# Qwen3-ForcedAligner 0.6B ยท OpenASR
**Word-level forced alignment for OpenASR transcripts -- a non-autoregressive Qwen3 audio+text model that refines per-word timestamps**
[](https://huggingface.co/Qwen/Qwen3-ForcedAligner-0.6B/blob/c7cbfc2048c462b0d63a45797104fc9db3ad62b7/LICENSE)
[](https://github.com/QuintinShaw/openasr)
[](https://openasr.org)
[](https://huggingface.co/Qwen/Qwen3-ForcedAligner-0.6B)
A capability-pack support model for the **[OpenASR](https://github.com/QuintinShaw/openasr)**
runtime โ pure-Rust inference, **no Python at inference time**. Not a standalone
transcription model: it augments another OpenASR ASR model's own decode path.
---
## โจ Highlights
- ๐ฏ **Refined word timestamps** โ consumes a finished transcript's text plus the source audio and replaces a model family's own approximate per-word timestamps with aligner-refined spans (`--word-timestamps=aligned`)
- โก **Non-autoregressive** โ a single forward pass over interleaved audio/text with argmax at `` positions (5000 80ms-wide bins), not incremental greedy decoding, so it is not dispatched through the qwen3-asr runtime
- ๐งฉ **Shares its backbone with Qwen3-ASR** โ the same audio-encoder + LM `thinker` tensor layout, byte-for-byte; only the final head differs (an independent 5000-way classification head instead of the tied vocabulary head)
- ๐ **Shared attribution dependency** โ used explicitly by `--word-timestamps=aligned` and internally when external diarization must split a coarse ASR segment at speaker changes
- ๐ฆ **Validated native Q4_K default** โ the public q4_k name follows OpenASR's unified tier naming, while boundary-sensitive audio, token-embedding, and timestamp-head matrices stay Q8_0; Q8_0 and FP16 remain available
- ๐ฆ **Native in OpenASR** โ `.oasr` packs run with no Python at inference, engineered for peak performance on CPU & GPU
## ๐ Quickstart
```bash
# 1. Install the OpenASR CLI ยท https://openasr.org
# 2. Pull the pack
openasr pull qwen3-forced-aligner-0.6b:q4
# 3. Use it as an opt-in refinement for another model's transcribe call
openasr transcribe meeting.wav --model --word-timestamps=aligned
```
## ๐ฆ Pack
| Quant | File (`.oasr`) | Size |
|:------|:---------------|-----:|
| fp16 | `qwen3-forced-aligner-0.6b-fp16.oasr` | 1.84 GB |
| q8_0 | `qwen3-forced-aligner-0.6b-q8_0.oasr` | 986 MB |
| q4_k | `qwen3-forced-aligner-0.6b-q4_k.oasr` | 765 MB |
## ๐ง About Qwen3-ForcedAligner 0.6B
Qwen3-ForcedAligner-0.6B is a **word-level forced-alignment** model from **Qwen**, sharing its
audio-encoder + LM `thinker` tensor layout byte-for-byte with **Qwen3-ASR** (same
`Qwen3ASRForConditionalGeneration` architecture). The only structural difference is the final
head: instead of a tied vocabulary `lm_head`, it uses an independent `Linear(hidden_size,
5000)` classification head over 80ms-wide timestamp bins. Given a transcript's text and its
source audio, it runs a single non-autoregressive forward pass and reads off word-boundary
timestamps at argmax `` positions -- refining a model family's own (typically
decode-time-approximate) per-word timestamps. This OpenASR repo repackages the weights as
`.oasr` packs that run natively in the OpenASR runtime -- no Python at inference, all decoding
local. OpenASR recommends the validated **q4_k** tier and also ships **q8_0** and an **fp16**
full-precision reference tier. The q4_k label is the catalog's unified product name, not a claim
that every matrix uses Q4_K: the audio encoder, token embedding, and timestamp head remain Q8_0,
and pack verification replays the exact per-tensor policy. Q3 and legacy all-Q4 packs are rejected
because small logit perturbations can move a word boundary across multiple 80ms bins.
**Not a standalone transcription model.** This pack cannot transcribe audio by itself; it is an
alignment dependency consumed explicitly via `openasr transcribe