Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Replace LRP gradient computation with VJP using Zygote.pullback #72

Merged
merged 2 commits into from
Jun 27, 2022
Merged

Replace LRP gradient computation with VJP using Zygote.pullback #72

merged 2 commits into from
Jun 27, 2022

Conversation

adrhill
Copy link
Member

@adrhill adrhill commented Jun 27, 2022 •
edited
Loading

Closes #68.

@adrhill adrhill added the run benchmark Run benchmark CI action label Jun 27, 2022
@github-actions
Copy link

Benchmark result

Judge result

Benchmark Report for /home/runner/work/ExplainableAI.jl/ExplainableAI.jl

Job Properties

  • Time of benchmarks:
    • Target: 27 Jun 2022 - 11:14
    • Baseline: 27 Jun 2022 - 11:18
  • Package commits:
    • Target: cf5369
    • Baseline: 0d21c1
  • Julia commits:
    • Target: 742b9a
    • Baseline: 742b9a
  • Julia command flags:
    • Target: None
    • Baseline: None
  • Environment variables:
    • Target: None
    • Baseline: None

Results

A ratio greater than 1.0 denotes a possible regression (marked with ❌), while a ratio less
than 1.0 denotes a possible improvement (marked with ✅). Only significant results - results
that indicate possible regressions or improvements - are shown below (thus, an empty table means that all
benchmark results remained invariant between builds).

ID time ratio memory ratio
["Layer", "Conv", "EpsilonRule"] 0.97 (5%) 0.95 (1%) ✅
["Layer", "Conv", "GammaRule"] 0.96 (5%) 0.95 (1%) ✅
["Layer", "Conv", "ZeroRule"] 0.96 (5%) 0.95 (1%) ✅
["Layer", "Dense", "EpsilonRule"] 1.07 (5%) ❌ 1.00 (1%)
["Layer", "Dense", "ZeroRule"] 1.16 (5%) ❌ 1.00 (1%)
["Layer", "MaxPool", "EpsilonRule"] 0.73 (5%) ✅ 0.76 (1%) ✅
["Layer", "MaxPool", "GammaRule"] 0.73 (5%) ✅ 0.76 (1%) ✅
["Layer", "MaxPool", "ZBoxRule"] 0.82 (5%) ✅ 0.76 (1%) ✅
["Layer", "MaxPool", "ZeroRule"] 0.79 (5%) ✅ 0.76 (1%) ✅
["Layer", "WrappedDense", "EpsilonRule"] 0.88 (5%) ✅ 0.97 (1%) ✅
["Layer", "WrappedDense", "ZeroRule"] 0.95 (5%) ✅ 0.97 (1%) ✅
["VGG", "Gradient", "construct analyzer"] 0.88 (5%) ✅ 1.00 (1%)
["VGG", "InputTimesGradient", "analyze"] 0.06 (5%) ✅ 0.63 (1%) ✅
["VGG", "InputTimesGradient", "construct analyzer"] 1.59 (5%) ❌ 1.00 (1%)
["VGG", "LRPCustom", "analyze"] 0.36 (5%) ✅ 0.88 (1%) ✅
["VGG", "LRPCustom", "construct analyzer"] 1.17 (5%) ❌ 1.00 (1%)
["VGG", "LRPZero", "analyze"] 0.24 (5%) ✅ 0.70 (1%) ✅
["VGG", "LRPZero", "construct analyzer"] 0.86 (5%) ✅ 1.00 (1%)
["VGG", "SmoothGrad", "analyze"] 0.87 (5%) ✅ 0.97 (1%) ✅
["VGG", "SmoothGrad", "construct analyzer"] 0.62 (5%) ✅ 1.00 (1%)

Benchmark Group List

Here's a list of all the benchmark groups executed by this job:

  • ["Layer", "Conv"]
  • ["Layer", "Dense"]
  • ["Layer", "MaxPool"]
  • ["Layer", "WrappedDense"]
  • ["VGG", "Gradient"]
  • ["VGG", "InputTimesGradient"]
  • ["VGG", "IntegratedGradients"]
  • ["VGG", "LRPCustom"]
  • ["VGG", "LRPZero"]
  • ["VGG", "SmoothGrad"]

Julia versioninfo

Target

Julia Version 1.7.3
Commit 742b9abb4d (2022-05-06 12:58 UTC)
Platform Info:
  OS: Linux (x86_64-pc-linux-gnu)
      Ubuntu 20.04.4 LTS
  uname: Linux 5.13.0-1031-azure #37~20.04.1-Ubuntu SMP Mon Jun 13 22:51:01 UTC 2022 x86_64 x86_64
  CPU: Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz: 
              speed         user         nice          sys         idle          irq
       #1  2095 MHz       3453 s          2 s        527 s       3929 s          0 s
       #2  2095 MHz       5385 s          1 s        540 s       2014 s          0 s
       
  Memory: 6.783607482910156 GB (1872.98828125 MB free)
  Uptime: 799.55 sec
  Load Avg:  1.2  1.3  0.89
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-12.0.1 (ORCJIT, skylake-avx512)

Baseline

Julia Version 1.7.3
Commit 742b9abb4d (2022-05-06 12:58 UTC)
Platform Info:
  OS: Linux (x86_64-pc-linux-gnu)
      Ubuntu 20.04.4 LTS
  uname: Linux 5.13.0-1031-azure #37~20.04.1-Ubuntu SMP Mon Jun 13 22:51:01 UTC 2022 x86_64 x86_64
  CPU: Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz: 
              speed         user         nice          sys         idle          irq
       #1  2095 MHz       5976 s          2 s        621 s       4014 s          0 s
       #2  2095 MHz       5730 s          1 s        642 s       4260 s          0 s
       
  Memory: 6.783607482910156 GB (1794.375 MB free)
  Uptime: 1069.78 sec
  Load Avg:  1.43  1.28  0.99
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-12.0.1 (ORCJIT, skylake-avx512)

Target result

Benchmark Report for /home/runner/work/ExplainableAI.jl/ExplainableAI.jl

Job Properties

  • Time of benchmark: 27 Jun 2022 - 11:14
  • Package commit: cf5369
  • Julia commit: 742b9a
  • Julia command flags: None
  • Environment variables: None

Results

Below is a table of this job's results, obtained by running the benchmarks.
The values listed in the ID column have the structure [parent_group, child_group, ..., key], and can be used to
index into the BaseBenchmarks suite to retrieve the corresponding benchmarks.
The percentages accompanying time and memory values in the below table are noise tolerances. The "true"
time/memory value for a given benchmark is expected to fall within this percentage of the reported value.
An empty cell means that the value was zero.

