jmsktm

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 and experience preferred (super rare at this point).

My own notes from a few months back.

Background

• Survey of LLMS
• Self-attention and transformer networks

jmsktm

1. download these two files to the same directory
2. docker-compose up
3. http://localhost should show you "it works" which is apache being reverse-proxied through nginx
4. docker sets up DNS for web based on compose service name so the nginx front-end can find http://web
5. proxy is set to listen on public port 80 on host so it'll receive incoming traffic, then push to httpd

jmsktm

#!/usr/bin/env python3

import os
path = "/home/james/Github"

def exec(command):
	print("Executing Command: {}".format(command))
	os.system(command)

def main():

jmsktm

#!/bin/bash

#Setup base paths
WORKSPACE_PATH="/home/student/capstone_ws"
PROJECT_NAME="project"

# Setup the base repository
rm -rf $WORKSPACE_PATH
mkdir -p $WORKSPACE_PATH && cd $WORKSPACE_PATH
git clone https://github.com/jmsktm/T2-CapstoneProject.git $PROJECT_NAME

jmsktm

echo 'deb http://http.debian.net/debian wheezy-backports main' > /etc/apt/sources.list.d/wheezy-backports-main.list
curl -s https://packagecloud.io/install/repositories/github/git-lfs/script.deb.sh | sudo bash
apt -y install git-lfs

#Setup base paths
WORKSPACE_PATH="/home/workspace/capstone_ws"
PROJECT_NAME="T2-CapstoneProject"

# Setup the base repository
rm -rf $WORKSPACE_PATH

jmsktm

#!/bin/bash

#Setup base paths
WORKSPACE_PATH="/home/workspace/capstone_ws"
PROJECT_NAME="project"

# Setup the base repository
rm -rf $WORKSPACE_PATH
mkdir -p $WORKSPACE_PATH && cd $WORKSPACE_PATH
git clone https://github.com/jmsktm/T2-CapstoneProject.git $PROJECT_NAME

jmsktm

#!/bin/bash

WORKSPACE_PATH="/home/workspace/capstone_ws"
PROJECT_NAME="project"

rm -rf $WORKSPACE_PATH
mkdir -p $WORKSPACE_PATH && cd $WORKSPACE_PATH
git clone https://github.com/jmsktm/T2-CapstoneProject.git $PROJECT_NAME
cd $PROJECT_NAME

jmsktm

pip install --upgrade pip
sudo apt-get update
sudo apt-get install python-catkin-pkg -y
pip install matplotlib

cd /home/workspace/CarND-Capstone/ros
catkin_make
source devel/setup.sh
roslaunch launch/styx.launch

jmsktm

# Procedure to install NVIDIA driver, Cuda and cuDNN

# Run update first
sudo apt-get update

echo '\n\n#### Processor:'
cat /proc/cpuinfo | grep 'model name' | uniq | cut -d ":" -f2 | awk '{$1=$1};1'
# AMD Ryzen Threadripper 2950X 16-Core Processor

echo '\n#### Linux version:'

jmsktm

"""
This function takes hough lines as input, and uses the information
to arrive at two straight lanes which correspond to the left and
the right lanes.
"""
def final_lines(lines, video_width, ymin, ymax):
    
    # Find the middle of screen (widthwise). We will use this to determine
    # whether a line/coordinate will be used to generate the left or
    # the right lane