`FineTune.list_events()` breaks when `stream=True`

[Yun-Kim]

Describe the bug

When I tried to list the events for my fine tune job through the Python API with streaming I get a correct response without error. However when I specify stream=True, this results in an exception being raised and crashing my application, even though the response was clearly received. Looking at the traceback, I can see that the response was received but failed to be parsed into JSON by openai.

Here's a traceback for the exception that I see:

Traceback (most recent call last):
  File "venv_py3105/lib/python3.10/site-packages/openai/api_requestor.py", line 677, in _interpret_response_line
    data = json.loads(rbody)
  File "/Users/yun.kim/.pyenv/versions/3.10.5/lib/python3.10/json/__init__.py", line 346, in loads
    return _default_decoder.decode(s)
  File "/Users/yun.kim/.pyenv/versions/3.10.5/lib/python3.10/json/decoder.py", line 337, in decode
    obj, end = self.raw_decode(s, idx=_w(s, 0).end())
  File "/Users/yun.kim/.pyenv/versions/3.10.5/lib/python3.10/json/decoder.py", line 355, in raw_decode
    raise JSONDecodeError("Expecting value", s, err.value) from None
json.decoder.JSONDecodeError: Expecting value: line 1 column 1 (char 0)

The above exception was the direct cause of the following exception:

File "openai_demo.py", line 216, in fine_tune_list_events
    resp = openai.FineTune.list_events(
  File "venv_py3105/lib/python3.10/site-packages/openai/api_resources/abstract/nested_resource_class_methods.py", line 123, in list_nested_resources
    return getattr(cls, resource_request_method)("get", url, **params)
  File "venv_py3105/lib/python3.10/site-packages/openai/api_resources/abstract/nested_resource_class_methods.py", line 44, in nested_resource_request
    response, _, api_key = requestor.request(
  File "venv_py3105/lib/python3.10/site-packages/openai/api_requestor.py", line 230, in request
    resp, got_stream = self._interpret_response(result, stream)
  File "venv_py3105/lib/python3.10/site-packages/openai/api_requestor.py", line 624, in _interpret_response
    self._interpret_response_line(
  File "venv_py3105/lib/python3.10/site-packages/openai/api_requestor.py", line 680, in _interpret_response_line
    raise error.APIError(
openai.error.APIError: HTTP code 200 from API (data: {"object": "fine-tune-event", "level": "info", "message": "Created fine-tune: ft-XXXXXXX", "created_at": 1684155377}

data: {"object": "fine-tune-event", "level": "info", "message": "Fine-tune costs $X.XX", "created_at": 1684155451}

data: {"object": "fine-tune-event", "level": "info", "message": "Fine-tune enqueued. Queue number: 1", "created_at": 1684155452}

data: {"object": "fine-tune-event", "level": "info", "message": "Fine-tune is in the queue. Queue number: 0", "created_at": 1684155452}

data: {"object": "fine-tune-event", "level": "info", "message": "Fine-tune started", "created_at": 1684155503}

data: {"object": "fine-tune-event", "level": "info", "message": "Completed epoch 1/4", "created_at": 1684155598}

data: {"object": "fine-tune-event", "level": "info", "message": "Completed epoch 2/4", "created_at": 1684155603}

data: {"object": "fine-tune-event", "level": "info", "message": "Completed epoch 3/4", "created_at": 1684155607}

data: {"object": "fine-tune-event", "level": "info", "message": "Completed epoch 4/4", "created_at": 1684155615}

data: {"object": "fine-tune-event", "level": "info", "message": "Uploaded model: davinci:ft-XXXXXXX-2023-05-15-13-00-56", "created_at": 1684155656}

data: {"object": "fine-tune-event", "level": "info", "message": "Uploaded result file: file-XXXXXXX", "created_at": 1684155658}

data: {"object": "fine-tune-event", "level": "info", "message": "Fine-tune succeeded", "created_at": 1684155658}

data: [DONE]

)

To Reproduce

I am using Python 3.10.5 with openai==0.27.6.

Here is my application code:

resp = openai.FineTune.list_events(
  "ft-XXXXX",  # fine tune job ID
  api_key=os.getenv("OPENAI_API_KEY"),
  stream=True
)

Code snippets

No response

OS

macOS

Python version

Python 3.10.5

Library version

v0.27.6

[LX-sys]

Have you solved the problem?

[RobertCraigie]

Our upcoming v1 beta supports streaming fine tunes events but it should be noted that this API is a legacy API and you should start using fine_tuning.jobs instead.

from openai import OpenAI

client = OpenAI()

stream = client.fine_tunes.list_events('id', stream=True)
for event in stream:
  print(event)