api: Shut up!
Browse files
app.py
CHANGED
@@ -8,6 +8,7 @@ import time
|
|
8 |
import json
|
9 |
import asyncio
|
10 |
import logging
|
|
|
11 |
from typing import Optional, List, Union, Dict, Any, Literal
|
12 |
from fastapi import FastAPI, HTTPException, Request, status
|
13 |
from fastapi.middleware.cors import CORSMiddleware
|
@@ -29,6 +30,7 @@ class SessionData:
|
|
29 |
self.history: List[Dict[str, Any]] = []
|
30 |
self.last_access: float = time.time()
|
31 |
self.active_tasks: Dict[str, asyncio.Task] = {}
|
|
|
32 |
|
33 |
class SessionManager:
|
34 |
def __init__(self):
|
@@ -48,55 +50,49 @@ class SessionManager:
|
|
48 |
for task_id, task in data.active_tasks.items():
|
49 |
if not task.done():
|
50 |
task.cancel()
|
51 |
-
logger.info(f"Cancelled active task {task_id} for expired session {sid}")
|
52 |
for user, sid in expired:
|
53 |
if user in self.sessions and sid in self.sessions[user]:
|
54 |
del self.sessions[user][sid]
|
55 |
if not self.sessions[user]:
|
56 |
del self.sessions[user]
|
57 |
-
logger.info(f"Session expired: user={user} session={sid}")
|
58 |
|
59 |
async def get_session(self, user: Optional[str], session_id: Optional[str]) -> (str, str, SessionData):
|
60 |
async with self.lock:
|
61 |
if not user:
|
62 |
user = str(uuid.uuid4())
|
63 |
-
logger.debug(f"Generated new user ID: {user}")
|
64 |
if user not in self.sessions:
|
65 |
self.sessions[user] = {}
|
66 |
if not session_id or session_id not in self.sessions[user]:
|
67 |
session_id = str(uuid.uuid4())
|
68 |
self.sessions[user][session_id] = SessionData()
|
69 |
-
logger.info(f"Created new session: user={user} session={session_id}")
|
70 |
session = self.sessions[user][session_id]
|
71 |
session.last_access = time.time()
|
72 |
-
logger.debug(f"Session accessed: user={user} session={session_id} history_length={len(session.history)}")
|
73 |
return user, session_id, session
|
74 |
|
75 |
session_manager = SessionManager()
|
76 |
|
77 |
async def refresh_client(app: FastAPI):
|
78 |
while True:
|
79 |
-
await asyncio.sleep(
|
80 |
async with app.state.client_lock:
|
81 |
-
if app.state.client is None:
|
82 |
-
|
83 |
-
|
84 |
-
|
85 |
-
|
86 |
-
|
87 |
-
|
88 |
-
|
89 |
-
|
90 |
-
|
91 |
-
|
92 |
-
|
93 |
-
|
94 |
-
|
95 |
-
|
96 |
-
|
97 |
-
|
98 |
-
|
99 |
-
break
|
100 |
|
101 |
@asynccontextmanager
|
102 |
async def lifespan(app: FastAPI):
|
@@ -104,13 +100,15 @@ async def lifespan(app: FastAPI):
|
|
104 |
app.state.client = None
|
105 |
app.state.client_lock = asyncio.Lock()
|
106 |
app.state.refresh_task = asyncio.create_task(refresh_client(app))
|
107 |
-
|
|
|
108 |
try:
|
109 |
yield
|
110 |
finally:
|
111 |
app.state.refresh_task.cancel()
|
|
|
|
|
112 |
await asyncio.sleep(0.1)
|
113 |
-
logger.info("App lifespan ended, refresh client task cancelled")
|
114 |
|
115 |
app = FastAPI(
|
116 |
title="J.A.R.V.I.S. OpenAI-Compatible API",
|
@@ -220,20 +218,16 @@ async def get_client(app: FastAPI) -> Client:
|
|
220 |
async def call_gradio(client: Client, params: dict):
|
221 |
for attempt in range(3):
|
222 |
try:
|
223 |
-
logger.debug(f"Calling Gradio attempt {attempt+1}")
|
224 |
return await asyncio.to_thread(lambda: client.submit(**params))
|
225 |
except Exception as e:
|
226 |
-
logger.warning(f"Gradio call failed attempt {attempt+1}: {e}", exc_info=True)
|
227 |
await asyncio.sleep(0.2 * (attempt + 1))
|
228 |
-
logger.error("Gradio upstream error after 3 attempts")
|
229 |
raise HTTPException(status_code=502, detail="Upstream Gradio app error")
|
230 |
|
231 |
async def stream_response(job, session_id: str, session_history: List[Dict[str, Any]], new_messages: List[Message], response_type: str):