ID time GC time memory allocations
["Layer", "Conv", "EpsilonRule"] 570.936 μs (5%) 1.36 MiB (1%) 131
["Layer", "Conv", "GammaRule"] 570.137 μs (5%) 1.36 MiB (1%) 140
["Layer", "Conv", "ZBoxRule"] 1.876 ms (5%) 4.46 MiB (1%) 495
["Layer", "Conv", "ZeroRule"] 569.736 μs (5%) 1.36 MiB (1%) 131
["Layer", "Dense", "EpsilonRule"] 36.803 μs (5%) 1.45 KiB (1%) 3
["Layer", "Dense", "GammaRule"] 102.207 μs (5%) 588.98 KiB (1%) 12
["Layer", "Dense", "ZBoxRule"] 340.022 μs (5%) 1017.84 KiB (1%) 152
["Layer", "Dense", "ZeroRule"] 40.102 μs (5%) 1.45 KiB (1%) 3
["Layer", "MaxPool", "EpsilonRule"] 56.103 μs (5%) 65.55 KiB (1%) 30
["Layer", "MaxPool", "GammaRule"] 56.704 μs (5%) 65.55 KiB (1%) 30
["Layer", "MaxPool", "ZBoxRule"] 56.404 μs (5%) 65.55 KiB (1%) 30
["Layer", "MaxPool", "ZeroRule"] 56.804 μs (5%) 65.55 KiB (1%) 30
["Layer", "WrappedDense", "EpsilonRule"] 83.606 μs (5%) 200.92 KiB (1%) 23
["Layer", "WrappedDense", "GammaRule"] 163.210 μs (5%) 788.45 KiB (1%) 32
["Layer", "WrappedDense", "ZBoxRule"] 329.721 μs (5%) 1017.84 KiB (1%) 152
["Layer", "WrappedDense", "ZeroRule"] 83.605 μs (5%) 200.92 KiB (1%) 23
["VGG", "Gradient", "analyze"] 1.165 s (5%) 11.652 ms 1.10 GiB (1%) 2012
["VGG", "Gradient", "construct analyzer"] 15.901 μs (5%) 26.66 KiB (1%) 189
["VGG", "InputTimesGradient", "analyze"] 1.145 s (5%) 11.758 ms 1.10 GiB (1%) 2015
["VGG", "InputTimesGradient", "construct analyzer"] 15.901 μs (5%) 26.66 KiB (1%) 189
["VGG", "IntegratedGradients", "analyze"] 10.267 s (5%) 81.998 ms 10.34 GiB (1%) 16942
["VGG", "IntegratedGradients", "construct analyzer"] 13.500 μs (5%) 24.69 KiB (1%) 187
["VGG", "LRPCustom", "analyze"] 1.984 s (5%) 20.510 ms 2.43 GiB (1%) 2300
["VGG", "LRPCustom", "construct analyzer"] 19.101 μs (5%) 45.88 KiB (1%) 159
["VGG", "LRPZero", "analyze"] 1.087 s (5%) 4.663 ms 761.54 MiB (1%) 1816
["VGG", "LRPZero", "construct analyzer"] 10.800 μs (5%) 43.88 KiB (1%) 82
["VGG", "SmoothGrad", "analyze"] 10.344 s (5%) 86.405 ms 10.33 GiB (1%) 16852
["VGG", "SmoothGrad", "construct analyzer"] 10.100 μs (5%) 24.69 KiB (1%) 187

Benchmark Group List

Here's a list of all the benchmark groups executed by this job:

  • ["Layer", "Conv"]
  • ["Layer", "Dense"]
  • ["Layer", "MaxPool"]
  • ["Layer", "WrappedDense"]
  • ["VGG", "Gradient"]
  • ["VGG", "InputTimesGradient"]
  • ["VGG", "IntegratedGradients"]
  • ["VGG", "LRPCustom"]
  • ["VGG", "LRPZero"]
  • ["VGG", "SmoothGrad"]

Julia versioninfo

Julia Version 1.7.3
Commit 742b9abb4d (2022-05-06 12:58 UTC)
Platform Info:
  OS: Linux (x86_64-pc-linux-gnu)
      Ubuntu 20.04.4 LTS
  uname: Linux 5.13.0-1031-azure #37~20.04.1-Ubuntu SMP Mon Jun 13 22:51:01 UTC 2022 x86_64 x86_64
  CPU: Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz: 
              speed         user         nice          sys         idle          irq
       #1  2095 MHz       3453 s          2 s        527 s       3929 s          0 s
       #2  2095 MHz       5385 s          1 s        540 s       2014 s          0 s
       
  Memory: 6.783607482910156 GB (1872.98828125 MB free)
  Uptime: 799.55 sec
  Load Avg:  1.2  1.3  0.89
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-12.0.1 (ORCJIT, skylake-avx512)

Baseline result

Benchmark Report for /home/runner/work/ExplainableAI.jl/ExplainableAI.jl

Job Properties

  • Time of benchmark: 27 Jun 2022 - 11:18
  • Package commit: 0d21c1
  • Julia commit: 742b9a
  • Julia command flags: None
  • Environment variables: None

Results

Below is a table of this job's results, obtained by running the benchmarks.
The values listed in the ID column have the structure [parent_group, child_group, ..., key], and can be used to
index into the BaseBenchmarks suite to retrieve the corresponding benchmarks.
The percentages accompanying time and memory values in the below table are noise tolerances. The "true"
time/memory value for a given benchmark is expected to fall within this percentage of the reported value.
An empty cell means that the value was zero.

ID time GC time memory allocations
["Layer", "Conv", "EpsilonRule"] 590.338 μs (5%) 1.43 MiB (1%) 169
["Layer", "Conv", "GammaRule"] 593.439 μs (5%) 1.44 MiB (1%) 181
["Layer", "Conv", "ZBoxRule"] 1.865 ms (5%) 4.46 MiB (1%) 492
["Layer", "Conv", "ZeroRule"] 592.339 μs (5%) 1.43 MiB (1%) 169
["Layer", "Dense", "EpsilonRule"] 34.502 μs (5%) 1.45 KiB (1%) 3
["Layer", "Dense", "GammaRule"] 101.406 μs (5%) 588.98 KiB (1%) 12
["Layer", "Dense", "ZBoxRule"] 341.022 μs (5%) 1017.84 KiB (1%) 152
["Layer", "Dense", "ZeroRule"] 34.502 μs (5%) 1.45 KiB (1%) 3
["Layer", "MaxPool", "EpsilonRule"] 76.905 μs (5%) 86.69 KiB (1%) 61
["Layer", "MaxPool", "GammaRule"] 77.205 μs (5%) 86.69 KiB (1%) 61
["Layer", "MaxPool", "ZBoxRule"] 68.804 μs (5%) 86.69 KiB (1%) 61
["Layer", "MaxPool", "ZeroRule"] 72.305 μs (5%) 85.97 KiB (1%) 61
["Layer", "WrappedDense", "EpsilonRule"] 95.207 μs (5%) 206.62 KiB (1%) 57
["Layer", "WrappedDense", "GammaRule"] 166.611 μs (5%) 794.16 KiB (1%) 66
["Layer", "WrappedDense", "ZBoxRule"] 341.121 μs (5%) 1017.84 KiB (1%) 152
["Layer", "WrappedDense", "ZeroRule"] 88.105 μs (5%) 206.19 KiB (1%) 57
["VGG", "Gradient", "analyze"] 1.180 s (5%) 12.731 ms 1.10 GiB (1%) 2012
["VGG", "Gradient", "construct analyzer"] 18.001 μs (5%) 26.66 KiB (1%) 189
["VGG", "InputTimesGradient", "analyze"] 20.459 s (5%) 221.956 ms 1.74 GiB (1%) 12117334
["VGG", "InputTimesGradient", "construct analyzer"] 10.001 μs (5%) 26.66 KiB (1%) 189
["VGG", "IntegratedGradients", "analyze"] 10.412 s (5%) 142.838 ms 10.34 GiB (1%) 16942
["VGG", "IntegratedGradients", "construct analyzer"] 13.701 μs (5%) 24.69 KiB (1%) 187
["VGG", "LRPCustom", "analyze"] 5.449 s (5%) 78.270 ms 2.76 GiB (1%) 6049173
["VGG", "LRPCustom", "construct analyzer"] 16.301 μs (5%) 45.88 KiB (1%) 159
["VGG", "LRPZero", "analyze"] 4.454 s (5%) 112.264 ms 1.06 GiB (1%) 5161651
["VGG", "LRPZero", "construct analyzer"] 12.501 μs (5%) 43.88 KiB (1%) 82
["VGG", "SmoothGrad", "analyze"] 11.861 s (5%) 150.591 ms 10.60 GiB (1%) 4876074
["VGG", "SmoothGrad", "construct analyzer"] 16.401 μs (5%) 24.69 KiB (1%) 187

