[GitHub] marcoabreu commented on issue #9596: Remove the gradle wrapper.

marcoabreu commented on issue #9596: Remove the gradle wrapper. URL: https://github.com/apache/incubator-mxnet/pull/9596#issuecomment-360967918 I'm afraid that this case is not covered by the Integration test. Could you please explain how you verified that the build is still working?

[GitHub] reminisce commented on issue #9552: [REQUEST FOR REVIEW | DO NOT MERGE] Model Quantization with Calibration

reminisce commented on issue #9552: [REQUEST FOR REVIEW | DO NOT MERGE] Model Quantization with Calibration URL: https://github.com/apache/incubator-mxnet/pull/9552#issuecomment-360965769 @marcoabreu Oh, I see. Since calibration is conducted offline, it's not constrained by the hardware

[GitHub] thinksanky opened a new pull request #9597: update copyright info

thinksanky opened a new pull request #9597: update copyright info URL: https://github.com/apache/incubator-mxnet/pull/9597 ## Description ## Updated copyright information in the footer This is an automated message from the

[GitHub] marcoabreu commented on issue #9584: bump up version to 1.1

marcoabreu commented on issue #9584: bump up version to 1.1 URL: https://github.com/apache/incubator-mxnet/pull/9584#issuecomment-360965685 You might want to change the version on the website as well: https://github.com/apache/incubator-mxnet/blob/master/docs/_static/mxnet-theme/index.html

[incubator-mxnet] branch master updated: Remove unnecessary RuntimeError (#9578)

This is an automated email from the ASF dual-hosted git repository. zhasheng pushed a commit to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git The following commit(s) were added to refs/heads/master by this push: new 45366a4 Remove unnecessary

[GitHub] szha closed pull request #9578: Remove unnecessary RuntimeError

szha closed pull request #9578: Remove unnecessary RuntimeError URL: https://github.com/apache/incubator-mxnet/pull/9578 This is a PR merged from a forked repository. As GitHub hides the original diff on merge, it is displayed below for the sake of provenance: As this is a foreign pull

[GitHub] marcoabreu commented on issue #9552: [REQUEST FOR REVIEW | DO NOT MERGE] Model Quantization with Calibration

marcoabreu commented on issue #9552: [REQUEST FOR REVIEW | DO NOT MERGE] Model Quantization with Calibration URL: https://github.com/apache/incubator-mxnet/pull/9552#issuecomment-360965066 Oh sorry, this question was targeted towards the ``num_bins`` question - the email messed it up.

[GitHub] reminisce commented on issue #9552: [REQUEST FOR REVIEW | DO NOT MERGE] Model Quantization with Calibration

reminisce commented on issue #9552: [REQUEST FOR REVIEW | DO NOT MERGE] Model Quantization with Calibration URL: https://github.com/apache/incubator-mxnet/pull/9552#issuecomment-360964981 @marcoabreu The optimal values are determined by the calibration datasets. So they are independent of

[GitHub] reminisce commented on issue #9552: [REQUEST FOR REVIEW | DO NOT MERGE] Model Quantization with Calibration

reminisce commented on issue #9552: [REQUEST FOR REVIEW | DO NOT MERGE] Model Quantization with Calibration URL: https://github.com/apache/incubator-mxnet/pull/9552#issuecomment-360964778 @jinhuang415 1. The parameters are quantized offline, which means the min/max values were

[GitHub] indhub opened a new pull request #9596: Remove the gradle wrapper.

indhub opened a new pull request #9596: Remove the gradle wrapper. URL: https://github.com/apache/incubator-mxnet/pull/9596 ## Description ## Remove the gradle wrapper. Apache RAT complains about the jar file being present in repo. Users can build using gradle directly. ##

[GitHub] thinksanky opened a new pull request #9595: fixed mail:to and the install typo

thinksanky opened a new pull request #9595: fixed mail:to and the install typo URL: https://github.com/apache/incubator-mxnet/pull/9595 ## Description ## Fixed mail:to and the install link typo This is an automated message

[GitHub] pengzhao-intel commented on issue #9545: Profiling discussion

pengzhao-intel commented on issue #9545: Profiling discussion URL: https://github.com/apache/incubator-mxnet/issues/9545#issuecomment-360964261 Yes, it's very important! From my experience of Theano profiling, the overhead is very small and maybe you can take a look.

