ComputeLibrary.git -

Age	Commit message (Expand)	Author
2022-11-01	Fix fixed-point quantized addition	Viet-Hoa Do
2022-11-01	Updateable weights in depthwise convolution	Milos Puzovic
2022-11-01	Add threshold for floating-point SOFT_RELU activation	Milos Puzovic
2022-11-01	Add check for Batch Matmul in GemmAssemblyDispatch	Mohammed Suhail Munshi
2022-10-27	Fix fixed-point quantized addition	Viet-Hoa Do
2022-10-20	Update reinterpret tensor as 1D for CPU add	Viet-Hoa Do
2022-10-20	Add test in GEMMLowp for batch matmul	Mohammed Suhail Munshi
2022-10-19	Fix FFTConvolutionLayer test	Viet-Hoa Do
2022-10-12	Optimize Neon™ Logistic Activation	Mohammed Suhail Munshi
2022-10-12	Adding documentation section explaining how BF16 is used	Ramy Elgammal
2022-10-10	Fix LUT-based activation layer	Viet-Hoa Do
2022-10-07	Optimize Neon™ SUB operator by squashing execution window	Jakub Sujak
2022-10-03	Fix Batch Matmul nightly failure	Adnan AlSinan
2022-10-03	Optimize CPU add layer on quantized data	Viet-Hoa Do
2022-09-26	Add FP32 Neon™ swish activation	Jonathan Deakin
2022-09-22	Fix unresolved symbol for target armv7a + Android	Pablo Marquez Tello
2022-09-16	Fix bug in QASYMM8_SIGNED to F32 cast layer	Viet-Hoa Do
2022-09-16	Optimize Quantized/Integer Bilinear Scale for Neon™	Gunes Bayir
2022-09-14	Interpreting tensor as 1D for CPU multiplication	Viet-Hoa Do
2022-09-14	Adding GELU activation	Murray Kornelsen
2022-09-14	INT8 Quantized MeanStdDevNorm (LayerNorm)	Murray Kornelsen
2022-09-12	Add test for NEGEMM to test a batched matrix multiplication with variable inp...	Adnan AlSinan
2022-09-09	Optimize FP32/16 Bilinear Scale Kernel for Neon™	Gunes Bayir
2022-09-08	Disable Winograd on fp16 if fast-math = false	Ramy Elgammal
2022-09-02	F16 Specialization for MeanStdDevNorm	Murray Kornelsen
2022-08-24	Fix add for tensors with non-matching strides	Jonathan Deakin
2022-08-18	Use Neon™ kernels for FP Bilinear Resize for SVE	Gunes Bayir
2022-08-17	Add LUT for quantized sigmoid function	Viet-Hoa Do
2022-08-08	Fix for AI benchmark ResNet regression	Viet-Hoa Do
2022-08-04	[ONCPUML-970] Fast math mode for fixed format kernels	Pablo Marquez Tello
2022-08-03	[ONCPUML-968] Fixed format kernel support in additional APIs	Milos Puzovic
2022-08-01	Optimize add layer by considering the input tensors as 1D array	Gunes Bayir
2022-07-27	Fix compilation error rasied in Nightly_NEW	Ramy Elgammal
2022-07-26	Fix for inclusion of "arm_gemm" from src into "Types.h" from core	Ramy Elgammal
2022-07-25	Enable march=armv8.6-a in non multi-isa builds	Pablo Marquez Tello
2022-07-19	[ONCPUML-951] Variable weight support for Convolution.	Francesco Petrogalli
2022-07-14	Integrate new winograd APIs from MLTech	ramelg01
2022-07-04	Fix build errors on armv8.6 SVE2 with NDK 23 and 24	Michalis Spyrou
2022-06-30	Wrong arguments for running activation function in CpuGemmDirectConv2d	Michalis Spyrou
2022-06-29	Add LUT-based leaky relu for QASYMM8 on CPU	Viet-Hoa Do
2022-06-24	Improve LUT Neon Hard-Swish	Pablo Marquez Tello
2022-06-23	Select neon LUT Hard-Swish kernel on all devices	Pablo Marquez Tello
2022-06-16	Fix SVE2 implementation of quantized SoftMax 1D	Viet-Hoa Do
2022-06-09	Fix crash in CpuActivationKernel	Pablo Marquez Tello
2022-06-07	Compute Hard-Swish with a Lookup table for qasymm8.	Pablo Marquez Tello
2022-05-24	[arm_gemm] Import fixed-format kernels from gemm_linux.	Francesco.Petrogalli@arm.com
2022-05-06	Use svcreate instead of list initializations.	Michalis Spyrou
2022-05-04	Add a missing validation check to CPU Pool3d	Adnan AlSinan
2022-04-25	Update Neon™ pooling kernel	ramelg01
2022-04-25	Add LU_BOUNDED_RELU support for QSYMM16	Pablo Marquez Tello