Update app.py

app.py

@@ -1,4 +1,5 @@
 import os
+import torch
 import gradio as gr
 import numpy as np
 import random
@@ -16,6 +17,20 @@ MAX_SEED = np.iinfo(np.int32).max
 HF_TOKEN = os.environ.get("HF_TOKEN")
 HF_TOKEN_UPSCALER = os.environ.get("HF_TOKEN_UPSCALER")
 
+if not os.path.exists('GFPGANv1.4.pth'):
+    os.system("wget https://github.com/TencentARC/GFPGAN/releases/download/v1.3.0/GFPGANv1.4.pth -P .")
+
+device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+model_path = 'GFPGANv1.4.pth'
+gfpgan = GFPGANer(
+    model_path=model_path,
+    upscale_factor=4,
+    arch='clean',
+    channel_multiplier=2,
+    model_name='GPFGAN',
+    device=device
+)
+
 def enable_lora(lora_add, basemodel):
     return basemodel if not lora_add else lora_add
 
@@ -41,7 +56,15 @@ def get_upscale_finegrain(prompt, img_path, upscale_factor):
         print(f"Error upscale image: {e}")
         return None
 
-async def gen(prompt, basemodel, width, height, scales, steps, seed, upscale_factor, process_upscale, lora_model, process_lora):
+def get_upscale_gfpgan(prompt, img_path):
+    try:
+        img = gfpgan.enhance(img_path)
+        return img
+    except Exception as e:
+        print(f"Error upscale image: {e}")
+        return None
+
+async def gen(prompt, basemodel, width, height, scales, steps, seed, upscale_factor, process_upscale, lora_model, process_lora, upscale_model):
     model = enable_lora(lora_model, basemodel) if process_lora else basemodel
     image, seed = await generate_image(prompt, model, "", width, height, scales, steps, seed)
     if image is None:
@@ -51,7 +74,10 @@ async def gen(prompt, basemodel, width, height, scales, steps, seed, upscale_fac
     image.save(image_path, format="JPEG")
     
     if process_upscale:
-        upscale_image = get_upscale_finegrain(prompt, image_path, upscale_factor)
+        if upscale_model == "FineGrain":
+            upscale_image = get_upscale_finegrain(prompt, image_path, upscale_factor)
+        elif upscale_model == "GPFGAN":
+            upscale_image = get_upscale_gfpgan(prompt, image_path)
         upscale_image_path = "upscale_image.jpg"
         upscale_image.save(upscale_image_path, format="JPEG")
         return [image_path, upscale_image_path]
@@ -74,6 +100,7 @@ with gr.Blocks(css=css, theme="Nymbo/Nymbo_Theme") as demo:
                 process_lora = gr.Checkbox(label="Procesar LORA")
                 process_upscale = gr.Checkbox(label="Procesar Escalador")
                 upscale_factor = gr.Radio(label="Factor de Escala", choices=[2, 4, 8], value=2)
+                upscale_model = gr.Radio(label="Modelo de Escalado", choices=["FineGrain", "GPFGAN"], value="FineGrain")
                 
                 with gr.Accordion(label="Opciones Avanzadas", open=False):
                     width = gr.Slider(label="Ancho", minimum=512, maximum=1280, step=8, value=1280)
@@ -90,7 +117,7 @@ with gr.Blocks(css=css, theme="Nymbo/Nymbo_Theme") as demo:
                     queue=False
                 ).then(
                     fn=gen,
-                    inputs=[prompt, basemodel_choice, width, height, scales, steps, seed, upscale_factor, process_upscale, lora_model_choice, process_lora],
+                    inputs=[prompt, basemodel_choice, width, height, scales, steps, seed, upscale_factor, process_upscale, lora_model_choice, process_lora, upscale_model],
                     outputs=[output_res]
                 )