Downloads · 30 days
87
23% of all-time downloads
beshkenadze/lang-id-voxlingua107-ecapa-mlx
lang-id-voxlingua107-ecapa-mlx is a audio classification model from beshkenadze. Use it for the audio classification task on the model card, and read the license before you ship it in a product. It is set up for mlx. The card lists the license as apache-2.0.
ECAPA-TDNN language identification model converted to MLX safetensors format for inference on Apple Silicon (Metal GPU).
Downloads · 30 days
87
23% of all-time downloads
All-time downloads
378
Public
Repo size
85.2 MB
Likes
0
Public
Click a slice to open those files.
.safetensors85.2 MB · 100%
From the Hugging Face model README
ECAPA-TDNN language identification model converted to MLX safetensors format for inference on Apple Silicon (Metal GPU).
Identifies 107 languages from audio. Original model: speechbrain/lang-id-voxlingua107-ecapa.
| Framework | Latency | Russian | English |
|---|---|---|---|
Python MLX (with mx.compile()) | 15.7ms | 99.6% | 99.9% |
Swift MLX (with compile() + GPU mel) | 14.8ms | 99.5% | 99.9% |
| CoreML GPU | 17ms | 99.7% | 98.6% |
# Clone the benchmark repo
git clone https://github.com/beshkenadze/lid-bench
cd lid-bench/mlx
# Setup
uv venv && uv pip install mlx numpy soundfile safetensors
# Run
python ecapa_tdnn_lid.py path/to/audio.wav --benchmark
Full implementation: github.com/beshkenadze/lid-bench
.ckpt → Conv1d axis swap → safetensorsWeights were converted from the original SpeechBrain checkpoint:
[out, in, kernel] → [out, kernel, in] (MLX convention)Convert script: mlx/convert_ecapa_weights.py in the lid-bench repo.