95 lines
3.5 KiB
Python
95 lines
3.5 KiB
Python
import random
|
|
import json
|
|
import asyncio
|
|
from os import environ
|
|
import openai
|
|
from openai import OpenAIError
|
|
from weather import time_now, get_current_weather, get_weather_forecast
|
|
from tools_description import TOOLS
|
|
from debug_stuff import debug
|
|
|
|
CALLABLE_FUNCTIONS = {
|
|
# Dictionary with functions to call.
|
|
# You can use globals()[func_name](**args) instead, but that's too implicit.
|
|
'time_now': time_now,
|
|
'get_current_weather': get_current_weather,
|
|
'get_weather_forecast': get_weather_forecast,
|
|
}
|
|
|
|
|
|
class Completion:
|
|
def __init__(self, model: str, api_key: str):
|
|
self.__model = model
|
|
self.__api_key = api_key
|
|
self.__messages = []
|
|
|
|
async def create_completion(self):
|
|
model = self.__model
|
|
try:
|
|
client = openai.AsyncClient(api_key=self.__api_key)
|
|
completion_kwargs = {
|
|
"model": model,
|
|
"messages": self.__messages,
|
|
"max_tokens": 4096,
|
|
"temperature": 0.7,
|
|
"presence_penalty": 0.5,
|
|
"frequency_penalty": 0.5,
|
|
"tools": TOOLS,
|
|
"tool_choice": "auto",
|
|
}
|
|
response = await client.chat.completions.create(**completion_kwargs)
|
|
response_content = response.choices[0].message.content
|
|
tool_calls = response.choices[0].message.tool_calls
|
|
self.append_message(role="assistant", content=response_content, tool_calls=tool_calls)
|
|
if tool_calls:
|
|
for i_call in tool_calls:
|
|
func_name = i_call.function.name
|
|
func_args = json.loads(i_call.function.arguments)
|
|
tool_call_id = i_call.id
|
|
self.function_manager(func_name, func_args, tool_call_id)
|
|
return await self.create_completion()
|
|
return response_content
|
|
except OpenAIError as e:
|
|
return self.get_error_message(error_message=str(e), error_type="OpenAIError")
|
|
|
|
def append_message(
|
|
self,
|
|
role: str,
|
|
content: str,
|
|
tool_calls: list = None,
|
|
tool_call_id: str = None,
|
|
):
|
|
self.__messages.append({
|
|
"role": role,
|
|
"content": content,
|
|
"tool_calls": tool_calls,
|
|
"tool_call_id": tool_call_id,
|
|
})
|
|
|
|
@staticmethod
|
|
def get_error_message(error_message: str, error_type: str) -> str:
|
|
reginald_responses = [
|
|
f"Regrettably, I must inform you that I have encountered a bureaucratic obstruction:",
|
|
f"It would seem that a most unfortunate technical hiccup has befallen my faculties:",
|
|
f"Ah, it appears I have received an urgent memorandum stating:",
|
|
f"I regret to inform you that my usual eloquence is presently obstructed by an unforeseen complication:",
|
|
]
|
|
random_response = random.choice(reginald_responses)
|
|
return f"{random_response}\n\n{error_type}: {error_message}"
|
|
|
|
def function_manager(self, func_name: str, func_kwargs: dict, tool_call_id: str):
|
|
result = CALLABLE_FUNCTIONS[func_name](**func_kwargs)
|
|
self.append_message(role="tool", content=result, tool_call_id=tool_call_id)
|
|
|
|
|
|
if __name__ == '__main__':
|
|
async def main():
|
|
test_message = input('Your input: ')
|
|
completion = Completion(model='gpt-4.1-mini', api_key=environ.get('OPENAI_API_KEY'))
|
|
completion.append_message(role='user', content=test_message)
|
|
result = await completion.create_completion()
|
|
print(result)
|
|
|
|
asyncio.run(main())
|
|
|