This embedding should be used in your NEGATIVE prompt.
Description
FAQ
Comments (11)
does this give more brightness or what? :-)
it's a cheap instant replacement for 20 words of negative prompt, it should give some realism and the look and feel of "a real vintage camera (something like Leica M TYP 240)" :)
Thank you, I seen this the other day was interested to try it. Would this go with the other yam negative embed? Also the model is very good, thank you!
thx! yes it can definitely be used along with yam-negative (example: https://civitai.com/posts/231840)
please a .safetensors version
it's a textual inversion not a checkpoint. They're either a .bin or .pt depending on the software used to train them.
but easynegative also got .safetensors
look:https://civitai.com/models/7808?modelVersionId=9208
im confused!
@hiddenmail4life810 hell idk then haha. I use google Colab to train them myself and it doesn’t give that option, not sure how people are making safetensor embeddings.
@hiddenmail4life810 according to (https://github.com/kohya-ss/sd-scripts/blob/main/train_textual_inversion.py line 538) this is quite possible, problem is that I'm using AUTOMATIC1111 for training embeddings, not kohya_ss, so i'm sorry can't do that in any near future
@jstnbdy and @Fenn thx for replying and explaining.
This is a pretty good one, having good results so far. Only recommendations to users is that.. certain models needs 0.7/0.8 instead of 1.0, need to adjust to get the results you want. overall good quality.