Benchmark Group List

Here's a list of all the benchmark groups executed by this job:

  • ["Layer", "Conv"]
  • ["Layer", "Dense"]
  • ["Layer", "MaxPool"]
  • ["Layer", "WrappedDense"]
  • ["VGG", "Gradient"]
  • ["VGG", "InputTimesGradient"]
  • ["VGG", "IntegratedGradients"]
  • ["VGG", "LRPCustom"]
  • ["VGG", "LRPZero"]
  • ["VGG", "SmoothGrad"]

Julia versioninfo

Julia Version 1.7.3
Commit 742b9abb4d (2022-05-06 12:58 UTC)
Platform Info:
  OS: Linux (x86_64-pc-linux-gnu)
      Ubuntu 20.04.4 LTS
  uname: Linux 5.13.0-1031-azure #37~20.04.1-Ubuntu SMP Mon Jun 13 22:51:01 UTC 2022 x86_64 x86_64
  CPU: Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz: 
              speed         user         nice          sys         idle          irq
       #1  2095 MHz       5976 s          2 s        621 s       4014 s          0 s
       #2  2095 MHz       5730 s          1 s        642 s       4260 s          0 s
       
  Memory: 6.783607482910156 GB (1794.375 MB free)
  Uptime: 1069.78 sec
  Load Avg:  1.43  1.28  0.99
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-12.0.1 (ORCJIT, skylake-avx512)

Runtime information

Runtime Info
BLAS #threads 2
BLAS.vendor() openblas64
Sys.CPU_THREADS 2

lscpu output:

Architecture:                    x86_64
CPU op-mode(s):                  32-bit, 64-bit
Byte Order:                      Little Endian
Address sizes:                   46 bits physical, 48 bits virtual
CPU(s):                          2
On-line CPU(s) list:             0,1
Thread(s) per core:              1
Core(s) per socket:              2
Socket(s):                       1
NUMA node(s):                    1
Vendor ID:                       GenuineIntel
CPU family:                      6
Model:                           85
Model name:                      Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
Stepping:                        4
CPU MHz:                         2095.189
BogoMIPS:                        4190.37
Hypervisor vendor:               Microsoft
Virtualization type:             full
L1d cache:                       64 KiB
L1i cache:                       64 KiB
L2 cache:                        2 MiB
L3 cache:                        35.8 MiB
NUMA node0 CPU(s):               0,1
Vulnerability Itlb multihit:     KVM: Mitigation: VMX unsupported
Vulnerability L1tf:              Mitigation; PTE Inversion
Vulnerability Mds:               Mitigation; Clear CPU buffers; SMT Host state unknown
Vulnerability Meltdown:          Mitigation; PTI
Vulnerability Mmio stale data:   Vulnerable: Clear CPU buffers attempted, no microcode; SMT Host state unknown
Vulnerability Spec store bypass: Vulnerable
Vulnerability Spectre v1:        Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2:        Mitigation; Retpolines, STIBP disabled, RSB filling
Vulnerability Srbds:             Not affected
Vulnerability Tsx async abort:   Mitigation; Clear CPU buffers; SMT Host state unknown
Flags:                           fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ss ht syscall nx pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology cpuid pni pclmulqdq ssse3 fma cx16 pcid sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm abm 3dnowprefetch invpcid_single pti fsgsbase bmi1 hle avx2 smep bmi2 erms invpcid rtm mpx avx512f avx512dq rdseed adx smap clflushopt avx512cd avx512bw avx512vl xsaveopt xsavec xsaves md_clear
Cpu Property Value
Brand Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
Vendor :Intel
Architecture :Skylake
Model Family: 0x06, Model: 0x55, Stepping: 0x04, Type: 0x00
Cores 2 physical cores, 2 logical cores (on executing CPU)
No Hyperthreading hardware capability detected
Clock Frequencies Not supported by CPU
Data Cache Level 1:3 : (32, 1024, 36608) kbytes
64 byte cache line size
Address Size 48 bits virtual, 46 bits physical
SIMD 512 bit = 64 byte max. SIMD vector size
Time Stamp Counter TSC is accessible via rdtsc
TSC increased at every clock cycle (non-invariant TSC)
Perf. Monitoring Performance Monitoring Counters (PMC) are not supported
Hypervisor Yes, Microsoft

@github-actions
Copy link

Benchmark result

Judge result

Benchmark Report for /home/runner/work/ExplainableAI.jl/ExplainableAI.jl

Job Properties

  • Time of benchmarks:
    • Target: 27 Jun 2022 - 11:14
    • Baseline: 27 Jun 2022 - 11:19
  • Package commits:
    • Target: cf5369
    • Baseline: 0d21c1
  • Julia commits:
    • Target: 742b9a
    • Baseline: 742b9a
  • Julia command flags:
    • Target: None
    • Baseline: None
  • Environment variables:
    • Target: None
    • Baseline: None

Results

A ratio greater than 1.0 denotes a possible regression (marked with ❌), while a ratio less
than 1.0 denotes a possible improvement (marked with ✅). Only significant results - results
that indicate possible regressions or improvements - are shown below (thus, an empty table means that all
benchmark results remained invariant between builds).

ID time ratio memory ratio
["Layer", "Conv", "EpsilonRule"] 0.97 (5%) 0.95 (1%) ✅
["Layer", "Conv", "GammaRule"] 0.96 (5%) 0.95 (1%) ✅
["Layer", "Conv", "ZeroRule"] 0.98 (5%) 0.95 (1%) ✅
["Layer", "Dense", "GammaRule"] 1.36 (5%) ❌ 1.00 (1%)
["Layer", "MaxPool", "EpsilonRule"] 0.95 (5%) ✅ 0.76 (1%) ✅
["Layer", "MaxPool", "GammaRule"] 0.95 (5%) 0.76 (1%) ✅
["Layer", "MaxPool", "ZBoxRule"] 0.94 (5%) ✅ 0.76 (1%) ✅
["Layer", "MaxPool", "ZeroRule"] 0.93 (5%) ✅ 0.76 (1%) ✅
["Layer", "WrappedDense", "EpsilonRule"] 0.95 (5%) 0.97 (1%) ✅
["Layer", "WrappedDense", "GammaRule"] 0.93 (5%) ✅ 0.99 (1%)
["Layer", "WrappedDense", "ZeroRule"] 0.99 (5%) 0.97 (1%) ✅
["VGG", "Gradient", "analyze"] 0.94 (5%) ✅ 1.00 (1%)
["VGG", "Gradient", "construct analyzer"] 0.75 (5%) ✅ 1.00 (1%)
["VGG", "InputTimesGradient", "analyze"] 0.06 (5%) ✅ 0.63 (1%) ✅
["VGG", "InputTimesGradient", "construct analyzer"] 1.17 (5%) ❌ 1.00 (1%)
["VGG", "IntegratedGradients", "construct analyzer"] 0.76 (5%) ✅ 1.00 (1%)
["VGG", "LRPCustom", "analyze"] 0.35 (5%) ✅ 0.88 (1%) ✅
["VGG", "LRPZero", "analyze"] 0.25 (5%) ✅ 0.70 (1%) ✅
["VGG", "LRPZero", "construct analyzer"] 1.06 (5%) ❌ 1.00 (1%)
["VGG", "SmoothGrad", "analyze"] 0.89 (5%) ✅ 0.97 (1%) ✅
["VGG", "SmoothGrad", "construct analyzer"] 1.14 (5%) ❌ 1.00 (1%)

