|
--- |
|
datasets: |
|
- tner/tweetner7 |
|
metrics: |
|
- f1 |
|
- precision |
|
- recall |
|
model-index: |
|
- name: tner/roberta-large-tweetner7-random |
|
results: |
|
- task: |
|
name: Token Classification |
|
type: token-classification |
|
dataset: |
|
name: tner/tweetner7 |
|
type: tner/tweetner7 |
|
args: tner/tweetner7 |
|
metrics: |
|
- name: F1 (test_2021) |
|
type: f1 |
|
value: 0.6632769652650823 |
|
- name: Precision (test_2021) |
|
type: precision |
|
value: 0.6554878048780488 |
|
- name: Recall (test_2021) |
|
type: recall |
|
value: 0.6712534690101758 |
|
- name: Macro F1 (test_2021) |
|
type: f1_macro |
|
value: 0.6096477771855761 |
|
- name: Macro Precision (test_2021) |
|
type: precision_macro |
|
value: 0.6042443991246051 |
|
- name: Macro Recall (test_2021) |
|
type: recall_macro |
|
value: 0.6191008735553379 |
|
- name: Entity Span F1 (test_2021) |
|
type: f1_entity_span |
|
value: 0.7900359938296291 |
|
- name: Entity Span Precision (test_2020) |
|
type: precision_entity_span |
|
value: 0.780713640469738 |
|
- name: Entity Span Recall (test_2021) |
|
type: recall_entity_span |
|
value: 0.7995836706372152 |
|
- name: F1 (test_2020) |
|
type: f1 |
|
value: 0.6439847577572129 |
|
- name: Precision (test_2020) |
|
type: precision |
|
value: 0.6771608471665712 |
|
- name: Recall (test_2020) |
|
type: recall |
|
value: 0.6139076284379865 |
|
- name: Macro F1 (test_2020) |
|
type: f1_macro |
|
value: 0.6008744778169367 |
|
- name: Macro Precision (test_2020) |
|
type: precision_macro |
|
value: 0.6358142893696356 |
|
- name: Macro Recall (test_2020) |
|
type: recall_macro |
|
value: 0.5742193301311931 |
|
- name: Entity Span F1 (test_2020) |
|
type: f1_entity_span |
|
value: 0.7552409474543968 |
|
- name: Entity Span Precision (test_2020) |
|
type: precision_entity_span |
|
value: 0.7943871706758304 |
|
- name: Entity Span Recall (test_2020) |
|
type: recall_entity_span |
|
value: 0.7197716658017644 |
|
|
|
pipeline_tag: token-classification |
|
widget: |
|
- text: "Get the all-analog Classic Vinyl Edition of `Takin' Off` Album from {@herbiehancock@} via {@bluenoterecords@} link below: {{URL}}" |
|
example_title: "NER Example 1" |
|
--- |
|
# tner/roberta-large-tweetner7-random |
|
|
|
This model is a fine-tuned version of [roberta-large](https://huggingface.co/roberta-large) on the |
|
[tner/tweetner7](https://huggingface.co/datasets/tner/tweetner7) dataset (`train_random` split). |
|
Model fine-tuning is done via [T-NER](https://github.com/asahi417/tner)'s hyper-parameter search (see the repository |
|
for more detail). It achieves the following results on the test set of 2021: |
|
- F1 (micro): 0.6632769652650823 |
|
- Precision (micro): 0.6554878048780488 |
|
- Recall (micro): 0.6712534690101758 |
|
- F1 (macro): 0.6096477771855761 |
|
- Precision (macro): 0.6042443991246051 |
|
- Recall (macro): 0.6191008735553379 |
|
|
|
|
|
|
|
The per-entity breakdown of the F1 score on the test set are below: |
|
- corporation: 0.5224148236700539 |
|
- creative_work: 0.45186640471512773 |
|
- event: 0.4894837476099427 |
|
- group: 0.6327722432153899 |
|
- location: 0.6692258477287268 |
|
- person: 0.838405036726128 |
|
- product: 0.6633663366336633 |
|
|
|
For F1 scores, the confidence interval is obtained by bootstrap as below: |
|
- F1 (micro): |
|
- 90%: [0.6546824558783396, 0.6722355436189195] |
|
- 95%: [0.6527609558375069, 0.6741666937877734] |
|
- F1 (macro): |
|
- 90%: [0.6546824558783396, 0.6722355436189195] |
|
- 95%: [0.6527609558375069, 0.6741666937877734] |
|
|
|
Full evaluation can be found at [metric file of NER](https://huggingface.co/tner/roberta-large-tweetner7-random/raw/main/eval/metric.json) |
|
and [metric file of entity span](https://huggingface.co/tner/roberta-large-tweetner7-random/raw/main/eval/metric_span.json). |
|
|
|
### Usage |
|
