tonytins

joined 1 year ago
[–] tonytins@pawb.social 1 points 1 year ago

🎶 Streaming killed TV star. 🎶

[–] tonytins@pawb.social 3 points 1 year ago (1 children)

A bit of both, actually. With such a diverse amount of instances, covering every possible type of social network, the best they're ever going to throttle the flagships. xD

[–] tonytins@pawb.social 3 points 1 year ago

Then don't put in any hidden fees.

[–] tonytins@pawb.social 3 points 1 year ago

OpenAI is really putting all their eggs in one basket with GPT. ಠ_ಠ It is great technology, don't get me wrong, but it only understands the frequency of words, not the actual meaning behind them.

[–] tonytins@pawb.social 40 points 1 year ago (11 children)

They're going to have a hard time throttling the Fediverse.

[–] tonytins@pawb.social 4 points 1 year ago

Shouldn't ChatGPT's code be used as a template, anyway?

[–] tonytins@pawb.social 1 points 1 year ago

I don't even...

[–] tonytins@pawb.social 14 points 1 year ago

I love my Steam Deck.

[–] tonytins@pawb.social 4 points 1 year ago

The thing about all GPT models is that they’re based on the frequency of the word to determine its usage. Which means the only way to get good results is if it's running on cutting edge equipment designed specifically for that job, while being almost a TB in size. Meanwhile, Diffusion models are only GB and run on the GPU but still produce masterpieces because they already know what that word is associated with.

[–] tonytins@pawb.social 7 points 1 year ago

I'm sure the social network also matters, too.

[–] tonytins@pawb.social 2 points 1 year ago

I have problems with it repeating certain words over and over again no matter how much I adjust the style and tone.

[–] tonytins@pawb.social 6 points 1 year ago

Already!? XD

view more: ‹ prev next ›