[GitHub] reminisce commented on a change in pull request #9552: [REQUEST FOR REVIEW | DO NOT MERGE] Model Quantization with Calibration

reminisce commented on a change in pull request #9552: [REQUEST FOR REVIEW | DO NOT MERGE] Model Quantization with Calibration URL: https://github.com/apache/incubator-mxnet/pull/9552#discussion_r164265094 ## File path: python/mxnet/quantization.py ## @@ -0,0 +1,467 @@ +#

[GitHub] zhreshold commented on a change in pull request #8918: Added in Large-Batch SGD with a warmup, and a LARS startegy. Also add?

zhreshold commented on a change in pull request #8918: Added in Large-Batch SGD with a warmup, and a LARS startegy. Also add? URL: https://github.com/apache/incubator-mxnet/pull/8918#discussion_r164265033 ## File path: python/mxnet/lr_scheduler.py ## @@ -136,3 +136,38 @@

[GitHub] zhreshold commented on a change in pull request #8918: Added in Large-Batch SGD with a warmup, and a LARS startegy. Also add?

zhreshold commented on a change in pull request #8918: Added in Large-Batch SGD with a warmup, and a LARS startegy. Also add? URL: https://github.com/apache/incubator-mxnet/pull/8918#discussion_r164265043 ## File path: python/mxnet/lr_scheduler.py ## @@ -136,3 +136,38 @@

[GitHub] zhreshold commented on a change in pull request #8918: Added in Large-Batch SGD with a warmup, and a LARS startegy. Also add?

zhreshold commented on a change in pull request #8918: Added in Large-Batch SGD with a warmup, and a LARS startegy. Also add? URL: https://github.com/apache/incubator-mxnet/pull/8918#discussion_r164265038 ## File path: python/mxnet/lr_scheduler.py ## @@ -136,3 +136,38 @@

[GitHub] reminisce commented on issue #9552: [REQUEST FOR REVIEW | DO NOT MERGE] Model Quantization with Calibration

reminisce commented on issue #9552: [REQUEST FOR REVIEW | DO NOT MERGE] Model Quantization with Calibration URL: https://github.com/apache/incubator-mxnet/pull/9552#issuecomment-360963839 @wentingj The quantized ops used in the benchmarks are convolution, fully-connected, avg_pooling,

[GitHub] zhreshold commented on issue #9361: infer_shape error for 'resnet-152'

zhreshold commented on issue #9361: infer_shape error for 'resnet-152' URL: https://github.com/apache/incubator-mxnet/issues/9361#issuecomment-360963671 You might want to checkout mxnet model serving @kevinthesun This is an

[GitHub] zhreshold commented on issue #9572: Return a NaN when using operator ( ** ) on Windows version with GPU

zhreshold commented on issue #9572: Return a NaN when using operator ( ** ) on Windows version with GPU URL: https://github.com/apache/incubator-mxnet/issues/9572#issuecomment-360963604 Interestingly, I found it my self on windows using GPU as well:

[GitHub] zhreshold closed issue #9571: ACCURACY IS USING NUMPY, URGENT FIX

zhreshold closed issue #9571: ACCURACY IS USING NUMPY, URGENT FIX URL: https://github.com/apache/incubator-mxnet/issues/9571 This is an automated message from the Apache Git Service. To respond to the message, please log on

[incubator-mxnet] branch master updated: use nd for accuracy calculation (#9583)

This is an automated email from the ASF dual-hosted git repository. zhreshold pushed a commit to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git The following commit(s) were added to refs/heads/master by this push: new f5f1b91 use nd for accuracy

[GitHub] zhreshold closed pull request #9583: use nd for accuracy calculation

zhreshold closed pull request #9583: use nd for accuracy calculation URL: https://github.com/apache/incubator-mxnet/pull/9583 This is a PR merged from a forked repository. As GitHub hides the original diff on merge, it is displayed below for the sake of provenance: As this is a foreign

[GitHub] zhreshold closed issue #9582: Misleading calculation of mxnet.metric.Accuracy

zhreshold closed issue #9582: Misleading calculation of mxnet.metric.Accuracy URL: https://github.com/apache/incubator-mxnet/issues/9582 This is an automated message from the Apache Git Service. To respond to the message,

[GitHub] eric-haibin-lin commented on issue #9581: Force the dtype of data to be float32 for test_correlation

eric-haibin-lin commented on issue #9581: Force the dtype of data to be float32 for test_correlation URL: https://github.com/apache/incubator-mxnet/pull/9581#issuecomment-360963176 @marcoabreu yes this could be a common problem to all other unit test using np random inputs

[GitHub] marcoabreu commented on issue #9581: Force the dtype of data to be float32 for test_correlation

marcoabreu commented on issue #9581: Force the dtype of data to be float32 for test_correlation URL: https://github.com/apache/incubator-mxnet/pull/9581#issuecomment-360963113 So am I right when I assume that this problem could also impact other tests because of a dtype mismatch? Maybe

