Speaker 1: Jeremy Howard
0:00 | Introduction to Language Models 1:31 | What is a Language Model? 6:27 | The ULMfit Algorithm 11:13 | Compression and Intelligence 16:36 | GPT-4: The Best Language Model 20:45 | GPT-4’s Capabilities and Limitations 24:41 | Getting Good Answers from Language Models 31:16 | Advanced Data Analysis (Code Interpreter) 37:53 | OpenAI API Pricing 46:35 | Building a Code Interpreter in Jupyter 53:32 | Using Language Models on Your Own Computer 59:00 | HuggingFace Transformers 1:00:44 | Model Leaderboards 1:01:32 | Llama2: A Popular Open Source Model 1:03:15 | Using 8-bit and GPTQ for Faster Inference 1:07:30 | Instruction Tuned Models 1:09:46 | Scaling Up with Larger Models 1:11:09 | Retrieval Augmented Generation (RAG) 1:14:23 | Sentence Transformers for Document Retrieval 1:16:52 | Vector Databases for RAG 1:20:49 | Fine Tuning Language Models 1:22:02 | Fine Tuning for SQL Generation 1:26:29 | Using Macs for Language Modeling 1:27:11 | MLC: Running Language Models on Macs and Mobile Devices 1:28:20 | Llama.cpp: Another Option for Macs and CUDA 1:29:37 | Choosing the Right Tools 1:30:12 | Conclusion and Call to Action