
5. Charlie Snell on DALL-E and CLIP
The Inside View
Dali
Dali usually when people mean end to end that means like you know the entire model is trained input to output. That's like mostly true for Dali except there's sort of this two stage training thing so like arguably it's not fully end to end. But I mostly meant in that sense end to end in the sense that it's trained directly for the task of text to image which is in contrast for the clip based stuff which is not necessarily trained for that. And then our lovely avocado shares. Iconic.
00:00
Transcript
Play full episode
Remember Everything You Learn from Podcasts
Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.