Module | Self CPU total | CPU total | CUDA total
-----------------------|----------------|-----------|-----------
EDVR_sr | | |
├── pre_deblur | | |
│├── conv_first | 1.612ms | 6.377ms | 59.683ms
│├── RB_L1_1 | | |
││├── conv1 | 88.769us | 309.065us | 165.163ms
││└── conv2 | 4.162ms | 16.600ms | 190.874ms
│├── RB_L1_2 | | |
││├── conv1 | 89.180us | 313.272us | 176.436ms
││└── conv2 | 6.353ms | 25.352ms | 199.747ms
│├── RB_L1_3 | | |
││├── conv1 | 134.412us | 467.793us | 173.243ms
││└── conv2 | 175.594us | 631.884us | 169.989ms
│├── RB_L1_4 | | |
││├── conv1 | 161.295us | 576.461us | 175.251ms
││└── conv2 | 146.771us | 524.259us | 175.441ms
│├── RB_L1_5 | | |
││├── conv1 | 142.661us | 509.228us | 175.278ms
││└── conv2 | 160.888us | 581.309us | 172.092ms
│├── RB_L2_1 | | |
││├── conv1 | 112.691us | 396.596us | 41.064ms
││└── conv2 | 137.734us | 495.700us | 40.801ms
│├── RB_L2_2 | | |
││├── conv1 | 103.887us | 367.672us | 40.584ms
││└── conv2 | 104.958us | 375.029us | 41.261ms
│├── RB_L3_1 | | |
││├── conv1 | 104.348us | 367.864us | 10.383ms
││└── conv2 | 185.083us | 669.036us | 10.590ms
│├── deblur_L2_conv | 108.267us | 382.931us | 60.463ms
│├── deblur_L3_conv | 1.899ms | 7.547ms | 22.711ms
│└── lrelu | 68.955us | 68.955us | 9.002ms
├── conv_1x1 | 168.538us | 605.097us | 72.055ms
├── feature_extraction | | |
│├── 0 | | |
││├── conv1 | 124.924us | 446.754us | 173.249ms
││└── conv2 | 173.828us | 626.203us | 175.450ms
│├── 1 | | |
││├── conv1 | 158.281us | 566.686us | 174.546ms
││└── conv2 | 125.162us | 451.988us | 175.009ms
│├── 2 | | |
││├── conv1 | 96.475us | 340.429us | 176.137ms
││└── conv2 | 84.347us | 295.080us | 175.810ms
│├── 3 | | |
││├── conv1 | 92.698us | 324.680us | 174.312ms
││└── conv2 | 166.583us | 599.656us | 173.517ms
│├── 4 | | |
││├── conv1 | 166.335us | 603.345us | 175.379ms
││└── conv2 | 145.906us | 526.894us | 175.359ms
├── fea_L2_conv1 | 169.420us | 614.830us | 64.473ms
├── fea_L2_conv2 | 163.854us | 594.608us | 43.207ms
├── fea_L3_conv1 | 147.875us | 533.039us | 16.817ms
├── fea_L3_conv2 | 147.033us | 533.180us | 11.032ms
├── pcd_align | | |
│├── L3_offset_conv1 | 940.072us | 3.313ms | 20.542ms
│├── L3_offset_conv2 | 600.781us | 2.065ms | 11.854ms
│├── L3_dcnpack | | |
││└── conv_offset_mask | 550.628us | 1.882ms | 34.724ms
│├── L2_offset_conv1 | 632.229us | 2.225ms | 73.263ms
│├── L2_offset_conv2 | 967.579us | 3.602ms | 74.843ms
│├── L2_offset_conv3 | 553.067us | 1.952ms | 42.602ms
│├── L2_dcnpack | | |
││└── conv_offset_mask | 472.621us | 1.653ms | 133.681ms
│├── L2_fea_conv | 567.537us | 1.996ms | 72.705ms
│├── L1_offset_conv1 | 639.844us | 2.261ms | 285.048ms
│├── L1_offset_conv2 | 4.343ms | 5.994ms | 286.848ms
│├── L1_offset_conv3 | 660.437us | 2.334ms | 164.148ms
│├── L1_dcnpack | | |
││└── conv_offset_mask | 757.924us | 2.694ms | 536.364ms
│├── L1_fea_conv | 1.246ms | 4.397ms | 283.717ms
│├── cas_offset_conv1 | 724.597us | 2.553ms | 283.504ms
│├── cas_offset_conv2 | 600.480us | 2.118ms | 164.786ms
│├── cas_dcnpack | | |
││└── conv_offset_mask | 644.824us | 2.239ms | 538.357ms
│└── lrelu | 1.801ms | 1.801ms | 50.527ms
├── tsa_fusion | | |
│├── tAtt_1 | 94.606us | 332.538us | 171.295ms
│├── tAtt_2 | 106.808us | 378.503us | 33.153ms
│├── fea_fusion | 179.067us | 646.294us | 31.584ms
│├── sAtt_1 | 144.178us | 515.133us | 31.732ms
│├── maxpool | 98.677us | 173.140us | 3.297ms
│├── avgpool | 51.820us | 51.820us | 1.520ms
│├── sAtt_2 | 143.973us | 513.542us | 4.779ms
│├── sAtt_3 | 86.342us | 301.920us | 8.692ms
│├── sAtt_4 | 92.452us | 323.349us | 3.568ms
│├── sAtt_5 | 86.526us | 300.206us | 34.319ms
│├── sAtt_L1 | 100.903us | 356.942us | 3.549ms
│├── sAtt_L2 | 121.861us | 441.587us | 4.061ms
│├── sAtt_L3 | 100.693us | 356.157us | 2.414ms
│├── sAtt_add_1 | 90.520us | 315.218us | 13.022ms
│├── sAtt_add_2 | 88.519us | 307.391us | 13.140ms
│└── lrelu | 206.242us | 206.242us | 5.810ms
├── recon_trunk | | |
│├── 0 | | |
││├── conv1 | 87.998us | 308.996us | 33.913ms
││└── conv2 | 121.220us | 432.030us | 34.123ms
│├── 1 | | |
