Update app.py

app.py

@@ -9,17 +9,14 @@ import os
 from datetime import datetime
 import time
 import re
-import tempfile
 
-# Try to import optional dependencies
 try:
     import google.generativeai as genai
     GEMINI_AVAILABLE = True
 except ImportError:
     GEMINI_AVAILABLE = False
-    print("Gemini AI not available - continuing without prompt enhancement")
+    print("Gemini AI not available")
 
-# Configure Gemini API if available
 GEMINI_API_KEY = os.getenv("GEMINI_API_KEY")
 if GEMINI_AVAILABLE and GEMINI_API_KEY:
     genai.configure(api_key=GEMINI_API_KEY)
@@ -29,27 +26,21 @@ if GEMINI_AVAILABLE and GEMINI_API_KEY:
         print(f"Error initializing Gemini: {e}")
         GEMINI_AVAILABLE = False
 
-# Hugging Face token
 HF_TOKEN = os.getenv("HUGGINGFACE_TOKEN") or os.getenv("HF_TOKEN")
 
-# Alternative text generation models to try
 TEXT_MODELS = [
     "microsoft/DialoGPT-medium",
     "gpt2",
-    "facebook/blenderbot-400M-distill",
-    "microsoft/DialoGPT-small"
+    "facebook/blenderbot-400M-distill"
 ]
 
-# Alternative image generation models to try
 IMAGE_MODELS = [
     "stabilityai/stable-diffusion-2-1",
     "runwayml/stable-diffusion-v1-5",
-    "CompVis/stable-diffusion-v1-4",
-    "stabilityai/stable-diffusion-2-1-base"
+    "CompVis/stable-diffusion-v1-4"
 ]
 
 def query_huggingface_text(payload, model_name):
-    """Query Hugging Face text generation API with better error handling"""
     API_URL = f"https://api-inference.huggingface.co/models/{model_name}"
     headers = {}
     
@@ -63,21 +54,17 @@ def query_huggingface_text(payload, model_name):
             result = response.json()
             return result
         elif response.status_code == 503:
-            print(f"Model {model_name} is loading, trying next model...")
+            print(f"Model {model_name} is loading")
             return None
         else:
-            print(f"Error {response.status_code} with model {model_name}: {response.text}")
+            print(f"Error {response.status_code} with model {model_name}")
             return None
             
-    except requests.exceptions.Timeout:
-        print(f"Timeout with model {model_name}")
-        return None
     except Exception as e:
         print(f"Error with model {model_name}: {str(e)}")
         return None
 
 def query_huggingface_image(payload, model_name):
-    """Query Hugging Face image generation API with better error handling"""
     API_URL = f"https://api-inference.huggingface.co/models/{model_name}"
     headers = {}
     
@@ -89,71 +76,47 @@ def query_huggingface_image(payload, model_name):
         
         if response.status_code == 200:
             return response.content
-        elif response.status_code == 503:
-            print(f"Image model {model_name} is loading, trying next model...")
-            return None
         else:
-            print(f"Error {response.status_code} with image model {model_name}")
+            print(f"Error with image model {model_name}")
             return None
             
-    except requests.exceptions.Timeout:
-        print(f"Timeout with image model {model_name}")
-        return None
     except Exception as e:
         print(f"Error with image model {model_name}: {str(e)}")
         return None
 
 def transcribe_audio(audio_file):
-    """Convert speech to text using speech recognition with better error handling"""
     if audio_file is None:
         return "No audio file provided"
     
     recognizer = sr.Recognizer()
     
     try:
-        # Handle different audio file types
         audio_path = str(audio_file)
         
-        # Load and process audio file
         with sr.AudioFile(audio_path) as source:
-            # Adjust for ambient noise if possible
-            try:
-                recognizer.adjust_for_ambient_noise(source, duration=0.2)
-            except:
-                pass  # Skip if adjustment fails
             audio = recognizer.record(source)
         