Benchmark Group List

Here's a list of all the benchmark groups executed by this job:

  • ["Layer", "Conv"]
  • ["Layer", "Dense"]
  • ["Layer", "MaxPool"]
  • ["Layer", "WrappedDense"]
  • ["VGG", "Gradient"]
  • ["VGG", "InputTimesGradient"]
  • ["VGG", "IntegratedGradients"]
  • ["VGG", "LRPCustom"]
  • ["VGG", "LRPZero"]
  • ["VGG", "SmoothGrad"]

Julia versioninfo

Target

Julia Version 1.7.3
Commit 742b9abb4d (2022-05-06 12:58 UTC)
Platform Info:
  OS: Linux (x86_64-pc-linux-gnu)
      Ubuntu 20.04.4 LTS
  uname: Linux 5.13.0-1031-azure #37~20.04.1-Ubuntu SMP Mon Jun 13 22:51:01 UTC 2022 x86_64 x86_64
  CPU: Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz: 
              speed         user         nice          sys         idle          irq
       #1  2095 MHz       3256 s          2 s        534 s       3928 s          0 s
       #2  2095 MHz       6013 s          1 s        524 s       1227 s          0 s
       
  Memory: 6.783607482910156 GB (1863.57421875 MB free)
  Uptime: 780.07 sec
  Load Avg:  1.14  1.26  0.89
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-12.0.1 (ORCJIT, skylake-avx512)

Baseline

Julia Version 1.7.3
Commit 742b9abb4d (2022-05-06 12:58 UTC)
Platform Info:
  OS: Linux (x86_64-pc-linux-gnu)
      Ubuntu 20.04.4 LTS
  uname: Linux 5.13.0-1031-azure #37~20.04.1-Ubuntu SMP Mon Jun 13 22:51:01 UTC 2022 x86_64 x86_64
  CPU: Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz: 
              speed         user         nice          sys         idle          irq
       #1  2095 MHz       5758 s          2 s        619 s       4180 s          0 s
       #2  2095 MHz       6522 s          1 s        633 s       3442 s          0 s
       
  Memory: 6.783607482910156 GB (1742.95703125 MB free)
  Uptime: 1064.0 sec
  Load Avg:  1.1  1.17  0.96
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-12.0.1 (ORCJIT, skylake-avx512)

Target result

Benchmark Report for /home/runner/work/ExplainableAI.jl/ExplainableAI.jl

Job Properties

  • Time of benchmark: 27 Jun 2022 - 11:14
  • Package commit: cf5369
  • Julia commit: 742b9a
  • Julia command flags: None
  • Environment variables: None

Results

Below is a table of this job's results, obtained by running the benchmarks.
The values listed in the ID column have the structure [parent_group, child_group, ..., key], and can be used to
index into the BaseBenchmarks suite to retrieve the corresponding benchmarks.
The percentages accompanying time and memory values in the below table are noise tolerances. The "true"
time/memory value for a given benchmark is expected to fall within this percentage of the reported value.
An empty cell means that the value was zero.

ID time GC time memory allocations
["Layer", "Conv", "EpsilonRule"] 715.245 μs (5%) 1.36 MiB (1%) 131
["Layer", "Conv", "GammaRule"] 715.047 μs (5%) 1.36 MiB (1%) 140
["Layer", "Conv", "ZBoxRule"] 2.256 ms (5%) 4.46 MiB (1%) 495
["Layer", "Conv", "ZeroRule"] 710.145 μs (5%) 1.36 MiB (1%) 131
["Layer", "Dense", "EpsilonRule"] 49.603 μs (5%) 1.45 KiB (1%) 3
["Layer", "Dense", "GammaRule"] 117.607 μs (5%) 588.98 KiB (1%) 12
["Layer", "Dense", "ZBoxRule"] 396.424 μs (5%) 1017.84 KiB (1%) 152
["Layer", "Dense", "ZeroRule"] 49.603 μs (5%) 1.45 KiB (1%) 3
["Layer", "MaxPool", "EpsilonRule"] 81.505 μs (5%) 65.55 KiB (1%) 30
["Layer", "MaxPool", "GammaRule"] 77.805 μs (5%) 65.55 KiB (1%) 30
["Layer", "MaxPool", "ZBoxRule"] 77.805 μs (5%) 65.55 KiB (1%) 30
["Layer", "MaxPool", "ZeroRule"] 78.905 μs (5%) 65.55 KiB (1%) 30
["Layer", "WrappedDense", "EpsilonRule"] 104.007 μs (5%) 200.92 KiB (1%) 23
["Layer", "WrappedDense", "GammaRule"] 177.211 μs (5%) 788.45 KiB (1%) 32
["Layer", "WrappedDense", "ZBoxRule"] 396.026 μs (5%) 1017.84 KiB (1%) 152
["Layer", "WrappedDense", "ZeroRule"] 104.007 μs (5%) 200.92 KiB (1%) 23
["VGG", "Gradient", "analyze"] 1.168 s (5%) 10.115 ms 1.10 GiB (1%) 2012
["VGG", "Gradient", "construct analyzer"] 15.801 μs (5%) 26.66 KiB (1%) 189
["VGG", "InputTimesGradient", "analyze"] 1.182 s (5%) 10.018 ms 1.10 GiB (1%) 2015
["VGG", "InputTimesGradient", "construct analyzer"] 16.001 μs (5%) 26.66 KiB (1%) 189
["VGG", "IntegratedGradients", "analyze"] 10.675 s (5%) 82.006 ms 10.34 GiB (1%) 16942
["VGG", "IntegratedGradients", "construct analyzer"] 14.801 μs (5%) 24.69 KiB (1%) 187
["VGG", "LRPCustom", "analyze"] 1.988 s (5%) 17.187 ms 2.43 GiB (1%) 2300
["VGG", "LRPCustom", "construct analyzer"] 23.001 μs (5%) 45.88 KiB (1%) 159
["VGG", "LRPZero", "analyze"] 1.175 s (5%) 9.108 ms 761.54 MiB (1%) 1816
["VGG", "LRPZero", "construct analyzer"] 14.800 μs (5%) 43.88 KiB (1%) 82
["VGG", "SmoothGrad", "analyze"] 10.971 s (5%) 80.141 ms 10.33 GiB (1%) 16852
["VGG", "SmoothGrad", "construct analyzer"] 14.501 μs (5%) 24.69 KiB (1%) 187

Benchmark Group List

Here's a list of all the benchmark groups executed by this job:

