Downloads · 30 days
0
jupitersong/madi-whisper-turbo-v3-q8
madi-whisper-turbo-v3-q8 is a automatic speech recognition model from jupitersong. Use it when you need speech turned into text. The card lists the license as mit.
Q8 quant of openai/whisper-large-v3-turbo.
Downloads · 30 days
0
Access
Public
Updated Jun 22, 2026
Repo size
867 MB
Likes
1
Public
Click a slice to open those files.
.safetensors867 MB · 100%
From the Hugging Face model README
Q8 quant of openai/whisper-large-v3-turbo.
This is a custom Q8 quantization of OpenAI's Whisper large-v3-turbo, produced for the Madi macOS transcription app (a pure Zig + Metal engine for Apple Silicon). The Madi app downloads this file on first run instead of bundling it, keeping the app ~37 MB.
| Base model | openai/whisper-large-v3-turbo (MIT) |
| Quantization | Q8_0 (per-32 block, int8 weights + f16 scales) |
| File | model.q8.safetensors |
| Size | 867,485,320 bytes (~827 MiB) |
| SHA-256 | 1014fd3ad4450a2e43e473eebbab485b165fd68cbe932372071d86c522bb5c8e |
| Size vs F16 | 1.86× smaller |
| Quality | Bit-identical transcripts to the F16 model (verified) |
The weights are pre-quantized exactly as the Madi engine quantizes them in memory
at load time — per-32-element blocks, scale = max(|w|) / 127 stored as f16,
q = clamp(round_half_away(w / scale), -127, 127) as int8. Linear/projection
weights are stored as K.qs (int8) + K.scales (f16); layernorm, bias, conv, and
positional-embedding tensors are copied in F16 unchanged.
Because the offline quantization is identical to the engine's in-memory path, loading this Q8 file produces transcripts that are byte-identical to loading the original F16 weights — the 1.86× size reduction is free of quality loss. This is a verified contract (F16-load vs Q8-load output equality), not an approximation.
This is not a general-purpose checkpoint. The safetensors file uses a layout
specific to the Madi Metal engine (.qs / .scales keys). It is not compatible
with 🤗 Transformers, whisper.cpp (GGML/GGUF), faster-whisper, or other Whisper
runtimes. Use one of those projects' own quantized checkpoints for those runtimes.
Released under the MIT License, inherited from the base model
openai/whisper-large-v3-turbo (also MIT). When redistributing, please retain
attribution to OpenAI's Whisper.