People

Research project

Kashmir AI Research

Faizan Ayoub

This site's own work: Kashmiri↔English translation research, an open evaluation platform with human ratings, a public Kashmiri→English model, and this catalogue.

ResearchHugging Face
1
Papers
1
Datasets
1
Models
1
Tools

Focus

Papers 1

paper
2026

Anatomy of Decoder-Only LLM Failure in Low-Resource Machine Translation

Faizan Ayoub, Neha Prerna Tigga · ICRTC-2026 (Springer) — Best Paper Award

Best Paper at ICRTC-2026. First quantified tokenizer fertility analysis for Kashmiri (4.06x overhead), a sharp cross-lingual transfer threshold near 30K pairs, and a deployable source-constrained decoding strategy worth +12.75 BLEU without retraining. Compares QLoRA Mistral-7B against NLLB-200 and IndicTrans2 on a 124,102-pair corpus.

Best PaperTokenizer FertilityQLoRA
Open Access
Featured

Datasets 1

dataset

Kashmir AI Research Parallel Corpus

Kashmir AI Research · Kashmir AI Research

A 124,102-pair Kashmiri–English training corpus aggregated from AI4Bharat BPCC and the SMUQamar corpus, deduplicated and filtered on length and script. The sentence pairs are the work of their original authors; the aggregation, filtering and split are ours. Not redistributed.

Parallel CorpusAggregatedFiltered
Featured

Models 1

model
2026

Kashmir AI Research NLLB-600M Kashmiri→English (CTranslate2)

Faizan Ayoub · Kashmir AI Research

NLLB-200 600M fine-tuned by Kashmir AI Research on 93K BPCC Kashmiri-English pairs and exported to CTranslate2 int8. It powers the translator at kashmirairesearch.online/translate.

Kashmiri→EnglishFine-tuned
CC-BY-NC-4.0

Tools 1

tool

Kashmir AI Research — Evaluation Platform

Faizan Ayoub · Kashmir AI Research

Blinded human evaluation of Kashmiri MT: native speakers score translations (adequacy, fluency, preference) and mark error spans. The current study runs at kashmirairesearch.online/consent.

EvaluationHuman-in-the-LoopError annotation
Featured

Profile 1

org
2026

Kashmir AI Research

Founded by Faizan Ayoub

This platform. Kashmiri→English machine translation research, a human evaluation platform for Kashmiri MT, a live Kashmiri→English translator, and this resource hub — built to give Kashmiri NLP a single home.

Research PlatformHuman EvaluationOpen
Featured

Worked with

  • Neha Prerna Tigga

Every work above is a record in the Kashmiri NLP resource hub, credited as published; profile details come from kashmirairesearch.online. Something missing or wrong? Submit it.