import os import gradio as gr import numpy as np import random from pathlib import Path from PIL import Image from huggingface_hub import AsyncInferenceClient, InferenceClient from gradio_client import Client, handle_file from gradio_imageslider import ImageSlider MAX_SEED = np.iinfo(np.int32).max HF_TOKEN = os.environ.get("HF_TOKEN") HF_TOKEN_UPSCALER = os.environ.get("HF_TOKEN_UPSCALER") client = AsyncInferenceClient() llm_client = InferenceClient("mistralai/Mixtral-8x7B-Instruct-v0.1") # Directorio de almacenamiento de imágenes DATA_PATH = Path("./data") DATA_PATH.mkdir(exist_ok=True) # Asegura que el directorio exista def enable_lora(lora_add, basemodel): return basemodel if not lora_add else lora_add async def generate_image(combined_prompt, model, width, height, scales, steps, seed): try: if seed == -1: seed = random.randint(0, MAX_SEED) seed = int(seed) image = await client.text_to_image( prompt=combined_prompt, height=height, width=width, guidance_scale=scales, num_inference_steps=steps, model=model ) return image, seed except Exception as e: return f"Error al generar imagen: {e}", None def get_upscale_finegrain(prompt, img_path, upscale_factor): try: client = Client("finegrain/finegrain-image-enhancer", hf_token=HF_TOKEN_UPSCALER) result = client.predict( input_image=handle_file(img_path), prompt=prompt, negative_prompt="", seed=42, upscale_factor=upscale_factor, controlnet_scale=0.6, controlnet_decay=1, condition_scale=6, tile_width=112, tile_height=144, denoise_strength=0.35, num_inference_steps=18, solver="DDIM", api_name="/process" ) return result[1] if isinstance(result, list) and len(result) > 1 else None except Exception as e: return None async def gen(prompt, basemodel, width, height, scales, steps, seed, upscale_factor, process_upscale, lora_model, process_lora): model = enable_lora(lora_model, basemodel) if process_lora else basemodel improved_prompt = await improve_prompt(prompt) combined_prompt = f"{prompt} {improved_prompt}" if seed == -1: seed = random.randint(0, MAX_SEED) seed = int(seed) image, seed = await generate_image(combined_prompt, model, width, height, scales, steps, seed) if isinstance(image, str) and image.startswith("Error"): return [image, None] image_path = DATA_PATH / f"image_{seed}.jpg" image.save(image_path, format="JPEG") if process_upscale: upscale_image_path = get_upscale_finegrain(combined_prompt, image_path, upscale_factor) if upscale_image_path: upscale_image = Image.open(upscale_image_path) upscale_image.save(DATA_PATH / f"upscale_image_{seed}.jpg", format="JPEG") return [image_path, DATA_PATH / f"upscale_image_{seed}.jpg"] else: return [image_path, image_path] else: return [image_path, image_path] async def improve_prompt(prompt): try: instruction = ("With this idea, describe in English a detailed img2vid prompt in a single paragraph of up to 200 characters maximum, developing atmosphere, characters, lighting, and cameras.") formatted_prompt = f"{prompt}: {instruction}" response = llm_client.text_generation(formatted_prompt, max_new_tokens=200) improved_text = response['generated_text'].strip() if 'generated_text' in response else response.strip() return improved_text except Exception as e: return f"Error mejorando el prompt: {e}" def get_storage(): files = [ { "name": str(file.resolve()), "size": file.stat().st_size, } for file in DATA_PATH.glob("*.jpg") if file.is_file() ] usage = sum([f['size'] for f in files]) return [file["name"] for file in files], f"Uso total: {usage/(1024.0 ** 3):.3f}GB" css = """ #col-container{ margin: 0 auto; max-width: 1024px;} """ with gr.Blocks(css=css, theme="Nymbo/Nymbo_Theme") as demo: with gr.Column(elem_id="col-container"): with gr.Row(): with gr.Column(scale=3): output_res = ImageSlider(label="Generadas / Escaladas") with gr.Column(scale=2): prompt = gr.Textbox(label="Descripción de imagen") basemodel_choice = gr.Dropdown( label="Modelo", choices=["black-forest-labs/FLUX.1-schnell", "black-forest-labs/FLUX.1-DEV"], value="black-forest-labs/FLUX.1-schnell" ) lora_model_choice = gr.Dropdown( label="LORA Realismo", choices=["Shakker-Labs/FLUX.1-dev-LoRA-add-details", "XLabs-AI/flux-RealismLora"], value="XLabs-AI/flux-RealismLora" ) with gr.Row(): process_lora = gr.Checkbox(label="Procesar LORA") process_upscale = gr.Checkbox(label="Procesar Escalador") improved_prompt = gr.Textbox(label="Prompt Mejorado", interactive=False) improve_btn = gr.Button("Mejorar prompt") improve_btn.click(fn=improve_prompt, inputs=[prompt], outputs=improved_prompt) with gr.Accordion(label="Opciones Avanzadas", open=False): width = gr.Slider(label="Ancho", minimum=512, maximum=1280, step=8, value=1280) height = gr.Slider(label="Alto", minimum=512, maximum=1280, step=8, value=768) upscale_factor = gr.Radio(label="Factor de Escala", choices=[2, 4, 8], value=2) scales = gr.Slider(label="Escalado", minimum=1, maximum=20, step=1, value=10) steps = gr.Slider(label="Pasos", minimum=1, maximum=100, step=1, value=20) seed = gr.Number(label="Semilla", value=-1) btn = gr.Button("Generar") btn.click( fn=gen, inputs=[prompt, basemodel_choice, width, height, scales, steps, seed, upscale_factor, process_upscale, lora_model_choice, process_lora], outputs=output_res ) with gr.Row(): with gr.Column(): file_list = gr.Gallery(label="Imágenes Guardadas") # Usar Gallery en lugar de Files storage_info = gr.Text(label="Uso de Almacenamiento") refresh_btn = gr.Button("Actualizar Galería") refresh_btn.click(fn=get_storage, inputs=None, outputs=[file_list, storage_info]) demo.launch(allowed_paths=[str(DATA_PATH)])