asahi417 commited on
Commit
eab0df8
1 Parent(s): c751c90

model update

Browse files
README.md ADDED
@@ -0,0 +1,176 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ datasets:
3
+ - tner/tweetner7
4
+ metrics:
5
+ - f1
6
+ - precision
7
+ - recall
8
+ model-index:
9
+ - name: tner/bert-large-tweetner7-random
10
+ results:
11
+ - task:
12
+ name: Token Classification
13
+ type: token-classification
14
+ dataset:
15
+ name: tner/tweetner7/test_2021
16
+ type: tner/tweetner7/test_2021
17
+ args: tner/tweetner7/test_2021
18
+ metrics:
19
+ - name: F1
20
+ type: f1
21
+ value: 0.6238958623895862
22
+ - name: Precision
23
+ type: precision
24
+ value: 0.6271028037383177
25
+ - name: Recall
26
+ type: recall
27
+ value: 0.6207215541165587
28
+ - name: F1 (macro)
29
+ type: f1_macro
30
+ value: 0.5754103658637805
31
+ - name: Precision (macro)
32
+ type: precision_macro
33
+ value: 0.5760445653768616
34
+ - name: Recall (macro)
35
+ type: recall_macro
36
+ value: 0.5751041088351385
37
+ - name: F1 (entity span)
38
+ type: f1_entity_span
39
+ value: 0.7648665930360984
40
+ - name: Precision (entity span)
41
+ type: precision_entity_span
42
+ value: 0.7689340813464236
43
+ - name: Recall (entity span)
44
+ type: recall_entity_span
45
+ value: 0.760841910489187
46
+ - task:
47
+ name: Token Classification
48
+ type: token-classification
49
+ dataset:
50
+ name: tner/tweetner7/test_2020
51
+ type: tner/tweetner7/test_2020
52
+ args: tner/tweetner7/test_2020
53
+ metrics:
54
+ - name: F1
55
+ type: f1
56
+ value: 0.6154274575327208
57
+ - name: Precision
58
+ type: precision
59
+ value: 0.6640625
60
+ - name: Recall
61
+ type: recall
62
+ value: 0.5734302023871303
63
+ - name: F1 (macro)
64
+ type: f1_macro
65
+ value: 0.5709159092071027
66
+ - name: Precision (macro)
67
+ type: precision_macro
68
+ value: 0.6168953196783556
69
+ - name: Recall (macro)
70
+ type: recall_macro
71
+ value: 0.5321784485961766
72
+ - name: F1 (entity span)
73
+ type: f1_entity_span
74
+ value: 0.732943469785575
75
+ - name: Precision (entity span)
76
+ type: precision_entity_span
77
+ value: 0.7908653846153846
78
+ - name: Recall (entity span)
79
+ type: recall_entity_span
80
+ value: 0.6829268292682927
81
+
82
+ pipeline_tag: token-classification
83
+ widget:
84
+ - text: "Get the all-analog Classic Vinyl Edition of `Takin' Off` Album from {{@Herbie Hancock@}} via {{USERNAME}} link below: {{URL}}"
85
+ example_title: "NER Example 1"
86
+ ---
87
+ # tner/bert-large-tweetner7-random
88
+
89
+ This model is a fine-tuned version of [bert-large-cased](https://huggingface.co/bert-large-cased) on the
90
+ [tner/tweetner7](https://huggingface.co/datasets/tner/tweetner7) dataset (`train_random` split).
91
+ Model fine-tuning is done via [T-NER](https://github.com/asahi417/tner)'s hyper-parameter search (see the repository
92
+ for more detail). It achieves the following results on the test set of 2021:
93
+ - F1 (micro): 0.6238958623895862
94
+ - Precision (micro): 0.6271028037383177
95
+ - Recall (micro): 0.6207215541165587
96
+ - F1 (macro): 0.5754103658637805
97
+ - Precision (macro): 0.5760445653768616
98
+ - Recall (macro): 0.5751041088351385
99
+
100
+
101
+
102
+ The per-entity breakdown of the F1 score on the test set are below:
103
+ - corporation: 0.49146005509641877
104
+ - creative_work: 0.3972413793103448
105
+ - event: 0.44788732394366193
106
+ - group: 0.5767073573078192
107
+ - location: 0.6721649484536083
108
+ - person: 0.8116810183451891
109
+ - product: 0.6307304785894207
110
+
111
+ For F1 scores, the confidence interval is obtained by bootstrap as below:
112
+ - F1 (micro):
113
+ - 90%: [0.6141711573096726, 0.6329835528622229]
114
+ - 95%: [0.6127880547187768, 0.6345997209553179]
115
+ - F1 (macro):
116
+ - 90%: [0.6141711573096726, 0.6329835528622229]
117
+ - 95%: [0.6127880547187768, 0.6345997209553179]
118
+
119
+ Full evaluation can be found at [metric file of NER](https://huggingface.co/tner/bert-large-tweetner7-random/raw/main/eval/metric.json)
120
+ and [metric file of entity span](https://huggingface.co/tner/bert-large-tweetner7-random/raw/main/eval/metric_span.json).
121
+
122
+ ### Usage
123
+ This model can be used through the [tner library](https://github.com/asahi417/tner). Install the library via pip
124
+ ```shell
125
+ pip install tner
126
+ ```
127
+ and activate model as below.
128
+ ```python
129
+ from tner import TransformersNER
130
+ model = TransformersNER("tner/bert-large-tweetner7-random")
131
+ model.predict(["Jacob Collier is a Grammy awarded English artist from London"])
132
+ ```
133
+ It can be used via transformers library but it is not recommended as CRF layer is not supported at the moment.
