jboner

Latency Comparison Numbers (~2012)
----------------------------------
L1 cache reference                           0.5 ns
Branch mispredict                            5   ns
L2 cache reference                           7   ns                      14x L1 cache
Mutex lock/unlock                           25   ns
Main memory reference                      100   ns                      20x L2 cache, 200x L1 cache
Compress 1K bytes with Zippy             3,000   ns        3 us
Send 1K bytes over 1 Gbps network       10,000   ns       10 us
Read 4K randomly from SSD*             150,000   ns      150 us          ~1GB/sec SSD

veekaybee

Anti-hype LLM reading list

Goals: Add links that are reasonable and good explanations of how stuff works. No hype and no vendor content if possible. Practical first-hand accounts of models in prod eagerly sought.

Foundational Concepts

Pre-Transformer Models

isaacs

# Hello, and welcome to makefile basics.
#
# You will learn why `make` is so great, and why, despite its "weird" syntax,
# it is actually a highly expressive, efficient, and powerful way to build
# programs.
#
# Once you're done here, go to
# http://www.gnu.org/software/make/manual/make.html
# to learn SOOOO much more.

JosiahParry

library(googledrive)

# authenticate yourself
drive_auth()

gdoc_to_rmd <- function(drive_id, output, verbose = TRUE, overwrite = FALSE) {
  
  # get the doc
  doc <- drive_get(as_id(drive_id))
  

kylebgorman

# autoloess.R: compute loess metaparameters automatically
# Kyle Gorman <gormanky@ohsu.edu>

aicc.loess <- function(fit) {
    # compute AIC_C for a LOESS fit, from:
    # 
    # Hurvich, C.M., Simonoff, J.S., and Tsai, C. L. 1998. Smoothing 
    # parameter selection in nonparametric regression using an improved 
    # Akaike Information Criterion. Journal of the Royal Statistical 
    # Society B 60: 271–293.

eliocamp

---
title: "Tweet thread"
author: "Elio Campitelli"
output: github_document
---


```{r}
knitr::opts_chunk$set(dev = "png", 
                      tweet_this = TRUE)

DavisVaughan

# This example demonstrates running furrr code distributed on 2 AWS instances ("nodes").

# The instances have already been created.

library(future)
library(furrr)

# Two t2.micro AWS instances
# Created from http://www.louisaslett.com/RStudio_AMI/
public_ip <- c("34.205.155.182", "34.201.26.217")

Rambou

If, like me, your are booting with UEFI (because having a triple boot ubuntu-windows-mac or because UEFI is the most modern type of bootloader and successor of EFI :p), you have to sign the proprietary modules each time they are recompiled (or upgrade kernel version) so that they are allowed to be loaded in the kernel.

1) Step one, create a self-signed certificate to sign nvidia driver:
sudo openssl req -new -x509 -newkey rsa:2048 -keyout UEFI.key -outform DER -out UEFI.der -nodes -days 36500 -subj "/CN=rambou_nvidia/"

2) step two load and store certificate in a supplementary key database MOC 
sudo mokutil --import UEFI.der

3) step three reboot your system
At this step after reboot you will be prompted to select your certificate to import in in key database. If you have inserted a password at certificate creation you'll be prompted to insert it. If you are not prompted, you may have to enter the BIOS by using function keys at boot time.

fmder

import numpy
from scipy.ndimage.interpolation import map_coordinates
from scipy.ndimage.filters import gaussian_filter

def elastic_transform(image, alpha, sigma, random_state=None):
    """Elastic deformation of images as described in [Simard2003]_.


    .. [Simard2003] Simard, Steinkraus and Platt, "Best Practices for
       Convolutional Neural Networks applied to Visual Document Analysis", in

charliejhadley

library("shiny")
library("tidyverse")
library("rnaturalearthdata")
library("sf")

# Using Africa instead of whole world as countries110 contains
# 3 invalid geometries. Also, too many examples use the USA everytime
africa_sf <- countries110 %>% 
  st_as_sf() %>% 
  st_transform(4326) %>%