[GitHub] eric-haibin-lin closed pull request #9535: Tutorials index page update

eric-haibin-lin closed pull request #9535: Tutorials index page update URL: https://github.com/apache/incubator-mxnet/pull/9535 This is a PR merged from a forked repository. As GitHub hides the original diff on merge, it is displayed below for the sake of provenance: As this is a foreign

[GitHub] eric-haibin-lin commented on issue #9412: Flaky Tests Tracking Issue

eric-haibin-lin commented on issue #9412: Flaky Tests Tracking Issue URL: https://github.com/apache/incubator-mxnet/issues/9412#issuecomment-360961378 #9581 fixes #9553 This is an automated message from the Apache Git

[GitHub] eric-haibin-lin commented on issue #9553: test_operator_gpu.test_correlation fails

eric-haibin-lin commented on issue #9553: test_operator_gpu.test_correlation fails URL: https://github.com/apache/incubator-mxnet/issues/9553#issuecomment-360961355 closed via #9581 This is an automated message from the

[GitHub] eric-haibin-lin closed issue #9553: test_operator_gpu.test_correlation fails

eric-haibin-lin closed issue #9553: test_operator_gpu.test_correlation fails URL: https://github.com/apache/incubator-mxnet/issues/9553 This is an automated message from the Apache Git Service. To respond to the message,

[GitHub] eric-haibin-lin closed pull request #9581: Force the dtype of data to be float32 for test_correlation

eric-haibin-lin closed pull request #9581: Force the dtype of data to be float32 for test_correlation URL: https://github.com/apache/incubator-mxnet/pull/9581 This is a PR merged from a forked repository. As GitHub hides the original diff on merge, it is displayed below for the sake of

[GitHub] eric-haibin-lin closed pull request #9577: Disable flaky test: test_correlation

eric-haibin-lin closed pull request #9577: Disable flaky test: test_correlation URL: https://github.com/apache/incubator-mxnet/pull/9577 This is a PR merged from a forked repository. As GitHub hides the original diff on merge, it is displayed below for the sake of provenance: As this is

[incubator-mxnet] branch master updated: Force the dtype of data to be float32 for test_correlation (#9581)

This is an automated email from the ASF dual-hosted git repository. haibin pushed a commit to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git The following commit(s) were added to refs/heads/master by this push: new f1ae225 Force the dtype of data to be

[GitHub] eric-haibin-lin commented on issue #9581: Force the dtype of data to be float32 for test_correlation

eric-haibin-lin commented on issue #9581: Force the dtype of data to be float32 for test_correlation URL: https://github.com/apache/incubator-mxnet/pull/9581#issuecomment-360961270 @szha Yes, we found a np seed/state that consistently fails the test. The random np input is in float64

[GitHub] eric-haibin-lin closed pull request #9575: replaced how_to with faq

eric-haibin-lin closed pull request #9575: replaced how_to with faq URL: https://github.com/apache/incubator-mxnet/pull/9575 This is a PR merged from a forked repository. As GitHub hides the original diff on merge, it is displayed below for the sake of provenance: As this is a foreign

[GitHub] eric-haibin-lin commented on issue #9575: replaced how_to with faq

eric-haibin-lin commented on issue #9575: replaced how_to with faq URL: https://github.com/apache/incubator-mxnet/pull/9575#issuecomment-360957382 Merging it for now so that we can get updated nightly test results. This is

[incubator-mxnet] branch master updated: replaced how_to with faq (#9575)

This is an automated email from the ASF dual-hosted git repository. haibin pushed a commit to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git The following commit(s) were added to refs/heads/master by this push: new 50850af replaced how_to with faq

[GitHub] szha commented on issue #9581: Force the dtype of data to be float32 for test_correlation

szha commented on issue #9581: Force the dtype of data to be float32 for test_correlation URL: https://github.com/apache/incubator-mxnet/pull/9581#issuecomment-360956732 How is the fix verified? Do you have an idea why it started failing?

[GitHub] ZiyueHuang commented on issue #9540: refactor regression ops to nnvm interface

ZiyueHuang commented on issue #9540: refactor regression ops to nnvm interface URL: https://github.com/apache/incubator-mxnet/pull/9540#issuecomment-360955082 Seems that I mistakenly checkout that test file... Added back.

