-
您好，这是一个非常有趣的工作，我想把他在我现在的任务上进行实验。我现在任务是Vicuna7B训练得到的。
请问如何合并呢？我看您的步骤是否可以按以下步骤：
Step 1: 将我训练的模型，得到DARE模型。
python inference_plms_glue.py --language_model_name roberta-base --weight_mask_rate 0.9 -…
-
HI, would it be possible for you to add other samplers that ComfyUI comes with, specifically Uni_pc_bh2. Also to be able to change to a different scheduler independently , like ddim_uniform or Karras
-
Can we get even more performance if we quantize the models first? If we can, it would be nice to add it. Thanks!
-
hello @GuiyeC
The apple/ml-stable-diffusion project has added SDXL support to DPMSolverMultistepScheduler some time ago.
[https://github.com/apple/ml-stable-diffusion/commit/94814cfa41935efd8151a43…
-
**LocalAI version:**
2.0.0-ffmpeg
**Environment, CPU architecture, OS, and Version:**
Linux srv-gpt 5.15.131-1-pve #1 SMP PVE 5.15.131-2 (2023-11-14T11:32Z) x86_64 x86_64 x86_64 GNU/Linux, Proxmo…
-
ever time i run a new image it adds one more "You shouldn't move a model when it is dispatched on multiple devices."
_From CMD_
100%|█████████| 5/5 [00:05
-
I know it's not easy but is there a change that you add TensorRT support for Fooocus?
-
Hi,
Thank you for such great efforts. I saw that there is a sample finetuning script for SDXL on huggingface (https://github.com/huggingface/diffusers/blob/main/examples/text_to_image/train_text_to…
-
/tmp/ccFr850C.s:9961: Error: selected processor does not support `vudot.u8 q4,q8,d4[0]' in ARM mode...
2.33 coupled with gcc versions 9-11.
ive tried your git checkout for xnnpack and the one re…
-
Seems like a silly question but I just can't find where to set the amount of steps and CFG. I see speed, quality, extreme speed, which seem to change the steps but I would like to put in a specific am…