What do timems, averageMs and percentage mean and how are they calculated?

[18liumin]

[zerollzeng]

It's layer profile. like this:

[05/08/2024-14:05:14] [I] === Profile (2094 iterations ) ===
[05/08/2024-14:05:14] [I]    Time(ms)     Avg.(ms)   Median(ms)   Time(%)   Layer
[05/08/2024-14:05:14] [I]       16.15       0.0077       0.0075       0.5   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/conv1_1 + node_of_gpu_0/res_conv1_bn_1 + node_of_gpu_0/res_conv1_bn_2
[05/08/2024-14:05:14] [I]       47.82       0.0228       0.0225       1.6   node_of_gpu_0/conv1_1 + node_of_gpu_0/res_conv1_bn_1 + node_of_gpu_0/res_conv1_bn_2
[05/08/2024-14:05:14] [I]       14.02       0.0067       0.0069       0.5   node_of_gpu_0/pool1_1
[05/08/2024-14:05:14] [I]       14.86       0.0071       0.0072       0.5   node_of_gpu_0/res2_0_branch2a_1 + node_of_gpu_0/res2_0_branch2a_bn_1 + node_of_gpu_0/res2_0_branch2a_bn_2
[05/08/2024-14:05:14] [I]       39.67       0.0189       0.0191       1.3   node_of_gpu_0/res2_0_branch2b_1 + node_of_gpu_0/res2_0_branch2b_bn_1 + node_of_gpu_0/res2_0_branch2b_bn_2
[05/08/2024-14:05:14] [I]       22.46       0.0107       0.0105       0.7   node_of_gpu_0/res2_0_branch1_1 + node_of_gpu_0/res2_0_branch1_bn_1
[05/08/2024-14:05:14] [I]       24.87       0.0119       0.0123       0.8   node_of_gpu_0/res2_0_branch2c_1 + node_of_gpu_0/res2_0_branch2c_bn_1 + node_of_gpu_0/res2_0_branch2c_bn_2 + node_of_gpu_0/res2_0_branch2c_bn_3
[05/08/2024-14:05:14] [I]       20.51       0.0098       0.0101       0.7   node_of_gpu_0/res2_1_branch2a_1 + node_of_gpu_0/res2_1_branch2a_bn_1 + node_of_gpu_0/res2_1_branch2a_bn_2
[05/08/2024-14:05:14] [I]       39.29       0.0188       0.0184       1.3   node_of_gpu_0/res2_1_branch2b_1 + node_of_gpu_0/res2_1_branch2b_bn_1 + node_of_gpu_0/res2_1_branch2b_bn_2
[05/08/2024-14:05:14] [I]       23.73       0.0113       0.0113       0.8   node_of_gpu_0/res2_1_branch2c_1 + node_of_gpu_0/res2_1_branch2c_bn_1 + node_of_gpu_0/res2_1_branch2c_bn_2 + node_of_gpu_0/res2_1_branch2c_bn_3
[05/08/2024-14:05:14] [I]       21.41       0.0102       0.0102       0.7   node_of_gpu_0/res2_2_branch2a_1 + node_of_gpu_0/res2_2_branch2a_bn_1 + node_of_gpu_0/res2_2_branch2a_bn_2
[05/08/2024-14:05:14] [I]       38.78       0.0185       0.0184       1.3   node_of_gpu_0/res2_2_branch2b_1 + node_of_gpu_0/res2_2_branch2b_bn_1 + node_of_gpu_0/res2_2_branch2b_bn_2
[05/08/2024-14:05:14] [I]       24.15       0.0115       0.0113       0.8   node_of_gpu_0/res2_2_branch2c_1 + node_of_gpu_0/res2_2_branch2c_bn_1 + node_of_gpu_0/res2_2_branch2c_bn_2 + node_of_gpu_0/res2_2_branch2c_bn_3
[05/08/2024-14:05:14] [I]       26.66       0.0127       0.0124       0.9   node_of_gpu_0/res3_0_branch2a_1 + node_of_gpu_0/res3_0_branch2a_bn_1 + node_of_gpu_0/res3_0_branch2a_bn_2
[05/08/2024-14:05:14] [I]       82.39       0.0393       0.0390       2.7   node_of_gpu_0/res3_0_branch2b_1 + node_of_gpu_0/res3_0_branch2b_bn_1 + node_of_gpu_0/res3_0_branch2b_bn_2
[05/08/2024-14:05:14] [I]       38.64       0.0185       0.0184       1.3   node_of_gpu_0/res3_0_branch1_1 + node_of_gpu_0/res3_0_branch1_bn_1
