* Small batch norm fix (cuda/no-mkldnn)
Signed-off-by: Alex Black <blacka101@gmail.com>
* Dropout fix for RnnOutputLayer
Signed-off-by: Alex Black <blacka101@gmail.com>
* Allow block size < 2 in batch_to_space_nd and space_to_batch_nd for import, in spite of what TF docs say
Signed-off-by: AlexDBlack <blacka101@gmail.com>
* Nd4j pad update
Signed-off-by: Ryan Nett <rnett@skymind.io>
* switched from guava Immutables to Collections.unmodifiableList/Map
Signed-off-by: Ryan Nett <rnett@skymind.io>
* javadoc
Signed-off-by: Ryan Nett <rnett@skymind.io>
* use new pad
Signed-off-by: Ryan Nett <rnett@skymind.io>
* conv tests use OpValidation
Signed-off-by: Ryan Nett <rnett@skymind.io>
* deconv3d overrides
Signed-off-by: Ryan Nett <rnett@skymind.io>
* test fix for the new pad method
Signed-off-by: Ryan Nett <rnett@skymind.io>
* more test fixes
Signed-off-by: Ryan Nett <rnett@skymind.io>
* more test fixes
Signed-off-by: Ryan Nett <rnett@skymind.io>
* rename SameDiff function methods to op (except for the actual SameDiff function ones)
Signed-off-by: Ryan Nett <rnett@skymind.io>
* more pad overloads, test fix
Signed-off-by: Ryan Nett <rnett@skymind.io>
* test updates
Signed-off-by: Ryan Nett <rnett@skymind.io>
* conv1d test
Signed-off-by: Ryan Nett <rnett@skymind.io>
* remove Conv1D tf import (there isn't a TF conv1d op)
Signed-off-by: Ryan Nett <rnett@skymind.io>
* remove numThreads from Nd4j
Signed-off-by: Ryan Nett <rnett@skymind.io>
* replace Old ops with their newer versions, deprecate ones that haven't already been deprecated
Signed-off-by: Ryan Nett <rnett@skymind.io>
* remove use of setNumThreads
Signed-off-by: Ryan Nett <rnett@skymind.io>
* fix for Reverse and ATan2
Signed-off-by: Ryan Nett <rnett@skymind.io>
* fix test for wrong equals type
Signed-off-by: Ryan Nett <rnett@skymind.io>
* well it works now
Signed-off-by: Ryan Nett <rnett@skymind.io>
* better javadocs
Signed-off-by: Ryan Nett <rnett@skymind.io>
* NonNulls
Signed-off-by: Ryan Nett <rnett@skymind.io>
* better array literal
Signed-off-by: Ryan Nett <rnett@skymind.io>
* re-add tf import stuff (will remove later)
Signed-off-by: Ryan Nett <rnett@skymind.io>
* conv1d config load fix
Signed-off-by: Ryan Nett <rnett@skymind.io>
* partial config usage changes
Signed-off-by: Ryan Nett <rnett@skymind.io>
* remove Old op classes
Signed-off-by: Ryan Nett <rnett@skymind.io>
* config property fixes
Signed-off-by: Ryan Nett <rnett@skymind.io>
* removed one too many ops
Signed-off-by: Ryan Nett <rnett@skymind.io>
* - start working on space_to_batch_nd
Signed-off-by: Yurii <yurii@skymind.io>
* - provide cpu helper for space_to_batch_nd op
Signed-off-by: Yurii <yurii@skymind.io>
* few typos fixed
Signed-off-by: raver119 <raver119@gmail.com>
* - add tests for space_to_batch and correct bugs
Signed-off-by: Yurii <yurii@skymind.io>
* - write cuda kernel for space_to_batch op
Signed-off-by: Yurii <yurii@skymind.io>
* - add order argument to shape::index2coords method in convolution cuda ops
Signed-off-by: Yurii <yurii@skymind.io>
* - restore some previous code
Signed-off-by: Yurii <yurii@skymind.io>
* old col2im kernel activated
Signed-off-by: raver119 <raver119@gmail.com>
* - change coords calculation in col2im kernel
Signed-off-by: Yurii <yurii@skymind.io>
* - restore old col2im kernel
Signed-off-by: Yurii <yurii@skymind.io>
* - add custom op for batch_to_space
Signed-off-by: Yurii <yurii@skymind.io>
* - provide cpu version for batch_to_space_nd op
Signed-off-by: Yurii <yurii@skymind.io>
* - provide cuda kernel for batch_to_space_nd op
Signed-off-by: Yurii <yurii@skymind.io>
* Jar packaging for maven
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Typo fixed
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* minimal viable prototype for SD
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Tests corrected
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* few fixes for bfloat16 in java and cpp (#114)
Signed-off-by: raver119 <raver119@gmail.com>
* Nd4j refactoring (#112)
* refactoring
Signed-off-by: Robert Altena <Rob@Ra-ai.com>
* wip
Signed-off-by: Robert Altena <Rob@Ra-ai.com>
* wip
Signed-off-by: Robert Altena <Rob@Ra-ai.com>
* wip
* fix: make test public.
