Research

We didn't scrape the internet. We scanned the shelf.

Most Assamese text on the open web is thin, repetitive, or badly OCR'd. So we built our own corpus the slower way: scanning printed sources page by page and cleaning what came out.

Abstract archival fibers transforming into computational pathways
Corpus research · 05
200M+tokens scanned to date
0tokens from raw internet scrapes
32Kvocabulary in our Assamese tokenizer
growingcorpus, scanned continuously
Why not just use the internet

Internet-scale data has an internet-scale problem.

Common web crawls are the default corpus for most language models. For a low-resource language like Assamese, the web is thin, and what exists is skewed toward short, repetitive, machine-translated, or informal text. That's a poor foundation for a model meant to write and reason in formal, literary, and technical Assamese.

What we scan instead
  • Published books and literary works
  • Magazines and periodicals
  • Newspapers, current and archival
  • Government publications and gazettes
  • Educational material and textbooks
What that buys us

Cleaner grammar, wider vocabulary, real editorial register, and coverage of formal Assamese that web text rarely contains. It also means a much smaller share of duplicate or near-duplicate content to filter out before training.

Pipeline

From scanned page to training token.

Scan & extract

Printed pages are digitised and run through our own OCR line, the same vision models behind Lipi Print and Lipi Hand.

Language identification

Assamese and Bengali are distinguished using script-level heuristics (most reliably the ৰ/ৱ distinction), backed by a fastText classifier.

Quality filtering

A perplexity model trained on Assamese Wikipedia flags low-quality or corrupted OCR output before it enters the corpus.

Deduplication

Document- and paragraph-level MinHash deduplication removes reprints and repeated boilerplate across sources.

Corpus explorer

One language. Multiple registers.

Explore representative, editorially written examples of the registers the corpus is designed to cover. These are previews, not excerpts from licensed training documents.

Books and literary works · curated source preview

Published literature

সাহিত্যই কেৱল সময়ৰ কথা নকয়; ই মানুহে নিজৰ পৃথিৱীখন কেনেদৰে বুজিছিল, তাৰো সাক্ষ্য বহন কৰে।
Open research

Eri is our proof of process.

We open-sourced Eri so the low-resource NLP community can see what training from scratch on a scanned, curated corpus actually produces, before we scale the same process to Muga and beyond.

See the model family →