3 ms·
Optimizing it to be cheaper to run (e.g. OpenAI saved a lot of money with the turbo variant that ChatGPT uses, relative to the original GPT-3 models).
by andai 4y ago
Optimizing it to be cheaper to run (e.g. OpenAI saved a lot of money with the turbo variant that ChatGPT uses, relative to the original GPT-3 models).