bddppq

---------- onnx coverage: ----------
Operators (passed/loaded/total): 21/21/70
------------------------------------
╒════════════════════╤════════════════════╕
│ Operator           │ Attributes         │
│                    │ (name: #values)    │
╞════════════════════╪════════════════════╡
│ Slice              │ axes: 2            │
│                    │ ends: 3            │
│                    │ starts: 3          │

bddppq

from argparse import ArgumentParser
import os
from timeit import Timer

import numpy as np
from caffe2.python import workspace
import onnx
from onnx.numpy_helper import to_array
import onnx_caffe2.backend
import torch

bddppq

jbai-mbp:/tmp/onnx-benchmark[1][2]$ python bench_caffe2.py --model_name resnet18 --batch_sizes 1 2 8 16 --runs 10
WARNING:root:This caffe2 python run does not have GPU support. Will run in CPU only mode.
WARNING:root:Debug message: No module named caffe2_pybind11_state_gpu
Batch Size=1
WARNING: Logging before InitGoogleLogging() is written to STDERR
I0925 20:12:10.023396 2958132160 net_simple.cc:81] Starting benchmark.
I0925 20:12:10.023424 2958132160 net_simple.cc:82] Running warmup runs.
I0925 20:12:10.221287 2958132160 net_simple.cc:92] Main runs.
I0925 20:12:11.880851 2958132160 net_simple.cc:103] Main run finished. Milliseconds per iter: 165.951. Iters per second: 6.02589
I0925 20:12:13.568706 2958132160 net_simple.cc:152] Operator #0 (105, Conv) 11.2651 ms/iter