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This repository has been archived by the owner on Nov 17, 2023.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

Numpy-compatible Infra #15581

Merged
merged 46 commits into from
Aug 8, 2019
Merged

Numpy-compatible Infra #15581

merged 46 commits into from
Aug 8, 2019

Commits on Aug 7, 2019

  1. [Do not review] [Do not merge] New numpy-compatible sum (apache#14739) 

    * Add numpy namespace and initial impl of np.sum (not complete)

* Clean up

* Fix import error

* numpy sum

* add test and backward data type support

* add license to test_numpy_op.py

* improve test to reduce flakiness

* fix sanity build

* extra numeric test and imperative test

* add error message for initial argument
    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    2e09926 View commit details
    Browse the repository at this point in the history

  2. [numpy] Infra for supporting numpy ops in imperative mode and Gluon A… 

    …PIs (apache#14758)

* Infra of new ndarray and symbol types for numpy operators

* Rename

* Fix import problem

* Refactor

* Remove redundant code

* Add docstring

* More on numpy ndarray and symbol

* Override unimplemented methdos for ndarray and _NumpySymbol

* Fix built-in methods of ndarray and _NumpySymbol

* Fix test and sanity check

* Fix pylint

* Address cr comments

* Add unit tests for ndarray and _NumpySymbol

* Add _true_divide

* Fix gpu build

* Add future import division

* More correct way of checking if an output is from a np compat op

* Fix gpu build

* Fix output ndarray/symbol types with at least one new ndarray/symbol

* Modify true_divide doc

* Fix flaky copying zero-size arrays via gpus

* Fix zero size in gluon hybridize and zeros/ones symbol not creating new symbol type

* Fix doc
    @reminisce @haojin2
    reminisce authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    c8127ae View commit details
    Browse the repository at this point in the history

  3. Enable np op compat check with name prefix (apache#14897)

    @reminisce @haojin2
    reminisce authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    3a437ad View commit details
    Browse the repository at this point in the history

  4. [numpy] Numpy dot (apache#14831) 

    * Numpy Dot case 1-4 + case 3.5 forward and 0.5 backward

* Backward computation and test coverage
    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    2657af1 View commit details
    Browse the repository at this point in the history

  5. numpy-compatible mean (apache#14859)

    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    6c461f2 View commit details
    Browse the repository at this point in the history

  6. [numpy] Some np ops for d2l (apache#14924) 

    * Add np transpose

More ops and namespaces for submodules

Add relu and sigmoid

Add reshape

Fix symbolic name mismatch

Add maximum and minimum

* Add convenience fluent method

* Add ndarray.item()

* Fix CI

* Fix lint

* Fix lint

* Fix reshape gpu

* Add example

* Remove python notebook outputs

* Remove notebook output

* Add one more example
    @reminisce @haojin2
    reminisce authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    162eaa7 View commit details
    Browse the repository at this point in the history

  7. [numpy] Refactor np modules (apache#14989) 

    * Refactor

* Initial refactoring

* Fix notebook

* Move numpy op check from backend to frontend

* Add homogeneous ndarray check

* Fix grouping inhomogeneous types of symbols

* Improve error handling of different types of symbols as outputs

* Fix test

* Fix numpy test

* Fix ci

* Try to fix gpu ci failure
    @reminisce @haojin2
    reminisce authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    aeaeb46 View commit details
    Browse the repository at this point in the history

  8. [numpy] Refactor np module (example runs through) (apache#15055) 

    * Refactor notebook

* notebook working with hybrid block

* More refactoring

* Remove unnecessary use_np_compat

* Use class decorator to initialize numpy ndarrays in parameter.py

* Clear notebook outputs

* Improve np decorator

* Remove npe op from optimizer

* Fix CI

* Fix functools.wraps issue in Python2

* Fix ci
    @reminisce @haojin2
    reminisce authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    846a335 View commit details
    Browse the repository at this point in the history

  9. Change np_compat to np_shape

    @reminisce @haojin2
    reminisce authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    fbd0a3b View commit details
    Browse the repository at this point in the history

  10. Temporarily disable test_amp

    @reminisce @haojin2
    reminisce authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    131dbe3 View commit details
    Browse the repository at this point in the history

  11. Numpy-compatible stack (apache#15027) 

    * numpy stack

* migrate to use_np_shape
    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    4fe6cad View commit details
    Browse the repository at this point in the history

  12. Numpy Unary Ops (apache#15010) 

    * Unary Ops

* new version of unit tests
    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    25277a5 View commit details
    Browse the repository at this point in the history

