Update app.py

app.py

@@ -18,6 +18,7 @@ def generate_text(prompt):
   output = model.generate(
       input_ids=input_ids,
       max_length=256,
+      num_beams=1,
       num_return_sequences=1,  # Generate only 1 sequence
       do_sample=True,  # Enable sampling for creativity
       top_k=50,  # Sample from the top 50 most likely tokens at each step
@@ -28,7 +29,6 @@ def generate_text(prompt):
 
   # Decode the generated token sequence back to text
   generated_text = tokenizer.decode(output[0], skip_special_tokens=True)
-  generated_text = generate_text.replace(prompt, "").strip()
   return generated_text
 
 # Define the Gradio interface