[jira] [Updated] (MXNET-702) Dangling outputs and dtype != float32: Gradient computation fails

[ https://issues.apache.org/jira/browse/MXNET-702?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Vandana Kannan updated MXNET-702: - Status: Done (was: In Progress) > Dangling outputs and dtype != float32: Gradient computation

[jira] [Updated] (MXNET-666) RNN BucketSentenceIter - Must load all data into memory?

[ https://issues.apache.org/jira/browse/MXNET-666?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Vandana Kannan updated MXNET-666: - Resolution: Feedback Received Status: Done (was: To Do) Submitter closed the issue as

[jira] [Created] (MXNET-830) Training with multiple features for LSTM

Ankit Khedia created MXNET-830: -- Summary: Training with multiple features for LSTM Key: MXNET-830 URL: https://issues.apache.org/jira/browse/MXNET-830 Project: Apache MXNet Issue Type:

[jira] [Commented] (MXNET-799) Gradient function not returning enough gradient

[ https://issues.apache.org/jira/browse/MXNET-799?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16581630#comment-16581630 ] Lin Yuan commented on MXNET-799: Not able to reproduce. > Gradient function not returning enough gradient

[jira] [Updated] (MXNET-799) Gradient function not returning enough gradient

[ https://issues.apache.org/jira/browse/MXNET-799?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Lin Yuan updated MXNET-799: --- Status: Done (was: In Progress) > Gradient function not returning enough gradient >

[jira] [Created] (MXNET-824) LSTM work for this multidimensional sequence

Ankit Khedia created MXNET-824: -- Summary: LSTM work for this multidimensional sequence Key: MXNET-824 URL: https://issues.apache.org/jira/browse/MXNET-824 Project: Apache MXNet Issue Type:

[jira] [Assigned] (MXNET-824) LSTM work for this multidimensional sequence

[ https://issues.apache.org/jira/browse/MXNET-824?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Ankit Khedia reassigned MXNET-824: -- Assignee: Ankit Khedia > LSTM work for this multidimensional sequence >

[jira] [Created] (MXNET-826) Autograd/Autodiff in R API

Anirudh Acharya created MXNET-826: - Summary: Autograd/Autodiff in R API Key: MXNET-826 URL: https://issues.apache.org/jira/browse/MXNET-826 Project: Apache MXNet Issue Type: New Feature

[jira] [Assigned] (MXNET-830) Training with multiple features for LSTM

[ https://issues.apache.org/jira/browse/MXNET-830?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Ankit Khedia reassigned MXNET-830: -- Assignee: Ankit Khedia > Training with multiple features for LSTM >

[jira] [Created] (MXNET-831) R-package with GPU, speed-up much less for regression example than classification

Anirudh Acharya created MXNET-831: - Summary: R-package with GPU, speed-up much less for regression example than classification Key: MXNET-831 URL: https://issues.apache.org/jira/browse/MXNET-831

[jira] [Updated] (MXNET-671) How to specify input format for charRNN model in MXNet-R

[ https://issues.apache.org/jira/browse/MXNET-671?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Ankit Khedia updated MXNET-671: --- Status: In Progress (was: To Do) > How to specify input format for charRNN model in MXNet-R >

[jira] [Updated] (MXNET-783) Describe input format for charRNN model

[ https://issues.apache.org/jira/browse/MXNET-783?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Ankit Khedia updated MXNET-783: --- Status: In Progress (was: Done) Resolution: (was: Fixed) > Describe input format for

[jira] [Assigned] (MXNET-827) Input format for mx.rnn

[ https://issues.apache.org/jira/browse/MXNET-827?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Ankit Khedia reassigned MXNET-827: -- Assignee: Ankit Khedia > Input format for mx.rnn > > >

[jira] [Created] (MXNET-828) Tutorial for Deep Reinforcement Learning

Anirudh Acharya created MXNET-828: - Summary: Tutorial for Deep Reinforcement Learning Key: MXNET-828 URL: https://issues.apache.org/jira/browse/MXNET-828 Project: Apache MXNet Issue Type:

[jira] [Created] (MXNET-827) Input format for mx.rnn

Ankit Khedia created MXNET-827: -- Summary: Input format for mx.rnn Key: MXNET-827 URL: https://issues.apache.org/jira/browse/MXNET-827 Project: Apache MXNet Issue Type: Improvement

[jira] [Updated] (MXNET-784) Need to have multiple loss function while training in R

[ https://issues.apache.org/jira/browse/MXNET-784?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Lin Yuan updated MXNET-784: --- Sprint: (was: ) > Need to have multiple loss function while training in R >

