Open mzwing opened 3 months ago
./sd -M convert -m Anything/anything_v30.ckpt --lora-model-dir lcm-lora-sdv1-5 -o Anything-GGUF/anything_v30_LCM.Q4_0.gguf -v --type q4_0
If I run the command above, it will neither merge the lora into the converted model nor convert it into gguf.
I want to merge the lora into the model and then convert the new model. Are there any ways to do that?
idk whether it is a good way to use LCM-LoRA :(
Env:
stable-diffusion.cpp version: sd-master-583cc5b with cublas and avx2
Base model link: https://civitai.com/models/9409?modelVersionId=34373
LoRA link: https://huggingface.co/latent-consistency/lcm-lora-sdv1-5/
Running on Google Colab (T4 GPU).
(Sorry for my poor English.)
If I run the command above, it will neither merge the lora into the converted model nor convert it into gguf.
I want to merge the lora into the model and then convert the new model. Are there any ways to do that?
idk whether it is a good way to use LCM-LoRA :(
Env:
stable-diffusion.cpp version: sd-master-583cc5b with cublas and avx2
Base model link: https://civitai.com/models/9409?modelVersionId=34373
LoRA link: https://huggingface.co/latent-consistency/lcm-lora-sdv1-5/
Running on Google Colab (T4 GPU).
(Sorry for my poor English.)