Upload app.py

app.py

@@ -474,7 +474,7 @@ def position_text_back(text, merged_coordinates, inpainted_image, font_path, fon
 def process(image, mask,
             prompt="background",
             negative_prompt="text",
-            num_inference_steps=10,
+            num_inference_steps=15,
             controlnet_conditioning_scale=0.9,
             guidance_scale=3.5,
             seed=124,
@@ -505,7 +505,7 @@ def process(image, mask,
 
 
 @spaces.GPU()
-def process_image(image, source_language, target_language, mode, font, font_color, num_inference_steps):
+def process_image(image, source_language, target_language, mode, font, font_color, num_inference_steps=15):
     """Main processing function for Gradio"""
     if image is None:
         return None, None, None, []
@@ -595,7 +595,6 @@ with gr.Blocks(title="Nativ - Demo") as demo:
             value='omniscript_bold',
                 label="Select Font"
             )
-            num_inference_steps = gr.Slider(minimum=1, step=1, maximum=50, value=10, label="num_inference_steps")
             font_color_picker_i = gr.ColorPicker(
                 value="#000000",  # Default color: black
                 label="Select Font Color"
@@ -627,8 +626,8 @@ with gr.Blocks(title="Nativ - Demo") as demo:
         )
         update_btn = gr.Button("Apply Changes")
     
-    def process_and_show_translations(image, source_lang, target_lang, mode, font, font_color, num_inference_steps):
-        boxes, bubbles, removed, final, translations, mask = process_image(image, source_lang, target_lang, mode, font, font_color, num_inference_steps)
+    def process_and_show_translations(image, source_lang, target_lang, mode, font, font_color):
+        boxes, bubbles, removed, final, translations, mask = process_image(image, source_lang, target_lang, mode, font, font_color)
         # Extract just the texts and join with newlines
         texts = '\n'.join(t['text'] for t in translations)
         return boxes, bubbles, removed, final, texts, translations
@@ -636,7 +635,7 @@ with gr.Blocks(title="Nativ - Demo") as demo:
     # Process button click
     process_btn.click(
         fn=process_and_show_translations,
-        inputs=[input_image, source_language, target_language, localization_mode, font_selector_i, font_color_picker_i, num_inference_steps],
+        inputs=[input_image, source_language, target_language, localization_mode, font_selector_i, font_color_picker_i],
         outputs=[detected_boxes, speech_bubbles, removed_text, final_output, translations_text, translations_state]
     )