File size: 11,816 Bytes
deb4f9b |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 |
import uuid
import json
import time
import os
import re
import random
import requests
from flask import Flask, request, Response, stream_with_context, jsonify
from flask_cors import CORS
from functools import lru_cache
from concurrent.futures import ThreadPoolExecutor
import base64
app = Flask(__name__)
CORS(app, resources={r"/*": {"origins": "*"}})
executor = ThreadPoolExecutor(max_workers=10)
# 使用LRU缓存,最大容量为10,用于缓存文件内容
@lru_cache(maxsize=10)
def read_file(filename):
try:
with open(filename, 'r') as f:
return f.read().strip()
except FileNotFoundError:
print(f"文件 {filename} 未找到")
return ""
except Exception as e:
print(f"读取文件 {filename} 时发生错误: {e}")
return ""
# 从环境变量或文件中获取信息
def get_env_or_file(env_var, filename):
return os.environ.get(env_var) or read_file(filename)
NOTDIAMOND_URLS = [
'https://chat.notdiamond.ai',
'https://chat.notdiamond.ai/mini-chat'
]
# 从预设的URL列表中随机选择一个
def get_notdiamond_url():
return random.choice(NOTDIAMOND_URLS)
# 模型信息字典,包含各模型的提供商和映射
MODEL_INFO = {
"gpt-4-turbo-2024-04-09": {
"provider": "openai",
"mapping": "gpt-4-turbo-2024-04-09"
},
"gemini-1.5-pro-exp-0801": {
"provider": "google",
"mapping": "models/gemini-1.5-pro-exp-0801"
},
"Meta-Llama-3.1-70B-Instruct-Turbo": {
"provider": "togetherai",
"mapping": "meta.llama3-1-70b-instruct-v1:0"
},
"Meta-Llama-3.1-405B-Instruct-Turbo": {
"provider": "togetherai",
"mapping": "meta.llama3-1-405b-instruct-v1:0"
},
"llama-3.1-sonar-large-128k-online": {
"provider": "perplexity",
"mapping": "llama-3.1-sonar-large-128k-online"
},
"gemini-1.5-pro-latest": {
"provider": "google",
"mapping": "models/gemini-1.5-pro-latest"
},
"claude-3-5-sonnet-20240620": {
"provider": "anthropic",
"mapping": "anthropic.claude-3-5-sonnet-20240620-v1:0"
},
"claude-3-haiku-20240307": {
"provider": "anthropic",
"mapping": "anthropic.claude-3-haiku-20240307-v1:0"
},
"gpt-4o-mini": {
"provider": "openai",
"mapping": "gpt-4o-mini"
},
"gpt-4o": {
"provider": "openai",
"mapping": "gpt-4o"
},
"mistral-large-2407": {
"provider": "mistral",
"mapping": "mistral.mistral-large-2407-v1:0"
}
}
# 生成系统指纹,用于追踪和安全目的
def generate_system_fingerprint():
return f"fp_{uuid.uuid4().hex[:10]}"
# 创建OpenAI格式的分块内容
def create_openai_chunk(content, model, finish_reason=None, usage=None):
system_fingerprint = generate_system_fingerprint()
chunk = {
"id": f"chatcmpl-{uuid.uuid4()}",
"object": "chat.completion.chunk",
"created": int(time.time()),
"model": model,
"system_fingerprint": system_fingerprint,
"choices": [
{
"index": 0,
"delta": {"content": content} if content else {},
"logprobs": None,
"finish_reason": finish_reason
}
]
}
if usage is not None:
chunk["usage"] = usage
return chunk
# 处理带美元符号的字符串
def process_dollars(s):
return s.replace('$$', '$')
def stream_notdiamond_response(response, model):
buffer = ""
last_content = ""
total_tokens = 0
for chunk in response.iter_content(chunk_size=1024):
if chunk:
buffer += chunk.decode('utf-8')
lines = buffer.split('\n')
buffer = lines.pop()
for line in lines:
if line.strip():
match = re.match(r'^([0-9a-zA-Z]+):(.*)$', line)
if match:
try:
_, content = match.groups()
data = json.loads(content)
content = ''
if 'output' in data and 'curr' in data['output']:
content = process_dollars(data['output']['curr'])
elif 'curr' in data:
content = process_dollars(data['curr'])
elif 'diff' in data and isinstance(data['diff'], list):
if len(data['diff']) > 1:
new_content = process_dollars(data['diff'][1])
content = last_content + new_content
elif len(data['diff']) == 1:
content = last_content
if content:
total_tokens += len(content.split())
last_content = content
yield create_openai_chunk(content, model)
except json.JSONDecodeError:
print(f"Error processing line: {line}") # 可考虑替换为日志记录
usage = {
"prompt_tokens": 0,
"completion_tokens": total_tokens,
"total_tokens": total_tokens
}
yield create_openai_chunk('', model, 'stop', usage=usage)
# 处理非流式响应
def handle_non_stream_response(response, model):
full_content = ""
for chunk in stream_notdiamond_response(response, model):
if chunk['choices'][0]['delta'].get('content'):
full_content += chunk['choices'][0]['delta']['content']
return jsonify({
"id": f"chatcmpl-{uuid.uuid4()}",
"object": "chat.completion",
"created": int(time.time()),
"model": model,
"system_fingerprint": generate_system_fingerprint(),
"choices": [
{
"index": 0,
"message": {
"role": "assistant",
"content": full_content
},
"finish_reason": "stop"
}
],
"usage": {
