AttributeError: 'NoneType' object has no attribute 'lower'

[ganchengguang]

When I run OPENAI_API_KEY=sk- python llm_judge/gen_judgment.py --mode pairwise-baseline --model-list llama2-7b-chat-

I get following error. Any suggestions or clue? Openai==0.28.1
Thank you very much.

| 2024-05-26 23:44:18 | INFO | Load questions
| 2024-05-26 23:44:19 | INFO | Load answers
| 2024-05-26 23:44:19 | INFO | Load reference answers
| 2024-05-26 23:44:19 | INFO | Load judge prompts
| 2024-05-26 23:44:19 | INFO | Make matches
| 2024-05-26 23:44:19 | INFO | Mode: pairwise-baseline
| 2024-05-26 23:44:19 | INFO | Judge model: gpt-4
| 2024-05-26 23:44:19 | INFO | Baseline model: openai--text-davinci-003
| 2024-05-26 23:44:19 | INFO | Total number of questions: 80
| 2024-05-26 23:44:19 | INFO | Total number of matches: 80
| 2024-05-26 23:44:19 | INFO | Total cost (estimated): $4
| 2024-05-26 23:44:19 | INFO | Output directory: /media/gan/2model/RIKEN_instruction/nejumi/ja-vicuna-qa-benchmark-main/data/jp_bench/model_judgment/pairwise/gpt-4
Press Enter to confirm...
| 2024-05-26 23:44:20 | INFO | Play matches
0%| | 0/80 [00:00<?, ?it/s]
Traceback (most recent call last):
File "/media/gan/2model/RIKEN_instruction/nejumi/ja-vicuna-qa-benchmark-main/llm_judge/gen_judgment.py", line 281, in
results.append(future.result())
File "/home/gan/anaconda3/envs/py39pytorch2cuda118/lib/python3.9/concurrent/futures/_base.py", line 433, in result
return self.__get_result()
File "/home/gan/anaconda3/envs/py39pytorch2cuda118/lib/python3.9/concurrent/futures/_base.py", line 389, in __get_result
raise self._exception
File "/home/gan/anaconda3/envs/py39pytorch2cuda118/lib/python3.9/concurrent/futures/thread.py", line 52, in run
result = self.fn(*self.args, **self.kwargs)
File "/media/gan/2model/RIKEN_instruction/nejumi/ja-vicuna-qa-benchmark-main/llm_judge/common.py", line 184, in play
g1_judgment = play(self.answer_1, self.answer_2)
File "/media/gan/2model/RIKEN_instruction/nejumi/ja-vicuna-qa-benchmark-main/llm_judge/common.py", line 182, in play
return self.judge.judge(**kwargs)
File "/media/gan/2model/RIKEN_instruction/nejumi/ja-vicuna-qa-benchmark-main/llm_judge/common.py", line 71, in judge
response = openai.ChatCompletion.create(**params)
File "/home/gan/anaconda3/envs/py39pytorch2cuda118/lib/python3.9/site-packages/openai/api_resources/chat_completion.py", line 25, in create
return super().create(*args, **kwargs)
File "/home/gan/anaconda3/envs/py39pytorch2cuda118/lib/python3.9/site-packages/openai/api_resources/abstract/engine_api_resource.py", line 151, in create
) = cls.__prepare_create_request(
File "/home/gan/anaconda3/envs/py39pytorch2cuda118/lib/python3.9/site-packages/openai/api_resources/abstract/engine_api_resource.py", line 82, in __prepare_create_request
typed_api_type = cls._get_api_type_and_version(api_type=api_type)[0]
File "/home/gan/anaconda3/envs/py39pytorch2cuda118/lib/python3.9/site-packages/openai/api_resources/abstract/api_resource.py", line 169, in _get_api_type_and_version
else ApiType.from_str(openai.api_type)
File "/home/gan/anaconda3/envs/py39pytorch2cuda118/lib/python3.9/site-packages/openai/util.py", line 35, in from_str
if label.lower() == "azure":
AttributeError: 'NoneType' object has no attribute 'lower'