Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Bug] BEVFusion无法训练 #2792

Open
3 tasks done
AndrewJSong opened this issue Oct 25, 2023 · 12 comments
Open
3 tasks done

[Bug] BEVFusion无法训练 #2792

AndrewJSong opened this issue Oct 25, 2023 · 12 comments

Comments

@AndrewJSong
Copy link

Prerequisite

Task

I'm using the official example scripts/configs for the officially supported tasks/models/datasets.

Branch

main branch https://github.com/open-mmlab/mmdetection3d

Environment

mmcv 2.0.1
mmdet 3.0.0rc5
mmdet3d 1.3.0
mmengine 0.9.0
mmsegmentation 1.0.0

Reproduces the problem - code sample

bash tools/dist_train.sh projects/BEVFusion/configs/bevfusion_lidar_voxel0075_second_secfpn_8xb4-cyclic-20e_nus-3d.py 1

Reproduces the problem - command or script

bash tools/dist_train.sh projects/BEVFusion/configs/bevfusion_lidar_voxel0075_second_secfpn_8xb4-cyclic-20e_nus-3d.py 1

Reproduces the problem - error message

File "/share/temp/mmdetection3d/mmdet3d/models/data_preprocessors/data_preprocessor.py", line 118, in init
batch_augments=batch_augments)
TypeError: init() got an unexpected keyword argument 'non_blocking'

直接注释掉 non_blocking=non_blocking;
新的报错:
File "/share/temp/mmdetection3d/projects/BEVFusion/bevfusion/transformer.py", line 30, in init
super().init(**kwargs)
TypeError: init() missing 2 required positional arguments: 'attn_cfgs' and 'feedforward_channels'

Additional information

bug真多

@xuehaipiaoxiang
Copy link

小弟 你解决了吗 23333333333333

@Gigalomanicx
Copy link

那个pkl是按照2.0生成的吗? 没改动的话(又是data_preprocessor出问题)应该是数据有问题?
我的BEVFusion复现成功了,有问题可以联系。

@seven-linglx
Copy link

那个pkl是按照2.0生成的吗? 没改动的话(又是data_preprocessor出问题)应该是数据有问题? 我的BEVFusion复现成功了,有问题可以联系。

pkl没有按照2.0的生成。请问具体怎么解决的?或者给个解决博客?

@seven-linglx
Copy link

seven-linglx commented Dec 29, 2023

我重新适配一下版本号解决了。
之前出问题时安装的版本号
mmcv 2.0.0rc4
mmdet 3.0.0rc5
mmdet3d 1.3.0
mmengine 0.10.1
更新了mmdet与mmengine的版本
mmdet 3.0.0rc5 -> mmdet 3.0.0rc6
mmengine 0.10.1 -> mmengine 0.10.2

@ddk1246
Copy link

ddk1246 commented Jan 3, 2024

@Gigalomanicx 你好,你复现出论文71.4 NDS (val) 的精度了吗?

@Mingqj
Copy link

Mingqj commented Jan 12, 2024

@ddk1246 您好,请问您复现的NDS (val)是多少呢

@YangChen1234567
Copy link

我重新适配一下版本号解决了。 之前出问题时安装的版本号 mmcv 2.0.0rc4 mmdet 3.0.0rc5 mmdet3d 1.3.0 mmengine 0.10.1 更新了mmdet与mmengine的版本 mmdet 3.0.0rc5 -> mmdet 3.0.0rc6 mmengine 0.10.1 -> mmengine 0.10.2

hello,请问你复现mmdet里bevfusion的性能了吗?

@ddk1246
Copy link

ddk1246 commented Feb 16, 2024

@ddk1246 您好,请问您复现的NDS (val)是多少呢

我只复现出了LiDAR-Only 68.8NDS(val),LiDAR-Camera 69.3NDS(val)。没有复现出本仓库在README中标识的71.4。

@YangChen1234567
Copy link

没有,我刚才在issue上回复了,您可以看一下   小手冰凉 @.***  

------------------ 原始邮件 ------------------ 发件人: @.>; 发送时间: 2024年2月16日(星期五) 下午2:49 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [open-mmlab/mmdetection3d] [Bug] BEVFusion无法训练 (Issue #2792) 我重新适配一下版本号解决了。 之前出问题时安装的版本号 mmcv 2.0.0rc4 mmdet 3.0.0rc5 mmdet3d 1.3.0 mmengine 0.10.1 更新了mmdet与mmengine的版本 mmdet 3.0.0rc5 -> mmdet 3.0.0rc6 mmengine 0.10.1 -> mmengine 0.10.2 hello,请问你复现mmdet里bevfusion的性能了吗? — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you were mentioned.Message ID: @.***>

好的,已看到,谢谢~

@Mingqj
Copy link

Mingqj commented Feb 22, 2024

@ddk1246 您好,请问您复现的NDS (val)是多少呢

我只复现出了LiDAR-Only 68.8NDS(val),LiDAR-Camera 69.3NDS(val)。没有复现出本仓库在README中标识的71.4。

我复现的LiDAR-Camera是70.78 NDS,也没有达到71.4 NDS

@lotusdaddy
Copy link

Hi, I also met this error:
TypeError: init() missing 2 required positional arguments: 'attn_cfgs' and 'feedforward_channels'
Can you tell me how do you fix it?

@JiankunShi
Copy link

@ddk1246 您好,请问您复现的NDS (val)是多少呢

我只复现出了LiDAR-Only 68.8NDS(val),LiDAR-Camera 69.3NDS(val)。没有复现出本仓库在README中标识的71.4。

我复现的LiDAR-Camera是70.78 NDS,也没有达到71.4 NDS

我复现的LiDAR-Camera,NDS是69.28,MAP是64.55,请问你的config是怎么设置的?我用的是源码提供的config,学习率和epoch都没改,只把batch-size从4改成了3。

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

9 participants