lora

lora copied to clipboard

Un-inject values from Unet

11

Right now, if we save_pretrained with the unet modifed for lora training, diffusion pipeline will refuse to load the unet later because of some "Linear" values. Is there a way...

d8ahazard

RuntimeError: Input type (c10::Half) and bias type (float) should be the same

5

First of all, congrats on the great work! I got this error in the middle of training on T4 in a colab: ``` ***** Running training ***** Num examples =...

qunash

Config.json not found

1

I got this error: `EntryNotFoundError(f"404 Client Error: Entry Not Found for url: {response.url}") transformers.utils.hub.EntryNotFoundError: 404 Client Error: Entry Not Found for url: https://huggingface.co/stabilityai/stable-diffusion-2-1-base/resolve/main/config.json ` I have manually confirmed that the...

invisiblestrangler

your calculation of time is based on steps, but you need to multiply steps times accumulations to get the correct eta estimate, so currently accumulation 2, and steps 1000, it...

Thomas-MMJ

would be nice to specify a checkpoint being generated ever 100 or 200 or whatever steps

Thomas-MMJ

I seem to be one of the few to get it working here are some thoughts and issues 1) might require bleeding edge diffusers and transformers and accelerate - i...

Thomas-MMJ

Everything is Working Great, quick question....

27

as seen here, (example outputs included) https://twitter.com/MushroomFleet/status/1602341447952437249 and below (just the merge experiments) What are your advised Training settings? I did use a bunch that i used in other Dreambooth...

MushroomFleet

Text encoder still not working correctly with LoRa Dreambooth training script

28

Hello, I am getting much better results using the `--train_text_encoder` flag with the Dreambooth script. However, the actual outputed LoRa `.pt` files from models trained with `train_text_encoder` gives very bad...

JohnnyRacer

hi, do you think you could take a look if it would be possible to train depth2image models with lora?

batrlatom

Gradient checkpointing blocks LoRA weight updates

14

Running with gradient checkpointing prevents LoRA weight updates. > Description: Ubuntu 18.04.6 LTS > diffusers==0.10.2 > lora-diffusion==0.0.3 > torchvision @ https://download.pytorch.org/whl/cu116/torchvision-0.14.0%2Bcu116-cp38-cp38-linux_x86_64.whl > transformers==4.25.1 > xformers @ https://github.com/brian6091/xformers-wheels/releases/download/0.0.15.dev0%2B4c06c79/xformers-0.0.15.dev0+4c06c79.d20221205-cp38-cp38-linux_x86_64.whl > > `Accelerate`...

brian6091