[05/08/2024-14:05:14] [I]       22.68       0.0108       0.0113       0.7   node_of_gpu_0/res3_0_branch2c_1 + node_of_gpu_0/res3_0_branch2c_bn_1 + node_of_gpu_0/res3_0_branch2c_bn_2 + node_of_gpu_0/res3_0_branch2c_bn_3
[05/08/2024-14:05:14] [I]       22.93       0.0110       0.0113       0.8   node_of_gpu_0/res3_1_branch2a_1 + node_of_gpu_0/res3_1_branch2a_bn_1 + node_of_gpu_0/res3_1_branch2a_bn_2
[05/08/2024-14:05:14] [I]        8.37       0.0040       0.0041       0.3   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/res3_1_branch2b_1 + node_of_gpu_0/res3_1_branch2b_bn_1 + node_of_gpu_0/res3_1_branch2b_bn_2
[05/08/2024-14:05:14] [I]       58.21       0.0278       0.0276       1.9   node_of_gpu_0/res3_1_branch2b_1 + node_of_gpu_0/res3_1_branch2b_bn_1 + node_of_gpu_0/res3_1_branch2b_bn_2
[05/08/2024-14:05:14] [I]        8.40       0.0040       0.0041       0.3   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/res3_1_branch2c_1 + node_of_gpu_0/res3_1_branch2c_bn_1 + node_of_gpu_0/res3_1_branch2c_bn_2 + node_of_gpu_0/res3_1_branch2c_bn_3
[05/08/2024-14:05:14] [I]       22.44       0.0107       0.0105       0.7   node_of_gpu_0/res3_1_branch2c_1 + node_of_gpu_0/res3_1_branch2c_bn_1 + node_of_gpu_0/res3_1_branch2c_bn_2 + node_of_gpu_0/res3_1_branch2c_bn_3
[05/08/2024-14:05:14] [I]       22.44       0.0107       0.0105       0.7   node_of_gpu_0/res3_2_branch2a_1 + node_of_gpu_0/res3_2_branch2a_bn_1 + node_of_gpu_0/res3_2_branch2a_bn_2
[05/08/2024-14:05:14] [I]        8.02       0.0038       0.0041       0.3   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/res3_2_branch2b_1 + node_of_gpu_0/res3_2_branch2b_bn_1 + node_of_gpu_0/res3_2_branch2b_bn_2
[05/08/2024-14:05:14] [I]       57.47       0.0274       0.0276       1.9   node_of_gpu_0/res3_2_branch2b_1 + node_of_gpu_0/res3_2_branch2b_bn_1 + node_of_gpu_0/res3_2_branch2b_bn_2
[05/08/2024-14:05:14] [I]        8.47       0.0040       0.0041       0.3   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/res3_2_branch2c_1 + node_of_gpu_0/res3_2_branch2c_bn_1 + node_of_gpu_0/res3_2_branch2c_bn_2 + node_of_gpu_0/res3_2_branch2c_bn_3
[05/08/2024-14:05:14] [I]       22.79       0.0109       0.0113       0.7   node_of_gpu_0/res3_2_branch2c_1 + node_of_gpu_0/res3_2_branch2c_bn_1 + node_of_gpu_0/res3_2_branch2c_bn_2 + node_of_gpu_0/res3_2_branch2c_bn_3
[05/08/2024-14:05:14] [I]       22.51       0.0108       0.0109       0.7   node_of_gpu_0/res3_3_branch2a_1 + node_of_gpu_0/res3_3_branch2a_bn_1 + node_of_gpu_0/res3_3_branch2a_bn_2
[05/08/2024-14:05:14] [I]        8.03       0.0038       0.0041       0.3   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/res3_3_branch2b_1 + node_of_gpu_0/res3_3_branch2b_bn_1 + node_of_gpu_0/res3_3_branch2b_bn_2
[05/08/2024-14:05:14] [I]       57.45       0.0274       0.0276       1.9   node_of_gpu_0/res3_3_branch2b_1 + node_of_gpu_0/res3_3_branch2b_bn_1 + node_of_gpu_0/res3_3_branch2b_bn_2
[05/08/2024-14:05:14] [I]        8.40       0.0040       0.0041       0.3   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/res3_3_branch2c_1 + node_of_gpu_0/res3_3_branch2c_bn_1 + node_of_gpu_0/res3_3_branch2c_bn_2 + node_of_gpu_0/res3_3_branch2c_bn_3
