File size: 2,946 Bytes
6c952ac
5188ea4
6c952ac
 
 
 
 
 
 
 
 
9709fd0
6c952ac
 
9709fd0
6c952ac
 
 
5188ea4
 
 
9709fd0
5188ea4
 
 
 
 
6c952ac
 
04b34dc
 
6c952ac
 
 
 
 
 
 
 
 
 
 
 
 
 
5188ea4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
---
license: cc-by-nc-4.0
base_model: numind/NuNER-multilingual-v0.1
pipeline_tag: token-classification
tags:
  - ner
  - multilingual
  - onnx
  - transformers.js
---

# Riidact NER — multilingual names & places

Token-classification model (PER / LOC / ORG) powering
[Riidact](https://riidact.com)'s on-device name detection. Runs
in the browser via transformers.js + ONNX Runtime; the quantized weights
(`onnx/model_quantized.onnx`, ~178 MB) are what ships.

## License

These fine-tuned weights are released under **CC-BY-NC-4.0**: free to use,
share, and adapt for **non-commercial** purposes, with attribution. For commercial use, contact Riidact.

The MIT-licensed base model (`numind/NuNER-multilingual-v0.1`) is unaffected
and remains available under its own terms — see the NOTICE below. The
non-commercial terms apply to this fine-tune, not to the underlying base.

## Coverage

Fine-tuned on 103 languages; **88 are certified** by a two-tier benchmark:
realistic-sentence recall with a zero-false-positive corpus (72 languages),
plus held-out WikiANN validation at PER F1 ≥ 0.80 (84 languages).

## Lineage

- **Backbone:** [numind/NuNER-multilingual-v0.1](https://huggingface.co/numind/NuNER-multilingual-v0.1) (MIT)
- **Training data:** WikiANN (Wikipedia-derived, PAN-X annotations), 501k
  sentences across 103 languages, plus a small synthetic supplement of
  conversational name shapes (honorific patterns) that Wikipedia-derived
  data lacks.

## Recommended inference floors

Span mean confidence, per entity: PER ≥ 0.55, LOC ≥ 0.9 (precision-first;
calibrated against the benchmark corpus above).

## NOTICE

This model is a derivative of `numind/NuNER-multilingual-v0.1`, licensed under
the MIT License:

```
MIT License

Copyright (c) NuMind

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
```

The MIT terms cover the base model. The additional non-commercial terms
(CC-BY-NC-4.0) apply to the fine-tuning contribution in these weights.