Thomas Capelle

Results 169 comments of Thomas Capelle

Of course! I am sure that Jeremy would also love to weight in.

If you want a working example check my repo: https://github.com/tcapelle/moving_mnist

Hey, i tried this today after reading your post and indeed it works almost identically. Sometimes bettter, sometimes worst. You disable moment regularization or another clever trick?

I took Adan for a spin today, it looks promising. I am just training my fav timm backbone `convnext_tiny` with nice results. The only downside I see is that it's...

It is possible. Take a look at this [tutorial](https://thinkmoult.com/basic-rendering-tutorial-radiance.html) on how to export a blender 3D model to radiance.

This would be really cool! Currently, `nbdev_test` does not produce any interpretable output, only lists the failing notebooks without the error message (like nbdev1 used to).

Ohh missed this! thanks guys, and I actually figured this out :P

Being trying to make my training not OOM and I am unable to make mixed precision work on dreambooth. ``` /home/tcapelle/mambaforge/envs/dream/lib/python3.10/site-packages/accelerate/accelerator.py:179: UserWarning: `log_with=tensorboard` was passed but no supported trackers are...

I know, but this is very cool as it produces a nice `wandb.workspace`, you can see the logging of the run here: https://wandb.ai/capecape/train_sd?workspace=user-capecape