Signed-off-by: Robert Altena <Rob@Ra-ai.com>
* make test public.
Signed-off-by: Robert Altena <Rob@Ra-ai.com>
* fixes read refactoring.
Signed-off-by: Robert Altena <Rob@Ra-ai.com>
* Enabled test
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Test copied from nd4j
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* [WIP] bitwise ops (#115)
* - cyclic_shift_bits + test
- shift_bits + test
Signed-off-by: raver119 <raver119@gmail.com>
* OMP_IF replacement
Signed-off-by: raver119 <raver119@gmail.com>
* Thin wrapper added
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Cleanup
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Shugeo cuda tests (#116)
* Added tests for get_seed/set_seed ops.
* Added missed tests for scatter_sub/mul/div ops.
* Added tests for hardsigmoid and hardsigmoid_bp.
* Added tests for hardtanh and hardtanh_bp ops.
* Added test for histogram op.
* Added tests for identity op.
* Refactored mergemaxindex op. Added tests for log1p,mergemaxindex, mod and mod_bp ops.
* Fixed tests for FloorDiv.
* Added test for rank op.
* Added tests for rationaltanh/rationaltanh_bp ops.
* Added tests for realdiv/realdiv_bp.
* Added tests for rectifiedtanh/_bp ops.
* Added tests for shapes_of op.
* Added tests for shapes_of op.
* Added tests for size op.
* Added tests for softplus/_bp ops.
* Added tests for softsign/_bp ops.
* Added tests for toggle_bits op. Fixed processing of OP_IMPL and so on defititions.
* Added test for truncatediv op.
* Added another test for truncatediv op.
* Added another test for histogram.
* Added tests for unstack_list op.
* Refactored to_int32/uint32/float16/float32/double/int64/uint64 ops and tests.
* Refactored mergemaxindex op helper for cuda platform and tests.
* Fixed cuda kernel for histogram op helper.
* Refactor skipgram to avoid early buffers shift.
* Fixed check up with non_max_suppression op cuda helper. Added cuda kernel implementation for skipgram op helpers.
* Added implementation of skipgram op helper for cuda platform. Working revision
* Fixed mergeMaxIndex kernel and move it to separate source file.
* Adding arithmetic
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Eliminated memory leaks and dropped waste prints with tests. (#117)
* Added tests
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* fix test
Signed-off-by: raver119 <raver119@gmail.com>
* no openmp for ClipByGlobalNorm
Signed-off-by: raver119 <raver119@gmail.com>
* Stubs for ops
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* [WIP] right shift ops (#118)
* right shift ops
Signed-off-by: raver119 <raver119@gmail.com>
* typo
Signed-off-by: raver119 <raver119@gmail.com>
* rotr test
Signed-off-by: raver119 <raver119@gmail.com>
* fix: IOException no longer thrown by read(). (#120)
Signed-off-by: Robert Altena <Rob@Ra-ai.com>
* Small fix in TensorflowConversion class (#121)
Signed-off-by: Alex Black <blacka101@gmail.com>
* Shyrma concat2 (#119)
* - rewrite/improve concat
Signed-off-by: Yurii <yurii@skymind.io>
* - ged rid of unnecessary argument in concat kernel
Signed-off-by: Yurii <yurii@skymind.io>
* InferenceSession additional validation for shape calc (#122)
Signed-off-by: Alex Black <blacka101@gmail.com>
* [WIP] build fix (#124)
* AffinityManager changes
Signed-off-by: raver119 <raver119@gmail.com>
* build fixes
Signed-off-by: raver119 <raver119@gmail.com>
* OP/CONFIGURABLE_OP shapefn fix (#125)
Signed-off-by: raver119 <raver119@gmail.com>
* Some ops added
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Nd4j refactoring (last one!) (#123)
* fix: IOException no longer thrown by read().
