修改 mmpose-0.28.1 train.md

openmmlab_test/mmclassification-0.24.1/train.md

@@ -55,7 +55,7 @@ configs/_base_/datasets/imagenet_bs32.py 中batch_size=samples_per_gpu*卡数,
 1.pytorch单机多卡训练
 
 ```python
-./tools/dist_train.sh configs/resnet/resnet18_b32x8_imagenet.py
+./tools/dist_train.sh configs/resnet/resnet18_b32x8_imagenet.py $GPUS
 ```
 2.mpirun单机多卡训练
 mpirun --allow-run-as-root --bind-to none -np 4 single_process.sh a03r3n15
@@ -65,9 +65,9 @@ a03r3n15为master节点ip
 
 1.pytorch多机多卡训练
 在第一台机器上:
-NODES=2 NODE_RANK=0 PORT=12345 MASTER_ADDR=10.1.3.56 sh tools/dist_train.sh configs/resnet/resnet18_b32x8_imagenet.py 4
+NODES=2 NODE_RANK=0 PORT=12345 MASTER_ADDR=10.1.3.56 sh tools/dist_train.sh configs/resnet/resnet18_b32x8_imagenet.py $GPUS
 在第二台机器上:
-NODES=2 NODE_RANK=1 PORT=12345 MASTER_ADDR=10.1.3.56 sh tools/dist_train.sh configs/resnet/resnet18_b32x8_imagenet.py 4
+NODES=2 NODE_RANK=1 PORT=12345 MASTER_ADDR=10.1.3.56 sh tools/dist_train.sh configs/resnet/resnet18_b32x8_imagenet.py $GPUS
 
 2.mpirun多机多卡训练
 mpirun --allow-run-as-root --hostfile hostfile --bind-to none -np 4 single_process.sh a03r3n15

openmmlab_test/mmdetection-2.25.2/train.md

@@ -74,7 +74,7 @@ configs/_base_/datasets/coco_detection.py 中batch_size=samples_per_gpu*卡数,
 
 ```python
 
-./tools/dist_train.sh configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco.py
+./tools/dist_train.sh configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco.py $GPUS
 
 ```
 
@@ -84,11 +84,11 @@ configs/_base_/datasets/coco_detection.py 中batch_size=samples_per_gpu*卡数,
 
 在第一台机器上:
 
-NODES=2 NODE_RANK=0 PORT=12345 MASTER_ADDR=10.1.3.56 sh tools/dist_train.sh configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco.py 4
+NODES=2 NODE_RANK=0 PORT=12345 MASTER_ADDR=10.1.3.56 sh tools/dist_train.sh configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco.py $GPUS
 
 在第二台机器上:
 
-NODES=2 NODE_RANK=1 PORT=12345 MASTER_ADDR=10.1.3.56 sh tools/dist_train.sh configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco.py 4
+NODES=2 NODE_RANK=1 PORT=12345 MASTER_ADDR=10.1.3.56 sh tools/dist_train.sh configs/faster_rcnn/faster_rcnn_r50_fpn_1x_coco.py $GPUS
 
 
 ### 半精度测试

openmmlab_test/mmpose-0.28.1/README.md

@@ -37,7 +37,7 @@
 
 <div align="center">
 
-English | [简体中文](README_CN.md)
+English | [简体中文](README_CN.md)  | [模型测试步骤](train.md)
 
