Gen AI Foundational Models for NLP & Language Understanding

Overview

This IBM course will teach you how to implement, train, and evaluate generative AI models for natural language processing (NLP). The course will help you acquire knowledge of NLP applications including document classification, language modeling, language translation, and fundamentals for building small and large language models. You will learn about converting words to features. You will understand one-hot encoding, bag-of-words, embedding, and embedding bags. You also will learn how Word2Vec embedding models are used for feature representation in text data. You will implement these capabilities using PyTorch. The course will teach you how to build, train, and optimize neural networks for document categorization. In addition, you will learn about the N-gram language model and sequence-to-sequence models. This course will help you evaluate the quality of generated text using metrics, such as BLEU. You will practice what you learn using Hands-on Labs and perform tasks such as implementing document classification using torchtext in PyTorch. You will gain the skills to build and train a simple language model with a neural network to generate text and integrate pre-trained embedding models, such as word2vec, for text analysis and classification. In addition, you will apply your new skills to develop sequence-to-sequence models in PyTorch and perform tasks such as language translation.

Syllabus

Fundamentals of Language Understanding

In this module, you will learn about one-hot encoding, bag-of-words, embeddings, and embedding bags. You will also gain knowledge of neural networks and their hyperparameters, cross-entropy loss, and optimization. You will then delve into the concept of language modeling with n-grams. The module also includes hands-on labs on document classification with PyTorch and building a simple language model with a neural network.

Word2Vec and Sequence-to-Sequence Models

In this module, you will learn about the word2vec embedding model and its types. You will also be introduced to sequence-to-sequence models and how they employ Recurrent neural networks (RNNs) to process variable-length input sequences and generate variable-length output sequences. You will gain insights about encoder-decoder RNN models, their architecture, and how to build them using PyTorch. The module will give you knowledge about evaluating the quality of text using perplexity, precision, and recall in text generation. In hands-on labs, you will integrate pre-trained embedding models for text analysis or classification and develop a sequence-to-sequence model for sequence transformation tasks.