CynthiaProtector opened a new issue #14917: Changing the learning rate affect 
the accuracy
URL: https://github.com/apache/incubator-mxnet/issues/14917
 
 
   ## Description
   During the training process, I tried to change the learning rate to 0.0 
after 200 steps, and I found that the accuracy experiences a dropoff. Why is 
that? 
   
   The following figure shows the training process of Lenet with the MNIST 
dataset. I use one server and one worker (two different nodes), and the worker 
contains 2 GPUs. I start the training task with the following command:
   
   `python ../../tools/launch.py -s 1 -n 1 --launcher ssh -H hosts python 
train_mnist.py --network lenet --kv-store dist_sync --gpus 0,1`
   
   
![lr-200change](https://user-images.githubusercontent.com/23467320/57387914-cb154c00-71e9-11e9-86cc-b4ca2458359f.png)

----------------------------------------------------------------
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