| import os |
| import json |
| import re |
| import base64 |
| from pathlib import Path |
| from openai import OpenAI |
| from PIL import Image |
| import numpy as np |
| import torch |
| from google import genai |
| try: |
| from colorama import Fore, Style, init |
| except Exception: |
| class _NoColor: |
| def __getattr__(self, _): |
| return "" |
| Fore = Style = _NoColor() |
| def init(*a, **k): |
| return None |
| from time import sleep |
|
|
| |
| client = genai.Client(api_key=os.environ.get("GEMINI_API_KEY") or os.environ.get("GOOGLE_API_KEY") or "EMPTY") |
|
|
| def encode_image_base64(image_path: str) -> str: |
| with open(image_path, "rb") as f: |
| return base64.b64encode(f.read()).decode("utf-8") |
|
|
| def transform_data_format(list:list) -> str: |
| |
| input = [] |
|
|
| for item in list: |
| dict = {} |
| if 'image' in item and isinstance(item['image'], str): |
| image = encode_image_base64(item['image']) |
| dict['inline_data'] = {"mime_type": "image/png", "data": image} |
| print(f"{Fore.YELLOW} Encoded image from path {Style.RESET_ALL} {item['image']}") |
| elif 'image_base64' in item: |
| image = item['image_base64'] |
| dict['inline_data'] = {"mime_type": "image/png", "data": image} |
| elif 'text' in item: |
| dict['text'] = item['text'] |
| print(f"{Fore.YELLOW} Added text {Style.RESET_ALL} {item['text']}") |
| input.append(dict) |
|
|
| return input |
|
|
| |
| |
| def generate_content(model_name, content): |
| input_data = transform_data_format(content) |
|
|
| for attempt in range(5): |
| try: |
| response = client.models.generate_content( |
| model=model_name, |
| contents={ |
| "role": "user", |
| "parts": input_data |
| } |
| ) |
| print(f"{Fore.YELLOW} The response from the model is: {Style.RESET_ALL}", response.text) |
| sleep(1) |
| return response.text |
|
|
| except: |
| print(f"{Fore.RED}503 服务端过载,等待重试(第 {attempt+1} 次)...{Style.RESET_ALL}") |
| sleep(2 ** attempt) |
| continue |
|
|
|
|
| print(f"{Fore.RED}重试次数已用完,仍然 503 过载{Style.RESET_ALL}") |
| return "error" |