本課程旨在介紹自然語言處理（NLP）和大型語言模型（LLM）的基礎知識和前瞻技
術，適合對自然語言技術感興趣的學生。隨著生成式人工智慧技術的快速發展，NLP在
各個領域中的應用日益廣泛。
本課程將提供學生NLP理論基礎，並結合實際應用，幫助學生掌握最新的NLP與LLM技
術。課程內容主要分為以下幾個部分：
1. 文字處理基礎：介紹NLP的基本概念和常用技術。教學基本的文字處理技術，如分
詞、詞性標註、命名實體識別等。
2. 機器學習模型：機器學習基本概念和算法，如線性回歸、決策樹、隨機森林等。介紹
如何將機器學習應用於NLP，包括文本分類、情感分析等。
3. 語言模型：語言模型的基本概念與原理，如N-gram模型、Word2Vec等。深度學習在語
言模型中的應用，如RNN、LSTM、Transformer等架構。詳細講解BERT、GPT等先進語
言模型，並探討其在不同NLP任務中的應用。
4. 前瞻大語言模型技術：介紹大型語言模型的發展歷程與最新研究進展，如GPT-3等。
探討這些模型的訓練方法、大規模資料集的使用，以及在不同領域中的能力。介紹輕量
化微調技術（PEFT），如LoRA，並說明其在提高訓練效率和效果方面的優勢。This course aims to introduce the basic knowledge and forward-looking technologies of natural language processing (NLP) and large language models (LLM).
technology, suitable for students interested in natural language technology. With the rapid development of generative artificial intelligence technology, NLP is
Applications in various fields are becoming increasingly widespread.
This course will provide students with the theoretical foundation of NLP and combine it with practical applications to help students master the latest NLP and LLM techniques.
technique. The course content is mainly divided into the following parts:
1. Word processing basics: Introduce the basic concepts and common techniques of NLP. Teaching basic word processing techniques such as
Words, part-of-speech tagging, named entity recognition, etc.
2. Machine learning model: basic concepts and algorithms of machine learning, such as linear regression, decision tree, random forest, etc. introduce
How to apply machine learning to NLP, including text classification, sentiment analysis, etc.
3. Language model: The basic concepts and principles of language model, such as N-gram model, Word2Vec, etc. Deep learning in Chinese
Applications in language models, such as RNN, LSTM, Transformer and other architectures. Detailed explanation of advanced languages such as BERT and GPT
language model and explore its application in different NLP tasks.
4. Forward-looking large language model technology: Introducing the development history and latest research progress of large language models, such as GPT-3, etc.
Explore the training methods of these models, the use of large-scale datasets, and their capabilities in different domains. Introducing lightweight
oriented fine-tuning technology (PEFT), such as LoRA, and illustrates its advantages in improving training efficiency and effectiveness.
.
.
| 評分項目 Grading Method | 配分比例 Grading percentage | 說明 Description |
|---|---|---|
| Homework x 5Homework x 5 homework next 5 |
75 | |
| Term Project x1Term Project x1 term project boots 1 |
25 |