[GitHub] anirudh2290 commented on issue #9380: nd.stack causes abort for differently sized arrays

anirudh2290 commented on issue #9380: nd.stack causes abort for differently sized arrays URL: https://github.com/apache/incubator-mxnet/issues/9380#issuecomment-360953495 Related to: #7335 This is an automated message from

[GitHub] anirudh2290 commented on issue #9567: nd.argmax cause "Kernel Died" error in Jupyter Notebook

anirudh2290 commented on issue #9567: nd.argmax cause "Kernel Died" error in Jupyter Notebook URL: https://github.com/apache/incubator-mxnet/issues/9567#issuecomment-360953466 Related to: #7335 This is an automated message

[GitHub] wkcn commented on issue #9511: set_lr_mult() or set_wd_mult() is invalid if not setting param_idx2name for the optimizer

wkcn commented on issue #9511: set_lr_mult() or set_wd_mult() is invalid if not setting param_idx2name for the optimizer URL: https://github.com/apache/incubator-mxnet/issues/9511#issuecomment-360952988 I think it's difficult to set_lr_mult() in the current implementation. I don't know

[GitHub] eric-haibin-lin commented on issue #9577: Disable flaky test: test_correlation

eric-haibin-lin commented on issue #9577: Disable flaky test: test_correlation URL: https://github.com/apache/incubator-mxnet/pull/9577#issuecomment-360951864 yes On 2018?1?26? ?? at 17:28 Yizhi Liu wrote: > I guess #9581

[GitHub] ashokei commented on a change in pull request #8918: Added in Large-Batch SGD with a warmup, and a LARS startegy. Also add?

ashokei commented on a change in pull request #8918: Added in Large-Batch SGD with a warmup, and a LARS startegy. Also add? URL: https://github.com/apache/incubator-mxnet/pull/8918#discussion_r164259824 ## File path: python/mxnet/lr_scheduler.py ## @@ -136,3 +136,42 @@

[GitHub] eric-haibin-lin commented on a change in pull request #9575: replaced how_to with faq

eric-haibin-lin commented on a change in pull request #9575: replaced how_to with faq URL: https://github.com/apache/incubator-mxnet/pull/9575#discussion_r164259422 ## File path: docs/community/index.md ## @@ -8,9 +8,9 @@ We track bugs and new feature requests in the

[GitHub] eric-haibin-lin commented on a change in pull request #9575: replaced how_to with faq

eric-haibin-lin commented on a change in pull request #9575: replaced how_to with faq URL: https://github.com/apache/incubator-mxnet/pull/9575#discussion_r164259265 ## File path: docs/tutorials/embedded/wine_detector.md ## @@ -37,9 +37,7 @@ To complete this tutorial, you

[GitHub] szha opened a new issue #9594: mx.metric.PearsonCorrelation is using numpy logic

szha opened a new issue #9594: mx.metric.PearsonCorrelation is using numpy logic URL: https://github.com/apache/incubator-mxnet/issues/9594 The metric module has been using numpy logic and is not benefiting from existing mxnet operators.

[GitHub] szha opened a new issue #9593: mx.metric.CrossEntropy/NegativeLogLikelihood is using numpy logic and partially duplicates mx.gluon.loss.SoftmaxCELoss

szha opened a new issue #9593: mx.metric.CrossEntropy/NegativeLogLikelihood is using numpy logic and partially duplicates mx.gluon.loss.SoftmaxCELoss URL: https://github.com/apache/incubator-mxnet/issues/9593 metric.CrossEntropy:

[GitHub] szha opened a new issue #9591: mx.metric.MAE is using numpy logic and duplicates mx.gluon.loss.L1Loss

szha opened a new issue #9591: mx.metric.MAE is using numpy logic and duplicates mx.gluon.loss.L1Loss URL: https://github.com/apache/incubator-mxnet/issues/9591 metric.MAE: https://github.com/apache/incubator-mxnet/blob/master/python/mxnet/metric.py#L725-L733 gluon.loss.L1Loss:

[GitHub] szha commented on a change in pull request #9583: use nd for accuracy calculation

szha commented on a change in pull request #9583: use nd for accuracy calculation URL: https://github.com/apache/incubator-mxnet/pull/9583#discussion_r164257186 ## File path: python/mxnet/metric.py ## @@ -380,23 +380,24 @@ def update(self, labels, preds):

[GitHub] zhreshold commented on a change in pull request #9583: use nd for accuracy calculation

zhreshold commented on a change in pull request #9583: use nd for accuracy calculation URL: https://github.com/apache/incubator-mxnet/pull/9583#discussion_r164257671 ## File path: python/mxnet/metric.py ## @@ -380,23 +380,24 @@ def update(self, labels, preds):

