Update app.py

app.py

@@ -10,8 +10,8 @@ from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStream
 
 HF_TOKEN = "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"  ##Llama Write Token
 
-MAX_MAX_NEW_TOKENS = 2048
-DEFAULT_MAX_NEW_TOKENS = 1024
+MAX_MAX_NEW_TOKENS = 8192
+DEFAULT_MAX_NEW_TOKENS = 4096
 MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
 
 DESCRIPTION = """\
@@ -33,7 +33,7 @@ def generate(
     message: str,
     chat_history: list[tuple[str, str]],
     system_prompt: str,
-    max_new_tokens: int = 1024,
+    max_new_tokens: int = 8192,
     temperature: float = 0.6,
     top_p: float = 0.9,
     top_k: int = 50,
@@ -95,8 +95,8 @@ def mask_with_protecto(text_for_prompt):
         # Parse the masked result from the API response and format it for display
         masked_result = response.json()
         final_result = json.dumps(masked_result, indent=4)
-        # return(str(masked_result["data"][0]["token_value"]))
-        return(str(masked_result))
+        return_value = str(masked_result["data"][0]["token_value"])
+        return(return_value)
     else:
         # Return an error message if the API request was not successful.
         return(str(response.status_code))