[05/08/2024-14:05:14] [I]       22.24       0.0106       0.0102       0.7   node_of_gpu_0/res3_3_branch2c_1 + node_of_gpu_0/res3_3_branch2c_bn_1 + node_of_gpu_0/res3_3_branch2c_bn_2 + node_of_gpu_0/res3_3_branch2c_bn_3
[05/08/2024-14:05:14] [I]       28.93       0.0138       0.0136       0.9   node_of_gpu_0/res4_0_branch2a_1 + node_of_gpu_0/res4_0_branch2a_bn_1 + node_of_gpu_0/res4_0_branch2a_bn_2
[05/08/2024-14:05:14] [I]      126.68       0.0605       0.0604       4.1   node_of_gpu_0/res4_0_branch2b_1 + node_of_gpu_0/res4_0_branch2b_bn_1 + node_of_gpu_0/res4_0_branch2b_bn_2
[05/08/2024-14:05:14] [I]       59.32       0.0283       0.0287       1.9   node_of_gpu_0/res4_0_branch1_1 + node_of_gpu_0/res4_0_branch1_bn_1
[05/08/2024-14:05:14] [I]       27.70       0.0132       0.0133       0.9   node_of_gpu_0/res4_0_branch2c_1 + node_of_gpu_0/res4_0_branch2c_bn_1 + node_of_gpu_0/res4_0_branch2c_bn_2 + node_of_gpu_0/res4_0_branch2c_bn_3
[05/08/2024-14:05:14] [I]       26.84       0.0128       0.0131       0.9   node_of_gpu_0/res4_1_branch2a_1 + node_of_gpu_0/res4_1_branch2a_bn_1 + node_of_gpu_0/res4_1_branch2a_bn_2
[05/08/2024-14:05:14] [I]        8.02       0.0038       0.0041       0.3   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/res4_1_branch2b_1 + node_of_gpu_0/res4_1_branch2b_bn_1 + node_of_gpu_0/res4_1_branch2b_bn_2
[05/08/2024-14:05:14] [I]       82.49       0.0394       0.0392       2.7   node_of_gpu_0/res4_1_branch2b_1 + node_of_gpu_0/res4_1_branch2b_bn_1 + node_of_gpu_0/res4_1_branch2b_bn_2
[05/08/2024-14:05:14] [I]        8.08       0.0039       0.0041       0.3   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/res4_1_branch2c_1 + node_of_gpu_0/res4_1_branch2c_bn_1 + node_of_gpu_0/res4_1_branch2c_bn_2 + node_of_gpu_0/res4_1_branch2c_bn_3
[05/08/2024-14:05:14] [I]       26.99       0.0129       0.0132       0.9   node_of_gpu_0/res4_1_branch2c_1 + node_of_gpu_0/res4_1_branch2c_bn_1 + node_of_gpu_0/res4_1_branch2c_bn_2 + node_of_gpu_0/res4_1_branch2c_bn_3
[05/08/2024-14:05:14] [I]       25.67       0.0123       0.0123       0.8   node_of_gpu_0/res4_2_branch2a_1 + node_of_gpu_0/res4_2_branch2a_bn_1 + node_of_gpu_0/res4_2_branch2a_bn_2
[05/08/2024-14:05:14] [I]        7.80       0.0037       0.0041       0.3   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/res4_2_branch2b_1 + node_of_gpu_0/res4_2_branch2b_bn_1 + node_of_gpu_0/res4_2_branch2b_bn_2
[05/08/2024-14:05:14] [I]       82.20       0.0393       0.0389       2.7   node_of_gpu_0/res4_2_branch2b_1 + node_of_gpu_0/res4_2_branch2b_bn_1 + node_of_gpu_0/res4_2_branch2b_bn_2
[05/08/2024-14:05:14] [I]        8.00       0.0038       0.0041       0.3   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/res4_2_branch2c_1 + node_of_gpu_0/res4_2_branch2c_bn_1 + node_of_gpu_0/res4_2_branch2c_bn_2 + node_of_gpu_0/res4_2_branch2c_bn_3
[05/08/2024-14:05:14] [I]       26.40       0.0126       0.0123       0.9   node_of_gpu_0/res4_2_branch2c_1 + node_of_gpu_0/res4_2_branch2c_bn_1 + node_of_gpu_0/res4_2_branch2c_bn_2 + node_of_gpu_0/res4_2_branch2c_bn_3
[05/08/2024-14:05:14] [I]       25.79       0.0123       0.0123       0.8   node_of_gpu_0/res4_3_branch2a_1 + node_of_gpu_0/res4_3_branch2a_bn_1 + node_of_gpu_0/res4_3_branch2a_bn_2
