ruDALL-E finetuned on Warhols

Nov 27, 2021 · 6:56 AM UTC

9
12
113
If you want to finetune your own very cool and sexy ruDALL-E, check out this here Kaggle notebook for an example of how to do so: kaggle.com/shonenkov/emojich…

Emojich-ruDALL-E

Explore and run machine learning code with Kaggle Notebooks | Using data from multiple data sources

kaggle.com
3
6
2
43
Replying to @MichaelFriese10
How well did it know Warhol beforehand?
1
*hand wobbles*
3
Replying to @MichaelFriese10
Looks interesting :) If you have a code ready, time, and desire, could you please try to finetune it on the WikiArt dataset? Just out of curiosity github.com/cs-chan/ArtGAN/bl…
1
Replying to @MichaelFriese10
I didn’t know you could fine tune rudalle.
1
Replying to @MichaelFriese10
wow
Replying to @MichaelFriese10
Impressive
Replying to @MichaelFriese10
Like it
Replying to @MichaelFriese10
Amazing!
This tweet is unavailable
N=231. Trained for three epochs, taking 17 minutes on a P100
4