31 Commits

Author	SHA1	Message	Date
Haoming	b3717313cc	Anima	2026-02-11 00:45:36 +08:00
Speeskees	68ede9f1a5	sync stream	2026-01-31 22:08:23 +08:00
Haoming	0e446f9c5b	mem	2026-01-21 17:01:13 +08:00
Haoming	9921cf0e74	Rewrite Backend (#559)	2026-01-19 17:11:59 +08:00
Haoming	6a4dee6794	update dynamic_args	2025-11-26 14:54:54 +08:00
Haoming	4e0e5e5483	9 to 4	2025-10-21 10:01:57 +08:00
Haoming	1518c77273	flux-kontext ref_latents	2025-10-08 18:02:46 +08:00
Haoming	286f830759	model_options	2025-10-08 16:11:06 +08:00
Haoming	689eb76f4d	cfgpp	2025-09-09 10:12:18 +08:00
Haoming	3a8b6470c6	mem	2025-09-04 14:50:14 +08:00
Haoming	04cf6f699d	sampling	2025-09-01 15:21:35 +08:00
Haoming	6afb284c60	scripts	2025-08-05 18:00:46 +08:00
Haoming	129faabc4c	kontext	2025-08-04 11:28:14 +08:00
Haoming	6dc921ecfc	Kontext Support	2025-08-04 11:04:19 +08:00
Haoming	3b84a5d299	lint	2025-07-28 16:43:27 +08:00
Haoming	115f93ecbc	lint	2025-07-28 16:41:35 +08:00
Panchovix	c13b26ba27	Rephrase low GPU warning (#1761) ... Make emphasis that it's performance degradation, and for this diffusion process.	2024-09-09 14:30:22 -03:00
layerdiffusion	33963f2d19	always compute on-the-fly lora weights when offload	2024-08-31 11:24:23 -07:00
layerdiffusion	d1d0ec46aa	Maintain patching related ... 1. fix several problems related to layerdiffuse not unloaded 2. fix several problems related to Fooocus inpaint 3. Slightly speed up on-the-fly LoRAs by precomputing them to computation dtype	2024-08-30 15:18:21 -07:00
layerdiffusion	2ab19f7f1c	revise lora patching	2024-08-22 11:59:43 -07:00
layerdiffusion	14eac6f2cf	add a way to empty cuda cache on the fly	2024-08-22 10:06:39 -07:00
layerdiffusion	0d8eb4c5ba	fix #1375	2024-08-21 11:01:59 -07:00
layerdiffusion	475524496d	revise	2024-08-19 18:54:54 -07:00
layerdiffusion	d7151b4dcd	add low vram warning	2024-08-19 11:08:01 -07:00
layerdiffusion	d38e560e42	Implement some rethinking about LoRA system ... 1. Add an option to allow users to use UNet in fp8/gguf but lora in fp16. 2. All FP16 loras do not need patch. Others will only patch again when lora weight change. 3. FP8 unet + fp16 lora are available (somewhat only available) in Forge now. This also solves some “LoRA too subtle” problems. 4. Significantly speed up all gguf models (in Async mode) by using independent thread (CUDA stream) to compute and dequant at the same time, even when low-bit weights are already on GPU. 5. View “online lora” as a module similar to ControlLoRA so that it is moved to GPU together with model when sampling, achieving significant speedup and perfect low VRAM management simultaneously.	2024-08-19 04:31:59 -07:00
layerdiffusion	396d9f378d	we do not need to waste 10 seconds on T5 when CFG=1	2024-08-08 00:19:16 -07:00
lllyasviel	a6baf4a4b5	revise kernel ... and add unused files	2024-08-07 16:51:24 -07:00
lllyasviel	a07c758658	intergrate k-diffusion	2024-08-07 15:05:42 -07:00
layerdiffusion	b7878058f9	improve backward combability #936	2024-08-06 01:06:24 -07:00
layerdiffusion	0863765173	rework sd1.5 and sdxl from scratch	2024-08-05 03:08:17 -07:00
layerdiffusion	bb5083f3c2	rework sample function	2024-08-03 13:27:23 -07:00