provider/OpenAIProvider.py

import configparser
import time
from typing import Any, Dict, List
from datetime import datetime

from utils.dict_utils import get_with_warning


class OpenAIProvider:
    
    def __init__(self, params) -> None:
        self.params = params
        
        self.model_path = get_with_warning(params, 'model')
        self.openai_api_key = get_with_warning(params, 'api_key')
        self.openai_api_base = get_with_warning(params, 'base_url')
  
        self.image_prompt_format = "openai"
        
        # Generation config
        self.top_p = float(get_with_warning(params, 'top_p', default=0.9))
        self.temperature = float(get_with_warning(params, 'temperature', default=0.8))
        self.max_new_tokens = int(get_with_warning(params, 'max_new_tokens', default=2048))
        
        self.prompt_tokens = 0
        self.completion_tokens = 0
        self.total_tokens = 0
    
    def reset(self):
        pass

    
    def get_tokens_usage(self):
        return {
            "prompt_tokens": self.prompt_tokens,
            "completion_tokens": self.completion_tokens,
            "total_tokens": self.total_tokens
        }
    
    def create_completion(self, message_prompts):
        
        print(datetime.now().strftime('%Y-%m-%d %H:%M:%S.%f')[:-3], f"  {self.model_path} create local completion...")
        
        from openai import OpenAI
        
        try:
            client = OpenAI(
                api_key=self.openai_api_key,
                base_url=self.openai_api_base,
            )

            chat_response = client.chat.completions.create(
                model=self.model_path,
                messages=message_prompts,
                temperature=self.temperature,
                top_p=self.top_p,
                max_tokens=self.max_new_tokens
            )
            
            response_text = chat_response.choices[0].message.content
            
            # print(chat_response)
            prompt_tokens = chat_response.usage.prompt_tokens
            completion_tokens = chat_response.usage.completion_tokens
            # 输出token使用统计
            total_tokens = prompt_tokens + completion_tokens
            print(f"Token Usage - Prompt Tokens: {prompt_tokens}, Completion Tokens: {completion_tokens}, Total Tokens: {total_tokens}")
            
            self.prompt_tokens += prompt_tokens
            self.completion_tokens += completion_tokens
            self.total_tokens += total_tokens
            
            return True, response_text    
            
        except Exception as e:
            
            print(time.strftime('%Y-%m-%d %H:%M:%S',time.localtime(time.time())), e)
            print(self.model_path)
            print(self.openai_api_key)
            print(self.openai_api_base)
            print(self.top_p)
            print(self.temperature)
            time.sleep(5)
            self.reset()
            
            return False, str(e)
                
        # print(f'response: {response_text}')