publish redimnet2-b6-cn OpenASR packs

.gitattributes

@@ -0,0 +1 @@
+*.oasr filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,105 @@
+---
+license: mit
+base_model: PalabraAI/redimnet2
+pipeline_tag: feature-extraction
+library_name: openasr
+tags:
+  - speaker-diarization
+  - openasr
+  - oasr
+---
+
+<div align="center">
+
+# ReDimNet2-B6 Speaker Embedder (CN-enhanced) · OpenASR
+
+**ReDimNet2-B6 speaker embedder for OpenASR diarization — 192-d CN-enhanced embeddings, fully on-device**
+
+[![License](https://img.shields.io/badge/license-MIT-2563eb.svg)](https://github.com/PalabraAI/redimnet2/blob/main/LICENSE)
+[![Format](https://img.shields.io/badge/format-.oasr-7c3aed.svg)](https://github.com/QuintinShaw/openasr)
+[![Runtime](https://img.shields.io/badge/runtime-OpenASR-111827.svg)](https://openasr.org)
+[![Base model](https://img.shields.io/badge/base-redimnet2-f59e0b.svg)](https://huggingface.co/PalabraAI/redimnet2)
+
+Speaker-diarization support pack for the **[OpenASR](https://github.com/QuintinShaw/openasr)** runtime —
+pure-Rust inference, **no Python at inference time**.
+
+</div>
+
+---
+
+## ✨ Highlights
+
+- 🗣️ **Preferred OpenASR speaker embedder** — when installed, diarization prefers this pack over WeSpeaker for anonymous speaker labels on any ASR family
+- 🧬 **192-dim ReDimNet2-B6** — PalabraAI's dimension-reshaping speaker net (12.5M params) with a Chinese-enhanced vb2+vox2+cnc2 training mix
+- 🔒 **Diarization, not identification** — anonymous session-relative labels; embeddings stay local and are discarded after the request unless you explicitly enroll a local profile
+- 🎯 **Parity-gated packaging** — ggml-graph forward pass matches the upstream Python reference at cosine ≥ 0.9999 on held-out fixtures
+- 🦀 **Native in OpenASR** — `.oasr` packs run with no Python at inference, engineered for peak performance on CPU & GPU
+
+## 🚀 Quickstart
+
+```bash
+# 1. Install the OpenASR CLI  ·  https://openasr.org
+# 2. Pull the pack
+openasr pull redimnet2-b6-cn:fp16
+
+# 3. Diarize any transcription (works with every OpenASR ASR model)
+openasr transcribe meeting.wav --model xasr-zh-en --diarize --format srt
+```
+
+## 📦 Pack
+
+| Quant | File (`.oasr`) | Size |
+|:------|:---------------|-----:|
+| fp16 | `redimnet2-b6-cn-fp16.oasr` | 28 MB |
+| q8_0 | `redimnet2-b6-cn-q8_0.oasr` | 27 MB |
+| f32 | `redimnet2-b6-cn-f32.oasr` | 50 MB |
+
+<sub>Single raw-**f32** build: the pure-Rust forward pass consumes f32 directly and the
+parity gates assert bit-exact outputs vs the upstream weights, so no integer
+quantization is produced.</sub>
+
+## 🧠 About ReDimNet2-B6 Speaker Embedder (CN-enhanced)
+
+ReDimNet2-B6 is PalabraAI's 12.5M-parameter speaker-embedding model from the
+ReDimNet2 family, trained on a VoxBlink2 + VoxCeleb2 + CN-Celeb2 mix so English and
+Chinese speakers share one embedding space. OpenASR packages the MIT-licensed
+checkpoint as a local `.oasr` capability pack and runs it through a ggml graph
+(not a pure-Rust hand-written forward). When this pack is installed it becomes the
+preferred speaker-embedding stage for diarization; the legacy WeSpeaker pack remains
+available as a fallback until a later quality gate retires it. Embeddings are 192-d
+cosine vectors with a ReDimNet-specific calibration profile, separate from WeSpeaker's
+256-d space.
+
+## ⚙️ How this pack was made
+
+Converted from [PalabraAI/redimnet2](https://huggingface.co/PalabraAI/redimnet2) with the OpenASR importer:
+
+```bash
+openasr model-pack import redimnet2 <src>.safetensors <out>.oasr \
+  --package-id redimnet2-b6-cn
+```
+
+The `.oasr` container is GGUF-backed; every tensor is stored as raw f32 so the
+pack round-trips bit-identically against the source weights.
+
+## ⚖️ License
+
+This pack **inherits the upstream model's license: MIT**
+([source](https://github.com/PalabraAI/redimnet2/blob/main/LICENSE)). OpenASR packaging retains the upstream copyright;
+the only modification is format conversion.
+
+## 🙏 Acknowledgements
+
+This pack redistributes **PalabraAI/redimnet2** checkpoint
+`b6-vb2+vox2+cnc2_v0-lm.pt` in OpenASR's `.oasr` runtime format. Credit for the
+model architecture, training, and original weights belongs to the upstream
+PalabraAI / ReDimNet2 authors (paper: ReDimNet2: Scaling Speaker Verification via
+Time-Pooled Dimension Reshaping). The upstream model is licensed under **MIT**;
+OpenASR packaging retains that license and attribution, with the only modification
+being format conversion for local ggml-graph loading.
+
+## 🔗 Links
+
+- 🦀 **OpenASR** — <https://github.com/QuintinShaw/openasr>
+- 🌐 **Website** — <https://openasr.org>
+- 🤗 **Upstream model** — [PalabraAI/redimnet2](https://huggingface.co/PalabraAI/redimnet2)

redimnet2-b6-cn-f32.oasr

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34f38adb26892a04b9bd39a3abc3de0606c3e406c133c25e20579d8547917678
+size 50048672

redimnet2-b6-cn-fp16.oasr

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b889eb50978957fcf5dae9dbcdb9aca7d3834bc72c3680c01e666b2d7e218727
+size 27775264

redimnet2-b6-cn-q8_0.oasr

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d489dede665dfbaaa7a35d38055dc3eb4df7d6bb8fc8cad625d8332d845c1db8
+size 26738464