add resnet18 310 infer in readme

model_zoo/official/cv/resnet/README.md

@@ -12,6 +12,10 @@
     - [Script Parameters](#script-parameters)
     - [Training Process](#training-process)
     - [Evaluation Process](#evaluation-process)
+- [Inference Process](#inference-process)
+    - [Export MindIR](#export-mindir)
+    - [Infer on Ascend310](#infer-on-ascend310)
+    - [result](#result)
 - [Model Description](#model-description)
     - [Performance](#performance)
         - [Evaluation Performance](#evaluation-performance)
@@ -479,6 +483,37 @@ result: {'top_5_accuracy': 0.9342589628681178, 'top_1_accuracy': 0.7680657810499
 
 ```
 
+## Inference Process
+
+### [Export MindIR](#contents)
+
+```shell
+python export.py --ckpt_file [CKPT_PATH] --file_name [FILE_NAME] --file_format [FILE_FORMAT]
+```
+
+The ckpt_file parameter is required,
+`EXPORT_FORMAT` should be in ["AIR", "MINDIR"]
+
+### Infer on Ascend310
+
+Before performing inference, the mindir file must bu exported by `export.py` script. We only provide an example of inference using MINDIR model.
+Current batch_Size can only be set to 1. The precision calculation process needs about 70G+ memory space, otherwise the process will be killed for execeeding memory limits.
+
+```shell
+# Ascend310 inference
+bash run_infer_310.sh [MINDIR_PATH] [DATA_PATH] [DEVICE_ID]
+```
+
+- `DEVICE_ID` is optional, default value is 0.
+
+### result
+
+Inference result is saved in current path, you can find result like this in acc.log file.
+
+```bash
+top1_accuracy:70.42, top5_accuracy:89.7
+```
+
 # [Model Description](#contents)
 
 ## [Performance](#contents)

model_zoo/official/cv/resnet/README_CN.md

@@ -14,6 +14,10 @@
     - [脚本参数](#脚本参数)
     - [训练过程](#训练过程)
     - [评估过程](#评估过程)
+    - [推理过程](#推理过程)
+        - [导出MindIR](#导出mindir)
+        - [在Ascend310执行推理](#在ascend310执行推理)
+        - [结果](#结果)
 - [模型描述](#模型描述)
     - [性能](#性能)
         - [评估性能](#评估性能)
@@ -446,6 +450,37 @@ result:{'top_5_accuracy':0.9342589628681178, 'top_1_accuracy':0.768065781049936}
 
 ```
 
+## 推理过程
+
+### [导出MindIR](#contents)
+
+```shell
+python export.py --ckpt_file [CKPT_PATH] --file_name [FILE_NAME] --file_format [FILE_FORMAT]
+```
+
+参数ckpt_file为必填项，
+`EXPORT_FORMAT` 必须在 ["AIR", "MINDIR"]中选择。
+
+### 在Ascend310执行推理
+
+在执行推理前，mindir文件必须通过`export.py`脚本导出。以下展示了使用minir模型执行推理的示例。
+目前仅支持batch_Size为1的推理。精度计算过程需要70G+的内存，否则进程将会因为超出内存被系统终止。
+
+```shell
+# Ascend310 inference
+bash run_infer_310.sh [MINDIR_PATH] [DATA_PATH] [DEVICE_ID]
+```
+
+- `DEVICE_ID` 可选，默认值为0。
+
+### 结果
+
+推理结果保存在脚本执行的当前路径，你可以在acc.log中看到以下精度计算结果。
+
+```bash
+top1_accuracy:70.42, top5_accuracy:89.7
+```
+
 # 模型描述
 
 ## 性能

model_zoo/official/cv/resnet/scripts/run_infer_310.sh

@@ -55,7 +55,7 @@ fi
 
 function compile_app()
 {
-    cd ../ascend310_infer/src/
+    cd ../ascend310_infer/src/ || exit
     if [ -f "Makefile" ]; then
         make clean
     fi
@@ -64,7 +64,7 @@ function compile_app()
 
 function infer()
 {
-    cd -
+    cd - || exit
     if [ -d result_Files ]; then
         rm -rf ./result_Files
     fi