---

license: cc-by-nc-sa-4.0

language:

• de

pipeline_tag: automatic-speech-recognition

tags:

• audio
• speech-recognition
• transcription
• gguf
• moonshine
• lightweight
• german

library_name: ggml

base_model: fidoriel/moonshine-tiny-de

---

Moonshine Tiny (German, fidoriel) -- GGUF

GGUF conversions and quantisations of fidoriel/moonshine-tiny-de for use with CrispStrobe/CrispASR.

Available variants

| File | Quant | Size | Notes |

|---|---|---|---|

| moonshine-tiny-de-fidoriel-f16.gguf | F16 | 52 MB | Half precision |

| moonshine-tiny-de-fidoriel-q4_k.gguf | Q4_K | 17 MB | Best size/quality tradeoff |

Model details

• Architecture: Conv1d stem + 6L transformer encoder + 6L transformer decoder (288d, 8 heads, partial RoPE, SiLU/GELU)
• Parameters: 27M
• Languages: German (fine-tuned from English moonshine-tiny)
• WER: 11.4% on Common Voice 22 German test set
• CER: 4.2%
• Output: Proper casing and punctuation
• License: CC-BY-NC-SA-4.0 (inherited from upstream)
• Source: fidoriel/moonshine-tiny-de

Usage with CrispASR

# Explicit model path
./build/bin/crispasr --backend moonshine -m moonshine-tiny-de-fidoriel-q4_k.gguf -f audio.wav

# Or via backend name (auto-download)
./build/bin/crispasr --backend moonshine-tiny-de -m auto -f audio.wav

Notes

• Moonshine models run on CPU only (GPU not needed for these small models)
• Tokenizer (tokenizer.bin) must be in the same directory as the model file
• Smaller/faster alternative to moonshine-base-de (17 MB vs 39 MB)

Provenance and EU AI Act Art. 53 note

• Upstream model: fidoriel/moonshine-tiny-de — published by fidoriel.
• Upstream licence: cc-by-nc-sa-4.0. This repository redistributes under the same terms; it grants no rights the upstream licence does not.
• What was done here: format conversion and/or quantisation only (GGUF/GGML). No training, no fine-tuning, no merging, no distillation, no change to architecture, vocabulary or capability. Only the numeric representation of the upstream weights differs.
• Training data: documented — where it is documented at all — by the upstream provider; see the upstream model card. No training data was used, added or selected by this repository. No training-content summary was found on the upstream model card at the time of writing; that documentation gap is upstream's and is not filled here.
• Provider status: under Regulation (EU) 2024/1689 the upstream authors remain the provider of this model. Converting the serialisation format does not make this repository the provider of a new general-purpose AI model, and no such claim is made. Questions about training content, copyright policy or model capability belong upstream.