Evan Chong evantkchong

Using GitHub Pages with a custom domain: Namecheap Edition

As often happens, I found the official documentation and forum answers to be "close, but no cigar", and so had to experiment a little to get things working.

The main problem for me was a lack of concrete configuration examples. That's not entirely GitHub's fault: having migrated from Google Domains to Namecheap in the middle of this project, I was once again reminded of how many different ways there are to do things in the name service universe [1].

Although you'd think the simplest setup would be to merely configure for the subdomain case (https://www.example.com), in my experience using the apex domain (https://example.com) instead resulted in fewer complications.

Procedure

So here's my recipe for using a custom domain with GitHub pages where Namecheap is the DNS provider:

Installing CUDA for TensorFlow GPU support on Ubuntu 20.04

These installation steps are adapted from the official Tensorflow documentation for installing CUDA on linux as well as the CUDA installation guide from Nvidia. Testing of this environment was performed on a machine with an NVIDIA GeForce RTX 2060 Mobile, Nvidia Driver 495.44 and TensorFlow 2.5.0.

Note that the installation steps below assume you already have a compatible Nvidia driver installed (>=450.80.02 for CUDA 11.4).

The choice of CUDA 11.4 is somewhat arbitrary as I initially believed this was the newest CUDA toolkit version supported by the newest version of TensorRT on Ubuntu 20.04 (8.2.0-1+cuda11.4 at the time of writing). However, this seems to not be the case (see Installation Myths).

Installation Myths

TensorFlow and CUDA version compatibility

Project Title

One Paragraph of project description goes here

Getting Started

These instructions will get you a copy of the project up and running on your local machine for development and testing purposes. See deployment for notes on how to deploy the project on a live system.

Prerequisites

VGG-Face model for Keras

This is the Keras model of VGG-Face.

It has been obtained through the following method:

vgg-face-keras:directly convert the vgg-face matconvnet model to keras model
vgg-face-keras-fc:first convert vgg-face caffe model to mxnet model,and then convert it to keras model

Details about the network architecture can be found in the following paper:

	/* Adapted from the css of the JetBrainsMono site at https://www.jetbrains.com/lp/mono */

	@font-face {
	font-family: JetBrainsMono;
	src: url("https://github.com/JetBrains/JetBrainsMono/raw/master/fonts/webfonts/JetBrainsMono-ExtraLight.woff2?raw=true") format("woff2"),
	url("https://github.com/JetBrains/JetBrainsMono/raw/master/fonts/ttf/JetBrainsMono-ExtraLight.ttf?raw=true") format("truetype");
	font-weight: 100;
	font-style: normal;
	}

	# 0 is too far from ` ;)
	set -g base-index 1

	# Automatically set window title
	set-window-option -g automatic-rename on
	set-option -g set-titles on

	#set -g default-terminal screen-256color
	set -g status-keys vi
	set -g history-limit 10000

	// Tweet (gif) : https://twitter.com/randompast/status/897526013437018114

	// $ budo --live cov2.js
	// Attempt at local convergence to a global solution via Brachistochrone
	// Intro:
	// https://www.youtube.com/watch?v=skvnj67YGmw
	// https://www.youtube.com/watch?v=Cld0p3a43fU
	// Mathier:
	// http://mathworld.wolfram.com/BrachistochroneProblem.html
	// https://www.youtube.com/watch?v=6HeQc7CSkZs&t=22s