-
Notifications
You must be signed in to change notification settings - Fork 333
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Default TPU optimization to much more common values `learning_rate=3e…
…-4`, `batch_size=64` (a common source of errors). This will change the behavior for NSynth training for the original ICLR2020 paper on TPU, but can be reset to 1e-5 and 128 manually. PiperOrigin-RevId: 321403423
- Loading branch information
1 parent
ed3764c
commit dd3fb52
Showing
3 changed files
with
6 additions
and
3 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -1,3 +1,7 @@ | ||
# -*-Python-*- | ||
include 'models/ae.gin' | ||
include 'datasets/nsynth.gin' | ||
|
||
# To recreate original experiment optimization params, uncomment lines below. | ||
# learning_rate = 1e-5 | ||
# batch_size = 128 |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
|
@@ -19,4 +19,4 @@ | |
pulling in all the dependencies in __init__.py. | ||
""" | ||
|
||
__version__ = '0.7.0' | ||
__version__ = '0.8.0' |