Zum Inhalt springen
ModelAufgelistetExtern
Herunterladen

Miocodec v2 44k GGUF

GGUF conversions of Aratako/MioCodec-25Hz-44.1kHz-v2 (MIT license, 133M params) for use with CrispASR.

Quellehuggingface.co/cstr/miocodec-v2-44k-GGUF/tree/587ccd1bb958c66623b19ad36304a4eaa527b397
ArtOtherBasismodellAratako/MioCodec-25Hz-44.1kHz-v2VersionRevision 587ccd1BBewertung

Modellquelle

Art
Other
Dateigröße
98,6–258 MB
Basismodell
Aratako/MioCodec-25Hz-44.1kHz-v2
Version
vRevision 587ccd1
Aktualisiert
12. Okt. 2026
Downloads
475
Quelle
Hugging Face

Quellenbeschreibung

MioCodec-25Hz-44.1kHz-v2 GGUF

GGUF conversions of Aratako/MioCodec-25Hz-44.1kHz-v2 (MIT license, 133M params) for use with CrispASR.

MioCodec is a lightweight neural audio codec: encode 44.1 kHz audio to 25 Hz discrete tokens (12800 vocab, 341 bps), decode tokens + speaker embedding back to waveform.

Files

FileSizeTypeNotes
miocodec-v2-44k-f16.gguf259 MBF16Full precision decode weights
miocodec-v2-44k-q8_0.gguf155 MBQ8_0~2x compression, cos=0.992 vs F32
miocodec-v2-44k-q4_k.gguf99 MBQ4_K~4x compression, cos=0.914 vs F32

All GGUFs contain the decoder-only weights (FSQ quantizer + wave_prenet + wave_decoder + wave_post_net + SnakeBeta upsampler + ISTFTHead). The WavLM encoder is separate.

Architecture

tokens(T) -> FSQ(T, 768) -> wave_prenet 6L Transformer(T, 512)
  -> ConvTranspose1d 2x -> interpolate -> ResNet x2
  -> wave_decoder 8L Transformer+AdaLN-Zero(T', 512) -> ResNet x2
  -> SnakeBeta Upsampler 9x -> ISTFTHead(394) -> 44.1kHz waveform
  • FSQ: Finite Scalar Quantization, levels=[8,8,8,5,5], 12800 codebook
  • Transformer: RoPE (normal/adjacent-pair), SwiGLU FFN, windowed attention
  • AdaLN-Zero: Speaker conditioning via 128-dim global embedding
  • SnakeBeta: x + exp(-beta) * sin^2(exp(alpha) * x) periodic activation
  • ISTFT: exp(log_mag) + cos/sin(phase) -> iFFT overlap-add, "same" padding

Parity (F16 vs Python F32 reference)

Stagecos
fsq_decoded1.000000
wave_prenet_out (6L Transformer)1.000000
wave_decoder_out (8L AdaLN Transformer)1.000000
wave_upsampler_out1.000000
istft_mag_phase1.000000
output_waveform (175518 samples)0.999998

Usage with CrispASR

# Diff harness (verify per-stage parity)
crispasr-diff miocodec miocodec-v2-44k-f16.gguf ref.gguf audio.wav

Conversion

python models/convert-miocodec-to-gguf.py \
  --model Aratako/MioCodec-25Hz-44.1kHz-v2 \
  --output miocodec-v2-44k-f16.gguf --dtype f16

# Quantize
crispasr-quantize miocodec-v2-44k-f16.gguf miocodec-v2-44k-q8_0.gguf q8_0
crispasr-quantize miocodec-v2-44k-f16.gguf miocodec-v2-44k-q4_k.gguf q4_k

License

MIT (same as the original model).

Provenance and EU AI Act Art. 53 note

  • Upstream model: Aratako/MioCodec-25Hz-44.1kHz-v2 — published by Aratako.
  • Upstream licence: mit. 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). 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.
  • 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.

Quellen

1 Quelle
Hugging FacePrimärVersion 587ccd1bb958c66623b19ad36304a4eaa527b397
huggingface.co/cstr/miocodec-v2-44k-GGUF/tree/587ccd1bb958c66623b19ad36304a4eaa527b397

Verifiziert 12. Okt.

