It's mentioned in the article. But for really, really long-horizon tasks, it might be reasonable that you don't want to have a small discount factor. For example, if you have really sparse rewards in a long-horizon task…
Flash attention is an implementation trick; you can implement MHA/GQA, for example, with flash attention.
It's mentioned in the article. But for really, really long-horizon tasks, it might be reasonable that you don't want to have a small discount factor. For example, if you have really sparse rewards in a long-horizon task…
Flash attention is an implementation trick; you can implement MHA/GQA, for example, with flash attention.