Deploy PuppyCompanion FastAPI 2025-06-02 10:08:59

README.md

@@ -1,6 +1,3 @@
-# Fichier README.md pour Hugging Face Spaces
-# À placer à la racine de votre Space
-
 ---
 title: PuppyCompanion
 emoji: 🐶
@@ -9,92 +6,8 @@ colorTo: purple
 sdk: docker
 pinned: false
 license: mit
-app_port: 7860
 ---
 
 # PuppyCompanion 🐶
 
-An intelligent AI assistant specialized in puppy care and training, powered by advanced RAG (Retrieval-Augmented Generation) technology and a modern FastAPI backend.
-
-## Features
-
-- **🐕 Specialized Knowledge Base**: Expert information from professional canine care resources
-- **🧠 Smart Agent Workflow**: LangGraph-powered decision making with tool selection
-- **🔍 Dual Search System**: 
-  - RAG system for specialized puppy knowledge from preprocessed chunks
-  - Web search integration via Tavily for up-to-date information
-- **📱 Modern Mobile-like Interface**: Responsive web interface with real-time updates
-- **🔧 Real-time Debug Console**: Watch the AI's decision process in real-time
-- **⚡ FastAPI Backend**: High-performance async API with WebSocket support
-- **📊 Detailed Source Attribution**: Shows exactly which knowledge sources were used
-
-## How it Works
-
-1. **Question Analysis**: The agent workflow analyzes your question and selects appropriate tools
-2. **RAG Search**: For dog-related questions, searches the specialized knowledge base using Qdrant vector database
-3. **Quality Evaluation**: Determines if the RAG response is comprehensive enough
-4. **Web Search Fallback**: If needed, uses Tavily to search for additional current information
-5. **Response Generation**: Combines knowledge sources to provide a comprehensive answer
-6. **Real-time Updates**: All processing steps are visible in the debug console via WebSocket
-
-## Usage
-
-Simply ask any question about puppy care, training, behavior, or health. Examples:
-- "How do I house train my 8-week-old puppy?"
-- "What vaccination schedule should I follow?"
-- "How to stop my puppy from biting furniture?"
-- "Best foods for a growing German Shepherd puppy?"
-
-## Technology Stack
-
-- **FastAPI**: Modern async web framework for the API backend
-- **LangGraph**: Agent workflow orchestration and tool selection
-- **OpenAI GPT-4**: Language model for natural language processing
-- **Qdrant**: Vector database for document embeddings and similarity search
-- **OpenAI Embeddings**: Text embeddings for semantic search
-- **Tavily**: Web search integration for real-time information
-- **WebSocket**: Real-time communication for debug console
-- **Docker**: Containerized deployment for HuggingFace Spaces
-
-## Configuration
-
-The application requires these environment variables:
-
-### Required
-- `OPENAI_API_KEY`: Your OpenAI API key for LLM processing and embeddings
-
-### Optional
-- `TAVILY_API_KEY`: For web search functionality (highly recommended)
-
-**⚠️ Important**: Configure these as secrets in your HuggingFace Space settings:
-1. Go to your space settings
-2. Add the API keys in the "Repository secrets" section
-3. Never commit API keys to your repository
-
-## Knowledge Base
-
-The application uses a preprocessed knowledge base of professional puppy care resources:
-- **File**: `all_books_preprocessed_chunks.json` (2.5MB)
-- **Content**: Chunked and processed expert knowledge about puppy training, health, and behavior
-- **Vector Store**: Automatically indexed in Qdrant on first startup
-
-## API Endpoints
-
-- `GET /`: Main web interface
-- `POST /chat`: Chat API endpoint for programmatic access
-- `WebSocket /ws`: Real-time debug console connection
-- `GET /health`: Application health check
-
-## Development
-
-The application features:
-- **Async Architecture**: Full async/await support for optimal performance  
-- **Persistent Vector Store**: Qdrant storage persists between restarts
-- **Connection Management**: WebSocket connection handling for multiple users
-- **Error Handling**: Comprehensive error handling and logging
-- **Health Monitoring**: Built-in health checks and monitoring
-
----
-
-*Transform the chaos of new puppy ownership into confidence with AI-powered expert assistance!* 🐾# Updated Lun  2 jui 2025 10:00:55 CEST
-# Force rebuild 2025-06-02 10:01:51
+AI-powered puppy care assistant!

