Update app.py

app.py

@@ -101,14 +101,14 @@ def generate_translation(input_text, model, tokenizer):
         # Set up the streamer for real-time output
         streamer = TextStreamer(tokenizer, skip_special_tokens=True)
         
-        # Generate translation with streaming
+        # Generate translation with streaming (disable beam search)
         model.eval()
         with torch.no_grad():
             outputs = model.generate(
                 input_ids=inputs["input_ids"],
                 attention_mask=inputs["attention_mask"],
                 max_length=512,
-                num_beams=5,
+                num_beams=1,  # Set to 1 to disable beam search for streaming
                 length_penalty=1.0,
                 early_stopping=True,
                 streamer=streamer,