|
所在平台: Udemy |
课程主页: https://www.udemy.com/course/openai-llm-fine-tuning-course/
课程评论：没有评论
课程名称：大语言模型微调基础 + 微调OpenAI GPT模型 课程概述：大语言模型（LLMs）已经彻底改变了人工智能行业，提供了前所未有的精准度，拓展了人工智能的可能性。然而，预训练的LLMs并不一定能满足组织的特定需求，因此需要微调这些模型，以便更好地适应独特的任务和要求。本课程旨在帮助您掌握LLM微调的技能。课程从微调的基本概念入手，强调微调在使LLM模型适应特定数据中的关键作用。随即，我们将深入实际操作，涵盖整个微调流程，以微调OpenAI GPT模型。通过实践环节，您将逐步学习如何准备和格式化数据集，执行OpenAI微调过程，以及评估模型结果。 课程内容： - 理论部分：介绍LLM和LLM微调的核心基础，讨论微调的必要性、工作原理、微调工作流程及所涉及的步骤；不同的微调技术，包括RLHF、PEFT、LoRA、QLoRA、标准、顺序和基于指令的微调；最佳实践。 - 实践部分：详细讲解OpenAI仪表盘和游乐场，让您全面了解OpenAI为生成性AI提供的各种工具；实际操作OpenAI微调工作流程，涵盖探索性数据分析（EDA）、数据预处理、数据格式化、创建微调任务和评估；理解OpenAI接受的专用JSONL格式，学习系统、用户、助手三个重要角色；使用Tiktoken库提前计算Token数量和微调费用；通过逐步的实践，在Python中使用自定义数据集微调OpenAI的GPT模型，评估微调模型的准确性和性能，与基础的预训练模型进行比较，评估微调的影响。 完成本课程后，您将能够熟练地微调OpenAI的GPT模型，以满足特定组织的需求，并开始您的LLM微调工程师的职业旅程。
"Large Language Models (LLMs) have revolutionized the AI industry, providing unprecedented precision, expanding the possibilities of artificial intelligence. However the pre-trained LLMs may not always meet the specific requirements of an organization, hence there is always a need to Fine tune the LLMs to tailor these models to your unique tasks and requirements."This comprehensive course is designed to equip you with the skills of LLM Fine tuning technique. It starts with a thorough introduction to the fundamentals of fine tuning while highlighting its critical role to make the LLM models adapt to your specific data. Then we will dive into Hands-on sessions covering the entire Fine-tuning workflow to Fine tune OpenAI GPT model. Through practical sessions, you'll step-by-step learn to prepare & format datasets, execute OpenAI fine tuning processes, and evaluate the model outcomes.By the end of this course, you will be proficient in Fine tuning the OpenAI's GPT model to meet specific organizational needs and start your career journey as LLM Fine tuning engineer. ____________________________________________________________________________________________What in nutshell is included in the course ?[Theory]We'll start with LLM and LLM Fine tuning's core basics and fundamentals.Discuss Why is Fine tuning needed, How it works, Workflow of Fine tuning and the Steps involved in it.Different types of LLM Fine tuning techniques including RLHF, PEFT, LoRA, QLoRA, Standard, Sequential and Instructions based LLM Fine tuning.Best practices for LLM Fine tuning.[Practicals]Get a detailed walkthrough of OpenAI Dashboard and Playground to get a holistic understanding of wide range of tools that OpenAI offers for Generative AI.Follow the OpenAI Fine tuning workflow in practical sessions covering Exploratory Data Analysis (EDA), Data preprocessing, Data formatting, Creating fine tuning job, Evaluation.Understand the OpenAI specialized JSONL format that it accepts for training & test data, and learn about 3 important roles - System, User, Assistant.Calculate the Token count and Fine tuning cost in advance using Tiktoken library.Gain Hands-on experience in fine tuning OpenAI's GPT model on a custom dataset using Python through step-by-step practical sessions.Assess the accuracy and performance of the fine-tuned model compared to the base pre-trained model to evaluate the impact of fine-tuning.
