e

Dockerfile

@@ -23,13 +23,6 @@ USER root
 USER root
 RUN apt-get update && \
     apt-get install -y nginx python3-pip curl gnupg lsb-release && \
-    export GCSFUSE_REPO=gcsfuse-$(lsb_release -c -s) && \
-    echo "deb http://packages.cloud.google.com/apt $GCSFUSE_REPO main" \
-      | tee /etc/apt/sources.list.d/gcsfuse.list && \
-    curl https://packages.cloud.google.com/apt/doc/apt-key.gpg \
-      | apt-key add - && \
-    apt-get update && \
-    apt-get install -y gcsfuse && \
     rm -rf /var/lib/apt/lists/* && \
     rm -f /etc/nginx/sites-enabled/default && \
     rm -f /etc/nginx/conf.d/default.conf
@@ -72,12 +65,6 @@ ENV HF_HOME=/tmp/hf_cache \
 RUN mkdir -p /tmp/hf_cache \
  && chmod 777 /tmp/hf_cache
 
-RUN mkdir -p /mnt/project/data \
-             /mnt/project/vectorstore_index \
-             /mnt/project/whoosh_index \
-             /mnt/project/cache \
- && chmod -R a+rwx /mnt/project
-
 # Install Python deps from requirements (ensures numpy/pandas compatibility), then ASGI
 # copy in your requirements
 COPY --from=backend-builder /app/backend/requirements.txt /tmp/requirements.txt

backend/main.py

@@ -44,9 +44,9 @@ logger = logging.getLogger(__name__)
 
 class Settings(BaseSettings):
     # Parquet + Whoosh/FAISS
-    parquet_path: str = "/mnt/project/data/consolidated_clean_pred.parquet"
-    whoosh_dir:    str = "/mnt/project/whoosh_index"
-    vectorstore_path: str = "/mnt/project/vectorstore_index"
+    parquet_path: str = "gs://mda_eu_project/data/consolidated_clean_pred.parquet"
+    whoosh_dir:    str = "gs://mda_eu_project/whoosh_index"
+    vectorstore_path: str = "gs://mda_eu_project/vectorstore_index"
     # Models
     embedding_model:     str = "sentence-transformers/LaBSE"
     llm_model:           str = "google/mt5-large"#"bigscience/bloom-3b"#"RedHatAI/Meta-Llama-3.1-8B-Instruct-quantized.w4a16"

run.sh

@@ -13,82 +13,26 @@
 #pkill -F http_server.pid
 #rm http_server.pid
 # Start nginx in foreground
-#echo "HF_SDK = $HF_SPACE_SDK, APP_PORT = $APP_PORT, PORT = $PORT"
-#echo "$GCP_SA_JSON" > /tmp/sa.json
-#chmod 600 /tmp/sa.json
-
-#export GOOGLE_APPLICATION_CREDENTIALS=/tmp/sa.json
-#nginx -g "daemon off;" &
-#NGINX_PID=$!
-
-# Serve static files via simple HTTP server on port 8000
-#env HTTP_SERVER_PORT=8000
-#python3 -m http.server --directory ./static --bind 0.0.0.0 ${HTTP_SERVER_PORT} &
-#HTTP_SERVER_PID=$!
-
-# Setup cleanup on exit
-#cleanup() {
-#  echo "Shutting down servers..."
-#  kill "${HTTP_SERVER_PID}" || true
-#  kill "${NGINX_PID}" || true
-#}
-#trap cleanup EXIT
-
-# Start FastAPI; ensure correct module path
-#uvicorn "app.main:app" --host 0.0.0.0 --port 7860
-
-#!/bin/bash
-set -e
-
-echo "Starting with HF_SDK=$HF_SPACE_SDK, APP_PORT=$APP_PORT, PORT=$PORT"
-
-# Write out GCP key and export
+echo "HF_SDK = $HF_SPACE_SDK, APP_PORT = $APP_PORT, PORT = $PORT"
 echo "$GCP_SA_JSON" > /tmp/sa.json
 chmod 600 /tmp/sa.json
-export GOOGLE_APPLICATION_CREDENTIALS=/tmp/sa.json
-
-# 1️⃣ Mount your GCS bucket under /mnt/project
-MOUNT_POINT=/mnt/project
-BUCKET_NAME=mda_eu_project
-#mkdir -p ${MOUNT_POINT}
-# allow_other so nginx, uvicorn, etc. (non-root) can write
-gcsfuse \
-  --implicit-dirs \
-  --file-mode=777 \
-  --dir-mode=777 \
-  ${BUCKET_NAME} \
-  "${MOUNT_POINT}"
-
-# 2️⃣ Ensure our four dirs exist
-#for d in data vectorstore_index whoosh_index cache; do
-#  mkdir -p ${MOUNT_POINT}/$d
-#done
-
-# 3️⃣ Point HF caches into bucket
-export HF_HOME=${MOUNT_POINT}/cache
-export TRANSFORMERS_CACHE=${MOUNT_POINT}/cache
-export HF_HUB_CACHE=${MOUNT_POINT}/cache
-export XDG_CACHE_HOME=${MOUNT_POINT}/cache
 
-# 4️⃣ (Optional) export paths for your app
-export DATA_DIR=${MOUNT_POINT}/data
-export VSTORE_DIR=${MOUNT_POINT}/vectorstore_index
-export WHOOSH_DIR=${MOUNT_POINT}/whoosh_index
-
-# 5️⃣ Start nginx + static server
+export GOOGLE_APPLICATION_CREDENTIALS=/tmp/sa.json
 nginx -g "daemon off;" &
 NGINX_PID=$!
 
-python3 -m http.server --directory ./static --bind 0.0.0.0 ${HTTP_SERVER_PORT:-8000} &
+# Serve static files via simple HTTP server on port 8000
+env HTTP_SERVER_PORT=8000
+python3 -m http.server --directory ./static --bind 0.0.0.0 ${HTTP_SERVER_PORT} &
 HTTP_SERVER_PID=$!
 
-# 6️⃣ Cleanup
-cleanup(){
-  echo "Shutting down…"
-  kill $HTTP_SERVER_PID || true
-  kill $NGINX_PID        || true
+# Setup cleanup on exit
+cleanup() {
+  echo "Shutting down servers..."
+  kill "${HTTP_SERVER_PID}" || true
+  kill "${NGINX_PID}" || true
 }
 trap cleanup EXIT
 
-# 7️⃣ Finally, launch FastAPI/uvicorn
-uvicorn "app.main:app" --host 0.0.0.0 --port ${PORT:-7860}
+# Start FastAPI; ensure correct module path
+uvicorn "app.main:app" --host 0.0.0.0 --port 7860