││├── conv***us | 346.483us | 33.972ms
││└── conv2 | 87.546us | 305.612us | 34.074ms
│├── 2 | | |
││├── conv1 | 89.788us | 316.117us | 33.768ms
││└── conv2 | 92.610us | 312.827us | 32.943ms
│├── 3 | | |
││├── conv1 | 85.813us | 299.170us | 32.873ms
││└── conv2 | 82.799us | 287.774us | 33.041ms
│├── 4 | | |
││├── conv1 | 80.457us | 280.554us | 32.980ms
││└── conv2 | 83.520us | 289.328us | 33.201ms
│├── 5 | | |
││├── conv1 | 81.458us | 282.397us | 33.177ms
││└── conv2 | 82.222us | 285.462us | 32.816ms
│├── 6 | | |
││├── conv1 | 83.040us | 285.797us | 33.029ms
││└── conv2 | 83.232us | 287.943us | 33.035ms
│├── 7 | | |
││├── conv1 | 88.126us | 291.320us | 33.111ms
││└── conv2 | 170.708us | 616.201us | 33.180ms
│├── 8 | | |
││├── conv1 | 127.680us | 452.590us | 33.324ms
││└── conv2 | 122.124us | 431.352us | 33.315ms
│└── 9 | | |
│ ├── conv1 | 110.738us | 386.950us | 33.013ms
│ └── conv2 | 104.227us | 364.880us | 32.997ms
├── upconv1 | 110.350us | 389.328us | 124.809ms
├── upconv2 | 196.124us | 714.129us | 127.092ms
├── upconv3 | 152.818us | 549.402us | 105.691ms
├── pixel_shuffle | 151.018us | 373.672us | 17.563ms
├── HRconv | 189.777us | 694.602us | 138.151ms
├── conv_last | 184.995us | 665.641us | 56.630ms
└── lrelu | 295.875us | 295.875us | 22.237ms
Module | Self CPU total | CPU total | CUDA total
------------------------------|----------------|-----------|-----------
EDVR_sr | | |
├── pre_deblur | | |
│├── conv_first | | |
││├── conv2d | 9.318us | 156.261us | 13.230ms
││├── convolution | 8.784us | 146.943us | 13.220ms
││├── _convolution | 22.409us | 138.159us | 13.211ms
││├── contiguous | 8.654us | 8.654us | 8.192us
││└── cudnn_convolution | 107.096us | 107.096us | 13.180ms
│├── RB_L1_1 | | |
││├── conv1 | | |
│││├── conv2d | 21.857us | 120.330us | 42.443ms
│││├── convolution | 6.205us | 98.473us | 42.425ms
│││├── _convolution | 14.580us | 92.268us | 42.418ms
│││├── contiguous | 3.428us | 3.428us | 3.072us
│││└── cudnn_convolution | 74.260us | 74.260us | 42.400ms
││├── conv2 | | |
│││├── conv2d | 10.033us | 6.635ms | 49.678ms
│││├── convolution | 9.170us | 6.625ms | 49.668ms
│││├── _convolution | 24.143us | 6.616ms | 49.660ms
│││├── contiguous | 4.954us | 4.954us | 6.144us
│││└── cudnn_convolution | 6.587ms | 6.587ms | 49.632ms
│├── RB_L1_2 | | |
││├── conv1 | | |
│││├── conv2d | 7.148us | 101.142us | 42.362ms
│││├── convolution | 6.885us | 93.994us | 42.355ms
│││├── _convolution | 16.903us | 87.109us | 42.349ms
│││├── contiguous | 3.717us | 3.717us | 3.072us
│││└── cudnn_convolution | 66.489us | 66.489us | 42.328ms
││├── conv2 | | |
│││├── conv2d | 8.024us | 14.657ms | 57.896ms
│││├── convolution | 7.941us | 14.649ms | 57.888ms
│││├── _convolution | 19.686us | 14.641ms | 57.880ms
│││├── contiguous | 4.343us | 4.343us | 4.096us
│││└── cudnn_convolution | 14.617ms | 14.617ms | 57.857ms
│├── RB_L1_3 | | |
││├── conv1 | | |
│││├── conv2d | 8.210us | 107.233us | 43.521ms
│││├── convolution | 7.353us | 99.023us | 43.513ms
│││├── _convolution | 19.626us | 91.670us | 43.507ms
│││├── contiguous | 4.056us | 4.056us | 3.072us
│││└── cudnn_convolution | 67.988us | 67.988us | 43.483ms
││├── conv2 | | |
│││├── conv2d | 6.195us | 94.322us | 42.967ms
│││├── convolution | 6.277us | 88.127us | 42.960ms
│││├── _convolution | 15.596us | 81.850us | 42.953ms
│││├── contiguous | 2.981us | 2.981us | 3.072us
│││└── cudnn_convolution | 63.273us | 63.273us | 42.934ms
│├── RB_L1_4 | | |
││├── conv1 | | |
│││├── conv2d | 6.511us | 90.633us | 43.143ms
│││├── convolution | 6.045us | 84.122us | 43.135ms
│││├── _convolution | 17.579us | 78.077us | 43.127ms
│││├── contiguous | 3.256us | 3.256us | 4.096us
│││└── cudnn_convolution | 57.242us | 57.242us | 43.108ms
││├── conv2 | | |
│││├── conv2d | 6.277us | 103.539us | 42.957ms
│││├── convolution | 6.154us | 97.262us | 42.950ms
│││├── _convolution | 14.865us | 91.108us | 42.942ms
│││├── contiguous | 5.784us | 5.784us | 7.168us
│││└── cudnn_convolution | 70.459us | 70.459us | 42.922ms
│├── RB_L1_5 | | |
││├── conv1 | | |
│││├── conv2d | 8.939us | 158.462us | 43.955ms
