Simplify app to avoid NumPy compatibility issues - remove voice features and problematic dependencies

app.py

@@ -1,13 +1,12 @@
 import torch
 import gradio as gr
 from diffusers import DiffusionPipeline, DPMSolverMultistepScheduler
-from diffusers.utils import export_to_video
 import numpy as np
 import os
 import logging
-from gtts import gTTS
-from moviepy.editor import VideoFileClip, AudioFileClip, CompositeAudioClip
 import tempfile
+import subprocess
+import json
 
 # Set up logging
 logging.basicConfig(level=logging.INFO)
@@ -20,7 +19,7 @@ class TextToVideoGenerator:
         self.device = "cuda" if torch.cuda.is_available() else "cpu"
         logger.info(f"Using device: {self.device}")
         
-        # Available models - including the advanced Wan2.1 model
+        # Available models - simplified for compatibility
         self.models = {
             "damo-vilab/text-to-video-ms-1.7b": {
                 "name": "DAMO Text-to-Video MS-1.7B",
@@ -37,70 +36,8 @@ class TextToVideoGenerator:
                 "fps": 6,
                 "quality": "Excellent",
                 "speed": "Medium"
-            },
-            "Wan-AI/Wan2.1-T2V-14B": {
-                "name": "Wan2.1-T2V-14B (SOTA)",
-                "description": "State-of-the-art text-to-video model with 14B parameters",
-                "max_frames": 32,
-                "fps": 8,
-                "quality": "SOTA",
-                "speed": "Medium",
-                "resolutions": ["480P", "720P"],
-                "features": ["Chinese & English text", "High motion dynamics", "Best quality"]
             }
         }
-        
-        # Voice options (gTTS only supports language, not gender/age)
-        self.voices = {
-            "Default (English)": "en"
-        }
-    
-    def generate_audio(self, text, voice_type):
-        """Generate audio from text using gTTS"""
-        try:
-            lang = self.voices[voice_type]
-            tts = gTTS(text=text, lang=lang)
-            with tempfile.NamedTemporaryFile(suffix=".mp3", delete=False) as temp_audio:
-                audio_path = temp_audio.name
-            tts.save(audio_path)
-            logger.info(f"Audio generated successfully: {audio_path}")
-            return audio_path
-        except Exception as e:
-            logger.error(f"Error generating audio: {str(e)}")
-            return None
-    
-    def merge_audio_video(self, video_path, audio_path, output_path):
-        """Merge audio and video using moviepy"""
-        try:
-            # Load video and audio
-            video_clip = VideoFileClip(video_path)
-            audio_clip = AudioFileClip(audio_path)
-            
-            # Ensure audio duration matches video duration
-            if audio_clip.duration > video_clip.duration:
-                audio_clip = audio_clip.subclip(0, video_clip.duration)
-            elif audio_clip.duration < video_clip.duration:
-                # Loop audio if it's shorter than video
-                loops_needed = int(video_clip.duration / audio_clip.duration) + 1
-                audio_clip = CompositeAudioClip([audio_clip] * loops_needed).subclip(0, video_clip.duration)
-            
-            # Merge audio and video
-            final_clip = video_clip.set_audio(audio_clip)
-            
-            # Write final video with audio
-            final_clip.write_videofile(output_path, codec='libx264', audio_codec='aac')
-            
-            # Clean up
-            video_clip.close()
-            audio_clip.close()
-            final_clip.close()
-            
-            logger.info(f"Audio and video merged successfully: {output_path}")
-            return output_path
-            
-        except Exception as e:
-            logger.error(f"Error merging audio and video: {str(e)}")
-            return None
     
     def load_model(self, model_id):
         """Load the specified model"""
@@ -114,22 +51,12 @@ class TextToVideoGenerator:
             if torch.cuda.is_available():
                 torch.cuda.empty_cache()
             
-            # Special handling for Wan2.1 model
-            if model_id == "Wan-AI/Wan2.1-T2V-14B":
-                # Wan2.1 requires specific configuration
-                self.pipeline = DiffusionPipeline.from_pretrained(
-                    model_id,
-                    torch_dtype=torch.float16 if self.device == "cuda" else torch.float32,
-                    variant="fp16" if self.device == "cuda" else None,
-                    use_safetensors=True
-                )
-            else:
-                # Standard loading for other models
-                self.pipeline = DiffusionPipeline.from_pretrained(
-                    model_id,
-                    torch_dtype=torch.float16 if self.device == "cuda" else torch.float32,
-                    variant="fp16" if self.device == "cuda" else None
-                )
+            # Standard loading for models
+            self.pipeline = DiffusionPipeline.from_pretrained(
+                model_id,
+                torch_dtype=torch.float16 if self.device == "cuda" else torch.float32,
+                variant="fp16" if self.device == "cuda" else None
+            )
             
             # Move to device
             self.pipeline = self.pipeline.to(self.device)
@@ -153,13 +80,9 @@ class TextToVideoGenerator:
             logger.error(f"Error loading model: {str(e)}")
             return f"Error loading model: {str(e)}"
     
