YoVDO

How Replit Trained Their Own LLMs - LLM Bootcamp

Offered By: The Full Stack via YouTube

Tags

LLM (Large Language Model) Courses Data Processing Courses Data Pipelines Courses

Course Description

Overview

Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only!
Explore the comprehensive process of training custom Large Language Models (LLMs) in this 32-minute conference talk by Reza Shabani from Replit. Gain insights into the entire workflow, from data processing to deployment, including the modern LLM stack, data pipelines using Databricks and Hugging Face, preprocessing techniques, tokenizer training, and running training with MosaicML and Weights & Biases. Learn about testing and evaluation methods using HumanEval and Hugging Face, as well as deployment strategies involving FasterTransformer, Triton Server, and Kubernetes. Discover valuable lessons on data-centrism, evaluation, and collaboration, and understand the qualities that make an effective LLM engineer.

Syllabus

Why train your own LLMs?
The Modern LLM Stack
Data Pipelines: Databricks & Hugging Face
Preprocessing
Tokenizer Training
Running Training: MosaicML, Weights & Biases
Testing & Evaluation: HumanEval, Hugging Face
Deployment: FasterTransformer, Triton Server, k8s
Lessons learned: data-centrism, eval, and collaboration
What makes a good LLM engineer?


Taught by

The Full Stack

Related Courses

Google Cloud Big Data and Machine Learning Fundamentals en Español
Google Cloud via Coursera
Data Analysis with Python
IBM via Coursera
Intro to TensorFlow 日本語版
Google Cloud via Coursera
TensorFlow on Google Cloud - Français
Google Cloud via Coursera
Freedom of Data with SAP Data Hub
SAP Learning