File size: 8,687 Bytes
c298231
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
---
library_name: span-marker
tags:
- span-marker
- token-classification
- ner
- named-entity-recognition
- generated_from_span_marker_trainer
metrics:
- precision
- recall
- f1
widget:
- text: The Bengal tiger is the most common subspecies of tiger, constituting approximately
    80% of the entire tiger population, and is found in Bangladesh, Bhutan, Myanmar,
    Nepal, and India.
- text: In other countries, it is a non-commissioned rank (e.g. Spain, Italy, France,
    the Netherlands and the Indonesian Police ranks).
- text: The filling consists of fish, pork and bacon, and is seasoned with salt (unless
    the pork is already salted).
- text: This stood until August 20, 1993 when it was beaten by one 1 / 100th of a
    second by Colin Jackson of Great Britain in Stuttgart, Germany, a subsequent record
    that stood for 13 years.
- text: Ann Patchett ’s novel " Bel Canto ", was another creative influence that helped
    her manage a plentiful cast of characters.
pipeline_tag: token-classification
model-index:
- name: SpanMarker
  results:
  - task:
      type: token-classification
      name: Named Entity Recognition
    dataset:
      name: Unknown
      type: unknown
      split: eval
    metrics:
    - type: f1
      value: 0.9130661114003124
      name: F1
    - type: precision
      value: 0.9148758606300855
      name: Precision
    - type: recall
      value: 0.9112635078969243
      name: Recall
---

# SpanMarker

This is a [SpanMarker](https://github.com/tomaarsen/SpanMarkerNER) model that can be used for Named Entity Recognition.

## Model Details

### Model Description
- **Model Type:** SpanMarker
<!-- - **Encoder:** [Unknown](https://huggingface.co/unknown) -->
- **Maximum Sequence Length:** 256 tokens
- **Maximum Entity Length:** 6 words
<!-- - **Training Dataset:** [Unknown](https://huggingface.co/datasets/unknown) -->
<!-- - **Language:** Unknown -->
<!-- - **License:** Unknown -->

### Model Sources

- **Repository:** [SpanMarker on GitHub](https://github.com/tomaarsen/SpanMarkerNER)
- **Thesis:** [SpanMarker For Named Entity Recognition](https://raw.githubusercontent.com/tomaarsen/SpanMarkerNER/main/thesis.pdf)

### Model Labels
| Label | Examples                                                                 |
|:------|:-------------------------------------------------------------------------|
| ANIM  | "vertebrate", "moth", "G. firmus"                                        |
| BIO   | "Aspergillus", "Cladophora", "Zythiostroma"                              |
| CEL   | "pulsar", "celestial bodies", "neutron star"                             |
| DIS   | "social anxiety disorder", "insulin resistance", "Asperger syndrome"     |
| EVE   | "Spanish Civil War", "National Junior Angus Show", "French Revolution"   |
| FOOD  | "Neera", "Bellini ( cocktail )", "soju"                                  |
| INST  | "Apple II", "Encyclopaedia of Chess Openings", "Android"                 |
| LOC   | "Kīlauea", "Hungary", "Vienna"                                           |
| MEDIA | "CSI : Crime Scene Investigation", "Big Comic Spirits", "American Idol"  |
| MYTH  | "Priam", "Oźwiena", "Odysseus"                                           |
| ORG   | "San Francisco Giants", "Arm Holdings", "RTÉ One"                        |
| PER   | "Amelia Bence", "Tito Lusiardo", "James Cameron"                         |
| PLANT | "vernal squill", "Sarracenia purpurea", "Drosera rotundifolia"           |
| TIME  | "prehistory", "Age of Enlightenment", "annual paid holiday"              |
| VEHI  | "Short 360", "Ferrari 355 Challenge", "Solution F / Chretien Helicopter" |

## Uses

### Direct Use for Inference

```python
from span_marker import SpanMarkerModel

# Download from the 🤗 Hub
model = SpanMarkerModel.from_pretrained("span_marker_model_id")
# Run inference
entities = model.predict("Ann Patchett ’s novel \" Bel Canto \", was another creative influence that helped her manage a plentiful cast of characters.")
```