 </div>
 

openmmlab_test/mmpose-0.28.1/train.md

@@ -2,7 +2,7 @@
 
 ## 测试前准备
 ### 数据集准备
-使用dummy数据集.
+使用COCO2017数据集.
 ### 环境部署
 ```python
 yum install python3
@@ -19,79 +19,56 @@ pip3 install opencv-python
 ```
 ### 安装python依赖包
 ```python
-pip3 install torch-1.10.0a0+gitcc7c9c7-cp36-cp36m-linux_x86_64.whl -i https://pypi.tuna.tsinghua.edu.cn/simple
-pip3 install torchvision-0.10.0a0+300a8a4-cp36-cp36m-linux_x86_64.whl -i https://pypi.tuna.tsinghua.edu.cn/simple
-pip3 install mmcv_full-1.3.16-cp36-cp36m-linux_x86_64.whl -i https://pypi.tuna.tsinghua.edu.cn/simple
+pip3 install torch-1.10.0a0+git2040069.dtk2210-cp37-cp37m-manylinux2014_x86_64.whl -i https://pypi.tuna.tsinghua.edu.cn/simple
+pip3 install torchvision-0.10.0a0+e04d001.dtk2210-cp37-cp37m-manylinux2014_x86_64.whl -i https://pypi.tuna.tsinghua.edu.cn/simple
+pip3 install mmcv_full-1.6.1+gitdebbc80.dtk2210-cp37-cp37m-manylinux2014_x86_64.whl -i https://pypi.tuna.tsinghua.edu.cn/simple
 poseval安装：
 cd poseval
 pip install -e .  
 mmpose安装：
-cd mmpose 
+cd mmpose-0.28.1 
 pip3 install -e .
 ```
-注：测试不同版本的dtk，需安装对应版本的库whl包，dtk22.04.1使用python3.7，dtk21.10.1使用python3.6
+注：测试不同版本的dtk，需安装对应版本的库whl包.
 ## ResNet50-Bottom-Up测试
 ### 单卡测试（单精度）
 ```python
-./sing_test.sh configs/speed_test/bottomup_hrnet_w32_coco_512x512_dummy.py
+./sing_test.sh configs/body/2d_kpt_sview_rgb_img/res50_coco_256x192_dummy.py
 ```
-若使用dtk22.04.1测试，需设置`export MIOPEN_FIND_MODE=1，export MIOPEN_USE_APPROXIMATE_PERFORMANCE=0`以下测试均相同.
-
 #### 参数说明
-configs/speed_test/bottomup_hrnet_w32_coco_512x512_dummy.py 中batch_size=samples_per_gpu*卡数,性能计算方法：batch_size/time
-![1659168500714](image/train/1659168500714.png)
+configs/body/2d_kpt_sview_rgb_img/res50_coco_256x192_dummy.py 中batch_size=samples_per_gpu*卡数,性能计算方法：batch_size/time
 #### 性能关注：time
-### 多卡测试（单精度）
-```python
-./multi_test.sh configs/speed_test/bottomup_hrnet_w32_coco_512x512_dummy.py
-```
-### 单卡测试（半精度）
-```python
-./sing_test.sh configs/speed_test/bottomup_hrnet_w32_coco_512x512_fp16_dummy.py
-```
-### 多卡测试（半精度）
-```python
-./multi_test.sh configs/speed_test/bottomup_hrnet_w32_coco_512x512_fp16_dummy.py
-```
-## ResNet50-Top-Down测试
-### 单卡测试（单精度）
-```python
-./sing_test.sh configs/speed_test/res50_coco_256x192_dummy.py
-```
-#### 参数说明
-configs/speed_test/res50_coco_256x192_dummy.py 中batch_size=samples_per_gpu*卡数,性能计算方法：batch_size/time
-![1659169127556](image/train/1659169127556.png)
-#### 性能关注：time
-### 多卡测试（单精度）
-```python
-./multi_test.sh configs/speed_test/res50_coco_256x192_dummy.py
-```
-### 单卡测试（半精度）
-```python
-./sing_test.sh configs/speed_test/res50_coco_256x192_fp16_dummy.py
-```
-### 多卡测试（半精度）
-```python
-./multi_test.sh configs/speed_test/res50_coco_256x192_fp16_dummy.py
-```
-## HrNet-Top-Down测试
-### 单卡测试（单精度）
-```python
-./sing_test.sh configs/speed_test/hrnet_w32_coco_256x192_dummy.py
-```
-#### 参数说明
-configs/speed_test/hrnet_w32_coco_256x192_dummy.py 中batch_size=samples_per_gpu*卡数,性能计算方法：batch_size/time
-![1659169253165](image/train/1659169253165.png)
-#### 性能关注：time
-### 多卡测试（单精度）
-```python
-./multi_test.sh configs/speed_test/hrnet_w32_coco_256x192_dummy.py
-```
-### 单卡测试（半精度）
-```python
-./sing_test.sh configs/speed_test/hrnet_w32_coco_256x192_fp16_dummy.py
-```
-### 多卡测试（半精度）
+
+#### 单机多卡训练
+
+1.pytorch单机多卡训练
+
 ```python
-./multi_test.sh configs/speed_test/hrnet_w32_coco_256x192_fp16_dummy.py
+./tools/dist_train.sh configs/body/2d_kpt_sview_rgb_img/res50_coco_256x192_dummy.py $GPUS
 ```
+
+#### 多机多卡训练
+
+1.pytorch多机多卡训练
+
+在第一台机器上:
+
+NODES=$NNODES NODE_RANK=$NODE_RANK PORT=$PORT MASTER_ADDR=$MASTER_ADDR sh tools/dist_train.sh configs/body/2d_kpt_sview_rgb_img/res50_coco_256x192_dummy.py $GPUS
+
+在第二台机器上:
+
+NODES=$NNODES NODE_RANK=$NODE_RANK PORT=$PORT MASTER_ADDR=$MASTER_ADDR sh tools/dist_train.sh configs/body/2d_kpt_sview_rgb_img/res50_coco_256x192_dummy.py $GPUS
+
+### 半精度测试
+
+修改configs文件，添加fp16 = dict(loss_scale=512.)，单机多卡和多机多卡测试与单精度测试方法相同。
+
+### 其他模型测试
+
+其他模型的测试步骤和ResNet50-Bottom-Up相同，只需修改对应的config文件即可，下面列出相关模型对应的config文件列表：
+
+| 模型               | configs                                                      |
+| ------------------ | ------------------------------------------------------------ |
+| ResNet50-Top-Down  | configs/body/2d_kpt_sview_rgb_img/res50_coco_256x192_dummy.py |
+| HrNet-Top-Down     | configs/body/2d_kpt_sview_rgb_img/topdown_heatmap/coco/hrnet_w32_coco_256x192.py |
+| ResNet50-Bottom-Up | configs/body/2d_kpt_sview_rgb_img/associative_embedding/coco/hrnet_w32_coco_512x512.py |