Update app.py

app.py

@@ -9,18 +9,14 @@ import uvicorn
 # Configure logging
 logging.basicConfig(
     level=logging.INFO,
-    format='%(asctime)s - %(levelname)s - %(message)s',
-    handlers=[
-        logging.StreamHandler(),
-        logging.FileHandler('app.log')
-    ]
+    format='%(asctime)s - %(levelname)s - %(message)s'
 )
 logger = logging.getLogger(__name__)
 
 # Configuration
 MODEL_ID = "google/gemma-1.1-2b-it"
 HF_TOKEN = os.getenv("HF_TOKEN", "")
-MAX_TOKENS = 400  # For ~200 word scripts
+MAX_TOKENS = 400
 DEVICE = "cpu"
 PORT = int(os.getenv("PORT", 7860))
 
@@ -45,61 +41,23 @@ class ScriptGenerator:
                 low_cpu_mem_usage=True
             ).to(DEVICE)
             
-            # Configure valid generation parameters
             self.generation_config = GenerationConfig(
                 max_new_tokens=MAX_TOKENS,
                 do_sample=True,
-                top_p=0.9,  # Replaces temperature
+                top_p=0.9,
                 num_beams=1,
                 no_repeat_ngram_size=2,
                 pad_token_id=self.tokenizer.eos_token_id
             )
             
             self.loaded = True
-            logger.info("Model and generation config loaded successfully")
+            logger.info("Model loaded | Port: %s", PORT)
         except Exception as e:
-            logger.error(f"Model loading failed: {str(e)}")
+            logger.error("Load failed: %s", str(e))
             raise
 
 generator = ScriptGenerator()
 
-def generate_script(topic: str) -> str:
-    """Generation with validated config"""
-    try:
-        if not generator.loaded:
-            generator.load_model()
-            
-        prompt = (
-            f"Generate a 1-minute (60s) video script about: {topic[:80]}\n"
-            "Required structure with timings:\n"
-            "[0:00-0:10] HOOK: Grab attention\n"
-            "[0:10-0:50] MAIN: 3 key points\n"
-            "[0:50-1:00] CTA: Clear action\n\n"
-            "Script:\n"
-        )
-        
-        inputs = generator.tokenizer(prompt, return_tensors="pt").to(DEVICE)
-        
-        # Use the pre-configured generation config
-        outputs = generator.model.generate(
-            **inputs,
-            generation_config=generator.generation_config
-        )
-        
-        script = generator.tokenizer.decode(outputs[0], skip_special_tokens=True)
-        
-        # Validate output structure
-        required_sections = ["HOOK:", "MAIN:", "CTA:"]
-        if not all(section in script for section in required_sections):
-            logger.warning("Script missing sections, adding template")
-            script += "\n\n[0:00-0:10] HOOK: [Add attention-grabber]\n[0:10-0:50] MAIN: [Add content]\n[0:50-1:00] CTA: [Add action]"
-            
-        return script
-        
-    except Exception as e:
-        logger.error(f"Generation failed: {str(e)}")
-        return f"Error: {str(e)}"
-
 app = FastAPI()
 
 @app.on_event("startup")
@@ -112,38 +70,40 @@ async def predict(request: Request):
         data = await request.json()
         topic = data.get("topic", "")
         
-        # Handle n8n's list input format
         if isinstance(topic, list):
             topic = topic[0] if len(topic) > 0 else ""
         topic = str(topic).strip()
         
-        if not topic:
-            return JSONResponse(
-                {"success": False, "error": "Empty topic"},
-                status_code=400
-            )
+        logger.info("Processing: %.30s...", topic)
+        
+        inputs = generator.tokenizer(
+            f"Create 1-minute script about {topic}:\n1) Hook\n2) Main\n3) CTA\n\nScript:",
+            return_tensors="pt"
+        ).to(DEVICE)
         
-        logger.info(f"Generating script for: {topic[:30]}...")
-        result = generate_script(topic)
+        outputs = generator.model.generate(
+            **inputs,
+            generation_config=generator.generation_config
+        )
         
-        return JSONResponse({
-            "success": not result.startswith("Error"),
-            "result": result,
-            "error": None if not result.startswith("Error") else result
-        })
+        script = generator.tokenizer.decode(outputs[0], skip_special_tokens=True)
+        return JSONResponse({"result": script})
         
     except Exception as e:
-        logger.error(f"API error: {str(e)}")
-        return JSONResponse(
-            {"success": False, "error": str(e)},
-            status_code=500
-        )
+        logger.error("API error: %s", str(e))
+        return JSONResponse({"error": str(e)}, status_code=500)
 
 if __name__ == "__main__":
+    # Hugging Face Spaces compatibility
+    if os.getenv("SPACES", "false").lower() == "true":
+        os.environ["GRADIO_SERVER_PORT"] = str(PORT)
+        os.environ["GRADIO_SERVER_NAME"] = "0.0.0.0"
+    
     uvicorn.run(
         app,
         host="0.0.0.0",
         port=PORT,
         log_level="info",
+        workers=1,
         timeout_keep_alive=30
     )