Artefact2

Which GGUF is right for me? (Opinionated)

Good question! I am collecting human data on how quantization affects outputs. See here for more information: ggerganov/llama.cpp#5962

In the meantime, use the largest that fully fits in your GPU. If you can comfortably fit Q4_K_S, try using a model with more parameters.

llama.cpp feature matrix

See the wiki upstream: https://github.com/ggerganov/llama.cpp/wiki/Feature-matrix

adam-hanna

Introduction

I often find myself ssh'ing into my servers and checking my systemd service logs with $ journalctl -f -u {name}.service. One day I got tired of this and wanted all of my important logs in once place (Amazon AWS Cloudwatch). To my dismay, there weren't any real good tutorials on how to do so. So, voilà.

Steps

Overall, it's a fairly simple process consisting of the following few steps.

1. Modify the service file

Open the service file with $ sudo vi /lib/systemd/system/{name}.service

Modify the [Service] section:

cjgiridhar

package main

import "fmt"

func main() {
	// shoppingList is a map that has a map inside it
	shoppingList := make(map[string]map[string]int)

	// veggies key points to veggiesMap
	veggiesMap := map[string]int{"onion": 2, "orka": 3}

qpwo

"""
A minimal implementation of Monte Carlo tree search (MCTS) in Python 3
Luke Harold Miles, July 2019, Public Domain Dedication
See also https://en.wikipedia.org/wiki/Monte_Carlo_tree_search
https://gist.github.com/qpwo/c538c6f73727e254fdc7fab81024f6e1
"""
from abc import ABC, abstractmethod
from collections import defaultdict
import math

dnozay

#!/usr/bin/env python

# This example shows how to use multiprocessing with an initializer function.
# We take advantage of that to make the workers each have a custom initial
# load. And in particular example, we will make the workers sleep.
# Because we make them sleep different amounts, one of them is going to be
# ready much before the others, and thus we can guess easily which worker
# will do most of the work.
#
# see https://stackoverflow.com/questions/26693797/python-multiprocessing-process-number

aras-p

// Just before switching jobs:
// Add one of these.
// Preferably into the same commit where you do a large merge.
//
// This started as a tweet with a joke of "C++ pro-tip: #define private public",
// and then it quickly escalated into more and more evil suggestions.
// I've tried to capture interesting suggestions here.
//
// Contributors: @r2d2rigo, @joeldevahl, @msinilo, @_Humus_,
// @YuriyODonnell, @rygorous, @cmuratori, @mike_acton, @grumpygiant,