change number of default annotations to 1

inference.py

@@ -12,7 +12,7 @@ def prepare_inputs(protein_tokenizer, term_tokenizer, aspect, protein_sequence):
 
 
 def run_inference(model, inputs, prompt, num_annotations):
-    outputs = model.generate(**inputs, max_new_tokens=20, num_return_sequences=num_annotations,
+    outputs = model.generate(**inputs, max_new_tokens=20, num_return_sequences=1,
                             return_dict_in_generate=True, output_scores=True, renormalize_logits=True,
                             **prompt)