Skip to content
This repository has been archived by the owner on Nov 17, 2023. It is now read-only.

[test] testcases for multi-devices #271

Merged
merged 5 commits into from
Oct 12, 2015
Merged

[test] testcases for multi-devices #271

merged 5 commits into from
Oct 12, 2015

Conversation

mli
Copy link
Member

@mli mli commented Oct 12, 2015

Not passed exactly...

~/mxnet/tests/python/multi-node $ ./local_lenet.py
[04:10:01] src/engine/engine.cc:32: MXNet start using engine: NaiveEngine
[04:10:02] src/io/iter_mnist.cc:94: MNISTIter: load 60000 images, shuffle=0, shape=(100, 1, 28, 28)
[04:10:02] src/io/iter_mnist.cc:94: MNISTIter: load 10000 images, shuffle=0, shape=(100, 1, 28, 28)
INFO:root:Start training with [gpu(0)]
INFO:root:Iteration[0] Train-accuracy=0.753656
INFO:root:Iteration[0] Time cost=3.405
INFO:root:Iteration[1] Train-accuracy=0.974367
INFO:root:Iteration[1] Time cost=3.407
INFO:root:Iteration[2] Train-accuracy=0.983900
INFO:root:Iteration[2] Time cost=3.415
INFO:root:Accuracy = 0.980200
[04:10:15] src/io/iter_mnist.cc:94: MNISTIter: load 60000 images, shuffle=0, shape=(100, 1, 28, 28)
[04:10:15] src/io/iter_mnist.cc:94: MNISTIter: load 10000 images, shuffle=0, shape=(100, 1, 28, 28)
INFO:root:Start training with [gpu(0)]
INFO:root:Iteration[0] Train-accuracy=0.753656
INFO:root:Iteration[0] Time cost=3.373
INFO:root:Iteration[1] Train-accuracy=0.974367
INFO:root:Iteration[1] Time cost=3.415
INFO:root:Iteration[2] Train-accuracy=0.983900
INFO:root:Iteration[2] Time cost=3.410
INFO:root:Accuracy = 0.980200
[04:10:27] src/io/iter_mnist.cc:94: MNISTIter: load 60000 images, shuffle=0, shape=(100, 1, 28, 28)
[04:10:27] src/io/iter_mnist.cc:94: MNISTIter: load 10000 images, shuffle=0, shape=(100, 1, 28, 28)
INFO:root:Start training with [gpu(0), gpu(1)]
INFO:root:Iteration[0] Train-accuracy=0.754073
INFO:root:Iteration[0] Time cost=5.362
INFO:root:Iteration[1] Train-accuracy=0.973933
INFO:root:Iteration[1] Time cost=5.448
INFO:root:Iteration[2] Train-accuracy=0.984817
INFO:root:Iteration[2] Time cost=5.473
INFO:root:Accuracy = 0.980700
[04:10:45] src/io/iter_mnist.cc:94: MNISTIter: load 60000 images, shuffle=0, shape=(100, 1, 28, 28)
[04:10:45] src/io/iter_mnist.cc:94: MNISTIter: load 10000 images, shuffle=0, shape=(100, 1, 28, 28)
INFO:root:Start training with [gpu(0), gpu(1)]
INFO:root:Iteration[0] Train-accuracy=0.754073
INFO:root:Iteration[0] Time cost=6.386
INFO:root:Iteration[1] Train-accuracy=0.973933
INFO:root:Iteration[1] Time cost=6.300
INFO:root:Iteration[2] Train-accuracy=0.984817
INFO:root:Iteration[2] Time cost=6.427
INFO:root:Accuracy = 0.980700
[04:11:06] src/io/iter_mnist.cc:94: MNISTIter: load 60000 images, shuffle=0, shape=(100, 1, 28, 28)
[04:11:06] src/io/iter_mnist.cc:94: MNISTIter: load 10000 images, shuffle=0, shape=(100, 1, 28, 28)
INFO:root:Start training with [gpu(0), gpu(1)]
INFO:root:Iteration[0] Train-accuracy=0.754073
INFO:root:Iteration[0] Time cost=5.842
INFO:root:Iteration[1] Train-accuracy=0.973933
INFO:root:Iteration[1] Time cost=5.898
INFO:root:Iteration[2] Train-accuracy=0.984817
INFO:root:Iteration[2] Time cost=5.973
INFO:root:Accuracy = 0.980700
[04:11:24] src/engine/naive_engine.cc:28: Engine shutdown