-    def generate_video(self, prompt, model_id, num_frames=16, fps=8, num_inference_steps=25, guidance_scale=7.5, seed=None, resolution="480P", voice_script="", voice_type="Default (English)", add_voice=True):
-        """Generate video from text prompt with optional voice"""
+    def generate_video(self, prompt, model_id, num_frames=16, fps=8, num_inference_steps=25, guidance_scale=7.5, seed=None):
+        """Generate video from text prompt"""
         try:
-            # Use prompt as voice script if voice_script is empty
-            if not voice_script.strip() and add_voice:
-                voice_script = prompt
-            
             # Load model if not already loaded
             if self.current_model != model_id:
                 load_result = self.load_model(model_id)
@@ -177,245 +100,85 @@ class TextToVideoGenerator:
             num_frames = min(num_frames, model_config["max_frames"])
             fps = model_config["fps"]
             
-            # Special handling for Wan2.1 model
-            if model_id == "Wan-AI/Wan2.1-T2V-14B":
-                # Wan2.1 specific parameters
-                if resolution == "720P":
-                    width, height = 1280, 720
-                else:  # 480P
-                    width, height = 832, 480
-                
-                logger.info(f"Generating Wan2.1 video with prompt: {prompt}")
-                logger.info(f"Parameters: frames={num_frames}, fps={fps}, steps={num_inference_steps}, resolution={resolution}")
-                
-                # Generate video with Wan2.1 specific settings
-                result = self.pipeline(
-                    prompt,
-                    num_inference_steps=num_inference_steps,
-                    guidance_scale=guidance_scale,
-                    num_frames=num_frames,
-                    width=width,
-                    height=height
-                )
-                video_frames = result['frames'] if isinstance(result, dict) else result.frames
-            else:
-                # Standard generation for other models
-                logger.info(f"Generating video with prompt: {prompt}")
-                logger.info(f"Parameters: frames={num_frames}, fps={fps}, steps={num_inference_steps}")
-                
-                result = self.pipeline(
-                    prompt,
-                    num_inference_steps=num_inference_steps,
-                    guidance_scale=guidance_scale,
-                    num_frames=num_frames
-                )
-                video_frames = result['frames'] if isinstance(result, dict) else result.frames
+            logger.info(f"Generating video with prompt: {prompt}")
+            logger.info(f"Parameters: frames={num_frames}, fps={fps}, steps={num_inference_steps}")
+            
+            # Generate video
+            result = self.pipeline(
+                prompt,
+                num_inference_steps=num_inference_steps,
+                guidance_scale=guidance_scale,
+                num_frames=num_frames
+            )
             
-            # Convert to numpy array
-            video_frames = np.array(video_frames)
+            # Extract frames
+            if hasattr(result, 'frames'):
+                video_frames = result.frames
+            elif isinstance(result, dict) and 'frames' in result:
+                video_frames = result['frames']
+            else:
+                video_frames = result
             
             # Save video
-            output_path = f"generated_video_{seed if seed else 'random'}.mp4"
+            output_path = tempfile.mktemp(suffix=".mp4")
+            
+            # Use diffusers export_to_video function
+            from diffusers.utils import export_to_video
             export_to_video(video_frames, output_path, fps=fps)
             
-            logger.info(f"Video saved to: {output_path}")
-            
-            # Add voice if requested
-            if add_voice and voice_script.strip():
-                logger.info(f"Generating voice for script: {voice_script}")
-                
-                # Generate audio
-                audio_path = self.generate_audio(voice_script, voice_type)
-                
-                if audio_path:
-                    # Create final output path with voice
-                    final_output_path = f"generated_video_with_voice_{seed if seed else 'random'}.mp4"
-                    
-                    # Merge audio and video
-                    final_path = self.merge_audio_video(output_path, audio_path, final_output_path)
-                    
-                    # Clean up temporary files
-                    try:
-                        os.unlink(audio_path)
-                        os.unlink(output_path)
-                    except:
-                        pass
-                    
-                    if final_path:
-                        return final_path, f"Video with voice generated successfully! Saved as {final_path}"
-                    else:
-                        return output_path, f"Video generated but voice merging failed. Saved as {output_path}"
-                else:
-                    return output_path, f"Video generated but voice generation failed. Saved as {output_path}"
-            else:
-                return output_path, f"Video generated successfully! Saved as {output_path}"
+            logger.info(f"Video generated successfully: {output_path}")
+            return output_path, f"Video generated successfully! Model: {model_config['name']}"
             
         except Exception as e:
             logger.error(f"Error generating video: {str(e)}")
             return None, f"Error generating video: {str(e)}"
     
     def get_available_models(self):
-        """Get list of available models"""
+        """Get list of available model IDs"""
         return list(self.models.keys())
     
     def get_model_info(self, model_id):
-        """Get information about a specific model"""
+        """Get detailed information about a model"""
         if model_id in self.models:
             return self.models[model_id]
-        return None
-    
-    def get_available_voices(self):
-        """Get list of available voices"""
-        return list(self.voices.keys())
-
-# Initialize the generator
-generator = TextToVideoGenerator()
+        return {"error": "Model not found"}
 
 def create_interface():
-    """Create Gradio interface"""
+    """Create the Gradio interface"""
+    generator = TextToVideoGenerator()
     
