I've thought about this for a while: I think the video solution is the cleanest and most reproducible *and* the least error prone.

Thankfully, #ATPESC has recordings from previous years from the same exact venue. I downloaded the audio with yt-dlp, converted it with ffmpeg, and then transcribed it with whisper.cpp.

So far, it seems like it might not be a good solution because the audio quality from the recording is much better than what I recorded.

There's no time like the present to make an #introduction!

I'm a recently defended #graduate #phd student. I studied data #visualization and how to apply algorithms written for #HPC systems so they'll work in the #cloud.

I just finished attending #ATPESC where we learned about all kinds of HPC techniques, libraries, and tools for 2 weeks.

My current #adhd hobby of the month (going on 3 months!) is #AI and large language models #LLM, specifically how to make them small enough to run locally.