Qwen3 ForcedAligner 0.6B 4bit (MLX) 4 bit quantized version of Qwen/Qwen3 ForcedAligner 0.6B for Apple Silicon inference via MLX. Predicts word level timestamps for audio+text pairs in a single non autoregressive forward pass. Model Details Component Config Audio encoder 24 layers, d model=1024, 16 heads, FFN=4096, float16 Text decoder 28 layers, hidden=1024, 16Q/8KV heads, 4 bit quantized (group size=64) Classify head Linear(1024, 5000), float16 Timestamp resolution 80ms per class (5000 classes = 400s max) Total size 979 MB (vs 1.84 GB bf16) How It Works Unlike ASR (autoregressive, token by token), the forced aligner runs the entire sequence in one forward pass through the decoder. The classify head predicts a timestamp class (0–4999) at each token position, which maps to time via class index × 80ms . Usage with Swift (MLX) This model is designed for use with speech swift: CLI Output: Quantization Text decoder (attention projections, MLP, embeddings) quantized to 4 bit using group quantization (group size=64). Audio encoder and classify head kept as float16 for accuracy. Converted with: Links Swift library : soniqo/speech swift — Swift Package for Qwen3 ASR, Qwen3 TTS, CosyVoice,…
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy