Commit Graph

7 Commits

Author SHA1 Message Date
rattus
c0370044cd MPDynamic: force load flux img_in weight (Fixes flux1 canny+depth lora crash) (#12446)
* lora: add weight shape calculations.

This lets the loader know if a lora will change the shape of a weight
so it can take appropriate action.

* MPDynamic: force load flux img_in weight

This weight is a bit special, in that the lora changes its geometry.
This is rather unique, not handled by existing estimate and doesn't
work for either offloading or dynamic_vram.

Fix for dynamic_vram as a special case. Ideally we can fully precalculate
these lora geometry changes at load time, but just get these models
working first.
2026-02-15 20:30:09 -05:00
Kohaku-Blueleaf
a97c98068f [Weight-adapter/Trainer] Bypass forward mode in Weight adapter system (#11958)
* Add API of bypass forward module

* bypass implementation

* add bypass fwd into nodes list/trainer
2026-01-24 22:56:22 -05:00
comfyanonymous
bdb10a583f Fix loras not working on mixed fp8. (#10899) 2025-11-26 00:07:58 -05:00
Kohaku-Blueleaf
7be2b49b6b Fix LoRA Trainer bugs with FP8 models. (#9854)
* Fix adapter weight init

* Fix fp8 model training

* Avoid inference tensor
2025-09-20 21:24:48 -04:00
flybirdxx
4c3e57b0ae Fixed an issue where qwenLora could not be loaded properly. (#9208) 2025-08-06 13:23:11 -04:00
Kohaku-Blueleaf
520eb77b72 LoRA Trainer: LoRA training node in weight adapter scheme (#8446) 2025-06-13 19:25:59 -04:00
Kohaku-Blueleaf
1f3fba2af5 Unified Weight Adapter system for better maintainability and future feature of Lora system (#7540) 2025-04-21 20:15:32 -04:00