listen-windy-lingua-ms-ct2

WindyWord.ai STT — Malay Lingua (CPU INT8 (CTranslate2))

Part of the Windstorm Labs open model catalogue. Scores, licence and attribution for every model: https://windytranslate.com/models/listen-windy-lingua-ms-ct2

Licence under review: the upstream model mesolitica/malaysian-whisper-small-v2 does not declare a licence on Hugging Face. Do not rely on this repository for commercial use until this note is removed.

Transcribes Malay speech (Austronesian > Malayo-Polynesian).

Quality

• FLEURS WER: 29.9% (50-sample audit)
• CER: 0.0713
• Band: ok
• Source: our 50-sample FLEURS screening (not a publication result)

About this variant

This is the ct2-int8 deployment format of our Malay Lingua STT model. Load it via the ct2-int8/ subfolder.

Part of the WindyWord.ai STT fleet — covering 35+ languages that commercial speech-to-text APIs underserve, with proper dialect / script disclosures where they matter.

Usage

from transformers import WhisperForConditionalGeneration, WhisperProcessor
processor = WhisperProcessor.from_pretrained("WindstormLabs/listen-windy-lingua-ms-ct2", subfolder="ct2-int8")
model = WhisperForConditionalGeneration.from_pretrained("WindstormLabs/listen-windy-lingua-ms-ct2", subfolder="ct2-int8")

Apps

The Windy Word apps are built on this model family.

---

Provenance & License

CTranslate2 INT8 conversion of WindstormLabs/listen-windy-lingua-ms, whose weights derive from mesolitica/malaysian-whisper-small-v2 (community fine-tune by mesolitica). Lineage inferred: same port batch, INT8 size matches the safetensors build of this language.

Licence under review: the upstream model mesolitica/malaysian-whisper-small-v2 does not declare a licence on Hugging Face. Do not rely on this repository for commercial use until this note is removed.