Skip to content

Instantly share code, notes, and snippets.

@nicolasvasilache

nicolasvasilache/gist:356c1be72370f8d82b0fa6cadd7676b2 Secret

Last active December 2, 2021 10:28
Show Gist options
  • Save nicolasvasilache/356c1be72370f8d82b0fa6cadd7676b2 to your computer and use it in GitHub Desktop.
Save nicolasvasilache/356c1be72370f8d82b0fa6cadd7676b2 to your computer and use it in GitHub Desktop.
Download ZIP
Conv 1d
Raw
gistfile1.txt
> export MLIR_RUNNER_UTILS_LIB=${IREE_LLVM_SANDBOX_BUILD_DIR}/lib/libmlir_runner_utils.so; cd ${IREE_LLVM_SANDBOX_SOURCE_DIR}; python -m python.examples.conv.conv_1d_bench
###############################################################
Compile-time problem size {'N': 8, 'W': 16, 'C': 32, 'KW': 3, 'F': 64, 'strides': [1], 'dilations': [1]}
Runtime problem size {'N': 8, 'W': 16, 'C': 32, 'KW': 3, 'F': 64, 'strides': [1], 'dilations': [1]}
Problem types [<class 'numpy.float32'>, <class 'numpy.float32'>, <class 'numpy.float32'>]
Compilation expert <python.examples.core.transform.TransformationList object at 0x7f2a224394c0>
compilation in 0.1124s
xxxxxxxxxx : 1000 iters time on 1 threads
------------------------------------------------------------------------------------------------------------------------
slowest p1 p10 p25 p50 p75 p90 p99 fastest unit
------------------------------------------------------------------------------------------------------------------------
4.6e-05 3.0e-05 1.4e-05 1.4e-05 1.2e-05 1.2e-05 1.1e-05 1.0e-05 1.0e-05 seconds
34.43 53.13 110.70 110.93 132.66 133.17 149.61 149.94 150.17 GFlops/s
2.38 3.67 7.64 7.66 9.16 9.19 10.32 10.35 10.36 GBs/s
###############################################################
Compile-time problem size {'N': 8, 'W': 16, 'C': 32, 'KW': 3, 'F': 64, 'strides': [1], 'dilations': [2]}
Runtime problem size {'N': 8, 'W': 16, 'C': 32, 'KW': 3, 'F': 64, 'strides': [1], 'dilations': [2]}
Problem types [<class 'numpy.float32'>, <class 'numpy.float32'>, <class 'numpy.float32'>]
Compilation expert <python.examples.core.transform.TransformationList object at 0x7f2a224394c0>
compilation in 0.1205s
xxxxxxxxxx : 1000 iters time on 1 threads
------------------------------------------------------------------------------------------------------------------------
slowest p1 p10 p25 p50 p75 p90 p99 fastest unit
------------------------------------------------------------------------------------------------------------------------
4.0e-05 2.7e-05 1.4e-05 1.3e-05 1.1e-05 1.0e-05 1.0e-05 1.0e-05 1.0e-05 seconds
39.78 57.89 111.30 122.64 149.71 150.25 150.46 150.77 151.00 GFlops/s
2.80 4.07 7.83 8.62 10.53 10.56 10.58 10.60 10.62 GBs/s
###############################################################
Compile-time problem size {'N': 8, 'W': 16, 'C': 32, 'KW': 3, 'F': 64, 'strides': [2], 'dilations': [1]}
Runtime problem size {'N': 8, 'W': 16, 'C': 32, 'KW': 3, 'F': 64, 'strides': [2], 'dilations': [1]}
Problem types [<class 'numpy.float32'>, <class 'numpy.float32'>, <class 'numpy.float32'>]
Compilation expert <python.examples.core.transform.TransformationList object at 0x7f2a224394c0>
compilation in 0.1314s
xxxxxxxxxx : 1000 iters time on 1 threads
------------------------------------------------------------------------------------------------------------------------
slowest p1 p10 p25 p50 p75 p90 p99 fastest unit