-    def generate_video_interface(prompt, model_id, num_frames, fps, num_inference_steps, guidance_scale, seed, resolution, voice_script, voice_type, add_voice):
+    def generate_video_interface(prompt, model_id, num_frames, fps, num_inference_steps, guidance_scale, seed):
+        """Interface function for video generation"""
         if not prompt.strip():
-            return None, "Please enter a prompt"
+            return None, "Please enter a video description"
         
-        return generator.generate_video(
+        video_path, status = generator.generate_video(
             prompt=prompt,
             model_id=model_id,
             num_frames=num_frames,
             fps=fps,
             num_inference_steps=num_inference_steps,
             guidance_scale=guidance_scale,
-            seed=seed,
-            resolution=resolution,
-            voice_script=voice_script,
-            voice_type=voice_type,
-            add_voice=add_voice
+            seed=seed
         )
+        
+        return video_path, status
     
-    # Custom CSS for professional styling
+    # Custom CSS for better styling
     custom_css = """
     .gradio-container {
         max-width: 1200px !important;
         margin: 0 auto !important;
     }
-    
-    .header {
-        text-align: center;
-        padding: 2rem 0;
-        background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
-        color: white;
-        border-radius: 15px;
-        margin-bottom: 2rem;
-    }
-    
-    .header h1 {
-        font-size: 2.5rem;
-        font-weight: 700;
-        margin: 0;
-        text-shadow: 2px 2px 4px rgba(0,0,0,0.3);
-    }
-    
-    .header p {
-        font-size: 1.1rem;
-        margin: 0.5rem 0 0 0;
-        opacity: 0.9;
-    }
-    
-    .feature-card {
-        background: white;
-        border-radius: 10px;
-        padding: 1.5rem;
-        box-shadow: 0 4px 6px rgba(0,0,0,0.1);
-        margin-bottom: 1rem;
-        border-left: 4px solid #667eea;
-    }
-    
-    .feature-card h3 {
-        color: #333;
-        margin: 0 0 0.5rem 0;
-        font-size: 1.2rem;
-    }
-    
-    .feature-card p {
-        color: #666;
-        margin: 0;
-        font-size: 0.9rem;
-    }
-    
-    .model-info {
-        background: #f8f9fa;
-        border-radius: 8px;
-        padding: 1rem;
-        border: 1px solid #e9ecef;
-    }
-    
-    .model-info h4 {
-        color: #495057;
-        margin: 0 0 0.5rem 0;
-        font-size: 1rem;
-    }
-    
-    .model-info p {
-        color: #6c757d;
-        margin: 0.25rem 0;
-        font-size: 0.85rem;
-    }
-    
     .generate-btn {
-        background: linear-gradient(135deg, #667eea 0%, #764ba2 100%) !important;
+        background: linear-gradient(45deg, #667eea 0%, #764ba2 100%) !important;
         border: none !important;
         color: white !important;
-        font-weight: 600 !important;
-        padding: 1rem 2rem !important;
-        border-radius: 10px !important;
-        font-size: 1.1rem !important;
-        transition: all 0.3s ease !important;
-    }
-    
-    .generate-btn:hover {
-        transform: translateY(-2px) !important;
-        box-shadow: 0 6px 12px rgba(102, 126, 234, 0.4) !important;
+        font-weight: bold !important;
     }
-    
-    .example-card {
-        background: #f8f9fa;
-        border-radius: 8px;
-        padding: 1rem;
-        margin: 0.5rem 0;
-        border: 1px solid #e9ecef;
-        cursor: pointer;
-        transition: all 0.2s ease;
-    }
-    
-    .example-card:hover {
-        background: #e9ecef;
-        transform: translateX(5px);
-    }
-    
     .status-box {
-        background: #e3f2fd;
-        border: 1px solid #2196f3;
-        border-radius: 8px;
-        padding: 1rem;
-    }
-    
-    .pricing-info {
-        background: linear-gradient(135deg, #ffecd2 0%, #fcb69f 100%);
-        border-radius: 10px;
-        padding: 1rem;
-        text-align: center;
-        margin: 1rem 0;
-    }
-    
-    .pricing-info h4 {
-        color: #d84315;
-        margin: 0 0 0.5rem 0;
-    }
-    
-    .pricing-info p {
-        color: #bf360c;
-        margin: 0;
-        font-size: 0.9rem;
+        background-color: #f8f9fa !important;
+        border: 1px solid #dee2e6 !important;
     }
     """
     
@@ -443,23 +206,13 @@ def create_interface():
                         container=True
                     )
                     
