fix: explicit licences in publish_hf.py (no CC-BY default), docs/licenses.md

README.md

@@ -559,4 +559,4 @@ Ettore Di Giacinto ([@mudler](https://github.com/mudler)).
 
 ## License
 
-parakeet.cpp is released under the [MIT License](LICENSE). The model weights are governed by the licenses of the original models, so check each model card on HuggingFace. The NVIDIA Parakeet models are mostly CC-BY-4.0 (nemotron-3.5-asr-streaming is OpenMDW-1.1). Moondream's [parakeet-ultra](https://huggingface.co/moondream/parakeet-ultra) and [parakeet-redux](https://huggingface.co/moondream/parakeet-redux) are CC-BY-4.0 too: credit Moondream and NVIDIA ([parakeet-tdt-0.6b-v3](https://huggingface.co/nvidia/parakeet-tdt-0.6b-v3)), link the [license](https://creativecommons.org/licenses/by/4.0/), and note that GGUF files made here are converted (and quantized, or dequantized for Redux) copies, not retrained models.
+parakeet.cpp is released under the [MIT License](LICENSE). The model weights are governed by the licenses of the original models, so check each model card on HuggingFace. The NVIDIA Parakeet models are mostly CC-BY-4.0 (nemotron-3.5-asr-streaming and Nemotron-3-Diarization are OpenMDW-1.1, parakeet_realtime_eou_120m-v1 is the NVIDIA Open Model License, and the Silero VAD files are MIT). See [docs/licenses.md](docs/licenses.md) for the full table. Moondream's [parakeet-ultra](https://huggingface.co/moondream/parakeet-ultra) and [parakeet-redux](https://huggingface.co/moondream/parakeet-redux) are CC-BY-4.0 too: credit Moondream and NVIDIA ([parakeet-tdt-0.6b-v3](https://huggingface.co/nvidia/parakeet-tdt-0.6b-v3)), link the [license](https://creativecommons.org/licenses/by/4.0/), and note that GGUF files made here are converted (and quantized, or dequantized for Redux) copies, not retrained models.

docs/licenses.md

@@ -0,0 +1,38 @@
+# Model licences
+
+The GGUF files published at `mudler/parakeet-cpp-gguf` are converted from other
+people's models, so each file family keeps the licence of its source model. The
+parakeet.cpp code is MIT-licensed.
+
+This table is the reference for `scripts/publish_hf.py`. The script has no default
+licence: a model id that is missing from its `LICENSES` table is an error. The
+check `tests/python/check_publish_hf_licences.py` (and
+`python3 scripts/publish_hf.py --check-licences`) fails when this table and the
+script differ, so change both together.
+
+| Source checkpoint | Licence | Licence text | Notice to keep |
+|---|---|---|---|
+| `moondream/parakeet-redux` | CC-BY-4.0 | <https://creativecommons.org/licenses/by/4.0/> |  |
+| `moondream/parakeet-ultra` | CC-BY-4.0 | <https://creativecommons.org/licenses/by/4.0/> |  |
+| `nvidia/Nemotron-3-Diarization` | OpenMDW 1.1 | <https://openmdw.ai/license/1-1/> |  |
+| `nvidia/nemotron-3.5-asr-streaming-0.6b` | OpenMDW 1.1 | <https://openmdw.ai/license/1-1/> |  |
+| `nvidia/parakeet-ctc-0.6b` | CC-BY-4.0 | <https://creativecommons.org/licenses/by/4.0/> |  |
+| `nvidia/parakeet-ctc-1.1b` | CC-BY-4.0 | <https://creativecommons.org/licenses/by/4.0/> |  |
+| `nvidia/parakeet-rnnt-0.6b` | CC-BY-4.0 | <https://creativecommons.org/licenses/by/4.0/> |  |
+| `nvidia/parakeet-rnnt-1.1b` | CC-BY-4.0 | <https://creativecommons.org/licenses/by/4.0/> |  |
+| `nvidia/parakeet-tdt-0.6b-v2` | CC-BY-4.0 | <https://creativecommons.org/licenses/by/4.0/> |  |
+| `nvidia/parakeet-tdt-0.6b-v3` | CC-BY-4.0 | <https://creativecommons.org/licenses/by/4.0/> |  |
+| `nvidia/parakeet-tdt-1.1b` | CC-BY-4.0 | <https://creativecommons.org/licenses/by/4.0/> |  |
+| `nvidia/parakeet-tdt_ctc-1.1b` | CC-BY-4.0 | <https://creativecommons.org/licenses/by/4.0/> |  |
+| `nvidia/parakeet-tdt_ctc-110m` | CC-BY-4.0 | <https://creativecommons.org/licenses/by/4.0/> |  |
+| `nvidia/parakeet_realtime_eou_120m-v1` | NVIDIA Open Model License | <https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/> | Licensed by NVIDIA Corporation under the NVIDIA Open Model License |
+| `snakers4/silero-vad` | MIT | <https://github.com/snakers4/silero-vad/blob/master/LICENSE> | Copyright (c) 2020-present Silero Team. Released under the MIT license |
+
+Notes:
+
+- The Moondream Ultra and Redux files are converted here, not trained. Credit
+  Moondream and NVIDIA (`parakeet-tdt-0.6b-v3`) when you use them. The Redux F16 and
+  Q8_0 files are dequantized from the ternary weights.
+- The Silero VAD files are converted from the official ONNX model, not trained.
+- Static card text for the files that `publish_hf.py` does not convert (diarization,
+  Moondream, Silero VAD, and the Nemotron 3.5 usage note) lives in `scripts/hf_card/`.

