That's a wrap! If you interested in: – Python – Data Science – Machine Learning – Maths for ML – MLOps – CV/NLP – LLMs Find me → @akshay_pachaar Everyday, I share tutorials on above topics! Check my tutorial on Self-Attention
LLMS
-
Stanford CS25 Transformers United Course Overview
By
–
Stanford CS25 – Transformers United Since their introduction in 2017, transformers have revolutionised the world of NLP! This course delves into the inner workings of transformers, exploring their applications with a focus on LLMs. Check this out
-
PyTorch Neural Networks Course for NLP Fundamentals
By
–
NLP A thorough introduction to cutting-edge neural networks for NLP. A hands-on course with plenty of assignments & a final project. PyTorch for all the code! Check this out https://
youtube.com/playlist?list=
PLoROMvodv4rOSH4v6133s9LFPRHjEmbmJ
… -
Stanford Offers Free World-Class Data Science Education Online
By
–
Stanford offers FREE Data Science Education! World-class courses on: – Machine Learning
– Computer Vision
– Statistical Learning
– Graph Neural Networks
– Reinforcement Learning
– Natural language processing (NLP)
– Large Language Models (LLMs) Read more… -
Improving Language Models Through Fine-Tuning and Data Quality
By
–
I think we need to increase the data and its quality, and as in LMs do fine tuning (aka grounding). The model already has compression bottlenecks, which we could improve sure, but the concepts are already there.
-
Transformers Excel at Interpolation but Struggle with Symbolic Learning
By
–
Not saying that Transformers are worse than RNNs, mind you — Transformers are *the best* at *what deep learning does* (generalizing via interpolation), specifically *because* of their strongly interpolative architecture prior (MHA). They are, however, worse at learning symbolic
-
Transformers’ Interpolative Architecture and Limitations for Symbolic Tasks
By
–
Ironically, Transformers are even worse in that regard — mostly due to their strongly interpolative architecture prior. Multi-head-attention literally hardcodes sample interpolation in latent space. Also, the fact that recurrence is a really helpful prior for symbolic programs.
-
Website Chatbot with LangChain GPT-4 and Streamlit
By
–
Chat with any Website using Python and Langchain Build a Website-Interacting Chatbot with LangChain, GPT-4 and Streamlit Full end-to-end application! Great tutorial to walk through on a lazy weekend
-
nanoGPT with MoEs and multimodal tokenization support
By
–
i would love for @karpathy to make a nanoGPT for MoEs with multimodal tokenization support