ChatGPT

-
ChatGPT is a free-to-use AI system.
Chat Generative Pre-Trained Transformer, is a sophisticated language model.
ChatGPT is a chatbot developed by OpenAI and launched on November 30, 2022.
Based on a large language model, it enables users to refine and steer a conversation towards a desired length, format, style, level of detail, and language.

ChatGPT technology is built on a transformer-based architecture known as the Generative Pre-trained Transformer (GPT). 
This architecture has transformed Natural Language Processing (NLP), allowing ChatGPT to interpret and create text in a more sophisticated manner.
Computer systems that imitate the human brain are known as Neural Networks.
Neural Networks are made up of layers of connected artificial neurons that work together to examine and interpret data. 
In the case of ChatGPT, these neural networks are specifically designed for text analysis and creation.

AI Tools
    1. Natural Language Processing (NLP)
        NLP Tools
            Tokenization
            Part-of-Speech Tagging
            Named Entity Recognition (NER)
            Sentiment Analysis
            Language Translation - OpenNMT
            Text Summmarization
            Chatbots
            Word Embeddings
            Synatx Anlysis
            Named Entity Linking(NEL)
    2. Computer Vision
    3. Machine Learning Frameworks
        TensorFlow
        PyTorch
        scikit-learn
    4. Data Analytics
        Tools:
            Pandas
            NumPy
            Tableau
    5. Speech Recognition
        Tools:
            Google's Speech=to-Text
            CMU Sphinx
    6. Virutal Assistants
        Siri
        Google Assistants
        Alexa
    7. Predictive Analytics
    8. Robotics
    9. AI Chatbots
    10. Reinforcement Learning

OpenNMT
OpenNMT is an open-source neural machine translation(NMT) framework that facilitates the development and deployment of machine translation models.
1. Neural Network Architecture
2. Preprocessing Tools
3. Model Training
4. Embeddings
5. Inference and Translation
6. REST API
7. Scalability
8. Support for Multiple Languages

ChatGPT is a language model developed by OpenAI. It is designed for natural language understanding and generation, specifically in a conversational context.

Generative AI application
Trained on billions of documents.
Based on Generative Pre-Trained Transformer (GPT) architecture, a type of neural network.
Web interface for end-users, API for developers.

Large Language Models (LLMs)
Large Language Models (LLMs) are powerful artificial intelligence models designedfor understanding and generating human-like text.
Neural network - Transformers

Prompt Engineering
A prompt is a specific question, command, or input that you provide to an AI system to request a particular reponse, information and action.
Prompt engineering is the process of crafting well-defined and structured input queries to interact with AI systems, in order to get accurate and relevant responses.

Embeddings
Embeddings are numerical representation of text.
They are essential for AI models to understand and work with human language effectively.


Fine Tunning
Fine-tuning a LLM is the process of adapting a pre-trained model to perform specific tasks or to cater to a particular domain more effectively.

The simulation of human intelligence processes by computer systems.

Machine Learning
Deep Learning
Natural Language Processing -> Large Language Models (LLMs), GPT (Generative Pretrained Transformer), ChatGPT
Computer Vision
Robotics
Speech Recognition

NLP/GPT/ChatGPT - Generate human-like text based on an input
DALL-E - image generation
WHISPER - generate English text

Building the prompt:

The 5W questions and HOW framework
5W are: what, who, where, why, which

what are you asking the system to perform?
what is the desired output?
what is the company's business?
who is asking?
who should the AI simulate?
who are the firm's clients?
where? based
why? explain the purpose.
Which aspects should be included?
How does the AI respond?


Github - copolit
AWS Code wishperer

GitHub Copilot
//prompt for required code example
eg:
//main - will give main method
/explain
/fix
/review and suggest potential
/help
/clear

Comments

Post a Comment

Popular posts from this blog

PL/SQL

-
PL/SQL PL/SQL stands for Procedural Language extensions to the Structured Query Language (SQL). PL/SQL can execute a number of queries in one block using single command. One can create a PL/SQL unit such as procedures, functions, packages, triggers, and types, which are stored in the database for reuse by applications. The PL/SQL engine resides in the Oracle engine. The Oracle engine can process not only single SQL statement but also block of many statements. Structure of PL/SQL Block: DECLARE     declaration statements; BEGIN     executable statements EXCEPTIONS     exception handling statements END; / Example : SET SERVEROUTPUT ON; SQL> DECLARE        var varchar2(40) := 'I love GeeksForGeeks' ;      BEGIN        dbms_output.put_line(var);      END;      / PL/SQL Functions PL/SQL functions are reusable blocks of code that can be used to perform specific tasks. They are...
Read more

JAVA8 Features

-
Java8 Features 1) Lambda Expressions : Concise functional code using ->. Lambda Expression basically expresses an instance of the functional interface, Lambda Expression provides a clear and concise way to represent a method of the functional interface using an expression. 2) Functional Interfaces : Single-method interfaces. An interface that contains only one abstract method is known as a functional interface. There is no restriction, that can have n number of default and static methods inside a functional interface. A functional interface is an interface that contains only one abstract method.  @FunctionalInterface annotation is used to ensure that the functional interface can’t have more than one abstract method. Runnable, ActionListener, and Comparable are some of the examples of functional interfaces.   Runnable –> This interface only contains the run() method.   Comparable –> This interface only contains the compareTo() method.   ActionListener –> ...
Read more

Build Automation

-
ANT: Ant was the first build tool released in 2000. Ant was developed on procedural programming idea. Ant was improved with an ability to accept plug-ins and dependency management over the network, with the help on Apache-IVY. Drawbacks: 1) XML is used as a format to write the build scripts. 2) Being hierarchical is not good for procedural programming, and 3) XML is relatively unmanageable.   MAVEN: Maven was introduced in 2004. Maven relied on the conventions and was able to download the dependencies over the network. Benefits : Life cycle of Maven, while following the same life cycle for multiple projects continuously. Problems: 1) Maven does not handle the conflicts between versions of the same library. 2) Complex customised build scripts are difficult to write in Maven, as compared to writing the build scripts in ANT. GRADLE : Gradle came into picture in 2012 Features  Declarative builds and build-by-convention 1) Gradle is available with separate Domain Specific Language ...
Read more