WindstormLabs/windy-core-ggml

Windy Core — ggml q5_1 (mobile build)

Windstorm Labs is the origin of record for this artifact. It is created here and promoted downstream; it is never edited anywhere else.

whisper.cpp / whisper.rn build of Windy Core, quantized to q5_1 so it runs on phones and tablets.

File	ggml-windy-core-q5_1.bin
Size	190.1 MB
SHA-256	924cf246d2a9773554f1546dbdc7bb272de6ec8fe6dca9a774ae60047b24ca7c
Quantization	q5_1
Languages	multilingual
Certified WER	5.85% (grade A)

Lineage

Stage	Value
Upstream base	openai/whisper-small (244M params, MIT)
Windy fine-tune	LoRA, merged — Clinic patient windy-core
Converted from	WindyWord/listen-windy-core @ 0781aae8f000814b6571a4c227fad6685b04de64
Converter	whisper.cpp @ 2ca53bb45e38, models/convert-h5-to-ggml.py -> whisper-quantize q5_1
Built on	Veron-1-5090, 2026-08-01

Promotion path

WindstormLabs (here, origin) -> WindyWord (product) -> WindyWordMobileApp (mobile).

Downstream copies are push-only. Fix weights here, never downstream.

Honesty note

Distinct weights from upstream (different SHA-256). On short clean English clips the smaller models can be indistinguishable from stock Whisper — a formal bake-off has not been run. Do not claim quality gains without one.

Clinic patient file: sneakyfree/Windy-Clinic -> stt-models/windy-core.json

Licence

Upstream base is apache-2.0 on Hugging Face, verified against the source repo we converted from. Note OpenAI's Whisper GitHub repo states MIT while its HF model cards state Apache-2.0; we follow the HF distribution point, since that is what we actually pulled. Windy weights are a derivative and are distributed under the same terms, with attribution upstream.