[05/08/2024-14:05:14] [I]        7.66       0.0037       0.0040       0.3   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/res4_3_branch2b_1 + node_of_gpu_0/res4_3_branch2b_bn_1 + node_of_gpu_0/res4_3_branch2b_bn_2
[05/08/2024-14:05:14] [I]       81.69       0.0390       0.0389       2.7   node_of_gpu_0/res4_3_branch2b_1 + node_of_gpu_0/res4_3_branch2b_bn_1 + node_of_gpu_0/res4_3_branch2b_bn_2
[05/08/2024-14:05:14] [I]        8.11       0.0039       0.0041       0.3   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/res4_3_branch2c_1 + node_of_gpu_0/res4_3_branch2c_bn_1 + node_of_gpu_0/res4_3_branch2c_bn_2 + node_of_gpu_0/res4_3_branch2c_bn_3
[05/08/2024-14:05:14] [I]       27.09       0.0129       0.0132       0.9   node_of_gpu_0/res4_3_branch2c_1 + node_of_gpu_0/res4_3_branch2c_bn_1 + node_of_gpu_0/res4_3_branch2c_bn_2 + node_of_gpu_0/res4_3_branch2c_bn_3
[05/08/2024-14:05:14] [I]       25.72       0.0123       0.0123       0.8   node_of_gpu_0/res4_4_branch2a_1 + node_of_gpu_0/res4_4_branch2a_bn_1 + node_of_gpu_0/res4_4_branch2a_bn_2
[05/08/2024-14:05:14] [I]        7.77       0.0037       0.0041       0.3   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/res4_4_branch2b_1 + node_of_gpu_0/res4_4_branch2b_bn_1 + node_of_gpu_0/res4_4_branch2b_bn_2
[05/08/2024-14:05:14] [I]       82.44       0.0394       0.0391       2.7   node_of_gpu_0/res4_4_branch2b_1 + node_of_gpu_0/res4_4_branch2b_bn_1 + node_of_gpu_0/res4_4_branch2b_bn_2
[05/08/2024-14:05:14] [I]        7.98       0.0038       0.0041       0.3   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/res4_4_branch2c_1 + node_of_gpu_0/res4_4_branch2c_bn_1 + node_of_gpu_0/res4_4_branch2c_bn_2 + node_of_gpu_0/res4_4_branch2c_bn_3
[05/08/2024-14:05:14] [I]       28.69       0.0137       0.0133       0.9   node_of_gpu_0/res4_4_branch2c_1 + node_of_gpu_0/res4_4_branch2c_bn_1 + node_of_gpu_0/res4_4_branch2c_bn_2 + node_of_gpu_0/res4_4_branch2c_bn_3
[05/08/2024-14:05:14] [I]       25.81       0.0123       0.0123       0.8   node_of_gpu_0/res4_5_branch2a_1 + node_of_gpu_0/res4_5_branch2a_bn_1 + node_of_gpu_0/res4_5_branch2a_bn_2
[05/08/2024-14:05:14] [I]        7.67       0.0037       0.0041       0.3   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/res4_5_branch2b_1 + node_of_gpu_0/res4_5_branch2b_bn_1 + node_of_gpu_0/res4_5_branch2b_bn_2
[05/08/2024-14:05:14] [I]       84.49       0.0404       0.0400       2.8   node_of_gpu_0/res4_5_branch2b_1 + node_of_gpu_0/res4_5_branch2b_bn_1 + node_of_gpu_0/res4_5_branch2b_bn_2
[05/08/2024-14:05:14] [I]        8.05       0.0038       0.0041       0.3   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/res4_5_branch2c_1 + node_of_gpu_0/res4_5_branch2c_bn_1 + node_of_gpu_0/res4_5_branch2c_bn_2 + node_of_gpu_0/res4_5_branch2c_bn_3
[05/08/2024-14:05:14] [I]       27.10       0.0129       0.0132       0.9   node_of_gpu_0/res4_5_branch2c_1 + node_of_gpu_0/res4_5_branch2c_bn_1 + node_of_gpu_0/res4_5_branch2c_bn_2 + node_of_gpu_0/res4_5_branch2c_bn_3
[05/08/2024-14:05:14] [I]       40.56       0.0194       0.0195       1.3   node_of_gpu_0/res5_0_branch2a_1 + node_of_gpu_0/res5_0_branch2a_bn_1 + node_of_gpu_0/res5_0_branch2a_bn_2
