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

Initialize training from frozen model #1149

Merged
merged 1 commit into from Jan 8, 2018

Conversation

Projects
None yet
2 participants
@reuben
Copy link
Member

commented Jan 4, 2018

This flag allows one to resume training from a frozen model, but as documented on the flag, it does not restore optimizer parameters, so adjusting the learning rate is necessary. I think it's an OK requirement from users.

@reuben reuben requested a review from kdavis-mozilla Jan 4, 2018

@kdavis-mozilla
Copy link
Collaborator

left a comment

LGTM.

if len(FLAGS.initialize_from_frozen_model) > 0:
log_info('Initializing from frozen model: {}'.format(FLAGS.initialize_from_frozen_model))
feed_dict = {}
model_feeder.set_data_set(feed_dict, model_feeder.train)

This comment has been minimized.

Copy link
@kdavis-mozilla

kdavis-mozilla Jan 5, 2018

Collaborator

Kind of odd that this has to be done as the run below shouldn't need any of the data, but ¯_(ツ)_/¯

@reuben reuben merged commit a03604d into master Jan 8, 2018

1 check passed

Taskcluster (pull_request) TaskGroup: success
Details

@reuben reuben deleted the initialize-from-frozen-model branch Jan 8, 2018

@lock

This comment has been minimized.

Copy link

commented Jan 3, 2019

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.

@lock lock bot locked and limited conversation to collaborators Jan 3, 2019

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
You can’t perform that action at this time.