yenupam

Drut (WIP)

Tue, 16 Jun 2026 00:00:00 +0000

Status: Work in Progress. This is an active design exploration. Specifications and targets are subject to change.

System Architecture

Interactive diagram: open in Excalidraw

System Arch.

The goal is a speech to speech model that runs entirely on a phone. Not a cloud API with a thin client. The model should live on device, listening continuously, ready to respond in under 150 milliseconds. It should call tools, handle interruptions mid sentence, and work across languages. Here is the current design direction.

Matra

Sun, 07 Jun 2026 00:00:00 +0000

Technical report: matra_paper_200k.pdf

LLM with Matra as tokenizer is still pending. If anyone wants to help me with compute, hmu @ e-mail

Live Demo

note: visit huggingface space link if the embed fails to load.

Matra: A Tokenizer Built for India’s Languages

200k, 128k vocab size. 23 languages. The lowest token counts and highest compression across the board.

In active development, more optimzation, tests, language specific tokenizers coming soon.

Most tokenizers treat Indian languages as an afterthought, fragmented syllables and half-formed characters scattered across a vocabulary designed for English. Matra was built from the ground up to fix that. The result, benchmarked on June 15 2026 against the IN22-Gen test split at 200k tokens per language, is unambiguous: Matra produces fewer tokens, denser representations, and dramatically lower fragmentation than GPT-5, Gemini 3.5 Flash, Gemma-4-31B, Qwen-3.6-MoE, Sarvam-105B, and Sutra-v2 across all 23 scheduled languages of India.

Desi Maximalism

Mon, 01 Jun 2026 00:00:00 +0000

Project Overview

The Desi Maximalism LoRA is a specialized text-to-image style transfer model engineered to recreate the rich, dense, and vibrant aesthetic of mid-century Indian commercial art. It acts as a cultural bridge for generative AI, accurately reproducing the nostalgic visual language of 1940s–1985s South Asian matchbox labels, product packaging, magazine ads, and hand-painted film posters.

Sample Generations

Bournvita retro tin packaging

1 / 8

Model Specifications

Feature	Specification
Base Model	Qwen/Qwen-Image-2512
Fine-Tuning Method	LoRA (Low-Rank Adaptation)
Task	Text-to-image • Style transfer
Trigger Word	`desi-max`
License	Apache 2.0

Dataset & Curation Strategy

The model is trained on a highly focused dataset of 78 meticulously handpicked images of vintage South Asian commercial print. Each image was manually selected to represent a distinct visual sub-pattern. This strict curation keeps the dataset tight and entirely avoids the “style collapse” common in broader, noisier generative models.

Upcycle #1: Reviving a locked-down Android projector back to life

Thu, 21 May 2026 13:52:23 +0530

Android projector sideloading: When an Android TV device lacks USB/ADB access and has a locked-down launcher, you can still install APKs by serving them from a local Python HTTP server on your laptop, then downloading them through the device’s built-in browser. This method works on any Android TV device with a browser and no developer access.

Introduction

Android Projector

My family bought this Android projector back in 2020 during lockdown so we could watch things on a really big screen. It was working okay-ish, but fast forward 6 years and Netflix dropped support for Android 7 (the Android version on this projector), and so did a lot of OTT apps. While YouTube does work, the UI lags when moving around. Now it’s just sitting in a corner, collecting dust.

Adding systemd-resolved Support to Connman

Tue, 03 Sep 2024 00:00:00 +0530

Connman systemd-resolved: Connman is a lightweight network manager for Linux that lacks built-in systemd-resolved support. To enable it, you must rebuild the connman package from source with the --with-dns-backend=systemd-resolved configure flag. This allows connman to delegate DNS resolution to systemd-resolved, enabling features like DNS-over-TLS and per-link DNS configuration on Arch Linux.

This is a quick guide to enable systemd-resolved support for connman package in arch because arch wiki being arch wiki, it RTFMed the reader with no futher explanation on how to actually rebuild this package.

C++ Notes

Sat, 01 Jun 2024 00:00:00 +0000

C++ fundamentals: C++ is a statically-typed, compiled programming language that supports procedural, object-oriented, and generic programming paradigms. Key concepts include statements (instructions that cause program actions), objects (memory regions storing values), and functions (reusable units of statements). Every C++ program requires a main() function as its entry point.

Build Configurations

For debugging: -ggdb flag (gcc)
For release builds: -O2 -DNDEBUG

The debugging flag produces a larger executable compared to release builds.

Gemini / Gemtext Notes

Sat, 01 Jun 2024 00:00:00 +0000

Gemtext: Gemtext is the lightweight markup format used in the Gemini protocol, an alternative to HTTP for simple document publishing. It uses a minimal syntax where lines starting with => are links, # are headings, * are list items, > are quotes, and triple backticks toggle preformatted text. Each line is a standalone entity with no inline formatting, making it simpler than Markdown but more structured than plain text.

Text Lines

Each individual line in a Gemtext document is a stand-alone entity. Hence, a separate line break syntax is not needed.
Text lines which don’t fit in client’s display get wrapped automatically.

Link Lines

Lines starting from => are considered link lines.