  13. [numpy] Fix np branch after rebase (apache#15086) 

    * Add np_array semantics for Gluon

Fix notebook

Fix sanity

Fix gluon deferred infer shape

Add np.random.uniform

Add random normal

Add boolean comparison ops

Add np.ndarray indexing

Reformat test ndarray indexing

Fix unit tests

Add one more test of indexing

Fix sanity

Enable amp test

Add np.arange

Revert cython unit test to ctypes

Delete unnecessary use_np_shape decorator from test

Rebase with numpy branch

support range as index

Fix python2 range type check

Add argmax

Disable clojure test

* Fix ci

* Add np.linalg.norm for ord='fro'

* Fix pylint
    @reminisce @haojin2
    reminisce authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    9dc5e0a View commit details
    Browse the repository at this point in the history

  14. numpy concatenate (apache#15104)

    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    5d8f125 View commit details
    Browse the repository at this point in the history

  15. [WIP][numpy] Fix for D2L Chapters 2/3/4 (apache#15139) 

    * Fix

* Fix linear regression gluon

* More fix

* Fix pylint

* Fix for chapter 4

* Add np.add mul div mod pow sub and shuffle

* Fix model selection, underfitting, overfitting

* Fix weight decay

* Fix dropout

* Fix

* Fix chapter 4
    @reminisce @haojin2
    reminisce authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    b4716a9 View commit details
    Browse the repository at this point in the history

  16. [numpy] Fix d2l performance regression (apache#15173) 

    * Add np array adapter decorator for layers

* Fix performance regression caused by too many conversions between nd.NDArray and np.ndarray

* Fix pylint

* Fix test backward compatibility issue

* Fix test_lambda
    @reminisce @haojin2
    reminisce authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    049ded2 View commit details
    Browse the repository at this point in the history

  17. Fix (apache#15188)

    @reminisce @haojin2
    reminisce authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    a584326 View commit details
    Browse the repository at this point in the history

  18. fix for chapter6 conv nn (apache#15224)

    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    0d0f284 View commit details
    Browse the repository at this point in the history

  19. [numpy] Fix d2l chapter8 (apache#15237) 

    * Add np op doc

* Fix several issues

* Add a N-D dot b 2D support

* Simplify array creation api

* Add swapaxes

* Fix rnn gluon

* More fix

* Fix pylint

* Delete

* Fix mp windows
    @reminisce @haojin2
    reminisce authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    1d47418 View commit details
    Browse the repository at this point in the history

  20. fix for ch11 (apache#15244)

    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    8a2b41f View commit details
    Browse the repository at this point in the history

  21. Numpy-compatible split (apache#15049) 

    * numpy split

* numpy split

* unit test

* unit test
    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    820752f View commit details
    Browse the repository at this point in the history

  22. [numpy] [DO NOT MERGE] Fix d2l chapters 9 and 13 (apache#15246) 

    * Add npx batch_dot and topk

* Text embedding uses numpy

* Fix SoftmaxCrossEntropyLoss with np

* Fix sentiment cnn

* Fix pylint

* Fix dot attention

* Fix seq2seq attention

* Add np.tile

* Fix transformer

* Fix ci

* Fix ci and rebase
    @reminisce @haojin2
    reminisce authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    21be6f8 View commit details
    Browse the repository at this point in the history

  23. [numpy] Fix d2l chapter 5 (apache#15264) 

    * Fix parameter initializer

* Add np.save and np.load

* Fix read-write

* Fix lint
    @reminisce @haojin2
    reminisce authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    a54a3f2 View commit details
    Browse the repository at this point in the history

  24. Numpy compatible max (apache#15161) 

    * numpy amax

* weird cu file diff

* fix the unit test error

* fix gpu bug

* minor fix

* fix lint

* remove scalar value check

* fix the bug on unit test

* fix the case () that breaks the kernel launch

* add zero dimension unit test

* revert the tuple change

* use mshadow maximum

* remove test zero

* change the macro for now

* change the cuda to use mashadow op

* fix the broadcast_reduce_op_value.cu wrong kernel

* add more logic in shape to detect the invalid situation

* change back to type swtich

* change to as_nd_ndarray

* add missing @npx.use_np_shape

* retrigger CI

* address the comment

* undo algorithm import

* remove the numeric gradient check
    @stu1130 @haojin2
    stu1130 authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    6bd552f View commit details
    Browse the repository at this point in the history