scripts/hf_card/diarization.md

@@ -0,0 +1,14 @@
+### nemotron-3-diarization
+
+Source: [nvidia/Nemotron-3-Diarization](https://huggingface.co/nvidia/Nemotron-3-Diarization) · Speaker diarization (Sortformer, up to 8 speakers, streaming speaker cache) · License: [OpenMDW 1.1](https://openmdw.ai/license/1-1/)
+
+| File | Variant | Size | Segments vs NeMo |
+|---|---|---:|---:|
+| `nemotron-3-diarization-f16.gguf` ← **recommended** | F16 | 200.7 MB | identical |
+| `nemotron-3-diarization-q8_0.gguf` | Q8_0 | 108.7 MB | identical on 2 of 3 clips, 99.5% of frames on the third |
+
+> Checked against NeMo on a 23.6 s and a 68.5 s two-speaker clip (offline and streaming: every segment identical, to the 10 ms frame) and a 12.3 min three-speaker clip (F16 100%, Q8_0 99.9% of speech frames). Q8_0 can flip a frame whose probability sits right at the 0.5 threshold, splitting a segment (99.5% of frames on a 31.5 s clip); use F16 when segment-exact output matters. Answers "who spoke when"; pair it with any ASR model above for speaker-attributed transcripts. See [diarization.md](https://github.com/mudler/parakeet.cpp/blob/master/docs/diarization.md).
+
+```bash
+build/examples/cli/diarize models/nemotron-3-diarization-f16.gguf meeting.wav
+```

