925fff26fd
alternative to `load_checkpoint_guess_config` that accepts a loaded state dict ( #4249 )
...
* make alternative fn
* add back ckpt path as 2nd argument?
7 months ago
75b9b55b22
Fix issues with #4302 and support loading diffusers format flux.
7 months ago
1765f1c60c
FLUX: Added full diffusers mapping for FLUX.1 schnell and dev. Adds full LoRA support from diffusers LoRAs. ( #4302 )
7 months ago
1de69fe4d5
Fix some issues with inference slowing down.
7 months ago
ae197f651b
Speed up hunyuan dit inference a bit.
7 months ago
1b5b8ca81a
Fix regression.
7 months ago
6678d5cf65
Fix regression.
7 months ago
e172564eea
Update controlnet.py to fix the default controlnet weight as constant ( #4285 )
7 months ago
a3cc326748
Better fix for lowvram issue.
7 months ago
86a97e91fc
Fix controlnet regression.
7 months ago
5acdadc9f3
Fix issue with some lowvram weights.
7 months ago
55ad9d5f8c
Fix regression.
7 months ago
a9f04edc58
Implement text encoder part of HunyuanDiT loras.
7 months ago
a475ec2300
Cleanup HunyuanDit controlnets.
...
Use the: ControlNetApply SD3 and HunyuanDiT node.
7 months ago
06eb9fb426
feat: add support for HunYuanDit ControlNet ( #4245 )
...
* add support for HunYuanDit ControlNet
* fix hunyuandit controlnet
* fix typo in hunyuandit controlnet
* fix typo in hunyuandit controlnet
* fix code format style
* add control_weight support for HunyuanDit Controlnet
* use control_weights in HunyuanDit Controlnet
* fix typo
7 months ago
413322645e
Raw torch is faster than einops?
7 months ago
11200de970
Cleaner code.
7 months ago
037c38eb0f
Try to improve inference speed on some machines.
7 months ago
1e11d2d1f5
Better prints.
7 months ago
66d4233210
Fix.
7 months ago
591010b7ef
Support diffusers text attention flux loras.
7 months ago
08f92d55e9
Partial model shift support.
7 months ago
8115d8cce9
Add Flux fp16 support hack.
7 months ago
6969fc9ba4
Make supported_dtypes a priority list.
7 months ago
cb7c4b4be3
Workaround for lora OOM on lowvram mode.
7 months ago
1208863eca
Fix "Comfy" lora keys.
...
They are in this format now:
diffusion_model.full.model.key.name.lora_up.weight
7 months ago
e1c528196e
Fix bundled embed.
7 months ago
17030fd4c0
Support for "Comfy" lora format.
...
The keys are just: model.full.model.key.name.lora_up.weight
It is supported by all comfyui supported models.
Now people can just convert loras to this format instead of having to ask
for me to implement them.
7 months ago
c19dcd362f
Controlnet code refactor.
7 months ago
1c08bf35b4
Support format for embeddings bundled in loras.
7 months ago
b334605a66
Fix OOMs happening in some cases.
...
A cloned model patcher sometimes reported a model was loaded on a device
when it wasn't.
7 months ago
c14ac98fed
Unload models and load them back in lowvram mode no free vram.
7 months ago
2d75df45e6
Flux tweak memory usage.
7 months ago
8edbcf5209
Improve performance on some lowend GPUs.
7 months ago
a178e25912
Fix Flux FP64 math on XPU ( #4210 )
7 months ago
78e133d041
Support simple diffusers Flux loras.
7 months ago
7afa985fba
Correct spelling 'token_weight_pars_t5' to 'token_weight_pairs_t5' ( #4200 )
7 months ago
3b71f84b50
ONNX tracing fixes.
7 months ago
0a6b008117
Fix issue with some custom nodes.
7 months ago
f7a5107784
Fix crash.
7 months ago
91be9c2867
Tweak lowvram memory formula.
7 months ago
03c5018c98
Lower lowvram memory to 1/3 of free memory.
7 months ago
2ba5cc8b86
Fix some issues.
7 months ago
1e68002b87
Cap lowvram to half of free memory.
7 months ago
ba9095e5bd
Automatically use fp8 for diffusion model weights if:
...
Checkpoint contains weights in fp8.
There isn't enough memory to load the diffusion model in GPU vram.
7 months ago
f123328b82
Load T5 in fp8 if it's in fp8 in the Flux checkpoint.
7 months ago
63a7e8edba
More aggressive batch splitting.
7 months ago
ea03c9dcd2
Better per model memory usage estimations.
7 months ago
3a9ee995cf
Tweak regular SD memory formula.
7 months ago
47da42d928
Better Flux vram estimation.
7 months ago
ljleb
comfyanonymous
Jaret Burkett
TTPlanetPig
来新璐
a-One-Fan
Silver