| .. |
|
integer_ops
|
Update source files with used includes.
|
2020-06-15 18:42:14 -07:00 |
|
sparse_ops
|
Add multi-threaded sparse FullyConnected kernel.
|
2020-06-10 11:05:53 -07:00 |
|
batch_matmul.h
|
Full int8 quantization BatchMatMul
|
2020-06-19 07:33:08 -07:00 |
|
cpu_check.cc
|
Drop the dependency on ruy:detect_arm, which is becoming private.
|
2020-05-13 21:13:35 -07:00 |
|
cpu_check.h
|
Drop the dependency on ruy:detect_arm, which is becoming private.
|
2020-05-13 21:13:35 -07:00 |
|
depthwiseconv_3x3_filter_common.h
|
Reference ruy from its new location as a separate GitHub project.
|
2020-04-03 11:17:08 -07:00 |
|
depthwiseconv_float.h
|
Reference ruy from its new location as a separate GitHub project.
|
2020-04-03 11:17:08 -07:00 |
|
depthwiseconv_multithread.h
|
Restore functionality of --define=tflite_with_ruy=true
|
2020-06-22 10:49:06 -07:00 |
|
depthwiseconv_uint8_3x3_filter.h
|
Reference ruy from its new location as a separate GitHub project.
|
2020-04-03 11:17:08 -07:00 |
|
depthwiseconv_uint8_transitional.h
|
Depthwise convolution 3x3 per-channel int8 for dot-product ARM (13).
|
2020-02-18 10:55:25 -08:00 |
|
depthwiseconv_uint8.h
|
Reference ruy from its new location as a separate GitHub project.
|
2020-04-03 11:17:08 -07:00 |
|
eigen_spatial_convolutions.h
|
|
|
|
eigen_tensor_reduced_instantiations_google.h
|
Update Eigen to 963ba1015bd448.
|
2019-12-11 11:40:07 -08:00 |
|
eigen_tensor_reduced_instantiations_oss.h
|
Update Eigen to 963ba1015bd448.
|
2019-12-11 11:40:07 -08:00 |
|
im2col_utils.h
|
Update source files with used includes.
|
2020-06-15 18:42:14 -07:00 |
|
legacy_optimized_ops.h
|
PR #34903: TransposeConv with Bias
|
2020-04-22 12:33:46 -07:00 |
|
multithreaded_conv.h
|
Port TFLite from the gemmlowp profiler to the ruy profiler.
|
2020-01-14 12:19:32 -08:00 |
|
neon_check.h
|
Re-enable warnings while including arm_neon_sse.h.
|
2020-01-31 15:20:22 -08:00 |
|
neon_tensor_utils_impl.h
|
Add optimized MatrixBatchVectorMultiplyAccumulate for asymmetric inputs for sse
|
2020-05-18 01:35:29 -07:00 |
|
neon_tensor_utils.cc
|
Add optimized MatrixBatchVectorMultiplyAccumulate for asymmetric inputs for sse
|
2020-05-18 01:35:29 -07:00 |
|
neon_tensor_utils.h
|
Add optimized MatrixBatchVectorMultiplyAccumulate for asymmetric inputs for sse
|
2020-05-18 01:35:29 -07:00 |
|
optimized_ops.h
|
optimize softmax quantized with two uint8 lookup table methods.
|
2020-06-08 00:50:47 -07:00 |
|
sse_check.h
|
Fix compilation error on x86 without SSSE3: use of undeclared identifier 'Portablefuncname'
|
2020-02-05 09:51:18 -08:00 |
|
sse_tensor_utils_impl.h
|
Add optimized MatrixBatchVectorMultiplyAccumulate for asymmetric inputs for sse
|
2020-05-18 01:35:29 -07:00 |
|
sse_tensor_utils.cc
|
Add optimized MatrixBatchVectorMultiplyAccumulate for asymmetric inputs for sse
|
2020-05-18 01:35:29 -07:00 |
|
sse_tensor_utils.h
|
Add optimized MatrixBatchVectorMultiplyAccumulate for asymmetric inputs for sse
|
2020-05-18 01:35:29 -07:00 |
