[GitHub] jeremiedb opened a new pull request #9022: R RNN API fixes and Optimizer clip gradient on NDArray

[GitBox <gi...@apache.org>]

jeremiedb opened a new pull request #9022: R RNN API fixes and Optimizer clip gradient on NDArray
URL: https://github.com/apache/incubator-mxnet/pull/9022
 
 
   ## Description ##
   Rework of the RNN API: 
   - Change the input to be of shape `seq_len X batch.size`, this will ease the use of existing arrayIter on NLP models. 
   - Add support for inference on explicit unrolled models (used when training on CPU rather than the `symbol.RNN`)
   - Inference support for both one-to-one (language model, time-series) and seq-to-one (classification). 
   
   Optimizers:
   For all optimizers, gradient clipping is now performed by `mx.nd.clip`, avoiding to back and forth with R array resulting in some speedup. 
   
   Following comments on lack of exmaples supporting the R package, exemples have been added: 
   https://jeremiedb.github.io/mxnet_R_bucketing/index.html
   If there's interest, idea to is add other examples, in the spirit of the gluon.mxnet.io notebooks
   
   ## Checklist ##
   ### Essentials ###
   - [ ] Passed code style checking (`make lint`)
   - [ ] Changes are complete (i.e. I finished coding on this PR)
   - [ ] All changes have test coverage:
   - Unit tests are added for small changes to verify correctness (e.g. adding a new operator)
   - Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore)
   - Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL)
   - [ ] Code is well-documented: 
   - For user-facing API changes, API doc string has been updated. 
   - For new C++ functions in header files, their functionalities and arguments are documented. 
   - For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable
   - [ ] To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change
   
   ! Need to fix example: `example\rnn\bucket_R\`

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on 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