Unixcoder finetuned returns only a .bin file called model.bin, how to use it for inference?

Hi everyone,
i’m a beginner in this community, trying to use bert based model UnixCoder in order to create embeddings that i will use in searching similiar source code snippets.
I already started using this model without any problem, but yesterday i tried to fine tune it on a dataset called “Poj-104”, i executed the script that you can find here. The output of this script is a single file called “model.bin”. I tried to use it, the same way, i used to use other pretrained model with this code below:
self.tokenizer = RobertaTokenizer.from_pretrained(“microsoft/unixcoder-base”)
self.model = RobertaModel.from_pretrained(local_name,local_files_only=True)
but it returns this error " does not appear to have a file named config.json in the directory".
Can you help me please with this problem, thank you in advance.