admin管理员组

文章数量:1122846

I am working with a DART-MX8M-PLUS and I want to run a TF-lite model on the VX Delegate. The model runs fine, but I encountered an issue that is quite peculiar.

I am interfacing via python with a proprietary program that has some time constraints in the run cycle and initialization functions (i.e., they can't be too long). The issue is that the model that I'am running on the VX Delegate require 5s to load the first time, and therefore the initialization of the system goes into timeout.

My question is: is there a way to preload the model on the delegate with a script, and run the inference with another?

I didn't find any related documentation, but I am curious if there is a workaround to the problem.

本文标签：

版权声明：本文标题：tensorflow - TFLite: preloading external delegate with a script and run inference on another script - Stack Overflow 内容由网友自发贡献，该文观点仅代表作者本人， 转载请联系作者并注明出处：http://www.betaflare.com/web/1736300883a1930980.html， 本站仅提供信息存储空间服务，不拥有所有权，不承担相关法律责任。如发现本站有涉嫌抄袭侵权/违法违规的内容，一经查实，本站将立刻删除。