juliusshufan commented on issue #12591: USE_MKLDNN=1 is default in make build 
(mkldnn must be explicitly turned off)
URL: https://github.com/apache/incubator-mxnet/pull/12591#issuecomment-424968127
 
 
   **RNN related data, including both accuracy, and performance benchmarking.**
   **Accuracy**
   1. GNMT model implemented by gluon-nlp (scripts\nmt\train_gnmt.py), IWMT2015 
dataset, en-vi translation. The decoder-encoder is a 2-layer LSTM, per the 
model implemenation, as gluon.rnncell used, the MKLDNN FC can be covered as it 
is gluon.rnncell is an unfused kernel, below figure is the ppl trends collected 
on both GPU and CPU, with same hyper-parameters, the two curves aligned very 
well.
   
![image](https://user-images.githubusercontent.com/33112206/46125781-3f077300-c25d-11e8-890f-45187e523b0f.png)
   

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
[email protected]


With regards,
Apache Git Services

Reply via email to