[GitHub] szha opened a new issue #9590: mx.metric.Perplexity should support calculating from cross-entropy

szha opened a new issue #9590: mx.metric.Perplexity should support calculating from cross-entropy URL: https://github.com/apache/incubator-mxnet/issues/9590 Currently part of the logic for calculating cross-entropy loss is in mx.metric.Perplexity. This means we would have repeated

[GitHub] yzhliu commented on issue #9577: Disable flaky test: test_correlation

yzhliu commented on issue #9577: Disable flaky test: test_correlation URL: https://github.com/apache/incubator-mxnet/pull/9577#issuecomment-360948391 I guess https://github.com/apache/incubator-mxnet/pull/9581 fixes the test?

[GitHub] szha commented on issue #9582: Misleading calculation of mxnet.metric.Accuracy

szha commented on issue #9582: Misleading calculation of mxnet.metric.Accuracy URL: https://github.com/apache/incubator-mxnet/issues/9582#issuecomment-360947976 Thanks for the feedback. This is an automated message from the

[GitHub] szha commented on a change in pull request #9583: use nd for accuracy calculation

szha commented on a change in pull request #9583: use nd for accuracy calculation URL: https://github.com/apache/incubator-mxnet/pull/9583#discussion_r164257186 ## File path: python/mxnet/metric.py ## @@ -380,23 +380,24 @@ def update(self, labels, preds):

[GitHub] zhreshold commented on a change in pull request #9583: use nd for accuracy calculation

zhreshold commented on a change in pull request #9583: use nd for accuracy calculation URL: https://github.com/apache/incubator-mxnet/pull/9583#discussion_r164257105 ## File path: python/mxnet/metric.py ## @@ -380,23 +380,24 @@ def update(self, labels, preds):

[GitHub] astonzhang commented on issue #9572: Return a NaN when using operator ( ** ) on Windows version with GPU

astonzhang commented on issue #9572: Return a NaN when using operator ( ** ) on Windows version with GPU URL: https://github.com/apache/incubator-mxnet/issues/9572#issuecomment-360947529 Related post: https://discuss.gluon.ai/t/topic/989 This issue has been raised by multiple

[GitHub] mbaijal commented on a change in pull request #9548: updated release version to 1.0.1 on the mainpage and re-arranged news?

mbaijal commented on a change in pull request #9548: updated release version to 1.0.1 on the mainpage and re-arranged news? URL: https://github.com/apache/incubator-mxnet/pull/9548#discussion_r164256910 ## File path: docs/_static/mxnet-theme/index.html ## @@ -17,27 +17,26

[GitHub] szha commented on issue #9588: metric should have TP FP TN FN Precision Recall F1 for both macro and micro versions

szha commented on issue #9588: metric should have TP FP TN FN Precision Recall F1 for both macro and micro versions URL: https://github.com/apache/incubator-mxnet/issues/9588#issuecomment-360947519 We may refer to the "average" option in

[GitHub] szha opened a new issue #9589: mx.metric should support top-K version of TP FP TN FN Precision Recall F1

szha opened a new issue #9589: mx.metric should support top-K version of TP FP TN FN Precision Recall F1 URL: https://github.com/apache/incubator-mxnet/issues/9589 This request is the same as #9588 with the addition of expanded match logic based on topk prediction.

[GitHub] mbaijal commented on a change in pull request #9548: updated release version to 1.0.1 on the mainpage and re-arranged news?

mbaijal commented on a change in pull request #9548: updated release version to 1.0.1 on the mainpage and re-arranged news? URL: https://github.com/apache/incubator-mxnet/pull/9548#discussion_r164256910 ## File path: docs/_static/mxnet-theme/index.html ## @@ -17,27 +17,26

[GitHub] astonzhang commented on issue #9572: Return a NaN when using operator ( ** ) on Windows version with GPU

astonzhang commented on issue #9572: Return a NaN when using operator ( ** ) on Windows version with GPU URL: https://github.com/apache/incubator-mxnet/issues/9572#issuecomment-360946922 @eric-haibin-lin Who do you think is the best person to look into this? Thanks.

[GitHub] szha opened a new issue #9588: metric should have TP FP TN FN Precision Recall F1 for both macro and micro versions

szha opened a new issue #9588: metric should have TP FP TN FN Precision Recall F1 for both macro and micro versions URL: https://github.com/apache/incubator-mxnet/issues/9588 Related to #9587. The current logic only supports what's called "macro F1" logic where we take the mean of F1

[GitHub] szha commented on issue #9572: Return a NaN when using operator ( ** ) on Windows version with GPU

szha commented on issue #9572: Return a NaN when using operator ( ** ) on Windows version with GPU URL: https://github.com/apache/incubator-mxnet/issues/9572#issuecomment-360947107 @yajiedesign @zhreshold would either of you take a look? Thanks.

