mirrors/Smart_container
1
0
Fork 0
mirror of https://github.com/ColugoMum/Smart_container.git synced 2025-06-03 21:54:04 +08:00
Code Issues Projects Releases Wiki Activity

Update README.md

Browse Source
This commit is contained in:
thomas-yanxin 2022-03-14 01:06:15 +08:00 committed by GitHub
parent 2ef56990f0
commit a60a6d8330
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 13 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

26
README.md
View File

@ -96,19 +96,19 @@ PP-ShiTu是一个实用的轻量级通用图像识别系统，主要由主体检
目前处理后的数据集已在[AIStudio](https://aistudio.baidu.com/aistudio/datasetdetail/113685)开源。 </font>
## 消融实验 ##
| model | num epoch | batch size/gpu cards | learning rate | use cutout | use ssld | top1 recall |
| :----: | :---- | :---- | :---- | :---- | :---- | :---- |
| PP_LCNet_x2_5 | 400 | 256/4 | 0.01 | N | N | 98.189% |
| PP_LCNet_x2_5 | 400 | 256/4 | 0.01 | Y | N | [98.21%](./exprements/log/98216.log) |
| PP_LCNet_x2_5 | 400 | 256/4| 0.005 | N | N | 98.201% |
| PP_LCNet_x2_5 | 400 | 256/4| 0.005 | Y | N | [98.29%](./exprements/log/98291.log) |
| PP_LCNet_x2_5 | 400 | 256/4 | 0.001 | Y | N | 98.26% |
| PP_LCNet_x2_5 | 400 | 64/4 | 0.005 | Y | Y | 98.30% |
| PP_LCNet_x2_5 | 400 | 64/4 | 0.0025 | Y | Y | [98.37%](./exprements/log/98379.log) |
| PP_LCNet_x2_5 | 400 | 64/4 | 0.002 | N | Y | [98.38%](./exprements/log/98383.log) |
| PP_LCNet_x2_5 | 400 | 64/4 | 0.002 | Y | Y | [98.39%]((./exprements/log/98395.log)) |
| PP_LCNet_x2_5 | 400 | 128/4 | 0.004 | N | Y | [98.44%](./exprements/log/98442.log) |
| PP_LCNet_x2_5 | 400 | 128/4 | 0.004 | Y | Y | 98.38% |
| model | num epoch | batch size/gpu cards | learning rate | use cutout | use ssld | top1 recall | 配置文件 |
| :----: | :---- | :---- | :---- | :---- | :---- | :---- | :---- |
| PP_LCNet_x2_5 | 400 | 256/4 | 0.01 | N | N | 98.189% | [配置文件](./exprements/PaddleClas/ppcls/configs/GeneralRecognition/GeneralRecognition_PPLCNet_x2_5_01.yaml) |
| PP_LCNet_x2_5 | 400 | 256/4 | 0.01 | Y | N | [98.21%](./exprements/log/98216.log) | [配置文件](./exprements/PaddleClas/ppcls/configs/GeneralRecognition/GeneralRecognition_PPLCNet_x2_5_01_cutout.yaml) |
| PP_LCNet_x2_5 | 400 | 256/4| 0.005 | N | N | 98.201% | [配置文件](./exprements/PaddleClas/ppcls/configs/GeneralRecognition/GeneralRecognition_PPLCNet_x2_5_005.yaml) |
| PP_LCNet_x2_5 | 400 | 256/4| 0.005 | Y | N | [98.29%](./exprements/log/98291.log) | [配置文件](./exprements/PaddleClas/ppcls/configs/GeneralRecognition/GeneralRecognition_PPLCNet_x2_5_005_cutout.yaml) |
| PP_LCNet_x2_5 | 400 | 256/4 | 0.001 | Y | N | 98.26% |配置文件|
| PP_LCNet_x2_5 | 400 | 64/4 | 0.005 | Y | Y | 98.30% | 配置文件|
| PP_LCNet_x2_5 | 400 | 64/4 | 0.0025 | Y | Y | [98.37%](./exprements/log/98379.log) | 配置文件 |
| PP_LCNet_x2_5 | 400 | 64/4 | 0.002 | N | Y | [98.38%](./exprements/log/98383.log) | [配置文件](./exprements/PaddleClas/ppcls/configs/GeneralRecognition/GeneralRecognition_PPLCNet_x2_5_dml_002.yaml) |
| PP_LCNet_x2_5 | 400 | 64/4 | 0.002 | Y | Y | [98.39%]((./exprements/log/98395.log)) | [配置文件](./exprements/PaddleClas/ppcls/configs/GeneralRecognition/GeneralRecognition_PPLCNet_x2_5_dml_002_cutout.yaml) |
| PP_LCNet_x2_5 | 400 | 128/4 | 0.004 | N | Y | [98.44%](./exprements/log/98442.log) | [配置文件](./exprements/PaddleClas/ppcls/configs/GeneralRecognition/GeneralRecognition_PPLCNet_x2_5_004.yaml) |
| PP_LCNet_x2_5 | 400 | 128/4 | 0.004 | Y | Y | 98.38% | [配置文件](./exprements/PaddleClas/ppcls/configs/GeneralRecognition/GeneralRecognition_PPLCNet_x2_5_004_cutout.yaml) |
**注**:
1. 本实验均基于 GPU:Tesla V100* 4; CPU:Inter Xeon* 32; RAM:DDR4 128GB 进行训练并测试；