|
232 |
partial = ""
|
233 |
try:
|
234 |
chunks = await asyncio.to_thread(lambda: list(job))
|
235 |
-
except Exception
|
236 |
-
logger.error(f"Streaming error: {e}", exc_info=True)
|
237 |
chunks = []
|
238 |
for chunk in chunks:
|
239 |
try:
|
@@ -258,8 +252,7 @@ async def stream_response(job, session_id: str, session_history: List[Dict[str,
|
|
258 |
"session_id": session_id
|
259 |
}
|
260 |
yield f"data: {json.dumps(data)}\n\n"
|
261 |
-
except Exception
|
262 |
-
logger.error(f"Chunk yield error: {e}", exc_info=True)
|
263 |
continue
|
264 |
session_history.extend([m.model_dump() for m in new_messages if m.role != "system"])
|
265 |
session_history.append({"role": "assistant", "content": partial})
|
@@ -271,9 +264,15 @@ async def stream_response(job, session_id: str, session_history: List[Dict[str,
|
|
271 |
}
|
272 |
yield f"data: {json.dumps(done_data)}\n\n"
|
273 |
|
|
|
|
|
274 |
@app.post("/v1/chat/completions")
|
275 |
async def chat_completions(req: ChatCompletionRequest):
|
276 |
user, session_id, session = await session_manager.get_session(req.user, req.session_id)
|
|
|
|
|
|
|
|
|
277 |
req.messages = sanitize_messages(req.messages)
|
278 |
for m in req.messages:
|
279 |
if m.role == "system":
|
@@ -299,7 +298,6 @@ async def chat_completions(req: ChatCompletionRequest):
|
|
299 |
"function_call": req.function_call or req.tool_choice,
|
300 |
}
|
301 |
params = {k: v for k, v in params.items() if v is not None}
|
302 |
-
logger.info(f"Chat completion request user={user} session={session_id} model={req.model} stream={req.stream}")
|
303 |
client = await get_client(app)
|
304 |
if req.stream:
|
305 |
job = await call_gradio(client, params)
|
@@ -309,12 +307,10 @@ async def chat_completions(req: ChatCompletionRequest):
|
|
309 |
loop = asyncio.get_running_loop()
|
310 |
try:
|
311 |
result = await loop.run_in_executor(None, lambda: client.predict(**params))
|
312 |
-
except Exception
|
313 |
-
logger.error(f"Gradio predict error: {e}", exc_info=True)
|
314 |
raise HTTPException(status_code=502, detail="Upstream Gradio app error")
|
315 |
session.history.extend([m.model_dump() for m in req.messages if m.role != "system"])
|
316 |
session.history.append({"role": "assistant", "content": result})
|
317 |
-
logger.info(f"Chat completion response sent user={user} session={session_id}")
|
318 |
return {
|
319 |
"id": str(uuid.uuid4()),
|
320 |
"object": "chat.completion",
|
@@ -324,7 +320,11 @@ async def chat_completions(req: ChatCompletionRequest):
|
|
324 |
|
325 |
@app.post("/v1/completions")
|
326 |
async def completions(req: CompletionRequest):
|
327 |
-
user, session_id,
|
|
|
|
|
|
|
|
|
328 |
prompt = req.prompt if isinstance(req.prompt, str) else "\n".join(req.prompt)
|
329 |
params = {
|
330 |
"message": prompt,
|
@@ -343,7 +343,6 @@ async def completions(req: CompletionRequest):
|
|
343 |
"seed": req.seed,
|
344 |
}
|
345 |
params = {k: v for k, v in params.items() if v is not None}
|
346 |
-
logger.info(f"Completion request user={user} session={session_id} model={req.model} stream={req.stream}")
|
347 |
client = await get_client(app)
|
348 |
if req.stream:
|
349 |
job = await call_gradio(client, params)
|
@@ -353,29 +352,24 @@ async def completions(req: CompletionRequest):
|
|
353 |
loop = asyncio.get_running_loop()
|
354 |
try:
|
355 |
result = await loop.run_in_executor(None, lambda: client.predict(**params))
|
356 |
-
except Exception
|
357 |
-
logger.error(f"Gradio predict error: {e}", exc_info=True)
|
358 |
raise HTTPException(status_code=502, detail="Upstream Gradio app error")
|
359 |
-