Signed-off-by: Robert Altena <Rob@Ra-ai.com>
* refactoring
* last refactorings
Signed-off-by: Robert Altena <Rob@Ra-ai.com>
* Advanced tests
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* [WIP] Java wrappers (#126)
* shift/rshift/rotl/rotr java/sd wrappers
Signed-off-by: raver119 <raver119@gmail.com>
* few additional wrappers
Signed-off-by: raver119 <raver119@gmail.com>
* minor naming tweak
Signed-off-by: raver119 <raver119@gmail.com>
* Test added
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* one more build fix
Signed-off-by: raver119 <raver119@gmail.com>
* Jar packaging for maven
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Typo fixed
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* minimal viable prototype for SD
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Tests corrected
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Enabled test
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Test copied from nd4j
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Thin wrapper added
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Cleanup
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Adding arithmetic
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Added tests
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Stubs for ops
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Some ops added
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Advanced tests
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Test added
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Ops added
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Small build fixes (#127)
* Small build fixes
Signed-off-by: Alex Black <blacka101@gmail.com>
* Fix RL4J
Signed-off-by: Alex Black <blacka101@gmail.com>
* Test fixes
Signed-off-by: Alex Black <blacka101@gmail.com>
* Another fix
Signed-off-by: Alex Black <blacka101@gmail.com>
* parent module name fix
Signed-off-by: raver119 <raver119@gmail.com>
* [WIP] Roll rewritten (#128)
* Process correct input vector.
* Added tests for roll.
* Refactored roll to conform with TF. Eliminated memory leaks with Roll op tests.
* no thread_local for cpu
Signed-off-by: raver119 <raver119@gmail.com>
* Jar packaging for maven
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Typo fixed
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* minimal viable prototype for SD
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Tests corrected
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Enabled test
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Test copied from nd4j
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Thin wrapper added
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Cleanup
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Adding arithmetic
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Added tests
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Stubs for ops
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Some ops added
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Advanced tests
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Test added
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Ops added
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Tests added
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Boolen logic ops
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Test added
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Shift operations
Signed-off-by: Alexander Stoyakin <alexander.stoyakin@gmail.com>
* Small build fixes
Signed-off-by: Alex Black <blacka101@gmail.com>
* Fix RL4J
Signed-off-by: Alex Black <blacka101@gmail.com>
* Test fixes
Signed-off-by: Alex Black <blacka101@gmail.com>
* Another fix
Signed-off-by: Alex Black <blacka101@gmail.com>
* fix: IOException no longer thrown by read().
Signed-off-by: Robert Altena <Rob@Ra-ai.com>
* refactoring
* last refactorings
Signed-off-by: Robert Altena <Rob@Ra-ai.com>
* Added tests for get_seed/set_seed ops.
* Added missed tests for scatter_sub/mul/div ops.
* Added tests for hardsigmoid and hardsigmoid_bp.
* Added tests for hardtanh and hardtanh_bp ops.
* Added test for histogram op.
* Added tests for identity op.
* Refactored mergemaxindex op. Added tests for log1p,mergemaxindex, mod and mod_bp ops.
* Fixed tests for FloorDiv.
* Added test for rank op.
* Added tests for rationaltanh/rationaltanh_bp ops.
* Added tests for realdiv/realdiv_bp.
* Added tests for rectifiedtanh/_bp ops.
* Added tests for shapes_of op.
* Added tests for shapes_of op.
* Added tests for size op.
* Added tests for softplus/_bp ops.
* Added tests for softsign/_bp ops.
* Added tests for toggle_bits op. Fixed processing of OP_IMPL and so on defititions.
* Added test for truncatediv op.
* Added another test for truncatediv op.
* Added another test for histogram.
* Added tests for unstack_list op.
* Refactored to_int32/uint32/float16/float32/double/int64/uint64 ops and tests.
* Refactored mergemaxindex op helper for cuda platform and tests.
* Fixed cuda kernel for histogram op helper.
* Refactor skipgram to avoid early buffers shift.
* Fixed check up with non_max_suppression op cuda helper. Added cuda kernel implementation for skipgram op helpers.
* Added implementation of skipgram op helper for cuda platform. Working revision
* Fixed mergeMaxIndex kernel and move it to separate source file.
Alex Black
raver119
Robert Altena
Ryan Nett
Alexander Stoyakin
Yurii Shyrma
shugeo