yoavg

Learning LLMs in 2025

So you know how the transformer works, and you know basic ML/DL, and you want to learn more about LLMs. One way to go is looking into the various "algorithmic" stuff (optimization algorithms, RL, DPO, etc). Lot's of materials on that. But the interesting stuff is (in my opinion at least) not there.

This is an attempt to collect a list of academic (or academic-like) materials that explore LLMs from other directions, and focus on the non-ML-algorithmic aspects.

Courses

• David Chiang's Theory of Neural Networks course.
• This is not primarily LLMs, but does have substantial section on Transformers. Formal/Theory. More of a book than a course.

gd3kr

"""
a simple script that reads tweets inside a json file, uses openai to compute embeddings and creates two files, metadata.tsv and output.tsv, which cam be used to visualise the tweets and their embeddings in TensorFlow Projector (https://projector.tensorflow.org/)
"""

# obtain tweets.json from https://gist.github.com/gd3kr/948296cf675469f5028911f8eb276dbc

import pandas as pd
import json
from openai import OpenAI

adtac

#!/usr/bin/env -S bash -c "docker run -p 8080:8080 -it --rm \$(docker build --progress plain -f \$0 . 2>&1 | tee /dev/stderr | grep -oP 'sha256:[0-9a-f]*')"

# syntax = docker/dockerfile:1.4.0

FROM node:20

WORKDIR /root

RUN npm install sqlite3

epurdy

NAME OF BELLIGERENT                                Rating +/- Deviation (sorted by Rating - Deviation)

Turkistan Islamic Party                            3587 +/- 113
Dadullah Front                                     3581 +/- 111
High Council of Afghanistan Islamic Emirate        3581 +/- 111
Syrian Arab Republic                               3429 +/- 161
ISIL-YP                                            3250 +/- 166
GPC                                                3233 +/- 166
Saleh                                              3163 +/- 163
Supreme Political Council                          3159 +/- 163

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

mihailik

//@ts-check

async function massBlock() {
  const dummy = {};

  const progressPopup = await showProgressPopup();
  const thumbsUp = '\uD83D\uDC4D';

  async function showProgressPopup() {
    const animationTimeMsec = 200;

ryanfb

#!/usr/bin/env ruby
# Based on: https://github.com/twitterdev/Twitter-API-v2-sample-code/blob/main/Bookmarks-lookup/bookmarks_lookup.rb
# See: https://github.com/ryanfb/twitter-bookmarks-export
require 'json'
require 'typhoeus'
require 'twitter_oauth2'

# First, you will need to enable OAuth 2.0 in your App’s auth settings in the Developer Portal to get your client ID.
# Inside your terminal you will need to set an enviornment variable
# export CLIENT_ID='your-client-id'

radekosmulski

Rocketknight1

# This is a new feature, so make sure to update to the latest version of transformers!
# You will also need to pip install tensorflow_text

import tensorflow as tf
from transformers import TFAutoModel, TFBertTokenizer


class EndToEndModel(tf.keras.Model):
    def __init__(self, checkpoint):
        super().__init__()

pmbaumgartner

from typing import Optional, Iterable, cast, List
from thinc.api import get_current_ops, Ops
from thinc.types import Ragged, Ints1d

from spacy.pipeline.spancat import Suggester
from spacy.tokens import Doc
from spacy.util import registry


@registry.misc("ngram_digits_suggester.v1")