I get an error when I set " 'executionEnvironment','parallel' " using LSTM layer

Question

ammar Alali 2018-3-29

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/391434-i-get-an-error-when-i-set-executionenvironment-parallel-using-lstm-layer

回答： Madhav Thakker 2020-12-29

I'm trying to train my network using the 31 workers I have on my local station (CPUs). Whenever I set the training option, "ExecutionEnvironment','Auto'"; I get this error: parallel does not support LSTM layer. I use LSTM layer in my network. Is there anyway to run the training job faster if I'm using LSTM layer ?

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

Madhav Thakker 2020-12-29

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/391434-i-get-an-error-when-i-set-executionenvironment-parallel-using-lstm-layer#answer_586382

Hi Ammar,

Using multiple workers are not yet supported for LSTMs or recurrent networks. Unfortunately, there is no workaround for training recurrent neural networks like LSTM in parallel. The developers are aware of this behaviour and will try to provide a workaround in future releases.

Hope this helps.