4 ms·
There are models trained via RNN, but LLMs usually use Transformer architecture with a bit of human feedback on top which sort of uses reinforcement learning li
by Jax_Hax 3y ago
There are models trained via RNN, but LLMs usually use Transformer architecture with a bit of human feedback on top which sort of uses reinforcement learning like AlphaGo