setting max model length to something reasonable, e.g. 4096

tokenizer_config.json

@@ -1746,7 +1746,7 @@
   "bos_token": "<bos>",
   "clean_up_tokenization_spaces": false,
   "eos_token": "<eos>",
-  "model_max_length": 1000000000000000019884624838656,
+  "model_max_length": 4096,
   "pad_token": "<pad>",
   "sp_model_kwargs": {},
   "spaces_between_special_tokens": false,