Fabrice Bellard is definitely outstanding programming genius. Every time he release something he prove that all modern software is awfully slow overbloated pile of useless shit.
This is what it make me to the "Most mice love cactus" seed phrase:
Most mice love cactus. It’s the best source of fiber, calcium, iron, and magnesium. And its crunchy seeds contain protein, zinc, potassium, and vitamin E.
But not all mice eat cactus. And if you’ve ever tried to feed cactus seeds to your pet, you probably know what they look like before they’re put in your mouth. They are so sharp and spiny, you wouldn’t want to eat them, even if your pet enjoys them. And most people don’t have much trouble imagining doing that, since we know that they can sometimes have a very bitter taste.
The problem is that the cactus seed does have a bitter taste. And because of that, these seeds are not good for your pet. For example, if you give your pet a handful of the cactus seeds, he or she may not even know that he or she has eaten them. But because they have a bitter taste, they
Exactly. "You don't need real disease at all, you just need an image of epidemic to scare people".
You also could get an opensource standalone text completion program with GPT-2 model from the same place - https://bellard.org/libnc/gpt2tc.html
Fabrice Bellard is definitely outstanding programming genius. Every time he release something he prove that all modern software is awfully slow overbloated pile of useless shit.
Don't know about this exact one, but if yout want to play with GPT models - https://bellard.org/textsynth/
GPT-J 6B is larger than GPT-3.
This is what it make me to the "Most mice love cactus" seed phrase: