Latest Changes

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+*.jpg filter=lfs diff=lfs merge=lfs -text

ui/layout.py

@@ -14,40 +14,54 @@ def build_ui(model, processor):
             style=style
         )
 
-    with gr.Blocks() as demo:
-        gr.Markdown("## 🖼️ BLIP Image Captioning (Zero-GPU)")
-        gr.Markdown(
-            "Generate captions or explanations from images using a CPU-only BLIP model."
-        )
+    with gr.Blocks(title="BLIP Image Captioning (Zero-GPU)") as demo:
+        with gr.Column(elem_id="container"):
+            gr.Markdown("## 🖼️ BLIP Image Captioning (Zero-GPU)")
+            gr.Markdown(
+                "Generate captions or explanations from images using a CPU-only BLIP model."
+            )
+
+            with gr.Row(equal_height=True):
+                with gr.Column():
+                    image_input = gr.Image(
+                        type="pil",
+                        label="Upload Image"
+                    )
+                    generate_btn = gr.Button("Generate")
 
-        with gr.Row():
-            image_input = gr.Image(
-                type="pil",
-                label="Upload Image"
+                with gr.Column():
+                    style_select = gr.Dropdown(
+                        choices=[
+                            "Short Caption",
+                            "Detailed Caption",
+                            "Creative Caption",
+                            "Image Explanation"
+                        ],
+                        value="Detailed Caption",
+                        label="Caption Style"
+                    )
+                    output_text = gr.Textbox(
+                        label="Generated Output",
+                        lines=5
+                    )
+
+            # ✅ Examples preload inputs ONLY
+            gr.Examples(
+                examples=[
+                    ["./assets/zebra.jpg", "Detailed Caption"],
+                    ["./assets/cat.jpg", "Image Explanation"],
+                    ["./assets/fridge.jpg", "Detailed Caption"],
+                    ["./assets/marriage.jpg", "Creative Caption"],
+                    ["./assets/giraffe.jpg", "Short Caption"]
+                ],
+                inputs=[image_input, style_select]
             )
 
-            with gr.Column():
-                style_select = gr.Dropdown(
-                    choices=[
-                        "Short Caption",
-                        "Detailed Caption",
-                        "Creative Caption",
-                        "Image Explanation"
-                    ],
-                    value="Detailed Caption",
-                    label="Caption Style"
-                )
-                output_text = gr.Textbox(
-                    label="Generated Output",
-                    lines=4
-                )
-
-        generate_btn = gr.Button("Generate")
-
-        generate_btn.click(
-            fn=run_caption,
-            inputs=[image_input, style_select],
-            outputs=output_text
-        )
+            # ✅ Single source of inference
+            generate_btn.click(
+                fn=run_caption,
+                inputs=[image_input, style_select],
+                outputs=output_text
+            )
 
     return demo