│││├── convolution | 7.693us | 149.523us | 43.946ms
│││├── _convolution | 21.187us | 141.830us | 43.937ms
│││├── contiguous | 8.700us | 8.700us | 8.192us
│││└── cudnn_convolution | 111.943us | 111.943us | 43.908ms
││├── conv2 | | |
│││├── conv2d | 14.083us | 222.727us | 43.662ms
│││├── convolution | 14.233us | 208.644us | 43.649ms
│││├── _convolution | 31.437us | 194.411us | 43.638ms
│││├── contiguous | 12.729us | 12.729us | 13.312us
│││└── cudnn_convolution | 150.245us | 150.245us | 43.598ms
│├── RB_L2_1 | | |
││├── conv1 | | |
│││├── conv2d | 19.361us | 110.939us | 10.327ms
│││├── convolution | 6.369us | 91.578us | 10.312ms
│││├── _convolution | 14.741us | 85.209us | 10.305ms
│││├── contiguous | 3.037us | 3.037us | 3.072us
│││└── cudnn_convolution | 67.431us | 67.431us | 10.286ms
││├── conv2 | | |
│││├── conv2d | 36.544us | 146.692us | 10.204ms
│││├── convolution | 6.505us | 110.148us | 10.189ms
│││├── _convolution | 15.300us | 103.643us | 10.182ms
│││├── contiguous | 3.193us | 3.193us | 3.072us
│││└── cudnn_convolution | 85.150us | 85.150us | 10.163ms
│├── RB_L2_2 | | |
││├── conv1 | | |
│││├── conv2d | 24.151us | 143.406us | 10.268ms
│││├── convolution | 6.172us | 119.255us | 10.249ms
│││├── _convolution | 24.430us | 113.083us | 10.241ms
│││├── contiguous | 6.131us | 6.131us | 6.144us
│││└── cudnn_convolution | 82.522us | 82.522us | 10.210ms
││├── conv2 | | |
│││├── conv2d | 27.964us | 137.110us | 10.291ms
│││├── convolution | 6.401us | 109.146us | 10.268ms
│││├── _convolution | 15.299us | 102.745us | 10.260ms
│││├── contiguous | 3.448us | 3.448us | 3.072us
│││└── cudnn_convolution | 83.998us | 83.998us | 10.242ms
│├── RB_L3_1 | | |
││├── conv1 | | |
│││├── conv2d | 6.981us | 158.955us | 2.645ms
│││├── convolution | 13.649us | 151.974us | 2.637ms
│││├── _convolution | 20.894us | 138.325us | 2.622ms
│││├── contiguous | 4.026us | 4.026us | 4.096us
│││└── cudnn_convolution | 113.405us | 113.405us | 2.597ms
││├── conv2 | | |
│││├── conv2d | 24.257us | 126.080us | 2.600ms
│││├── convolution | 6.277us | 101.823us | 2.586ms
│││├── _convolution | 20.825us | 95.546us | 2.579ms
│││├── contiguous | 3.460us | 3.460us | 4.096us
│││└── cudnn_convolution | 71.261us | 71.261us | 2.554ms
│├── deblur_L2_conv | | |
││├── conv2d | 6.698us | 103.381us | 15.064ms
││├── convolution | 5.844us | 96.683us | 15.057ms
││├── _convolution | 14.993us | 90.839us | 15.049ms
││├── contiguous | 3.638us | 3.638us | 3.072us
││└── cudnn_convolution | 72.208us | 72.208us | 15.031ms
│├── deblur_L3_conv | | |
││├── conv2d | 8.108us | 2.156ms | 5.816ms
││├── convolution | 8.828us | 2.148ms | 5.807ms
││├── _convolution | 23.828us | 2.139ms | 5.799ms
││├── contiguous | 3.841us | 3.841us | 4.096us
││└── cudnn_convolution | 2.112ms | 2.112ms | 5.773ms
│├── lrelu | | |
││└── leaky_relu_ | 37.138us | 37.138us | 460.800us
├── conv_1x1 | | |
│├── conv2d | 9.036us | 196.808us | 17.956ms
│├── convolution | 7.938us | 187.772us | 17.947ms
│├── _convolution | 23.807us | 179.834us | 17.937ms
│├── contiguous | 3.854us | 3.854us | 4.096us
│└── cudnn_convolution | 152.173us | 152.173us | 17.911ms
├── feature_extraction | | |
│├── 0 | | |
││├── conv1 | | |
│││├── conv2d | 11.376us | 215.781us | 43.786ms
│││├── convolution | 10.447us | 204.405us | 43.775ms
│││├── _convolution | 25.849us | 193.958us | 43.766ms
│││├── contiguous | 4.750us | 4.750us | 5.120us
│││└── cudnn_convolution | 163.359us | 163.359us | 43.737ms
││├── conv2 | | |
│││├── conv2d | 7.644us | 122.787us | 44.091ms
│││├── convolution | 7.517us | 115.143us | 44.085ms
│││├── _convolution | 19.648us | 107.626us | 44.078ms
│││├── contiguous | 4.384us | 4.384us | 5.120us
│││└── cudnn_convolution | 83.594us | 83.594us | 44.056ms
│├── 1 | | |
││├── conv1 | | |
│││├── conv2d | 6.314us | 96.065us | 44.264ms
│││├── convolution | 6.074us | 89.751us | 44.257ms
│││├── _convolution | 16.637us | 83.677us | 44.251ms
│││├── contiguous | 3.073us | 3.073us | 3.072us
│││└── cudnn_convolution | 63.967us | 63.967us | 44.231ms
││├── conv2 | | |
