Downloads · 30 days
22
19% of all-time downloads
fdt-disfluency-tiny-4m
stillerman/fdt-disfluency-tiny-4m
fdt-disfluency-tiny-4m is a token classification model from stillerman. Use it when you need labels on individual words, such as names. It is set up for transformers.js. The card lists the license as apache-2.0.
Disfluency deletion tagger for live speech transcripts: tags every whitespace word KEEP / DELETE / KEEPSTRIPCOMMA / KEEPCAPITALIZE, then a ~15-line reconstruction turns tags into cleaned text. Deletion-only by constru…
All-time downloads
113
Public
Parameters
4.4M
21.9 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors17.5 MB · 76%
At a glance
- Task
- Token Classification
- Library
- transformers.js
- License
- apache-2.0
- Model type
- bert
- Access
- Public
- Created
- Jul 7, 2026
- Updated
- Jul 7, 2026
- SHA
- ba012f0a
Try a prompt
Model card
From the Hugging Face model README
stillerman/fdt-disfluency-tiny-4m
Disfluency deletion tagger for live speech transcripts: tags every
whitespace word KEEP / DELETE / KEEP_STRIP_COMMA / KEEP_CAPITALIZE,
then a ~15-line reconstruction turns tags into cleaned text. Deletion-only by
construction — it cannot rephrase, hallucinate, or alter names and numbers.
- Architecture: BERT L2/H128 (4.4M params), v1 data mix
- Val metrics: exact-match 0.5980, DELETE-F1 0.8892
- Training data: synthetic disfluency injection over conversational corpora — see stillerman/fdt-disfluency-synthetic
onnx/model_quantized.onnx(int8) is ready for transformers.js (device: "webgpu",dtype: "q8"); runs at ~10–50 ms per utterance in-browser.- ⚠️ Trained partly on DailyDialog (CC BY-NC-SA): treat as research artifact, not for commercial deployment as-is.
Trained on a DGX Spark as part of the FluencyAI digital-twin project.
- Task
- Token Classification
- Library
- transformers.js
- Type
- bert
- License
- apache-2.0
- Created
- Jul 7, 2026
- Updated
- Jul 7, 2026