  • ["Layer", "Conv"]
  • ["Layer", "Dense"]
  • ["Layer", "MaxPool"]
  • ["Layer", "WrappedDense"]
  • ["VGG", "Gradient"]
  • ["VGG", "InputTimesGradient"]
  • ["VGG", "IntegratedGradients"]
  • ["VGG", "LRPCustom"]
  • ["VGG", "LRPZero"]
  • ["VGG", "SmoothGrad"]

Julia versioninfo

Julia Version 1.7.3
Commit 742b9abb4d (2022-05-06 12:58 UTC)
Platform Info:
  OS: Linux (x86_64-pc-linux-gnu)
      Ubuntu 20.04.4 LTS
  uname: Linux 5.13.0-1031-azure #37~20.04.1-Ubuntu SMP Mon Jun 13 22:51:01 UTC 2022 x86_64 x86_64
  CPU: Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz: 
              speed         user         nice          sys         idle          irq
       #1  2095 MHz       3256 s          2 s        534 s       3928 s          0 s
       #2  2095 MHz       6013 s          1 s        524 s       1227 s          0 s
       
  Memory: 6.783607482910156 GB (1863.57421875 MB free)
  Uptime: 780.07 sec
  Load Avg:  1.14  1.26  0.89
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-12.0.1 (ORCJIT, skylake-avx512)

Baseline result

Benchmark Report for /home/runner/work/ExplainableAI.jl/ExplainableAI.jl

Job Properties

  • Time of benchmark: 27 Jun 2022 - 11:19
  • Package commit: 0d21c1
  • Julia commit: 742b9a
  • Julia command flags: None
  • Environment variables: None

Results

Below is a table of this job's results, obtained by running the benchmarks.
The values listed in the ID column have the structure [parent_group, child_group, ..., key], and can be used to
index into the BaseBenchmarks suite to retrieve the corresponding benchmarks.
The percentages accompanying time and memory values in the below table are noise tolerances. The "true"
time/memory value for a given benchmark is expected to fall within this percentage of the reported value.
An empty cell means that the value was zero.

ID time GC time memory allocations
["Layer", "Conv", "EpsilonRule"] 738.047 μs (5%) 1.43 MiB (1%) 169
["Layer", "Conv", "GammaRule"] 748.047 μs (5%) 1.44 MiB (1%) 181
["Layer", "Conv", "ZBoxRule"] 2.306 ms (5%) 4.46 MiB (1%) 492
["Layer", "Conv", "ZeroRule"] 724.846 μs (5%) 1.43 MiB (1%) 169
["Layer", "Dense", "EpsilonRule"] 48.203 μs (5%) 1.45 KiB (1%) 3
["Layer", "Dense", "GammaRule"] 86.706 μs (5%) 588.98 KiB (1%) 12
["Layer", "Dense", "ZBoxRule"] 396.225 μs (5%) 1017.84 KiB (1%) 152
["Layer", "Dense", "ZeroRule"] 48.203 μs (5%) 1.45 KiB (1%) 3
["Layer", "MaxPool", "EpsilonRule"] 85.805 μs (5%) 86.69 KiB (1%) 61
["Layer", "MaxPool", "GammaRule"] 81.505 μs (5%) 86.69 KiB (1%) 61
["Layer", "MaxPool", "ZBoxRule"] 82.905 μs (5%) 86.69 KiB (1%) 61
["Layer", "MaxPool", "ZeroRule"] 85.006 μs (5%) 85.97 KiB (1%) 61
["Layer", "WrappedDense", "EpsilonRule"] 109.107 μs (5%) 206.62 KiB (1%) 57
["Layer", "WrappedDense", "GammaRule"] 190.212 μs (5%) 794.16 KiB (1%) 66
["Layer", "WrappedDense", "ZBoxRule"] 399.225 μs (5%) 1017.84 KiB (1%) 152
["Layer", "WrappedDense", "ZeroRule"] 104.907 μs (5%) 206.19 KiB (1%) 57
["VGG", "Gradient", "analyze"] 1.241 s (5%) 11.950 ms 1.10 GiB (1%) 2012
["VGG", "Gradient", "construct analyzer"] 21.102 μs (5%) 26.66 KiB (1%) 189
["VGG", "InputTimesGradient", "analyze"] 21.210 s (5%) 223.530 ms 1.74 GiB (1%) 12124565
["VGG", "InputTimesGradient", "construct analyzer"] 13.701 μs (5%) 26.66 KiB (1%) 189
["VGG", "IntegratedGradients", "analyze"] 10.970 s (5%) 138.230 ms 10.34 GiB (1%) 16942
["VGG", "IntegratedGradients", "construct analyzer"] 19.501 μs (5%) 24.69 KiB (1%) 187
["VGG", "LRPCustom", "analyze"] 5.743 s (5%) 79.578 ms 2.76 GiB (1%) 6050812
["VGG", "LRPCustom", "construct analyzer"] 23.201 μs (5%) 45.88 KiB (1%) 159
["VGG", "LRPZero", "analyze"] 4.788 s (5%) 113.795 ms 1.06 GiB (1%) 5182525
["VGG", "LRPZero", "construct analyzer"] 13.901 μs (5%) 43.88 KiB (1%) 82
["VGG", "SmoothGrad", "analyze"] 12.381 s (5%) 149.222 ms 10.60 GiB (1%) 4942857
["VGG", "SmoothGrad", "construct analyzer"] 12.701 μs (5%) 24.69 KiB (1%) 187

Benchmark Group List

Here's a list of all the benchmark groups executed by this job:

  • ["Layer", "Conv"]
  • ["Layer", "Dense"]
  • ["Layer", "MaxPool"]
  • ["Layer", "WrappedDense"]
  • ["VGG", "Gradient"]
  • ["VGG", "InputTimesGradient"]
  • ["VGG", "IntegratedGradients"]
  • ["VGG", "LRPCustom"]
  • ["VGG", "LRPZero"]
  • ["VGG", "SmoothGrad"]

Julia versioninfo

Julia Version 1.7.3
Commit 742b9abb4d (2022-05-06 12:58 UTC)
Platform Info:
  OS: Linux (x86_64-pc-linux-gnu)
      Ubuntu 20.04.4 LTS
  uname: Linux 5.13.0-1031-azure #37~20.04.1-Ubuntu SMP Mon Jun 13 22:51:01 UTC 2022 x86_64 x86_64
  CPU: Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz: 
              speed         user         nice          sys         idle          irq
       #1  2095 MHz       5758 s          2 s        619 s       4180 s          0 s
       #2  2095 MHz       6522 s          1 s        633 s       3442 s          0 s
       
  Memory: 6.783607482910156 GB (1742.95703125 MB free)
  Uptime: 1064.0 sec
  Load Avg:  1.1  1.17  0.96
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-12.0.1 (ORCJIT, skylake-avx512)

Runtime information

Runtime Info
BLAS #threads 2
BLAS.vendor() openblas64
Sys.CPU_THREADS 2

lscpu output:

