Joan webeng

Reinforcement Learning for Language Models

Yoav Goldberg, April 2023.

Why RL?

With the release of the ChatGPT model and followup large language models (LLMs), there was a lot of discussion of the importance of "RLHF training", that is, "reinforcement learning from human feedback". I was puzzled for a while as to why RL (Reinforcement Learning) is better than learning from demonstrations (a.k.a supervised learning) for training language models. Shouldn't learning from demonstrations (or, in language model terminology "instruction fine tuning", learning to immitate human written answers) be sufficient? I came up with a theoretical argument that was somewhat convincing. But I came to realize there is an additional argumment which not only supports the case of RL training, but also requires it, in particular for models like ChatGPT. This additional argument is spelled out in (the first half of) a talk by John Schulman from OpenAI. This post pretty much

##VGG19 model for Keras

This is the Keras model of the 19-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

Cheat sheet for using Selenium2TestCase v1.4 extension PHPUnit

Prepare browser

Full screen browser

class ClassTest extends \PHPUnit_Extensions_Selenium2TestCase
{
    public static $browsers = array(
        array(

	// ==UserScript==
	// @name Get rid of EU Cookie Banners
	// @namespace http://blog.deobald.org/
	// @description Get rid of EU Cookie Banners
	// @license http://creativecommons.org/licenses/by-nc-sa/3.0/
	// @downloadURL https://gist.githubusercontent.com/BlaM/6d826d6e9d77d2d77bf9a92fdad55788/raw/cookiebanner-go-away.user.js
	// @homepage https://gist.github.com/BlaM/6d826d6e9d77d2d77bf9a92fdad55788
	// @version 0.3.4
	// @author Dominik Deobald
	// @match http:///*

	# Follow instructions here: http://flask.pocoo.org/snippets/65/
	# You'll need to make sure you have the right Python version and any packages installed: http://docs.webfaction.com/software/python.html
	# This tweak to index.py is what made it actually work: http://stackoverflow.com/questions/3696606/how-to-solve-import-errors-while-trying-to-deploy-flask-using-wsgi-on-apache2

	import sys
	yourappname = "/home/username/webapps/yourappname/htdocs"
	if not yourappname in sys.path:
	sys.path.insert(0, yourappname)

	from yourappname import app as application