Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
GaggiX
4 days ago
|
parent
|
context
|
favorite
| on:
Faster convergence for diffusion models
I wonder how well this technique works if the distribution of the training dataset between the diffusion model and the image encoder is quite different, for example if you use DinoV2 as the encoder but train the diffusion model on anime.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: