update

app.py

@@ -78,7 +78,7 @@ def rag_pipeline(prompt, index="pubmed", num_docs=3, model_name="HuggingFaceH4/z
     for message in messages:
         print('MSG', message)
 
-    max_new_tokens = 512
+    max_new_tokens = 1024
 
     if model_name.startswith('openai/'):
         openai_model_name = model_name.split('/')[1]
@@ -107,7 +107,7 @@ def rag_pipeline(prompt, index="pubmed", num_docs=3, model_name="HuggingFaceH4/z
 
         # Define the generation_kwargs with stopping criteria
         generation_kwargs = {
-            # "max_new_tokens": max_new_tokens,
+            "max_new_tokens": max_new_tokens,
             "generation_kwargs": {"stopping_criteria": stopping_criteria},
             "return_full_text": False
         }