[AUTOMATED] Migration to Gradio 6.0

README.md

@@ -4,7 +4,7 @@ emoji: 💻
 colorFrom: indigo
 colorTo: red
 sdk: gradio
-sdk_version: 5.44.1
+sdk_version: 6.0.0
 app_file: app.py
 pinned: false
 license: apache-2.0

app.py

@@ -303,8 +303,81 @@ def predict(image, model_selection, top_k, threshold):
 
 
 # --- Gradio Interface ---
-with gr.Blocks(
-    css="""
+with gr.Blocks() as demo:
+
+    gr.Markdown("# Kaloscope Artist Style Classification")
+
+    with gr.Row():
+        with gr.Column():
+            image_input = gr.Image(type="pil", label="Upload Image", elem_id="image-upload")
+
+        with gr.Column():
+            submit_btn = gr.Button("Predict")
+
+            tags_output = gr.Textbox(label="Predicted Tags", buttons=["copy"])
+            prettier_output = gr.DataFrame(
+                elem_id="results-table-wrapper",
+                # value=[
+                #    [
+                #        1,
+                #        "[Samplaaaae Artist](https://example.com)",
+                #        "<span class='copy-btn' data-copy='Samplaaaae Artist'>📋</span>",
+                #        "95.00%",
+                #    ],
+                #    [
+                #        2,
+                #        "[Another Artist](https://example.com)",
+                #        "<span class='copy-btn' data-copy='Another Artist'>📋</span>",
+                #        "90.00%",
+                #    ],
+                #    [
+                #        3,
+                #        "[Third Artist](https://example.com)",
+                #        "<span class='copy-btn' data-copy='Third Artist'>📋</span>",
+                #        "85.00%",
+                #    ],
+                # ],
+                interactive=False,
+                datatype=["number", "markdown", "html", "str"],
+                headers=["Rank", "Artist", "", "Score"],
+            )
+            json_accordion = gr.Accordion("JSON Output", open=False)
+            with json_accordion:
+                json_output = gr.Code(language="json", show_label=False, lines=7)
+
+            with gr.Group():
+                model_selection = gr.Dropdown(
+                    choices=[
+                        (
+                            f"{name}",
+                            # f"{name} | Repo: {MODELS[name].get('repo_id') or 'local'}",
+                            name,
+                        )
+                        for name in MODELS
+                    ],
+                    value=list(MODELS.keys())[0],
+                    label="Select Model",
+                )
+                with gr.Row(elem_id="slider-row-container"):
+                    top_k_slider = gr.Slider(minimum=1, maximum=25, value=5, step=1, label="Top K")
+                    threshold_slider = gr.Slider(minimum=0.0, maximum=1.0, value=0.0, step=0.01, label="Threshold")
+                time_display = gr.Markdown()  # populated after prediction
+
+    gr.Markdown(
+        "Models sourced from [heathcliff01/Kaloscope](https://huggingface.co/heathcliff01/Kaloscope) & [heathcliff01/Kaloscope2.0](https://huggingface.co/heathcliff01/Kaloscope2.0) (Original PyTorch releases) "
+        + "and [DraconicDragon/Kaloscope-onnx](https://huggingface.co/DraconicDragon/Kaloscope-onnx) (ONNX converted and EMA weights).  \n"
+        + "OpenVINO™ will be used to accelerate ONNX CPU inference with ONNX CPUExecutionProvider as fallback."
+    )
+
+    submit_btn.click(
+        fn=predict,
+        inputs=[image_input, model_selection, top_k_slider, threshold_slider],
+        outputs=[tags_output, prettier_output, json_output, time_display],
+    )
+
+if __name__ == "__main__":
+    demo.launch(
+        css="""
     * { box-sizing: border-box; }
     @media (max-width: 1022px) {
         #slider-row-container {
@@ -393,7 +466,7 @@ with gr.Blocks(
         background: rgba(128, 128, 128, 0.1);
     }
 """,
-    js="""
+        js="""
     function() {
         document.addEventListener('click', function(e) {
             if (e.target.classList.contains('copy-btn')) {
@@ -428,77 +501,4 @@ with gr.Blocks(
         }, true);
     }
     """,
-) as demo:
-
-    gr.Markdown("# Kaloscope Artist Style Classification")
-
-    with gr.Row():
-        with gr.Column():
-            image_input = gr.Image(type="pil", label="Upload Image", elem_id="image-upload")
-
-        with gr.Column():
-            submit_btn = gr.Button("Predict")
-
-            tags_output = gr.Textbox(label="Predicted Tags", show_copy_button=True)
-            prettier_output = gr.DataFrame(
-                elem_id="results-table-wrapper",
-                # value=[
-                #    [
-                #        1,
-                #        "[Samplaaaae Artist](https://example.com)",
-                #        "<span class='copy-btn' data-copy='Samplaaaae Artist'>📋</span>",
-                #        "95.00%",
-                #    ],
-                #    [
-                #        2,
-                #        "[Another Artist](https://example.com)",
-                #        "<span class='copy-btn' data-copy='Another Artist'>📋</span>",
-                #        "90.00%",
-                #    ],
-                #    [
-                #        3,
-                #        "[Third Artist](https://example.com)",
-                #        "<span class='copy-btn' data-copy='Third Artist'>📋</span>",
-                #        "85.00%",
-                #    ],
-                # ],
-                interactive=False,
-                datatype=["number", "markdown", "html", "str"],
-                headers=["Rank", "Artist", "", "Score"],
-            )
-            json_accordion = gr.Accordion("JSON Output", open=False)
-            with json_accordion:
-                json_output = gr.Code(language="json", show_label=False, lines=7)
-
-            with gr.Group():
-                model_selection = gr.Dropdown(
-                    choices=[
-                        (
-                            f"{name}",
-                            # f"{name} | Repo: {MODELS[name].get('repo_id') or 'local'}",
-                            name,
-                        )
-                        for name in MODELS
-                    ],
-                    value=list(MODELS.keys())[0],
-                    label="Select Model",
-                )
-                with gr.Row(elem_id="slider-row-container"):
-                    top_k_slider = gr.Slider(minimum=1, maximum=25, value=5, step=1, label="Top K")
-                    threshold_slider = gr.Slider(minimum=0.0, maximum=1.0, value=0.0, step=0.01, label="Threshold")
-                time_display = gr.Markdown()  # populated after prediction
-
-    gr.Markdown(
-        "Models sourced from [heathcliff01/Kaloscope](https://huggingface.co/heathcliff01/Kaloscope) & [heathcliff01/Kaloscope2.0](https://huggingface.co/heathcliff01/Kaloscope2.0) (Original PyTorch releases) "
-        + "and [DraconicDragon/Kaloscope-onnx](https://huggingface.co/DraconicDragon/Kaloscope-onnx) (ONNX converted and EMA weights).  \n"
-        + "OpenVINO™ will be used to accelerate ONNX CPU inference with ONNX CPUExecutionProvider as fallback."
-    )
-
-    submit_btn.click(
-        fn=predict,
-        inputs=[image_input, model_selection, top_k_slider, threshold_slider],
-        outputs=[tags_output, prettier_output, json_output, time_display],
-    )
-
-if __name__ == "__main__":
-    demo.launch()
+    )