refactor: Move EasyOCR models to lfs (#329)

* Extend LFS support to prebuilt_artifacts/models

* Move EasyOCR model artifacts to Git LFS

* fix: EasyOCR model installation script

* fix: update EasyOCR model paths in installation script

* fix: simplify EasyOCR model installation script by removing existing file check

* Fix easyocr model installation script to work with LFS models

* Update EasyOCR model installation script print statement for clarity

Author: baasitsharief

.gitattributes

@@ -1 +1,2 @@
 prebuilt_artifacts/* filter=lfs diff=lfs merge=lfs -text
+prebuilt_artifacts/models/* filter=lfs diff=lfs merge=lfs -text

llm-service/runtime.Dockerfile

@@ -11,8 +11,8 @@ WORKDIR /app
 RUN uv sync -n -p /usr/local/bin/python3.12
 
 RUN wget https://corretto.aws/downloads/latest/amazon-corretto-21-x64-linux-jdk.tar.gz -O amazon-corretto-21-x64-linux-jdk.tar.gz
-RUN wget https://github.com/cloudera/CML_AMP_RAG_Studio/releases/download/model_download/craft_mlt_25k.pth -O craft_mlt_25k.pth
-RUN wget https://github.com/cloudera/CML_AMP_RAG_Studio/releases/download/model_download/latin_g2.pth -O latin_g2.pth
+COPY --chown=cdsw:cdsw ../prebuilt_artifacts/models/craft_mlt_25k.pth /app/craft_mlt_25k.pth
+COPY --chown=cdsw:cdsw ../prebuilt_artifacts/models/latin_g2.pth /app/latin_g2.pth
 RUN wget https://github.com/qdrant/qdrant/releases/download/v1.11.3/qdrant-x86_64-unknown-linux-musl.tar.gz -O qdrant.tar.gz
 
 USER root

prebuilt_artifacts/models/craft_mlt_25k.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4a5efbfb48b4081100544e75e1e2b57f8de3d84f213004b14b85fd4b3748db17
+size 83152330

prebuilt_artifacts/models/latin_g2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aaa95be1c4a9cb3496879bed7c520886ce1164f89e026f0c54488394e74e8c55
+size 15406141

scripts/01_install_base.py

@@ -64,4 +64,4 @@
         ["bash scripts/install_easyocr_model.sh"], shell=True, check=True
     )
 )
-print("Downloading EASYOCR models complete")
+print("Installing EASYOCR models complete")

scripts/install_easyocr_model.sh

@@ -37,20 +37,16 @@
 # BUSINESS ADVANTAGE OR UNAVAILABILITY, OR LOSS OR CORRUPTION OF
 # DATA.
 #
-
-CRAFT_MLT_MODEL_URL=https://github.com/cloudera/CML_AMP_RAG_Studio/releases/download/model_download/craft_mlt_25k.pth
-LATIN_G2_URL=https://github.com/cloudera/CML_AMP_RAG_Studio/releases/download/model_download/latin_g2.pth
-
-mkdir -p .EasyOCR/model
-cd .EasyOCR/model
-
-if [ -f /app/craft_mlt_25k.pth ]; then
-    cp /app/craft_mlt_25k.pth craft_mlt_25k.pth
-    cp /app/latin_g2.pth latin_g2.pth
+if [ -z "$IS_COMPOSABLE" ]; then
+  CRAFT_MLT_MODEL=/home/cdsw/prebuilt_artifacts/models/craft_mlt_25k.pth
+  LATIN_G2_MODEL=/home/cdsw/prebuilt_artifacts/models/latin_g2.pth
 else
-    wget --no-verbose -O craft_mlt_25k.pth ${CRAFT_MLT_MODEL_URL}
-    wget --no-verbose -O latin_g2.pth ${LATIN_G2_URL}
+  CRAFT_MLT_MODEL=/home/cdsw/rag-studio/prebuilt_artifacts/models/craft_mlt_25k.pth
+  LATIN_G2_MODEL=/home/cdsw/rag-studio/prebuilt_artifacts/models/latin_g2.pth
 fi
 
+mkdir -p .EasyOCR/model
+cd .EasyOCR/model
 
-
+cp ${CRAFT_MLT_MODEL} craft_mlt_25k.pth
+cp ${LATIN_G2_MODEL} latin_g2.pth