------------------------------------------------------------------------------------------------------------------------
4.6e-05 3.3e-05 1.8e-05 1.8e-05 1.5e-05 1.4e-05 1.4e-05 1.4e-05 1.4e-05 seconds
34.19 47.95 85.16 85.66 102.48 115.12 115.60 115.88 116.21 GFlops/s
2.69 3.78 6.71 6.75 8.07 9.07 9.11 9.13 9.15 GBs/s
###############################################################
Compile-time problem size {'N': 8, 'W': 16, 'C': 32, 'KW': 3, 'F': 64, 'strides': [2], 'dilations': [2]}
Runtime problem size {'N': 8, 'W': 16, 'C': 32, 'KW': 3, 'F': 64, 'strides': [2], 'dilations': [2]}
Problem types [<class 'numpy.float32'>, <class 'numpy.float32'>, <class 'numpy.float32'>]
Compilation expert <python.examples.core.transform.TransformationList object at 0x7f2a224394c0>
compilation in 0.1353s
xxxxxxxxxx : 1000 iters time on 1 threads
------------------------------------------------------------------------------------------------------------------------
slowest p1 p10 p25 p50 p75 p90 p99 fastest unit
------------------------------------------------------------------------------------------------------------------------
4.0e-05 3.1e-05 1.8e-05 1.5e-05 1.4e-05 1.4e-05 1.2e-05 1.1e-05 1.1e-05 seconds
39.81 50.28 85.43 101.80 115.02 115.52 136.71 137.14 137.31 GFlops/s
3.19 4.03 6.84 8.15 9.21 9.25 10.95 10.98 11.00 GBs/s
###############################################################
Compile-time problem size {'N': 8, 'W': 16, 'C': 32, 'KW': 3, 'F': 64, 'strides': [2], 'dilations': [3]}
Runtime problem size {'N': 8, 'W': 16, 'C': 32, 'KW': 3, 'F': 64, 'strides': [2], 'dilations': [3]}
Problem types [<class 'numpy.float32'>, <class 'numpy.float32'>, <class 'numpy.float32'>]
Compilation expert <python.examples.core.transform.TransformationList object at 0x7f2a224394c0>
compilation in 0.1348s
xxxxxxxxxx : 1000 iters time on 1 threads
------------------------------------------------------------------------------------------------------------------------
slowest p1 p10 p25 p50 p75 p90 p99 fastest unit
------------------------------------------------------------------------------------------------------------------------
4.0e-05 3.1e-05 1.9e-05 1.9e-05 1.8e-05 1.8e-05 1.7e-05 1.7e-05 1.7e-05 seconds
39.03 51.13 84.03 84.28 88.27 88.69 92.79 93.04 93.28 GFlops/s
3.18 4.16 6.84 6.86 7.18 7.22 7.55 7.57 7.59 GBs/s
###############################################################
Compile-time problem size {'N': 8, 'W': 16, 'C': 32, 'KW': 3, 'F': 64, 'strides': [3], 'dilations': [2]}
Runtime problem size {'N': 8, 'W': 16, 'C': 32, 'KW': 3, 'F': 64, 'strides': [3], 'dilations': [2]}
Problem types [<class 'numpy.float32'>, <class 'numpy.float32'>, <class 'numpy.float32'>]
Compilation expert <python.examples.core.transform.TransformationList object at 0x7f2a224394c0>
compilation in 0.1359s
xxxxxxxxxx : 1000 iters time on 1 threads
------------------------------------------------------------------------------------------------------------------------
slowest p1 p10 p25 p50 p75 p90 p99 fastest unit
------------------------------------------------------------------------------------------------------------------------
4.8e-05 3.2e-05 1.9e-05 1.9e-05 1.6e-05 1.4e-05 1.4e-05 1.4e-05 1.4e-05 seconds
32.60 48.63 83.88 84.24 100.89 113.47 113.67 113.85 114.12 GFlops/s
2.93 4.37 7.54 7.57 9.06 10.19 10.21 10.23 10.25 GBs/s
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment