add gpt-4 result

RESULTS

@@ -55,6 +55,7 @@ gpt-3.5-turbo-0301.run2     0.4551
 gpt-3.5-turbo-0301          0.4038
 text-curie-001              0.1217
 text-davinci-003            0.2884
+gpt-4                       0.7788
 
 
 FrenchMedMCQA: A French Multiple-Choice Question Answering Dataset for Medical domain

run_api.py

@@ -5,7 +5,7 @@ import backoff
 openai.api_key = OPENAI_TOKEN
 
 openai_lm_models = ['text-davinci-003', 'text-davinci-002', 'code-davinci-002', 'code-cushman-001', 'text-curie-001', 'text-babbage-001', 'text-ada-001', 'davinci', 'curie', 'babage', 'ada']
-openai_chat_models = ['gpt-3.5-turbo-0301', 'gpt-3.5-turbo']
+openai_chat_models = ['gpt-3.5-turbo-0301', 'gpt-3.5-turbo', 'gpt-4', 'gpt-4-0314', 'gpt-4-32k', 'gpt-4-32k-0314']
 
 @backoff.on_exception(backoff.expo, (openai.error.RateLimitError, openai.error.APIError))
 def openai_complete(prompt, model='text-davinci-003'):
@@ -40,6 +40,7 @@ def main(result_path: str, corpus_path: str, model: str = 'openai/gpt-3.5-turbo-
     def generate(input_string):
         if api == 'openai':
             if llm in openai_chat_models:
+                #print('This is a chat model')
                 return openai_chat(input_string, llm)
             else:
                 return openai_complete(input_string, llm)

runs

@@ -8,6 +8,8 @@ python run_opt-iml.py output/opt-iml-30b_prompt0.txt ../json/dev.json opt-iml-30
 
 python run_api.py output/code-cushman-001_prompt0.txt ../json/dev.json openai/code-cushman-001 | tee logs/code-cushman-001_prompt0.txt
 python run_api.py output/code-davinci-002_prompt0.txt ../json/dev.json openai/code-davinci-002 | tee logs/code-davinci-002_prompt0.txt
+python run_api.py output/gpt-4-0314_prompt0.txt ../json/dev.json openai/gpt-4-0314 | tee logs/gpt-4-0314_prompt0.txt
+
 python run_api.py output/j1-jumbo_prompt0.txt ../json/dev.json ai21/j1-jumbo | tee logs/j1-jumbo_prompt0.txt
 
 python run_bloomz.py output/en_bloomz-560m_prompt0.txt ../json/dev-en.json bloomz-560m en/0 | tee logs/en_bloomz-560m_prompt0.txt