"prompt_tokens": len(full_content) // 4,
"completion_tokens": len(full_content) // 4,
"total_tokens": len(full_content) // 2
}
})
# 生成流式响应
def generate_stream_response(response, model):
for chunk in stream_notdiamond_response(response, model):
yield f"data: {json.dumps(chunk)}\n\n"
yield "data: [DONE]\n\n"
# 获取模型列表的API
@app.route('/ai/v1/models', methods=['GET'])
def proxy_models():
models = [
{
"id": model_id,
"object": "model",
"provider": info["provider"]
} for model_id, info in MODEL_INFO.items()
]
return jsonify({
"object": "list",
"data": models
})
# 缓存cookie和过期时间
cached_cookie = None
expires_at = 0
# 获取cookie
def get_cookie(email, password):
global cached_cookie, expires_at
current_time = int(time.time())
if cached_cookie and expires_at > current_time:
return cached_cookie
auth_url = "https://spuckhogycrxcbomznwo.supabase.co/auth/v1/token?grant_type=password"
auth_headers = {
'accept': '*/*',
'accept-language': 'zh-CN,zh;q=0.9',
'apikey': 'eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpc3MiOiJzdXBhYmFzZSIsInJlZiI6InNwdWNraG9neWNyeGNib216bndvIiwicm9sZSI6ImFub24iLCJpYXQiOjE3MDcyNDYwMzksImV4cCI6MjAyMjgyMjAzOX0.tvlGT7NZY8bijMjNIu1WhAtPnSKuDeYhtveo4DRt6xg',
'authorization': 'Bearer eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpc3MiOiJzdXBhYmFzZSIsInJlZiI6InNwdWNraG9neWNyeGNib216bndvIiwicm9sZSI6ImFub24iLCJpYXQiOjE3MDcyNDYwMzksImV4cCI6MjAyMjgyMjAzOX0.tvlGT7NZY8bijMjNIu1WhAtPnSKuDeYhtveo4DRt6xg',
'content-type': 'application/json;charset=UTF-8',
'origin': 'https://chat.notdiamond.ai',
'priority': 'u=1, i',
'referer': 'https://chat.notdiamond.ai/',
'sec-ch-ua': '"Chromium";v="128", "Not;A=Brand";v="24", "Google Chrome";v="128"',
'sec-ch-ua-mobile': '?0',
'sec-ch-ua-platform': '"Windows"',
'sec-fetch-dest': 'empty',
'sec-fetch-mode': 'cors',
'sec-fetch-site': 'cross-site',
'user-agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/128.0.0.0 Safari/537.36',
'x-client-info': 'supabase-ssr/0.4.0',
'x-supabase-api-version': '2024-01-01'
}
auth_data = {
"email": email,
"password": password,
"gotrue_meta_security": {}
}
response = requests.post(auth_url, headers=auth_headers, json=auth_data)
response.raise_for_status()
auth_response = response.json()
expires_at = auth_response['expires_at']
encoded_token = base64.b64encode(json.dumps(auth_response).encode('utf-8')).decode('utf-8')
cached_cookie = f"sb-spuckhogycrxcbomznwo-auth-token=base64-{encoded_token}"
return cached_cookie
# 处理请求的API
@app.route('/ai/v1/chat/completions', methods=['POST'])
def handle_request():
try:
request_data = request.get_json()
messages = request_data.get('messages')
model_id = request_data.get('model', '')
model = MODEL_INFO.get(model_id, {}).get('mapping', model_id)
stream = request_data.get('stream', False)
# 从请求头中获取Authorization信息
auth_header = request.headers.get('Authorization')
if not auth_header or not auth_header.startswith('Bearer '):
return jsonify({'error': 'Missing or invalid Authorization header'}), 401
auth_info = auth_header.split(' ')[1]
email, password = auth_info.split('|')
# 获取cookie
cookie = get_cookie(email, password)
payload = {
"messages": messages,
"model": model,
"stream": stream,
"frequency_penalty": request_data.get('frequency_penalty', 0),
"presence_penalty": request_data.get('presence_penalty', 0),
"temperature": request_data.get('temperature', 0.8),
"top_p": request_data.get('top_p', 1)
}
headers = {
'accept': 'text/event-stream',
'accept-language': 'zh-CN,zh;q=0.9',
'content-type': 'application/json',
'next-action': get_env_or_file('NEXT_ACTION', 'next_action.txt'),
'user-agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/128.0.0.0 Safari/537.36',
'cookie': cookie
}
url = get_notdiamond_url()
future = executor.submit(requests.post, url, headers=headers, json=[payload], stream=True)
response = future.result()
response.raise_for_status()
if stream:
return Response(stream_with_context(generate_stream_response(response, model)), content_type='text/event-stream')
else:
return handle_non_stream_response(response, model)
except json.JSONDecodeError as e:
return jsonify({
'error': 'Invalid JSON format',
'details': str(e)
}), 400 # 返回400错误,因为请求数据格式不正确
except requests.exceptions.RequestException as e:
return jsonify({
'error': 'Request failed',
'details': str(e)
}), 502 # 返回502错误,表示请求失败
except Exception as e:
return jsonify({
'error': 'Internal Server Error',
'details': str(e)
}), 500
if __name__ == '__main__':
port = int(os.environ.get("PORT", 8000))
app.run(debug=True, host='0.0.0.0', port=port, threaded=True) |