Figure 1. Jeronim with his 1070TI GPU, that has fans from aliexpress that don't really fit so he has to reapply the duct tape every now and then.
I am currently working as a research engineer at tensorpix. We are developing models for video superresolution and video restoration. How it feels when your multiple discriminator GAN doen’t diverge into NaNs: me
Big fan of CLIP
On the link below you can find my ML blog + short comments on various deep learning projects i have done over the years, reach out about anything if you have any questions
I sometimes talk about ML papers. If i told you you can find the slides on my site, here they are. Most of these, but not all are from our paperclub
Scalable Diffusion Models with Transformers
FreeU : Free Lunch in Diffusion U-Net
Patch n’ Pack: NaViT, a Vision Transformer for any Aspect Ratio and Resolution
Lottery ticket hypothesis, but actually the Han et al. pruning paper