Architecture:                    x86_64
CPU op-mode(s):                  32-bit, 64-bit
Byte Order:                      Little Endian
Address sizes:                   46 bits physical, 48 bits virtual
CPU(s):                          2
On-line CPU(s) list:             0,1
Thread(s) per core:              1
Core(s) per socket:              2
Socket(s):                       1
NUMA node(s):                    1
Vendor ID:                       GenuineIntel
CPU family:                      6
Model:                           85
Model name:                      Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
Stepping:                        4
CPU MHz:                         2095.172
BogoMIPS:                        4190.34
Hypervisor vendor:               Microsoft
Virtualization type:             full
L1d cache:                       64 KiB
L1i cache:                       64 KiB
L2 cache:                        2 MiB
L3 cache:                        35.8 MiB
NUMA node0 CPU(s):               0,1
Vulnerability Itlb multihit:     KVM: Mitigation: VMX unsupported
Vulnerability L1tf:              Mitigation; PTE Inversion
Vulnerability Mds:               Mitigation; Clear CPU buffers; SMT Host state unknown
Vulnerability Meltdown:          Mitigation; PTI
Vulnerability Mmio stale data:   Vulnerable: Clear CPU buffers attempted, no microcode; SMT Host state unknown
Vulnerability Spec store bypass: Vulnerable
Vulnerability Spectre v1:        Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2:        Mitigation; Retpolines, STIBP disabled, RSB filling
Vulnerability Srbds:             Not affected
Vulnerability Tsx async abort:   Mitigation; Clear CPU buffers; SMT Host state unknown
Flags:                           fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ss ht syscall nx pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology cpuid pni pclmulqdq ssse3 fma cx16 pcid sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm abm 3dnowprefetch invpcid_single pti fsgsbase bmi1 hle avx2 smep bmi2 erms invpcid rtm mpx avx512f avx512dq rdseed adx smap clflushopt avx512cd avx512bw avx512vl xsaveopt xsavec xsaves md_clear
Cpu Property Value
Brand Intel(R) Xeon(R) Platinum 8171M CPU @ 2.60GHz
Vendor :Intel
Architecture :Skylake
Model Family: 0x06, Model: 0x55, Stepping: 0x04, Type: 0x00
Cores 2 physical cores, 2 logical cores (on executing CPU)
No Hyperthreading hardware capability detected
Clock Frequencies Not supported by CPU
Data Cache Level 1:3 : (32, 1024, 36608) kbytes
64 byte cache line size
Address Size 48 bits virtual, 46 bits physical
SIMD 512 bit = 64 byte max. SIMD vector size
Time Stamp Counter TSC is accessible via rdtsc
TSC increased at every clock cycle (non-invariant TSC)
Perf. Monitoring Performance Monitoring Counters (PMC) are not supported
Hypervisor Yes, Microsoft

@github-actions
Copy link

Benchmark result

Judge result

Benchmark Report for /home/runner/work/ExplainableAI.jl/ExplainableAI.jl

Job Properties

  • Time of benchmarks:
    • Target: 27 Jun 2022 - 11:34
    • Baseline: 27 Jun 2022 - 11:39
  • Package commits:
    • Target: 04593f
    • Baseline: 0d21c1
  • Julia commits:
    • Target: 742b9a
    • Baseline: 742b9a
  • Julia command flags:
    • Target: None
    • Baseline: None
  • Environment variables:
    • Target: None
    • Baseline: None

Results

A ratio greater than 1.0 denotes a possible regression (marked with ❌), while a ratio less
than 1.0 denotes a possible improvement (marked with ✅). Only significant results - results
that indicate possible regressions or improvements - are shown below (thus, an empty table means that all
benchmark results remained invariant between builds).

ID time ratio memory ratio
["Layer", "Conv", "EpsilonRule"] 0.98 (5%) 0.95 (1%) ✅
["Layer", "Conv", "GammaRule"] 0.98 (5%) 0.95 (1%) ✅
["Layer", "Conv", "ZeroRule"] 0.98 (5%) 0.95 (1%) ✅
["Layer", "Dense", "GammaRule"] 1.05 (5%) ❌ 1.00 (1%)
["Layer", "MaxPool", "EpsilonRule"] 0.89 (5%) ✅ 0.76 (1%) ✅
["Layer", "MaxPool", "GammaRule"] 0.89 (5%) ✅ 0.76 (1%) ✅
["Layer", "MaxPool", "ZBoxRule"] 0.87 (5%) ✅ 0.76 (1%) ✅
["Layer", "MaxPool", "ZeroRule"] 0.88 (5%) ✅ 0.76 (1%) ✅
["Layer", "WrappedDense", "EpsilonRule"] 1.09 (5%) ❌ 0.97 (1%) ✅
["Layer", "WrappedDense", "ZBoxRule"] 1.13 (5%) ❌ 1.00 (1%)
["Layer", "WrappedDense", "ZeroRule"] 1.07 (5%) ❌ 0.97 (1%) ✅
["VGG", "Gradient", "construct analyzer"] 0.75 (5%) ✅ 1.00 (1%)
["VGG", "InputTimesGradient", "analyze"] 0.06 (5%) ✅ 0.63 (1%) ✅
["VGG", "InputTimesGradient", "construct analyzer"] 1.39 (5%) ❌ 1.00 (1%)
["VGG", "IntegratedGradients", "construct analyzer"] 0.75 (5%) ✅ 1.00 (1%)
["VGG", "LRPCustom", "analyze"] 0.41 (5%) ✅ 0.88 (1%) ✅
["VGG", "LRPZero", "analyze"] 0.28 (5%) ✅ 0.70 (1%) ✅
["VGG", "LRPZero", "construct analyzer"] 1.19 (5%) ❌ 1.00 (1%)
["VGG", "SmoothGrad", "analyze"] 0.89 (5%) ✅ 0.97 (1%) ✅
["VGG", "SmoothGrad", "construct analyzer"] 1.37 (5%) ❌ 1.00 (1%)

Benchmark Group List

Here's a list of all the benchmark groups executed by this job:

  • ["Layer", "Conv"]
  • ["Layer", "Dense"]
  • ["Layer", "MaxPool"]
  • ["Layer", "WrappedDense"]
  • ["VGG", "Gradient"]
  • ["VGG", "InputTimesGradient"]
  • ["VGG", "IntegratedGradients"]
  • ["VGG", "LRPCustom"]
  • ["VGG", "LRPZero"]
  • ["VGG", "SmoothGrad"]

Julia versioninfo

Target

Julia Version 1.7.3
Commit 742b9abb4d (2022-05-06 12:58 UTC)
Platform Info:
  OS: Linux (x86_64-pc-linux-gnu)
      Ubuntu 20.04.4 LTS
  uname: Linux 5.13.0-1031-azure #37~20.04.1-Ubuntu SMP Mon Jun 13 22:51:01 UTC 2022 x86_64 x86_64
  CPU: Intel(R) Xeon(R) CPU E5-2673 v4 @ 2.30GHz: 
              speed         user         nice          sys         idle          irq
       #1  2294 MHz       4324 s          1 s        535 s       2932 s          0 s
       #2  2294 MHz       4794 s          1 s        500 s       2517 s          0 s
       
  Memory: 6.783607482910156 GB (865.23046875 MB free)
  Uptime: 785.81 sec
  Load Avg:  1.2  1.46  0.99
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-12.0.1 (ORCJIT, broadwell)

Baseline

Julia Version 1.7.3
Commit 742b9abb4d (2022-05-06 12:58 UTC)
Platform Info:
  OS: Linux (x86_64-pc-linux-gnu)
      Ubuntu 20.04.4 LTS
  uname: Linux 5.13.0-1031-azure #37~20.04.1-Ubuntu SMP Mon Jun 13 22:51:01 UTC 2022 x86_64 x86_64
  CPU: Intel(R) Xeon(R) CPU E5-2673 v4 @ 2.30GHz: 
              speed         user         nice          sys         idle          irq
       #1  2294 MHz       4822 s          1 s        640 s       5121 s          0 s
       #2  2294 MHz       7314 s          1 s        581 s       2713 s          0 s
       
  Memory: 6.783607482910156 GB (1779.73828125 MB free)
  Uptime: 1065.63 sec
  Load Avg:  1.1  1.27  1.03
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-12.0.1 (ORCJIT, broadwell)

Target result

Benchmark Report for /home/runner/work/ExplainableAI.jl/ExplainableAI.jl

Job Properties

  • Time of benchmark: 27 Jun 2022 - 11:34
  • Package commit: 04593f
  • Julia commit: 742b9a
  • Julia command flags: None
  • Environment variables: None

Results

Below is a table of this job's results, obtained by running the benchmarks.
The values listed in the ID column have the structure [parent_group, child_group, ..., key], and can be used to
index into the BaseBenchmarks suite to retrieve the corresponding benchmarks.
The percentages accompanying time and memory values in the below table are noise tolerances. The "true"
time/memory value for a given benchmark is expected to fall within this percentage of the reported value.
An empty cell means that the value was zero.