[GitHub] astonzhang commented on issue #9572: Return a NaN when using operator ( ** ) on Windows version with GPU

astonzhang commented on issue #9572: Return a NaN when using operator ( ** ) on Windows version with GPU URL: https://github.com/apache/incubator-mxnet/issues/9572#issuecomment-360946922 @eric-haibin-lin Who do you think is the best person to look into this? Thanks.

[GitHub] szha opened a new issue #9587: metric.F1 doc needs clarification

szha opened a new issue #9587: metric.F1 doc needs clarification URL: https://github.com/apache/incubator-mxnet/issues/9587 The current logic only supports what's called "macro F1" logic where we take the mean of F1 scores for mini-batches, which needs to be clarified in doc.

[GitHub] yzhliu commented on issue #9579: Reverting namespace change

yzhliu commented on issue #9579: Reverting namespace change URL: https://github.com/apache/incubator-mxnet/pull/9579#issuecomment-360946658 Thanks @nswamy This is an automated message from the Apache Git Service. To respond

[GitHub] szha opened a new issue #9586: mx.metric F1 is using numpy logic

szha opened a new issue #9586: mx.metric F1 is using numpy logic URL: https://github.com/apache/incubator-mxnet/issues/9586 The metric module has been using numpy logic and is not benefiting from existing mxnet operators.

[GitHub] szha opened a new issue #9585: TopKAccuracy is using numpy logic

szha opened a new issue #9585: TopKAccuracy is using numpy logic URL: https://github.com/apache/incubator-mxnet/issues/9585 The metric module has been using numpy logic and is not benefiting from existing mxnet operators.

[GitHub] eric-haibin-lin commented on a change in pull request #9548: updated release version to 1.0.1 on the mainpage and re-arranged news?

eric-haibin-lin commented on a change in pull request #9548: updated release version to 1.0.1 on the mainpage and re-arranged news? URL: https://github.com/apache/incubator-mxnet/pull/9548#discussion_r164255583 ## File path: docs/_static/mxnet-theme/index.html ## @@

[GitHub] eric-haibin-lin commented on a change in pull request #9548: updated release version to 1.0.1 on the mainpage and re-arranged news?

eric-haibin-lin commented on a change in pull request #9548: updated release version to 1.0.1 on the mainpage and re-arranged news? URL: https://github.com/apache/incubator-mxnet/pull/9548#discussion_r164255462 ## File path: docs/_static/mxnet-theme/index.html ## @@

[GitHub] Ishitori commented on issue #9582: Misleading calculation of mxnet.metric.Accuracy

Ishitori commented on issue #9582: Misleading calculation of mxnet.metric.Accuracy URL: https://github.com/apache/incubator-mxnet/issues/9582#issuecomment-360945239 It shouldn't be a problem under a correct usage, but there is a way to use it incorrectly and system doesn't report that.

[GitHub] eric-haibin-lin opened a new pull request #9584: bump up version to 1.1

eric-haibin-lin opened a new pull request #9584: bump up version to 1.1 URL: https://github.com/apache/incubator-mxnet/pull/9584 This is an automated message from the Apache Git Service. To respond to the message, please log

[GitHub] szha commented on issue #9582: Misleading calculation of mxnet.metric.Accuracy

szha commented on issue #9582: Misleading calculation of mxnet.metric.Accuracy URL: https://github.com/apache/incubator-mxnet/issues/9582#issuecomment-360944566 Is there a case where casting would be a problem under proper usage?

[GitHub] Ishitori commented on issue #9582: Misleading calculation of mxnet.metric.Accuracy

Ishitori commented on issue #9582: Misleading calculation of mxnet.metric.Accuracy URL: https://github.com/apache/incubator-mxnet/issues/9582#issuecomment-360944095 Agree, it is an incorrect usage of a class. Yet, silent casting is happening, which makes it hard to spot a problem.

