Update app.py

app.py

@@ -18,7 +18,17 @@ model = AutoModelForCausalLM.from_pretrained(model_name)
 # Define inference function
 def generate_text(input_text):
     inputs = tokenizer(input_text, return_tensors="pt")
-    outputs = model.generate(inputs["input_ids"], max_length=256, num_return_sequences=1)
+    #outputs = model.generate(inputs["input_ids"], max_length=256, num_return_sequences=1)
+    output = model.generate(
+    inputs["input_ids"],
+    num_return_sequences=1,
+    max_length=256,            # Set max length for output
+    temperature=0.8,           # Control randomness (higher is more random)
+    top_k=50,                  # Top-k sampling to limit vocabulary to top 50 choices
+    top_p=0.95,                # Nucleus sampling to choose tokens with 95% cumulative probability
+    repetition_penalty=1.2,    # Penalize repetition; increase if repetitions persist
+    do_sample=True             # Enable sampling for non-deterministic output
+    )
     response = tokenizer.decode(outputs[0], skip_special_tokens=True)
     return response