Best LLMs cheatsheet you’ll ever find ✅

Covers concepts, finetuning, evaluations...

Here's what you'll learn.

1️⃣ Introduction

↳ Gen AI and Large Language Models

↳ Tokens, Embeddings, and Self-Attention

↳ Transformers and Parallel Processing

2️⃣ Transformers Architecture and Variants

↳ BERT (Encoder-only models)

↳ GPT (Decoder-only models)

↳ T5 and BART (Encoder-Decoder models)

3️⃣ Large Language Models and Training

↳ Fine-tuning (Task-specific, Multi-task, Instruction-based)

↳ Parameter Efficient Fine-Tuning (PEFT, LoRA)

↳ Preference Tuning (RLHF, DPO, IPO)

↳ Scaling and Optimization (Quantization, Distillation, Pruning)

4️⃣ Applications

↳ LLM-as-a-Judge (Evaluation and Benchmarking)

↳ RAG (Retrieval-Augmented Generation)

↳ Agents (ReAct, PAL)

↳ Reasoning Models (Chain-of-Thought, Scaling Laws)

Crips and to the point explanations by Dataiku team.

♻ Repost if you found this useful.