voicetide/whisper-phone

Whisper Large v3 Turbo for phones, whisper.cpp ggml (q4_0)

A whisper.cpp ggml file of openai/whisper-large-v3-turbo, quantized to q4_0. Voice Tide's iPhone app downloads it, optionally, as the speech model for its extra dictation languages. The download needs no account, and nothing is sent back.

File	Size (bytes)	SHA-256
ggml-large-v3-turbo-q4_0.bin	473,992,235	2ceb779ca61ce46ed6e9aa43c4a0f8a89a37f683df3c0ba3119238fcd5852155

Origin

• Creator: OpenAI.
• Original model: https://huggingface.co/openai/whisper-large-v3-turbo
• Source file: ggml-large-v3-turbo.bin from ggerganov/whisper.cpp at revision 5359861c.
• Changes: quantized to q4_0 with whisper-quantize from whisper.cpp v1.9.4 (927cfce).

Licence

OpenAI licenses the original model under the MIT License. The copyright notice and the licence are in LICENSE. The file here is a modified version: it was quantized to q4_0.