File size: 17,690 Bytes
935f280 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 |
from fastapi import FastAPI, HTTPException, Header
from fastapi.middleware.cors import CORSMiddleware
from fastapi.responses import StreamingResponse
from pydantic import BaseModel
import openai
from typing import List, Optional, Union
import logging
import httpx
import uuid
import time
import json
from datetime import datetime, timezone
import requests
import uvicorn
import random
logging.basicConfig(
level=logging.INFO, format="%(asctime)s - %(levelname)s - %(message)s"
)
logger = logging.getLogger(__name__)
app = FastAPI()
app.add_middleware(
CORSMiddleware,
allow_origins=["*"],
allow_credentials=True,
allow_methods=["*"],
allow_headers=["*"],
)
MAX_RETRIES = 3
class ChatRequest(BaseModel):
messages: List[dict]
model: str
temperature: Optional[float] = 0.7
stream: Optional[bool] = False
tools: Optional[List[dict]] = []
tool_choice: Optional[str] = "auto"
class EmbeddingRequest(BaseModel):
input: Union[str, List[str]]
model: str
encoding_format: Optional[str] = "float"
async def verify_authorization(authorization: str = Header(None)):
print("Authorization header:", authorization)
if not authorization:
logger.error("Missing Authorization header")
raise HTTPException(status_code=401, detail="Missing Authorization header")
if not authorization.startswith("Bearer "):
logger.error("Invalid Authorization header format")
raise HTTPException(
status_code=401, detail="Invalid Authorization header format"
)
token = authorization.replace("Bearer ", "")
return token
def get_openai_models(api_keys):
api_key = random.choice(api_keys)
try:
client = openai.OpenAI(api_key=api_key)
models = client.models.list()
return models.model_dump()
except Exception as e:
logger.error(f"Error getting models from OpenAI with key {api_key}: {e}")
return {"error": str(e)}
def get_gemini_models(api_keys):
api_key = random.choice(api_keys)
base_url = "https://generativelanguage.googleapis.com/v1beta"
url = f"{base_url}/models?key={api_key}"
try:
response = requests.get(url)
if response.status_code == 200:
gemini_models = response.json()
return convert_to_openai_models_format(gemini_models)
else:
logger.error(f"Error getting models from Gemini with key {api_key}: {response.status_code} - {response.text}")
return {"error": f"Gemini API error: {response.status_code} - {response.text}"}
except requests.RequestException as e:
logger.error(f"Request failed: {e}")
return {"error": f"Request failed: {e}"}
def convert_to_openai_models_format(gemini_models):
openai_format = {"object": "list", "data": []}
for model in gemini_models.get("models", []):
openai_model = {
"id": model["name"].split("/")[-1],
"object": "model",
"created": int(datetime.now(timezone.utc).timestamp()),
"owned_by": "google",
"permission": [],
"root": model["name"],
"parent": None,
}
openai_format["data"].append(openai_model)
return openai_format
def convert_messages_to_gemini_format(messages):
gemini_messages = []
for msg in messages:
role = "user" if msg["role"] == "user" else "model"
parts = []
if isinstance(msg["content"], str):
parts.append({"text": msg["content"]})
elif isinstance(msg["content"], list):
for content in msg["content"]:
if isinstance(content, str):
parts.append({"text": content})
elif isinstance(content, dict) and content["type"] == "text":
parts.append({"text": content["text"]})
elif isinstance(content, dict) and content["type"] == "image_url":
image_url = content["image_url"]["url"]
if image_url.startswith("data:image"):
parts.append(
{
"inline_data": {
"mime_type": "image/jpeg",
"data": image_url.split(",")[1],
}
}
)
else:
parts.append(
{
"image_url": {
"url": image_url,
}
}
)
gemini_messages.append({"role": role, "parts": parts})
return gemini_messages
async def convert_gemini_response_to_openai(response, model, stream=False):
if stream:
chunk = response
if not chunk["candidates"]:
return None
return {
"id": "chatcmpl-" + str(uuid.uuid4()),
"object": "chat.completion.chunk",
"created": int(time.time()),
"model": model,
"choices": [
{
"index": 0,
"delta": {
"content": chunk["candidates"][0]["content"]["parts"][0]["text"]
},
"finish_reason": None,
}
],
}
else:
content = response["candidates"][0]["content"]["parts"][0]["text"]
return {
"id": "chatcmpl-" + str(uuid.uuid4()),
"object": "chat.completion",
"created": int(time.time()),
"model": model,
"choices": [
{
"index": 0,
"message": {
"role": "assistant",
"content": content,
},
"finish_reason": "stop",
}
],
"usage": {"prompt_tokens": 0, "completion_tokens": 0, "total_tokens": 0},
}
@app.get("/v1/models")
@app.get("/hf/v1/models")
async def list_models(authorization: str = Header(None)):
token = await verify_authorization(authorization)
api_keys = [key.strip() for key in token.split(',')]
all_models = []
error_messages = []
for api_key in api_keys:
if api_key.startswith("sk-"):
response = get_openai_models([api_key])
else:
response = get_gemini_models([api_key])
if "error" in response:
error_messages.append(response["error"])
else:
if isinstance(response, dict) and 'data' in response:
all_models.extend(response['data'])
else:
logger.warning(f"Unexpected response format from model list API for key {api_key}: {response}")
if error_messages and not all_models:
raise HTTPException(status_code=500, detail=f"Errors encountered: {', '.join(error_messages)}")
return {"data": all_models, "object": "list"}
