FilterHN

From 300KB to 69KB per Token: How LLM Architectures Solve the KV Cache Problem

32 points

by future-shock-ai

2 days ago

| 1 comment

| news.future-shock.ai

| HN

1 hour ago

[-]

Unrelated, but 69KB is how much RAM Voyager 1 has.

5 minutes ago

[-]

Voyager as a token of curiosity