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

russmo

docker run -d \
    --name=crashplan-pro \
    -h $HOSTNAME \
    -e USER_ID=0 \
    -e GROUP_ID=0 \
    -e TZ=“America/Los_Angeles” \
    -p 5800:5800 \
    -p 5900:5900 \
    -v /share/CACHEDEV1_DATA/Container/config/crashplanpro:/config:rw \
    -v /share/CACHEDEV1_DATA:/storage:rw \

htoyryla

# git clone from https://github.com/tkarras/progressive_growing_of_gans
# download the snapshot from their Google drive
# use the following code in the same directory to generate random faces
import os
import sys
import time
import glob
import shutil
import operator
import theano

victor-shepardson

Code

https://gist.github.com/victor-shepardson/5b3d3087dc2b4817b9bffdb8e87a57c4

Setup

I'm using Ubuntu 16.04 with a GTX 1060

install CUDA

brannondorsey

Image-to-Image Translation with Conditional Adversarial Networks

Notes from arXiv:1611.07004v1 [cs.CV] 21 Nov 2016

• Euclidean distance between predicted and ground truth pixels is not a good method of judging similarity because it yields blurry images.
• GANs learn a loss function rather than using an existing one.
• GANs learn a loss that tries to classify if the output image is real or fake, while simultaneously training a generative model to minimize this loss.
• Conditional GANs (cGANs) learn a mapping from observed image x and random noise vector z to y: y = f(x, z)
• The generator G is trained to produce outputs that cannot be distinguished from "real" images by an adversarially trained discrimintor, D which is trained to do as well as possible at detecting the generator's "fakes".
• The discriminator D, learns to classify between real and synthesized pairs. The generator learns to fool the discriminator.
• Unlike an unconditional GAN, both th

EderSantana

Code for Keras plays catch blog post

Train

python qlearn.py

Test

1. Generate figures

oglops

#!/usr/bin/env python
try:
        # for python newer than 2.7
    from collections import OrderedDict
except ImportError:
        # use backport from pypi
    from ordereddict import OrderedDict

import yaml

matsen

For sane jupyter notebook diffs

• Install the nbconvert package, though you probably already have it if you are using jupyter.
• Put the nb2md script below in your path and make executable
• Add the following to your .gitattributes file, which can be in your home directory (use nb2md for all projects) or in the root of your project:

*.ipynb diff=nb2md

• Run

baraldilorenzo

##VGG16 model for Keras

This is the Keras model of the 16-layer network used by the VGG team in the ILSVRC-2014 competition.

It has been obtained by directly converting the Caffe model provived by the authors.

Details about the network architecture can be found in the following arXiv paper:

Very Deep Convolutional Networks for Large-Scale Image Recognition

K. Simonyan, A. Zisserman

karpathy

"""
Minimal character-level Vanilla RNN model. Written by Andrej Karpathy (@karpathy)
BSD License
"""
import numpy as np

# data I/O
data = open('input.txt', 'r').read() # should be simple plain text file
chars = list(set(data))
data_size, vocab_size = len(data), len(chars)