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

svpino

Running TensorFlow on Apple Silicon

1. Install Homebrew The Missing Package Manager for macOS (or Linux) — Homebrew

2. Download Miniforge3

3. Install Miniforge3 and restart your terminal as soon as the installation finishes:

maslychm

Install PyTorch 1.12.0 with Mac M1 GPU support (MPS device: for Metal Performance Shaders)

TLDR: Dowload package directly from anaconda.org and install over the current torch version

Explanation

Wrong version (1.10.2) gets installed for me when I run conda install pytorch -c pytorch-nightly.
To fix, install the latest pytorch version from the stable channel, then manually download package for your python version, and install it over the current pytorch. Newest Pytorch Nightly releases here: Anaconda: Pytorch Nightly Files.

ortegatron

from detectron2.engine.hooks import HookBase
from detectron2.evaluation import inference_context
from detectron2.utils.logger import log_every_n_seconds
from detectron2.data import DatasetMapper, build_detection_test_loader
import detectron2.utils.comm as comm
import torch
import time
import datetime

class LossEvalHook(HookBase):

ylashin

import flask
from flask_cors import CORS
from flask import request, jsonify
from detectron2.engine import DefaultPredictor
from detectron2.config import get_cfg
from detectron2.data import MetadataCatalog
import cv2
import requests
import numpy as np

swedishmike

## Create the virtual environment
conda create -n 'environment_name'

## Activate the virtual environment
conda activate 'environment_name'

## Make sure that ipykernel is installed
pip install --user ipykernel

## Add the new virtual environment to Jupyter

MuhsinFatih

I am assuming you are here because like me, you installed a bazillion different python interpreters on mac and the whole thing is a spagetti. Today, I finally fixed my python installation. Whatever I install for python2 or python3 using pip JUST.WORKS.. My god! finally.

What the hell?

Here is what I had messed up, which you also probably did:

• I had too many different python interpreters
• Too many different symlinks which I lost track of
• almost no package I installed with pip worked without a headache
• any attempt to fix using online resources made it worse.

mschae16

To prepare for the Docker 101 workshop, please install Docker and follow the steps in creating a Dockerhub account.

To prepare for the Kubernetes 101 workshop, please install the following:

• Docker
• minikube

Verify your minikube install by running the following commands in your terminal:

$ minikube start and

everling-prime

SSARCandy

#!/bin/bash

function show_gpu_user {
        pid=$(pstree -sg $1 | grep -Eo 'bash\([0-9]*\)' | head -1 | grep -Eo '[0-9]*');
        docker ps -q | xargs docker inspect --format "{{.Name}}  {{.State.Pid}}" | grep $pid | awk '{printf "%-24s", $1}';
        ps aux | grep $1 | grep -v grep | awk '{ for(i=1;i<=NF;i++) {if ( i >= 11 ) printf $i" "}; printf "\n" }';
}


echo " ";