Yoav Goldberg, Nov 24, 2024
This piece started with a pair of twitter and bluesky posts:
let's talk about "agents" (in the LLM sense). there's a lot of buzz around "multi-agent" systems where agents collaborate but... i don't really get how it differs from a thinking of a single agent with multiple modes of operation. what are the benefits of modeling as multi-agent?
— (((ل()(ل() 'yoav))))👾 (@yoavgo) November 23, 2024
Develop an AI prompt that solves random 12-token instances of the A::B problem (defined here), with 90%+ success rate.
We'll use your prompt as the SYSTEM PROMPT, and a specific instance of problem as the PROMPT, inside XML tags. Example:
Yoav Goldberg, April 2023.
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
Audience: I assume you heard of chatGPT, maybe played with it a little, and was imressed by it (or tried very hard not to be). And that you also heard that it is "a large language model". And maybe that it "solved natural language understanding". Here is a short personal perspective of my thoughts of this (and similar) models, and where we stand with respect to language understanding.
Around 2014-2017, right within the rise of neural-network based methods for NLP, I was giving a semi-academic-semi-popsci lecture, revolving around the story that achieving perfect language modeling is equivalent to being as intelligent as a human. Somewhere around the same time I was also asked in an academic panel "what would you do if you were given infinite compute and no need to worry about labour costs" to which I cockily responded "I would train a really huge language model, just to show that it doesn't solve everything!". We
Yoav Goldberg, Jan 23, 2021.
The FAccT paper "On the Dangers of Stochastic Parrots: Can Languae Models be Too Big" by Bender, Gebru, McMillan-Major and Shmitchell has been the center of a controversary recently. The final version is now out, and, owing a lot to this controversary, would undoubtly become very widely read. I read an earlier draft of the paper, and I think that the new and updated final version is much improved in many ways: kudos for the authors for this upgrade. I also agree with and endorse most of the content. This is important stuff, you should read it.
However, I do find some aspects of the paper (and the resulting discourse around it and around technology) to be problematic. These weren't clear to me when initially reading the first draft several months ago, but they became very clear to me now. These points are for the most part
| /* inline formatting, link targets and [[ ]] disappears if not active line*/ | |
| div:not(.CodeMirror-activeline) > .CodeMirror-line span.cm-formatting, | |
| div:not(.CodeMirror-activeline) > .CodeMirror-line span.cm-string.cm-url, | |
| div:not(.CodeMirror-activeline) > .CodeMirror-line span.cm-formatting-link, | |
| div:not(.CodeMirror-activeline)>.CodeMirror-line span.cm-hmd-barelink, | |
| div:not(.CodeMirror-activeline)>.CodeMirror-line span.cm-comment | |
| { display: none; } | |
| /* hide all html tags -- IT IS COMMENTED OUT BY DEFAULT */ | |
| /* div:not(.CodeMirror-activeline) > .CodeMirror-line span.cm-tag{ display: none; } */ |
If you haven’t worked with JavaScript in the last few years, these three points should give you enough knowledge to feel comfortable reading the React documentation:
let and const statements. For the purposes of the React documentation, you can consider them equivalent to var.class keyword to define JavaScript classes. There are two things worth remembering about them. Firstly, unlike with objects, you don't need to put commas between class method definitions. Secondly, unlike many other languages with classes, in JavaScript the value of this in a method [depends on how it is called](https://developer.mozilla.org/en-US/docs/Web/Jav
Experiment on the effect of different SaveMode and Cassandra starting from a populated table
If the cassandra table that spark targets exists then
SaveMode.Append will update itSaveMode.Overwrite will truncate and insert (but it requires option "confirm.truncate" -> "true")SaveMode.Ignore will not perform any action on existing tableSaveMode.ErrorIfExists (default) will throw the following exception:
https://github.com/datastax/spark-cassandra-connector/blob/v2.0.6/spark-cassandra-connector/src/main/scala/org/apache/spark/sql/cassandra/DefaultSource.scala#L93-L96
| *** Cluster Setup for Google Container Engine *** | |
| 0/ Install and configure local gcloud and kubectl: https://cloud.google.com/sdk/docs/ | |
| > gcloud components install kubectl | |
| 1/ Configure Google Cloud account: | |
| > gcloud config set account YOUR_EMAIL_ADDRESS | |
| > gcloud config set project YOUR_PROJECT_ID | |
| > gcloud config set compute/zone us-west1-a | |
| > gcloud config set container/cluster example |