[GitHub] szha commented on issue #9582: Misleading calculation of mxnet.metric.Accuracy

szha commented on issue #9582: Misleading calculation of mxnet.metric.Accuracy URL: https://github.com/apache/incubator-mxnet/issues/9582#issuecomment-360943676 The accuracy class is not intended to be used for one-hot labels. I'm clarifying that in the api doc in the above PR. In terms

[GitHub] szha opened a new pull request #9583: use nd for accuracy calculation

szha opened a new pull request #9583: use nd for accuracy calculation URL: https://github.com/apache/incubator-mxnet/pull/9583 ## Description ## (Brief description on what this PR is about) ## Checklist ## ### Essentials ### - [x] Passed code style checking (`make lint`)

[GitHub] gpu commented on issue #9572: Return a NaN when using operator ( ** ) on Windows version with GPU

gpu commented on issue #9572: Return a NaN when using operator ( ** ) on Windows version with GPU URL: https://github.com/apache/incubator-mxnet/issues/9572#issuecomment-360942737 Thanks for the notifications. But I'm afraid I cannot help you here :o)

[GitHub] szha commented on issue #8444: /include/mxnet-cpp/operator.hpp:157:31: error: expected ';' at end of declaration std::vector outputs{output};

szha commented on issue #8444: /include/mxnet-cpp/operator.hpp:157:31: error: expected ';' at end of declaration std::vector outputs{output}; URL: https://github.com/apache/incubator-mxnet/issues/8444#issuecomment-360941899 @apache/mxnet-committers: This issue has been inactive

[GitHub] szha commented on issue #8438: HybridBlocks don't work with constant NDArray initialized Parameters

szha commented on issue #8438: HybridBlocks don't work with constant NDArray initialized Parameters URL: https://github.com/apache/incubator-mxnet/issues/8438#issuecomment-360941902 @apache/mxnet-committers: This issue has been inactive for the past 90 days. It has no label and needs

[GitHub] szha commented on issue #8448: How can I get symbol shape of a layer as input of another layer in run time? Not with ?infer_shape?

szha commented on issue #8448: How can I get symbol shape of a layer as input of another layer in run time? Not with ?infer_shape? URL: https://github.com/apache/incubator-mxnet/issues/8448#issuecomment-360941895 @apache/mxnet-committers: This issue has been inactive for the past 90 days.

[GitHub] pracheer commented on issue #8749: [FIX ATTACHED] Not working example on Iterators tutorial - SimpleIter

pracheer commented on issue #8749: [FIX ATTACHED] Not working example on Iterators tutorial - SimpleIter URL: https://github.com/apache/incubator-mxnet/issues/8749#issuecomment-360941877 This issues has been already addressed by https://github.com/apache/incubator-mxnet/pull/9460.

[GitHub] Ishitori opened a new issue #9582: Misleading calculation of mxnet.metric.Accuracy

Ishitori opened a new issue #9582: Misleading calculation of mxnet.metric.Accuracy URL: https://github.com/apache/incubator-mxnet/issues/9582 ## Description When passes "preds" and "labels" arguments of the same shape to mxnet.metric.Accuracy.update() method, they get silently

[GitHub] anirudh2290 opened a new pull request #9581: Force the dtype of data to be float32 for test_correlation

anirudh2290 opened a new pull request #9581: Force the dtype of data to be float32 for test_correlation URL: https://github.com/apache/incubator-mxnet/pull/9581 ## Description ## Force the dtype of img1 and img2 to be float32. This should fix the intermittent breaking of test:

[GitHub] szha commented on a change in pull request #9575: replaced how_to with faq

szha commented on a change in pull request #9575: replaced how_to with faq URL: https://github.com/apache/incubator-mxnet/pull/9575#discussion_r164245730 ## File path: docs/build_version_doc/build_doc.sh ## @@ -70,6 +70,14 @@ then cp -a "$local_build/." "$web_folder"

[GitHub] zhanghang1989 closed issue #9580: BatchNorm Not Backward Correctly in Evaluation Mode

zhanghang1989 closed issue #9580: BatchNorm Not Backward Correctly in Evaluation Mode URL: https://github.com/apache/incubator-mxnet/issues/9580 This is an automated message from the Apache Git Service. To respond to the

[GitHub] zhanghang1989 commented on issue #9580: BatchNorm Not Backward Correctly in Evaluation Mode

zhanghang1989 commented on issue #9580: BatchNorm Not Backward Correctly in Evaluation Mode URL: https://github.com/apache/incubator-mxnet/issues/9580#issuecomment-360928440 Solved. Should use ``loss1.backward(train_mode=False)``

[GitHub] zhreshold commented on a change in pull request #8918: Added in Large-Batch SGD with a warmup, and a LARS startegy. Also add?

zhreshold commented on a change in pull request #8918: Added in Large-Batch SGD with a warmup, and a LARS startegy. Also add? URL: https://github.com/apache/incubator-mxnet/pull/8918#discussion_r164242034 ## File path: python/mxnet/lr_scheduler.py ## @@ -136,3 +136,42 @@

[GitHub] aaronmarkham commented on issue #9535: Tutorials index page update

aaronmarkham commented on issue #9535: Tutorials index page update URL: https://github.com/apache/incubator-mxnet/pull/9535#issuecomment-360927660 Based on feedback, I simplified the selector to reflect the current arrangement and ordering of the 22 Gluon/MXNet (v1.0) tutorials.