-        # Try Google Speech Recognition (free tier)
         try:
-            text = recognizer.recognize_google(audio, language='en-US')
-            if text.strip():
-                return text
-            else:
-                return "No speech detected in the audio"
+            text = recognizer.recognize_google(audio)
+            return text
         except sr.UnknownValueError:
-            return "Could not understand the audio. Please speak more clearly and try again."
+            return "Could not understand the audio"
         except sr.RequestError as e:
-            return f"Speech recognition service temporarily unavailable: {str(e)}"
+            return f"Speech recognition error: {str(e)}"
         
     except Exception as e:
-        return f"Error processing audio file: {str(e)}. Please check your audio format."
+        return f"Error processing audio: {str(e)}"
 
 def enhance_prompt_with_gemini(text):
-    """Enhance the prompt using Gemini API for better results"""
     if not (GEMINI_AVAILABLE and GEMINI_API_KEY):
         return text, text
     
     try:
         prompt = f"""
-        Enhance this prompt for AI content and image generation. Make it more detailed and creative while keeping the original intent:
-        
+        Enhance this prompt for content and image generation:
         Original: {text}
         
-        Please provide:
-        1. An enhanced text generation prompt
-        2. An enhanced image generation prompt
-        
-        Format your response as:
+        Provide:
         TEXT: [enhanced text prompt]
         IMAGE: [enhanced image prompt]
         """
@@ -161,7 +124,6 @@ def enhance_prompt_with_gemini(text):
         response = gemini_model.generate_content(prompt)
         enhanced = response.text
         
-        # Parse the response
         text_match = re.search(r'TEXT:\s*(.+?)(?=IMAGE:|$)', enhanced, re.DOTALL)
         image_match = re.search(r'IMAGE:\s*(.+?)$', enhanced, re.DOTALL)
         
@@ -170,37 +132,29 @@ def enhance_prompt_with_gemini(text):
         
         return enhanced_text, enhanced_image
     except Exception as e:
-        print(f"Gemini enhancement error: {str(e)}")
+        print(f"Gemini error: {str(e)}")
         return text, text
 
 def generate_text_content(prompt, content_type="blog"):
-    """Generate text content using Hugging Face models"""
-    
-    # Enhance prompt with Gemini if available
     if GEMINI_AVAILABLE and GEMINI_API_KEY:
         enhanced_text, _ = enhance_prompt_with_gemini(prompt)
         prompt = enhanced_text
     
-    # Adjust prompt based on content type
     content_templates = {
-        "blog": f"Write a detailed blog post about: {prompt}\n\nBlog post:",
-        "social": f"Write an engaging social media post about: {prompt}\n\nPost:",
-        "caption": f"Write a creative caption for: {prompt}\n\nCaption:",
-        "story": f"Write a short story about: {prompt}\n\nStory:"
+        "blog": f"Write a blog post about: {prompt}\n\nPost:",
+        "social": f"Write a social media post about: {prompt}\n\nPost:",
+        "caption": f"Write a caption for: {prompt}\n\nCaption:",
+        "story": f"Write a story about: {prompt}\n\nStory:"
     }
     
     full_prompt = content_templates.get(content_type, prompt)
     
-    # Try different models until one works
     for model in TEXT_MODELS:
         payload = {
             "inputs": full_prompt,
             "parameters": {
                 "max_length": 200,
-                "temperature": 0.7,
-                "do_sample": True,
-                "top_p": 0.9,
-                "repetition_penalty": 1.1
+                "temperature": 0.7
             }
         }
         
@@ -215,7 +169,6 @@ def generate_text_content(prompt, content_type="blog"):
                 else:
                     continue
                 
-                # Clean up the response
                 if generated_text and generated_text.startswith(full_prompt):
                     generated_text = generated_text[len(full_prompt):].strip()
                 
@@ -223,76 +176,57 @@ def generate_text_content(prompt, content_type="blog"):
                     return generated_text
                     
             except Exception as e:
-                print(f"Error processing result from {model}: {e}")
+                print(f"Error processing result: {e}")
                 continue
     
