Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
AnthusAI
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
4 ms
·
1.
▲
by
AnthusAI
3y ago
Fine tuning is generally not the best way to teach an LLM new knowledge. RAG is still more appropriate. Fine tuning is generally more effective for controlling the format of the responses but it's not going to teach the model a lot