singa-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From GitBox <...@apache.org>
Subject [GitHub] [singa] chrishkchris opened a new pull request #579: hotfix: bugs in autograd.py and also update test case
Date Wed, 22 Jan 2020 06:58:09 GMT
chrishkchris opened a new pull request #579: hotfix: bugs in autograd.py and also update test
case
URL: https://github.com/apache/singa/pull/579
 
 
   I have fixed the bugs in autograd.py and update test case mentioned in issue #576 
   
   The results are ok now:
   
   ```
   ubuntu@ip-172-31-24-48:~/singa/test/python$ python3 test_operation.py
   .............................................................................................................................
   ----------------------------------------------------------------------
   Ran 125 tests in 0.624s
   
   OK
   ubuntu@ip-172-31-24-48:~/singa/test/python$ cd ..
   ubuntu@ip-172-31-24-48:~/singa/test$ cd ..
   ubuntu@ip-172-31-24-48:~/singa$ cd examples
   ubuntu@ip-172-31-24-48:~/singa/examples$ cd autograd
   ubuntu@ip-172-31-24-48:~/singa/examples/autograd$ python3 mlp.py
   train_data_shape: (400, 2)
   train_label_shape: (400, 2)
   training loss =  0.6905591
   training loss =  0.5654273
   training loss =  0.54077435
   training loss =  0.5085985
   training loss =  0.42543384
   training loss =  0.31906518
   training loss =  0.25143874
   training loss =  0.20494391
   training loss =  0.17236656
   training loss =  0.14908642
   training loss =  0.13166472
   ubuntu@ip-172-31-24-48:~/singa/examples/autograd$ python3 download_mnist.py
   Downloading http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz
   Downloading http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz
   Downloading http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz
   Downloading http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz
   ubuntu@ip-172-31-24-48:~/singa/examples/autograd$ python3 mnist_cnn.py
   Starting Epoch 0:
   Training loss = 582.361877, training accuracy = 0.795024
   Evaluation accuracy = 0.945012, Elapsed Time = 2.526412s
   Starting Epoch 1:
   Training loss = 232.189026, training accuracy = 0.922142
   Evaluation accuracy = 0.957131, Elapsed Time = 2.488854s
   Starting Epoch 2:
   Training loss = 163.937531, training accuracy = 0.945804
   Evaluation accuracy = 0.973558, Elapsed Time = 2.490621s
   Starting Epoch 3:
   Training loss = 137.145462, training accuracy = 0.954592
   Evaluation accuracy = 0.972456, Elapsed Time = 2.501071s
   Starting Epoch 4:
   Training loss = 116.372910, training accuracy = 0.961229
   Evaluation accuracy = 0.972756, Elapsed Time = 2.501080s
   Starting Epoch 5:
   Training loss = 103.669510, training accuracy = 0.965331
   Evaluation accuracy = 0.974960, Elapsed Time = 2.512660s
   Starting Epoch 6:
   Training loss = 95.173775, training accuracy = 0.967499
   Evaluation accuracy = 0.973057, Elapsed Time = 2.500494s
   Starting Epoch 7:
   Training loss = 84.533409, training accuracy = 0.971551
   Evaluation accuracy = 0.983073, Elapsed Time = 2.499969s
   Starting Epoch 8:
   Training loss = 80.991859, training accuracy = 0.972936
   Evaluation accuracy = 0.979768, Elapsed Time = 2.500226s
   Starting Epoch 9:
   Training loss = 74.402122, training accuracy = 0.974536
   Evaluation accuracy = 0.984675, Elapsed Time = 2.502320s
   ubuntu@ip-172-31-24-48:~/singa/examples/autograd$ python3 resnet.py
   Start intialization............
   100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████|
100/100 [00:29<00:00,  3.45it/s]
   Throughput = 110.192340941104 per second
   Total=0.29040130853652957, forward=0.09273585081100463, softmax=0.0013292169570922852,
backward=0.19633624076843265, sgd=0.009238913059234619
   ```

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