ARM-software/ComputeLibrary
The Compute Library is a set of computer vision and machine learning functions optimised for both Arm CPUs and GPUs using SIMD technologies.
C++MIT
Issues
- 4
Arm NN compute library build failure on Raspberry PI 4
#1110 opened by pauljainta - 2
Unable to execute MobilenetV2 on Mali-G710 GPU.
#1108 opened by somasundaram1702 - 3
NEPooling3dLayer performance issue
#1107 opened by alvoron - 4
Depthwise convolution fp16 performance drop
#1109 opened by alvoron - 0
- 3
CPUInfo::get().has_fp16() returns true on RaspberryPi 4
#1096 opened by alvoron - 3
- 1
problem with graph_vgg16.cpp
#1105 opened by RavikumarLav - 8
linear+gelu fused operator is not supported in ACL
#1083 opened by snadampal - 0
File not found conv1_weights.npy in Example graph_resnet50.cpp in computeLibrary
#1104 opened by RavikumarLav - 0
- 5
Problem with graph_alexnet.cpp
#1094 opened by sdavejones - 0
Question about quantization gemm example
#1102 opened by allnes - 3
Uninitialized Regs used causes unintended behavior
#1093 opened by Nitesh8998 - 0
why L1_cache_size and L2_cache_size are constant value
#1101 opened by dervon - 3
Compiling error on Raspberry Pi 5:
#1097 opened by wachhu - 3
Do you have group deconvolution layer?
#1091 opened by allnes - 9
- 5
sparse gemm kernels are not supported in ACL
#1084 opened by snadampal - 1
Unit test failure CPU/UNIT/Context/CpuCapabilities on WoA
#1098 opened by morgolock - 1
Segfault when running DWC tests on WoA
#1099 opened by morgolock - 1
Fix ACL WoA native build compiler errors
#1100 opened by morgolock - 2
Latency for Conv2d and Depthwise
#1074 opened by wenhyan - 12
ACL implementation for fp32 to bf16 conversion and weights pre-packing (reordering) is missing
#1060 opened by snadampal - 5
Problem with deconvolution layer validation
#1061 opened by allnes - 3
[OpenVINO 2023.2.0] ComputeLibrary/libarm_compute-static.a] sh: Argument list too long
#1089 opened by saininav - 1
Performance in NCHW layout operations
#1088 opened by allnes - 4
- 2
ACL operators need to be made stateless to avoid runtime initialization overhead
#1085 opened by snadampal - 2
why compile_commands.json not generated?
#1086 opened by Shipley1105 - 2
ACL implementation for Filtering functions like FIR/IIR
#1072 opened by musicwei - 0
Question about thread-safety
#1073 opened by allnes - 2
ohos
#1087 opened by gejianhui1 - 3
The size limitation for CLFFT1D
#1081 opened by ZhangErliang - 2
Convolution performance issues: NCHW slower than NHWC
#1079 opened by alvoron - 3
Crash when using softmax
#1078 opened by abcdrm - 1
Benchmark quantify test
#1077 opened by wenhyan - 1
Errors while build configuration using cmake.
#1075 opened by sachindia86 - 2
Pure image processing operations
#1068 opened by ismailkocdemir - 1
Add support for self-attention layer for GPU
#1069 opened by atrah22 - 2
How to use fixed format kernel?
#1070 opened by Waylon-Zhu - 2
How to add a new operator to ACL
#1071 opened by zxros10 - 10
Benchmark for cpu one core and one thread
#1062 opened by wenhyan - 1
ACL performance gap between baremetal and linux.
#1065 opened by goldhandsome - 2
Documentation isn't work for 23.08 release
#1067 opened by allnes - 5
- 8
Help Wanted: Accelerating Scikit-Learn Algorithms with Arm Compute Library (ACL) and OpenBLAS
#1063 opened by shashank-fujitsu - 2
- 1
How to add a new backend?
#1059 opened by DandanXu1007 - 1