# FILE: app_complete.py # DESCRIPTION: Gradio web interface for the LTX-Video generation service. # Provides a user-friendly, step-by-step workflow for creating videos. import gradio as gr import traceback import sys # ============================================================================== # --- BACKEND SERVICES IMPORT --- # ============================================================================== # Encapsulate imports in a try-except block for robust error handling at startup. try: # This assumes the backend file is named 'ltx_server_refactored_complete.py' # and is in a reachable path (e.g., 'api/'). from api.ltx_server_refactored_complete import video_generation_service # Placeholder for SeedVR server. # from api.seedvr_server import SeedVRServer # seedvr_inference_server = SeedVRServer() seedvr_inference_server = None print("Backend services imported successfully.") except ImportError as e: print(f"FATAL ERROR: Could not import backend services. Ensure the backend file is accessible. Details: {e}") sys.exit(1) except Exception as e: print(f"FATAL ERROR: An unexpected error occurred during backend initialization. Details: {e}") sys.exit(1) # ============================================================================== # --- UI WRAPPER FUNCTIONS --- # These functions act as a bridge between the Gradio UI and the backend service. # They handle data conversion, error catching, and UI updates. # ============================================================================== def run_generate_base_video( generation_mode: str, prompt: str, neg_prompt: str, start_img: str, height: int, width: int, duration: float, seed: int, randomize_seed: bool, fp_guidance_preset: str, fp_guidance_scale_list: str, fp_stg_scale_list: str, progress=gr.Progress(track_tqdm=True) ) -> tuple: """ Wrapper to call the backend for generating the initial low-resolution video. It decides whether to use the 'narrative' or 'single' generation mode. """ try: print(f"[UI] Request received for base video generation. Mode: {generation_mode}") initial_conditions = [] if start_img: # Estimate total frames for conditioning context num_frames_estimate = int(duration * 24) items_list = [[start_img, 0, 1.0]] # [[media, frame, weight]] initial_conditions = video_generation_service.prepare_condition_items( items_list, height, width, num_frames_estimate ) # Package advanced LTX settings for the backend ltx_configs = { "guidance_preset": fp_guidance_preset, "guidance_scale_list": fp_guidance_scale_list, "stg_scale_list": fp_stg_scale_list, } # Select the appropriate backend function based on UI mode if generation_mode == "Narrativa (Múltiplos Prompts)": func_to_call = video_generation_service.generate_narrative_low else: func_to_call = video_generation_service.generate_single_low video_path, tensor_path, final_seed = func_to_call( prompt=prompt, negative_prompt=neg_prompt, height=height, width=width, duration=duration, seed=None if randomize_seed else int(seed), initial_conditions=initial_conditions, ltx_configs_override=ltx_configs ) if not video_path: raise RuntimeError("Backend failed to return a valid video path.") # Update the session state with the results new_state = {"low_res_video": video_path, "low_res_latents": tensor_path, "used_seed": final_seed} print(f"[UI] Base video generation successful. Path: {video_path}") return video_path, new_state, gr.update(visible=True) except Exception as e: error_message = f"❌ An error occurred during base generation:\n{e}" print(f"{error_message}\nDetails: {traceback.format_exc()}") raise gr.Error(error_message) def run_ltx_refinement( state: dict, prompt: str, neg_prompt: str, progress=gr.Progress(track_tqdm=True) ) -> tuple: """Wrapper to call the LTX texture refinement and upscaling backend function.""" if not state or not state.get("low_res_latents"): raise gr.Error("Error: Please generate a base video in Step 1 before refining.") try: print("[UI] Request received for LTX refinement.") video_path, tensor_path = video_generation_service.generate_upscale_denoise( latents_path=state["low_res_latents"], prompt=prompt, negative_prompt=neg_prompt, seed=state["used_seed"] ) # Update state with refined assets state["refined_video_ltx"] = video_path