How to use GPU on model that was imported from the "Deep Learning Toolbox Importer for TensorFlow-Keras Models" ?

Question

Tuong Nguyen Minh 2020-8-15

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/579984-how-to-use-gpu-on-model-that-was-imported-from-the-deep-learning-toolbox-importer-for-tensorflow-ke

回答： Srivardhan Gadila 2020-8-18

I have trained a network model in Keras and has successfully been able to import such model into Mathlab 2020a. How ever I am not sure how to use GPU for that model to accelerate the speed ?

Please help me thank you !

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Srivardhan Gadila 2020-8-18

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/579984-how-to-use-gpu-on-model-that-was-imported-from-the-deep-learning-toolbox-importer-for-tensorflow-ke#answer_481536

You can use the following Name-Value pair arguments: ExecutionEnvironment & Acceleration of the predict function in order to use the GPU for prediction.

Make sure that you are using a SeriesNetwork or a DAGNetwork object with the predict function. If you have used importKerasNetwork then no issues. If you have used importKerasLayers then convert the layers/layerGraph to a SeriesNetwork or a DAGNetwork object using the assembleNetwork function.