scripts/hf_card/moondream.md

@@ -0,0 +1,25 @@
+### ultra and redux (Moondream)
+
+Source: [moondream/parakeet-ultra](https://huggingface.co/moondream/parakeet-ultra) and [moondream/parakeet-redux](https://huggingface.co/moondream/parakeet-redux) by [Moondream](https://huggingface.co/moondream), derived from NVIDIA's [nvidia/parakeet-tdt-0.6b-v3](https://huggingface.co/nvidia/parakeet-tdt-0.6b-v3) · TDT transducer (FastConformer), multilingual, with a voice-activity head (`transcribe --vad`) · heads: TDT · license: [CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/)
+
+| File | Variant | Size | Runs on |
+|---|---|---:|---|
+| `ultra-f16.gguf` | F16 | 1441.9 MB | any backend |
+| `ultra-q8_0.gguf` | Q8_0 | 941.5 MB | any backend |
+| `redux-packed.gguf` | packed ternary encoder | 213.3 MB | CPU only, offline only |
+| `redux-f16.gguf` | F16, dequantized | 1441.9 MB | any backend |
+| `redux-q8_0.gguf` | Q8_0, dequantized | 941.5 MB | any backend |
+
+> There is no NeMo reference for these models, so there is no WER column. Each file was checked to decode the `tests/fixtures/speech.wav` clip to the expected sentence. See [ternary.md](https://github.com/mudler/parakeet.cpp/blob/master/docs/ternary.md) for the measurements.
+
+- **Ultra** has ordinary F16 weights and runs on any backend, like the other v3 files.
+- **Redux packed** keeps the encoder linear layers as ternary weights (-1, 0 or +1 times a per-group scale) and runs a native CPU kernel. It does not run on GPU backends or in streaming mode, and parakeet.cpp refuses to load it there. For GPU use, take `redux-f16.gguf` or `redux-q8_0.gguf`.
+- **Redux F16 and Q8_0** are dequantized: the ternary weights were expanded to ordinary weights and then stored as F16 or Q8_0. They are not packed.
+- **Changes:** these files are converted here, not trained. Nothing was trained or fine-tuned by the parakeet.cpp project. The models were trained by NVIDIA (the base) and Moondream (Ultra and Redux).
+
+```bash
+huggingface-cli download mudler/parakeet-cpp-gguf redux-packed.gguf --local-dir models/
+build/examples/cli/parakeet-cli transcribe --model models/redux-packed.gguf --input audio.wav
+# Long audio: cut at pauses with the model's voice-activity head (offline only).
+build/examples/cli/parakeet-cli transcribe --model models/ultra-q8_0.gguf --input long.wav --vad
+```

scripts/hf_card/nemotron-3.5-note.md

@@ -0,0 +1,6 @@
+> This model runs both offline and as a cache-aware streaming model, and it is the only one here that takes a target language. Pass `--lang <locale>` (for example `en-US`, `de-DE`, `es-ES`, `ja-JP`), or leave it at the default `auto` to let the model detect the language. The WER column is measured against NeMo, offline, on the languages en, de and auto; streaming and offline output also match NeMo for en, de, es, ja-JP and auto. See [parity.md](https://github.com/mudler/parakeet.cpp/blob/master/docs/parity.md). The small prompt layers, the LSTM and the feature extractor stay F32 in every quantization.
+
+```bash
+huggingface-cli download mudler/parakeet-cpp-gguf nemotron-3.5-asr-streaming-0.6b-f16.gguf --local-dir models/
+build/examples/cli/parakeet-cli transcribe --model models/nemotron-3.5-asr-streaming-0.6b-f16.gguf --input audio.wav --lang de-DE
+```

scripts/hf_card/silero-vad.md

@@ -0,0 +1,13 @@
+### silero-vad (Silero Team)
+
+Source: [snakers4/silero-vad](https://github.com/snakers4/silero-vad) v6.2.3 by the [Silero Team](https://github.com/snakers4) · a small voice-activity detector, not a speech recognizer · 8 kHz and 16 kHz in one file · license: [MIT](https://github.com/snakers4/silero-vad/blob/master/LICENSE)
+
+| File | Variant | Size |
+|---|---|---:|
+| `silero-vad-f32.gguf` | F32 | 2.2 MB |
+| `silero-vad-f16.gguf` | F16 weights, widened to F32 at load | 1.3 MB |
+
+- The files are converted from the official ONNX model with `scripts/convert_silero_vad_to_gguf.py` in parakeet.cpp. The GGUF records the source version and the ONNX sha256. They were converted here, not trained.
+- They need a parakeet.cpp build that has the standalone VAD API. That code is not in a release yet, so check the parakeet.cpp repository before relying on it.
+- Use it as a stand-alone detector (`parakeet-cli vad --model silero-vad-f16.gguf --input audio.wav`) or to cut long audio before transcribing with any model, including those that have no VAD head (`parakeet-cli transcribe --model tdt-0.6b-v3-q8_0.gguf --input long.wav --vad --vad-model silero-vad-f16.gguf`).
+- Probabilities match onnxruntime to about 1e-6 (F32) and 1e-3 (F16) on the test clips.