Muhammad Rizqi Nur

> no error with [new xformers build](https://github.com/camenduru/stable-diffusion-webui-colab/releases/tag/0.0.14) you can use like this `!pip install -qq https://github.com/camenduru/stable-diffusion-webui-colab/releases/download/0.0.14/xformers-0.0.14.dev0-cp37-cp37m-linux_x86_64.whl` Thank you. Your wheel works for google colab. I can't build from source with...

> This can be closed now. Just add --xformers to command line args. --xformers won't automatically install xformers on google colab, at least for me #3259

Idk what caused it, but some people say it's fixed with newer xformer. I can't build xformer on colab though, it just stucks. And the new version isn't available on...

> Auto1111 seems to install xformers 0.0.12 instead of 0.0.13 so that's my issue? Does anyone know where I could change the line that install xformers? xformers 0.0.13+ on pip...

> The way that loss is calculated is it slightly adds noise to an image, then asks the AI to guess what parts are noise and should be removed (which...

Wait, the model and image preview are logged every x step instead of epoch. So they're logged and saved in the middle of an epoch? It's understandable for training the...

I just realized that the loss in CSV is set in the settings. It's 100 for me, might be 32 for you. If I set this and the model log...

Wait, no. It's true that the loss is now logged every 1 epoch, but it's late by 1 step. It seems to be logging the very first 1 step, then...

To save free space, you can use symbolic links instead of actually duplicating VAEs.

+1 Embedding selection would be very helpful for embedding training, for trying the logged models. After having trained models logged, I'd pick ones that seem good, then try them out...