- Aug 19, 2024
-
-
Signed-off-by:
Jakub Sujak <jakub.sujak@arm.com> Approved-by:
Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>
-
- Aug 16, 2024
-
-
* The LHS matrix is Quantized (Q) Symmetric (S) Signed 8-bit (I8) with per-block quantization (D32) quantization parameters * The RHS matrix is Quantized (Q) Symmetric (S) Signed 4-bit (I4) with per-block quantization(C32) F16 scale factors, * The destination is F32 * Implement micro-kernels to perform the matrix multiplication * Implement a micro-kernel to pack the LHS and RHS matrices * Added unit tests Signed-off-by:
Gian Marco <Iodice gianmarco.iodice@arm.com> Signed-off-by:
Anitha <Raj Anitha.Raj@arm.com> Signed-off-by:
Viet-Hoa Do <viet-hoa.do@arm.com> Reviewed-by:
Viet-Hoa Do <viet-hoa.do@arm.com> Reviewed-by:
Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com> Reviewed-by:
Jakub Sujak <jakub.sujak@arm.com> Approved-by:
Jakub Sujak <jakub.sujak@arm.com>
-
- Aug 15, 2024
-
-
Viet-Hoa Do authored
* Tested kernels: - lhs_quant_pack_qai8dxp_f32 - rhs_pack_nxk_qsi4cxp_qsu4cxs1s0 - matmul_clamp_f32_qai8dxp4x8_qsi4cxp8x8_8x8x32_neon_i8mm * Only entire matrix is tested - support for testing arbitrary portion of the output will come later. * Update test framework to accommodate the new tests. - The new test no longer relies on DataFormat concept. Now the packing function for each format is explicitly called. Signed-off-by:
Viet-Hoa Do <viet-hoa.do@arm.com> Reviewed-by:
Viet-Hoa Do <viet-hoa.do@arm.com> Approved-by:
Jakub Sujak <jakub.sujak@arm.com>
-
Viet-Hoa Do authored
Signed-off-by:
Viet-Hoa Do <viet-hoa.do@arm.com> Approved-by:
Anton Bondarenko <anton.bondarenko@arm.com>
-
Anton Bondarenko authored
Signed-off-by:
Anton Bondarenko <anton.bondarenko@arm.com> Reviewed-by:
Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com> Approved-by:
Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>
-
- Aug 14, 2024
-
-
Viet-Hoa Do authored
This reverts commit 94aa71e6. Signed-off-by:
Viet-Hoa Do <viet-hoa.do@arm.com> Approved-by:
Jakub Sujak <jakub.sujak@arm.com>
-
Viet-Hoa Do authored
Signed-off-by:
Viet-Hoa Do <viet-hoa.do@arm.com> Approved-by:
Anton Bondarenko <anton.bondarenko@arm.com> Approved-by:
Jakub Sujak <jakub.sujak@arm.com>
-
- Aug 13, 2024
-
-
Felix Johnny Thomasmathibalan authored
Block size: 6x8 Signed-off-by:
Felix Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com> Reviewed-by:
Viet-Hoa Do <viet-hoa.do@arm.com> Approved-by:
Jakub Sujak <jakub.sujak@arm.com>
-
Felix Johnny Thomasmathibalan authored
Update so CMake and Bazel have the same '-march' compiler option for files. Signed-off-by:
Felix Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com> Approved-by:
Jakub Sujak <jakub.sujak@arm.com>
-
- Aug 06, 2024
-
-
Jakub Sujak authored
* Compute the general matrix-vector (GEMV) multiplication between an FP16 LHS and RHS and accumulate into FP16 output. The RHS packs FP16 weights and biases together. * Optimized for Arm® Neon™ using MLA instructions. * Add accompanying tests. Signed-off-by:
Jakub Sujak <jakub.sujak@arm.com> Approved-by:
Viet-Hoa Do <viet-hoa.do@arm.com>
-
- Aug 01, 2024
-
-
Felix Johnny Thomasmathibalan authored
Signed-off-by:
Felix Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com> Approved-by:
Viet-Hoa Do <viet-hoa.do@arm.com>
-
Signed-off-by:
Michael Kozlov <michael.kozlov@arm.com> Approved-by:
Viet-Hoa Do <viet-hoa.do@arm.com>
-
- Jul 30, 2024
-
-
Viet-Hoa Do authored
