Use asyncio.gather for __query_ml_predict tasks

main.py

@@ -492,9 +492,11 @@ async def generate_questions(request):
     # print(__ps)
     
     for z, _p in enumerate(__ps):
+        # __query_ml_predict is an awaitable
         ptasks.append(__query_ml_predict(qtype=(QType.STMT if z%2==1 else QType.WH), content=_p['content'], header=_p['header'], l=request.path_params.get('lang', lang.VI_VN), num_qs=request.path_params.get('num_qs', 5 * _p.get('count', 1)), token_limit = int(1024 * _p.get('count', 1))))
 
-    __raw_outputs = [await p for p in ptasks]
+    # __raw_outputs = [await p for p in ptasks]
+    __raw_outputs = asyncio.gather(*ptasks)
     
     for pgph_i, o in enumerate(__raw_outputs):
         # print(o)