logger.info(f"Completion response sent user={user} session={session_id}")
|
360 |
return {"id": str(uuid.uuid4()), "object": "text_completion", "choices": [{"text": result}]}
|
361 |
|
362 |
@app.post("/v1/embeddings")
|
363 |
async def embeddings(req: EmbeddingRequest):
|
364 |
inputs = req.input if isinstance(req.input, list) else [req.input]
|
365 |
embeddings = [[0.0] * 768 for _ in inputs]
|
366 |
-
logger.info(f"Embedding request model={req.model} inputs_count={len(inputs)}")
|
367 |
return {"object": "list", "data": [{"embedding": emb, "index": i} for i, emb in enumerate(embeddings)]}
|
368 |
|
369 |
@app.get("/v1/models")
|
370 |
async def get_models():
|
371 |
-
logger.info("Models list requested")
|
372 |
return {"object": "list", "data": [{"id": "Q8_K_XL", "object": "model", "owned_by": "J.A.R.V.I.S."}]}
|
373 |
|
374 |
@app.get("/v1/history")
|
375 |
async def get_history(user: Optional[str] = None, session_id: Optional[str] = None):
|
376 |
user = user or "anonymous"
|
377 |
sessions = session_manager.sessions
|
378 |
-
logger.info(f"History requested user={user} session={session_id}")
|
379 |
if user in sessions and session_id and session_id in sessions[user]:
|
380 |
return {"user": user, "session_id": session_id, "history": sessions[user][session_id].history}
|
381 |
return {"user": user, "session_id": session_id, "history": []}
|
@@ -384,7 +378,6 @@ async def get_history(user: Optional[str] = None, session_id: Optional[str] = No
|
|
384 |
async def cancel_response(user: Optional[str], session_id: Optional[str], task_id: Optional[str]):
|
385 |
user = user or "anonymous"
|
386 |
if not task_id:
|
387 |
-
logger.warning(f"Cancel response missing task_id user={user} session={session_id}")
|
388 |
raise HTTPException(status_code=400, detail="Missing task_id for cancellation")
|
389 |
async with session_manager.lock:
|
390 |
if user in session_manager.sessions and session_id in session_manager.sessions[user]:
|
@@ -392,9 +385,7 @@ async def cancel_response(user: Optional[str], session_id: Optional[str], task_i
|
|
392 |
task = session.active_tasks.get(task_id)
|
393 |
if task and not task.done():
|
394 |
task.cancel()
|
395 |
-
logger.info(f"Cancelled task {task_id} for user={user} session={session_id}")
|
396 |
return {"message": f"Cancelled task {task_id}"}
|
397 |
-
logger.warning(f"Task not found or already completed task_id={task_id} user={user} session={session_id}")
|
398 |
raise HTTPException(status_code=404, detail="Task not found or already completed")
|
399 |
|
400 |
@app.api_route("/v1", methods=["POST", "GET", "OPTIONS", "HEAD"])
|
@@ -403,15 +394,12 @@ async def router(request: Request):
|
|
403 |
try:
|
404 |
body_json = await request.json()
|
405 |
except Exception:
|
406 |
-
logger.error("Invalid JSON body in router POST")
|
407 |
raise HTTPException(status_code=400, detail="Invalid JSON body")
|
408 |
try:
|
409 |
body = RouterRequest(**body_json)
|
410 |
except ValidationError as e:
|
411 |
-
logger.error(f"Validation error in router POST: {e.errors()}")
|
412 |
raise HTTPException(status_code=422, detail=e.errors())
|
413 |
endpoint = body.endpoint or "chat/completions"
|
414 |
-
logger.info(f"Router POST to endpoint={endpoint}")
|
415 |
if endpoint == "chat/completions":
|
416 |
if not body.model or not body.messages:
|
417 |
raise HTTPException(status_code=422, detail="Missing 'model' or 'messages'")
|
@@ -432,12 +420,10 @@ async def router(request: Request):
|
|
432 |
elif endpoint == "history":
|
433 |
return await get_history(body.user, body.session_id)
|
434 |
elif endpoint == "responses/cancel":
|
435 |
-
return await cancel_response(body.user, body.session_id, body.