-    # Fallback content if all models fail
     fallback_content = {
-        "blog": f"# {prompt}\n\nThis is an interesting topic that deserves exploration. Here are some key points to consider:\n\n• The fundamental concepts and principles\n• Practical applications and use cases\n• Benefits and potential challenges\n• Future developments and trends\n\nThis topic offers many opportunities for further discussion and research.",
-        "social": f"🌟 Excited to share thoughts on {prompt}! This is such an important topic that deserves more attention. What are your thoughts? #AI #Innovation",
-        "caption": f"✨ {prompt} ✨ Sometimes the most beautiful moments come from the simplest ideas. 📸 #inspiration #creativity",
-        "story": f"Once upon a time, there was something special about {prompt}. It captured the imagination of everyone who encountered it, leading to unexpected adventures and new discoveries. The end was just the beginning of something even more wonderful."
+        "blog": f"# About {prompt}\n\nThis is an interesting topic with many aspects to explore. Here are key points:\n\n• Main concepts and principles\n• Practical applications\n• Future possibilities\n\nThis topic offers great potential for discussion.",
+        "social": f"Excited to share thoughts about {prompt}! This is such an important topic. What are your thoughts? #inspiration",
+        "caption": f"✨ {prompt} ✨ Beautiful moments from simple ideas. #creativity #inspiration",
+        "story": f"There was something special about {prompt}. It captured everyone's imagination, leading to wonderful adventures and discoveries."
     }
     
-    return fallback_content.get(content_type, f"Content generated for: {prompt}")
+    return fallback_content.get(content_type, f"Content about: {prompt}")
 
 def generate_image_from_text(prompt):
-    """Generate image using Hugging Face Stable Diffusion models"""
-    
-    # Enhance prompt with Gemini if available
     if GEMINI_AVAILABLE and GEMINI_API_KEY:
         _, enhanced_image = enhance_prompt_with_gemini(prompt)
         prompt = enhanced_image
     
-    # Add some style enhancements to the prompt
-    enhanced_prompt = f"{prompt}, high quality, detailed, artistic, professional, masterpiece"
+    enhanced_prompt = f"{prompt}, high quality, detailed, artistic"
     
-    # Try different image models until one works
     for model in IMAGE_MODELS:
-        payload = {
-            "inputs": enhanced_prompt,
-            "parameters": {
-                "num_inference_steps": 20,
-                "guidance_scale": 7.5
-            }
-        }
+        payload = {"inputs": enhanced_prompt}
         
         image_bytes = query_huggingface_image(payload, model)
         
         if image_bytes:
             try:
                 image = Image.open(io.BytesIO(image_bytes))
-                # Ensure image is in RGB mode
                 if image.mode != 'RGB':
                     image = image.convert('RGB')
                 return image
             except Exception as e:
-                print(f"Error opening image from {model}: {str(e)}")
+                print(f"Error opening image: {str(e)}")
                 continue
     
-    # Return a placeholder image if all models fail
     placeholder = Image.new('RGB', (512, 512), color='lightblue')
     return placeholder
 
 def process_voice_input(audio_file, content_type):
-    """Main function to process voice input and generate content"""
-    
     if audio_file is None:
         return "Please record some audio first", None, ""
     
-    # Transcribe audio
     transcribed_text = transcribe_audio(audio_file)
     
     if transcribed_text.startswith("Error") or transcribed_text.startswith("Could not"):
         return transcribed_text, None, transcribed_text
     
-    # Generate text content
     try:
         text_content = generate_text_content(transcribed_text, content_type)
     except Exception as e:
         text_content = f"Error generating text: {str(e)}"
     
-    # Generate image
     try:
         image = generate_image_from_text(transcribed_text)
     except Exception as e:
@@ -302,18 +236,14 @@ def process_voice_input(audio_file, content_type):
     return text_content, image, transcribed_text
 
 def process_text_input(text_input, content_type):
-    """Process direct text input"""
-    
     if not text_input.strip():
         return "Please enter some text", None
     
