@gerrymcgovern The article confuses two very different things - the process of training an LLM, which is computationally intensive and uses a lot of energy but is done once, and the process of using the trained LLM, which isn't.
Top-level
@gerrymcgovern The article confuses two very different things - the process of training an LLM, which is computationally intensive and uses a lot of energy but is done once, and the process of using the trained LLM, which isn't. 2 comments
|
@bontchev @gerrymcgovern No, it doesn't, it calls out both, but doesn't confuse the two.