Downloads · 30 days
0
talatapp/wespeaker-voxceleb-resnet34-LM-onnx
wespeaker-voxceleb-resnet34-LM-onnx is a machine learning model from talatapp. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.
Re-export of Wespeaker/wespeaker-voxceleb-resnet34-LM to a single ONNX file whose I/O contract matches FluidInference's CoreML wespeakerv2.mlmodelc:
Downloads · 30 days
0
Access
Public
Updated May 16, 2026
Repo size
60.1 MB
Likes
2
Public
Click a slice to open those files.
.onnx26.6 MB · 100%
From the Hugging Face model README
Re-export of Wespeaker/wespeaker-voxceleb-resnet34-LM to a single ONNX file whose I/O contract matches FluidInference's CoreML wespeaker_v2.mlmodelc:
| role | name | dtype | shape |
|---|---|---|---|
| input | waveform | f32 | [3, 160_000] |
| input | mask | f32 | [3, 589] |
| output | embedding | f32 | [3, 256] |
The wrapper bakes the full preprocessing chain into the ONNX so the consumer hands it raw 16 kHz mono PCM and gets back a 256-dim embedding per speaker slot:
snip_edges=False.Slots whose mask.sum() == 0 get an all-zero embedding so the downstream matcher can detect "no signal" without seeing whatever the encoder produces from all-zero input.
| File | Size | Use |
|---|---|---|
voxceleb_resnet34_LM.onnx | ~26 MB | full-precision fp32 |
voxceleb_resnet34_LM.int8.onnx | ~7 MB | per-channel int8 dynamic quant — Windows manifest target |
The Mac side of talat uses FluidInference's CoreML conversion of WeSpeaker, which bakes the preprocessing chain into the graph. The upstream Wespeaker/wespeaker-voxceleb-resnet34-LM ONNX export does NOT — it expects pre-computed fbank features instead. So Windows hitting the same upstream ONNX blew up at engine startup with "missing 'waveform' input".
This re-export gives both platforms the same I/O surface so the Rust talat-inference::EmbeddingExtractor works unchanged across Mac (CoreML) and Windows (ONNX). The asymmetric quant story matches the rest of the in-process model bundle: fp16 CoreML on Apple silicon (ANE-native), int8 ONNX on Windows (DirectML / CPU EP).
The export script is at scripts/wespeaker-onnx-export/ in the talat repo — convert_wespeaker.py is heavily commented if you need to adapt it.
The cosine similarity against FluidInference's CoreML version is close-but-not-bit-exact because:
The speaker matcher compares embeddings by cosine similarity in a 256-dim space, which is forgiving of small drift. If cluster decisions diverge between Mac and Windows on the same audio, the masking strategy in the wrapper is the first thing to revisit.
Inherits the WeSpeaker (Apache 2.0) and VoxCeleb (CC-BY-4.0 / academic) licenses of the upstream checkpoint.