transformers
Add z-loss to Bamba for v2
#37842
Merged

Add z-loss to Bamba for v2 #37842

daviswer
daviswer Remove const
6989543b
daviswer Fix arg ref
f8878532
daviswer Sharded save
b4e73f72
daviswer Merge branch 'huggingface:main' into main
5e29bdf4
daviswer Add z_loss flag
11a4dee4
daviswer Add modeling zloss
570f9a3d
daviswer Demodularize clm forward for zloss
11406796
daviswer Also demodularize init for z_loss flag
aa4ec14f
fabianlim
fabianlim commented on 2025-04-28
fabianlim
fabianlim commented on 2025-04-28
fabianlim
fabianlim commented on 2025-04-28
fabianlim
fabianlim commented on 2025-04-28
vasqu
vasqu commented on 2025-04-28
daviswer PR comments (mostly modularizing right)
5ebda1f2
daviswer Demodularize forward
26bd7d90
daviswer
vasqu
vasqu commented on 2025-05-07
daviswer Better name zloss and explain typematch
210613a5
vasqu
vasqu commented on 2025-05-07
vasqu
daviswer Fully propagate coeff name
f321778e
daviswer style fixes
aad9194f
vasqu
vasqu
vasqu commented on 2025-05-07
vasqu
vasqu commented on 2025-05-07
daviswer zloss default float
b1d3a735
vasqu
daviswer daviswer marked this pull request as ready for review 1 year ago
daviswer Remove conflicting annotations
109e7727
daviswer Merge branch 'main' into bamba-add-zloss
1547ccdf
Cyrilvallez
Cyrilvallez approved these changes on 2025-05-28
Cyrilvallez Cyrilvallez enabled auto-merge (squash) 1 year ago
Cyrilvallez Merge branch 'main' into bamba-add-zloss
f9e01335
HuggingFaceDocBuilderDev
fabianlim
disabled auto-merge 1 year ago
Manually disabled by user
Cyrilvallez
Cyrilvallez Merge branch 'main' into bamba-add-zloss
ee01e0fc
Cyrilvallez Cyrilvallez merged a6f0e2b6 into main 1 year ago
daviswer daviswer deleted the bamba-add-zloss branch 1 year ago

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone