app.py

import os
import uuid
import httpx
import torch
import logging
import json
import asyncio
from typing import Dict, Optional
from fastapi import FastAPI, Request, BackgroundTasks, HTTPException, Depends
from fastapi.responses import JSONResponse
from fastapi.security import HTTPBearer, HTTPAuthorizationCredentials
from transformers import AutoTokenizer, AutoModelForCausalLM
import uvicorn
from contextlib import asynccontextmanager

# Configuration - NOW WORKING!
MODEL_ID = "google/gemma-1.1-2b-it"
HF_TOKEN = os.getenv("HF_TOKEN", "")
API_KEY = os.getenv("API_KEY", "default-key-123")
MAX_TOKENS = int(os.getenv("MAX_TOKENS", "450"))
DEVICE = os.getenv("DEVICE", "cpu")
PORT = int(os.getenv("PORT", "7860"))

# Setup logging
logging.basicConfig(
    level=logging.INFO,
    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s'
)
logger = logging.getLogger(__name__)

# Security
security = HTTPBearer()

# Job storage
jobs: Dict[str, dict] = {}

class AIGenerator:
    def __init__(self):
        self.tokenizer = None
        self.model = None
        self.loaded = False
        self.load_error = None
        
    def load_model(self):
        """Load the AI model with authentication"""
        if self.loaded:
            return True
            
        logger.info(f"🚀 Loading model: {MODEL_ID}")
        
        if not HF_TOKEN:
            logger.error("❌ HF_TOKEN is not set!")
            self.load_error = "HF_TOKEN environment variable is not set"
            return False
            
        try:
            # Load tokenizer with authentication
            logger.info("📥 Loading tokenizer...")
            self.tokenizer = AutoTokenizer.from_pretrained(
                MODEL_ID,
                token=HF_TOKEN  # Key change: use 'token' parameter
            )
            
            # Set padding token
            if self.tokenizer.pad_token is None:
                self.tokenizer.pad_token = self.tokenizer.eos_token
                
            logger.info("✅ Tokenizer loaded")
            
            # Load model with authentication
            logger.info("📥 Loading model...")
            self.model = AutoModelForCausalLM.from_pretrained(
                MODEL_ID,
                torch_dtype=torch.float32,
                token=HF_TOKEN,  # Key change: use 'token' parameter
                device_map=None
            )
            
            # Move to device
            self.model = self.model.to(DEVICE)
            self.model.eval()
            
            self.loaded = True
            logger.info("🎉 Model loaded successfully!")
            return True
            
        except Exception as e:
            self.load_error = str(e)
            logger.error(f"❌ Model loading failed: {str(e)}")
            return False

# Global generator instance
generator = AIGenerator()

async def verify_api_key(credentials: HTTPAuthorizationCredentials = Depends(security)):
    """Verify API key"""
    if credentials.credentials != API_KEY:
        raise HTTPException(status_code=401, detail="Invalid API key")
    return True

@asynccontextmanager
async def lifespan(app: FastAPI):
    """Lifespan manager - preload model on startup"""
    logger.info("🚀 Starting AI API Server...")
    logger.info(f"📊 Config: Model={MODEL_ID}, Device={DEVICE}, MaxTokens={MAX_TOKENS}")
    
    # Try to preload model (non-blocking)
    try:
        generator.load_model()
    except Exception as e:
        logger.warning(f"Model preloading failed, will load on first request: {e}")
    
    yield

app = FastAPI(lifespan=lifespan)

def generate_text(prompt: str, max_tokens: int = None) -> str:
    """Generate text based on prompt"""
    try:
        if not generator.loaded:
            if not generator.load_model():
                raise Exception(f"Model failed to load: {generator.load_error}")
        
        logger.info(f"📝 Generating text for prompt: '{prompt[:50]}...'")
        
        # Tokenize
        inputs = generator.tokenizer(
            prompt, 
            return_tensors="pt", 
            truncation=True,
            max_length=512
        )
        
        inputs = {k: v.to(DEVICE) for k, v in inputs.items()}
        
        # Generate
        with torch.no_grad():
            outputs = generator.model.generate(
                **inputs,
                max_new_tokens=max_tokens or MAX_TOKENS,
                do_sample=True,
                top_p=0.9,
                temperature=0.8,
                pad_token_id=generator.tokenizer.pad_token_id,
                repetition_penalty=1.1
            )
        
        # Decode
        generated_text = generator.tokenizer.decode(outputs[0], skip_special_tokens=True)
        
        # Remove prompt if included
        if prompt in generated_text:
            generated_text = generated_text.replace(prompt, "").strip()
        
        logger.info(f"✅ Generated {len(generated_text)} characters")
        return generated_text
        
    except Exception as e:
        logger.error(f"❌ Generation failed: {str(e)}")
        raise