* Make it clearer and remove redundant unnecessary text. Signed-off-by:
Viet-Hoa Do <viet-hoa.do@arm.com> Reviewed-by:
Viet-Hoa Do <viet-hoa.do@arm.com> Reviewed-by:
Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com> Approved-by:
Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>
-
- Jul 29, 2024
-
-
Jakub Sujak authored
Signed-off-by:
Jakub Sujak <jakub.sujak@arm.com> Approved-by:
Viet-Hoa Do <viet-hoa.do@arm.com>
-
Viet-Hoa Do authored
* It informs the user of potential security weaknesses that their product must mitigate. Signed-off-by:
Viet-Hoa Do <viet-hoa.do@arm.com> Reviewed-by:
Viet-Hoa Do <viet-hoa.do@arm.com> Reviewed-by:
Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com> Approved-by:
Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>
-
- Jul 24, 2024
-
-
Anton Bondarenko authored
* Use baremetal compiler to compile FVP image components * Add NDK for x86_64 target Signed-off-by:
Anton Bondarenko <anton.bondarenko@arm.com> Approved-by:
Jens Elofsson <jens.elofsson@arm.com>
-
- Jul 22, 2024
-
-
Viet-Hoa Do authored
Signed-off-by:
Viet-Hoa Do <viet-hoa.do@arm.com> Approved-by:
Jakub Sujak <jakub.sujak@arm.com>
-
- Jul 18, 2024
-
-
Anton Bondarenko authored
Curl is an alternative for wget and used by some external tools. So it's good to have both for flexibility. Signed-off-by:
Anton Bondarenko <anton.bondarenko@arm.com> Reviewed-by:
Viet-Hoa Do <viet-hoa.do@arm.com> Approved-by:
Viet-Hoa Do <viet-hoa.do@arm.com>
-
- Jul 16, 2024
-
-
Anton Bondarenko authored
New hook uses Bazel tools v6.4.0 and supports common development environments. Signed-off-by:
Anton Bondarenko <anton.bondarenko@arm.com> Approved-by:
Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>
-
- Jul 15, 2024
-
-
Anton Bondarenko authored
Allows to pass external arguments by using kwargs to cc_library rule. Signed-off-by:
Anton Bondarenko <anton.bondarenko@arm.com> Approved-by:
Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>
-
Felix Johnny Thomasmathibalan authored
Signed-off-by:
Felix Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com> Approved-by:
Jens Elofsson <jens.elofsson@arm.com>
-
Felix Johnny Thomasmathibalan authored
Signed-off-by:
Felix Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com> Approved-by:
Anton Bondarenko <anton.bondarenko@arm.com>
-
Anton Bondarenko authored
Bazel is one of the build system used in the modern software world and especually in ML OSS projects. Support for Bazel build would allow simpler intergration into such frameworks. Currently supported: * Build microkernels with required CPU micro-architecture * Combine into common library (matmul) * Tests compilation and execution Signed-off-by:
Anton Bondarenko <anton.bondarenko@arm.com> Reviewed-by:
Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com> Approved-by:
Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>
-
- Jul 12, 2024
-
-
Viet-Hoa Do authored
* Fix condition for left-over rows store. Resolves: COMPMID-7245 Signed-off-by:
Viet-Hoa Do <viet-hoa.do@arm.com> Reviewed-by:
Viet-Hoa Do <viet-hoa.do@arm.com> Approved-by:
Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>
-
- Jul 08, 2024
-
-
Jakub Sujak authored
* Add instructions for building for various platforms using CMake. * Provide a CMake toolchain file for Arm GNU Toolchain. Signed-off-by:
Jakub Sujak <jakub.sujak@arm.com> Approved-by:
Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com> Approved-by:
Jens Elofsson <jens.elofsson@arm.com>
