MCPcopy Create free account
hub / github.com/PaddlePaddle/FastDeploy / RequestFuncInput

Class RequestFuncInput

benchmarks/backend_request_func.py:37–57  ·  view source on GitHub ↗

Input for requesting LLMs via API

Source from the content-addressed store, hash-verified

35
36@dataclass
37class RequestFuncInput:
38 """Input for requesting LLMs via API"""
39
40 no: int
41 prompt: str
42 history_QA: Optional[dict]
43 hyper_parameters: dict
44 api_url: str
45 prompt_len: int
46 output_len: int
47 model: str
48 model_name: Optional[str] = None
49 logprobs: Optional[int] = None
50 extra_body: Optional[dict] = None
51 multi_modal_content: Optional[dict] = None
52 ignore_eos: bool = False
53 language: Optional[str] = None
54 debug: bool = False
55 pd_metrics: bool = False
56 response_format: Optional[dict] = None
57 random_flag: bool = False
58
59
60@dataclass

Callers 2

benchmarkFunction · 0.90
benchmarkFunction · 0.90

Calls

no outgoing calls

Tested by

no test coverage detected