### Downstream Use
You can finetune this model on your own dataset.

<details><summary>Click to expand</summary>

```python
from span_marker import SpanMarkerModel, Trainer

# Download from the 🤗 Hub
model = SpanMarkerModel.from_pretrained("span_marker_model_id")

# Specify a Dataset with "tokens" and "ner_tag" columns
dataset = load_dataset("conll2003") # For example CoNLL2003

# Initialize a Trainer using the pretrained model & dataset
trainer = Trainer(
    model=model,
    train_dataset=dataset["train"],
    eval_dataset=dataset["validation"],
)
trainer.train()
trainer.save_model("span_marker_model_id-finetuned")
```
</details>

<!--
### Out-of-Scope Use

*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->

<!--
## Bias, Risks and Limitations

*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->

<!--
### Recommendations

*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->

## Training Details

### Training Set Metrics
| Training set          | Min | Median  | Max |
|:----------------------|:----|:--------|:----|
| Sentence length       | 2   | 21.6493 | 237 |
| Entities per sentence | 0   | 1.5369  | 36  |

### Training Hyperparameters
- learning_rate: 1e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 1
- mixed_precision_training: Native AMP

### Training Results
| Epoch  | Step  | Validation Loss | Validation Precision | Validation Recall | Validation F1 | Validation Accuracy |
|:------:|:-----:|:---------------:|:--------------------:|:-----------------:|:-------------:|:-------------------:|
| 0.0576 | 1000  | 0.0142          | 0.8714               | 0.7729            | 0.8192        | 0.9698              |
| 0.1153 | 2000  | 0.0107          | 0.8316               | 0.8815            | 0.8558        | 0.9744              |
| 0.1729 | 3000  | 0.0092          | 0.8717               | 0.8797            | 0.8757        | 0.9780              |
| 0.2306 | 4000  | 0.0082          | 0.8811               | 0.8886            | 0.8848        | 0.9798              |
| 0.2882 | 5000  | 0.0084          | 0.8523               | 0.9163            | 0.8831        | 0.9790              |
| 0.3459 | 6000  | 0.0079          | 0.8700               | 0.9113            | 0.8902        | 0.9802              |
| 0.4035 | 7000  | 0.0070          | 0.9107               | 0.8859            | 0.8981        | 0.9822              |
| 0.4611 | 8000  | 0.0069          | 0.9259               | 0.8797            | 0.9022        | 0.9827              |
| 0.5188 | 9000  | 0.0067          | 0.9061               | 0.8965            | 0.9013        | 0.9829              |
| 0.5764 | 10000 | 0.0066          | 0.9034               | 0.8996            | 0.9015        | 0.9829              |
| 0.6341 | 11000 | 0.0064          | 0.9160               | 0.8996            | 0.9077        | 0.9839              |
| 0.6917 | 12000 | 0.0066          | 0.8952               | 0.9121            | 0.9036        | 0.9832              |
| 0.7494 | 13000 | 0.0062          | 0.9165               | 0.9009            | 0.9086        | 0.9841              |
| 0.8070 | 14000 | 0.0062          | 0.9010               | 0.9121            | 0.9065        | 0.9835              |
| 0.8647 | 15000 | 0.0062          | 0.9084               | 0.9127            | 0.9105        | 0.9842              |
| 0.9223 | 16000 | 0.0060          | 0.9151               | 0.9098            | 0.9125        | 0.9846              |
| 0.9799 | 17000 | 0.0060          | 0.9149               | 0.9113            | 0.9131        | 0.9848              |

### Framework Versions
- Python: 3.8.16
- SpanMarker: 1.5.0
- Transformers: 4.29.0.dev0
- PyTorch: 1.10.1
- Datasets: 2.15.0
- Tokenizers: 0.13.2

## Citation

### BibTeX
```
@software{Aarsen_SpanMarker,
    author = {Aarsen, Tom},
    license = {Apache-2.0},
    title = {{SpanMarker for Named Entity Recognition}},
    url = {https://github.com/tomaarsen/SpanMarkerNER}
}
```

<!--
## Glossary

*Clearly define terms in order to be accessible across audiences.*
-->

<!--
## Model Card Authors

*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->

<!--
## Model Card Contact

*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->