-
- Jul 05, 2024
-
-
Jakub Sujak authored
Signed-off-by:
Jakub Sujak <jakub.sujak@arm.com> Approved-by:
Anton Bondarenko <anton.bondarenko@arm.com>
-
Runs on Linux® and Android™. Tests the different variations of the matmul_clamp_f32_qai8dxp_qsi4cxp microkernel with user-specified m-, n-, and k-values. Signed-off-by:
Jens Elofsson <jens.elofsson@arm.com> Approved-by:
Jakub Sujak <jakub.sujak@arm.com>
-
- Jul 04, 2024
-
-
Signed-off-by:
Felix Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com> Approved-by:
Jakub Sujak <jakub.sujak@arm.com>
-
Felix Johnny Thomasmathibalan authored
Signed-off-by:
Felix Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com> Approved-by:
Jakub Sujak <jakub.sujak@arm.com>
-
Jakub Sujak authored
Signed-off-by:
Jakub Sujak <jakub.sujak@arm.com> Approved-by:
Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>
-
Signed-off-by:
Jakub Sujak <jakub.sujak@arm.com> Approved-by:
Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>
-
Jakub Sujak authored
Signed-off-by:
Jakub Sujak <jakub.sujak@arm.com> Reviewed-by:
Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com> Approved-by:
Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>
-
Jakub Sujak authored
This example demonstrates the usage of the FP16 packing and matmul routines which: 1. Packs the bias and the weights together into a single tensor. 2. Performs a matrix multiplication of the activations and the packed tensor. All tensors are in half precision floating-point (FP16) data type. Signed-off-by:
Jakub Sujak <jakub.sujak@arm.com> Reviewed-by:
Jakub Sujak <jakub.sujak@arm.com> Approved-by:
Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>
-
- Jul 03, 2024
-
-
Anton Bondarenko authored
Signed-off-by:
Anton Bondarenko <anton.bondarenko@arm.com> Approved-by:
Jakub Sujak <jakub.sujak@arm.com>
-
- Jul 02, 2024
-
-
- Adjust the int4 matmul micro-kernels (qsi4cx) to skip the bias Signed-off-by:
Gian Marco Iodice <gianmarco.iodice@arm.com> Signed-off-by:
Felix Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com> Signed-off-by:
Jakub Sujak <jakub.sujak@arm.com> Approved-by:
Jakub Sujak <jakub.sujak@arm.com>
-
Signed-off-by:
Jens Elofsson <jens.elofsson@arm.com> Approved-by:
Jakub Sujak <jakub.sujak@arm.com> Approved-by:
Anton Bondarenko <anton.bondarenko@arm.com>
-
- Jul 01, 2024
-
-
Jens Elofsson authored
Add file and line information to KAI_ERROR-macro. Signed-off-by:
Jens Elofsson <jens.elofsson@arm.com> Approved-by:
Jakub Sujak <jakub.sujak@arm.com>
-
Anton Bondarenko authored
* Prepare and upload JUnit report for tests execution * Simplify pre-commit job * Replace dependencies with needs to unblock jobs Signed-off-by:
Anton Bondarenko <anton.bondarenko@arm.com> Reviewed-by:
Viet-Hoa Do <viet-hoa.do@arm.com> Approved-by:
Jakub Sujak <jakub.sujak@arm.com>
-
Anton Bondarenko authored
New hook would be able to automatically download specified version of the Buildifier tool instead of relying on local one. Signed-off-by:
Anton Bondarenko <anton.bondarenko@arm.com> Approved-by:
Jakub Sujak <jakub.sujak@arm.com>
-
- Jun 28, 2024
-
-
Jakub Sujak authored
* Add compiler version checks in CMake before adding feature specific (e.g. FEAT_DOTPROD) micro-kernels. * Enable compiling scalar routines using MSVC. Signed-off-by:
Felix Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com> Signed-off-by:
Jakub Sujak <jakub.sujak@arm.com> Approved-by:
Jens Elofsson <jens.elofsson@arm.com>
-