Updated Dockerfile for Siamese (#209)

* changed standardize output to be json compatible

Signed-off-by: mmerler <michele.merler@gmail.com>

* allow input fields containerize and clustering to be either strings (from API) or dict (from command line)

Signed-off-by: mmerler <michele.merler@gmail.com>

* fixed threshold reading from .ini config files in run_models

Signed-off-by: mmerler <michele.merler@gmail.com>

* updated dev base image, added batching option for siamese model inference, set siamese model and tokenizer to local, updated github actions versions

Signed-off-by: mmerler <michele.merler@gmail.com>

* updated Dockerfile for Siamese default

Signed-off-by: mmerler <michele.merler@gmail.com>

---------

Signed-off-by: mmerler <michele.merler@gmail.com>

Dockerfile

@@ -34,7 +34,11 @@ RUN  python -m pip install --upgrade pip wheel build setuptools; \
      cd entity_standardizer; python -m build; pip install dist/entity_standardizer_tca-1.0-py3-none-any.whl; cd ..; \
      pip install -r /app/requirements.txt; \
      python benchmarks/generate_data.py; \
-     python benchmarks/run_models.py;
+     mkdir -p models/deploy; \
+     wget https://www.dropbox.com/s/bobcey7ufklw7mr/siamese_model.zip -O "./models/deploy/siamese_model.zip"; \
+     unzip models/deploy/siamese_model.zip -d models/deploy/; \
+     rm models/deploy/siamese_model.zip; \
+     python benchmarks/run_models.py -model_type siamese -batch_size 5;
 
 RUN chown -R 1001:0 ./