│││├── conv2d | 6.772us | 111.351us | 44.092ms
│││├── convolution | 6.846us | 104.579us | 44.084ms
│││├── _convolution | 14.405us | 97.733us | 44.077ms
│││├── contiguous | 5.208us | 5.208us | 5.120us
│││└── cudnn_convolution | 78.120us | 78.120us | 44.058ms
│├── 2 | | |
││├── conv1 | | |
│││├── conv2d | 7.708us | 157.944us | 43.433ms
│││├── convolution | 9.167us | 150.236us | 43.424ms
│││├── _convolution | 21.579us | 141.069us | 43.414ms
│││├── contiguous | 3.633us | 3.633us | 3.072us
│││└── cudnn_convolution | 115.857us | 115.857us | 43.389ms
││├── conv2 | | |
│││├── conv2d | 9.008us | 179.216us | 43.230ms
│││├── convolution | 7.889us | 170.208us | 43.221ms
│││├── _convolution | 21.151us | 162.319us | 43.213ms
│││├── contiguous | 9.768us | 9.768us | 9.216us
│││└── cudnn_convolution | 131.400us | 131.400us | 43.182ms
│├── 3 | | |
││├── conv1 | | |
│││├── conv2d | 7.634us | 136.037us | 43.597ms
│││├── convolution | 7.679us | 128.403us | 43.588ms
│││├── _convolution | 22.238us | 120.724us | 43.579ms
│││├── contiguous | 3.532us | 3.532us | 4.096us
│││└── cudnn_convolution | 94.954us | 94.954us | 43.554ms
││├── conv2 | | |
│││├── conv2d | 6.669us | 117.395us | 43.105ms
│││├── convolution | 7.096us | 110.726us | 43.098ms
│││├── _convolution | 17.094us | 103.630us | 43.090ms
│││├── contiguous | 6.015us | 6.015us | 6.144us
│││└── cudnn_convolution | 80.521us | 80.521us | 43.067ms
│├── 4 | | |
││├── conv1 | | |
│││├── conv2d | 7.145us | 123.248us | 43.091ms
│││├── convolution | 7.850us | 116.103us | 43.082ms
│││├── _convolution | 21.869us | 108.253us | 43.073ms
│││├── contiguous | 3.383us | 3.383us | 4.096us
│││└── cudnn_convolution | 83.001us | 83.001us | 43.048ms
││├── conv2 | | |
│││├── conv2d | 8.278us | 138.504us | 42.950ms
│││├── convolution | 7.706us | 130.226us | 42.940ms
│││├── _convolution | 18.838us | 122.520us | 42.932ms
│││├── contiguous | 7.850us | 7.850us | 8.192us
│││└── cudnn_convolution | 95.832us | 95.832us | 42.907ms
├── fea_L2_conv1 | | |
│├── conv2d | 8.490us | 166.028us | 15.597ms
│├── convolution | 7.756us | 157.538us | 15.587ms
│├── _convolution | 26.123us | 149.782us | 15.578ms
│├── contiguous | 3.878us | 3.878us | 4.096us
│└── cudnn_convolution | 119.781us | 119.781us | 15.549ms
├── fea_L2_conv2 | | |
│├── conv2d | 7.226us | 151.246us | 10.659ms
│├── convolution | 8.174us | 144.020us | 10.652ms
│├── _convolution | 19.413us | 135.846us | 10.641ms
│├── contiguous | 4.555us | 4.555us | 6.144us
│└── cudnn_convolution | 111.878us | 111.878us | 10.617ms
├── fea_L3_conv1 | | |
│├── conv2d | 8.003us | 150.190us | 4.018ms
│├── convolution | 7.317us | 142.187us | 4.009ms
│├── _convolution | 21.456us | 134.870us | 4.002ms
│├── contiguous | 4.265us | 4.265us | 4.096us
│└── cudnn_convolution | 109.149us | 109.149us | 3.977ms
├── fea_L3_conv2 | | |
│├── conv2d | 7.713us | 140.694us | 2.675ms
│├── convolution | 10.297us | 132.981us | 2.666ms
│├── _convolution | 19.058us | 122.684us | 2.654ms
│├── contiguous | 3.232us | 3.232us | 4.096us
│└── cudnn_convolution | 100.394us | 100.394us | 2.632ms
├── pcd_align | | |
│├── L3_offset_conv1 | | |
││├── conv2d | 6.055us | 94.496us | 962.560us
││├── convolution | 6.224us | 88.441us | 955.392us
││├── _convolution | 18.030us | 82.217us | 947.200us
││├── contiguous | 3.334us | 3.334us | 4.096us
││└── cudnn_convolution | 60.853us | 60.853us | 927.744us
│├── L3_offset_conv2 | | |
││├── conv2d | 6.211us | 88.924us | 582.656us
││├── convolution | 5.781us | 82.713us | 576.512us
││├── _convolution | 15.500us | 76.932us | 568.320us
││├── contiguous | 2.787us | 2.787us | 3.072us
││└── cudnn_convolution | 58.645us | 58.645us | 551.936us
│├── L3_dcnpack | | |
││├── conv_offset_mask | | |
│││├── conv2d | 5.954us | 86.833us | 1.701ms
│││├── convolution | 6.067us | 80.879us | 1.694ms
│││├── _convolution | 13.366us | 74.812us | 1.687ms
│││├── contiguous | 6.001us | 6.001us | 4.096us
│││└── cudnn_convolution | 55.445us | 55.445us | 1.668ms
│├── L2_offset_conv1 | | |
││├── conv2d | 10.050us | 95.027us | 3.595ms
││├── convolution | 9.453us | 84.977us | 3.584ms
