Gitiles
Code Review Sign In
review.mlplatform.org / ml / ComputeLibrary / 840a72cc745c60eccbd26fe192b035ec68b2ee41 / src / runtime
  1. 840a72c COMPMID-3773: Fix the accumulation of projection in QLSTM by Sang-Hoon Park · 3 years, 10 months ago
  2. 10df0b6 COMPMID-3778: Investigate CLGEMM failures on S10plus by Michele Di Giorgio · 3 years, 10 months ago
  3. 245ba05 COMPMID-3783: Segfaulting in Reduction layer with NEON on Mate20 with HFNet by Sheri Zhang · 3 years, 10 months ago
  4. 9e698d5 COMPMID-3757: (u)int8: Don't select the 16-bit route on A53 for cases with very few rows by David Mansell · 4 years ago
  5. fa269bb MLCE-229: Fixed requantization per channel in asm kernel by morgolock · 3 years, 11 months ago
  6. 7d0adc6 COMPMID-3151: Remove NEDepthwiseConvolutionLayer3x3Kernel by Georgios Pinitas · 4 years ago
  7. 2d22139 COMPMID-3154: Deprecate NE/CL LocallyConnectedLayer by Georgios Pinitas · 4 years ago
  8. ec4dee8 COMPMID-3750: Disable asm kernels when shifts are negative. by morgolock · 4 years ago
  9. 12e0209 COMPMID-3772: Update GEMM selection heuristic for Mali-G76 (F32) by Gian Marco Iodice · 4 years ago
  10. 026d045 COMPMID-3770: Add batch size in the OpenCL GEMM kernel selection by Gian Marco Iodice · 4 years ago
  11. 96209c7 COMPMID-3694 COMPMID-3695 COMPMID-3458: Softmax Axis by SiCong Li · 4 years ago
  12. 5111264 COMPMID-3661: Added multidimension support to OMP scheduler. by morgolock · 4 years ago
  13. 0bc80da MLCE-229: Support for negative shifts in asm kernels by morgolock · 4 years ago
  14. c58f0ad COMPMID-3502: Add support of different quantization input/output for ReduceMean by Manuel Bottini · 4 years ago
  15. a45abfd COMPMID-3687: Remove deprecated functions in 20.05 release by Sang-Hoon Park · 4 years ago
  16. 9c7fed8 COMPMID-3656: Disabled reduce_axis in LOG_SOFTMAX and SOFTMAX by morgolock · 4 years ago
  17. 187558f Added missing parameter num_groups to the validate call of NEConvolutionLayer. by Alexander Jung · 4 years ago
  18. 4d81d75 COMPMID-2450: Implement CLMaxUnpoolingLayer by Gian Marco Iodice · 4 years, 1 month ago
  19. 0499dff COMPMID-3392: Collapse TensorMaps into a single TensorPack by Georgios Pinitas · 4 years ago
  20. 7a81d2a COMPMID-3624: CTS failure on Resize quantized in Neon and CL by Michele Di Giorgio · 4 years ago
  21. a179798 COMPMID-3585: Android R while_fib_n_5_quant8 failure on CpuAcc by Michele Di Giorgio · 4 years ago
  22. 42e5736 COMPMID-2078: Remove legacy TODOs by Georgios Pinitas · 4 years ago
  23. 1009e87 COMPMID-3385: Async support to CLArithmetic* kernels/functions Pt.2 by Michalis Spyrou · 4 years ago
  24. 3351f2a COMPMID-3575: Mixed preicision in NEInstanceNormalizationLayerKernel by Sang-Hoon Park · 4 years, 1 month ago
  25. ad7515d COMPMID-3385: Async support to CLArithmetic* kernels/functions Pt.1 by Michalis Spyrou · 4 years ago
  26. 09cad72 COMPMID-3393: Minor tweaks on memory injection interface by Georgios Pinitas · 4 years ago
  27. 318c9f4 COMPMID-3578: Update FP32/int8 kernel selection. by David Mansell · 4 years, 1 month ago
  28. f932d2c COMPMID-3386: Support memory injection in CLConcatenate functions/kernels by Michele Di Giorgio · 4 years, 1 month ago
  29. f20d6d6 COMPMID-3390: Async support to CLStridedSliceLayerKernel kernels/functions by Michalis Spyrou · 4 years, 1 month ago
  30. c6eaec3 COMPMID-3326; Update heuristic for GEMMReshaped and GEMMReshapedOnlyRHS by Gian Marco Iodice · 4 years ago
  31. f6f7876 COMPMID-3532: Align data type support between doxygen and implementation - CL by Michele Di Giorgio · 4 years, 1 month ago
  32. 9c82e01 COMPMID-3604: Graph failures during tuning by Georgios Pinitas · 4 years ago
  33. a084b46 [ONCPUML-97]: Implement "int8" support for 2D decomposition at high core counts by Aleksandr Nikolaev · 4 years, 1 month ago
  34. f738fe6 COMPMID-3389: Async support to CLElementwiseUnaryLayerKernel kernels/functions by Michalis Spyrou · 4 years, 1 month ago
