[Fix] Fix Enet export and infer problem. #1919
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Collaborator
shiyutang
commented
Mar 24, 2022
shiyutang
commented
- Add docs about set input_shape.
- Add cityscape_val.list to decrease the infer time of enet.
|
The inference time is as follows by running prepare.sh and test_train_infer_python.sh |
|
一次cpu预测就要2.5min，整个lite_train_lite_infer跑下来需要多长时间呢？会不会超过15min呢？ |
LutaoChu
requested changes
Mar 25, 2022
| @@ -16,7 +16,7 @@ | |||
|
|
|||
| 在终端输入以下命令进行预测: | |||
| ```shell | |||
| python deploy/python/infer.py --config /path/to/deploy.yaml --image_path | |||
| python deploy/python/infer.py --config /path/to/deploy.yaml --image_path # 如果导出指定了input_shape， 此处的图片大小需要和 input_shape 保持一致 | |||
There was a problem hiding this comment.
如果deploy.yml的transform可以保证输入图片尺寸一样也是可以的。『此处的图片大小』写得有点模糊，到底是哪处呢
| @@ -30,6 +30,7 @@ python export.py \ | |||
| --config configs/bisenet/bisenet_cityscapes_1024x1024_160k.yml \ | |||
| --model_path bisenet/model.pdparams\ | |||
| --save_dir output | |||
| --input_shape 1 3 512 1024 | |||
There was a problem hiding this comment.
这个是通过示例说明固定输入尺寸的方式，不代表bisenet需要固定
docs/model_export_cn.md
Outdated
| @@ -38,10 +39,12 @@ python export.py \ | |||
| |config|配置文件|是|-| | |||
| |model_path|预训练权重的路径|否|配置文件中指定的预训练权重路径| | |||
| |save_dir|保存预测模型的路径|否|output| | |||
| |input_shape| 设置导出模型的输入shape，比如传入`--input_shape 1 3 1024 1024`。如果不设置input_shape，默认导出模型的输入shape是[-1, 3, -1, -1] | 否 | None | | |||
| |input_shape| 设置导出模型的输入shape，比如传入`--input_shape 1 3 1024 1024`。如果不设置input_shape，默认导出模型的输入shape是[-1, 3, -1, -1] | 否（最好指定） | None | | |||
There was a problem hiding this comment.
改成 如果预测shape固定，建议指定input_shape参数
test_tipc/prepare.sh
Outdated
| @@ -75,3 +75,7 @@ else | |||
|
|
|||
| fi | |||
| fi | |||
|
|
|||
| if [ ${model_name} == "enet" ];then | |||
| mv ./test_tipc/data/cityscapes_val_5.list ./test_tipc/data/cityscapes | |||
|
训练、评估、导出都比较快，单个 infer 时间2.5分钟。 |
整个流程还包括多个cpu infer和gpu infer，看下整体耗时满不满足15min把 |
cpu infer比较慢，gpu快很多 |
|
根据日志，训练到最后一次预测，一共耗时十分钟： |
LutaoChu
approved these changes
Mar 29, 2022
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.