ID time GC time memory allocations
["Layer", "Conv", "EpsilonRule"] 840.206 μs (5%) 1.36 MiB (1%) 131
["Layer", "Conv", "GammaRule"] 842.707 μs (5%) 1.36 MiB (1%) 140
["Layer", "Conv", "ZBoxRule"] 2.644 ms (5%) 4.46 MiB (1%) 495
["Layer", "Conv", "ZeroRule"] 840.906 μs (5%) 1.36 MiB (1%) 131
["Layer", "Dense", "EpsilonRule"] 71.001 μs (5%) 1.45 KiB (1%) 3
["Layer", "Dense", "GammaRule"] 179.301 μs (5%) 588.98 KiB (1%) 12
["Layer", "Dense", "ZBoxRule"] 563.804 μs (5%) 1017.84 KiB (1%) 152
["Layer", "Dense", "ZeroRule"] 71.001 μs (5%) 1.45 KiB (1%) 3
["Layer", "MaxPool", "EpsilonRule"] 86.400 μs (5%) 65.55 KiB (1%) 30
["Layer", "MaxPool", "GammaRule"] 87.001 μs (5%) 65.55 KiB (1%) 30
["Layer", "MaxPool", "ZBoxRule"] 85.901 μs (5%) 65.55 KiB (1%) 30
["Layer", "MaxPool", "ZeroRule"] 86.100 μs (5%) 65.55 KiB (1%) 30
["Layer", "WrappedDense", "EpsilonRule"] 154.602 μs (5%) 200.92 KiB (1%) 23
["Layer", "WrappedDense", "GammaRule"] 265.202 μs (5%) 788.45 KiB (1%) 32
["Layer", "WrappedDense", "ZBoxRule"] 565.804 μs (5%) 1017.84 KiB (1%) 152
["Layer", "WrappedDense", "ZeroRule"] 154.701 μs (5%) 200.92 KiB (1%) 23
["VGG", "Gradient", "analyze"] 1.345 s (5%) 15.755 ms 1.10 GiB (1%) 2012
["VGG", "Gradient", "construct analyzer"] 15.300 μs (5%) 26.66 KiB (1%) 189
["VGG", "InputTimesGradient", "analyze"] 1.342 s (5%) 14.993 ms 1.10 GiB (1%) 2015
["VGG", "InputTimesGradient", "construct analyzer"] 15.600 μs (5%) 26.66 KiB (1%) 189
["VGG", "IntegratedGradients", "analyze"] 12.593 s (5%) 114.950 ms 10.34 GiB (1%) 16942
["VGG", "IntegratedGradients", "construct analyzer"] 13.000 μs (5%) 24.69 KiB (1%) 187
["VGG", "LRPCustom", "analyze"] 2.319 s (5%) 30.376 ms 2.43 GiB (1%) 2300
["VGG", "LRPCustom", "construct analyzer"] 27.700 μs (5%) 45.88 KiB (1%) 159
["VGG", "LRPZero", "analyze"] 1.258 s (5%) 12.295 ms 761.54 MiB (1%) 1816
["VGG", "LRPZero", "construct analyzer"] 18.700 μs (5%) 43.88 KiB (1%) 82
["VGG", "SmoothGrad", "analyze"] 12.575 s (5%) 117.647 ms 10.33 GiB (1%) 16852
["VGG", "SmoothGrad", "construct analyzer"] 12.900 μs (5%) 24.69 KiB (1%) 187

Benchmark Group List

Here's a list of all the benchmark groups executed by this job:

  • ["Layer", "Conv"]
  • ["Layer", "Dense"]
  • ["Layer", "MaxPool"]
  • ["Layer", "WrappedDense"]
  • ["VGG", "Gradient"]
  • ["VGG", "InputTimesGradient"]
  • ["VGG", "IntegratedGradients"]
  • ["VGG", "LRPCustom"]
  • ["VGG", "LRPZero"]
  • ["VGG", "SmoothGrad"]

Julia versioninfo

Julia Version 1.7.3
Commit 742b9abb4d (2022-05-06 12:58 UTC)
Platform Info:
  OS: Linux (x86_64-pc-linux-gnu)
      Ubuntu 20.04.4 LTS
  uname: Linux 5.13.0-1031-azure #37~20.04.1-Ubuntu SMP Mon Jun 13 22:51:01 UTC 2022 x86_64 x86_64
  CPU: Intel(R) Xeon(R) CPU E5-2673 v4 @ 2.30GHz: 
              speed         user         nice          sys         idle          irq
       #1  2294 MHz       4324 s          1 s        535 s       2932 s          0 s
       #2  2294 MHz       4794 s          1 s        500 s       2517 s          0 s
       
  Memory: 6.783607482910156 GB (865.23046875 MB free)
  Uptime: 785.81 sec
  Load Avg:  1.2  1.46  0.99
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-12.0.1 (ORCJIT, broadwell)

Baseline result

Benchmark Report for /home/runner/work/ExplainableAI.jl/ExplainableAI.jl

Job Properties

  • Time of benchmark: 27 Jun 2022 - 11:39
  • Package commit: 0d21c1
  • Julia commit: 742b9a
  • Julia command flags: None
  • Environment variables: None

Results

Below is a table of this job's results, obtained by running the benchmarks.
The values listed in the ID column have the structure [parent_group, child_group, ..., key], and can be used to
index into the BaseBenchmarks suite to retrieve the corresponding benchmarks.
The percentages accompanying time and memory values in the below table are noise tolerances. The "true"
time/memory value for a given benchmark is expected to fall within this percentage of the reported value.
An empty cell means that the value was zero.

