Add Silero VAD, WeSpeaker, Sortformer 4spk; per-file license README

README.md

@@ -1,27 +1,44 @@
 ---
 license: other
+license_name: mixed-per-file
+license_link: https://huggingface.co/TalkBar/voice-models#per-file-licenses
 tags:
-  - talkbar
   - speech-recognition
-  - whisper
-  - parakeet
-  - canary
-  - gigaam
+  - asr
   - onnx
   - ggml
-pretty_name: TalkBar Voice Models
 ---
 
-# TalkBar voice models (app download catalog)
-
-Public redistributions of ASR weights used by the **TalkBar** desktop app.
-Files are byte-identical to the previous Handy CDN host and are pinned by
-SHA-256 in the TalkBar client.
-
-TalkBar does **not** claim authorship of these model weights. This repository
-exists so the app can download from a TalkBar-controlled origin.
-
-## Files & integrity
+# TalkBar voice models — ASR download host
+
+First-party download host for the TalkBar desktop app's speech-recognition
+weights. Every file is a byte-exact repackage of a permissively licensed
+upstream model (packaging: tar.gz bundles matching the app's install
+layout, or the upstream file as-is). The app downloads at a pinned commit
+and verifies SHA-256.
+
+## Per-file licenses
+
+| File | Upstream model | Upstream license |
+| --- | --- | --- |
+| `parakeet-v3-int8.tar.gz` | [nvidia/parakeet-tdt-0.6b-v3](https://huggingface.co/nvidia/parakeet-tdt-0.6b-v3) (int8 ONNX export bundle) | [CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/) |
+| `ggml-small.bin` | [OpenAI Whisper](https://github.com/openai/whisper) small, GGML conversion | [MIT](https://github.com/openai/whisper/blob/main/LICENSE) |
+| `whisper-medium-q4_1.bin` | OpenAI Whisper medium, GGML q4_1 quantization | MIT |
+| `ggml-large-v3-turbo.bin` | OpenAI Whisper large-v3-turbo, GGML conversion | MIT |
+| `ggml-large-v3-q5_0.bin` | OpenAI Whisper large-v3, GGML q5_0 quantization | MIT |
+| `giga-am-v3.int8.onnx` | [salute-developers/GigaAM](https://github.com/salute-developers/GigaAM) v3, int8 ONNX export | [MIT](https://github.com/salute-developers/GigaAM/blob/main/LICENSE) |
+| `canary-180m-flash.tar.gz` | [nvidia/canary-180m-flash](https://huggingface.co/nvidia/canary-180m-flash) (ONNX bundle) | [CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/) |
+| `canary-1b-v2.tar.gz` | [nvidia/canary-1b-v2](https://huggingface.co/nvidia/canary-1b-v2) (ONNX bundle) | [CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/) |
+| `silero_vad.onnx` | [snakers4/silero-vad](https://github.com/snakers4/silero-vad) | [MIT](https://github.com/snakers4/silero-vad/blob/master/LICENSE) |
+| `voxceleb_resnet34_LM.onnx` | [WeSpeaker](https://github.com/wenet-e2e/wespeaker) voxceleb ResNet34-LM | [Apache-2.0](https://github.com/wenet-e2e/wespeaker/blob/master/LICENSE) |
+| `diar_streaming_sortformer_4spk-v2.1.onnx` | [nvidia/diar_streaming_sortformer_4spk-v2.1](https://huggingface.co/nvidia/diar_streaming_sortformer_4spk-v2.1) (ONNX export, no weight changes) | [NVIDIA Open Model License](https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-open-model-license/) |
+
+Attribution: NVIDIA (Parakeet, Canary), OpenAI (Whisper), SberDevices /
+salute-developers (GigaAM). No weight modifications beyond quantization /
+format conversion / archive packaging; all redistribution under the
+upstream licenses above with attribution.
+
+## Integrity
 
 | File | Bytes | SHA-256 |
 | --- | ---: | --- |
@@ -33,15 +50,6 @@ exists so the app can download from a TalkBar-controlled origin.
 | `giga-am-v3.int8.onnx` | 224,893,347 | `2e3fcb7a7b66030336fd10c2fcfb033bd1dc7e1bf238fe5cfd83b1d0cfc9d28e` |
 | `canary-180m-flash.tar.gz` | 153,194,652 | `6d9cfca6118b296e196eaedc1c8fa9788305a7b0f1feafdb6dc91932ab6e53f7` |
 | `canary-1b-v2.tar.gz` | 725,374,161 | `02305b2a25f9cf3e7deaffa7f94df00efa44f442cd55c101c2cb9c000f904666` |
-
-## Provenance (high level)
-
-- **Whisper GGML** bins: OpenAI Whisper lineage (community quantized / packaged builds).
-- **Parakeet / Canary** archives: NVIDIA NeMo family ASR packages.
-- **GigaAM** ONNX: GigaAM v3 Russian ASR (int8). Vocab for the app bundle is embedded in the TalkBar client, not this file.
-
-See upstream project licenses and model cards for redistribution terms.
-
-## Related
-
-- Large-meeting diarizer (8spk): [`TalkBar/ultra-sortformer-8spk`](https://huggingface.co/TalkBar/ultra-sortformer-8spk)
+| `silero_vad.onnx` | 1,807,522 | `a35ebf52fd3ce5f1469b2a36158dba761bc47b973ea3382b3186ca15b1f5af28` |
+| `voxceleb_resnet34_LM.onnx` | 26,530,309 | `7bb2f06e9df17cdf1ef14ee8a15ab08ed28e8d0ef5054ee135741560df2ec068` |
+| `diar_streaming_sortformer_4spk-v2.1.onnx` | 492,243,003 | `f010a62654fd6d2adb70166265df2e53f5a766e354d2cce152b4c7966a04d02d` |

diar_streaming_sortformer_4spk-v2.1.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f010a62654fd6d2adb70166265df2e53f5a766e354d2cce152b4c7966a04d02d
+size 492243003

silero_vad.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a35ebf52fd3ce5f1469b2a36158dba761bc47b973ea3382b3186ca15b1f5af28
+size 1807522

voxceleb_resnet34_LM.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7bb2f06e9df17cdf1ef14ee8a15ab08ed28e8d0ef5054ee135741560df2ec068
+size 26530309