kinoc

#  So now you want to finetune that GPT-J-6B on a 3090/TITAN GPU ... okay 
#  More exploratory coding. It uses the Huggingface model port, deepspeed and reads all text/md files from a target directory
#  It is a fragment of a larger system with remote editing, but that's another story
#  This is the raw, training tester. Items to look out for:
#  - uses DeepSpeed and has a DS config
#  - to save space uses SGD instead of ADAM
#  - uses gradient checkpointing
#  - freezes 25% of the layers to fit

# Assumes you can already run https://gist.github.com/kinoc/2d636a68876cd3de7b6e9c9452b61089

joshuaskorich

/*
* (un)comment correct payload first (x86 or x64)!
* 
* $ gcc cowroot.c -o cowroot -pthread
* $ ./cowroot
* DirtyCow root privilege escalation
* Backing up /usr/bin/passwd.. to /tmp/bak
* Size of binary: 57048
* Racing, this may take a while..
* /usr/bin/passwd overwritten

celoyd

Himawari-8 animation tutorial

Here’s how to make animations like this one. It requires intermediate Unix command-line knowledge, to install some tools and to debug if they don’t work. You’ll need these utilities:

• curl (or you can translate to wget)
• convert and montage, part of ImageMagick
• ffmpeg, plus whatever codecs
• parallel, for iteration that’s nicer than shell for loops or xargs
• run everything in zsh for leading 0s in numerical ranges to work

henrik

tmux cheatsheet

As configured in my dotfiles.

start new:

tmux

start new with session name: