LLM tinkering
Getting familiar with running Llama locally via LM Studio. I’m seeing whether I can use it to augment my reflection practice. The idea is to feed it some of my private data like journals and see if it can surface insights that are not immediately apparent to me.
Also getting accustomed with vector databases like Chroma. And trying to understand the attraction of using LangChain. In even more addition, how do I even decide how to chunk large documents so that they fit into a context window while still maintaining meaning.
Right now, I’m a little stalled on converting my freeplane mindmaps to a format that is easily consumed by the LLM while retaining some level of token economy.
Accountability Cohorts
Also forming accountability cohorts to help people achieve their most ambitious goals through the power of mild social pressure. I’m thinking of turning this into a membership to help people build, maintain and gracefully exit accountability relationships.
And also looking for an accountability partner or group that can match my level of intensity or dedication (feels a bit wanky when I write it out though! 😂)
Last updated 2025-03-05