ID time GC time memory allocations
["Layer", "Conv", "EpsilonRule"] 858.807 μs (5%) 1.43 MiB (1%) 169
["Layer", "Conv", "GammaRule"] 863.206 μs (5%) 1.44 MiB (1%) 181
["Layer", "Conv", "ZBoxRule"] 2.540 ms (5%) 4.46 MiB (1%) 492
["Layer", "Conv", "ZeroRule"] 857.206 μs (5%) 1.43 MiB (1%) 169
["Layer", "Dense", "EpsilonRule"] 70.801 μs (5%) 1.45 KiB (1%) 3
["Layer", "Dense", "GammaRule"] 170.201 μs (5%) 588.98 KiB (1%) 12
["Layer", "Dense", "ZBoxRule"] 560.304 μs (5%) 1017.84 KiB (1%) 152
["Layer", "Dense", "ZeroRule"] 70.800 μs (5%) 1.45 KiB (1%) 3
["Layer", "MaxPool", "EpsilonRule"] 97.201 μs (5%) 86.69 KiB (1%) 61
["Layer", "MaxPool", "GammaRule"] 98.200 μs (5%) 86.69 KiB (1%) 61
["Layer", "MaxPool", "ZBoxRule"] 98.801 μs (5%) 86.69 KiB (1%) 61
["Layer", "MaxPool", "ZeroRule"] 97.501 μs (5%) 85.97 KiB (1%) 61
["Layer", "WrappedDense", "EpsilonRule"] 141.301 μs (5%) 206.62 KiB (1%) 57
["Layer", "WrappedDense", "GammaRule"] 272.102 μs (5%) 794.16 KiB (1%) 66
["Layer", "WrappedDense", "ZBoxRule"] 501.903 μs (5%) 1017.84 KiB (1%) 152
["Layer", "WrappedDense", "ZeroRule"] 144.601 μs (5%) 206.19 KiB (1%) 57
["VGG", "Gradient", "analyze"] 1.337 s (5%) 17.386 ms 1.10 GiB (1%) 2012
["VGG", "Gradient", "construct analyzer"] 20.500 μs (5%) 26.66 KiB (1%) 189
["VGG", "InputTimesGradient", "analyze"] 20.816 s (5%) 237.364 ms 1.74 GiB (1%) 12125962
["VGG", "InputTimesGradient", "construct analyzer"] 11.200 μs (5%) 26.66 KiB (1%) 189
["VGG", "IntegratedGradients", "analyze"] 12.719 s (5%) 179.364 ms 10.34 GiB (1%) 16942
["VGG", "IntegratedGradients", "construct analyzer"] 17.300 μs (5%) 24.69 KiB (1%) 187
["VGG", "LRPCustom", "analyze"] 5.651 s (5%) 91.423 ms 2.76 GiB (1%) 6049107
["VGG", "LRPCustom", "construct analyzer"] 26.701 μs (5%) 45.88 KiB (1%) 159
["VGG", "LRPZero", "analyze"] 4.457 s (5%) 135.487 ms 1.06 GiB (1%) 5172159
["VGG", "LRPZero", "construct analyzer"] 15.700 μs (5%) 43.88 KiB (1%) 82
["VGG", "SmoothGrad", "analyze"] 14.084 s (5%) 188.047 ms 10.60 GiB (1%) 4904284
["VGG", "SmoothGrad", "construct analyzer"] 9.400 μs (5%) 24.69 KiB (1%) 187

Benchmark Group List

Here's a list of all the benchmark groups executed by this job:

  • ["Layer", "Conv"]
  • ["Layer", "Dense"]
  • ["Layer", "MaxPool"]
  • ["Layer", "WrappedDense"]
  • ["VGG", "Gradient"]
  • ["VGG", "InputTimesGradient"]
  • ["VGG", "IntegratedGradients"]
  • ["VGG", "LRPCustom"]
  • ["VGG", "LRPZero"]
  • ["VGG", "SmoothGrad"]

Julia versioninfo

Julia Version 1.7.3
Commit 742b9abb4d (2022-05-06 12:58 UTC)
Platform Info:
  OS: Linux (x86_64-pc-linux-gnu)
      Ubuntu 20.04.4 LTS
  uname: Linux 5.13.0-1031-azure #37~20.04.1-Ubuntu SMP Mon Jun 13 22:51:01 UTC 2022 x86_64 x86_64
  CPU: Intel(R) Xeon(R) CPU E5-2673 v4 @ 2.30GHz: 
              speed         user         nice          sys         idle          irq
       #1  2294 MHz       4822 s          1 s        640 s       5121 s          0 s
       #2  2294 MHz       7314 s          1 s        581 s       2713 s          0 s
       
  Memory: 6.783607482910156 GB (1779.73828125 MB free)
  Uptime: 1065.63 sec
  Load Avg:  1.1  1.27  1.03
  WORD_SIZE: 64
  LIBM: libopenlibm
  LLVM: libLLVM-12.0.1 (ORCJIT, broadwell)

Runtime information

Runtime Info
BLAS #threads 2
BLAS.vendor() openblas64
Sys.CPU_THREADS 2

lscpu output:

Architecture:                    x86_64
CPU op-mode(s):                  32-bit, 64-bit
Byte Order:                      Little Endian
Address sizes:                   46 bits physical, 48 bits virtual
CPU(s):                          2
On-line CPU(s) list:             0,1
Thread(s) per core:              1
Core(s) per socket:              2
Socket(s):                       1
NUMA node(s):                    1
Vendor ID:                       GenuineIntel
CPU family:                      6
Model:                           79
Model name:                      Intel(R) Xeon(R) CPU E5-2673 v4 @ 2.30GHz
Stepping:                        1
CPU MHz:                         2294.687
BogoMIPS:                        4589.37
Hypervisor vendor:               Microsoft
Virtualization type:             full
L1d cache:                       64 KiB
L1i cache:                       64 KiB
L2 cache:                        512 KiB
L3 cache:                        50 MiB
NUMA node0 CPU(s):               0,1
Vulnerability Itlb multihit:     KVM: Mitigation: VMX unsupported
Vulnerability L1tf:              Mitigation; PTE Inversion
Vulnerability Mds:               Mitigation; Clear CPU buffers; SMT Host state unknown
Vulnerability Meltdown:          Mitigation; PTI
Vulnerability Mmio stale data:   Vulnerable: Clear CPU buffers attempted, no microcode; SMT Host state unknown
Vulnerability Spec store bypass: Vulnerable
Vulnerability Spectre v1:        Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2:        Mitigation; Retpolines, STIBP disabled, RSB filling
Vulnerability Srbds:             Not affected
Vulnerability Tsx async abort:   Mitigation; Clear CPU buffers; SMT Host state unknown
Flags:                           fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ss ht syscall nx pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology cpuid pni pclmulqdq ssse3 fma cx16 pcid sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm abm 3dnowprefetch invpcid_single pti fsgsbase bmi1 hle avx2 smep bmi2 erms invpcid rtm rdseed adx smap xsaveopt md_clear
Cpu Property Value
Brand Intel(R) Xeon(R) CPU E5-2673 v4 @ 2.30GHz
Vendor :Intel
Architecture :Broadwell
Model Family: 0x06, Model: 0x4f, Stepping: 0x01, Type: 0x00
Cores 2 physical cores, 2 logical cores (on executing CPU)
No Hyperthreading hardware capability detected
Clock Frequencies Not supported by CPU
Data Cache Level 1:3 : (32, 256, 51200) kbytes
64 byte cache line size
Address Size 48 bits virtual, 46 bits physical
SIMD 256 bit = 32 byte max. SIMD vector size
Time Stamp Counter TSC is accessible via rdtsc
TSC increased at every clock cycle (non-invariant TSC)
Perf. Monitoring Performance Monitoring Counters (PMC) are not supported
Hypervisor Yes, Microsoft

@adrhill adrhill merged commit 02a4bb2 into master Jun 27, 2022
@adrhill adrhill deleted the ah/vjp branch June 27, 2022 11:48
@adrhill adrhill mentioned this pull request Jun 29, 2022
Closed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
run benchmark Run benchmark CI action
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.

Refactor gradient methods to make use of VJPs
1 participant
@adrhill