-                    with gr.Row():
-                        model_id = gr.Dropdown(
-                            choices=generator.get_available_models(),
-                            value=generator.get_available_models()[0],
-                            label="🤖 AI Model",
-                            info="Choose the AI model for video generation",
-                            container=True
-                        )
-                        
-                        resolution = gr.Dropdown(
-                            choices=["480P", "720P"],
-                            value="480P",
-                            label="📐 Resolution (Wan2.1 only)",
-                            info="Select video resolution",
-                            visible=False,
-                            container=True
-                        )
+                    model_id = gr.Dropdown(
+                        choices=generator.get_available_models(),
+                        value=generator.get_available_models()[0],
+                        label="🤖 AI Model",
+                        info="Choose the AI model for video generation",
+                        container=True
+                    )
                     
                     with gr.Row():
                         num_frames = gr.Slider(
@@ -506,34 +259,6 @@ def create_interface():
                         container=True
                     )
                 
-                # Voice Section
-                with gr.Group():
-                    gr.Markdown("## 🎤 Voice & Audio")
-                    
-                    with gr.Row():
-                        add_voice = gr.Checkbox(
-                            label="🎵 Add Voice Narration",
-                            value=True,
-                            info="Enable to add professional voice-over"
-                        )
-                        
-                        voice_type = gr.Dropdown(
-                            choices=generator.get_available_voices(),
-                            value="Default (English)",
-                            label="🗣️ Voice Type",
-                            info="Select the voice for narration",
-                            container=True
-                        )
-                    
-                    voice_script = gr.Textbox(
-                        label="📜 Narration Script (Optional)",
-                        placeholder="Enter your narration script here... (Leave blank to use video description)",
-                        lines=2,
-                        max_lines=3,
-                        info="If left blank, the video description will be used as narration",
-                        container=True
-                    )
-                
                 # Generate Button
                 generate_btn = gr.Button("🚀 Generate Professional Video", variant="primary", size="lg")
                 
@@ -549,21 +274,6 @@ def create_interface():
                     gr.Markdown("## 🤖 AI Model Details")
                     model_info = gr.JSON(label="Current Model Specifications")
                 
-                # Pricing Information
-                with gr.Group():
-                    gr.Markdown("## 💰 Pricing")
-                    gr.Markdown("""
-                    **Free Tier:** 5 videos per day
-                    
-                    **Pro Plan:** $9.99/month
-                    - Unlimited videos
-                    - Priority processing
-                    - HD quality
-                    - Advanced features
-                    
-                    **Enterprise:** Contact us
-                    """)
-                
                 # Examples
                 with gr.Group():
                     gr.Markdown("## 💡 Inspiration Examples")
@@ -575,7 +285,6 @@ def create_interface():
                     • A futuristic city with flying cars and neon lights
                     • A butterfly emerging from a cocoon in a garden
                     • A rocket launching into space with fire and smoke
-                    • Two anthropomorphic cats in comfy boxing gear and bright gloves fight intensely on a spotlighted stage
                     • A majestic dragon soaring through a mystical forest with glowing mushrooms
                     """)
                 
@@ -587,13 +296,8 @@ def create_interface():
                     - State-of-the-art video generation
                     - Quality vs speed options
                     
-                    🎤 **Professional Voice-Over**
-                    - Multiple voice types
-                    - Custom narration scripts
-                    
                     🎨 **Advanced Controls**
                     - Quality settings
-                    - Resolution options
                     - Reproducible results
                     
                     ⚡ **Fast Processing**
@@ -604,7 +308,7 @@ def create_interface():
         # Event handlers
         generate_btn.click(
             fn=generate_video_interface,
-            inputs=[prompt, model_id, num_frames, fps, num_inference_steps, guidance_scale, seed, resolution, voice_script, voice_type, add_voice],
+            inputs=[prompt, model_id, num_frames, fps, num_inference_steps, guidance_scale, seed],
             outputs=[video_output, status_text]
         )
         
@@ -613,25 +317,12 @@ def create_interface():
             info = generator.get_model_info(model_id)
             return info
         
-        # Show/hide resolution selector based on model
-        def update_resolution_visibility(model_id):
-            if model_id == "Wan-AI/Wan2.1-T2V-14B":
-                return gr.Dropdown(visible=True)
-            else:
-                return gr.Dropdown(visible=False)
-        
         model_id.change(
             fn=update_model_info,
             inputs=model_id,
             outputs=model_info
         )
         
-        model_id.change(
-            fn=update_resolution_visibility,
-            inputs=model_id,
-            outputs=resolution
-        )
-        
         # Load initial model info
         interface.load(lambda: generator.get_model_info(generator.get_available_models()[0]), outputs=model_info)
     

requirements.txt

@@ -11,6 +11,4 @@ gradio==4.25.0
 huggingface-hub==0.23.0
 xformers==0.0.25
 imageio==2.34.0
-imageio-ffmpeg==0.4.9
-gTTS==2.5.1
-moviepy==1.0.3 
+imageio-ffmpeg==0.4.9 

text-to-video-generator/app.py

@@ -1,13 +1,12 @@
 import torch
 import gradio as gr
 from diffusers import DiffusionPipeline, DPMSolverMultistepScheduler
-from diffusers.utils import export_to_video
 import numpy as np
 import os
 import logging
-from gtts import gTTS
-from moviepy.editor import VideoFileClip, AudioFileClip, CompositeAudioClip
 import tempfile
+import subprocess
+import json
 
