Ga - UNICEF WCARO NLP

🌍 General Information

ISO 639-3gaa

ISO 639-1—

French Namega

Glottocodegaaa1244

Language FamilyNiger–Congo

Writing SystemLatin (Ga alphabet) Ghanaian braille

L1 Speakers745,000

L2 Speakers—

Population820000 (10⁵)

Endangermentsafe

CountriesGhana

Also Known AsAccra, Acra, Amina, Gain, Gamei

WikipediaGa on Wikipedia

🗺️ Language Use

🗺️

Language use data not yet available for Ga in CLEAR Global's platform.

See the sections below for other available data.

👥 Actors (3)

AdwumaTech AI (Startup)

AdwumaTech AI provides enterprise-grade AI training data infrastructure powered by African engineering talent. Core offerings include RLHF for code ge...

Farmerline / Darli AI (Startup)

Ghanaian agritech company building technology solutions for smallholder farmers. Core platform Mergdata is a cloud-based farm management system used b...

Ghana NLP / Khaya AI (Research Community)

Ghana NLP is an open-source initiative focused on Natural Language Processing for Ghanaian and African languages. They developed Khaya, the first AI-p...

🤖 NLP & Tech Resources

✓ AfroLID

✓ Google Translate: 2024

✓ Fineweb2: 💬

✓ MMS-LID

✓ SimbaBench

Benchmarks

ASR (Speech-to-Text)

SimbaBench (FinancialSpeech)

Model	WER	CER	Source
facebook/mms-1b-all	135.4	53.98	reported
openai/whisper-large-v3	725.03	527.18	reported
openai/whisper-large-v3-turbo	260.3	180.07	reported
facebook/seamless-m4t-v2-large	333.08	244.96	reported
UBC-NLP/Simba-S	10.42	7.17	reported
UBC-NLP/Simba-M	19.88	7.41	reported
UBC-NLP/omniASR_LLM_7B_v2	136.49	74.42	reported
UBC-NLP/omniASR_LLM_1B_v2	135.23	66.47	reported
UBC-NLP/omniASR_LLM_300M_v2	133.05	80.66	reported
UBC-NLP/omniASR_CTC_7B_v2	118.59	45.35	reported
UBC-NLP/omniASR_CTC_3B_v2	115.48	43.67	reported
UBC-NLP/omniASR_CTC_1B_v2	108.15	43.95	reported

📝 Notes

Observations

Use Cases Document: Ghana AGOO platform priority language (important for Accra/coastal areas)
Ghana has active UNICEF NLP work

Data

GhanaNLP Parallel Corpora — 11,652 human-translated Ga-English sentence pairs (second largest in collection). Multi-domain. CC BY-NC-SA 4.0. Datasets on HuggingFace but not tagged with language IDs: https://huggingface.co/collections/Ghana-NLP/data-collection-project

ASR Models (3)

Model	Size	Downloads	Likes
espnet/xeus	—	60	146
UBC-NLP/Simba-SLID-49	94.6M	15	0
AyikuBen/whisper-ga-finetuned-demo	37.8M	2	0

View on HuggingFace: gaa: 3

TTS Models (0)

No TTS models found.

MT Models (17)

Model	Downloads	Likes
Helsinki-NLP/opus-mt-fr-gaa	42	0
Helsinki-NLP/opus-mt-sv-gaa	41	0
Helsinki-NLP/opus-mt-fi-gaa	38	0
Helsinki-NLP/opus-mt-gaa-sv	38	0
Helsinki-NLP/opus-mt-gaa-de	37	0
Helsinki-NLP/opus-mt-gaa-en	36	0
Helsinki-NLP/opus-mt-gaa-fi	36	0
Helsinki-NLP/opus-mt-es-gaa	33	0
Helsinki-NLP/opus-mt-gaa-fr	33	0
Helsinki-NLP/opus-mt-gaa-es	30	0
Helsinki-NLP/opus-mt-de-gaa	26	0
Helsinki-NLP/opus-mt-en-gaa	24	0
masakhane/africomet-stl-1.1	0	5
masakhane/africomet-qe-stl-1.1	0	5
McGill-NLP/ssa-comet-mtl	0	3
McGill-NLP/ssa-comet-stl	0	1
McGill-NLP/ssa-comet-qe	0	1

View on HuggingFace: gaa: 17

LLM Models (1)

Model	Size	Downloads	Likes
ghananlpcommunity/opani-chat_1b-merged-16bit	1B	2	0

View on HuggingFace: gaa: 1

Datasets

ASR Datasets (2)

Dataset	Rows	Downloads	Likes
UBC-NLP/SimbaBench_dataset	748.0K	975	0
ghananlpcommunity/ga-speech-text-parallel-90k	98.3K	221	0

View on HuggingFace: gaa: 2

TTS Datasets (2)

Dataset	Rows	Downloads	Likes
UBC-NLP/SimbaBench_dataset	748.0K	975	0
ghananlpcommunity/ga-speech-text-parallel-90k	98.3K	221	0

View on HuggingFace: gaa: 2

Translation Datasets (8)

Dataset	Rows	Downloads	Likes
gtak1/panlex-meanings	64.1M	6.3K	0
google/smol	798.0K	2.8K	94
cointegrated/panlex-meanings	157.0M	255	13
community-datasets/udhr	—	123	2
cis-lmu/GlotStoryBook	80.1K	67	7
FrancophonIA/gatitos	—	61	4
kargaranamir/GlotStoryBook-backup	80.1K	21	0
Bretagne/PolyWrite_clean	155	11	0

View on HuggingFace: gaa: 8