Text-to-Speech
Piper
ONNX
Twi
mirror

Kasanoma Twi (Piper, mirror)

An unmodified mirror of the Twi Piper voice from the michsethowusu/kasanoma release v1, asset kasanoma_model-twi.zip.

We did not train this voice and we did not change it. The two files are the two files inside that zip, byte for byte, with their sha256 below.

Read this before you use it: the licence is not stated

The source names no licence. Its README says, in full:

All Kasanoma models are released under open-source licenses.

There is no LICENSE file in the repository, none in the release zip, and GET /repos/michsethowusu/kasanoma/license answers 404. "Open-source licenses" is not a licence: it does not say which one, and the terms that reach you are therefore unknown.

The metadata here reads license: other for that reason, and not because some other licence applies. If you need defined terms, ask the author at kasanoma@kasanoma.org before you use this. This mirror exists so the files stay reachable and checkable, and it grants nothing the source has not granted.

Attribution

Files

file bytes sha256
model.onnx 63516050 a2e7f349fcca7b445b2c5c18d4a7009e128ea07c39bc4fc7dd6f0d9894afb191
model.onnx.json 4835 06de269e3f7b3afa0c313c7ca9185e199cd9b877dc41a6505a308a2bc87711de

Piper format 1.3.0, 16 kHz.

Downloads last month
12
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support