Generative AI Posts
Repository contains LinkedIn posts about Generative AI knowledge sharing, learning resources and research explanations. Although the repository currently contains only my posts, but it doesn't necessarily have to. Please raise PR to add relevant high-quality linkedin post from other people too.
If you find this repository helpful, please leave to appreciate. It will help it reach more people.
August, 2024
- How does Uber predict the Expected Time of Arrival (ETA) for the trips?
- Multi-agentic RAG is all you need to build scalable enterprise-level applications using LLM.
- Do you know why TikTok's recommendation algorithm is so good?
- LangChain recently released LangGraph studio
July, 2024
- How to evaluate planning capabilities of LLM?
- Few applied research directions to ensure low latency for LLMs
- How does deep learning frameworks like TensorFlow work?
- Difference between thinking pattern of Early career ML engineer vs Experienced ML engineer
- Designing ChatGPT like AI assistant from scratch - ML system design
- Different ways of prompting an LLM
- How to transition to AI/ML from any field?
- My experience & Learnings as ML Engineer at TikTok
- How to give memory to your AI assistant just like ChatGPT?
- What is Calibration and why it is important for industry scale ML products?
- Will opensource models like Meta's Llama be able to defeat closed source models like OpenAI's GPT series?
June, 2024
- Different variants of LoRA finetuning
- Popular evaluation metrics for LLMs
- Different strategies to implement Retrieval Augmented Generation (RAG)
- Popular algorithms to perform embedding retrieval for RAG
- Matryoshka embeddings for RAG
- Popular fast embedding generators for RAG
- Popular Guardrails methods for LLMs
- Popular Decoding Strategies for LLMs
- How to create custom instruction tuning dataset for SFT?
- What is Quantization in LLMs?
- What is Model Merging in LLMs and what are some popular methods for model merging?
- How to create document embedding store for Retrieval-Augmented Generation (RAG)?
- Metrics to evaluate RAG system