││├── _convolution | 13.454us | 75.524us | 3.576ms
││├── contiguous | 3.325us | 3.325us | 3.072us
││└── cudnn_convolution | 58.745us | 58.745us | 3.558ms
│├── L2_offset_conv2 | | |
││├── conv2d | 6.491us | 89.329us | 3.588ms
││├── convolution | 6.344us | 82.838us | 3.581ms
││├── _convolution | 16.443us | 76.494us | 3.573ms
││├── contiguous | 3.284us | 3.284us | 3.072us
││└── cudnn_convolution | 56.767us | 56.767us | 3.559ms
│├── L2_offset_conv3 | | |
││├── conv2d | 6.647us | 92.284us | 2.098ms
││├── convolution | 6.505us | 85.637us | 2.091ms
││├── _convolution | 15.973us | 79.132us | 2.084ms
││├── contiguous | 2.976us | 2.976us | 4.096us
││└── cudnn_convolution | 60.183us | 60.183us | 2.067ms
│├── L2_dcnpack | | |
││├── conv_offset_mask | | |
│││├── conv2d | 5.967us | 88.569us | 6.537ms
│││├── convolution | 5.856us | 82.602us | 6.531ms
│││├── _convolution | 16.713us | 76.746us | 6.524ms
│││├── contiguous | 3.150us | 3.150us | 4.096us
│││└── cudnn_convolution | 56.883us | 56.883us | 6.506ms
│├── L2_fea_conv | | |
││├── conv2d | 6.719us | 92.066us | 3.618ms
││├── convolution | 6.079us | 85.347us | 3.611ms
││├── _convolution | 13.478us | 79.268us | 3.602ms
││├── contiguous | 7.669us | 7.669us | 6.144us
││└── cudnn_convolution | 58.121us | 58.121us | 3.583ms
│├── L1_offset_conv1 | | |
││├── conv2d | 7.217us | 100.946us | 14.104ms
││├── convolution | 5.827us | 93.729us | 14.096ms
││├── _convolution | 15.532us | 87.902us | 14.090ms
││├── contiguous | 5.686us | 5.686us | 6.144us
││└── cudnn_convolution | 66.684us | 66.684us | 14.070ms
│├── L1_offset_conv2 | | |
││├── conv2d | 6.488us | 101.949us | 14.048ms
││├── convolution | 5.877us | 95.461us | 14.042ms
││├── _convolution | 20.203us | 89.584us | 14.034ms
││├── contiguous | 3.263us | 3.263us | 4.096us
││└── cudnn_convolution | 66.118us | 66.118us | 14.011ms
│├── L1_offset_conv3 | | |
││├── conv2d | 6.203us | 102.860us | 8.202ms
││├── convolution | 5.782us | 96.657us | 8.194ms
││├── _convolution | 20.732us | 90.875us | 8.188ms
││├── contiguous | 3.199us | 3.199us | 3.072us
││└── cudnn_convolution | 66.944us | 66.944us | 8.163ms
│├── L1_dcnpack | | |
││├── conv_offset_mask | | |
│││├── conv2d | 6.664us | 126.178us | 26.949ms
│││├── convolution | 12.563us | 119.514us | 26.941ms
│││├── _convolution | 18.239us | 106.951us | 26.928ms
│││├── contiguous | 3.435us | 3.435us | 3.072us
│││└── cudnn_convolution | 85.277us | 85.277us | 26.908ms
│├── L1_fea_conv | | |
││├── conv2d | 12.681us | 231.611us | 14.135ms
││├── convolution | 8.732us | 218.930us | 14.122ms
││├── _convolution | 35.428us | 210.198us | 14.112ms
││├── contiguous | 6.409us | 6.409us | 6.144us
││└── cudnn_convolution | 168.361us | 168.361us | 14.077ms
│├── cas_offset_conv1 | | |
││├── conv2d | 10.470us | 207.881us | 14.045ms
││├── convolution | 12.908us | 197.411us | 14.036ms
││├── _convolution | 27.568us | 184.503us | 14.026ms
││├── contiguous | 4.611us | 4.611us | 5.120us
││└── cudnn_convolution | 152.324us | 152.324us | 13.995ms
│├── cas_offset_conv2 | | |
││├── conv2d | 9.517us | 167.142us | 8.250ms
││├── convolution | 9.720us | 157.625us | 8.242ms
││├── _convolution | 26.968us | 147.905us | 8.232ms
││├── contiguous | 3.883us | 3.883us | 5.120us
││└── cudnn_convolution | 117.054us | 117.054us | 8.206ms
│├── cas_dcnpack | | |
││├── conv_offset_mask | | |
│││├── conv2d | 7.952us | 144.740us | 26.983ms
│││├── convolution | 8.212us | 136.788us | 26.975ms
│││├── _convolution | 24.998us | 128.576us | 26.967ms
│││├── contiguous | 4.024us | 4.024us | 5.120us
│││└── cudnn_convolution | 99.554us | 99.554us | 26.942ms
│├── lrelu | | |
││└── leaky_relu_ | 26.338us | 26.338us | 1.383ms
├── tsa_fusion | | |
│├── tAtt_1 | | |
││├── conv2d | 7.405us | 131.884us | 41.526ms
││├── convolution | 9.148us | 124.479us | 41.519ms
││├── _convolution | 23.505us | 115.331us | 41.510ms
││├── contiguous | 4.219us | 4.219us | 4.096us
││└── cudnn_convolution | 87.607us | 87.607us | 41.483ms
│├── tAtt_2 | | |
││├── conv2d | 10.400us | 196.848us | 8.344ms
