Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
|
@@ -357,6 +357,7 @@ def create_gradio_interface():
|
|
| 357 |
|
| 358 |
def setup_api_endpoints(demo):
|
| 359 |
from fastapi import FastAPI, HTTPException
|
|
|
|
| 360 |
from pydantic import BaseModel
|
| 361 |
from typing import Optional
|
| 362 |
|
|
@@ -372,8 +373,9 @@ def setup_api_endpoints(demo):
|
|
| 372 |
timestamp: str
|
| 373 |
model_used: str
|
| 374 |
|
|
|
|
| 375 |
@demo.app.post("/api/chat")
|
| 376 |
-
|
| 377 |
try:
|
| 378 |
if not request.message.strip():
|
| 379 |
raise HTTPException(status_code=400, detail="Message cannot be empty")
|
|
@@ -386,13 +388,13 @@ def setup_api_endpoints(demo):
|
|
| 386 |
response = ai_system.process_query(request.message, request.user_id, request.conversation_id)
|
| 387 |
processing_time = time.time() - start_time
|
| 388 |
|
| 389 |
-
return
|
| 390 |
-
response
|
| 391 |
-
processing_time
|
| 392 |
-
conversation_id
|
| 393 |
-
timestamp
|
| 394 |
-
model_used
|
| 395 |
-
)
|
| 396 |
|
| 397 |
except HTTPException:
|
| 398 |
raise
|
|
@@ -400,32 +402,46 @@ def setup_api_endpoints(demo):
|
|
| 400 |
logger.error(f"API chat error: {e}")
|
| 401 |
raise HTTPException(status_code=500, detail="Internal server error")
|
| 402 |
|
|
|
|
| 403 |
@demo.app.get("/api/health")
|
| 404 |
-
|
| 405 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 406 |
|
| 407 |
@demo.app.get("/api/models")
|
| 408 |
-
|
| 409 |
-
|
| 410 |
"available_models": ["microsoft/Phi-3.5-mini-instruct"],
|
| 411 |
"current_model": Config.MODEL_NAME,
|
| 412 |
"quantization": "Q4_K_M",
|
| 413 |
"context_length": 4096,
|
| 414 |
"parameters": "3.8B"
|
| 415 |
}
|
|
|
|
| 416 |
|
| 417 |
@demo.app.get("/api/stats")
|
| 418 |
-
|
| 419 |
if not monitor:
|
| 420 |
-
return
|
|
|
|
|
|
|
|
|
|
| 421 |
|
| 422 |
-
|
| 423 |
"total_requests": len(monitor.inference_metrics),
|
| 424 |
"average_response_time": monitor.get_average_response_time(),
|
| 425 |
"error_rate": monitor.get_error_rate(),
|
| 426 |
"uptime": monitor.get_uptime(),
|
| 427 |
"system_health": get_system_status()
|
| 428 |
}
|
|
|
|
| 429 |
|
| 430 |
if __name__ == "__main__":
|
| 431 |
logger.info("🎵 Starting Saem's Tunes AI on Hugging Face Spaces...")
|
|
|
|
| 357 |
|
| 358 |
def setup_api_endpoints(demo):
|
| 359 |
from fastapi import FastAPI, HTTPException
|
| 360 |
+
from fastapi.responses import JSONResponse
|
| 361 |
from pydantic import BaseModel
|
| 362 |
from typing import Optional
|
| 363 |
|
|
|
|
| 373 |
timestamp: str
|
| 374 |
model_used: str
|
| 375 |
|
| 376 |
+
# FIX: Remove async from endpoints that don't need it
|
| 377 |
@demo.app.post("/api/chat")
|
| 378 |
+
def api_chat(request: ChatRequest): # Remove async
|
| 379 |
try:
|
| 380 |
if not request.message.strip():
|
| 381 |
raise HTTPException(status_code=400, detail="Message cannot be empty")
|
|
|
|
| 388 |
response = ai_system.process_query(request.message, request.user_id, request.conversation_id)
|
| 389 |
processing_time = time.time() - start_time
|
| 390 |
|
| 391 |
+
return JSONResponse(content={
|
| 392 |
+
"response": response,
|
| 393 |
+
"processing_time": processing_time,
|
| 394 |
+
"conversation_id": request.conversation_id or f"conv_{int(time.time())}",
|
| 395 |
+
"timestamp": datetime.now().isoformat(),
|
| 396 |
+
"model_used": Config.MODEL_NAME
|
| 397 |
+
})
|
| 398 |
|
| 399 |
except HTTPException:
|
| 400 |
raise
|
|
|
|
| 402 |
logger.error(f"API chat error: {e}")
|
| 403 |
raise HTTPException(status_code=500, detail="Internal server error")
|
| 404 |
|
| 405 |
+
# FIX: Remove async and use direct dict return
|
| 406 |
@demo.app.get("/api/health")
|
| 407 |
+
def api_health(): # Remove async
|
| 408 |
+
try:
|
| 409 |
+
status_data = get_system_status()
|
| 410 |
+
return status_data # FastAPI automatically converts dict to JSON
|
| 411 |
+
except Exception as e:
|
| 412 |
+
logger.error(f"Health endpoint error: {e}")
|
| 413 |
+
return JSONResponse(
|
| 414 |
+
content={"status": "error", "error": str(e)},
|
| 415 |
+
status_code=500
|
| 416 |
+
)
|
| 417 |
|
| 418 |
@demo.app.get("/api/models")
|
| 419 |
+
def api_models(): # Remove async
|
| 420 |
+
models_info = {
|
| 421 |
"available_models": ["microsoft/Phi-3.5-mini-instruct"],
|
| 422 |
"current_model": Config.MODEL_NAME,
|
| 423 |
"quantization": "Q4_K_M",
|
| 424 |
"context_length": 4096,
|
| 425 |
"parameters": "3.8B"
|
| 426 |
}
|
| 427 |
+
return models_info # FastAPI automatically converts dict to JSON
|
| 428 |
|
| 429 |
@demo.app.get("/api/stats")
|
| 430 |
+
def api_stats(): # Remove async
|
| 431 |
if not monitor:
|
| 432 |
+
return JSONResponse(
|
| 433 |
+
content={"error": "Monitoring system not available"},
|
| 434 |
+
status_code=503
|
| 435 |
+
)
|
| 436 |
|
| 437 |
+
stats_data = {
|
| 438 |
"total_requests": len(monitor.inference_metrics),
|
| 439 |
"average_response_time": monitor.get_average_response_time(),
|
| 440 |
"error_rate": monitor.get_error_rate(),
|
| 441 |
"uptime": monitor.get_uptime(),
|
| 442 |
"system_health": get_system_status()
|
| 443 |
}
|
| 444 |
+
return stats_data # FastAPI automatically converts dict to JSON
|
| 445 |
|
| 446 |
if __name__ == "__main__":
|
| 447 |
logger.info("🎵 Starting Saem's Tunes AI on Hugging Face Spaces...")
|