-    # Generate text content
     try:
         text_content = generate_text_content(text_input, content_type)
     except Exception as e:
         text_content = f"Error generating text: {str(e)}"
     
-    # Generate image
     try:
         image = generate_image_from_text(text_input)
     except Exception as e:
@@ -323,61 +253,36 @@ def process_text_input(text_input, content_type):
     return text_content, image
 
 def create_interface():
-    """Create the main Gradio interface optimized for Hugging Face Spaces"""
-    
-    # Custom CSS for better appearance
-    custom_css = """
-    .gradio-container {
-        max-width: 1200px !important;
-    }
-    .main-header {
-        text-align: center;
-        background: linear-gradient(45deg, #FF6B6B, #4ECDC4);
-        -webkit-background-clip: text;
-        -webkit-text-fill-color: transparent;
-        font-size: 2.5em;
-        font-weight: bold;
-        margin-bottom: 20px;
-    }
-    """
-    
-    with gr.Blocks(title="VociArt - Voice AI Creator", theme=gr.themes.Soft(), css=custom_css) as app:
-        
-        gr.HTML("""
-        <div class="main-header">
-            🎙️ VociArt - Voice AI Creator
-        </div>
-        """)
+    with gr.Blocks(title="VociArt - Voice AI Creator", theme=gr.themes.Soft()) as app:
         
         gr.Markdown("""
-        Transform your voice into AI-generated content and stunning visuals! 🚀
+        # 🎙️ VociArt - Voice AI Creator
+        
+        Transform your voice into AI-generated content and images!
         
-        **✨ Features:** Voice-to-text • AI content generation • Image creation • Multiple content types
+        **Features:** Voice-to-text • Content generation • Image creation
         """)
         
         with gr.Tab("🎙️ Voice Input"):
             with gr.Row():
-                with gr.Column(scale=1):
+                with gr.Column():
                     audio_input = gr.Audio(
                         sources=["microphone"],
                         type="filepath",
-                        label="🎤 Record Your Voice",
-                        show_download_button=False
+                        label="🎤 Record Your Voice"
                     )
                     
                     content_type = gr.Dropdown(
                         choices=["blog", "social", "caption", "story"],
                         value="blog",
-                        label="📝 Content Type",
-                        info="Choose the type of content to generate"
+                        label="📝 Content Type"
                     )
                     
-                    voice_submit_btn = gr.Button("🚀 Generate from Voice", variant="primary", size="lg")
+                    voice_submit_btn = gr.Button("🚀 Generate from Voice", variant="primary")
                 
-                with gr.Column(scale=1):
+                with gr.Column():
                     transcribed_output = gr.Textbox(
                         label="📝 What You Said",
-                        placeholder="Your transcribed speech will appear here...",
                         lines=3
                     )
             
@@ -385,16 +290,13 @@ def create_interface():
                 with gr.Column():
                     text_output = gr.Textbox(
                         label="📄 Generated Content",
-                        lines=8,
-                        placeholder="AI-generated content will appear here...",
-                        show_copy_button=True
+                        lines=8
                     )
                 
                 with gr.Column():
                     image_output = gr.Image(
                         label="🎨 Generated Image",
-                        type="pil",
-                        show_download_button=True
+                        type="pil"
                     )
         
         with gr.Tab("⌨️ Text Input"):
@@ -402,7 +304,6 @@ def create_interface():
                 with gr.Column():
                     text_input = gr.Textbox(
                         label="💭 Enter Your Idea",
-                        placeholder="Type your creative idea here...",
                         lines=3
                     )
                     
@@ -412,101 +313,63 @@ def create_interface():
                         label="📝 Content Type"
                     )
                     
-                    text_submit_btn = gr.Button("🚀 Generate from Text", variant="primary", size="lg")
+                    text_submit_btn = gr.Button("🚀 Generate from Text", variant="primary")
             
             with gr.Row():
                 with gr.Column():
                     text_output_2 = gr.Textbox(
                         label="📄 Generated Content",
-                        lines=8,
-                        placeholder="AI-generated content will appear here...",
-                        show_copy_button=True
+                        lines=8
                     )
                 
                 with gr.Column():
                     image_output_2 = gr.Image(
                         label="🎨 Generated Image",
-                        type="pil",
-                        show_download_button=True
+                        type="pil"
                     )
         