mli added a commit that referenced this pull request Oct 12, 2015
@mli
Merge pull request #271 from mli/master
0125990 
[test] testcases for multi-devices
@mli mli merged commit 0125990 into apache:master Oct 12, 2015
eric-haibin-lin pushed a commit to eric-haibin-lin/mxnet that referenced this pull request Dec 2, 2017
@tqchen
fix lint (apache#271)
e7701ab
iblislin pushed a commit to iblislin/incubator-mxnet that referenced this pull request Mar 18, 2018
@vchuravy
Merge pull request apache#271 from iblis17/fix-travis
f34643b 
travis : set make -j4 and travis_wait
joseph-wakeling-sociomantic added a commit to joseph-wakeling-sociomantic/mxnet that referenced this pull request Mar 23, 2018
@joseph-wakeling-sociomantic
Merge stable 1.1.0 release of MXNet
184b9de 
https://github.com/apache/incubator-mxnet/releases/tag/1.1.0

A quite extended amount of manual work was needed here to resolve all
the conflicts, almost none due to the tsunami fork!

Some diffs seemed to survive the merge, and were fixed via manually
checking out the files in question from the upstream 1.1.0 tag.

* 3rdparty/cub ()...05eb57fa(05eb57fa) (1 commits)
  > Merge pull request sociomantic-tsunami#1 from ptrendx/update

* 3rdparty/googletest ()...release-1.8.0(ec44c6c) (1 commits)
  > Merge pull request apache#821 from mazong1123/master

* 3rdparty/openmp ()...37c7212(37c7212) (1 commits)
  > Merging r317115:

* 3rdparty/cub ()...05eb57fa(05eb57fa) (1 commits)
  > Merge pull request sociomantic-tsunami#1 from ptrendx/update

* 3rdparty/googletest ()...release-1.8.0(ec44c6c) (1 commits)
  > Merge pull request apache#821 from mazong1123/master

* 3rdparty/openmp ()...37c7212(37c7212) (1 commits)
  > Merging r317115:

* dmlc-core 87b7ffa(87b7ffa)...6389c10(6389c10) (20 commits)
  > Fix symbol demangling on stacktraces under Linux. (apache#356)
  > Add more environment variables to support 3rd-party S3 implementations (apache#354)
  > add qid as ranklib format (apache#317)
  > Make choice of archiver configurable (apache#352)
  > Fix warnings in unittest_lockfree.cc, fix travis CI (apache#347)
  (...)

* mshadow 2d7780c(2d7780c)...16ac8cd(16ac8cd) (5 commits)
  > Add fully qualified type on define default_real_t (apache#317)
  > fix AddTakeGradLargeBatch for CPU. close apache#235 (apache#316)
  > fix float16 min and max values (apache#315)
  > [Windows] make Packet::size constexpr (apache#313)
  > add GetRndEngine (apache#308)

* nnvm e4a138a(e4a138a)...7a052d6(7a052d6) (3 commits)
  > [SYMBOL] Add __iter__ and GetChildren for symbol (apache#268)
  > API call to get symbol output count (apache#270)
  > [FRONTEND] Fix mxnet multi outputs (apache#271)
@joseph-wakeling-sociomantic joseph-wakeling-sociomantic mentioned this pull request Mar 23, 2018
Merged
eric-haibin-lin pushed a commit to eric-haibin-lin/mxnet that referenced this pull request Apr 4, 2018
@zhreshold @tqchen
[FRONTEND] Fix mxnet multi outputs (apache#271)
ecb02ae 
* fix mxnet multi outputs

* add test case for multi outputs

* fix

* fix

* fix

* fix index

* use json hack

* fix test cases

* fix test cases

* fix test cases

* fix test cases

* fix test cases

* fix test cases
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Reviewers
No reviews
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
1 participant
@mli