Update app.py

app.py

@@ -1,23 +1,36 @@
 import gradio as gr
 from transformers import AutoModelForSequenceClassification, AutoTokenizer
 
-model_name = "alperugurcan/Contradictory"
-tokenizer = AutoTokenizer.from_pretrained(model_name)
-model = AutoModelForSequenceClassification.from_pretrained(model_name)
+# Load model
+model = AutoModelForSequenceClassification.from_pretrained("alperugurcan/Contradictory")
+tokenizer = AutoTokenizer.from_pretrained("alperugurcan/Contradictory")
 
 def predict(premise, hypothesis):
+    # Tokenize and predict
     inputs = tokenizer(premise, hypothesis, return_tensors="pt", truncation=True)
     outputs = model(**inputs)
     prediction = outputs.logits.softmax(-1)[0]
-    return {"Entailment": prediction[0].item(),
-            "Neutral": prediction[1].item(),
-            "Contradiction": prediction[2].item()}
+    
+    # Return results
+    return {
+        "Entailment": float(prediction[0]),
+        "Neutral": float(prediction[1]),
+        "Contradiction": float(prediction[2])
+    }
 
+# Create interface
 demo = gr.Interface(
     fn=predict,
-    inputs=["text", "text"],
-    outputs="label",
-    title="NLI Classifier"
+    inputs=[
+        gr.Textbox(label="Premise"),
+        gr.Textbox(label="Hypothesis")
+    ],
+    outputs=gr.Label(),
+    title="NLI Classifier",
+    examples=[
+        ["The cat is sleeping.", "The cat is awake."],
+        ["It's raining.", "The ground is wet."]
+    ]
 )
 
 demo.launch()