-        with gr.Tab("ℹ️ About & Tips"):
+        with gr.Tab("ℹ️ About"):
             gr.Markdown("""
-            ## 🌟 About VociArt
-            
-            VociArt transforms your spoken ideas into professional content and stunning visuals using cutting-edge AI technology.
+            ## About VociArt
             
-            ### 🎯 How to Use:
-            1. **Voice Tab**: Click the microphone, speak your idea clearly, select content type, then click generate
-            2. **Text Tab**: Type your idea directly, choose content type, and generate
+            Transform spoken ideas into content and visuals using AI!
             
-            ### 📝 Content Types:
-            - **📰 Blog**: Detailed articles and posts
-            - **📱 Social**: Engaging social media content  
-            - **📸 Caption**: Creative image captions
-            - **📚 Story**: Short narratives and tales
+            ### How to Use:
+            1. **Voice**: Record your idea, select content type, generate
+            2. **Text**: Type your idea, choose type, generate
             
-            ### 💡 Pro Tips:
-            - **Speak Clearly**: Use a quiet environment and speak at normal pace
-            - **Be Specific**: Detailed prompts create better results
-            - **Try Different Types**: Each content type has unique characteristics
-            - **Use Keywords**: Include relevant terms for better image generation
+            ### Content Types:
+            - **Blog**: Articles and posts
+            - **Social**: Social media content  
+            - **Caption**: Image captions
+            - **Story**: Short stories
             
-            ### 🔧 Technical Features:
-            - **Free AI Models**: Powered by Hugging Face's free inference API
-            - **Speech Recognition**: Google Speech Recognition for transcription
-            - **Smart Fallbacks**: Multiple models ensure reliability
-            - **Gemini Enhancement**: Optional prompt improvement (if API key provided)
+            ### Tips:
+            - Speak clearly in a quiet environment
+            - Be specific with your ideas
+            - Try different content types
             
-            ### 🎨 Example Prompts:
-            - *"A futuristic city with flying cars at sunset"*
-            - *"Write about the benefits of morning meditation"*
-            - *"Create a social media post about healthy cooking"*
-            - *"A magical forest with glowing mushrooms"*
-            
-            ---
-            💝 **Made with love using free AI models** - Perfect for creators, marketers, and storytellers!
+            Made with free AI models from Hugging Face!
             """)
         
-        # Event handlers with better error handling
         voice_submit_btn.click(
             fn=process_voice_input,
             inputs=[audio_input, content_type],
-            outputs=[text_output, image_output, transcribed_output],
-            api_name="voice_generate"
+            outputs=[text_output, image_output, transcribed_output]
         )
         
         text_submit_btn.click(
             fn=process_text_input,
             inputs=[text_input, text_content_type],
-            outputs=[text_output_2, image_output_2],
-            api_name="text_generate"
-        )
-        
-        # Add examples
-        gr.Examples(
-            examples=[
-                ["A peaceful mountain landscape with a lake", "caption"],
-                ["The future of artificial intelligence in education", "blog"],
-                ["Delicious homemade pizza recipe", "social"],
-                ["A brave knight on a quest for the golden crown", "story"]
-            ],
-            inputs=[text_input, text_content_type],
-            outputs=[text_output_2, image_output_2],
-            fn=process_text_input,
-            cache_examples=False
+            outputs=[text_output_2, image_output_2]
         )
     
     return app
 
-# Launch the application
 if __name__ == "__main__":
-    print("🚀 Starting VociArt...")
+    print("Starting VociArt...")
     app = create_interface()
     app.launch(
         server_name="0.0.0.0",
-        server_port=7860,
-        share=False,  # Set to False for Hugging Face Spaces
-        show_error=True,
-        quiet=False
+        server_port=7860
     )