Create app.py

app.py

@@ -0,0 +1,45 @@
+import torch
+from transformers import AutoImageProcessor, SiglipForImageClassification
+from PIL import Image
+import torch.nn.functional as F
+import gradio as gr
+
+# Load model and processor from Hugging Face Hub
+model_path = "Ateeqq/nsfw-image-detection"
+processor = AutoImageProcessor.from_pretrained(model_path)
+model = SiglipForImageClassification.from_pretrained(model_path)
+model.eval()
+
+def predict(image):
+    # Convert to RGB and preprocess
+    image = Image.fromarray(image).convert("RGB")
+    inputs = processor(images=image, return_tensors="pt")
+
+    # Inference
+    with torch.no_grad():
+        logits = model(**inputs).logits
+    probs = F.softmax(logits, dim=1)[0].tolist()
+
+    # Prepare output dict
+    labels = [model.config.id2label[i] for i in range(len(probs))]
+    result = {labels[i]: float(f"{probs[i]:.6f}") for i in range(len(labels))}
+    predicted_idx = int(torch.argmax(logits, dim=1)[0])
+    result["predicted_label"] = model.config.id2label[predicted_idx]
+    return result
+
+# Gradio Interface
+def main():
+    description = "NSFW Image Detection using SigLIP2 Safety Classifier"
+    iface = gr.Interface(
+        fn=predict,
+        inputs=gr.Image(type="numpy", label="Upload Image"),
+        outputs=gr.Label(num_top_classes=3, label="Predictions"),
+        title="NSFW Image Detector",
+        description=description,
+        examples=[],
+        allow_flagging="never"
+    )
+    iface.launch()
+
+if __name__ == "__main__":
+    main()