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

nuImage relation parameters #401

Closed
Shaoqing26 opened this issue Apr 1, 2021 · 2 comments
Closed

nuImage relation parameters #401

Shaoqing26 opened this issue Apr 1, 2021 · 2 comments

Comments

@Shaoqing26
Copy link

a general question of nuImage models,:
HTC | R-50 | IN+COCO-20e,is that means backbone is first pre-trained on ImageNet, and then the detector is pre-trained on COCO train2017 dataset by 20 epochs?and how many epochs is the model training on nuImage dataset?

AND Is there has any plan to fix it and prove traning code of nuImage dataset,I found that there just has inference code in project,and also have no evaluate code,is there any plan to complete it?

@ZwwWayne
Copy link
Collaborator

ZwwWayne commented Apr 3, 2021

Yes, the number of epochs on nuImages dataset is indicated by the lr schedule column.

@ZwwWayne ZwwWayne closed this as completed Apr 3, 2021
@yezhen17
Copy link
Collaborator

yezhen17 commented Apr 3, 2021

@Shaoqing26

You can take a look at the tutorials to understand the configs.

For example, htc_r50_fpn_1x_nuim.py inherits from htc_without_semantic_r50_fpn_1x_nuim.py, which inherits the learning schedule from mmdet_schedule_1x.py, where you can find the answer to your first question.

As for your second question, please refer to #402 .

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

3 participants