commit
d7d5353bee
|
@ -99,6 +99,9 @@ Currently there are 32 pretrained models of the mobile series open source by Pad
|
|||
| ShuffleNetV2_x1_5 | 19.352 | 14.000 |
|
||||
| ShuffleNetV2_x2_0 | 34.770 | 28.000 |
|
||||
| ShuffleNetV2_swish | 16.023 | 9.100 |
|
||||
| GhostNet_x0_5 | 5.714 | 10.000 |
|
||||
| GhostNet_x1_0 | 13.558 | 20.000 |
|
||||
| GhostNet_x1_3 | 19.982 | 29.000 |
|
||||
|
||||
|
||||
## Inference speed based on T4 GPU
|
||||
|
|
Binary file not shown.
Before Width: | Height: | Size: 148 KiB After Width: | Height: | Size: 141 KiB |
|
@ -100,6 +100,9 @@ GhosttNet是华为于2020年提出的一种全新的轻量化网络结构,通
|
|||
| ShuffleNetV2_x1_5 | 19.352 | 14.000 |
|
||||
| ShuffleNetV2_x2_0 | 34.770 | 28.000 |
|
||||
| ShuffleNetV2_swish | 16.023 | 9.100 |
|
||||
| GhostNet_x0_5 | 5.714 | 10.000 |
|
||||
| GhostNet_x1_0 | 13.558 | 20.000 |
|
||||
| GhostNet_x1_3 | 19.982 | 29.000 |
|
||||
|
||||
|
||||
## 基于T4 GPU的预测速度
|
||||
|
|
Loading…
Reference in New Issue