update entity detection (#134)

* Fix requirements.txt (#84) * fix itsdangerous requirements * pin itsdangerous requirements for all flask==1.1.1 servers * update entity detection * rm redundant files * reformat * reformat * codestyle * update requirements * codestyle * add requirements * fixes * fix: analyze_downloads handling incorrect path for agent container * refactor: black analyze_downloads * refactor: ambiguous variable name in analyze_downloads * fixes * codestyle * fixes * Update requirements.txt * fixes * fix pyopenssl requirement * fixes * fixes * fixes * Update test_entity_detection.py * handle case with empty input string * resolve * Update analyze_downloads.py * codestyle * input fixes * logs Co-authored-by: Andrii.Hura <[email protected]> Co-authored-by: mtalimanchuk <[email protected]> Co-authored-by: Dilyara Baymurzina <[email protected]> Co-authored-by: Fedor Ignatov <[email protected]>
deeppavlov · Apr 29, 2022 · cffbe57 · cffbe57
1 parent e76f25c
commit cffbe57
Show file tree

Hide file tree

Showing 24 changed files with 1,315 additions and 2,657 deletions.
diff --git a/annotators/entity_detection/Dockerfile b/annotators/entity_detection/Dockerfile
@@ -1,26 +1,35 @@
-FROM tensorflow/tensorflow:1.15.2-gpu
+FROM deeppavlov/base-gpu:0.12.1
 
 RUN apt-get update && apt-get install git -y
 
+ARG SEQ_TAG_CONFIG
+ARG EL_TAG_CONFIG
 ARG CONFIG
-ARG COMMIT=0.13.0
+ARG FINEGRAINED
+ARG LOWERCASE
 ARG PORT
 ARG SRC_DIR
 ARG SED_ARG=" | "
 
+ENV SEQ_TAG_CONFIG=$SEQ_TAG_CONFIG
+ENV EL_TAG_CONFIG=$EL_TAG_CONFIG
 ENV CONFIG=$CONFIG
+ENV FINEGRAINED=$FINEGRAINED
+ENV LOWERCASE=$LOWERCASE
 ENV PORT=$PORT
 
 COPY ./annotators/entity_detection/requirements.txt /src/requirements.txt
 RUN pip install -r /src/requirements.txt
 
-RUN pip install git+https://github.com/deepmipt/DeepPavlov.git@${COMMIT}
+RUN python -m spacy download en_core_web_sm
 
 COPY $SRC_DIR /src
 
 WORKDIR /src
-RUN python -m deeppavlov install $CONFIG
+RUN python -m deeppavlov install $SEQ_TAG_CONFIG
+RUN python -m deeppavlov install $EL_TAG_CONFIG
 
-RUN sed -i "s|$SED_ARG|g" "$CONFIG"
+RUN sed -i "s|$SED_ARG|g" "$SEQ_TAG_CONFIG"
+RUN sed -i "s|$SED_ARG|g" "$EL_TAG_CONFIG"
 
 CMD gunicorn  --workers=1 --timeout 500 server:app -b 0.0.0.0:8103
diff --git a/annotators/entity_detection/Dockerfile_new b/annotators/entity_detection/Dockerfile_new
diff --git a/annotators/entity_detection/bert_ner_preprocessor.py b/annotators/entity_detection/bert_ner_preprocessor.py