MCPcopy Create free account
hub / github.com/NVIDIA/TensorRT-LLM / submit

Method submit

tensorrt_llm/executor/rpc_proxy_mixin.py:82–99  ·  view source on GitHub ↗
(self, request: GenerationRequest)

Source from the content-addressed store, hash-verified

80 atexit.register(self.shutdown)
81
82 def submit(self, request: GenerationRequest) -> GenerationResult:
83 request.set_id(self._get_next_client_id())
84 logprob_params = self._get_logprob_params(request)
85
86 # submit is a fire-and-forget operation, don't need to wait for response
87 with nvtx_range_debug("RPCExecutor.submit", color="green", category="Proxy"):
88 self.rpc_client.submit(request).remote(need_response=False)
89
90 result = GenerationResult(
91 request,
92 background_error_handler=self._handle_background_error,
93 executor=self,
94 disaggregated_params=request.disaggregated_params,
95 logprob_params=logprob_params,
96 )
97 self._results[request.id] = result
98
99 return result
100
101 def handle_responses(self, responses: list[GenerationResult]) -> bool:
102 async_queues = []

Callers

nothing calls this directly

Calls 6

nvtx_range_debugFunction · 0.85
GenerationResultClass · 0.85
set_idMethod · 0.80
_get_next_client_idMethod · 0.80
_get_logprob_paramsMethod · 0.80
remoteMethod · 0.80

Tested by

no test coverage detected