A programming framework for agentic AI
cleanup
Chi Wang0c60bc51118d
diff --git a/autogen/code_utils.py b/autogen/code_utils.pyindex bcc9be1a6..88ce23195 100644--- a/autogen/code_utils.py+++ b/autogen/code_utils.py@@ -444,7 +444,7 @@ def eval_function_completions(# assertion generatorassertions, gen_cost = assertions(definition)else:- gen_cost = 0+ assertions, gen_cost = None, 0if n > 1 or test is None:for i in range(n):response = responses[i] = _remove_check(responses[i])diff --git a/autogen/oai/openai_utils.py b/autogen/oai/openai_utils.pyindex 9df981b56..0215eeeea 100644--- a/autogen/oai/openai_utils.py+++ b/autogen/oai/openai_utils.py@@ -75,7 +75,6 @@ def config_list_openai_aoai("""if "OPENAI_API_KEY" not in os.environ and exclude != "openai":try:- print(f"{key_file_path}/{openai_api_key_file}")with open(f"{key_file_path}/{openai_api_key_file}") as key_file:os.environ["OPENAI_API_KEY"] = key_file.read().strip()except FileNotFoundError:diff --git a/test/oai/test_completion.py b/test/oai/test_completion.pyindex 7eb28bc3e..9ae952a02 100644--- a/test/oai/test_completion.py+++ b/test/oai/test_completion.py@@ -41,9 +41,12 @@ def test_filter():except ImportError as exc:print(exc)return+ config_list = autogen.config_list_from_models(+ KEY_LOC, exclude="aoai", model_list=["text-ada-001", "gpt-3.5-turbo", "text-davinci-003"]+ )response = autogen.Completion.create(context={"yes_or_no_choice": True},- config_list=[{"model": "text-ada-001"}, {"model": "gpt-3.5-turbo"}, {"model": "text-davinci-003"}],+ config_list=config_list,prompt="Is 37 a prime number? Please answer 'Yes.' or 'No.'",filter_func=yes_or_no_filter,)@@ -54,13 +57,13 @@ def test_filter():)response = autogen.Completion.create(context={"yes_or_no_choice": False},- config_list=[{"model": "text-ada-001"}, {"model": "gpt-3.5-turbo"}, {"model": "text-davinci-003"}],+ config_list=config_list,prompt="Is 37 a prime number?",filter_func=yes_or_no_filter,)assert response["model"] == "text-ada-001"response = autogen.Completion.create(- config_list=[{"model": "text-ada-001"}, {"model": "gpt-3.5-turbo"}, {"model": "text-davinci-003"}],+ config_list=config_list,prompt="How to construct a json request to Bing API to search for 'latest AI news'? Return the JSON request.",filter_func=valid_json_filter,)@@ -123,7 +126,12 @@ def test_nocontext():print(exc)returnresponse = autogen.Completion.create(- model="text-ada-001", prompt="1+1=", max_tokens=1, use_cache=False, request_timeout=10+ model="text-ada-001",+ prompt="1+1=",+ max_tokens=1,+ use_cache=False,+ request_timeout=10,+ config_list=autogen.config_list_openai_aoai(KEY_LOC, exclude="aoai"),)print(response)code, _ = generate_code(@@ -165,7 +173,7 @@ def test_nocontext():)def test_humaneval(num_samples=1):gpt35_config_list = autogen.config_list_from_json(- env_or_file="OAI_CONFIG_LIST",+ env_or_file=OAI_CONFIG_LIST,filter_dict={"model": {"gpt-3.5-turbo",@@ -177,6 +185,7 @@ def test_humaneval(num_samples=1):"gpt",},},+ file_location=KEY_LOC,)assertions = partial(generate_assertions, config_list=gpt35_config_list)eval_with_generated_assertions = partial(@@ -215,13 +224,14 @@ def test_humaneval(num_samples=1):# no error should be raisedresponse = autogen.Completion.create(context=test_data[0],- config_list=[{"model": "gpt-3.5-turbo"}],+ config_list=autogen.config_list_from_models(KEY_LOC, model_list=["gpt-3.5-turbo"]),prompt="",max_tokens=1,retry_timeout=0,raise_on_ratelimit_or_timeout=False,)# assert response == -1+ config_list = autogen.config_list_openai_aoai(KEY_LOC, exclude="aoai")# a minimal tuning exampleconfig, _ = autogen.Completion.tune(data=tune_data,@@ -231,8 +241,9 @@ def test_humaneval(num_samples=1):n=1,prompt="{definition}",allow_format_str_template=True,+ config_list=config_list,)- response = autogen.Completion.create(context=test_data[0], **config)+ response = autogen.Completion.create(context=test_data[0], config_list=config_list, **config)# a minimal tuning example for tuning chat completion models using the Completion classconfig, _ = autogen.Completion.tune(data=tune_data,@@ -243,8 +254,9 @@ def test_humaneval(num_samples=1):model="text-davinci-003",prompt="{definition}",allow_format_str_template=True,+ config_list=config_list,)- response = autogen.Completion.create(context=test_data[0], **config)+ response = autogen.Completion.create(context=test_data[0], config_list=config_list, **config)# a minimal tuning example for tuning chat completion models using the ChatCompletion classconfig_list = autogen.config_list_openai_aoai(KEY_LOC)config, _ = autogen.ChatCompletion.tune(@@ -297,12 +309,12 @@ def test_humaneval(num_samples=1):print(config2)print(analysis.best_result)print(test_data[0])- response = autogen.Completion.create(context=test_data[0], **config2)+ response = autogen.Completion.create(context=test_data[0], config_list=config_list, **config2)print(response)autogen.Completion.data = test_data[:num_samples]result = autogen.Completion._eval(analysis.best_config, prune=False, eval_only=True)print("result without pruning", result)- result = autogen.Completion.test(test_data[:num_samples], **config2)+ result = autogen.Completion.test(test_data[:num_samples], config_list=config_list, **config2)print(result)try:code, cost, selected = implement(@@ -364,6 +376,7 @@ def test_math(num_samples=-1):]autogen.Completion.set_cache(seed)+ config_list = autogen.config_list_openai_aoai(KEY_LOC, exclude="aoai")vanilla_config = {"model": "text-davinci-003","temperature": 1,@@ -373,11 +386,12 @@ def test_math(num_samples=-1):"stop": "###",}test_data_sample = test_data[0:3]- result = autogen.Completion.test(test_data_sample, eval_math_responses, **vanilla_config)+ result = autogen.Completion.test(test_data_sample, eval_math_responses, config_list=config_list, **vanilla_config)result = autogen.Completion.test(test_data_sample,eval_math_responses,agg_method="median",+ config_list=config_list,**vanilla_config,)@@ -418,9 +432,10 @@ def test_math(num_samples=-1):num_samples=num_samples,prompt=prompts, # the prompt templates to choose fromstop="###", # the stop sequence+ config_list=config_list,)print("tuned config", config)- result = autogen.Completion.test(test_data_sample, config_list=autogen.config_list_openai_aoai(KEY_LOC), **config)+ result = autogen.Completion.test(test_data_sample, config_list=config_list, **config)print("result from tuned config:", result)print("empty responses", eval_math_responses([], None))@@ -432,9 +447,9 @@ if __name__ == "__main__":assert len(config_list) >= 3, config_listopenai.api_key = os.environ["OPENAI_API_KEY"]- test_filter()- test_chatcompletion()- test_multi_model()- test_nocontext()+ # test_filter()+ # test_chatcompletion()+ # test_multi_model()+ # test_nocontext()test_humaneval(1)- test_math(1)+ # test_math(1)