Other Parts Discussed in Thread: PROCESSOR-SDK-J721S2 请注意,本文内容源自机器翻译,可能存在语法或其它翻译错误,仅供参考。如需获取准确内容,请参阅链接中的英语原文或自行翻译。 https://e2e.ti.com/support/processors-group/processors/f/processors-forum/1232355/processor-sdk-j721s2-how-to-import-pre-quantized-network…
Thank you for the question.
Generally speaking, you will expect some performance hit when you increase the setting from 8bit to 16bit.
My understanding is that you are using mixed-precision model, and the inference performance (speed) is not good enough…