ComputeLibrary.git -

Age	Commit message (Expand)	Author
2023-04-14	Fix dynamic weights for CPU connected layer	Viet-Hoa Do
2023-04-13	Implement MatMul Function and Operator with Floating Point support for CPU	Mohammed Suhail Munshi
2023-04-04	Support dynamic weights for Fully Connected layers on GPU	Jakub Sujak
2023-04-03	Implement MatMul Function	Ramy Elgammal
2023-04-03	Add Cropping to CLBatchToSpace	Omar Al Khatib
2023-03-30	Add cropping support to NEBatchToSpace	SiCong Li
2023-03-21	Add dynamic weights for CPU fully connected layer	Viet-Hoa Do
2023-03-14	Add CropInfo to BatchToSpace reference and fixture	SiCong Li
2023-03-06	Fix LWS search space used by CLTuner	SiCong Li
2023-03-03	Add weights_info as optional input for NEDeconvolutionLayer	Annop Wongwathanarat
2023-02-14	Extend skip upsampling for deconvolution for non-1x1 kernels	Annop Wongwathanarat
2023-02-09	Fix performance regression in Transposed Convolution	Gunes Bayir
2023-02-01	Add new operator AddMulAdd for Neon™ backend for Float/Quantized types	Gunes Bayir
2023-02-01	Remove fixed format strides hack	Jonathan Deakin
2023-01-30	Skip upsampling for deconvolution when not needed	Annop Wongwathanarat
2023-01-23	Fix performance regression when stride equal to kernel size	Gunes Bayir
2023-01-20	Add enable_fast_math for NEDeconvolutionLayer	Annop Wongwathanarat
2023-01-18	Revert "Update the heuristic for CLDepthwiseConvolutionNative kernel"	Gian Marco Iodice
2023-01-12	Update the heuristic for CLDepthwiseConvolutionNative kernel	Gian Marco Iodice
2022-12-30	Move DWC native heuristic into the heuristic folder	Gian Marco Iodice
2022-12-29	Update the ClConv2d heuristic	Gian Marco Iodice
2022-12-21	Optimize MeanReduce by integer acc. and removing upfront dequant.	Omar Al Khatib
2022-12-14	Optimize Transposed Convolution for CL backend (Quantized)	Gunes Bayir
2022-12-09	Use heuristics for setting dynamic fusion direct conv2d tile sizes	Ramy Elgammal
2022-12-09	Implement the OpenCL kernel to compute the indirect convolution	Gian Marco Iodice
2022-11-22	Remove dynamic fusion prototype with tests and examples	SiCong Li
2022-11-18	Add num_threads_to_use to OMPScheduler based on workload size	cfRod
2022-11-14	Optimize Transposed Convolution for CL backend (FP32/16)	Gunes Bayir
2022-10-06	Rework DepthwiseConvolution heuristic on OpenCL	Gian Marco Iodice
2022-09-16	Optimize Quantized/Integer Bilinear Scale for Neon™	Gunes Bayir
2022-09-09	Optimize FP32/16 Bilinear Scale Kernel for Neon™	Gunes Bayir
2022-09-07	Optimize depthwise convolution on OpenCL	Gian Marco Iodice
2022-09-01	Use parent buffer in CLSubTensor. This avoids calling enqueueMapBuffer repeat...	Murray Kornelsen
2022-08-24	Fix validation problem in CLQLSTMLayer	Pablo Marquez Tello
2022-08-11	Fix CTS/SLTS failure related to Depthwise Convolution	Gunes Bayir
2022-08-03	[ONCPUML-968] Fixed format kernel support in additional APIs	Milos Puzovic
2022-08-01	Fix for OpenMP scheduler work breakdown	Milos Puzovic
2022-07-26	Fix for inclusion of "arm_gemm" from src into "Types.h" from core	Ramy Elgammal
2022-07-19	[ONCPUML-951] Variable weight support for Convolution.	Francesco Petrogalli
2022-07-14	Integrate new winograd APIs from MLTech	ramelg01
2022-07-13	Add Gemm MMUL Reshaped Only Rhs Support for FP32/FP16	Gunes Bayir
2022-07-07	Add missing flag when building cl graph examples and fix	Michalis Spyrou
2022-05-31	Fix compilation error in 32-bit with logging build	Viet-Hoa Do
2022-05-06	Integrate Dynamic Fusion patches	SiCong Li
2022-05-06	QLSTM add support for different qinfo	Mike Kelly
2022-04-21	NEQLSTM: perform type conversion in prepare method.	Pablo Marquez Tello
2022-04-14	Enable dynamic cl tuning for dynamically fused kernels	SiCong Li
2022-04-13	Add DirectConvolution2D kernel component for dynamic fusion	Gunes Bayir
2022-04-01	Add CPU Pool3d FP16/32 implementation	Adnan AlSinan
2022-03-29	NEQLSTM: Add support for QASYMM8_SIGNED for input_to_forget_weights	Pablo Marquez Tello