Skip to main content

Blog @ Codonomics

Democratizing Knowledge & Experience

Search This Blog

My Footprints On Internet

Open Source
Android Play Store
LinkedIn
Medium
GoodReads
StackOverflow
Testimonial
My Quotes
About Me

Showing posts from July, 2025

Posts

Buy @ Amazon

Lessons From Failing To Fine-tune A Small LLM On My Laptop

on July 27, 2025

AI AI Agent AIOps Fine-Tuning FinOps LLM Performance Engineering Prompt Engineering

+

Get link
Facebook
X
Pinterest
Email
Other Apps

Ollama Cheatsheet

on July 16, 2025

Cheatsheet Gen AI LLM Ollama

+

Get link
Facebook
X
Pinterest
Email
Other Apps

Subscribe

Posts

Posts

All Comments

All Comments

Followers

Archives

2025 22

2024 22
- October 6
- September 3
- July 1
- June 4
- May 1
- March 1
- February 1
- January 5
2023 22
- December 1
- September 3
- August 4
- July 6
- May 3
- April 3
- March 1
- January 1
2022 9
- September 1
- August 2
- July 1
- June 2
- March 1
- January 2
2021 28
- December 3
- November 5
- October 8
- September 1
- August 2
- July 3
- June 1
- May 1
- April 1
- March 1
- February 2
2020 45
- December 2
- November 2
- October 7
- September 5
- August 10
- July 6
- June 2
- May 3
- April 2
- March 2
- February 1
- January 3
2019 31
- December 1
- October 1
- September 4
- August 2
- July 1
- June 1
- May 1
- April 2
- March 5
- February 6
- January 7
2018 22
- December 7
- November 2
- August 5
- April 3
- March 5
2017 20
- December 2
- November 1
- October 1
- September 4
- August 7
- July 3
- June 1
- March 1
2016 12
- August 3
- July 1
- April 2
- March 3
- February 2
- January 1
2015 5
2014 33
- September 4
- August 6
- May 3
- March 5
- February 7
- January 8
2013 9
- December 3
- July 1
- May 5
2012 7
- December 1
- October 1
- August 2
- May 3
2011 26
- November 4
- October 5
- September 3
- June 4
- May 1
- April 3
- March 2
- February 2
- January 2
2010 12
- December 1
- November 2
- October 2
- September 3
- August 3
- July 1

Show more Show less

Labels

10 commandments
12 Factor App Principles
Aadhar
AB Testing
Acer Predator Triton 300
achievement
Adopt AWS Rightly
Advice
Agile
agile adoption

Show more Show less

Stack Exchange User Flair

Popular Posts

The Truth About AI Coding Agents

The Truth About AI Coding Agents What They Can and Can’t Do for Your Engineering Teams Image generated by Microsoft's Copilot inbuilt into Edge Browser A not so recent "warning" email from the CEO of a leading tech company to its employees, proclaiming that AI was already handling a significant percentage of development work and would inevitably lead to widespread developer redundancy, ignited a wildfire of anxiety among developers and unfortunate rash decisions by department heads. Suddenly, CEOs of other companies were following suit, downsizing its development team and pouring investments into the promise of a magical AI coding agent. In our age of virality, it seems bad news and exaggerated claims spread with alarming speed and influence. Let's be clear: this was, at best, a brilliant marketing play by a CEO aiming to dominate the burgeoning AI coding agent market – and it undeniably paid off handsomely. In my conversations with executive leadership, I advise them...

Lessons From Failing To Fine-tune A Small LLM On My Laptop

I recently shared my frustration in my LI feed something like below: Getting supposedly small things done by monster of an LLM to me is still an expensive affair in terms of money. Getting the same damn small thing done by quantized LLM is seeming to be expensive in terms of time. Prompt Engineering they say is the new language. The reality is LLMs still haven't matured enough to select right tools with simple one-shot or few-shot prompting. I didn't struggle teaching my pet dog to pick the right tool, as much as I am doing teaching my relatively small LLM running on my laptop to select right tool from a given set of tools to generate appropriate answers. I love and do bet on GenAI but am cognizant of the cost vs effort tradeoff as with anything else in software engineering, but more blatant in the Generative AI ecosystem. Yes, it is relatively much much easier to leverage an LLM with 70 b...

Essential Vim For CKAD Or CKA Exam

If you are preparing for CKAD or CKA exam and are new to Vim/Vi, this post will help you know everything that is essential to being productive with Vim during your exam. Vim has 3 modes as depicted in picture below: In Command Mode , you do things like navigating your cursor from one point to another in your text file, or delete text or replace text or select text or things like that. When you open Vim, you will be in this mode to begin with. In Insert Mode , you insert text and delete text with [backspace] key :) In Execution Mode , you do things like search text in file, set tab length, set line numbers in editor and stuff like that.

Connecting to Expo Dev Server in WSL 2 over LAN

This post drafts things I did to connect my Android device to Expo dev server running in WSL 2 on Windows 10 over LAN connection (in my case using Wi-Fi). The set-up looks like below: Note: Expo-Client is installed on my android device. Note: Expo-CLI npm version used is 3.27.14 Assumption: IPv4 address is used over IPv4. When you run ` ip address ` in your WSL 2 terminal, you will your ip-address of your VM as something like 172.x.x.x . When you check the ip address of your Windows 10 host, by running ` Get-NetIPConfiguration ` in your PowerShell terminal you will find your host ip to be something like 192.168.x.x . When you connect your Android device to the same LAN over Wi-Fi, your ip address would be similar (something like 192.168.x.x). The fact that WSL 2 isn't sharing the same ip address as its host is an issue to be resolved. This needs to be fixed as first thing, because otherwise how do we even reach the server running in WSL over LAN? The easy fix to this is t...

Android Activity's onDestroy() Ain't Your Reliable Friend

Look at the Android Activity life-cycle events in the picture beside. The lifecycle isn't something new. You're familiar with it right? You're also that good Android developer that writes clean code, wherein you consciously override onDestroy() method to include resources clean-up code and save some state, right? Per Android doc , it is okay and usually the practice to write cleanup code in this method but not anything beyond that. So if you are to write code to persist state in this method, or anything else beyond resource clean-up (like threads associated with an activity), you're better off to push it up the life-cycle, either in onStop() or onPause() method. That is how I settled in the first iteration of new learning but wasn't happy. Why? To quote Android Doc:

My Past Speaking Engagements

Powered by Blogger

Copyright 2021