Update app.py

app.py

@@ -85,7 +85,7 @@ def resize_image(image, resolution):
     new_h = int(h * ratio)
     return image.resize((new_w, new_h), Image.LANCZOS)
 
-def text_to_image(text, size=72, position="middle-center"):
+def text_to_image(text, size=72, position="top-center"):
     width, height = 1024, 576
     image = Image.new("RGB", (width, height), "white")
     draw = ImageDraw.Draw(image)
@@ -137,7 +137,7 @@ def text_to_image(text, size=72, position="middle-center"):
     return image
 
 @spaces.GPU
-def infer_canny(prompt, text_for_image, 
+def infer_canny(prompt, text_for_image, text_position, font_size,
           negative_prompt = "nsfw, facial shadows, low resolution, jpeg artifacts, blurry, bad quality, dark face, neon lights", 
           seed = 397886929, 
           randomize_seed = False,
@@ -155,7 +155,7 @@ def infer_canny(prompt, text_for_image,
     generator = torch.Generator().manual_seed(seed)
     
     # Generate text image
-    init_image = text_to_image(text_for_image)
+    init_image = text_to_image(text_for_image, size=font_size, position=text_position)
     init_image = resize_image(init_image, MAX_IMAGE_SIZE)
     
     pipe = pipe_canny.to("cuda")
@@ -179,6 +179,16 @@ css = """
 footer {
     visibility: hidden;
 }
+.text-position-grid {
+    display: grid;
+    grid-template-columns: repeat(3, 1fr);
+    gap: 5px;
+    margin-bottom: 10px;
+}
+.text-position-grid button {
+    aspect-ratio: 1;
+    padding: 0;
+}
 """
 
 with gr.Blocks(theme="Nymbo/Nymbo_Theme", css=css) as Kolors:
@@ -196,6 +206,28 @@ with gr.Blocks(theme="Nymbo/Nymbo_Theme", css=css) as Kolors:
                     placeholder="Enter text to be converted into an image",
                     lines=3
                 )
+            with gr.Row():
+                with gr.Column():
+                    gr.Markdown("Text Position")
+                    with gr.Row(elem_classes="text-position-grid"):
+                        text_positions = ["top-left", "top-center", "top-right", 
+                                          "middle-left", "middle-center", "middle-right", 
+                                          "bottom-left", "bottom-center", "bottom-right"]
+                        text_position = gr.Radio(
+                            choices=text_positions,
+                            value="top-center",
+                            label="",
+                            show_label=False,
+                            elem_classes="text-position-grid"
+                        )
+                with gr.Column():
+                    font_size = gr.Slider(
+                        label="Font Size",
+                        minimum=12,
+                        maximum=144,
+                        step=1,
+                        value=72
+                    )
             with gr.Accordion("Advanced Settings", open=False):
                 negative_prompt = gr.Textbox(
                     label="Negative prompt",
@@ -258,7 +290,7 @@ with gr.Blocks(theme="Nymbo/Nymbo_Theme", css=css) as Kolors:
 
     canny_button.click(
         fn = infer_canny,
-        inputs = [prompt, text_for_image, negative_prompt, seed, randomize_seed, guidance_scale, num_inference_steps, controlnet_conditioning_scale, control_guidance_end, strength],
+        inputs = [prompt, text_for_image, text_position, font_size, negative_prompt, seed, randomize_seed, guidance_scale, num_inference_steps, controlnet_conditioning_scale, control_guidance_end, strength],
         outputs = [result, seed_used]
     )