| .. |
|
attnlstm
|
operator code cleanup (#4228)
|
2020-06-13 14:47:44 -07:00 |
|
bert
|
Update Attention op to support attention mask for GPT-2 (#4330)
|
2020-06-29 23:26:23 -07:00 |
|
quantization
|
Handle cases which produce an empty output in the MatMul op family (#4573)
|
2020-07-21 17:57:47 -07:00 |
|
activations.cc
|
Remove FastGelu from activations.
|
2020-04-09 19:17:54 -07:00 |
|
activations.h
|
Parallel all the activations ops (#3722)
|
2020-05-05 01:18:17 -07:00 |
|
cdist.cc
|
Rework CDist (#3393)
|
2020-04-09 14:05:25 +10:00 |
|
cdist.h
|
Rework CDist (#3393)
|
2020-04-09 14:05:25 +10:00 |
|
conv_transpose_with_dynamic_pads.cc
|
Adding custom op ConvTransposeWithDynamicPads. (#638)
|
2019-05-31 11:48:43 -07:00 |
|
conv_transpose_with_dynamic_pads.h
|
Adding custom op ConvTransposeWithDynamicPads. (#638)
|
2019-05-31 11:48:43 -07:00 |
|
cpu_contrib_kernels.cc
|
Use single OpKernel for u8u8 and u8s8 types (#4414)
|
2020-07-02 18:23:58 -07:00 |
|
cpu_contrib_kernels.h
|
move cpu/cuda related files to coresponding cpu/cuda folder (#3668)
|
2020-04-24 00:12:02 -07:00 |
|
crop.cc
|
|
|
|
crop.h
|
fix xcode alerts (#4470)
|
2020-07-16 10:20:34 -07:00 |
|
crop_and_resize.cc
|
operator code cleanup (#4228)
|
2020-06-13 14:47:44 -07:00 |
|
crop_and_resize.h
|
Replace GSL with GSL-LITE submodule and fix up refs (#1920)
|
2019-10-01 12:43:29 -07:00 |
|
dynamicslice.cc
|
Fix dynamicslice.cc after merge.
|
2020-04-09 15:17:21 -07:00 |
|
element_wise_ops.cc
|
move all contrib ops to contrib ops namespace (#1190)
|
2019-06-24 10:19:01 -07:00 |
|
element_wise_ops.h
|
move all contrib ops to contrib ops namespace (#1190)
|
2019-06-24 10:19:01 -07:00 |
|
expand_dims.cc
|
|
|
|
expand_dims.h
|
|
|
|
fused_activation.cc
|
implement conv+clip fusion (#1412)
|
2019-07-17 12:16:45 -07:00 |
|
fused_activation.h
|
cleanup fused conv activation handling (#1403)
|
2019-07-14 16:34:16 -07:00 |
|
fused_conv.cc
|
cleanup fused conv activation handling (#1403)
|
2019-07-14 16:34:16 -07:00 |
|
fused_gemm.cc
|
Parallel all the activations ops (#3722)
|
2020-05-05 01:18:17 -07:00 |
|
image_scaler.cc
|
|
|
|
image_scaler.h
|
operator code cleanup (#4228)
|
2020-06-13 14:47:44 -07:00 |
|
inverse.cc
|
Implement Inverse(12) for CPU and CUDA (#3485)
|
2020-04-18 17:10:21 -07:00 |
|
layer_norm.cc
|
Thread pool changes (#3153)
|
2020-03-30 12:18:40 -07:00 |
|
layer_norm.h
|
Add layernorm operator (#1967)
|
2019-10-03 11:32:13 -07:00 |
|
matmul_integer16.cc
|
Handle cases which produce an empty output in the MatMul op family (#4573)
|
2020-07-21 17:57:47 -07:00 |
|
matmul_integer16.h
|
Add support for int8 x uint8 for MatMulInteger, and int16 x int16 custom op (#1391)
|
2019-08-28 21:40:24 -07:00 |
|
maxpool_with_mask.cc
|
|
|
|
maxpool_with_mask.h
|
MaxpoolWithMask (#3831)
|
2020-05-05 22:19:35 -07:00 |
|
mean_variance_normalization_exp.cc
|
|
|
|
murmur_hash3.cc
|
Support more types in MurmurhashOp (#3827)
|
2020-05-07 20:43:11 -07:00 |
|
murmur_hash3.h
|
Improve code readability and performance. (#2257)
|
2019-10-25 16:19:59 -07:00 |
|
nchwc_ops.cc
|
Nnapi, add auto_pad support for Conv/GlobalAveragePool/AveragePool/GlobalMaxPool/MaxPool operators (#4499)
|
2020-07-15 00:21:42 -07:00 |
|
nchwc_ops.h
|
NCHWc Upsample/Mul optimizations (#3116)
|
2020-03-02 14:40:49 -08:00 |
|
qlinear_binary_op.cc
|
Fix avx2 load 32 bytes buffer overrun. (#4455)
|
2020-07-09 15:54:31 -07:00 |
|
qlinear_binary_op.h
|
qlinaradd for arm/sse2/avx2 using intrinsic, enable binary broadcasting parallel (#4216)
|
2020-07-01 11:54:44 -07:00 |
|
qlinear_lookup_table.cc
|
Implement QLinearLeakyRelu (#3648)
|
2020-05-14 14:52:55 -07:00 |
|
qlinear_lookup_table.h
|
Implement QLinearLeakyRelu (#3648)
|
2020-05-14 14:52:55 -07:00 |
|
sample.cc
|
|
|
|
sample.h
|
|
|
|
skip_layer_norm.cc
|
operator code cleanup (#4228)
|
2020-06-13 14:47:44 -07:00 |
|
skip_layer_norm.h
|
Add Epsilon Attribute to Skip and Embed Layer Normalization (#3768)
|
2020-05-05 16:36:26 -07:00 |
|
tokenizer.cc
|
Introduce container type runtime checks and other improvements (#2522)
|
2019-12-04 16:04:17 -08:00 |
|
transpose_matmul.cc
|
Handle cases which produce an empty output in the MatMul op family (#4573)
|
2020-07-21 17:57:47 -07:00 |
|
transpose_matmul.h
|
address PR comments (#3312)
|
2020-03-25 19:35:12 -07:00 |
|
unique.cc
|
Implement contrib kernels for Pad (changed interface) and Unique (new ONNX op) (#1006)
|
2019-05-13 13:10:18 -07:00 |
|
unique.h
|
Implement contrib kernels for Pad (changed interface) and Unique (new ONNX op) (#1006)
|
2019-05-13 13:10:18 -07:00 |
|
word_conv_embedding.cc
|
operator code cleanup (#4228)
|
2020-06-13 14:47:44 -07:00 |
|
word_conv_embedding.h
|
Let mlas use session thread pool (#1609)
|
2019-08-16 13:21:15 -07:00 |
