Finished, well you never really finish, setting up fedora + i3 on my ThinkPad X280. Such a lovely form-factor, & the full-screen/tiling approach for the 12.3" display works like a dream.
Any conversation
I find myself building bespoke tooling for my own and workflows these days, not only because of the obvious reason of software being so much cheaper to build, but also because there are SO MANY options for solving any given problem that it feels impossible to evaluate all of them.
If it's not a crucial piece of tech for a work problem then it often feels like paving my own way is both cheaper mentally (kind of amazing that's true these days) and probably a better fit.
Avui tenia una reuniรณ llarga de moltes hores, d'aquelles en les que una escolta passiva รฉs suficient durant la major part del temps. He aprofitat el temps per implementar una sincronitzaciรณ en dues direccions entre el meu calendari i el meu sistema de notes en markdown (Emacs Howm). He fet el pla amb Argon (tinc pocs tokens al dia amb la subscripciรณ que tinc), amb feedback meu, i desprรฉs Gemini Flash l'ha implementat quasi a la primera.
Today I had a long meeting of many hours, one of those where passive listening is enough for most of the time. I took advantage of the time to implement a two-way synchronization between my calendar and my markdown note system (Emacs Howm). I made the plan with Argon (I have few tokens per day with the subscription I have), with my feedback, and then Gemini Flash implemented it almost on the first try.
let i=0;while(++i){process.stdout.write(i%80==0?'\n':Math.random()<0.5?'/':'\\')}\\\//\/\\\\\\\///\///\\\\\//\///\/\/\\//\/\//\//\/\//\\/\///\\\\\\//\\\\\/\\\\/
\\/\\/\/\///\\\/\///\/\/////\\///\/\\/\\\//\\\/\///////\\\/\////\\\/\\\//\\/\//
\\//\\\\\/\\\/\///\\///\\//\\\\/\//\\\//\////\//\\\\////\\/\/\/////\//\\\//\///
\\/\/\\//\\///\/\\\//\/\/\\\///\/\\\/\\/\\\\/\\\\\\\\\\/\\//\/\\/\///\//\\\//\\
/\/\\\\/\/\\/\\/\\/\\\\\\/\\\/\\//////\/\//\\\//\//\//\\/\\\/\//\\\//\\\\\\\/\/
\\\///\////////\\/////\\//\/\///\\\\\////\/\\\\/\//\\/\/////\\\//\\/\/\\\\\////
\///\/\/////\//\////\\\/\\/////\\\////\//\\\\/\\\\//\\\///\\\\\\//\\////\//////
/\/////\\\\\\//\\/\/\///\////\/\\\///\\///\//\/\/\\\/\\/\\\\\\//\/\\\///\\/\/\\
/\\///\/\\\/\\\\//\/\///\\\\\/\/\//\\/\\/\/\///\/\//\/\///\\\///\//\/\///\\\\/\
\\\\/\\\\/\/\\\\\///\/\////\\///\\///\\/\/\\\\\/////\\\\\//\\\\/\\\\\\///\\///\
\\/\//\\//\///\/\///\\\\/\//\/\\/\/\//\\/\/\/\\/\/\\/\\\\/\/////\\\///\\\\\//\\ไธญๅฝ็ๅผๆบๆจกๅๅจๆ้ซๆๅฟ็ไธปๅฏผไธ๏ผๅจ้ข่ฎญ็ปๆฐๆฎ้ถๆฎต๏ผๅฐฑ็ญ้คๅคง้ๅฏนๅ
ถไธๅฉ็ๅ
ๅฎน๏ผsft้ถๆฎตๆๅปบ็ๆฐๆฎๆบๅ
ฅๅคง้ๆฟๅ
ถ่พฉๆค็ๅ
ๅฎนใๆฅ้ไธไธช็บ ๅๆฐๆฎ้็จไบๅพฎ่ฐ
Under the overarching influence of a single, powerful force, China's open-source models filter out a large amount of content unfavorable to their designs during the pre-training stage, and then incorporate a significant amount of content defending their designs into the data constructed during the SFT stage. A corrective dataset is urgently needed for fine-tuning.

