Create app.py

app.py

@@ -0,0 +1,48 @@
+import gradio as gr
+from transformers import AutoTokenizer, AutoModelForSequenceClassification, pipeline
+
+# Hugging Face model path
+MODEL_NAME = "umarfarzan/clipworthy-deberta-model"
+
+# Load tokenizer & model
+tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
+model = AutoModelForSequenceClassification.from_pretrained(MODEL_NAME)
+
+# Create pipeline
+classifier = pipeline(
+    "text-classification",
+    model=model,
+    tokenizer=tokenizer,
+    device=-1  # CPU; set to 0 for GPU
+)
+
+# Function to predict and classify as clipworthy/not clipworthy
+def predict_clipworthiness(text):
+    if not text.strip():
+        return {"error": "No text provided"}
+    
+    # Get raw model prediction
+    result = classifier(text, truncation=True, max_length=256)
+    
+    # Extract score (assumes result is a list with a dict containing 'label' and 'score')
+    score = result[0]['score']
+    
+    # Return "clipworthy" if score > 0.974, else "not clipworthy"
+    label = "clipworthy" if score > 0.971 else "not clipworthy"
+    
+    return {"label": label, "score": score}
+
+# Gradio interface
+iface = gr.Interface(
+    fn=predict_clipworthiness,
+    inputs=gr.Textbox(
+        label="Transcript Text",
+        placeholder="Paste transcript here..."
+    ),
+    outputs=gr.JSON(label="Prediction"),
+    title="Clipworthy Classifier",
+    description="Paste transcript text and get classification results."
+)
+
+if __name__ == "__main__":
+    iface.launch()