[BUG] [GenAI-Perf] openai-fronted server with --endpoint-type completions

[jihyeonRyu]

Description
I’m currently running perf-analyzer with the --endpoint-type completions option. But then I encountered the following error:

Thread [0] had error: OpenAI response returns HTTP code 400: {"detail":"only single string input is supported"}

Could you please advise on how to resolve this issue?

*The curl to completions has no problem.
*Only genai-perf has problem.

Triton Information
What version of Triton are you using?

Triton Server Version 2.54.0
TensoRT-LLM v0.17.0
genai-perf 0.0.10 (PIP)
server/python/openai: main branch (18a9c51)

Are you using the Triton container or did you build it yourself?

To Reproduce
Steps to reproduce the behavior.

Describe the models (framework, inputs, outputs), ideally include the model configuration file (if using an ensemble include the model configuration file for that as well).

Expected behavior
A clear and concise description of what you expected to happen.