││├── convolution | 8.956us | 186.448us | 8.333ms
││├── _convolution | 40.433us | 177.492us | 8.325ms
││├── contiguous | 5.728us | 5.728us | 5.120us
││└── cudnn_convolution | 131.331us | 131.331us | 8.288ms
│├── fea_fusion | | |
││├── conv2d | 7.499us | 160.805us | 7.869ms
││├── convolution | 7.077us | 153.306us | 7.861ms
││├── _convolution | 20.843us | 146.229us | 7.853ms
││├── contiguous | 7.063us | 7.063us | 8.192us
││└── cudnn_convolution | 118.323us | 118.323us | 7.826ms
│├── sAtt_1 | | |
││├── conv2d | 6.553us | 93.752us | 7.867ms
││├── convolution | 5.817us | 87.199us | 7.861ms
││├── _convolution | 17.110us | 81.382us | 7.855ms
││├── contiguous | 3.151us | 3.151us | 3.072us
││└── cudnn_convolution | 61.121us | 61.121us | 7.835ms
│├── maxpool | | |
││├── max_pool2d | 10.519us | 47.668us | 362.496us
││└── max_pool2d_with_indices | 37.149us | 37.149us | 354.304us
│├── avgpool | | |
││└── avg_pool2d | 21.864us | 21.864us | 323.584us
│├── sAtt_2 | | |
││├── conv2d | 6.221us | 107.818us | 1.165ms
││├── convolution | 6.039us | 101.597us | 1.159ms
││├── _convolution | 16.297us | 95.558us | 1.152ms
││├── contiguous | 5.670us | 5.670us | 5.120us
││└── cudnn_convolution | 73.591us | 73.591us | 1.129ms
│├── sAtt_3 | | |
││├── conv2d | 6.515us | 89.578us | 2.175ms
││├── convolution | 5.974us | 83.063us | 2.169ms
││├── _convolution | 16.266us | 77.089us | 2.163ms
││├── contiguous | 2.815us | 2.815us | 3.072us
││└── cudnn_convolution | 58.008us | 58.008us | 2.144ms
│├── sAtt_4 | | |
││├── conv2d | 6.043us | 98.866us | 896.000us
││├── convolution | 6.026us | 92.823us | 888.832us
││├── _convolution | 13.996us | 86.797us | 882.688us
││├── contiguous | 5.669us | 5.669us | 6.144us
││└── cudnn_convolution | 67.132us | 67.132us | 863.232us
│├── sAtt_5 | | |
││├── conv2d | 6.185us | 99.557us | 8.245ms
││├── convolution | 6.185us | 93.372us | 8.238ms
││├── _convolution | 14.620us | 87.187us | 8.231ms
││├── contiguous | 6.909us | 6.909us | 6.144us
││└── cudnn_convolution | 65.658us | 65.658us | 8.211ms
│├── sAtt_L1 | | |
││├── conv2d | 6.162us | 99.165us | 906.240us
││├── convolution | 5.806us | 93.003us | 900.096us
││├── _convolution | 19.679us | 87.197us | 891.904us
││├── contiguous | 3.147us | 3.147us | 4.096us
││└── cudnn_convolution | 64.371us | 64.371us | 870.400us
│├── sAtt_L2 | | |
││├── conv2d | 6.104us | 119.743us | 1.028ms
││├── convolution | 6.470us | 113.639us | 1.022ms
││├── _convolution | 15.277us | 107.169us | 1.014ms
││├── contiguous | 3.093us | 3.093us | 3.072us
││└── cudnn_convolution | 88.799us | 88.799us | 996.352us
│├── sAtt_L3 | | |
││├── conv2d | 6.170us | 91.031us | 604.160us
││├── convolution | 6.037us | 84.861us | 598.016us
││├── _convolution | 12.947us | 78.824us | 590.848us
││├── contiguous | 2.900us | 2.900us | 3.072us
││└── cudnn_convolution | 62.977us | 62.977us | 575.488us
│├── sAtt_add_1 | | |
││├── conv2d | 6.262us | 97.919us | 3.289ms
││├── convolution | 5.596us | 91.657us | 3.282ms
││├── _convolution | 16.465us | 86.061us | 3.276ms
││├── contiguous | 3.281us | 3.281us | 3.072us
││└── cudnn_convolution | 66.315us | 66.315us | 3.256ms
│├── sAtt_add_2 | | |
││├── conv2d | 8.874us | 90.650us | 3.220ms
││├── convolution | 8.176us | 81.776us | 3.213ms
││├── _convolution | 12.985us | 73.600us | 3.204ms
││├── contiguous | 2.861us | 2.861us | 3.072us
││└── cudnn_convolution | 57.754us | 57.754us | 3.189ms
│├── lrelu | | |
││└── leaky_relu_ | 19.715us | 19.715us | 1.379ms
├── recon_trunk | | |
│├── 0 | | |
││├── conv1 | | |
│││├── conv2d | 6.481us | 87.151us | 8.216ms
│││├── convolution | 5.962us | 80.670us | 8.209ms
│││├── _convolution | 13.675us | 74.708us | 8.202ms
│││├── contiguous | 2.911us | 2.911us | 3.072us
│││└── cudnn_convolution | 58.122us | 58.122us | 8.187ms
││├── conv2 | | |
│││├── conv2d | 6.114us | 88.595us | 8.217ms
│││├── convolution | 5.658us | 82.481us | 8.208ms
│││├── _convolution | 16.776us | 76.823us | 8.203ms
│││├── contiguous | 2.941us | 2.941us | 3.072us
│││└── cudnn_convolution | 57.106us | 57.106us | 8.184ms
│├── 1 | | |
