NPU (HUAWEI Ascend)¶
Usage¶
Please refer to the building documentation of MMCV to install MMCV and MMEngine on NPU devices.
Here we use 4 NPUs on your computer to train the model with the following command:
bash tools/dist_train.sh configs/deeplabv3/deeplabv3_r50-d8_512x1024_40k_cityscapes.py 4
Also, you can use only one NPU to train the model with the following command:
python tools/train.py configs/deeplabv3/deeplabv3_r50-d8_512x1024_40k_cityscapes.py
Models Results¶
Model | mIoU | Config | Download |
---|---|---|---|
deeplabv3 | 78.92 | config | log |
deeplabv3plus | 79.68 | config | log |
hrnet | 77.09 | config | log |
fcn | 72.69 | config | log |
pspnet | 78.07 | config | log |
unet | 69.00 | config | log |
apcnet | 78.07 | config | log |
upernet | 78.15 | config | log |
Notes:
If not specially marked, the results on NPU with amp are the basically same as those on the GPU with FP32.
All above models are provided by Huawei Ascend group.