf /opt/ghc
-          sudo rm -rf "/usr/local/share/boost"
-          sudo rm -rf "$AGENT_TOOLSDIRECTORY"
-          sudo rm -rf /opt/hostedtoolcache
+      - name: Maximize build disk space
+        uses: easimon/maximize-build-space@v10
+        with:
+           root-reserve-mb: 512
+           swap-size-mb: 1024
+           remove-dotnet: 'true'
+#      - name: Create more disk space
+#        run: |
+#          sudo rm -rf /usr/share/dotnet
+#          sudo rm -rf /opt/ghc
+#          sudo rm -rf "/usr/local/share/boost"
+#          sudo rm -rf "$AGENT_TOOLSDIRECTORY"
+#          sudo rm -rf /opt/hostedtoolcache
       - uses: actions/checkout@v4
       - name: Build and push
         id: docker_build

Dockerfile.datastet

@@ -34,14 +34,9 @@ RUN ./gradlew installModels  --no-daemon --info --stacktrace \
 # Preparing distribution
 WORKDIR /opt/grobid
 RUN unzip -o /opt/grobid/datastet-source/build/distributions/datastet-*.zip -d datastet_distribution \
-  && mv datastet_distribution/datastet-* datastet
+  && mv datastet_distribution/datastet-* datastet \
+  && rm -rf /opt/grobid/datastet-source/build
 
-# Cleanup
-RUN rm -rf datastet-source/.git \
-    && rm -rf /opt/grobid/datastet-source/build/distributions/datastet-*.zip \
-    && rm -rf datastet-source/build
-
-WORKDIR /opt
 
 # -------------------
 # build runtime image
@@ -82,7 +77,7 @@ WORKDIR /opt/grobid
 #RUN mv elmo_2x4096_512_2048cnn_2xhighway_5.5B_weights.hdf5 /opt/elmo/
 
 # this will build and load embeddings on the image forever (only if required by the config) :)
-RUN #python3 preload_embeddings.py --registry ./resources-registry.json --embedding word2vec
+RUN python3 preload_embeddings.py --registry ./resources-registry.json --embedding word2vec
 
 ARG GROBID_VERSION
 ENV GROBID_VERSION=${GROBID_VERSION:-latest}