state["refined_latents_ltx"] = tensor_path print(f"[UI] LTX refinement successful. Path: {video_path}") return video_path, state except Exception as e: error_message = f"❌ An error occurred during LTX Refinement:\n{e}" print(f"{error_message}\nDetails: {traceback.format_exc()}") raise gr.Error(error_message) def run_seedvr_upscaling( state: dict, seed: int, resolution: int, batch_size: int, fps: int, progress=gr.Progress(track_tqdm=True) ) -> tuple: """Wrapper to call the SeedVR upscaling backend service.""" if not state or not state.get("low_res_video"): raise gr.Error("Error: Please generate a base video in Step 1 before upscaling.") if not seedvr_inference_server: raise gr.Error("Error: The SeedVR upscaling server is not available.") try: print("[UI] Request received for SeedVR upscaling.") def progress_wrapper(p, desc=""): progress(p, desc=desc) output_filepath = seedvr_inference_server.run_inference( file_path=state["low_res_video"], seed=seed, resolution=resolution, batch_size=batch_size, fps=fps, progress=progress_wrapper ) status_message = f"✅ Upscaling complete!\nSaved to: {output_filepath}" print(f"[UI] SeedVR upscaling successful. Path: {output_filepath}") return gr.update(value=output_filepath), gr.update(value=status_message) except Exception as e: error_message = f"❌ An error occurred during SeedVR Upscaling:\n{e}" print(f"{error_message}\nDetails: {traceback.format_exc()}") return None, gr.update(value=error_message) # ============================================================================== # --- UI BUILDER --- # Functions dedicated to creating parts of the Gradio interface. # ============================================================================== def build_ui(): """Constructs the entire Gradio application UI.""" with gr.Blocks(theme=gr.themes.Soft(primary_hue="blue")) as demo: # App state persists across interactions within a session app_state = gr.State(value={"low_res_video": None, "low_res_latents": None, "used_seed": None}) gr.Markdown("# LTX Video - Geração e Pós-Produção por Etapas", elem_id="main-title") ui_components = {} # Dictionary to hold all key UI components with gr.Row(): with gr.Column(scale=1): # Build the main generation controls (Step 1) _build_generation_controls(ui_components) with gr.Column(scale=1): gr.Markdown("### Vídeo Base Gerado") ui_components['low_res_video_output'] = gr.Video( label="O resultado da Etapa 1 aparecerá aqui", interactive=False ) # Build the post-production section (Step 2), initially hidden _build_postprod_controls(ui_components) # Connect all UI events to their corresponding functions _register_event_handlers(app_state, ui_components) return demo def _build_generation_controls(ui: dict): """Builds the UI components for Step 1: Base Video Generation.""" gr.Markdown("### Etapa 1: Configurações de Geração") ui['generation_mode'] = gr.Radio( label="Modo de Geração", choices=["Simples (Prompt Único)", "Narrativa (Múltiplos Prompts)"], value="Narrativa (Múltiplos Prompts)", info="Simples para uma ação contínua, Narrativa para uma sequência de cenas (uma por linha)." ) ui['prompt'] = gr.Textbox(label="Prompt(s)", value="Um leão majestoso caminha pela savana\nEle sobe em uma grande pedra e olha o horizonte", lines=4) ui['neg_prompt'] = gr.Textbox(label="Negative Prompt", value="blurry, low quality, bad anatomy, deformed", lines=2) ui['start_image'] = gr.Image(label="Imagem de Início (Opcional)", type="filepath", sources=["upload"]) with gr.Accordion("Parâmetros Principais", open=True): ui['duration'] = gr.Slider(label="Duração Total (s)", value=4, step=1, minimum=1, maximum=30) with gr.Row(): ui['height'] = gr.Slider(label="Height", value=432, step=16, minimum=256, maximum=1024) ui['width'] = gr.Slider(label="Width", value=768, step=16, minimum=256, maximum=1024) with gr.Row(): ui['seed'] = gr.Number(label="Seed", value=42, precision=0) ui['randomize_seed'] = gr.Checkbox(label="Randomize Seed", value=True) with gr.Accordion("Opções Avançadas de Guiagem (First Pass)", open=False): ui['fp_guidance_preset'] = gr.Dropdown( label="Preset de Guiagem", choices=["Padrão (Recomendado)", "Agressivo", "Suave", "Customizado"], value="Padrão (Recomendado)", info="Controla como a guiagem de texto se comporta