| """ | |
| The most atomic way to train and run inference for a GPT in pure, dependency-free Python. | |
| This file is the complete algorithm. | |
| Everything else is just efficiency. | |
| @karpathy | |
| """ | |
| import os # os.path.exists | |
| import math # math.log, math.exp |
Terminals should generate the 256-color palette from the user's base16 theme.
If you've spent much time in the terminal, you've probably set a custom base16 theme. They work well. You define a handful of colors in one place and all your programs use them.
The drawback is that 16 colors is limiting. Complex and color-heavy programs struggle with such a small palette.
This article is a response to mfiano’s From Common Lisp to Julia which might also convey some developments happening in Common Lisp. I do not intend to suggest that someone coming from a Matlab, R, or Python background should pickup Common Lisp. Julia is a reasonably good language when compared to what it intends to replace. You should pickup Common Lisp only if you are interested in programming in general, not limited to scientific computing, and envision yourself writing code for the rest of your life. It will expand your mind to what is possible, and that goes beyond the macro system. Along the same lines though, you should also pickup C, Haskell, Forth, and perhaps a few other languages that have some noteworthy things to teach, and that I too have been to lazy to learn.
/I also do not intend to offend anyone. I’m okay with criticizing Common Lisp (I myself have done it below!), but I want t
| """ | |
| stable diffusion dreaming | |
| creates hypnotic moving videos by smoothly walking randomly through the sample space | |
| example way to run this script: | |
| $ python stablediffusionwalk.py --prompt "blueberry spaghetti" --name blueberry | |
| to stitch together the images, e.g.: | |
| $ ffmpeg -r 10 -f image2 -s 512x512 -i blueberry/frame%06d.jpg -vcodec libx264 -crf 10 -pix_fmt yuv420p blueberry.mp4 |
2019-05-18
It's uncanny how closely this talk follows a line of reasoning I've been working on lately. If you're copied here, we may have discussed one or more of its aspects
https://www.youtube.com/watch?v=pW-SOdj4Kkk
(Jonathan Blow is an independent video game developer, speaking here at a conference in Moscow.)
I have included working code examples that can be run throughout, as well as graphs. I hope this helps make this easier to understand in a more hands on way.
Suppose you know that there are 10 balls in an urn, some are red and some are blue. So there are 11 different possible models for this situation:
I've started using Helix as my main editor, and I wanted a place to jot down notes as I run into things. So this is that spot. For anyone stumbling on this, please keep in mind that over-all I'm really liking Helix, which isn't well reflected in the notes below.
Please don't add comments to this gist. I do want to discuss these items, but not here. This is just a place for my personal notes, to help me keep track of my thoughts. (Unfortunately, Github doesn't support disabling comments on gists.)
If you're reading this and want to comment on something, feel free to open up an issue on the Helix repo if it's significant enough, or just ping me in the Helix matrix room otherwise.
This post was adapted from an earlier Twitter thread.
It's incredible how many collective developer hours have been wasted on pushing through the turd that is ES Modules (often mistakenly called "ES6 Modules"). Causing a big ecosystem divide and massive tooling support issues, for... well, no reason, really. There are no actual advantages to it. At all.
It looks shiny and new and some libraries use it in their documentation without any explanation, so people assume that it's the new thing that must be used. And then I end up having to explain to them why, unlike CommonJS, it doesn't actually work everywhere yet, and may never do so. For example, you can't import ESM modules from a CommonJS file! (Update: I've released a module that works around this issue.)
And then there's Rollup, which apparently requires ESM to be u
| :def hoogle \x -> return $ ":!hoogle --count=15 \"" ++ x ++ "\"" | |
| :def doc \x -> return $ ":!hoogle --info \"" ++ x ++ "\"" | |
| :set -Wall | |
| :set -fno-warn-type-defaults -ferror-spans -freverse-errors -fprint-expanded-synonyms | |
| :set prompt "\ESC[0;32m%s\n\ESC[m[ghci]\ESC[38;5;172mλ \ESC[m" | |
| :set prompt-cont " \ESC[38;5;172m> \ESC[m" |
