upload

Dockerfile

@@ -1,29 +1,36 @@
 # Use Python 3.9 slim image as base
 FROM python:3.9-slim
 
-# Set working directory
+# Create app directory
 WORKDIR /app
 
-# Set environment variables
-ENV PYTHONUNBUFFERED=1 \
-    PYTHONDONTWRITEBYTECODE=1
-
-# Install system dependencies
-RUN apt-get update && \
-    apt-get install -y --no-install-recommends \
-    build-essential \
-    curl \
-    && rm -rf /var/lib/apt/lists/*
-
-# Copy requirements file
+# Copy requirements and install dependencies
 COPY requirements.txt .
-
-# Install Python dependencies
-RUN pip install --no-cache-dir -r requirements.txt
-
-# Copy the rest of the application
+RUN pip install -r requirements.txt
+
+# Create cache directories with proper permissions
+RUN mkdir -p /tmp/huggingface \
+    && mkdir -p /tmp/transformers \
+    && mkdir -p /tmp/sentence-transformers \
+    && mkdir -p /tmp/matplotlib \
+    && chmod 777 /tmp/huggingface \
+    && chmod 777 /tmp/transformers \
+    && chmod 777 /tmp/sentence-transformers \
+    && chmod 777 /tmp/matplotlib
+
+# Copy application code
 COPY . .
 
+# Set environment variables
+ENV HF_HOME=/tmp/huggingface
+ENV TRANSFORMERS_CACHE=/tmp/transformers
+ENV SENTENCE_TRANSFORMERS_HOME=/tmp/sentence-transformers
+ENV MPLCONFIGDIR=/tmp/matplotlib
+ENV TF_ENABLE_ONEDNN_OPTS=0
+
+# Preload models during build
+RUN python -c "from services.api.chatbot.model_init import embedding_model"
+
 # Expose the port
 EXPOSE 7860
 

services/api/chatbot/model_init.py

@@ -0,0 +1,19 @@
+import os
+from sentence_transformers import SentenceTransformer
+from langchain_community.embeddings import HuggingFaceEmbeddings
+
+# Set cache directories
+os.environ['HF_HOME'] = '/tmp/huggingface'
+os.environ['TRANSFORMERS_CACHE'] = '/tmp/transformers'
+os.environ['SENTENCE_TRANSFORMERS_HOME'] = '/tmp/sentence-transformers'
+
+# Initialize model singleton
+def init_embedding_model():
+    return HuggingFaceEmbeddings(
+        model_name="sentence-transformers/all-MiniLM-L6-v2",
+        cache_folder="/tmp/transformers",
+        model_kwargs={'device': 'cpu'}  # Force CPU usage for better compatibility
+    )
+
+# Create global instance
+embedding_model = init_embedding_model()

services/api/chatbot/retrieval.py

@@ -24,7 +24,7 @@ from .config import (
     SENTENCE_TRANSFORMERS_HOME,
     MPLCONFIGDIR
 )
-
+from .model_init import embedding_model
 
 load_dotenv()
 
@@ -35,12 +35,6 @@ MYSQL_HOST = os.getenv("MYSQL_HOST")
 MYSQL_DB = os.getenv("MYSQL_DB")
 MYSQL_PORT = int(os.getenv("MYSQL_PORT", 3306))
 
-# Initialize embedding model with a more compatible configuration
-embedding_model = HuggingFaceEmbeddings(
-    model_name="sentence-transformers/all-MiniLM-L6-v2",
-    model_kwargs={'device': 'cpu'},
-    encode_kwargs={'normalize_embeddings': True}
-)
 client = qdrant_client.QdrantClient(QDRANT_HOST, api_key=QDRANT_API_KEY)
 
 def connect_db():