I’m trying to follow the fine tuning guide, OpenAI API and am able to produce a JSONL file by running the following command in Gitbash:
openai tools fine_tunes.prepare_data -f out.csv
The output of that command helpfully gives me another command, to “use that file when fine-tuning”:
openai api fine_tunes.create -t “out_prepared.jsonl” --no_packing --batch_size 0.25
However, that command fails, at first because --batch size is supposed to be an integer, but then because I get “openai.error.InvalidRequestError: No such File object: out_prepared.jsonl”. The error output also says something about "UnicodeDecodeError: ‘charmap’ codec can’t decode byte 0x9d in position 1546: character maps to "
Since I was able to generate the JSONL file using the provided formatting tool, I don’t understand why the command it recommends I use next fails.
My command succeeded, but then seems to be having problems with file encoding when printing about the success to the console.
[2021-12-20 14:45:58] Fine-tune succeeded
Traceback (most recent call last):
File “C:\Users\mkessler\AppData\Roaming\Python\Python310\Scripts\openai”, line 72, in
sys.exit(main())
File “C:\Users\mkessler\AppData\Roaming\Python\Python310\Scripts\openai”, line 61, in main
args.func(args)
File “C:\Users\mkessler\AppData\Roaming\Python\Python310\site-packages\openai\cli.py”, line 364, in follow
cls._stream_events(args.id)
File “C:\Users\mkessler\AppData\Roaming\Python\Python310\site-packages\openai\cli.py”, line 407, in _stream_events
sys.stdout.write("\nJob complete! Status: succeeded \U0001f389")
File “C:\Python310\lib\encodings\cp1252.py”, line 19, in encode
return codecs.charmap_encode(input,self.errors,encoding_table)[0]
UnicodeEncodeError: ‘charmap’ codec can’t encode character ‘\U0001f389’ in position 34: character maps to