Create app.py

app.py

@@ -0,0 +1,35 @@
+import gradio as gr
+import torch
+from transformers import AutoImageProcessor, AutoModelForImageClassification
+from PIL import Image
+
+# Load model + processor (auto cached inside Spaces)
+processor = AutoImageProcessor.from_pretrained("prithivMLmods/Realistic-Gender-Classification")
+model = AutoModelForImageClassification.from_pretrained("prithivMLmods/Realistic-Gender-Classification")
+
+def predict(image):
+    # Preprocess
+    inputs = processor(images=image, return_tensors="pt")
+    with torch.no_grad():
+        outputs = model(**inputs)
+        probs = torch.nn.functional.softmax(outputs.logits, dim=-1)[0].cpu().numpy()
+
+    # Labels
+    labels = list(model.config.id2label.values())
+
+    # Clean dict for FlutterFlow
+    result = {
+        "female": float(probs[labels.index("female portrait")]),
+        "male": float(probs[labels.index("male portrait")])
+    }
+    return result
+
+# Gradio interface (Spaces auto-hosts this)
+demo = gr.Interface(
+    fn=predict,
+    inputs=gr.Image(type="pil"),
+    outputs=gr.JSON()
+)
+
+if __name__ == "__main__":
+    demo.launch()