Urgent.News

What's breaking now, across thousands of outlets.

Tech

Which Copy Do I Keep? Sloth-Powered Folder Tidying with a Local Gemma

This is a submission for the Hacktoberfest Open-Source AI Challenge: Week 3 — Slow and Steady . Your Downloads folder has contract.pdf , contract_final.pdf , contract_final_v2(1).pdf , and contract_FINAL (copy).pdf . You know one of them is the real one. You don't know which one. So you open all four, squint at the footer, find the signed copy, close three tabs — or worse, you pick wrong and sign…

The Hacktoberfest Open-Source AI Challenge presents "Keeper", an AI tool for sorting and managing folders. Imagine having a sloth's patience for tedious file organization tasks; Keeper could automate the process, thanks to its unique approach. It scans folders, identifies duplicates and near-duplicates, and decides which copy is the legitimate one to keep.

Instead of deleting files, it moves the losers into a quarantine folder, leaving the original untouched. This tool is built on an open-weight model, ensuring no privacy concerns or exfiltration risks. It's lightweight, runs on low-end devices, and doesn't require any cloud dependencies. Keeper is not only efficient but also fully reversible, ensuring your folder's original state can be restored at any time.

Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at dev.to →

More in Tech

The KV cache is eating your GPU. MQA and GQA are how models fight back

Quick recap from the last post: the KV cache saves you from recomputing old tokens, and it's the reason generation is fast. But it has a bill. And the bill is memory.

  • KV cache consumes 512 KB per token for Llama-2 7B
  • MQA reduces memory usage by 32x compared to traditional MHA
  • GQA balances MQA and MHA, reducing cache size by 4x

Tokenization is why your agent bill looks weird

Models don't read text. They read integers. Between your string and the model there's a tokenizer, and it quietly decides how much you pay, how much fits in the context window, and why your model is…

  • Tokenization divides strings into vocabulary pieces from 30k-200k entries
  • Byte Pair Encoding (BPE) merges frequent character pairs into new symbols
  • Hindi tokenization differs significantly from other languages, using 24 tokens vs 111

More from Tuesday 20 October →