mindspore-Ecosystem
/
mindspore
21
0
Fork 27
Code Issues Pull Requests 6 Projects Releases Wiki Activity

fix I42MO2

This commit is contained in:
l00486551 2021-07-29 15:05:33 +08:00
parent d6ee1b8407
commit 42f90db3c1
1 changed files with 37 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

49
model_zoo/research/cv/resnetv2/README_CN.md
View File

@ -65,26 +65,38 @@ ResNet系列模型是在2015年提出的，该网络创新性的提出了残差
```Shell
# 分布式训练
用法：sh run_distribute_train.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [RANK_TABLE_FILE] [DATASET_PATH]
用法：
cd ./scripts
sh run_distribute_train.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [RANK_TABLE_FILE] [DATASET_PATH]
# 单机训练
用法：sh run_standalone_train.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [DATASET_PATH]
用法：
cd ./scripts
sh run_standalone_train.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [DATASET_PATH]
# 运行评估示例
用法：sh run_eval.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [DATASET_PATH] [CHECKPOINT_PATH]
用法：
cd ./scripts
sh run_eval.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [DATASET_PATH] [CHECKPOINT_PATH]
```
- GPU处理器环境运行
```shell
# 分布式训练
用法：sh run_distribute_train_gpu.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [RANK_TABLE_FILE] [DATASET_PATH]
用法：
cd ./scripts
sh run_distribute_train_gpu.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [RANK_TABLE_FILE] [DATASET_PATH]
# 单机训练
用法：sh run_standalone_train_gpu.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [DATASET_PATH]
用法：
cd ./scripts
sh run_standalone_train_gpu.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [DATASET_PATH]
# 运行评估示例
用法：sh run_eval_gpu.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [DATASET_PATH] [CHECKPOINT_PATH]
用法：
cd ./scripts
sh run_eval_gpu.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [DATASET_PATH] [CHECKPOINT_PATH]
```
# 脚本说明
@ -168,10 +180,14 @@ ResNet系列模型是在2015年提出的，该网络创新性的提出了残差
```Shell
# 分布式训练
用法：sh run_distribute_train.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [RANK_TABLE_FILE] [DATASET_PATH]
用法：
cd ./scripts
sh run_distribute_train.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [RANK_TABLE_FILE] [DATASET_PATH]
# 单机训练
用法：sh run_standalone_train.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [DATASET_PATH]
用法：
cd ./scripts
sh run_standalone_train.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [DATASET_PATH]
```
分布式训练需要提前创建JSON格式的HCCL配置文件。
@ -182,10 +198,14 @@ ResNet系列模型是在2015年提出的，该网络创新性的提出了残差
```shell
# 分布式训练
用法：sh run_distribute_train_gpu.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [RANK_TABLE_FILE] [DATASET_PATH]
用法：
cd ./scripts
sh run_distribute_train_gpu.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [RANK_TABLE_FILE] [DATASET_PATH]
# 单机训练
用法：sh run_standalone_train_gpu.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [DATASET_PATH]
用法：
cd ./scripts
sh run_standalone_train_gpu.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [DATASET_PATH]
```
## 结果
@ -224,14 +244,18 @@ epoch time: 813347.102 ms, per step time: 325.075 ms
```Shell
# 评估
用法：sh run_eval.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [DATASET_PATH] [CHECKPOINT_PATH]
用法：
cd ./scripts
sh run_eval.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [DATASET_PATH] [CHECKPOINT_PATH]
```
### GPU处理器环境运行
```shell
# 运行评估示例
用法：sh run_eval_gpu.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [DATASET_PATH] [CHECKPOINT_PATH]
用法：
cd ./scripts
sh run_eval_gpu.sh [resnetv2_50|resnetv2_101|resnetv2_152] [cifar10|imagenet2012] [DATASET_PATH] [CHECKPOINT_PATH]
```
## 结果
@ -267,6 +291,7 @@ python export.py --ckpt_file [CKPT_PATH] --file_name [FILE_NAME] --file_format [
```shell
# Ascend310 inference
cd ./scripts
bash run_infer_310.sh [MINDIR_PATH] [DATASET] [DATA_PATH] [DEVICE_ID]
```