[jira] [Updated] (MXNET-823) Make NDArrays mutable in the R API

[ https://issues.apache.org/jira/browse/MXNET-823?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Anirudh Acharya updated MXNET-823: -- Description: Python API's ndarray objects are mutable, whereas in the R API the ndarray

[jira] [Created] (MXNET-823) Make NDArrays mutable in the R API

Anirudh Acharya created MXNET-823: - Summary: Make NDArrays mutable in the R API Key: MXNET-823 URL: https://issues.apache.org/jira/browse/MXNET-823 Project: Apache MXNet Issue Type:

[jira] [Updated] (MXNET-633) Fix the begin_round for epochs while training in R

[ https://issues.apache.org/jira/browse/MXNET-633?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Ankit Khedia updated MXNET-633: --- Resolution: Fixed Status: Done (was: To Do) > Fix the begin_round for epochs while training

[jira] [Updated] (MXNET-803) Inconsistent / wrong output from sum

[ https://issues.apache.org/jira/browse/MXNET-803?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Hao Jin updated MXNET-803: -- Status: In Review (was: In Progress) > Inconsistent / wrong output from sum >

[jira] [Updated] (MXNET-802) Autograd fails when using `take` operator repeatedly

[ https://issues.apache.org/jira/browse/MXNET-802?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Hao Jin updated MXNET-802: -- Status: In Progress (was: To Do) > Autograd fails when using `take` operator repeatedly >

[jira] [Updated] (MXNET-801) non-deterministic backward of scatter_nd

[ https://issues.apache.org/jira/browse/MXNET-801?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Hao Jin updated MXNET-801: -- Resolution: Not A Bug Status: Done (was: To Do) > non-deterministic backward of scatter_nd >

[jira] [Updated] (MXNET-802) Autograd fails when using `take` operator repeatedly

[ https://issues.apache.org/jira/browse/MXNET-802?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Hao Jin updated MXNET-802: -- Status: In Review (was: In Progress) > Autograd fails when using `take` operator repeatedly >

[jira] [Updated] (MXNET-799) Gradient function not returning enough gradient

[ https://issues.apache.org/jira/browse/MXNET-799?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Lin Yuan updated MXNET-799: --- Status: In Progress (was: To Do) > Gradient function not returning enough gradient >

[jira] [Updated] (MXNET-811) test_arange failure - Jetson TX2 (CPU)

[ https://issues.apache.org/jira/browse/MXNET-811?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Hao Jin updated MXNET-811: -- Status: In Progress (was: To Do) > test_arange failure - Jetson TX2 (CPU) >

[jira] [Updated] (MXNET-804) nd.softmax() doesn't support grad_req='add'

[ https://issues.apache.org/jira/browse/MXNET-804?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Hao Jin updated MXNET-804: -- Status: In Progress (was: To Do) > nd.softmax() doesn't support grad_req='add' >

[jira] [Updated] (MXNET-811) test_arange failure - Jetson TX2 (CPU)

[ https://issues.apache.org/jira/browse/MXNET-811?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Hao Jin updated MXNET-811: -- Status: In Review (was: In Progress) > test_arange failure - Jetson TX2 (CPU) >

[jira] [Updated] (MXNET-803) Inconsistent / wrong output from sum

[ https://issues.apache.org/jira/browse/MXNET-803?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Hao Jin updated MXNET-803: -- Status: In Progress (was: To Do) > Inconsistent / wrong output from sum >

[jira] [Updated] (MXNET-786) Specify layer specific learning rate while training in R

[ https://issues.apache.org/jira/browse/MXNET-786?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Anirudh Acharya updated MXNET-786: -- Priority: Major (was: Critical) > Specify layer specific learning rate while training in R >

[jira] [Updated] (MXNET-786) Specify layer specific learning rate while training in R

[ https://issues.apache.org/jira/browse/MXNET-786?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Anirudh Acharya updated MXNET-786: -- Issue Type: New Feature (was: Bug) > Specify layer specific learning rate while training in R

[jira] [Updated] (MXNET-785) How to create new operators in mxnet-R

[ https://issues.apache.org/jira/browse/MXNET-785?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Anirudh Acharya updated MXNET-785: -- Priority: Major (was: Critical) > How to create new operators in mxnet-R >

[jira] [Assigned] (MXNET-831) R-package with GPU, speed-up much less for regression example than classification

[ https://issues.apache.org/jira/browse/MXNET-831?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Ankit Khedia reassigned MXNET-831: -- Assignee: Ankit Khedia > R-package with GPU, speed-up much less for regression example than >