 # Set up logging
 logging.basicConfig(level=logging.INFO)
@@ -20,7 +19,7 @@ class TextToVideoGenerator:
         self.device = "cuda" if torch.cuda.is_available() else "cpu"
         logger.info(f"Using device: {self.device}")
         
-        # Available models - including the advanced Wan2.1 model
+        # Available models - simplified for compatibility
         self.models = {
             "damo-vilab/text-to-video-ms-1.7b": {
                 "name": "DAMO Text-to-Video MS-1.7B",
@@ -37,70 +36,8 @@ class TextToVideoGenerator:
                 "fps": 6,
                 "quality": "Excellent",
                 "speed": "Medium"
-            },
-            "Wan-AI/Wan2.1-T2V-14B": {
-                "name": "Wan2.1-T2V-14B (SOTA)",
-                "description": "State-of-the-art text-to-video model with 14B parameters",
-                "max_frames": 32,
-                "fps": 8,
-                "quality": "SOTA",
-                "speed": "Medium",
-                "resolutions": ["480P", "720P"],
-                "features": ["Chinese & English text", "High motion dynamics", "Best quality"]
             }
         }
-        
-        # Voice options (gTTS only supports language, not gender/age)
-        self.voices = {
-            "Default (English)": "en"
-        }
-    
-    def generate_audio(self, text, voice_type):
-        """Generate audio from text using gTTS"""
-        try:
-            lang = self.voices[voice_type]
-            tts = gTTS(text=text, lang=lang)
-            with tempfile.NamedTemporaryFile(suffix=".mp3", delete=False) as temp_audio:
-                audio_path = temp_audio.name
-            tts.save(audio_path)
-            logger.info(f"Audio generated successfully: {audio_path}")
-            return audio_path
-        except Exception as e:
-            logger.error(f"Error generating audio: {str(e)}")
-            return None
-    
-    def merge_audio_video(self, video_path, audio_path, output_path):
-        """Merge audio and video using moviepy"""
-        try:
-            # Load video and audio
-            video_clip = VideoFileClip(video_path)
-            audio_clip = AudioFileClip(audio_path)
-            
-            # Ensure audio duration matches video duration
-            if audio_clip.duration > video_clip.duration:
-                audio_clip = audio_clip.subclip(0, video_clip.duration)
-            elif audio_clip.duration < video_clip.duration:
-                # Loop audio if it's shorter than video
-                loops_needed = int(video_clip.duration / audio_clip.duration) + 1
-                audio_clip = CompositeAudioClip([audio_clip] * loops_needed).subclip(0, video_clip.duration)
-            
-            # Merge audio and video
-            final_clip = video_clip.set_audio(audio_clip)
-            
-            # Write final video with audio
-            final_clip.write_videofile(output_path, codec='libx264', audio_codec='aac')
-            
-            # Clean up
-            video_clip.close()
-            audio_clip.close()
-            final_clip.close()
-            
-            logger.info(f"Audio and video merged successfully: {output_path}")
-            return output_path
-            
-        except Exception as e:
-            logger.error(f"Error merging audio and video: {str(e)}")
-            return None
     
     def load_model(self, model_id):
         """Load the specified model"""
@@ -114,22 +51,12 @@ class TextToVideoGenerator:
             if torch.cuda.is_available():
                 torch.cuda.empty_cache()
             
-            # Special handling for Wan2.1 model
-            if model_id == "Wan-AI/Wan2.1-T2V-14B":
-                # Wan2.1 requires specific configuration
-                self.pipeline = DiffusionPipeline.from_pretrained(
-                    model_id,
-                    torch_dtype=torch.float16 if self.device == "cuda" else torch.float32,
-                    variant="fp16" if self.device == "cuda" else None,
-                    use_safetensors=True
-                )
-            else:
-                # Standard loading for other models
-                self.pipeline = DiffusionPipeline.from_pretrained(
-                    model_id,
-                    torch_dtype=torch.float16 if self.device == "cuda" else torch.float32,
-                    variant="fp16" if self.device == "cuda" else None
-                )
+            # Standard loading for models
+            self.pipeline = DiffusionPipeline.from_pretrained(
+                model_id,
+                torch_dtype=torch.float16 if self.device == "cuda" else torch.float32,
+                variant="fp16" if self.device == "cuda" else None
+            )
             
             # Move to device
             self.pipeline = self.pipeline.to(self.device)
@@ -153,13 +80,9 @@ class TextToVideoGenerator:
             logger.error(f"Error loading model: {str(e)}")
             return f"Error loading model: {str(e)}"
     