deploy_hf.sh.backup

@@ -0,0 +1,263 @@
+#!/bin/bash
+
+# HuggingFace deployment script for PuppyCompanion FastAPI
+# Usage: ./deploy_hf.sh [--space-name SPACE_NAME]
+
+set -e
+
+# Colors
+GREEN='\033[0;32m'
+YELLOW='\033[1;33m'
+RED='\033[0;31m'
+BLUE='\033[0;34m'
+NC='\033[0m'
+
+log() {
+    echo -e "${GREEN}[DEPLOY]${NC} $1"
+}
+
+warn() {
+    echo -e "${YELLOW}[WARN]${NC} $1"
+}
+
+error() {
+    echo -e "${RED}[ERROR]${NC} $1"
+}
+
+info() {
+    echo -e "${BLUE}[INFO]${NC} $1"
+}
+
+# Default configuration
+SPACE_NAME="puppycompanion-v3"
+APP_DIR="."
+HF_USERNAME="JTh34"
+
+# Parse arguments
+while [[ $# -gt 0 ]]; do
+    case $1 in
+        --space-name)
+            SPACE_NAME="$2"
+            shift 2
+            ;;
+        *)
+            error "Unknown option: $1"
+            exit 1
+            ;;
+    esac
+done
+
+# Ask for space name if not provided (with default)
+if [ -z "$SPACE_NAME" ]; then
+    read -p "HuggingFace space name [puppycompanion-v3]: " SPACE_NAME
+    SPACE_NAME=${SPACE_NAME:-puppycompanion-v3}
+fi
+
+if [ -z "$SPACE_NAME" ]; then
+    error "Space name required"
+    exit 1
+fi
+
+# Check that the app directory exists
+if [ ! -d "$APP_DIR" ]; then
+    error "Application directory not found: $APP_DIR"
+    exit 1
+fi
+
+cd "$APP_DIR"
+
+log "Preparing deployment for FastAPI space: $HF_USERNAME/$SPACE_NAME"
+
+# Pre-deployment checks
+log "Pre-deployment checks..."
+
+# Check required files for FastAPI application
+REQUIRED_FILES=(
+    "main.py"
+    "Dockerfile"
+    "README.MD"
+    "all_books_preprocessed_chunks.json"
+    "rag_system.py"
+    "agent_workflow.py"
+    "embedding_models.py"
+    "books_config.json"
+    "static/index.html"
+)
+
+for file in "${REQUIRED_FILES[@]}"; do
+    if [ ! -f "$file" ]; then
+        error "Missing required file: $file"
+        exit 1
+    fi
+done
+
+log "✅ All required files found"
+
+# Check that API keys are not in files
+ENV_FILES=(".env" ".env.local" ".env.prod" ".env.development" ".env.staging")
+for env_file in "${ENV_FILES[@]}"; do
+    if [ -f "$env_file" ]; then
+        if grep -q "sk-\|OPENAI_API_KEY\|TAVILY_API_KEY" "$env_file" 2>/dev/null; then
+            warn "API keys detected in $env_file - file will be excluded from deployment"
+            warn "Make sure to configure your secrets on HuggingFace Spaces"
+        fi
+    fi
+done
+
+# Check for PDF files that should not be deployed
+if ls data/*.pdf >/dev/null 2>&1; then
+    warn "PDF files detected in data/ directory"
+    warn "These will be excluded from deployment for security"
+fi
+
+# Validate chunks file size
+if [ -f "all_books_preprocessed_chunks.json" ]; then
+    FILE_SIZE=$(stat -f%z "all_books_preprocessed_chunks.json" 2>/dev/null || stat -c%s "all_books_preprocessed_chunks.json" 2>/dev/null)
+    FILE_SIZE_MB=$((FILE_SIZE / 1024 / 1024))
+    log "Chunks file size: ${FILE_SIZE_MB}MB"
+    if [ $FILE_SIZE_MB -gt 100 ]; then
+        warn "Large chunks file detected (${FILE_SIZE_MB}MB)"
+        warn "This may cause slower deployment and startup times"
+    fi
+fi
+
+# Create requirements.txt from pyproject.toml if needed and requirements.txt doesn't exist
+if [ ! -f "requirements.txt" ] && [ -f "pyproject.toml" ]; then
+    log "Generating requirements.txt from pyproject.toml..."
+    if command -v pip-compile &> /dev/null; then
+        pip-compile pyproject.toml
+    else
+        warn "pip-compile not found. Basic requirements.txt generation..."
+        # Basic extraction of dependencies from pyproject.toml
+        if command -v python &> /dev/null; then
+            python -c "
+import tomllib
+with open('pyproject.toml', 'rb') as f:
+    data = tomllib.load(f)
+deps = data.get('project', {}).get('dependencies', [])
+with open('requirements.txt', 'w') as f:
+    for dep in deps:
+        f.write(dep + '\n')
