We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
There was an error while loading. Please reload this page.
1 parent 1e304ae commit 98ea1beCopy full SHA for 98ea1be
docs/changelogs/v3.4.2.md
@@ -5,10 +5,20 @@
5
* Implement `SCION` optimizer. (#348, #352)
6
* [Training Deep Learning Models with Norm-Constrained LMOs](https://arxiv.org/abs/2502.07529)
7
8
+### Update
9
+
10
+* Update ScheduleFreeSGD, AdamW, RAdam optimizers with the latest. (#351, #353)
11
+* Remove `use_palm` variant in ScheduleFree optimizer due to instability. (#353)
12
+* Ranger25 optimizer. (#353)
13
14
+### Fix
15
16
+* Remove `weight decouple` parameter in ScheduleFree optimizers. (#351, #353)
17
18
### Docs
19
20
* Fix `AliG` optimizer visualization. (#350)
21
22
### Contributions
23
-thanks to @AidinHamedi
24
+thanks to @AidinHamedi, @hatonosuke
0 commit comments