-    def generate_video(self, prompt, model_id, num_frames=16, fps=8, num_inference_steps=25, guidance_scale=7.5, seed=None, resolution="480P", voice_script="", voice_type="Default (English)", add_voice=True):
-        """Generate video from text prompt with optional voice"""
+    def generate_video(self, prompt, model_id, num_frames=16, fps=8, num_inference_steps=25, guidance_scale=7.5, seed=None):
+        """Generate video from text prompt"""
         try:
-            # Use prompt as voice script if voice_script is empty
-            if not voice_script.strip() and add_voice:
-                voice_script = prompt
-            
             # Load model if not already loaded
             if self.current_model != model_id:
                 load_result = self.load_model(model_id)
@@ -177,245 +100,85 @@ class TextToVideoGenerator:
             num_frames = min(num_frames, model_config["max_frames"])
             fps = model_config["fps"]
             
-            # Special handling for Wan2.1 model
-            if model_id == "Wan-AI/Wan2.1-T2V-14B":
-                # Wan2.1 specific parameters
-                if resolution == "720P":
-                    width, height = 1280, 720
-                else:  # 480P
-                    width, height = 832, 480
-                
-                logger.info(f"Generating Wan2.1 video with prompt: {prompt}")
-                logger.info(f"Parameters: frames={num_frames}, fps={fps}, steps={num_inference_steps}, resolution={resolution}")
-                
-                # Generate video with Wan2.1 specific settings
-                result = self.pipeline(
-                    prompt,
-                    num_inference_steps=num_inference_steps,
-                    guidance_scale=guidance_scale,
-                    num_frames=num_frames,
-                    width=width,
-                    height=height
-                )
-                video_frames = result['frames'] if isinstance(result, dict) else result.frames
-            else:
-                # Standard generation for other models
-                logger.info(f"Generating video with prompt: {prompt}")
-                logger.info(f"Parameters: frames={num_frames}, fps={fps}, steps={num_inference_steps}")
-                
-                result = self.pipeline(
-                    prompt,
-                    num_inference_steps=num_inference_steps,
-                    guidance_scale=guidance_scale,
-                    num_frames=num_frames
-                )
-                video_frames = result['frames'] if isinstance(result, dict) else result.frames
+            logger.info(f"Generating video with prompt: {prompt}")
+            logger.info(f"Parameters: frames={num_frames}, fps={fps}, steps={num_inference_steps}")
+            
+            # Generate video
+            result = self.pipeline(
+                prompt,
+                num_inference_steps=num_inference_steps,
+                guidance_scale=guidance_scale,
+                num_frames=num_frames
+            )
             
-            # Convert to numpy array
-            video_frames = np.array(video_frames)
+            # Extract frames
+            if hasattr(result, 'frames'):
+                video_frames = result.frames
+            elif isinstance(result, dict) and 'frames' in result:
+                video_frames = result['frames']
+            else:
+                video_frames = result
             
             # Save video
-            output_path = f"generated_video_{seed if seed else 'random'}.mp4"
+            output_path = tempfile.mktemp(suffix=".mp4")
+            
+            # Use diffusers export_to_video function
+            from diffusers.utils import export_to_video
             export_to_video(video_frames, output_path, fps=fps)
             
-            logger.info(f"Video saved to: {output_path}")
-            
-            # Add voice if requested
-            if add_voice and voice_script.strip():
-                logger.info(f"Generating voice for script: {voice_script}")
-                
-                # Generate audio
-                audio_path = self.generate_audio(voice_script, voice_type)
-                
-                if audio_path:
-                    # Create final output path with voice
-                    final_output_path = f"generated_video_with_voice_{seed if seed else 'random'}.mp4"
-                    
-                    # Merge audio and video
-                    final_path = self.merge_audio_video(output_path, audio_path, final_output_path)
-                    
-                    # Clean up temporary files
-                    try:
-                        os.unlink(audio_path)
-                        os.unlink(output_path)
-                    except:
-                        pass
-                    
-                    if final_path:
-                        return final_path, f"Video with voice generated successfully! Saved as {final_path}"
-                    else:
-                        return output_path, f"Video generated but voice merging failed. Saved as {output_path}"
-                else:
-                    return output_path, f"Video generated but voice generation failed. Saved as {output_path}"
-            else:
-                return output_path, f"Video generated successfully! Saved as {output_path}"
+            logger.info(f"Video generated successfully: {output_path}")
+            return output_path, f"Video generated successfully! Model: {model_config['name']}"
             
         except Exception as e:
             logger.error(f"Error generating video: {str(e)}")
             return None, f"Error generating video: {str(e)}"
     
     def get_available_models(self):
-        """Get list of available models"""
+        """Get list of available model IDs"""
         return list(self.models.keys())
     
     def get_model_info(self, model_id):
-        """Get information about a specific model"""
+        """Get detailed information about a model"""
         if model_id in self.models:
             return self.models[model_id]
-        return None
-    
-    def get_available_voices(self):
-        """Get list of available voices"""
-        return list(self.voices.keys())
-
-# Initialize the generator
-generator = TextToVideoGenerator()
+        return {"error": "Model not found"}
 
 def create_interface():
-    """Create Gradio interface"""
+    """Create the Gradio interface"""
+    generator = TextToVideoGenerator()
     
