Suraj Patil comments

Results 240 comments of


                                            Suraj Patil

[Dreambooth] Add a function to log interim generation

Adding this to my todo list.

Training dreambooth on multiple GPUs with deepspeed

From the stack trace, it looks like deepspeed is not installed correctly, Could you try re-installing it ? Also, note that deepspeed support is experimental, I would recommend using `xformers`...

AttributeError: Can't pickle local object 'main.<locals>.<lambda>'

Similar issue https://github.com/huggingface/diffusers/issues/2041 and fix #2070

Strange behaviour and issues with SD 2.1 on multi-gpu (4090) and xformers attention

@linyu0219 what GPU did you use, 4090 or something else? Also would be nice if you could share the code snippet.

[textual_inversion] Add an option for only saving the embeddings

Gently pinging @allo- , let me know if I can help here :)

Training memory optimizations not working on AMD hardware

I'm not really familiar with AMD GPUs, maybe @NouamaneTazi has some ideas here :)

[Feature Request] Dreambooth - Save intermediate checkpoints

The script is now updated to save intermediate checkpoints cf https://github.com/huggingface/diffusers/blob/main/examples/dreambooth/train_dreambooth.py#L675 just pass the `--save_steps` argument to the command. Will close this issue now.

Support for multiple fine-tuned concepts & classes on one model with Dreambooth

Sorry to be so late here, don't have much time to work on it at the moment, but will revisit next week. But if someone wants to open a PR...

Optimize VRAM use in textual inversion training

Sorry for being late again, I've posted instructions in this comment https://github.com/huggingface/diffusers/pull/687#pullrequestreview-1165354545 @Ttl LMK if you are busy, then I'll make the necessary changes and merge :)

Optimize VRAM use in textual inversion training

Thanks, will make open a PR then :)