@app.post("/v1/chat/completions")
@app.post("/hf/v1/chat/completions")
async def chat_completion(request: ChatRequest, authorization: str = Header(None)):
token = await verify_authorization(authorization)
api_keys = [key.strip() for key in token.split(',')]
logger.info(f"Chat completion request - Model: {request.model}")
retries = 0
while retries < MAX_RETRIES:
api_key = random.choice(api_keys)
try:
logger.info(f"Attempt {retries + 1} with API key: {api_key}")
if api_key.startswith("sk-"):
client = openai.OpenAI(api_key=api_key)
if request.stream:
logger.info("Streaming response enabled")
async def generate():
try:
stream_response = client.chat.completions.create(
model=request.model,
messages=request.messages,
temperature=request.temperature,
stream=True,
)
for chunk in stream_response:
chunk_json = chunk.model_dump_json()
yield f"data: {chunk_json}\n\n"
yield "data: [DONE]\n\n"
except Exception as e:
logger.error(f"Stream error: {str(e)}")
raise
return StreamingResponse(content=generate(), media_type="text/event-stream")
else:
response = client.chat.completions.create(
model=request.model,
messages=request.messages,
temperature=request.temperature,
)
logger.info("Chat completion successful")
return response.model_dump()
else:
gemini_messages = convert_messages_to_gemini_format(request.messages)
payload = {
"contents": gemini_messages,
"generationConfig": {
"temperature": request.temperature,
}
}
if request.stream:
logger.info("Streaming response enabled")
async def generate():
nonlocal api_key, retries, api_keys
while retries < MAX_RETRIES:
try:
async with httpx.AsyncClient() as client:
stream_url = f"https://generativelanguage.googleapis.com/v1beta/models/{request.model}:streamGenerateContent?alt=sse&key={api_key}"
async with client.stream("POST", stream_url, json=payload, timeout=60.0) as response:
if response.status_code == 429:
logger.warning(f"Rate limit reached for key: {api_key}")
retries += 1
if retries >= MAX_RETRIES:
yield f"data: {json.dumps({'error': 'Max retries reached'})}\n\n"
break
api_keys.remove(api_key)
if not api_keys:
yield f"data: {json.dumps({'error': 'All API keys exhausted'})}\n\n"
break
api_key = random.choice(api_keys)
logger.info(f"Retrying with a new API key: {api_key}")
continue
if response.status_code != 200:
logger.error(f"Error in streaming response with key {api_key}: {response.status_code} - {response.text}")
retries += 1
if retries >= MAX_RETRIES:
yield f"data: {json.dumps({'error': 'Max retries reached'})}\n\n"
break
api_keys.remove(api_key)
if not api_keys:
yield f"data: {json.dumps({'error': 'All API keys exhausted'})}\n\n"
break
api_key = random.choice(api_keys)
logger.info(f"Retrying with a new API key: {api_key}")
continue
async for line in response.aiter_lines():
if line.startswith("data: "):
try:
chunk = json.loads(line[6:])
if not chunk.get("candidates"):
continue
content = chunk["candidates"][0]["content"]["parts"][0]["text"]
new_chunk = {
"id": "chatcmpl-" + str(uuid.uuid4()),
"object": "chat.completion.chunk",
"created": int(time.time()),
"model": request.model,
"choices": [
{
"index": 0,
"delta": {
"content": content
},
"finish_reason": None,
}
],
}
yield f"data: {json.dumps(new_chunk)}\n\n"
except json.JSONDecodeError:
continue
yield "data: [DONE]\n\n"
return
except Exception as e:
logger.error(f"Stream error: {str(e)}")
retries += 1
if retries >= MAX_RETRIES:
yield f"data: {json.dumps({'error': 'Max retries reached'})}\n\n"
break
api_keys.remove(api_key)
if not api_keys:
yield f"data: {json.dumps({'error': 'All API keys exhausted'})}\n\n"
break
api_key = random.choice(api_keys)
logger.info(f"Retrying with a new API key: {api_key}")
continue
return StreamingResponse(content=generate(), media_type="text/event-stream")
else:
async with httpx.AsyncClient() as client:
non_stream_url = f"https://generativelanguage.googleapis.com/v1beta/models/{request.model}:generateContent?key={api_key}"
response = await client.post(non_stream_url, json=payload)
if response.status_code != 200:
logger.error(f"Error in non-streaming response with key {api_key}: {response.status_code} - {response.text}")
retries += 1
if retries >= MAX_RETRIES:
raise HTTPException(status_code=500, detail="Max retries reached")
api_keys.remove(api_key)
if not api_keys:
raise HTTPException(status_code=500, detail="All API keys exhausted")
api_key = random.choice(api_keys)
logger.info(f"Retrying with a new API key: {api_key}")
continue
gemini_response = response.json()
logger.info("Chat completion successful")
return await convert_gemini_response_to_openai(gemini_response, request.model)
except Exception as e:
logger.error(f"Error in chat completion: {str(e)}")
if isinstance(e, HTTPException):
raise e
retries += 1
if retries >= MAX_RETRIES:
logger.error("Max retries reached, giving up")
raise HTTPException(status_code=500, detail="Max retries reached")
api_keys.remove(api_key)
if not api_keys:
raise HTTPException(status_code=500, detail="All API keys exhausted")
api_key = random.choice(api_keys)
logger.info(f"Retrying with a new API key: {api_key}")
continue
raise HTTPException(status_code=500, detail="Unexpected error in chat completion")
@app.get("/health")
@app.get("/")
async def health_check():
logger.info("Health check endpoint called")
return {"status": "healthy"}
if __name__ == "__main__":
uvicorn.run(app, host="0.0.0.0", port=8080) |