  25. Numpy compatible multinomial (apache#15219) 

    * draft of multinomial

* rename to more concise name

* finish shape

* complete the forward function

* complete forward without handle 0 dimension & scalar

* handle 0 dimension

* add new line

* fix lint

* fix the build error

* fix lint

* finish unit test

* change the registration

* make multinomial support pvals as mx.ndarray

* delete newline

* fix lint error

* support input as list, mx.ndarray, np.ndarray & unit test

* fix lint

* fix the include error

* fix lint

* refactor & pass the tensor instead of tuple to kernel

* fix lint

* updata the doc

* address the comment
    @stu1130 @haojin2
    stu1130 authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    42d6760 View commit details
    Browse the repository at this point in the history

  26. Numpy compatible linspace (apache#15256) 

    * draft

* finish linspace implementation

* finish linspace

* delete newline

* fix pylint

* add more unit test

* address comment

* add more test case

* disable too-many-arguments

* resolve confliction

* add ctx
    @stu1130 @haojin2
    stu1130 authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    9f8d4a4 View commit details
    Browse the repository at this point in the history

  27. numpy-compatible cumsum (apache#15309)

    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    12aab7a View commit details
    Browse the repository at this point in the history

  28. [numpy] Misc fix for other chapters (apache#15332) 

    * Add np.prod

* Fix ndarray.reshape accepting positional integers as arguments

* Rebase

* Fix rebase error

* Add np.ndarray.flatten

* Fix

* Add broadcast_to

* Add meshgrid and broadcast_arrays

* Fix sin, cos, sinh, cosh not supporting scalars

* Add more unary ops supporting python scalars

* Fix

* Fix

* Fix ci

* Fix sanity
    @reminisce @haojin2
    reminisce authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    9cc355f View commit details
    Browse the repository at this point in the history

  29. [numpy] Change d2l chapters cv and gan to use numpy (apache#15368) 

    * Change op name style to lower case underscore

* Add ops under image to npx

* Add image submodule to npx

* Fix split_and_load use np

* Fix fine tuning

* Fix bbox and anchor

* Fix odd

* Fix ssd and rcnn

* Remove restriction on binary element-wise scalar

* Fix gan

* Fix sanity

* Try to fix website build failure

* Add npx.random.seed

* Fix doc
    @reminisce @haojin2
    reminisce authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    78c541f View commit details
    Browse the repository at this point in the history

  30. add doc for multinomial, dot, cumsum, clip, abs, exp, arctan (apache#… 

    …15386)
    @hzfan @haojin2
    hzfan authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    5218a09 View commit details
    Browse the repository at this point in the history

  31. [numpy] Fix several places in numpy (apache#15398) 

    * Fix

* More fix
    @reminisce @haojin2
    reminisce authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    99a9b0a View commit details
    Browse the repository at this point in the history

  32. [numpy] fix cython (apache#15418) 

    * add cython support for numpy

* stay with original API for backward compatibility
    @hzfan @haojin2
    hzfan authored and haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    7caacd8 View commit details
    Browse the repository at this point in the history

  33. fix after rebase

    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    a8869b6 View commit details
    Browse the repository at this point in the history

  34. get rid of coverage in clang60 mkldnn

    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    33f1cfb View commit details
    Browse the repository at this point in the history

  35. fix lint issues

    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    56ac957 View commit details
    Browse the repository at this point in the history

  36. fix flaky test and get rid of extra print

    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    ba54b26 View commit details
    Browse the repository at this point in the history

  37. remove numpy examples

    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    6b84d53 View commit details
    Browse the repository at this point in the history

  38. revert apache#15309 apache#15256 apache#15219 apache#15161

    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    7e8deab View commit details
    Browse the repository at this point in the history

  39. remove numpy docs

    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    0b6f2c8 View commit details
    Browse the repository at this point in the history

  40. remove changes to contrib/text/embedding.py

    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    b2e48d9 View commit details
    Browse the repository at this point in the history

  41. remove numpy changes to gluon peripherals

    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    7ede112 View commit details
    Browse the repository at this point in the history

  42. Revert "remove numpy docs" 

    This reverts commit c104695.
    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    024bedb View commit details
    Browse the repository at this point in the history

  43. get rid of most operators

    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    0e955a3 View commit details
    Browse the repository at this point in the history

  44. Revert "get rid of coverage in clang60 mkldnn" 

    This reverts commit 77dc905.
    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    3262591 View commit details
    Browse the repository at this point in the history

  45. remove np-compatible from mxnet.image mxnet.initializer

    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    275b063 View commit details
    Browse the repository at this point in the history

  46. address comments

    @haojin2
    haojin2 committed Aug 7, 2019
    Configuration menu
    Copy the full SHA
    345c522 View commit details
    Browse the repository at this point in the history