[Feature] add docs with NPU backend (open-mmlab#9333)

* update npu doc

* add ssdlite_mbv2 result

* update url

configs/retinanet/retinanet_r50_fpn_fp16_1x_coco.py

@@ -1,3 +1,7 @@
 _base_ = './retinanet_r50_fpn_1x_coco.py'
 # fp16 settings
 fp16 = dict(loss_scale=512.)
+
+# set grad_norm for stability during mixed-precision training
+optimizer_config = dict(
+    _delete_=True, grad_clip=dict(max_norm=35, norm_type=2))

configs/ssd/ssd512_fp16_coco.py

@@ -0,0 +1,9 @@
+_base_ = ['./ssd512_coco.py']
+# fp16 settings
+fp16 = dict(loss_scale='dynamic')
+
+# learning policy
+# In order to avoid non-convergence in the early stage of
+# mixed-precision training, the warmup in the lr_config is set to linear,
+# warmup_iters increases and warmup_ratio decreases.
+lr_config = dict(warmup='linear', warmup_iters=1000, warmup_ratio=1.0 / 10)

docs/en/device/npu.md

@@ -0,0 +1,35 @@
+# NPU (HUAWEI Ascend)
+
+## Usage
+
+Please refer to [link](https://github.com/open-mmlab/mmcv/blob/master/docs/zh_cn/get_started/build.md) installing mmcv on NPU Devices.
+
+Here we use 8 NPUs on your computer to train the model with the following command:
+
+```shell
+bash tools/dist_train.sh configs/ssd/ssd300_coco.py 8
+```
+
+Also, you can use only one NPU to train the model with the following command:
+
+```shell
+python tools/train.py configs/ssd/ssd300_coco.py
+```
+
+## Verified Models
+
+|        Model         | box AP | mask AP | Config                                                                                                                        | Download                                                                                                     |
+| :------------------: | :----: | :-----: | :---------------------------------------------------------------------------------------------------------------------------- | :----------------------------------------------------------------------------------------------------------- |
+|     [ssd300](<>)     |  25.6  |   ---   | [config](https://github.com/open-mmlab/mmdetection/blob/master/configs/ssd/ssd300_fp16_coco.py)                               | [log](https://download.openmmlab.com/mmdetection/v2.0/npu/ssd300_coco.log.json)                              |
+|     [ssd512](<>)     |  29.4  |   ---   | [config](https://github.com/open-mmlab/mmdetection/blob/master/configs/ssd/ssd512_fp16_coco.py)                               | [log](https://download.openmmlab.com/mmdetection/v2.0/npu/ssd512_coco.log.json)                              |
+| [\*ssdlite-mbv2](<>) |  20.2  |   ---   | [config](https://github.com/open-mmlab/mmdetection/blob/master/configs/ssd/ssdlite_mobilenetv2_scratch_600e_coco.py)          | [log](https://download.openmmlab.com/mmdetection/v2.0/npu/ssdlite_mobilenetv2_scratch_600e_coco.log.json)    |
+| [retinanet-r50](<>)  |  36.6  |   ---   | [config](https://github.com/open-mmlab/mmdetection/blob/master/configs/retinanet/retinanet_r50_fpn_fp16_1x_coco.py)           | [log](https://download.openmmlab.com/mmdetection/v2.0/npu/retinanet_r50_fpn_1x_coco.log.json)                |
+|   [\*fcos-r50](<>)   |  36.1  |   ---   | [config](https://github.com/open-mmlab/mmdetection/blob/master/configs/fcos/fcos_r50_caffe_fpn_gn-head_fp16_1x_bs8x8_coco.py) | [log](https://download.openmmlab.com/mmdetection/v2.0/npu/fcos_r50_caffe_fpn_gn-head_1x_coco_bs8x8.log.json) |
+|   [solov2-r50](<>)   |  ---   |  34.7   | [config](https://github.com/open-mmlab/mmdetection/blob/master/configs/solov2/solov2_r50_fpn_1x_coco.py)                      | [log](https://download.openmmlab.com/mmdetection/v2.0/npu/solov2_r50_fpn_1x_coco.log.json)                   |
+
+**Notes:**
+
+- If not specially marked, the results are same between results on the NPU and results on the GPU with FP32.
+- (\*) The results on the NPU of these models are aligned with the results of the mixed-precision training on the GPU, but are lower than the results of the FP32. This situation is mainly related to the phase of the model itself in mixed-precision training, users please adjust the hyperparameters to achieve the best result by self.
+
+**All above models are provided by Huawei Ascend group.**

docs/en/index.rst

@@ -50,6 +50,12 @@ Welcome to MMDetection's documentation!
 
    api.rst
 
+.. toctree::
+   :maxdepth: 1
+   :caption: Device Support
+
+   device/npu.md
+
 Indices and tables
 ==================