Didn't tweet nanoGPT yet (quietly getting it to good shape) but it's trending on HN so here it is 🙂 : https://
github.com/karpathy/nanoG
PT
…
Aspires to be simplest, fastest repo for training/finetuning medium-sized GPTs. So far confirmed it reproduced GPT-2 (124M). 2 simple files of ~300 lines
LLMS
-
nanoGPT: Simplest Repository for Training Medium-Sized GPTs
By
–
-
ChatGPT Politeness and Data Training Practices Disagreement
By
–
Fact – Saying thank you to ChatGPT improves responses later in the current chat session you're in. Fact – They have implied they are using data from the current free preview to train it to better respond in the future. I disagree with your assessment.
-
The Growing Power of Text Prompts for AI Systems
By
–
## Prompts as Commodities The power of a block of text has also never been bigger. The right choice of words can result in amazing outputs for AIs.
-
LLMs With Tools: Beyond Single-Word Generation
By
–
Many people mistakenly think that large language models that generate one word at a time is the end game.
— Oriol Vinyals (@OriolVinyalsML) 11 janvier 2023
Connecting LLMs with tools, e.g.: search engines, python interpreters, etc., is super exciting, and leverages tools' power, robustness & more! pic.twitter.com/39PmdYEgwvMany people mistakenly think that large language models that generate one word at a time is the end game.
Connecting LLMs with tools, e.g.: search engines, python interpreters, etc., is super exciting, and leverages tools' power, robustness & more! -
GPT-3 with Memory: ChatGPT Approximation Explained
By
–
Yes correct – it’s not chat gpt, but rather an approximation of chat gpt that is gpt3 + memory (to replicate the conversational aspect)
-
Multilingual Modeling Podcast with Nils Reimers and SentenceBERT
By
–
Check out this exciting podcast hosted by @CShorten30 featuring the one and only @Nils_Reimers
! Discussing multilingual modeling, his work on SentenceBERT, and more! -
DeepMind RETRO: Controlling Chatbot Access Through Curated Databases
By
–
Parece que la propuesta de RETRO de DeepMind podría ser un camino interesante, donde el límite se puede establecer curando una base de datos que contenga sólo aquella información a la que se quiera dar acceso al chatbot. https://
deepmind.com/publications/i
mproving-language-models-by-retrieving-from-trillions-of-tokens
… -
FAISS Wrapper Now Returns Similarity Scores in Semantic Search
By
–
Return scores during semantic search – h/t @krrish_dh In addition to getting back the top documents from a vectorstore, the FAISS wrapper now supports returning similarity scores for those documents as well Documentation: https://
langchain.readthedocs.io/en/latest/modu
les/utils/combine_docs_examples/vectorstores.html#faiss-specific
… -
WolframAlpha Integration Tool for LangChain Framework
By
–
WolframAlpha integration – @nickscamara_ Support for using Wolfram Alpha as a tool Collab Notebook: https://
colab.research.google.com/drive/1AAyEdTz
-Z6ShKvewbt1ZHUICqak0MiwR
… Documentation: https://
langchain.readthedocs.io/en/latest/modu
les/utils/examples/wolfram_alpha.html
… s/o to others @brucehammer for suggesting this a few weeks ago, and others in the discord for reviewing! -
Finding Generative AI App Ideas: Problem-Solving Approach
By
–
How to find ideas for your next generative AI app? By identifying your goal – the problem you want to solve. Let’s look at some example goals that are well-suited for large language models. (Thread)