ao longo da difusão." ) with gr.Group(visible=False) as ui['custom_guidance_group']: gr.Markdown("⚠️ Edite as listas em formato JSON. Ex: `[1.0, 2.5, 3.0]`") ui['fp_guidance_scale_list'] = gr.Textbox(label="Lista de Guidance Scale", value="[1, 1, 6, 8, 6, 1, 1]") ui['fp_stg_scale_list'] = gr.Textbox(label="Lista de STG Scale (Movimento)", value="[0, 0, 4, 4, 4, 2, 1]") ui['generate_low_btn'] = gr.Button("1. Gerar Vídeo Base", variant="primary") def _build_postprod_controls(ui: dict): """Builds the UI components for Step 2: Post-Production.""" with gr.Group(visible=False) as ui['post_prod_group']: gr.Markdown("--- \n## Etapa 2: Pós-Produção", elem_id="postprod-title") with gr.Tabs(): with gr.TabItem("🚀 Upscaler de Textura (LTX)"): with gr.Row(): with gr.Column(scale=1): gr.Markdown("Usa o prompt e a semente originais para refinar o vídeo, adicionando detalhes e texturas de alta qualidade.") ui['ltx_refine_btn'] = gr.Button("2. Aplicar Refinamento LTX", variant="primary") with gr.Column(scale=1): ui['ltx_refined_video_output'] = gr.Video(label="Vídeo com Textura Refinada", interactive=False) with gr.TabItem("✨ Upscaler de Resolução (SeedVR)"): is_seedvr_available = seedvr_inference_server is not None if not is_seedvr_available: gr.Markdown("🔴 *O serviço SeedVR não está disponível nesta instância.*") with gr.Row(): with gr.Column(scale=1): ui['seedvr_seed'] = gr.Slider(minimum=0, maximum=999999, value=42, step=1, label="Seed") ui['seedvr_resolution'] = gr.Slider(minimum=720, maximum=1440, value=1072, step=8, label="Resolução Vertical") ui['seedvr_batch_size'] = gr.Slider(minimum=1, maximum=16, value=4, step=1, label="Batch Size por GPU") ui['seedvr_fps'] = gr.Number(label="FPS de Saída (0 = original)", value=0) ui['run_seedvr_btn'] = gr.Button("2. Iniciar Upscaling SeedVR", variant="primary", interactive=is_seedvr_available) with gr.Column(scale=1): ui['seedvr_video_output'] = gr.Video(label="Vídeo com Upscale SeedVR", interactive=False) ui['seedvr_status_box'] = gr.Textbox(label="Status", value="Aguardando...", lines=3, interactive=False) # ============================================================================== # --- EVENT HANDLERS --- # Connects UI component events (like clicks) to the wrapper functions. # ============================================================================== def _register_event_handlers(app_state: gr.State, ui: dict): """Registers all Gradio event handlers.""" # --- Handler for custom guidance visibility --- def toggle_custom_guidance(preset_choice: str) -> gr.update: return gr.update(visible=(preset_choice == "Customizado")) ui['fp_guidance_preset'].change( fn=toggle_custom_guidance, inputs=ui['fp_guidance_preset'], outputs=ui['custom_guidance_group'] ) # --- Handler for the main "Generate" button --- gen_inputs = [ ui['generation_mode'], ui['prompt'], ui['neg_prompt'], ui['start_image'], ui['height'], ui['width'], ui['duration'], ui['seed'], ui['randomize_seed'], ui['fp_guidance_preset'], ui['fp_guidance_scale_list'], ui['fp_stg_scale_list'] ] gen_outputs = [ ui['low_res_video_output'], app_state, ui['post_prod_group'] ] ui['generate_low_btn'].click(fn=run_generate_base_video, inputs=gen_inputs, outputs=gen_outputs) # --- Handler for the LTX Refine button --- refine_inputs = [app_state, ui['prompt'], ui['neg_prompt']] refine_outputs = [ui['ltx_refined_video_output'], app_state] ui['ltx_refine_btn'].click(fn=run_ltx_refinement, inputs=refine_inputs, outputs=refine_outputs) # --- Handler for the SeedVR Upscale button --- if 'run_seedvr_btn' in ui: seedvr_inputs = [app_state, ui['seedvr_seed'], ui['seedvr_resolution'], ui['seedvr_batch_size'], ui['seedvr_fps']] seedvr_outputs = [ui['seedvr_video_output'], ui['seedvr_status_box']] ui['run_seedvr_btn'].click(fn=run_seedvr_upscaling, inputs=seedvr_inputs, outputs=seedvr_outputs) # ============================================================================== # --- APPLICATION ENTRY POINT --- # ============================================================================== if __name__ == "__main__": print("Building Gradio UI...") gradio_app = build_ui() print("Launching Gradio app...") gradio_app.queue().launch( server_name="0.0.0.0", server_port=7860, debug=True, show_error=True )