README.md

---
pretty_name: Kare Code-Switching Illustrative Samples
task_categories:
  - automatic-speech-recognition
language:
  - yo
  - ha
  - ig
  - en
tags:
  - code-switching
  - nigerian-languages
  - yoruba
  - hausa
  - igbo
  - pidgin
  - clinical
  - text-to-speech
size_categories:
  - n<1K
---

# Kare — Code-Switching Illustrative Samples

Eight short audio clips demonstrating the English/Yoruba/Hausa/Igbo/Pidgin
**code-switching** that [Kare](https://kare-health.vercel.app), a voice-first
AI health assistant for Nigeria, is built to understand — submitted as part
of Kare's entry to the **Sahara CodeSwitch Africa Challenge**.

## What this is — and isn't

- **Is:** eight original sentences, written by the Kare team specifically
  for this demonstration, mixing everyday Yoruba / Hausa / Igbo / Nigerian
  Pidgin with embedded English clinical terms — the exact phenomenon Kare's
  [ASR benchmark](https://raw.githubusercontent.com/Vikky-Ajayi/Kare/main/docs/Kare_Benchmark_Report.pdf)
  exists to measure. Audio was synthesized with **Sahara (Intron) TTS**, the
  speech API this challenge provides.
- **Isn't** the benchmark corpus itself. Kare's actual benchmark evaluates
  five ASR models against Intron's own `AfriSwitchCare` and `AfriSwitch`
  datasets — real, professionally-produced datasets gated on Hugging Face
  (contact-approval and manual-approval respectively). Those aren't ours to
  redistribute; this repo links to the source instead:
  - https://huggingface.co/datasets/intronhealth/AfriSwitchCare
  - https://huggingface.co/datasets/intronhealth/AfriSwitch

  This dataset is a small, original, fully-owned set built to give reviewers
  something to actually listen to without going through that gate. No human
  speaker's voice or identity is involved — every clip is machine-synthesized
  from text we wrote — so there is no consent or de-identification question
  to begin with.

## Contents

8 clips, ~35 seconds total, 2 per language (Yoruba, Hausa, Igbo, Nigerian
Pidgin), each pairing everyday matrix-language grammar with embedded English
clinical vocabulary (symptom names, drug names). See `metadata.csv` for the
full table (`file_name`, `language_code`, `language`, `duration_s`, `text`,
`english_gloss`, `voice`).

| Clip | Language | Duration | Text |
|---|---|---|---|
| yoruba-01 | Yoruba | 3.48s | Doctor, ara mi gbóná, and mo ní headache láti àárọ̀. |
| yoruba-02 | Yoruba | 3.05s | Mo ti mu paracetamol, but the fever ò tíì lọ. |
| hausa-01 | Hausa | 4.79s | Likita, ina jin zafi a ciki na, and I've been vomiting since last night. |
| hausa-02 | Hausa | 3.44s | Na sha the malaria drugs, but zazzabi bai tafi ba. |
| igbo-01 | Igbo | 4.57s | Dọkịta, ahụ na-esi m ike kemgbe ụtụtụ, and my whole body feels weak. |
| igbo-02 | Igbo | 2.74s | Ejila m ọgwụ ahụ, but the swelling ka nọ. |
| pidgin-01 | Nigerian Pidgin | 7.14s | Doctor, the malaria don come back. My body dey hot since morning, and the sharp pain for my side never stop. |
| pidgin-02 | Nigerian Pidgin | 5.40s | I don dey take the tablet wey you give me, but the swelling never really go down. |

English glosses for all eight are in `metadata.csv`.

## A note on accuracy

The Pidgin and Yoruba lines were written and checked by the team directly;
Hausa and Igbo constructions were written in-house for this demo and are
common, everyday phrasing, but have **not** been reviewed by a native or
professional linguist. If something reads oddly to a fluent speaker, treat
it as a demo-quality artifact, not a claim of linguistic authority.

## Provenance & reuse

- **Text**: original, written by the Kare team for this submission — free
  to reuse or adapt.
- **Audio**: synthesized via Sahara (Intron) TTS under Kare's access to that
  API for the Sahara CodeSwitch Africa Challenge; shared here for challenge
  review. If you want to reuse the audio itself outside that context, check
  with Intron on Sahara TTS output terms first.
- **Loading**: standard Hugging Face `audiofolder` layout —
  `load_dataset("vikkyblacq/kare-codeswitch-samples")` picks up
  `metadata.csv` automatically.

Built for the [Sahara CodeSwitch Africa Challenge](https://huggingface.co/datasets/intronhealth/AfriSwitchCare).
Kare's full benchmark, code and submission materials:
https://github.com/Vikky-Ajayi/Kare