[jira] [Assigned] (MXNET-787) Call im2Rec in R API

[ https://issues.apache.org/jira/browse/MXNET-787?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Ankit Khedia reassigned MXNET-787: -- Assignee: Ankit Khedia > Call im2Rec in R API > > > Key:

[jira] [Assigned] (MXNET-787) Call im2Rec in R API

[ https://issues.apache.org/jira/browse/MXNET-787?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Anirudh Acharya reassigned MXNET-787: - Assignee: (was: Anirudh Acharya) > Call im2Rec in R API > > >

[jira] [Assigned] (MXNET-833) Fix char RNN example

[ https://issues.apache.org/jira/browse/MXNET-833?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Ankit Khedia reassigned MXNET-833: -- Assignee: Ankit Khedia > Fix char RNN example > > > Key:

[jira] [Updated] (MXNET-831) R-package with GPU, speed-up much less for regression example than classification

[ https://issues.apache.org/jira/browse/MXNET-831?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Anirudh Acharya updated MXNET-831: -- Priority: Critical (was: Major) > R-package with GPU, speed-up much less for regression

[jira] [Updated] (MXNET-829) MXNET CPU WORKER THREADS usage in R

[ https://issues.apache.org/jira/browse/MXNET-829?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Anirudh Acharya updated MXNET-829: -- Summary: MXNET CPU WORKER THREADS usage in R (was: MXNET COU WORKER THREADS usage in R) >

[jira] [Updated] (MXNET-829) MXNET CPU WORKER THREADS usage in R

[ https://issues.apache.org/jira/browse/MXNET-829?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Anirudh Acharya updated MXNET-829: -- Priority: Critical (was: Major) > MXNET CPU WORKER THREADS usage in R >

[jira] [Created] (MXNET-833) Fix char RNN example

Ankit Khedia created MXNET-833: -- Summary: Fix char RNN example Key: MXNET-833 URL: https://issues.apache.org/jira/browse/MXNET-833 Project: Apache MXNet Issue Type: Improvement

[jira] [Assigned] (MXNET-636) R crashes while training/transfer-learning with vgg16

[ https://issues.apache.org/jira/browse/MXNET-636?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Ankit Khedia reassigned MXNET-636: -- Assignee: Ankit Khedia > R crashes while training/transfer-learning with vgg16 >

[jira] [Assigned] (MXNET-787) Call im2Rec in R API

[ https://issues.apache.org/jira/browse/MXNET-787?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Anirudh Acharya reassigned MXNET-787: - Assignee: Anirudh Acharya > Call im2Rec in R API > > >

[jira] [Assigned] (MXNET-829) MXNET CPU WORKER THREADS usage in R

[ https://issues.apache.org/jira/browse/MXNET-829?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Anirudh Acharya reassigned MXNET-829: - Assignee: Anirudh Acharya > MXNET CPU WORKER THREADS usage in R >

[jira] [Commented] (MXNET-672) Newly built mxnet binaries encounters out of memory error with mx.nd.internal.as.array(nd)

[ https://issues.apache.org/jira/browse/MXNET-672?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16581728#comment-16581728 ] Anirudh Acharya commented on MXNET-672: --- Will be part of fixing C-GAN tutorial > Newly built mxnet

[jira] [Assigned] (MXNET-672) Newly built mxnet binaries encounters out of memory error with mx.nd.internal.as.array(nd)

[ https://issues.apache.org/jira/browse/MXNET-672?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Anirudh Acharya reassigned MXNET-672: - Assignee: Anirudh Acharya > Newly built mxnet binaries encounters out of memory error

[jira] [Updated] (MXNET-824) make LSTM work for this multidimensional sequence

[ https://issues.apache.org/jira/browse/MXNET-824?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Ankit Khedia updated MXNET-824: --- Summary: make LSTM work for this multidimensional sequence (was: LSTM work for this multidimensional

[jira] [Updated] (MXNET-824) make LSTM work for this multidimensional sequence

[ https://issues.apache.org/jira/browse/MXNET-824?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Anirudh Acharya updated MXNET-824: -- Priority: Critical (was: Major) > make LSTM work for this multidimensional sequence >

[jira] [Updated] (MXNET-827) Input format for mx.rnn

[ https://issues.apache.org/jira/browse/MXNET-827?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Anirudh Acharya updated MXNET-827: -- Priority: Critical (was: Major) > Input format for mx.rnn > > >

[jira] [Updated] (MXNET-785) How to create new operators in mxnet-R

[ https://issues.apache.org/jira/browse/MXNET-785?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Anirudh Acharya updated MXNET-785: -- Issue Type: Task (was: Bug) > How to create new operators in mxnet-R >