Tyler

Software Engineer | App Developer

[Barstool Baseball @StoolBaseball]

WILD BILL TO THE MOON

[Adam McCalvy @AdamMcCalvy]

Brewers probable starters for the Cubs series:

Friday: Jacob Misiorowski
Saturday: Kyle Harrison
Sunday: Brandon Woodruff

[Milwaukee Bucks @Bucks]

With the 15th selection in the 2013 NBA Draft the Milwaukee Bucks select Giannis Antetokounmpo. #BucksDraft

[Ryan @BrewPack8]

They can never take this away from us

[SportsCenter @SportsCenter]

Giannis' time in Milwaukee comes to an end 🦌

The Greek Freak left a legacy 💪

[Ryan @BrewPack8]

Thank you Giannis…

For keeping the Bucks in Milwaukee
For staying throughout all the noise
For a championship
For 13 seasons
For being the greatest to ever put on a Bucks uniform
For everything

[Milwaukee Brewers @Brewers]

Your Milwaukee Brewers:
🔵 45 wins in 71 games
🔵 Season-high 19 games above .500
🔵 113 runs in June, most in MLB
🔵 Second fewest runs allowed (263) in NL
🔵 Most home wins (25) in MLB

[SleeperBrewers @SleeperBrewers]

This might be the most insane pimp job on a HR I’ve seen in a long time

[Jonathan Lucroy @JLucroy20]

Last week I made a tweet about Uribe and his actions on the field. I made a comment about the lack of veteran leadership in the Brewers clubhouse.

I want to make a public statement that I very clearly misspoke and was wrong for saying it the way I did.

Yeli and Woody are the leaders of that clubhouse. They are both legendary @Brewers players and represent themselves, the Brewers, and the community very well. They have had a ton of success, and are very respected amongst the players and the league as a whole.

I reached out to Murph, Yeli, and Woody and conveyed my apologies to them for disrespecting them. In no way did I mean for that to happen, however, I was clearly wrong and I did a terrible job at conveying my thoughts using the words I did.

Being a distraction to the team is the last thing I want to be. I want nothing but success and championships for this team, city, and the @Brewers organization.

I will be better and will clarify my words better moving forward.

I talk a lot about taking responsibility and having humility in the face of a mistake.
This was a mistake that I made and am making it right.

-Luc

[Ryan @BrewPack8]

The Brewers will enter June with a 4.5 game lead (+ current tiebreaker) in the Central

[Milwaukee Brewers @Brewers]

The Milwaukee Brewers in May:

📈 19-7 (best in MLB)
🔥 2.52 ERA (best in MLB)
7️⃣ Series wins
🧹 Three sweeps

[Sam Dekker @dekker]

This is not hyperbole at all; Contreras has a case for MVP so far.

[Sam Dekker @dekker]

@JLucroy20 The team is in 1st place after their 3 best hitters missed most of the first month-plus. Teams with “no leadership” don’t do that. Murph addressed it, the vets addressed it, he apologized publicly and they’re moving on. I don’t know how you could paint this as a clubhouse issue.

[Ryan @BrewPack8]

BREWERS WIN

SERIES WIN

FLY THE L

FIRST PLACE

[Meteorologist Sam Kuffel @SamKuffelWx]

lol. Lmao.

[Milwaukee Brewers @Brewers]

SET THE TONE

#ThisIsMyCrew x @UWCreditUnion

[Milwaukee Brewers @Brewers]

This win is dedicated to the beautiful city of Green Bay

[Ryan @BrewPack8]

FIRST SWEEP OF THE YANKEES SINCE 1989. THEY WERE THE HOTTEST TEAM IN BASEBALL AND THE CREW SWEEPS EM OUTTA MILWAUKEE WOW WOW WOW

[Andrej Karpathy @karpathy]

LLM Knowledge Bases

Something I'm finding very useful recently: using LLMs to build personal knowledge bases for various topics of research interest. In this way, a large fraction of my recent token throughput is going less into manipulating code, and more into manipulating knowledge (stored as markdown and images). The latest LLMs are quite good at it. So:

Data ingest:
I index source documents (articles, papers, repos, datasets, images, etc.) into a raw/ directory, then I use an LLM to incrementally "compile" a wiki, which is just a collection of .md files in a directory structure. The wiki includes summaries of all the data in raw/, backlinks, and then it categorizes data into concepts, writes articles for them, and links them all. To convert web articles into .md files I like to use the Obsidian Web Clipper extension, and then I also use a hotkey to download all the related images to local so that my LLM can easily reference them.

IDE:
I use Obsidian as the IDE "frontend" where I can view the raw data, the the compiled wiki, and the derived visualizations. Important to note that the LLM writes and maintains all of the data of the wiki, I rarely touch it directly. I've played with a few Obsidian plugins to render and view data in other ways (e.g. Marp for slides).

Q&A:
Where things get interesting is that once your wiki is big enough (e.g. mine on some recent research is ~100 articles and ~400K words), you can ask your LLM agent all kinds of complex questions against the wiki, and it will go off, research the answers, etc. I thought I had to reach for fancy RAG, but the LLM has been pretty good about auto-maintaining index files and brief summaries of all the documents and it reads all the important related data fairly easily at this ~small scale.

Output:
Instead of getting answers in text/terminal, I like to have it render markdown files for me, or slide shows (Marp format), or matplotlib images, all of which I then view again in Obsidian. You can imagine many other visual output formats depending on the query. Often, I end up "filing" the outputs back into the wiki to enhance it for further queries. So my own explorations and queries always "add up" in the knowledge base.

Linting:
I've run some LLM "health checks" over the wiki to e.g. find inconsistent data, impute missing data (with web searchers), find interesting connections for new article candidates, etc., to incrementally clean up the wiki and enhance its overall data integrity. The LLMs are quite good at suggesting further questions to ask and look into.

Extra tools:
I find myself developing additional tools to process the data, e.g. I vibe coded a small and naive search engine over the wiki, which I both use directly (in a web ui), but more often I want to hand it off to an LLM via CLI as a tool for larger queries.

Further explorations:
As the repo grows, the natural desire is to also think about synthetic data generation + finetuning to have your LLM "know" the data in its weights instead of just context windows.

TLDR: raw data from a given number of sources is collected, then compiled by an LLM into a .md wiki, then operated on by various CLIs by the LLM to do Q&A and to incrementally enhance the wiki, and all of it viewable in Obsidian. You rarely ever write or edit the wiki manually, it's the domain of the LLM. I think there is room here for an incredible new product instead of a hacky collection of scripts.

[Milwaukee Brewers @Brewers]

Calm down