-    def generate_video_interface(prompt, model_id, num_frames, fps, num_inference_steps, guidance_scale, seed, resolution, voice_script, voice_type, add_voice):
+    def generate_video_interface(prompt, model_id, num_frames, fps, num_inference_steps, guidance_scale, seed):
+        """Interface function for video generation"""
         if not prompt.strip():
-            return None, "Please enter a prompt"
+            return None, "Please enter a video description"
         
-        return generator.generate_video(
+        video_path, status = generator.generate_video(
             prompt=prompt,
             model_id=model_id,
             num_frames=num_frames,
             fps=fps,
             num_inference_steps=num_inference_steps,
             guidance_scale=guidance_scale,
-            seed=seed,
-            resolution=resolution,
-            voice_script=voice_script,
-            voice_type=voice_type,
-            add_voice=add_voice
+            seed=seed
         )
+        
+        return video_path, status
     
-    # Custom CSS for professional styling
+    # Custom CSS for better styling
     custom_css = """
     .gradio-container {
         max-width: 1200px !important;
         margin: 0 auto !important;
     }
-    
-    .header {
-        text-align: center;
-        padding: 2rem 0;
-        background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
-        color: white;
-        border-radius: 15px;
-        margin-bottom: 2rem;
-    }
-    
-    .header h1 {
-        font-size: 2.5rem;
-        font-weight: 700;
-        margin: 0;
-        text-shadow: 2px 2px 4px rgba(0,0,0,0.3);
-    }
-    
-    .header p {
-        font-size: 1.1rem;
-        margin: 0.5rem 0 0 0;
-        opacity: 0.9;
-    }
-    
-    .feature-card {
-        background: white;
-        border-radius: 10px;
-        padding: 1.5rem;
-        box-shadow: 0 4px 6px rgba(0,0,0,0.1);
-        margin-bottom: 1rem;
-        border-left: 4px solid #667eea;
-    }
-    
-    .feature-card h3 {
-        color: #333;
-        margin: 0 0 0.5rem 0;
-        font-size: 1.2rem;
-    }
-    
-    .feature-card p {
-        color: #666;
-        margin: 0;
-        font-size: 0.9rem;
-    }
-    
-    .model-info {
-        background: #f8f9fa;
-        border-radius: 8px;
-        padding: 1rem;
-        border: 1px solid #e9ecef;
-    }
-    
-    .model-info h4 {
-        color: #495057;
-        margin: 0 0 0.5rem 0;
-        font-size: 1rem;
-    }
-    
-    .model-info p {
-        color: #6c757d;
-        margin: 0.25rem 0;
-        font-size: 0.85rem;
-    }
-    
     .generate-btn {
-        background: linear-gradient(135deg, #667eea 0%, #764ba2 100%) !important;
+        background: linear-gradient(45deg, #667eea 0%, #764ba2 100%) !important;
         border: none !important;
         color: white !important;
-        font-weight: 600 !important;
-        padding: 1rem 2rem !important;
-        border-radius: 10px !important;
-        font-size: 1.1rem !important;
-        transition: all 0.3s ease !important;
-    }
-    
-    .generate-btn:hover {
-        transform: translateY(-2px) !important;
-        box-shadow: 0 6px 12px rgba(102, 126, 234, 0.4) !important;
+        font-weight: bold !important;
     }
-    
-    .example-card {
-        background: #f8f9fa;
-        border-radius: 8px;
-        padding: 1rem;
-        margin: 0.5rem 0;
-        border: 1px solid #e9ecef;
-        cursor: pointer;
-        transition: all 0.2s ease;
-    }
-    
-    .example-card:hover {
-        background: #e9ecef;
-        transform: translateX(5px);
-    }
-    
     .status-box {
-        background: #e3f2fd;
-        border: 1px solid #2196f3;
-        border-radius: 8px;
-        padding: 1rem;
-    }
-    
-    .pricing-info {
-        background: linear-gradient(135deg, #ffecd2 0%, #fcb69f 100%);
-        border-radius: 10px;
-        padding: 1rem;
-        text-align: center;
-        margin: 1rem 0;
-    }
-    
-    .pricing-info h4 {
-        color: #d84315;
-        margin: 0 0 0.5rem 0;
-    }
-    
-    .pricing-info p {
-        color: #bf360c;
-        margin: 0;
-        font-size: 0.9rem;
+        background-color: #f8f9fa !important;
+        border: 1px solid #dee2e6 !important;
     }
     """
     
@@ -443,23 +206,13 @@ def create_interface():
                         container=True
                     )
                     