This model can be used through the [tner library](https://github.com/asahi417/tner). Install the library via pip. |
|
```shell |
|
pip install tner |
|
``` |
|
[TweetNER7](https://huggingface.co/datasets/tner/tweetner7) pre-processed tweets where the account name and URLs are |
|
converted into special formats (see the dataset page for more detail), so we process tweets accordingly and then run the model prediction as below. |
|
|
|
```python |
|
import re |
|
from urlextract import URLExtract |
|
from tner import TransformersNER |
|
|
|
extractor = URLExtract() |
|
|
|
def format_tweet(tweet): |
|
# mask web urls |
|
urls = extractor.find_urls(tweet) |
|
for url in urls: |
|
tweet = tweet.replace(url, "{{URL}}") |
|
# format twitter account |
|
tweet = re.sub(r"\b(\s*)(@[\S]+)\b", r'\1{\2@}', tweet) |
|
return tweet |
|
|
|
|
|
text = "Get the all-analog Classic Vinyl Edition of `Takin' Off` Album from @herbiehancock via @bluenoterecords link below: http://bluenote.lnk.to/AlbumOfTheWeek" |
|
text_format = format_tweet(text) |
|
model = TransformersNER("tner/roberta-large-tweetner7-random") |
|
model.predict([text_format]) |
|
``` |
|
It can be used via transformers library but it is not recommended as CRF layer is not supported at the moment. |
|
|
|
### Training hyperparameters |
|
|
|
The following hyperparameters were used during training: |
|
- dataset: ['tner/tweetner7'] |
|
- dataset_split: train_random |
|
- dataset_name: None |
|
- local_dataset: None |
|
- model: roberta-large |
|
- crf: True |
|
- max_length: 128 |
|
- epoch: 30 |
|
- batch_size: 32 |
|
- lr: 1e-05 |
|
- random_seed: 0 |
|
- gradient_accumulation_steps: 1 |
|
- weight_decay: 1e-07 |
|
- lr_warmup_step_ratio: 0.15 |
|
- max_grad_norm: 1 |
|
|
|
The full configuration can be found at [fine-tuning parameter file](https://huggingface.co/tner/roberta-large-tweetner7-random/raw/main/trainer_config.json). |
|
|
|
### Reference |
|
If you use any resource from T-NER, please consider to cite our [paper](https://aclanthology.org/2021.eacl-demos.7/). |
|
|
|
``` |
|
|
|
@inproceedings{ushio-camacho-collados-2021-ner, |
|
title = "{T}-{NER}: An All-Round Python Library for Transformer-based Named Entity Recognition", |
|
author = "Ushio, Asahi and |
|
Camacho-Collados, Jose", |
|
booktitle = "Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations", |
|
month = apr, |
|
year = "2021", |
|
address = "Online", |
|
publisher = "Association for Computational Linguistics", |
|
url = "https://aclanthology.org/2021.eacl-demos.7", |
|
doi = "10.18653/v1/2021.eacl-demos.7", |
|
pages = "53--62", |
|
abstract = "Language model (LM) pretraining has led to consistent improvements in many NLP downstream tasks, including named entity recognition (NER). In this paper, we present T-NER (Transformer-based Named Entity Recognition), a Python library for NER LM finetuning. In addition to its practical utility, T-NER facilitates the study and investigation of the cross-domain and cross-lingual generalization ability of LMs finetuned on NER. Our library also provides a web app where users can get model predictions interactively for arbitrary text, which facilitates qualitative model evaluation for non-expert programmers. We show the potential of the library by compiling nine public NER datasets into a unified format and evaluating the cross-domain and cross- lingual performance across the datasets. The results from our initial experiments show that in-domain performance is generally competitive across datasets. However, cross-domain generalization is challenging even with a large pretrained LM, which has nevertheless capacity to learn domain-specific features if fine- tuned on a combined dataset. To facilitate future research, we also release all our LM checkpoints via the Hugging Face model hub.", |
|
} |
|
|
|
``` |
|
|
|
|