|
436 |
else:
|
437 |
-
logger.warning(f"Router POST unknown endpoint: {endpoint}")
|
438 |
raise HTTPException(status_code=404, detail="Endpoint not found")
|
439 |
else:
|
440 |
-
logger.info(f"Router {request.method} called - only POST supported with JSON body")
|
441 |
return JSONResponse({"message": "Send POST request with JSON body"}, status_code=status.HTTP_405_METHOD_NOT_ALLOWED)
|
442 |
|
443 |
@app.get("/")
|
|
|
8 |
import json
|
9 |
import asyncio
|
10 |
import logging
|
11 |
+
import os
|
12 |
from typing import Optional, List, Union, Dict, Any, Literal
|
13 |
from fastapi import FastAPI, HTTPException, Request, status
|
14 |
from fastapi.middleware.cors import CORSMiddleware
|
|
|
30 |
self.history: List[Dict[str, Any]] = []
|
31 |
self.last_access: float = time.time()
|
32 |
self.active_tasks: Dict[str, asyncio.Task] = {}
|
33 |
+
self.last_request_time: float = 0.0
|
34 |
|
35 |
class SessionManager:
|
36 |
def __init__(self):
|
|
|
50 |
for task_id, task in data.active_tasks.items():
|
51 |
if not task.done():
|
52 |
task.cancel()
|
|
|
53 |
for user, sid in expired:
|
54 |
if user in self.sessions and sid in self.sessions[user]:
|
55 |
del self.sessions[user][sid]
|
56 |
if not self.sessions[user]:
|
57 |
del self.sessions[user]
|
|
|
58 |
|
59 |
async def get_session(self, user: Optional[str], session_id: Optional[str]) -> (str, str, SessionData):
|
60 |
async with self.lock:
|
61 |
if not user:
|
62 |
user = str(uuid.uuid4())
|
|
|
63 |
if user not in self.sessions:
|
64 |
self.sessions[user] = {}
|
65 |
if not session_id or session_id not in self.sessions[user]:
|
66 |
session_id = str(uuid.uuid4())
|
67 |
self.sessions[user][session_id] = SessionData()
|
|
|
68 |
session = self.sessions[user][session_id]
|
69 |
session.last_access = time.time()
|
|
|
70 |
return user, session_id, session
|
71 |
|
72 |
session_manager = SessionManager()
|
73 |
|
74 |
async def refresh_client(app: FastAPI):
|
75 |
while True:
|
76 |
+
await asyncio.sleep(15 * 60)
|
77 |
async with app.state.client_lock:
|
78 |
+
if app.state.client is not None:
|
79 |
+
try:
|
80 |
+
old_client = app.state.client
|
81 |
+
app.state.client = None
|
82 |
+
del old_client
|
83 |
+
app.state.client = Client("https://hadadrjt-ai.hf.space/")
|
84 |
+
logger.info("Refreshed Gradio client connection")
|
85 |
+
except Exception as e:
|
86 |
+
logger.error(f"Error refreshing Gradio client: {e}", exc_info=True)
|
87 |
+
app.state.client = None
|
88 |
+
|
89 |
+
async def clear_terminal_periodically():
|
90 |
+
while True:
|
91 |
+
await asyncio.sleep(300)
|
92 |
+
if os.name == "nt":
|
93 |
+
os.system("cls")
|
94 |
+
else:
|
95 |
+
print("\033c", end="", flush=True)
|
|
|
96 |
|
97 |
@asynccontextmanager
|
98 |
async def lifespan(app: FastAPI):
|
|
|
100 |
app.state.client = None
|
101 |
app.state.client_lock = asyncio.Lock()
|
102 |
app.state.refresh_task = asyncio.create_task(refresh_client(app))
|
103 |
+
app.state.cleanup_task = asyncio.create_task(session_manager.cleanup())
|
104 |
+
app.state.clear_log_task = asyncio.create_task(clear_terminal_periodically())
|
105 |
try:
|
106 |
yield
|
107 |
finally:
|
108 |
app.state.refresh_task.cancel()
|
109 |
+
app.state.cleanup_task.cancel()
|
110 |
+
app.state.clear_log_task.cancel()
|
