singa-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From GitBox <...@apache.org>
Subject [GitHub] [singa] chrishkchris commented on issue #651: Add new example APIs
Date Tue, 07 Apr 2020 04:35:33 GMT
chrishkchris commented on issue #651: Add new example APIs
URL: https://github.com/apache/singa/pull/651#issuecomment-610168094
 
 
   now it supports cifar 100 as well, we can train (mlp|cnn|resnet) + (cifar10|cifar100|mnist)
   
   ```
   root@71ac539cda77:~/dcsysh/singa/examples/autograd# mpiexec -np 2 python3 train_mpi.py
resnet cifar100 --lr 0.1
   Starting Epoch 0:
   Training loss = 3123.622803, training accuracy = 0.094952
   Evaluation accuracy = 0.151743, Elapsed Time = 71.773229s
   Starting Epoch 1:
   Training loss = 2569.560547, training accuracy = 0.197817
   Evaluation accuracy = 0.273538, Elapsed Time = 72.339331s
   Starting Epoch 2:
   Training loss = 2142.290039, training accuracy = 0.299519
   Evaluation accuracy = 0.367688, Elapsed Time = 72.516621s
   Starting Epoch 3:
   Training loss = 1766.168091, training accuracy = 0.395593
   Evaluation accuracy = 0.455929, Elapsed Time = 72.540777s
   Starting Epoch 4:
   Training loss = 1488.601074, training accuracy = 0.478446
   Evaluation accuracy = 0.503806, Elapsed Time = 72.693376s
   Starting Epoch 5:
   Training loss = 1302.042725, training accuracy = 0.535337
   Evaluation accuracy = 0.536859, Elapsed Time = 72.012540s
   Starting Epoch 6:
   Training loss = 1148.359131, training accuracy = 0.583213
   Evaluation accuracy = 0.583133, Elapsed Time = 72.194807s
   Starting Epoch 7:
   Training loss = 1035.075928, training accuracy = 0.620353
   Evaluation accuracy = 0.606370, Elapsed Time = 72.698941s
   Starting Epoch 8:
   Training loss = 924.950745, training accuracy = 0.654067
   Evaluation accuracy = 0.621194, Elapsed Time = 72.009050s
   Starting Epoch 9:
   Training loss = 834.806763, training accuracy = 0.685857
   Evaluation accuracy = 0.643129, Elapsed Time = 72.860034s
   ```
   

----------------------------------------------------------------
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:
users@infra.apache.org


With regards,
Apache Git Services

Mime
View raw message