4 ms·
Continual feedback means continual training. No way around it. So you’d have to scope down the functional unit to a fairly small lora in order to get reasonable
by freeone3000 2y ago
Continual feedback means continual training. No way around it. So you’d have to scope down the functional unit to a fairly small lora in order to get reasonable re-training costs here.
- nico 2y agoOr maybe figure out a different architecture Either way, the end user experience would be vastly improved
- regularfry 2y agoThat's not quite true. The system prompt is state that you can use for "training" in a way that fits the problem here. It's not differentiable so you're in slightly alien territory, but it's also more comprehensible than gradient-descending a bunch of weights.
- immibis 2y agoIf you treat it as vectors instead of words, it might be differentiable.