  35. 534fdea [ONCPUML-96] FP16 support for 2D decomposition at high core counts. by cfRod · 4 years, 1 month ago
  36. 4667ddd COMPMID-3374: Remove memory state from NEConcatenateLayer kernels by Georgios Pinitas · 4 years, 1 month ago
  37. 06e890b COMPMID-3565: Exposes interface to enable thread binding by Georgios Pinitas · 4 years, 1 month ago
  38. d9eaf61 COMPMID-3324: Adjusting capitalization of Arm copyright claim to reflect Arm preferred presentation by Michele Di Giorgio · 4 years, 1 month ago
  39. f9b595a COMPMID-3532: Align data type support between doxygen and implementation - NEON by Michele Di Giorgio · 4 years, 1 month ago
  40. 0cc50ed COMPMID-3324: Remove pretransposed support from NEON backend by Georgios Pinitas · 4 years, 1 month ago
  41. ab23dd0 COMPMID-3387: Support memory injection in CLActivationLayer by Georgios Pinitas · 4 years, 1 month ago
  42. 6eb7345 COMPMID-3373: Async support to NEArithmetic* kernels/functions (Pt. 2) by Michalis Spyrou · 4 years, 1 month ago
  43. 5aa1a0b COMPID-3324: Clean GEMM kernels by Georgios Pinitas · 4 years, 1 month ago
  44. 94d5051 COMPMID-3539: Change indexing for nearest neighbor with aligned corners by Sang-Hoon Park · 4 years, 1 month ago
  45. 2aad21a COMPMID-3388: Async support to CLReshapeLayerKernel kernels/functions by Michalis Spyrou · 4 years, 1 month ago
  46. d13931d COMPMID-3534: CLGEMMConvolutionLayer doesn't support QASYMM8_SIGNED properly by morgolock · 4 years, 1 month ago
  47. fcf6f4e COMPMID-3477: Remove padding from NEPixelWiseMultiplicationKernel by Sheri Zhang · 4 years, 1 month ago
  48. 6b6a16f COMPMID-3501 Modify heuristics for f16+fastmath NEON Winograd Conv by SiCong Li · 4 years, 2 months ago
  49. f2c714b COMPMID-3153: Remove padding from NENormalizationLayerKernel by Manuel Bottini · 4 years, 2 months ago
  50. 3687ee1 COMPMID-3539: Ignore align_corners for scaled size of 1 by Sang-Hoon Park · 4 years, 1 month ago
  51. c5b6d88 COMPMID-3562: Support QASYMM8_SIGNED in CLArgMinMaxLayerKernel by Sheri Zhang · 4 years, 1 month ago
  52. 173ba9b COMPMID-3373: Async support to NEArithmetic* kernels/functions (Pt. 1) by Michalis Spyrou · 4 years, 1 month ago
  53. 0f954eb COMPMID-3324: Remove pretranspose_B support from NEGEMM by Georgios Pinitas · 4 years, 1 month ago
  54. 1732da8 COMPMID-3384: Async support to NEStridedSliceLayerKernel kernels/functions by Michalis Spyrou · 4 years, 1 month ago
  55. ce0c675 COMPMID-3377: Async support to NEElementwiseUnaryLayerKernel kernels/functions by Michalis Spyrou · 4 years, 1 month ago
  56. 1902383 Revert "COMPMID-3480: Perform in-place computations in NEArithmeticAdditionKernel" by Michele Di Giorgio · 4 years, 1 month ago
  57. 6ad60af COMPMID-3520: Move ndrange.hpp header from arm_gemm to assembly by Michele Di Giorgio · 4 years, 2 months ago
  58. 1fd2c80 COMPMID-3375: Port NEActivationLayer functions/kernels to run on by Georgios Pinitas · 4 years, 2 months ago
  59. bcd2352 COMPMID-3391: Implement Async interfaces by Michalis Spyrou · 4 years, 2 months ago
  60. 37722d9 COMPMID-2449: Implement NEUnPoolLayer by morgolock · 4 years, 4 months ago
  61. 4a61653 COMPMID-3480: Perform in-place computations in NEArithmeticAdditionKernel by Michele Di Giorgio · 4 years, 2 months ago
  62. ccd9496 COMPMID-3364: use ScaleKernelInfo for Scale on OpenCL and GLES by Sang-Hoon Park · 4 years, 2 months ago
  63. 11c562c COMPMID-3527: QLSTM should allow nullptr for optional projection bias by Michele Di Giorgio · 4 years, 2 months ago
  64. d004a7a COMPMID-3510 [Interface change] Fix definition of "axis" in NESoftmaxLayer and CLSoftmaxLayer by SiCong Li · 4 years, 2 months ago
  65. 25ef721 COMPMID-3180: Remove padding from NEThreshold by Georgios Pinitas · 4 years, 2 months ago
  66. f025a77 COMPMID-3363, COMPMID-3364: Add align_corners support to nearest neighbor by Sang-Hoon Park · 4 years, 2 months ago
