Merge pull request #325 from open-edge-platform/update-branch

bug: fix rag toolkit fix vulnerability (#868)

Author: gooishin

usecases/ai/microservices/speech-to-text/Dockerfile

@@ -47,7 +47,8 @@ ENV PATH="/usr/src/.venv/bin:$PATH"
 
 # Install python dependencies
 COPY --chown=openvino:openvino requirements.txt .
-RUN python3 -m pip install -r requirements.txt
+RUN python3 -m pip install -U pip \
+    && python3 -m pip install -r requirements.txt
 
 COPY --chown=openvino:openvino main.py .
 COPY --chown=openvino:openvino utils.py .

usecases/ai/microservices/text-to-speech/Dockerfile

@@ -54,7 +54,7 @@ RUN mkdir -p /usr/src/app/data/models \
 # Install Melo TTS
 COPY requirements.txt .
 RUN python3 -m pip install --upgrade pip \
-    && python3 -m pip install git+https://github.com/zhaohb/MeloTTS-OV.git@1962eee \
+    && python3 -m pip install git+https://github.com/zhaohb/MeloTTS-OV.git@1962eee --extra-index-url https://download.pytorch.org/whl/cpu \
     && python3 -m pip install -r requirements.txt \
     && python3 -m unidic download
 

usecases/ai/microservices/text-to-speech/requirements.txt

@@ -1,8 +1,11 @@
 fastapi[all]
 uvicorn
 pydantic<2
+soxr
 
 --extra-index-url https://download.pytorch.org/whl/cpu
-optimum-intel[openvino,nncf]==1.21.0
-openvino==2024.6.0
+torch==2.8.0
+torchaudio==2.8.0
+optimum-intel[openvino,nncf]==1.26.1
+openvino==2025.3.0
 transformers

usecases/ai/rag-toolkit/backend/Dockerfile

@@ -3,7 +3,7 @@
 
 FROM debian:12-slim
 ARG DEBIAN_FRONTEND=noninteractive
-ARG OPENVINO_VERSION=2025.1.0
+ARG OPENVINO_VERSION=2025.3.0
 SHELL ["/bin/bash", "-o", "pipefail", "-c"]
 
 USER root
@@ -58,7 +58,7 @@ RUN python3 -m venv /usr/src/.venv
 ENV PATH="/usr/src/.venv/bin:$PATH"
 
 # Install python dependencies
-RUN python3 -m pip install fastapi[all]==0.111.0 \
+RUN python3 -m pip install fastapi[all] \
         sse_starlette==2.1.2 \
         scipy \
         soundfile \
@@ -75,10 +75,10 @@ RUN python3 -m pip install fastapi[all]==0.111.0 \
         cached_path==1.6.3 \
         python-magic
 RUN python3 -m pip install --extra-index-url https://download.pytorch.org/whl/cpu \
-        'torch>=2.6.0' \
-        'torchaudio>=2.6.0' \
-        openvino==2025.1.0 \
-        optimum-intel[openvino,nncf]==1.22.0
+        torch==2.8.0 \
+        torchaudio==2.8.0 \
+        openvino==2025.3.0 \
+        optimum-intel[openvino,nncf]==1.26.1
 
 COPY --chown=openvino:openvino backend .
 

usecases/ai/rag-toolkit/backend/main.py

@@ -294,6 +294,10 @@ def _formatting_rag_fusion_result(retrieval_list, query):
     if 'rag' in data:
         data.pop('rag')
 
+    if 'max_tokens' in data:
+        data['max_completion_tokens'] = data.pop('max_tokens')
+
+    print(f"Data: {data}")
     try:
         llm = requests.post(
             endpoint,

usecases/ai/rag-toolkit/docker-compose.yml

@@ -15,7 +15,7 @@ services:
     networks:
       - app-network
     ports:
-      - "127.0.0.1:8010:3000"
+      - "${HOST:-127.0.0.1}:8010:3000"
 
   backend:
     build:
@@ -31,10 +31,10 @@ services:
     networks:
       - app-network
     ports:
-      - "127.0.0.1:8011:8011"
+      - "${HOST:-127.0.0.1}:8011:8011"
     environment:
       - OPENAI_BASE_URL=http://serving:8000/v1
-      - ALLOWED_CORS=["http://localhost:8010"]
+      - ALLOWED_CORS=["http://localhost:8010","http://localhost:3000"]
       - EMBEDDING_DEVICE=${EMBEDDING_DEVICE:-CPU}
       - RERANKER_DEVICE=${RERANKER_DEVICE:-CPU}
     depends_on:
@@ -61,7 +61,7 @@ services:
     networks:
       - app-network
     ports:
-      - "127.0.0.1:8012:8000"
+      - "${HOST:-127.0.0.1}:8012:8000"
     environment:
       - DEFAULT_MODEL_ID=Qwen/Qwen2.5-7B-Instruct
       - MODEL_PATH=./data/models/llm
@@ -92,7 +92,7 @@ services:
       - BERT_DEVICE=${BERT_DEVICE:-CPU}
       - ALLOWED_CORS=["*"]
     ports:
-      - 127.0.0.1:8013:5995
+      - ${HOST:-127.0.0.1}:8013:5995
     devices:
       - /dev/dri:/dev/dri
     volumes:
@@ -110,11 +110,11 @@ services:
     networks:
       - app-network
     environment:
-      - DEFAULT_MODEL_ID=openai/whisper-tiny 
+      - DEFAULT_MODEL_ID=openai/whisper-base
       - STT_DEVICE=CPU
       - ALLOWED_CORS=["*"]
     ports:
-      - 127.0.0.1:8014:5996
+      - ${HOST:-127.0.0.1}:8014:5996
     devices:
       - /dev/dri:/dev/dri
     volumes: