Mmdetection: What are the differences between 'load_from' and 'resume_from' in config files?

[yeliudev]

For example, in configs/faster_rcnn_r50_fpn_1x.py, we have

load_from = None
resume_from = None

I have no idea what are the differences between these two properties.

Anyway, thx a lot for the great work! 😊

[mosheliv]

from the GETTING_STARTED.md:

Difference between resume_from and load_from: resume_from loads both the
model weights and optimizer status, and the epoch is also inherited from
the specified checkpoint. It is usually used for resuming the training
process that is interrupted accidentally. load_from only loads the model
weights and the training epoch starts from 0. It is usually used for
finetuning.

On Wed, Jul 17, 2019 at 2:44 PM Ye Liu [e-mail] wrote:

For example, in configs/faster_rcnn_r50_fpn_1x.py, we have
load_from = None resume_from = None
I have no idea what are the differences between these two properties.
Anyway, thx a lot for the great work!

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
https://github.com/open-mmlab/mmdetection/issues/1012?email_source=notifications&email_token=AC7IWC5EGCTDWHO2UCSKGBLP72BSJA5CNFSM4IEJ73PKYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4G7TXDJQ,
or mute the thread
https://github.com/notifications/unsubscribe-auth/AC7IWCZ5V6HWDUCBCYJ7YM3P72BSJANCNFSM4IEJ73PA
.

[yeliudev]

from the GETTING_STARTED.md: Difference between resume_from and load_from: resume_from loads both the model weights and optimizer status, and the epoch is also inherited from the specified checkpoint. It is usually used for resuming the training process that is interrupted accidentally. load_from only loads the model weights and the training epoch starts from 0. It is usually used for finetuning.
…
On Wed, Jul 17, 2019 at 2:44 PM Ye Liu @.*> wrote: For example, in configs/faster_rcnn_r50_fpn_1x.py, we have load_from = None resume_from = None I have no idea what are the differences between these two properties. Anyway, thx a lot for the great work! — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#1012?email_source=notifications&email_token=AC7IWC5EGCTDWHO2UCSKGBLP72BSJA5CNFSM4IEJ73PKYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4G7TXDJQ>, or mute the thread https://github.com/notifications/unsubscribe-auth/AC7IWCZ5V6HWDUCBCYJ7YM3P72BSJANCNFSM4IEJ73PA .

Got it. I missed the message in GETTING_STARTED.md.
Thank you very much!