││├── conv1 | | |
│││├── conv2d | 5.808us | 85.810us | 8.242ms
│││├── convolution | 5.975us | 80.002us | 8.235ms
│││├── _convolution | 16.448us | 74.027us | 8.230ms
│││├── contiguous | 2.934us | 2.934us | 3.072us
│││└── cudnn_convolution | 54.645us | 54.645us | 8.211ms
││├── conv2 | | |
│││├── conv2d | 12.929us | 92.926us | 8.244ms
│││├── convolution | 5.744us | 79.997us | 8.238ms
│││├── _convolution | 16.542us | 74.253us | 8.232ms
│││├── contiguous | 2.880us | 2.880us | 3.072us
│││└── cudnn_convolution | 54.831us | 54.831us | 8.214ms
│├── 2 | | |
││├── conv1 | | |
│││├── conv2d | 6.413us | 100.062us | 8.188ms
│││├── convolution | 5.759us | 93.649us | 8.181ms
│││├── _convolution | 19.227us | 87.890us | 8.175ms
│││├── contiguous | 2.965us | 2.965us | 3.072us
│││└── cudnn_convolution | 65.698us | 65.698us | 8.151ms
││├── conv2 | | |
│││├── conv2d | 7.536us | 120.075us | 8.232ms
│││├── convolution | 7.033us | 112.539us | 8.224ms
│││├── _convolution | 22.225us | 105.506us | 8.217ms
│││├── contiguous | 3.210us | 3.210us | 4.096us
│││└── cudnn_convolution | 80.071us | 80.071us | 8.192ms
│├── 3 | | |
││├── conv1 | | |
│││├── conv2d | 6.992us | 121.204us | 8.267ms
│││├── convolution | 7.037us | 114.212us | 8.260ms
│││├── _convolution | 22.940us | 107.175us | 8.252ms
│││├── contiguous | 3.105us | 3.105us | 3.072us
│││└── cudnn_convolution | 81.130us | 81.130us | 8.226ms
││├── conv2 | | |
│││├── conv2d | 7.523us | 124.388us | 8.291ms
│││├── convolution | 6.900us | 116.865us | 8.283ms
│││├── _convolution | 21.735us | 109.965us | 8.275ms
│││├── contiguous | 2.935us | 2.935us | 3.072us
│││└── cudnn_convolution | 85.295us | 85.295us | 8.251ms
│├── 4 | | |
││├── conv1 | | |
│││├── conv2d | 6.900us | 121.657us | 8.257ms
│││├── convolution | 6.654us | 114.757us | 8.248ms
│││├── _convolution | 22.959us | 108.103us | 8.240ms
│││├── contiguous | 3.058us | 3.058us | 4.096us
│││└── cudnn_convolution | 82.086us | 82.086us | 8.214ms
││├── conv2 | | |
│││├── conv2d | 6.019us | 101.440us | 8.199ms
│││├── convolution | 6.450us | 95.421us | 8.193ms
│││├── _convolution | 20.097us | 88.971us | 8.185ms
│││├── contiguous | 3.012us | 3.012us | 4.096us
│││└── cudnn_convolution | 65.862us | 65.862us | 8.162ms
│├── 5 | | |
││├── conv1 | | |
│││├── conv2d | 6.511us | 101.070us | 8.187ms
│││├── convolution | 6.306us | 94.559us | 8.179ms
│││├── _convolution | 28.783us | 88.253us | 8.172ms
│││├── contiguous | 3.326us | 3.326us | 4.096us
│││└── cudnn_convolution | 56.144us | 56.144us | 8.141ms
││├── conv2 | | |
│││├── conv2d | 6.081us | 88.464us | 8.236ms
│││├── convolution | 6.151us | 82.383us | 8.229ms
│││├── _convolution | 17.028us | 76.232us | 8.222ms
│││├── contiguous | 3.286us | 3.286us | 4.096us
│││└── cudnn_convolution | 55.918us | 55.918us | 8.202ms
│├── 6 | | |
││├── conv1 | | |
│││├── conv2d | 6.063us | 86.803us | 8.236ms
│││├── convolution | 5.701us | 80.740us | 8.228ms
│││├── _convolution | 16.474us | 75.039us | 8.220ms
│││├── contiguous | 3.305us | 3.305us | 3.072us
│││└── cudnn_convolution | 55.260us | 55.260us | 8.201ms
││├── conv2 | | |
│││├── conv2d | 6.083us | 91.663us | 8.257ms
│││├── convolution | 5.939us | 85.580us | 8.249ms
│││├── _convolution | 18.140us | 79.641us | 8.243ms
│││├── contiguous | 2.786us | 2.786us | 4.096us
│││└── cudnn_convolution | 58.715us | 58.715us | 8.223ms
│├── 7 | | |
││├── conv1 | | |
│││├── conv2d | 6.953us | 125.802us | 8.209ms
│││├── convolution | 7.485us | 118.849us | 8.200ms
│││├── _convolution | 26.536us | 111.364us | 8.187ms
│││├── contiguous | 3.843us | 3.843us | 4.096us
│││└── cudnn_convolution | 80.985us | 80.985us | 8.162ms
││├── conv2 | | |
│││├── conv2d | 6.630us | 125.386us | 8.237ms
│││├── convolution | 7.316us | 118.756us | 8.230ms
│││├── _convolution | 21.805us | 111.440us | 8.221ms
│││├── contiguous | 3.496us | 3.496us | 4.096us
│││└── cudnn_convolution | 86.139us | 86.139us | 8.196ms
│├── 8 | | |
││├── conv1 | | |
│││├── conv2d | 6.592us | 121.497us | 8.211ms
│││├── convolution | 7.503us | 114.905us | 8.203ms
