Ravi Kumar Blog Posts

Streamlit is an open-source Python framework for building data applications with minimal code. It turns Python scripts into interactive web apps without requiring HTML, CSS, or JavaScript knowledge. The idea behind this UX framework is like write a Python script, and Streamlit automatically creates a web interface. The fundamental concept in Streamlit is like the script reruns on every interaction and think like the script is a function that gets called on every user interaction. Streamlit is gaining traction among developers working with large language models (LLMs) for its ability to streamline web application development. P.C: Google Images Installation & Running: # Install C:\Users\Admin> pip install streamlit # Create a file: app.py # Run it C:\Users\Admin> streamlit run app.py The app opens automatically in your browser at http://localhost:8501 The simple python script used to display the UI elements in the browser are shown below. We can build the simple graphs using ...

DuckDB is an open-source, in-process analytical database management system (OLAP) designed for fast analytical queries. It's often described as "SQLite for analytics" and was first released in 2019 by the Database Architectures group at CWI (Centrum Wiskunde & Informatica) in the Netherlands. DuckDB is a powerful and high-performance in-process database for fast analytical processing. It is useful for data science and engineering-based applications and provides a full-featured database without the deployment of a separate dedicated database server. It can interface with multiple programming languages and scripts including SQL, Python, R, Java, and Node.js. Pic Credit : Google Images Key Features: In-process database: Runs embedded within your application (no separate server process) Column-oriented storage: Optimized for analytical queries that scan large amounts of data Zero dependencies: Single file deployment, no external dependencies Rich SQL support: PostgreSQL-c...

We learned about Ollama and how to work on a model locally in the previous post:  Click Now we can extend our knowledge to build a simple ChatBOT using Python 3.14 and the respective langchain packages as per earlier post. Before working on this exercise, we need to make sure that Ollama server up and running and make sure to pull the llama3.2 model locally.  Pic Credits : Google Images Here is the sample Python program in Visual Studio environment to build a conversation (text) between you and the Bot.  Now we can run the program from Visual Studio terminal or we can run it from the external command line prompt like below. The good part of this exercise is like we don't need an OpenAI key, any 3rd party service model and we can run this python script from our computer locally with no internet connectivity required. Why because, we already installed or pulled the needed model in our local sys and use that LLM from our sys using Ollama server running locally.

Ollama is used to run and manage large language models (LLMs) locally on your own computer for offline, private, and efficient AI operations. It simplifies downloading, running, and customizing open-source LLMs like Llama 3 or Mistral through a simple command-line interface, making it a valuable tool for developers, researchers, and anyone concerned with data privacy. Ollama server is used to serve large language models (LLMs) locally on a user's machine, making them accessible via an API, a web UI, or through various applications like code assistants. It provides a simple way to download, run, and manage models for tasks such as building chatbots, developing applications, and conducting research, with a focus on local execution for data privacy and lower latency. Here are simple steps to work on this exercise: 1) Install Ollama: Download and install Ollama from Ollama Download 2) Start the server: Run ollama serve in a separate terminal to make the server available. (This is of...