+" 2>/dev/null || {
+                warn "Could not generate requirements.txt automatically"
+                warn "Please create the requirements.txt file manually"
+                exit 1
+            }
+        fi
+    fi
+elif [ -f "requirements.txt" ]; then
+    log "✅ requirements.txt found"
+fi
+
+# Validate Dockerfile for FastAPI
+if grep -q "main.py" Dockerfile && grep -q "7860" Dockerfile; then
+    log "✅ Dockerfile configured for FastAPI"
+else
+    warn "Dockerfile may not be properly configured for FastAPI deployment"
+fi
+
+# Check SSH connection to HuggingFace
+log "Checking HuggingFace SSH connection..."
+if ! ssh -T [email protected] -o ConnectTimeout=10 -o BatchMode=yes 2>/dev/null; then
+    warn "SSH connection to HuggingFace failed"
+    info "Check your SSH configuration or use: ssh-keygen -t ed25519 -C '[email protected]'"
+    info "Then add the public key to your HuggingFace profile"
+    
+    # Fallback to HTTPS if available
+    if command -v huggingface-cli &> /dev/null && huggingface-cli whoami &> /dev/null; then
+        warn "Using HTTPS as fallback..."
+        USE_SSH=false
+    else
+        error "No authentication method available"
+        exit 1
+    fi
+else
+    log "✅ HuggingFace SSH connection OK"
+    USE_SSH=true
+fi
+
+# Create the space if it does not exist (requires huggingface-cli)
+if command -v huggingface-cli &> /dev/null; then
+    log "Creating/updating HuggingFace space..."
+    huggingface-cli repo create "$SPACE_NAME" --type space --space_sdk docker 2>/dev/null || true
+else
+    warn "huggingface-cli not available, make sure the space exists"
+fi
+
+# Clone or update the repo
+TEMP_DIR="/tmp/hf_deploy_$$"
+log "Cloning repository..."
+
+if [ "$USE_SSH" = true ]; then
+    # Use SSH
+    git clone "[email protected]:spaces/$HF_USERNAME/$SPACE_NAME" "$TEMP_DIR" || {
+        error "SSH clone failed. Check username and space name."
+        exit 1
+    }
+else
+    # Use HTTPS
+    git clone "https://huggingface.co/spaces/$HF_USERNAME/$SPACE_NAME" "$TEMP_DIR" || {
+        error "HTTPS clone failed. Check username and space name."
+        exit 1
+    }
+fi
+
+# Copy files with FastAPI-specific exclusions
+log "Copying FastAPI application files..."
+rsync -av \
+    --exclude='.git' \
+    --exclude='venv_*' \
+    --exclude='.env*' \
+    --exclude='__pycache__' \
+    --exclude='.chainlit' \
+    --exclude='*.pdf' \
+    --exclude='*.PDF' \
+    --exclude='data/*.pdf' \
+    --exclude='data/*.PDF' \
+    --exclude='.DS_Store' \
+    --exclude='*.log' \
+    --exclude='qdrant_storage' \
+    --exclude='deploy_hf.sh' \
+    --exclude='.gitattributes' \
+    --exclude='uv.lock' \
+    --exclude='document_loader_preproc.py' \
+    ./ "$TEMP_DIR/"
+
+# Go to temp directory
+cd "$TEMP_DIR"
+
+# Configure Git if needed
+git config user.email "[email protected]" || true
+git config user.name "Deploy Script" || true
+
+# Verify critical files are present
+log "Verifying deployment files..."
+CRITICAL_FILES=("main.py" "Dockerfile" "all_books_preprocessed_chunks.json")
+for file in "${CRITICAL_FILES[@]}"; do
+    if [ ! -f "$file" ]; then
+        error "Critical file missing after copy: $file"
+        exit 1
+    fi
+done
+
+log "✅ All critical files verified"
+
+# Add and commit
+log "Committing changes..."
+git add .
+git commit -m "Deploy PuppyCompanion FastAPI $(date '+%Y-%m-%d %H:%M:%S')" || {
+    warn "No changes detected"
+}
+
+# Push to HuggingFace
+log "Pushing to HuggingFace..."
+git push
+
+# Clean up
+cd - > /dev/null
+rm -rf "$TEMP_DIR"
+
+log "🚀 FastAPI deployment completed successfully!"
+info "Your app will be available at: https://huggingface.co/spaces/$HF_USERNAME/$SPACE_NAME"
+info "Deployment may take a few minutes to build and start..."
+info "Check the logs on HuggingFace Spaces for any issues"