Modellartefakte

3 Artefakte

miocodec-v2-44k-f16.gguf

gguf · 258 MB · SHA-256 19441c391418…9c4f · Hugging Face

Herunterladen

miocodec-v2-44k-q4_k.gguf

gguf · 98,6 MB · SHA-256 44cb477f8c94…2041 · Hugging Face

Herunterladen

miocodec-v2-44k-q8_0.gguf

gguf · 154 MB · SHA-256 b593993556d3…817c · Hugging Face

Herunterladen

Quellenauszüge

3 Auszüge
Readme2. Aug.huggingface.co/cstr/miocodec-v2-44k-GGUF

--- license: mit tags: - audio - codec - gguf - crispasr base_model: Aratako/MioCodec-25Hz-44.1kHz-v2 pipeline_tag: audio-to-audio --- # MioCodec-25Hz-44.1kHz-v2 GGUF GGUF conversions of [Aratako/MioCodec-25Hz-44.1kHz-v2](https://huggingface.co/Aratako/MioCodec-25Hz-44.1kHz-v2) (MIT license, 133M params) for use with [CrispASR](https://github.com/CrispStrobe/CrispASR). MioCodec is a lightweight neural audio codec: encode 44.1 kHz audio to 25 Hz discrete tokens (12800 vocab, 341 bps), decode tokens + speaker embedding back to waveform. ## Files | File | Size | Type | Notes | |------|------|------|-------| | `miocodec-v2-44k-f16.gguf` | 259 MB | F16 | Full precision decode weights | | `miocodec-v2-44k-q8_0.gguf` | 155 MB | Q8_0 | ~2x compression, cos=0.992 vs F32 | | `miocodec-v2-44k-q4_k.gguf` | 99 MB | Q4_K | ~4x compression, cos=0.914 vs F32 | All GGUFs contain the **decoder-only** weights (FSQ quantizer + wave_prenet + wave_decoder + wave_post_net + SnakeBeta upsampler + ISTFTHead). The WavLM encoder is separate. ## Architecture ``` tokens(T) -> FSQ(T, 768) -> wave_prenet 6L Transformer(T, 512) -> ConvTranspose1d 2x -> interpolate -> ResNet x2 -> wave_decoder 8L Transformer+AdaLN-Zero(T', 512) -> ResNet x2 -> SnakeBeta Upsampler 9x -> ISTFTHead(394) -> 44.1kHz waveform ``` - **FSQ**: Finite Scalar Quantization, levels=[8,8,8,5,5], 12800 codebook - **Transformer**: RoPE (normal/adjacent-pair), SwiGLU FFN, windowed attention - **AdaLN-Zero**: Speaker conditioning via 128-dim global embedding - **SnakeBeta**: `x + exp(-beta) * sin^2(exp(alpha) * x)` periodic activation - **ISTFT**: exp(log_mag) + cos/sin(phase) -> iFFT overlap-add, "same" padding ## Parity (F16 vs Python F32 reference) | Stage | cos | |-------|-----| | fsq_decoded | 1.000000 | | wave_prenet_out (6L Transformer) | 1.000000 | | wave_decoder_out (8L AdaLN Transformer) | 1.000000 | | wave_upsampler_out | 1.000000 | | istft_mag_phase | 1.000000 | | output_waveform (175518 samples) | 0.999998 | ## Usage with CrispASR ```bash # Diff harness (verify per-stage parity) crispasr-diff miocodec miocodec-v2-44k-f16.gguf ref.gguf audio.wav ``` ## Conversion ```bash python models/convert-miocodec-to-gguf.py \ --model Aratako/MioCodec-25Hz-44.1kHz-v2 \ --output miocodec-v2-44k-f16.gguf --dtype f16 # Quantize crispasr-quantize miocodec-v2-44k-f16.gguf miocodec-v2-44k-q8_0.gguf q8_0...

Note12. Okt.huggingface.co/cstr/miocodec-v2-44k-GGUF

Source context: 475 downloads · 1 likes · Pipeline audio-to-audio · Repo cstr/miocodec-v2-44k-GGUF

Note2. Aug.huggingface.co/cstr/miocodec-v2-44k-GGUF

Source context: 350 downloads · 1 likes · Pipeline audio-to-audio · Repo cstr/miocodec-v2-44k-GGUF