[GitHub] zhanghang1989 opened a new issue #9580: BatchNorm Not Backward Correctly in Evaluation Mode

zhanghang1989 opened a new issue #9580: BatchNorm Not Backward Correctly in Evaluation Mode URL: https://github.com/apache/incubator-mxnet/issues/9580 ## Description I found the evaluation mode backward is already supported in the source code

[incubator-mxnet] branch master updated: fix vision datasets doc (#9576)

This is an automated email from the ASF dual-hosted git repository. haibin pushed a commit to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git The following commit(s) were added to refs/heads/master by this push: new 7fb8154 fix vision datasets doc

[GitHub] eric-haibin-lin closed pull request #9576: fix vision datasets doc

eric-haibin-lin closed pull request #9576: fix vision datasets doc URL: https://github.com/apache/incubator-mxnet/pull/9576 This is a PR merged from a forked repository. As GitHub hides the original diff on merge, it is displayed below for the sake of provenance: As this is a foreign

[GitHub] marcoabreu commented on issue #9579: Reverting namespace change

marcoabreu commented on issue #9579: Reverting namespace change URL: https://github.com/apache/incubator-mxnet/pull/9579#issuecomment-360923184 @srochel This is an automated message from the Apache Git Service. To respond to

[GitHub] thomelane commented on issue #7943: Request: Visualization for Gluon neural networks

thomelane commented on issue #7943: Request: Visualization for Gluon neural networks URL: https://github.com/apache/incubator-mxnet/issues/7943#issuecomment-360896662 You're able to plot if you've got a HybridBlock, since you can convert to a Symbol by hybridizing and passing an input

[GitHub] thomelane commented on issue #7943: Request: Visualization for Gluon neural networks

thomelane commented on issue #7943: Request: Visualization for Gluon neural networks URL: https://github.com/apache/incubator-mxnet/issues/7943#issuecomment-360896662 Until this is added, you might be able to work with `_cached_graph` property of HybridBlocks. ``` import mxnet

[GitHub] nswamy opened a new pull request #9579: Reverting namespace change

nswamy opened a new pull request #9579: Reverting namespace change URL: https://github.com/apache/incubator-mxnet/pull/9579 Reverting namespace change since it breaks backward compatibility and needs a major version bump up. Revert "Refactor scala code/tests/examples/docs to use

[GitHub] zheng-da commented on a change in pull request #8302: Refactor operators & MKLDNN

zheng-da commented on a change in pull request #8302: Refactor operators & MKLDNN URL: https://github.com/apache/incubator-mxnet/pull/8302#discussion_r164232807 ## File path: src/kvstore/kvstore_dist.h ## @@ -32,11 +32,6 @@ #include "mxnet/engine.h" #include "ps/ps.h"

[GitHub] anirudh2290 commented on a change in pull request #9570: Add checksum for file download in test_consistency

anirudh2290 commented on a change in pull request #9570: Add checksum for file download in test_consistency URL: https://github.com/apache/incubator-mxnet/pull/9570#discussion_r164232269 ## File path: tests/python/gpu/test_forward.py ## @@ -41,8 +42,19 @@ def

[GitHub] anirudh2290 opened a new pull request #9578: Remove non-necessary RuntimeError

anirudh2290 opened a new pull request #9578: Remove non-necessary RuntimeError URL: https://github.com/apache/incubator-mxnet/pull/9578 ## Description ## utils.download already raises an Exception if the checksum doesn't match. Not needed to do it twice. @szha @eric-haibin-lin

[GitHub] szha commented on issue #9576: fix vision datasets doc

szha commented on issue #9576: fix vision datasets doc URL: https://github.com/apache/incubator-mxnet/pull/9576#issuecomment-360912904 Before: https://mxnet.incubator.apache.org/versions/master/api/python/gluon/data.html#vision After:

[GitHub] eric-haibin-lin opened a new pull request #9577: Disable flaky test: test_correlation

eric-haibin-lin opened a new pull request #9577: Disable flaky test: test_correlation URL: https://github.com/apache/incubator-mxnet/pull/9577 #9553 @marcoabreu @anirudh2290 This is an automated message from the Apache

  1   2   >