The NLP & LLM Crash Course - Build your AI Chatbot Quickly

Name: The NLP & LLM Crash Course - Build your AI Chatbot Quickly
Rating: 4.6 (4 reviews)

Master NLP and Large Language Models (LLM): Build and deploy your own ChatGPT-like chatbot with Python in record time.

Created byYersel Hurtado

Last updated 9/2024

English

English [Auto],

What you'll learn

Understand how NLP & LLMs and their architecture work
Implement Sentiment Analysis models
Implement Named Entity Recognition (NER) models
Implement Question-Answering models
Learn how to provision your own space with a GPU
Learn how to create a chatbot interface
Learn how to create your own AI chatbot from scratch in just 2 hours
Learn how to use Open Source models like Llama 3.1, BERT, and others

Course content

4 sections • 23 lectures • 1h 42m total length

Introduction1:26
Learn the basics of large language models and natural language processing, explore transformers and pipelines, and deploy your ai chatbot to the cloud with a user interface.
What is an LLM?1:59
Large language models are NLP-based AIs trained on vast data to understand and generate language; private models like ChatGPT contrast with open source options GPT-2, llama 3, and Pi 3.
LLM Architecture3:20
Explore encoder, decoder, and sequence-to-sequence architectures and when to use them. Learn bidirectional encoders for classification, unidirectional decoders for generation, and seq2seq for translation and summarization.
Type of Tasks for LLMs3:42
Explore NLP task types, including sentence classification (sentiment analysis, spam detection) and sentence similarity. Learn word-level tagging (grammar components, ner) and text generation tasks like completion and summarization.

What is the Transformers library?1:23
Learn how the transformer library from Hugging Face enables easy interaction with AI models via pre-trained models, cross-framework support with PyTorch and TensorFlow, and a simple pipeline-based sentiment analysis example.
Link of the Notebook: Sentiment Analysis0:51
Using Pipeline for Sentiment Analysis7:34
Explore how to use the transformers pipeline class for sentiment analysis in Colab, loading Distilbert models and classifying English text as positive or negative.
How to search and pick an NLP model6:54
Learn to search, filter, and select an NLP model for sentiment analysis using Huggingface, matching task, language, and inference capabilities to build a text classification pipeline.
Link of the Notebook: Named Entity Recognition0:29
Using Pipeline for NER8:04
apply named entity recognition with a transformers pipeline in Colab, identify persons, organizations, and locations, and group entities using grouped_entities true for accurate results.
Link of the Notebook: Question & Answer0:29
Using AutoModel and AutoTokenizer for QA8:19
Instantiate the auto model and auto tokenizer for question answering, select a pre-trained model, provide a context and question, and tokenize inputs for PyTorch tensors.
Generate and process the response of Q&A model7:31
Learn to generate and process a q&a model's response by tokenizing input, selecting start and end logits, and converting token IDs to tokens and back to text.

Select the model to implement the chatbot2:27
Navigate to the model tab, select the text generation task, choose the meta llama tree model, read terms and conditions, and submit an access request to implement your chatbot.
Provisiong an space with GPU6:04
Learn to provision a cloud GPU space for an 8-billion-parameter chatbot, compare CPU vs GPU needs, and deploy via Gradio using zero GPU on Hugging Face.
Create the Authentication Token3:35
Generate a read-only authentication token on Hugging Face, name it, and copy it securely since it appears only once. Then store it as a secret in your profile settings.
Setup Llama model7:36
Configure a llama model for text generation in VSCode by loading auto model and tokenizer. Authenticate with a token, specify the model name, and enable cuda if available for performance.
Process and tokenize the inputs7:01
Learn to tokenize inputs for a large language model using gpu-backed huggingface spaces, combine system, history, and user messages into a tokenized chat, and configure max tokens and top probability.
Define terminators and generate the output of the model7:30
Define terminators to signal sentence ends and feed input ids to the model. Decode outputs, skip special tokens, and generate the model's response using max tokens, temperature, and top p.
Create the UI of the chatbot6:37
Build the chatbot user interface with the Gradio library, defining a chat interface and a response function with inputs for system message, max tokens, temperature, and top p.
Deploy your LLM chatbot9:21
Deploy your LLM chatbot by debugging the deployment in the repository, rebuilding environments, installing requirements, and fixing issues like transformer and tokenizer failures, then testing the chat UI.
Repository of our Chatbot0:22

Requirements

Basic Knowlegde of Python

Description

Master NLP and Large Language Models (LLM): Build and deploy your ChatGPT-like chatbot with Python in record time.

Would you like to dive into artificial intelligence and create your own chatbot in just 2 hours? This is possible with our intensive course on NLP & LLM and Generative AI. We will teach you from scratch what a Large Language Model (LLM) is and how to leverage its power to develop innovative applications.

What You'll Learn:

Natural Language Processing (NLP) & Large Language Models (LLMs): Understand the architecture and inner workings of LLMs like GPT.
Transformers Library: Harness pipelines for sentiment analysis and entity recognition—key skills in natural language processing.
AutoClass Models: Get hands-on with AutoModel and AutoTokenizer to build question-answering systems.
Advanced Environments: Set up GPU configurations and create authentication tokens to work with sophisticated AI models.
Build a User Interface for the Chatbot: Create an intuitive chat-style interface to test your chatbot.
Open Source Models: Learn how to choose the right model based on the specific task at hand.
Chatbot Development: Build the chat logic, design an engaging user interface, and deploy your very own LLM-powered chatbot.

What You Need:
All you need is a basic knowledge of Python and a computer to start building your own chatbot.

Who this course is for:

Everyone with basic knowledge of Python

The NLP & LLM Crash Course - Build your AI Chatbot Quickly

What you'll learn

Explore related topics

Course content

Introduction4 lectures • 10min

Transformers9 lectures • 42min

LLM Chatbot Implementation9 lectures • 51min

Course Farewell1 lecture • 1min

Requirements

Description

Who this course is for: