1a1f97263d
CHORE Replace deprecated torch_dtype with dtype ( #2837 )
...
Note: Diffusers is left as is for now, might need an update later.
2025-10-16 14:59:09 +02:00
9b8cf2a0c3
FIX Handle embed scale for trainable tokens, LoRA ( #2825 )
...
Resolves #2809
Some models like Gemma3 apply a scalar to the embedding output. It needs
to be taken into account when using trainable tokens or LoRA applied to
the embedding layer.
2025-10-14 12:35:31 +02:00
49b29c1d1a
ENH XPU support for boft/controlnet example ( #2674 )
...
---------
Signed-off-by: Liu, Kaixuan <kaixuan.liu@intel.com >
2025-08-04 11:15:36 +02:00
b3130c9edb
Use HF Papers ( #2542 )
...
Replaced all arxiv.org/pdf links with HF papers.
2025-05-27 13:48:53 +02:00
09358aad30
Chore: Docs markdown formatting ( #1899 )
2024-07-03 18:12:53 +02:00
811169939f
BOFT: Orthogonal Finetuning via Butterfly Factorization ( #1326 )
...
Implements https://hf.co/papers/2311.06243 .
---------
Co-authored-by: Zeju Qiu <zeju.qiu@gmail.com >
Co-authored-by: Yuliang Xiu <yuliangxiu@gmail.com >
Co-authored-by: Yao Feng <yaofeng1995@gmail.com >
2024-04-12 13:04:09 +02:00