[05/08/2024-14:05:14] [I]      270.12       0.1290       0.1290       8.8   node_of_gpu_0/res5_0_branch2b_1 + node_of_gpu_0/res5_0_branch2b_bn_1 + node_of_gpu_0/res5_0_branch2b_bn_2
[05/08/2024-14:05:14] [I]      114.63       0.0547       0.0545       3.8   node_of_gpu_0/res5_0_branch1_1 + node_of_gpu_0/res5_0_branch1_bn_1
[05/08/2024-14:05:14] [I]       49.27       0.0235       0.0236       1.6   node_of_gpu_0/res5_0_branch2c_1 + node_of_gpu_0/res5_0_branch2c_bn_1 + node_of_gpu_0/res5_0_branch2c_bn_2 + node_of_gpu_0/res5_0_branch2c_bn_3
[05/08/2024-14:05:14] [I]       53.11       0.0254       0.0256       1.7   node_of_gpu_0/res5_1_branch2a_1 + node_of_gpu_0/res5_1_branch2a_bn_1 + node_of_gpu_0/res5_1_branch2a_bn_2
[05/08/2024-14:05:14] [I]        8.11       0.0039       0.0041       0.3   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/res5_1_branch2b_1 + node_of_gpu_0/res5_1_branch2b_bn_1 + node_of_gpu_0/res5_1_branch2b_bn_2
[05/08/2024-14:05:14] [I]      184.84       0.0883       0.0881       6.0   node_of_gpu_0/res5_1_branch2b_1 + node_of_gpu_0/res5_1_branch2b_bn_1 + node_of_gpu_0/res5_1_branch2b_bn_2
[05/08/2024-14:05:14] [I]        8.17       0.0039       0.0041       0.3   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/res5_1_branch2c_1 + node_of_gpu_0/res5_1_branch2c_bn_1 + node_of_gpu_0/res5_1_branch2c_bn_2 + node_of_gpu_0/res5_1_branch2c_bn_3
[05/08/2024-14:05:14] [I]       51.69       0.0247       0.0246       1.7   node_of_gpu_0/res5_1_branch2c_1 + node_of_gpu_0/res5_1_branch2c_bn_1 + node_of_gpu_0/res5_1_branch2c_bn_2 + node_of_gpu_0/res5_1_branch2c_bn_3
[05/08/2024-14:05:14] [I]       52.93       0.0253       0.0256       1.7   node_of_gpu_0/res5_2_branch2a_1 + node_of_gpu_0/res5_2_branch2a_bn_1 + node_of_gpu_0/res5_2_branch2a_bn_2
[05/08/2024-14:05:14] [I]        7.66       0.0037       0.0041       0.3   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/res5_2_branch2b_1 + node_of_gpu_0/res5_2_branch2b_bn_1 + node_of_gpu_0/res5_2_branch2b_bn_2
[05/08/2024-14:05:14] [I]      183.24       0.0875       0.0872       6.0   node_of_gpu_0/res5_2_branch2b_1 + node_of_gpu_0/res5_2_branch2b_bn_1 + node_of_gpu_0/res5_2_branch2b_bn_2
[05/08/2024-14:05:14] [I]        8.08       0.0039       0.0041       0.3   Reformatting CopyNode for Input Tensor 0 to node_of_gpu_0/res5_2_branch2c_1 + node_of_gpu_0/res5_2_branch2c_bn_1 + node_of_gpu_0/res5_2_branch2c_bn_2 + node_of_gpu_0/res5_2_branch2c_bn_3
[05/08/2024-14:05:14] [I]       49.10       0.0234       0.0236       1.6   node_of_gpu_0/res5_2_branch2c_1 + node_of_gpu_0/res5_2_branch2c_bn_1 + node_of_gpu_0/res5_2_branch2c_bn_2 + node_of_gpu_0/res5_2_branch2c_bn_3
[05/08/2024-14:05:14] [I]       11.31       0.0054       0.0051       0.4   node_of_gpu_0/pool5_1
[05/08/2024-14:05:14] [I]       76.69       0.0366       0.0369       2.5   node_of_gpu_0/pred_1 + (Unnamed Layer* 178) [ElementWise]
[05/08/2024-14:05:14] [I]        9.48       0.0045       0.0042       0.3   node_of_gpu_0/softmax_1
[05/08/2024-14:05:14] [I]     3056.46       1.4596       1.4544     100.0   Total
[05/08/2024-14:05:14] [I]
&&&& PASSED TensorRT.trtexec [TensorRT v100001] # trtexec --onnx=ResNet50.onnx --dumpProfile