3 ms·
I tried it out and I am super impressed by something so small. It hallucinated a lot, but formulated its incorrect answers wonderfully. It's certainly in the r
by 7734128 3y ago
I tried it out and I am super impressed by something so small. It hallucinated a lot, but formulated its incorrect answers wonderfully.
It's certainly in the realm of being able to voice a NPC in an RPG game by now, as occasional hallucinations doesn't really matter there.
- brucethemoose2 3y agoThese are not even the best 7B models for an RPG. The RP Llama/Mistral finetunes should do much better.
- MuffinFlavored 3y ago> It hallucinated a lot How do we (the "LLM community") solve this/overcome this?
- behnamoh 3y agomillion dollar startup idea.
- 7734128 3y agoI suggest that we make a universal set of truths of everything which the LLM can compare its answer to. Perhaps a relational database? Or just one huge JSON?
- brucethemoose2 3y agoThis is kind of an inherent property of the llama architecture. Maybe new architectures will be better, but companies will likely be making the best foundational models. In the meantime, you can hook it up to a vectordb and ask the llm to check itself with zero temperature, but that only helps so much.