-                    with gr.Row():
-                        model_id = gr.Dropdown(
-                            choices=generator.get_available_models(),
-                            value=generator.get_available_models()[0],
-                            label="🤖 AI Model",
-                            info="Choose the AI model for video generation",
-                            container=True
-                        )
-                        
-                        resolution = gr.Dropdown(
-                            choices=["480P", "720P"],
-                            value="480P",
-                            label="📐 Resolution (Wan2.1 only)",
-                            info="Select video resolution",
-                            visible=False,
-                            container=True
-                        )
+                    model_id = gr.Dropdown(
+                        choices=generator.get_available_models(),
+                        value=generator.get_available_models()[0],
+                        label="🤖 AI Model",
+                        info="Choose the AI model for video generation",
+                        container=True
+                    )
                     
                     with gr.Row():
                         num_frames = gr.Slider(
@@ -506,34 +259,6 @@ def create_interface():
                         container=True
                     )
                 
-                # Voice Section
-                with gr.Group():
-                    gr.Markdown("## 🎤 Voice & Audio")
-                    
-                    with gr.Row():
-                        add_voice = gr.Checkbox(
-                            label="🎵 Add Voice Narration",
-                            value=True,
-                            info="Enable to add professional voice-over"
-                        )
-                        
-                        voice_type = gr.Dropdown(
-                            choices=generator.get_available_voices(),
-                            value="Default (English)",
-                            label="🗣️ Voice Type",
-                            info="Select the voice for narration",
-                            container=True
-                        )
-                    
-                    voice_script = gr.Textbox(
-                        label="📜 Narration Script (Optional)",
-                        placeholder="Enter your narration script here... (Leave blank to use video description)",
-                        lines=2,
-                        max_lines=3,
-                        info="If left blank, the video description will be used as narration",
-                        container=True
-                    )
-                
                 # Generate Button
                 generate_btn = gr.Button("🚀 Generate Professional Video", variant="primary", size="lg")
                 
@@ -549,21 +274,6 @@ def create_interface():
                     gr.Markdown("## 🤖 AI Model Details")
                     model_info = gr.JSON(label="Current Model Specifications")
                 
-                # Pricing Information
-                with gr.Group():
-                    gr.Markdown("## 💰 Pricing")
-                    gr.Markdown("""
-                    **Free Tier:** 5 videos per day
-                    
-                    **Pro Plan:** $9.99/month
-                    - Unlimited videos
-                    - Priority processing
-                    - HD quality
-                    - Advanced features
-                    
-                    **Enterprise:** Contact us
-                    """)
-                
                 # Examples
                 with gr.Group():
                     gr.Markdown("## 💡 Inspiration Examples")
@@ -575,7 +285,6 @@ def create_interface():
                     • A futuristic city with flying cars and neon lights
                     • A butterfly emerging from a cocoon in a garden
                     • A rocket launching into space with fire and smoke
-                    • Two anthropomorphic cats in comfy boxing gear and bright gloves fight intensely on a spotlighted stage
                     • A majestic dragon soaring through a mystical forest with glowing mushrooms
                     """)
                 
@@ -587,13 +296,8 @@ def create_interface():
                     - State-of-the-art video generation
                     - Quality vs speed options
                     
-                    🎤 **Professional Voice-Over**
-                    - Multiple voice types
-                    - Custom narration scripts
-                    
                     🎨 **Advanced Controls**
                     - Quality settings
-                    - Resolution options
                     - Reproducible results
                     
                     ⚡ **Fast Processing**
@@ -604,7 +308,7 @@ def create_interface():
         # Event handlers
         generate_btn.click(
             fn=generate_video_interface,
-            inputs=[prompt, model_id, num_frames, fps, num_inference_steps, guidance_scale, seed, resolution, voice_script, voice_type, add_voice],
+            inputs=[prompt, model_id, num_frames, fps, num_inference_steps, guidance_scale, seed],
             outputs=[video_output, status_text]
         )
         
@@ -613,25 +317,12 @@ def create_interface():
             info = generator.get_model_info(model_id)
             return info
         
-        # Show/hide resolution selector based on model
-        def update_resolution_visibility(model_id):
-            if model_id == "Wan-AI/Wan2.1-T2V-14B":
-                return gr.Dropdown(visible=True)
-            else:
-                return gr.Dropdown(visible=False)
-        
         model_id.change(
             fn=update_model_info,
             inputs=model_id,
             outputs=model_info
         )
         
-        model_id.change(
-            fn=update_resolution_visibility,
-            inputs=model_id,
-            outputs=resolution
-        )
-        
         # Load initial model info
         interface.load(lambda: generator.get_model_info(generator.get_available_models()[0]), outputs=model_info)
     

text-to-video-generator/requirements.txt

@@ -11,6 +11,4 @@ gradio==4.25.0
 huggingface-hub==0.23.0
 xformers==0.0.25
 imageio==2.34.0
-imageio-ffmpeg==0.4.9
-gTTS==2.5.1
-moviepy==1.0.3 
+imageio-ffmpeg==0.4.9