111 |
await asyncio.sleep(0.1)
|
|
|
112 |
|
113 |
app = FastAPI(
|
114 |
title="J.A.R.V.I.S. OpenAI-Compatible API",
|
|
|
218 |
async def call_gradio(client: Client, params: dict):
|
219 |
for attempt in range(3):
|
220 |
try:
|
|
|
221 |
return await asyncio.to_thread(lambda: client.submit(**params))
|
222 |
except Exception as e:
|
|
|
223 |
await asyncio.sleep(0.2 * (attempt + 1))
|
|
|
224 |
raise HTTPException(status_code=502, detail="Upstream Gradio app error")
|
225 |
|
226 |
async def stream_response(job, session_id: str, session_history: List[Dict[str, Any]], new_messages: List[Message], response_type: str):
|
227 |
partial = ""
|
228 |
try:
|
229 |
chunks = await asyncio.to_thread(lambda: list(job))
|
230 |
+
except Exception:
|
|
|
231 |
chunks = []
|
232 |
for chunk in chunks:
|
233 |
try:
|
|
|
252 |
"session_id": session_id
|
253 |
}
|
254 |
yield f"data: {json.dumps(data)}\n\n"
|
255 |
+
except Exception:
|
|
|
256 |
continue
|
257 |
session_history.extend([m.model_dump() for m in new_messages if m.role != "system"])
|
258 |
session_history.append({"role": "assistant", "content": partial})
|
|
|
264 |
}
|
265 |
yield f"data: {json.dumps(done_data)}\n\n"
|
266 |
|
267 |
+
RATE_LIMIT_SECONDS = 1.0
|
268 |
+
|
269 |
@app.post("/v1/chat/completions")
|
270 |
async def chat_completions(req: ChatCompletionRequest):
|
271 |
user, session_id, session = await session_manager.get_session(req.user, req.session_id)
|
272 |
+
now = time.time()
|
273 |
+
if now - session.last_request_time < RATE_LIMIT_SECONDS:
|
274 |
+
raise HTTPException(status_code=429, detail="Too many requests, please slow down")
|
275 |
+
session.last_request_time = now
|
276 |
req.messages = sanitize_messages(req.messages)
|
277 |
for m in req.messages:
|
278 |
if m.role == "system":
|
|
|
298 |
"function_call": req.function_call or req.tool_choice,
|
299 |
}
|
300 |
params = {k: v for k, v in params.items() if v is not None}
|
|
|
301 |
client = await get_client(app)
|
302 |
if req.stream:
|
303 |
job = await call_gradio(client, params)
|
|
|
307 |
loop = asyncio.get_running_loop()
|
308 |
try:
|
309 |
result = await loop.run_in_executor(None, lambda: client.predict(**params))
|
310 |
+
except Exception:
|
|
|
311 |
raise HTTPException(status_code=502, detail="Upstream Gradio app error")
|
312 |
session.history.extend([m.model_dump() for m in req.messages if m.role != "system"])
|
313 |
session.history.append({"role": "assistant", "content": result})
|
|
|
314 |
return {
|
315 |
"id": str(uuid.uuid4()),
|
316 |
"object": "chat.completion",
|
|
|
320 |
|
321 |
@app.post("/v1/completions")
|
322 |
async def completions(req: CompletionRequest):
|
323 |
+
user, session_id, session = await session_manager.get_session(req.user, req.session_id)
|
324 |
+
now = time.time()
|
325 |
+
if now - session.last_request_time < RATE_LIMIT_SECONDS:
|
326 |
+
raise HTTPException(status_code=429, detail="Too many requests, please slow down")
|
327 |
+
session.last_request_time = now
|
328 |
prompt = req.prompt if isinstance(req.prompt, str) else "\n".join(req.prompt)
|
329 |
params = {
|
330 |
"message": prompt,
|
|
|
343 |
"seed": req.seed,
|
344 |
}
|
345 |
params = {k: v for k, v in params.items() if v is not None}
|
|
|
346 |
client = await get_client(app)
