mikeobr commented on issue #16173: Saving and loading cudNN autotune and graph 
optimization
URL: 
https://github.com/apache/incubator-mxnet/issues/16173#issuecomment-558341889
 
 
   This feature would be very useful for us. 
   
   We deploy our models via MXNet Model Server: 
https://github.com/awslabs/multi-model-server as a custom service. Each worker 
has an instance of the model, so we experience memory instability if concurrent 
autotuning is happening at once. This forces us to either under utilize the 
available GPU  or risk errors at startup. Being able to cache autotune would 
help with both cold starts and production-izing models.

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to 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