Skip to content

Commit f773e86

Browse files
authored
cherry pick polish doc (#329)
1 parent feef394 commit f773e86

File tree

6 files changed

+6
-4
lines changed

6 files changed

+6
-4
lines changed

docs/MODEL_ZOO.md

+1
Original file line numberDiff line numberDiff line change
@@ -160,6 +160,7 @@ improved performance mainly by using L1 loss in bounding box width and height re
160160
randomly color distortion, randomly cropping, randomly expansion, randomly interpolation method, randomly flippling. YOLO v3 used randomly
161161
reshaped minibatch in training, inferences can be performed on different image sizes with the same model weights, and we provided evaluation
162162
results of image size 608/416/320 above. Deformable conv is added on stage 5 of backbone.
163+
- YOLO v3 enhanced model improves the precision to 43.2 involved with deformable conv, dropblock and IoU loss. See more details in [YOLOv3_ENHANCEMENT](./featured_model/YOLOv3_ENHANCEMENT.md)
163164

164165
### RetinaNet
165166

docs/MODEL_ZOO_cn.md

+1
Original file line numberDiff line numberDiff line change
@@ -153,6 +153,7 @@ Paddle提供基于ImageNet的骨架网络预训练模型。所有预训练模型
153153

154154
- 上表中也提供了原论文[YOLOv3](https://arxiv.org/abs/1804.02767)中YOLOv3-DarkNet53的精度,我们的实现版本主要从在bounding box的宽度和高度回归上使用了L1损失,图像mixup和label smooth等方法优化了其精度。
155155
- YOLO v3在8卡,总batch size为64下训练270轮。数据增强包括:mixup, 随机颜色失真,随机剪裁,随机扩张,随机插值法,随机翻转。YOLO v3在训练阶段对minibatch采用随机reshape,可以采用相同的模型测试不同尺寸图片,我们分别提供了尺寸为608/416/320大小的测试结果。deformable卷积作用在骨架网络5阶段。
156+
- YOLO v3增强版模型通过引入可变形卷积,dropblock和IoU loss,将精度进一步提升至43.2, 详情见[YOLOv3增强模型](./featured_model/YOLOv3_ENHANCEMENT.md)
156157

157158
### RetinaNet
158159

docs/advanced_tutorials/CONFIG.md

+1-1
Original file line numberDiff line numberDiff line change
@@ -184,7 +184,7 @@ A small utility (`tools/configure.py`) is included to simplify the configuration
184184
4. `generate`: Generate a configuration template for a given list of modules. By default it generates a complete configuration file, which can be quite verbose; if a `--minimal` flag is given, it generates a template that only contain non optional settings. For example, to generate a configuration for Faster R-CNN architecture with `ResNet` backbone and `FPN`, run:
185185

186186
```shell
187-
python tools/configure.py generate FasterRCNN ResNet RPNHead RoIAlign BBoxAssigner BBoxHead FasterRCNNTrainFeed FasterRCNNTestFeed LearningRate OptimizerBuilder
187+
python tools/configure.py generate FasterRCNN ResNet RPNHead RoIAlign BBoxAssigner BBoxHead LearningRate OptimizerBuilder
188188
```
189189

190190
For a minimal version, run:

docs/advanced_tutorials/CONFIG_cn.md

+1-1
Original file line numberDiff line numberDiff line change
@@ -174,7 +174,7 @@ pip install typeguard http://github.com/willthefrog/docstring_parser/tarball/mas
174174
4. `generate`: 根据给出的模块列表生成配置文件,默认生成完整配置,如果指定 `--minimal` ,生成最小配置,即省略所有默认配置项。例如,执行下列命令可以生成Faster R-CNN (`ResNet` backbone + `FPN`) 架构的配置文件:
175175

176176
```shell
177-
python tools/configure.py generate FasterRCNN ResNet RPNHead RoIAlign BBoxAssigner BBoxHead FasterRCNNTrainFeed FasterRCNNTestFeed LearningRate OptimizerBuilder
177+
python tools/configure.py generate FasterRCNN ResNet RPNHead RoIAlign BBoxAssigner BBoxHead LearningRate OptimizerBuilder
178178
```
179179

180180
如需最小配置,运行:

docs/advanced_tutorials/inference/INFERENCE.md

+1-1
Original file line numberDiff line numberDiff line change
@@ -2,7 +2,7 @@
22

33
本篇教程使用Python API对[导出模型](EXPORT_MODEL.md)保存的inference_model进行预测。
44

5-
在PaddlePaddle中预测引擎和训练引擎底层有着不同的优化方法,代码走不通的分支,两者都可以进行预测。在入门教程的训练/评估/预测流程中介绍的预测流程,即tools/infer.py是使用训练引擎分支的预测流程。保存的inference_model,可以通过`fluid.io.load_inference_model`接口,走训练引擎分支预测。本文档也同时介绍通过预测引擎的Python API进行预测,一般而言这种方式的速度优于前者。
5+
在PaddlePaddle中预测引擎和训练引擎底层有着不同的优化方法,代码走不同的分支,两者都可以进行预测。在入门教程的训练/评估/预测流程中介绍的预测流程,即tools/infer.py是使用训练引擎分支的预测流程。保存的inference_model,可以通过`fluid.io.load_inference_model`接口,走训练引擎分支预测。本文档也同时介绍通过预测引擎的Python API进行预测,一般而言这种方式的速度优于前者。
66

77

88
这篇教程介绍的Python API预测示例,除了可视化部分依赖PaddleDetection外,预处理、模型结构、执行流程均不依赖PaddleDetection。

ppdet/utils/check.py

+1-1
Original file line numberDiff line numberDiff line change
@@ -55,7 +55,7 @@ def check_version():
5555
"Please make sure the version is good with your code." \
5656

5757
try:
58-
fluid.require_version('1.6.0')
58+
fluid.require_version('1.7.0')
5959
except Exception as e:
6060
logger.error(err)
6161
sys.exit(1)

0 commit comments

Comments
 (0)