Skip to content
This repository has been archived by the owner on Oct 31, 2023. It is now read-only.

how to set multi-scale training? #88

Closed
zimenglan-sysu-512 opened this issue Nov 1, 2018 · 3 comments
Closed

how to set multi-scale training? #88

zimenglan-sysu-512 opened this issue Nov 1, 2018 · 3 comments

Comments

@zimenglan-sysu-512
Copy link
Contributor

❓ Questions and Help

when i set MIN_SIZE_TRAIN: (600,700,800,900,1000), i met the problem: ValueError: Type mismatch (<class 'int'> vs. <class 'tuple'>) with values (800 vs. (600, 700, 800, 900, 1000)) for config key: INPUT.MIN_SIZE_TRAIN, so how to set the yml file to multi-scale trainging?

thanks.

@fmassa
Copy link
Contributor

fmassa commented Nov 1, 2018

Multi-scale training support will be added in this PR #69

If you need it for now, you could try cherry-pick the relevant changes there, or directly work from that branch

@zimenglan-sysu-512
Copy link
Contributor Author

thanks @fmassa

@JoyHuYY1412
Copy link

Multi-scale training support will be added in this PR #69

If you need it for now, you could try cherry-pick the relevant changes there, or directly work from that branch

Hi, is this multi-scale training also support mask rcnn?

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

No branches or pull requests

3 participants