|
|
|
|
|
|
by YeGoblynQueenne
2354 days ago
|
|
|
Leaving the question of machine translation etc aside for the moment, this is about playing chess from textual examples of play. There is no reason to assume that, even if GPT-2 was really any good at machine translation, that it would be any good at chess. I guess people think "it's a powerful model so it should do well in any task" but that's typically not the case for neural nets. I know what OpenAI claims about how it can do a little bit of everything, machine translation benchmarks are borked and I bet so are question answering ones (which I confess I don't know much about). |
|
|