# Provenance — vokra/wavtokenizer-large-speech-75token

| Field | Value |
|---|---|
| Upstream source | novateur/WavTokenizer-large-speech-75token (single-codebook FSQ audio codec, 24 kHz, hop 320 → 75 tok/s, arXiv:2408.16532, MIT) |
| Upstream licence | `mit` |
| Architecture | `wavtokenizer` |
| Tensors | 1091 |
| SHA-256 | `99b7dce0426266f7f2f6615091d832cea71387ce57edfae66666143a5c33a36b` |
| Converted by | vokra-core 0.1.0-alpha.0 |
| GGUF schema generation | 1 |

Every row is read from the artifact's own `vokra.*` metadata; none of it
is supplied by hand, so this file cannot disagree with the weight it
describes.

## Reproducing

```bash
vokra-cli convert --model <kind> --input <upstream> --output model.gguf
shasum -a 256 model.gguf   # expect 99b7dce0426266f7f2f6615091d832cea71387ce57edfae66666143a5c33a36b
```