Upload app.py with huggingface_hub

app.py

@@ -0,0 +1,43 @@
+import gradio as gr
+import torch
+from transformers import AutoModelForSequenceClassification, AutoTokenizer
+
+MODEL_NAME = "Ankit74990/TruthX-DISTILBERT"
+
+tokenizer = None
+model = None
+
+def load_model():
+    global tokenizer, model
+    if tokenizer is None:
+        tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
+        model = AutoModelForSequenceClassification.from_pretrained(MODEL_NAME)
+        model.eval()
+
+def predict(text):
+    load_model()
+    inputs = tokenizer(text, return_tensors="pt", truncation=True, max_length=512, padding=True)
+    with torch.no_grad():
+        outputs = model(**inputs)
+    probs = torch.nn.functional.softmax(outputs.logits, dim=-1)
+    fake_prob = probs[0][0].item()
+    real_prob = probs[0][1].item()
+    label = "Real News" if fake_prob < real_prob else "Fake News"
+    confidence = max(fake_prob, real_prob)
+    return {"label": label, "confidence": f"{confidence:.2%}", "fake_probability": f"{fake_prob:.2%}", "real_probability": f"{real_prob:.2%}"}
+
+demo = gr.Interface(
+    fn=predict,
+    inputs="text",
+    outputs=gr.Label(label="Prediction Result"),
+    title="TruthX - Fake News Detector",
+    description="Enter a news article or headline to determine if it's real or fake news.",
+    examples=[
+        ["Breaking: Scientists discover new cure for cancer"],
+        ["Government announces new policy changes for education"],
+        ["Massive scandal revealed involving major corporation"],
+    ]
+)
+
+if __name__ == "__main__":
+    demo.launch()