Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
55555
on April 26, 2023
|
parent
|
context
|
favorite
| on:
DeepFloyd IF: open-source text-to-image model
So this one can create perfect text in images? If true, that’s insane
GaggiX
on April 26, 2023
[–]
LDM-400M was already able to generate text (predecessor of Stable Diffusion), thanks to the fact that every token in the text encoder (trained from scratch) was available in the attention layer.
flangola7
on April 26, 2023
|
parent
[–]
>thanks to the fact that every token in the text encoder (trained from scratch) was available in the attention layer.
>ChatGPT explain this like I'm 5
coolspot
on April 26, 2023
|
root
|
parent
[–]
“Every word in the text can be used to help create the image.”
Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