  67. c261798 COMPMID-3363: Create ScaleKernelInfo by Sang-Hoon Park · 4 years, 2 months ago
  68. 7889291 COMPMID-3069: Add return statement in ST scheduler by Michalis Spyrou · 4 years, 2 months ago
  69. c8bdeaa COMPMID-3069: Fix reduction in GEMMLowp by Michalis Spyrou · 4 years, 2 months ago
  70. 387259a COMPMID-3069: Removing deprecated functions and classes from 20.05 release by Manuel Bottini · 4 years, 2 months ago
  71. e6bd70c COMPMID-3069: Fix CLQLSTM configure argument order by Michalis Spyrou · 4 years, 2 months ago
  72. 11bfe0e COMPMID-3491: ReduceMean not properly validating quantizations by Manuel Bottini · 4 years, 2 months ago
  73. 5f39091 COMPMID-3176: Remove padding from NEArithmeticSubtractionKernel by Michalis Spyrou · 4 years, 3 months ago
  74. ee4833d COMPMID-3497: Add checks for zero scale values to QLSTMLayer by Sang-Hoon Park · 4 years, 2 months ago
  75. 0905028 COMPMID-3069: Align graph convolution implementation for CL and NEON. by Georgios Pinitas · 4 years, 3 months ago
  76. 4efe5dc COMPMID-3484: Regression in Transpose convolution Android R CTS test. by Manuel Bottini · 4 years, 3 months ago
  77. ea2ce17 Introduce CL kernel cache feature by Inki Dae · 4 years, 4 months ago
  78. 5ce897f COMPMID-3108: Add Winograd 3x3,4x4 FP16 support for NEON by Georgios Pinitas · 4 years, 3 months ago
  79. a7431ae COMPMID-3439: Fix peephole and projection in CLQLSTMLayer by Sang-Hoon Park · 4 years, 3 months ago
  80. 1f567af COMPMID-3442: Add support of negative axis in NESoftmaxLayer and reference code by Sheri Zhang · 4 years, 3 months ago
  81. beb2d45 COMPMID-3470: Modify NE/CLQLSTMLayer interface to provide 3 outputs by Michele Di Giorgio · 4 years, 3 months ago
  82. 6e10aa3 COMPMID-3316: NEDeconvolutionLayer failing for a big input by Manuel Bottini · 4 years, 3 months ago
  83. d5c020a COMPMID-3239: Fix projection and peephole in NEQLSTMLayer by Sang-Hoon Park · 4 years, 3 months ago
  84. 6f8b17d [ONCPUML-7] arm_compute support for ND parallelism by Joseph Dobson · 4 years, 6 months ago
  85. c6aef87 COMPMID-3069: Validate bias dimensionality in NEFullyConnectedLayer by Georgios Pinitas · 4 years, 3 months ago
  86. bc0e1c9 COMPMID-3069: Allow different quantization info in NEElementwiseMin/Max by Michele Di Giorgio · 4 years, 3 months ago
  87. 3a35398 COMPMID-3240: Add support for layer normalization to CLQLSTMLayer by Sheri Zhang · 4 years, 3 months ago
  88. 2b84be5 COMPMID-3280: Make all ML primitives for CL use the new interface - Part 2 by Manuel Bottini · 4 years, 4 months ago
  89. 9230e27 COMPMID-3241: Add Layer Normalization to NEQLSTMLayer by Sang-Hoon Park · 4 years, 3 months ago
  90. 1c1b3aa COMPMID-3236: Implement CLQLSTMLayer by Michele Di Giorgio · 4 years, 4 months ago
  91. e631681 COMPMID: Fixed issue in the scheduler of NEConcatenateLayer by Gian Marco Iodice · 4 years, 3 months ago
  92. 30b46a6 COMPMID-3241: Fix hidden scale in NEQLSTMLayer by Sang-Hoon Park · 4 years, 3 months ago
  93. eb65f6d COMPMID-3304: Update OpenCL GEMM heuristic for Int8 by Gian Marco Iodice · 4 years, 4 months ago
  94. 47a8990 COMPMID-3237: Implement NEQLSTMLayer by Michele Di Giorgio · 4 years, 5 months ago
  95. 4c6bd51 COMPMID-3280: Make all ML primitives for CL use the new interface - Part 1 by Manuel Bottini · 4 years, 4 months ago
  96. f6ffe99 COMPMID-3349: Fixed performance regression Mali-G71 by Gian Marco Iodice · 4 years, 4 months ago
  97. 5a4fe19 COMPMID-3112: Reworking heuristic for CLGEMM - part1 by Gian Marco Iodice · 4 years, 5 months ago
  98. 11d4918 COMPMID-3279: Create CLCompiler interface by Michalis Spyrou · 4 years, 4 months ago
  99. f64d336 COMPMID-3236: Extend CLGEMMLowpReduction kernels to multiply by a scalar value by Michele Di Giorgio · 4 years, 4 months ago
  100. fedefc3 COMPMID-2765 Add support for QASYMM8_SIGNED in NEDeconvolutionLayer by Luca Foschiani · 4 years, 6 months ago