Downloads · 30 days
10
17% of all-time downloads
ettin-17m-sponsor-v1-android
CuriousDragon/ettin-17m-sponsor-v1-android
ettin-17m-sponsor-v1-android is a token classification model from CuriousDragon. Use it when you need labels on individual words, such as names. It is set up for onnxruntime. The card lists the license as cc-by-nc-sa-4.0.
ONNX Runtime package used by Flow for on-device sponsor-span detection. Quantized from CuriousDragon/ettin-17m-sponsor-v1 (embeddingint8perchannel, ORT format, opset 18).
All-time downloads
60
Public
Repo size
127 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.onnx98.3 MB · 75%
At a glance
- Task
- Token Classification
- Library
- onnxruntime
- License
- cc-by-nc-sa-4.0
- Model type
- modernbert
- Access
- Public
- Created
- Sep 3, 2026
- Updated
- Sep 21, 2026
- SHA
- 79dbe15e
Try a prompt
Base models
Model card
From the Hugging Face model README
Ettin 17M sponsor detector (Android INT8)
ONNX Runtime package used by Flow for on-device sponsor-span detection.
Quantized from CuriousDragon/ettin-17m-sponsor-v1
(embedding_int8_per_channel, ORT format, opset 18).
- Primary runtime file:
sponsor_detector_v1.int8.ort(~28 MB) - Also included: FP32 ONNX and INT8 ONNX for desktop/debug
- Tokenizer: same Ettin / ModernBERT tokenizer as the PyTorch checkpoints
- Confidence threshold in the Android package:
0.0
Files
| File | Role |
|---|---|
sponsor_detector_v1.int8.ort | Shipping Android model |
sponsor_detector_v1.int8.onnx | INT8 ONNX before ORT packing |
sponsor_detector_v1.fp32.onnx | Unquantized ONNX |
tokenizer.json / tokenizer_config.json | Tokenizer |
required_operators_and_types.config | Reduced ORT operator config |
tokenizer_goldens.json | Tokenization goldens |
sponsor_feedback_v1.schema.json | Local training-journal schema |
android_manifest.json | Export manifest and hashes |
Limitations
- English transcripts only.
- Target is external paid sponsorships. Self-promotion, memberships, merchandise, and interaction reminders are treated as negatives.
- SponsorBlock labels are noisy and incomplete. Evaluations against them are weak labels, not ground truth.
- Full-video reconstruction interpolates timestamps inside caption cues.
- Window-level scores do not by themselves skip video; Flow stitches overlapping 768-token windows with a 128-token overlap.
License
These weights are released under CC BY-NC-SA 4.0.
The base encoder jhu-clsp/ettin-encoder-17m
is MIT. Fine-tuning used SponsorBlock labels
(CC BY-NC-SA 4.0) and the pinned Xenova/sponsorblock-768
window corpus. Non-commercial use and share-alike apply to this derivative.
This project is not affiliated with SponsorBlock.
- Task
- Token Classification
- Library
- onnxruntime
- Type
- modernbert
- License
- cc-by-nc-sa-4.0
- Languages
- en
- Created
- Sep 3, 2026
- Updated
- Sep 21, 2026