@app.post("/api/generate-sync")
async def generate_sync(
    request: Request,
    auth: bool = Depends(verify_api_key)
):
    """
    Synchronous text generation
    Body: {"prompt": "your text", "max_tokens": 100}
    """
    try:
        data = await request.json()
        
        if not data.get("prompt"):
            raise HTTPException(status_code=400, detail="Prompt is required")
            
        prompt = data["prompt"]
        max_tokens = data.get("max_tokens")
        
        logger.info(f"📥 Sync request: '{prompt[:50]}...'")
        
        generated_text = generate_text(prompt, max_tokens)
        
        return JSONResponse({
            "status": "success",
            "result": generated_text,
            "prompt": prompt,
            "text_length": len(generated_text),
            "model": MODEL_ID
        })
        
    except Exception as e:
        logger.error(f"❌ Sync generation error: {str(e)}")
        raise HTTPException(status_code=500, detail=str(e))

@app.post("/api/generate")
async def generate_async(
    request: Request,
    background_tasks: BackgroundTasks,
    auth: bool = Depends(verify_api_key)
):
    """
    Asynchronous text generation (for longer tasks)
    Body: {"prompt": "your text", "max_tokens": 100, "callback_url": "optional"}
    """
    try:
        data = await request.json()
        job_id = str(uuid.uuid4())
        
        if not data.get("prompt"):
            raise HTTPException(status_code=400, detail="Prompt is required")
            
        prompt = data["prompt"]
        max_tokens = data.get("max_tokens")
        callback_url = data.get("callback_url")
        
        logger.info(f"📥 Async request {job_id}")
        
        jobs[job_id] = {
            "status": "processing",
            "prompt": prompt
        }
        
        # Process in background
        background_tasks.add_task(
            process_job_async,
            job_id,
            prompt,
            max_tokens,
            callback_url
        )
        
        return JSONResponse({
            "job_id": job_id,
            "status": "queued",
            "message": "Generation started",
            "model": MODEL_ID
        })
        
    except Exception as e:
        logger.error(f"❌ Async request error: {str(e)}")
        raise HTTPException(status_code=400, detail=str(e))

async def process_job_async(job_id: str, prompt: str, max_tokens: int = None, callback_url: str = None):
    """Background processing for async jobs"""
    try:
        logger.info(f"🔄 Processing async job {job_id}")
        
        generated_text = generate_text(prompt, max_tokens)
        
        jobs[job_id] = {
            "status": "complete",
            "result": generated_text,
            "prompt": prompt,
            "text_length": len(generated_text)
        }
        
        logger.info(f"✅ Completed async job {job_id}")
        
        # Send callback if provided
        if callback_url:
            try:
                async with httpx.AsyncClient(timeout=30.0) as client:
                    await client.post(
                        callback_url,
                        json={
                            "job_id": job_id,
                            "status": "complete",
                            "result": generated_text,
                            "prompt": prompt
                        }
                    )
            except Exception as e:
                logger.error(f"❌ Callback failed: {e}")
                
    except Exception as e:
        error_msg = str(e)
        logger.error(f"❌ Async job {job_id} failed: {error_msg}")
        jobs[job_id] = {
            "status": "failed",
            "error": error_msg,
            "prompt": prompt
        }

@app.get("/api/status/{job_id}")
async def get_status(job_id: str, auth: bool = Depends(verify_api_key)):
    """Check job status"""
    if job_id not in jobs:
        raise HTTPException(status_code=404, detail="Job not found")
    
    return JSONResponse(jobs[job_id])

@app.get("/health")
async def health_check():
    """Health check endpoint"""
    return JSONResponse({
        "status": "healthy",
        "model_loaded": generator.loaded,
        "model": MODEL_ID,
        "device": DEVICE,
        "max_tokens": MAX_TOKENS
    })

@app.get("/model-info")
async def model_info():
    """Model information"""
    return JSONResponse({
        "model": MODEL_ID,
        "loaded": generator.loaded,
        "error": generator.load_error,
        "device": DEVICE,
        "requires_auth": True,
        "token_available": bool(HF_TOKEN)
    })

@app.get("/")
async def root():
    """Root endpoint"""
    return JSONResponse({
        "message": "🤖 AI Text Generation API",
        "version": "1.0",
        "model": MODEL_ID,
        "status": "operational" if generator.loaded else "model_loading",
        "endpoints": {
            "generate_sync": "POST /api/generate-sync",
            "generate_async": "POST /api/generate",
            "check_status": "GET /api/status/{job_id}",
            "health": "GET /health",
            "model_info": "GET /model-info"
        },
        "usage": 'curl -X POST /api/generate-sync -H "Authorization: Bearer YOUR_KEY" -d \'{"prompt":"Hello"}\''
    })

if __name__ == "__main__":
    uvicorn.run(
        app,
        host="0.0.0.0",
        port=PORT,
        log_level="info"
    )