Victor Jotham Ashioya ashioyajotham

Purpose

Bootstrap knowledge of LLMs ASAP. With a bias/focus to GPT.

Avoid being a link dump. Try to provide only valuable well tuned information.

Prelude

Neural network links before starting with transformers.

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

Description

Cheatsheet for LaTex, using Markdown for markup. I use this with atom.io and 📦markdown-preview-plus to write math stuff. 📦keyboard-localization is necessary when using an international layout (like [swiss] german).

Further Reference and source: ftp://ftp.ams.org/pub/tex/doc/amsmath/short-math-guide.pdf

Example expressions / functions

	import re, string, unicodedata
	import nltk
	import contractions
	import inflect
	from nltk import word_tokenize, sent_tokenize
	from nltk.corpus import stopwords
	from nltk.stem import LancasterStemmer, WordNetLemmatizer

	def replace_contractions(text):
	"""Replace contractions in string of text"""

	# Credit 🙏: I just used the example from langchain docs and it works quite well: https://python.langchain.com/en/latest/use_cases/question_answering.html
	# Note 2: The Arxiv -> PDF logic is a bit messy, I'm sure it can be done better
	# Note 3: Please install the following:

	# To run:

	# Save this in a `app.py`
	# pip install arxiv PyPDF2 langchain chromadb

	# The chat feature was shipped in H2O nightly this week, we will need to install from nightly link:

	import asyncio
	import json
	from time import time, sleep

	import aiohttp
	import pandas as pd
	from bs4 import BeautifulSoup
	from tqdm import tqdm

	# Ващет СИКРЕТНА!!!!! Но код хранится в private репозитории, так что пох

	class MyStreamListener(tweepy.StreamListener):
	def __init__(self, api=None):
	super(MyStreamListener, self).__init__()
	self.num_tweets = 0
	self.file = open("tweets.txt", "w")

	def on_status(self, status):
	tweet = status._json
	self.file.write( json.dumps(tweet) + '\n' )
	self.num_tweets += 1