|
347 |
if req.stream:
|
348 |
job = await call_gradio(client, params)
|
|
|
352 |
loop = asyncio.get_running_loop()
|
353 |
try:
|
354 |
result = await loop.run_in_executor(None, lambda: client.predict(**params))
|
355 |
+
except Exception:
|
|
|
356 |
raise HTTPException(status_code=502, detail="Upstream Gradio app error")
|
|
|
357 |
return {"id": str(uuid.uuid4()), "object": "text_completion", "choices": [{"text": result}]}
|
358 |
|
359 |
@app.post("/v1/embeddings")
|
360 |
async def embeddings(req: EmbeddingRequest):
|
361 |
inputs = req.input if isinstance(req.input, list) else [req.input]
|
362 |
embeddings = [[0.0] * 768 for _ in inputs]
|
|
|
363 |
return {"object": "list", "data": [{"embedding": emb, "index": i} for i, emb in enumerate(embeddings)]}
|
364 |
|
365 |
@app.get("/v1/models")
|
366 |
async def get_models():
|
|
|
367 |
return {"object": "list", "data": [{"id": "Q8_K_XL", "object": "model", "owned_by": "J.A.R.V.I.S."}]}
|
368 |
|
369 |
@app.get("/v1/history")
|
370 |
async def get_history(user: Optional[str] = None, session_id: Optional[str] = None):
|
371 |
user = user or "anonymous"
|
372 |
sessions = session_manager.sessions
|
|
|
373 |
if user in sessions and session_id and session_id in sessions[user]:
|
374 |
return {"user": user, "session_id": session_id, "history": sessions[user][session_id].history}
|
375 |
return {"user": user, "session_id": session_id, "history": []}
|
|
|
378 |
async def cancel_response(user: Optional[str], session_id: Optional[str], task_id: Optional[str]):
|
379 |
user = user or "anonymous"
|
380 |
if not task_id:
|
|
|
381 |
raise HTTPException(status_code=400, detail="Missing task_id for cancellation")
|
382 |
async with session_manager.lock:
|
383 |
if user in session_manager.sessions and session_id in session_manager.sessions[user]:
|
|
|
385 |
task = session.active_tasks.get(task_id)
|
386 |
if task and not task.done():
|
387 |
task.cancel()
|
|
|
388 |
return {"message": f"Cancelled task {task_id}"}
|
|
|
389 |
raise HTTPException(status_code=404, detail="Task not found or already completed")
|
390 |
|
391 |
@app.api_route("/v1", methods=["POST", "GET", "OPTIONS", "HEAD"])
|
|
|
394 |
try:
|
395 |
body_json = await request.json()
|
396 |
except Exception:
|
|
|
397 |
raise HTTPException(status_code=400, detail="Invalid JSON body")
|
398 |
try:
|
399 |
body = RouterRequest(**body_json)
|
400 |
except ValidationError as e:
|
|
|
401 |
raise HTTPException(status_code=422, detail=e.errors())
|
402 |
endpoint = body.endpoint or "chat/completions"
|
|
|
403 |
if endpoint == "chat/completions":
|
404 |
if not body.model or not body.messages:
|
405 |
raise HTTPException(status_code=422, detail="Missing 'model' or 'messages'")
|
|
|
420 |
elif endpoint == "history":
|
421 |
return await get_history(body.user, body.session_id)
|
422 |
elif endpoint == "responses/cancel":
|
423 |
+
return await cancel_response(body.user, body.session_id, body.tool_choice if isinstance(body.tool_choice, str) else None)
|
424 |
else:
|
|
|
425 |
raise HTTPException(status_code=404, detail="Endpoint not found")
|
426 |
else:
|
|
|
427 |
return JSONResponse({"message": "Send POST request with JSON body"}, status_code=status.HTTP_405_METHOD_NOT_ALLOWED)
|
428 |
|
429 |
@app.get("/")
|