Update main.py

main.py

@@ -79,7 +79,7 @@ async def inference(request: InferenceRequest):
         inferenceTokenizer.encode(input_text))
 
     results = inferenceTranslator.translate_batch(
-        [input_tokens], max_input_length=0, max_decoding_length=max_length, num_hypotheses=1, repetition_penalty=1.3, sampling_topk=40, sampling_temperature=0.2, use_vmap=False)
+        [input_tokens], beam_size=1, max_input_length=0, max_decoding_length=max_length, num_hypotheses=1, repetition_penalty=1.3, sampling_topk=40, sampling_temperature=0.7, use_vmap=False)
 
     output_tokens = results[0].hypotheses[0]
     output_text = inferenceTokenizer.decode(