│││├── _convolution | 22.946us | 107.402us | 8.195ms
│││├── contiguous | 3.376us | 3.376us | 3.072us
│││└── cudnn_convolution | 81.080us | 81.080us | 8.169ms
││├── conv2 | | |
│││├── conv2d | 7.066us | 115.946us | 8.255ms
│││├── convolution | 9.911us | 108.880us | 8.248ms
│││├── _convolution | 18.125us | 98.969us | 8.242ms
│││├── contiguous | 2.911us | 2.911us | 3.072us
│││└── cudnn_convolution | 77.933us | 77.933us | 8.221ms
│├── 9 | | |
││├── conv1 | | |
│││├── conv2d | 6.471us | 101.256us | 8.272ms
│││├── convolution | 8.887us | 94.785us | 8.264ms
│││├── _convolution | 15.858us | 85.898us | 8.255ms
│││├── contiguous | 2.916us | 2.916us | 4.096us
│││└── cudnn_convolution | 67.124us | 67.124us | 8.237ms
││└── conv2 | | |
││ ├── conv2d | 6.526us | 92.084us | 8.215ms
││ ├── convolution | 8.615us | 85.558us | 8.208ms
││ ├── _convolution | 14.271us | 76.943us | 8.201ms
││ ├── contiguous | 2.763us | 2.763us | 3.072us
││ └── cudnn_convolution | 59.909us | 59.909us | 8.186ms
├── upconv1 | | |
│├── conv2d | 6.078us | 103.623us | 31.029ms
│├── convolution | 9.666us | 97.545us | 31.022ms
│├── _convolution | 14.785us | 87.879us | 31.015ms
│├── contiguous | 3.103us | 3.103us | 3.072us
│└── cudnn_convolution | 69.991us | 69.991us | 30.998ms
├── upconv2 | | |
│├── conv2d | 6.107us | 97.631us | 32.029ms
│├── convolution | 6.589us | 91.524us | 32.023ms
│├── _convolution | 18.494us | 84.935us | 32.014ms
│├── contiguous | 2.931us | 2.931us | 4.096us
│└── cudnn_convolution | 63.510us | 63.510us | 31.997ms
├── upconv3 | | |
│├── conv2d | 7.143us | 122.706us | 26.778ms
│├── convolution | 9.899us | 115.563us | 26.770ms
│├── _convolution | 18.894us | 105.664us | 26.762ms
│├── contiguous | 3.208us | 3.208us | 4.096us
│└── cudnn_convolution | 83.562us | 83.562us | 26.741ms
├── pixel_shuffle | | |
│├── pixel_shuffle | 20.867us | 126.565us | 5.932ms
│├── reshape | 15.966us | 72.601us | 5.879ms
│├── as_strided | 11.647us | 11.647us | 12.288us
│├── permute | 10.313us | 10.313us | 9.216us
│├── clone | 42.860us | 42.860us | 5.863ms
│└── _unsafe_view | 13.775us | 13.775us | 3.072us
├── HRconv | | |
│├── conv2d | 6.791us | 104.156us | 34.608ms
│├── convolution | 8.782us | 97.365us | 34.600ms
│├── _convolution | 14.569us | 88.583us | 34.590ms
│├── contiguous | 3.312us | 3.312us | 3.072us
│└── cudnn_convolution | 70.702us | 70.702us | 34.571ms
├── conv_last | | |
│├── conv2d | 7.488us | 139.605us | 14.136ms
│├── convolution | 17.199us | 132.117us | 14.130ms
│├── _convolution | 18.994us | 114.918us | 14.110ms
│├── contiguous | 3.522us | 3.522us | 4.096us
│└── cudnn_convolution | 92.402us | 92.402us | 14.086ms
└── lrelu | | |
└── leaky_relu_ | 30.614us | 30.614us | 5.475ms
Trace(path=('EDVR_sr', 'pre_deblur', 'conv_first'), leaf=True, module=Conv2d(3, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)))
--------------------- --------------- --------------- --------------- --------------- --------------- --------------- --------------- --------------- --------------- -----------------------------------
Name Self CPU total % Self CPU total CPU total % CPU total CPU time avg CUDA total % CUDA total CUDA time avg Number of Calls Input Shapes
--------------------- --------------- --------------- --------------- --------------- --------------- --------------- --------------- --------------- --------------- -----------------------------------
conv2d 5.96% 9.318us 100.00% 156.261us 156.261us 25.03% 13.230ms 13.230ms 1 []
convolution 5.62% 8.784us 94.04% 146.943us 146.943us 25.01% 13.220ms 13.220ms 1 []
_convolution 14.34% 22.409us 88.42% 138.159us 138.159us 25.00% 13.211ms 13.211ms 1 []
contiguous 5.54% 8.654us 5.54% 8.654us 8.654us 0.02% 8.192us 8.192us 1 []
cudnn_convolution 68.54% 107.096us 68.54% 107.096us 107.096us 24.94% 13.180ms 13.180ms 1 []
--------------------- --------------- --------------- --------------- --------------- --------------- --------------- --------------- --------------- --------------- -----------------------------------
Self CPU time total: 156.261us
CUDA time total: 52.849ms
百度成长空间 412人发布