134
+
135
+ ### Training hyperparameters
136
+
137
+ The following hyperparameters were used during training:
138
+ - dataset: ['tner/tweetner7']
139
+ - dataset_split: train_random
140
+ - dataset_name: None
141
+ - local_dataset: None
142
+ - model: bert-large-cased
143
+ - crf: True
144
+ - max_length: 128
145
+ - epoch: 30
146
+ - batch_size: 32
147
+ - lr: 0.0001
148
+ - random_seed: 0
149
+ - gradient_accumulation_steps: 1
150
+ - weight_decay: 1e-07
151
+ - lr_warmup_step_ratio: 0.15
152
+ - max_grad_norm: 1
153
+
154
+ The full configuration can be found at [fine-tuning parameter file](https://huggingface.co/tner/bert-large-tweetner7-random/raw/main/trainer_config.json).
155
+
156
+ ### Reference
157
+ If you use any resource from T-NER, please consider to cite our [paper](https://aclanthology.org/2021.eacl-demos.7/).
158
+
159
+ ```
160
+
161
+ @inproceedings{ushio-camacho-collados-2021-ner,
162
+ title = "{T}-{NER}: An All-Round Python Library for Transformer-based Named Entity Recognition",
163
+ author = "Ushio, Asahi and
164
+ Camacho-Collados, Jose",
165
+ booktitle = "Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations",
166
+ month = apr,
167
+ year = "2021",
168
+ address = "Online",
169
+ publisher = "Association for Computational Linguistics",
170
+ url = "https://aclanthology.org/2021.eacl-demos.7",
171
+ doi = "10.18653/v1/2021.eacl-demos.7",
172
+ pages = "53--62",
173
+ 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.",
174
+ }
175
+
176
+ ```
eval/metric.json DELETED
@@ -1 +0,0 @@
1
- {"random.dev": {"micro/f1": 0.619047619047619, "micro/f1_ci": {}, "micro/recall": 0.596903363587827, "micro/precision": 0.6428982173663025, "macro/f1": 0.5686540193943024, "macro/f1_ci": {}, "macro/recall": 0.547210195927133, "macro/precision": 0.5921601321577861, "per_entity_metric": {"corporation": {"f1": 0.5444743935309972, "f1_ci": {}, "precision": 0.5674157303370787, "recall": 0.5233160621761658}, "creative_work": {"f1": 0.4155844155844156, "f1_ci": {}, "precision": 0.4383561643835616, "recall": 0.3950617283950617}, "event": {"f1": 0.3593073593073593, "f1_ci": {}, "precision": 0.3824884792626728, "recall": 0.33877551020408164}, "group": {"f1": 0.6086956521739131, "f1_ci": {}, "precision": 0.632398753894081, "recall": 0.5867052023121387}, "location": {"f1": 0.5993485342019543, "f1_ci": {}, "precision": 0.6388888888888888, "recall": 0.5644171779141104}, "person": {"f1": 0.830945558739255, "f1_ci": {}, "precision": 0.8463035019455253, "recall": 0.8161350844277674}, "product": {"f1": 0.6222222222222221, "f1_ci": {}, "precision": 0.639269406392694, "recall": 0.6060606060606061}}}, "2021.test": {"micro/f1": 0.6238958623895862, "micro/f1_ci": {"90": [0.6141711573096726, 0.6329835528622229], "95": [0.6127880547187768, 0.6345997209553179]}, "micro/recall": 0.6207215541165587, "micro/precision": 0.6271028037383177, "macro/f1": 0.5754103658637805, "macro/f1_ci": {"90": [0.565187314674676, 0.5855711118073963], "95": [0.563829302210402, 0.5869723763958868]}, "macro/recall": 0.5751041088351385, "macro/precision": 0.5760445653768616, "per_entity_metric": {"corporation": {"f1": 0.49146005509641877, "f1_ci": {"90": [0.46653944279830356, 0.5156783797722426], "95": [0.4614505665499631, 0.5195247572390868]}, "precision": 0.48743169398907105, "recall": 0.4955555555555556}, "creative_work": {"f1": 0.3972413793103448, "f1_ci": {"90": [0.3643388832934659, 0.431475468975469], "95": [0.357765951797727, 0.4368833727344366]}, "precision": 0.40055632823365783, "recall": 0.39398084815321477}, "event": {"f1": 0.44788732394366193, "f1_ci": {"90": [0.4252921892767356, 0.4712090747178809], "95": [0.42052975286951233, 0.4741356529398811]}, "precision": 0.4626576139670223, "recall": 0.4340309372156506}, "group": {"f1": 0.5767073573078192, "f1_ci": {"90": [0.5553909855557407, 0.5986221015634983], "95": [0.5507048431429983, 0.6039963050868477]}, "precision": 0.5776602775941837, "recall": 0.5757575757575758}, "location": {"f1": 0.6721649484536083, "f1_ci": {"90": [0.6437046042939083, 0.6989743577005144], "95": [0.6382294156836619, 0.7034515355421547]}, "precision": 0.6617050067658998, "recall": 0.6829608938547486}, "person": {"f1": 0.8116810183451891, "f1_ci": {"90": [0.800425398613887, 0.8226651182198794], "95": [0.7974815277796216, 0.8242620017446516]}, "precision": 0.8243346007604563, "recall": 0.799410029498525}, "product": {"f1": 0.6307304785894207, "f1_ci": {"90": [0.6096523222472238, 0.6528996533628258], "95": [0.6046932305919911, 0.6557772793770326]}, "precision": 0.6179664363277394, "recall": 0.6440329218106996}}}, "2020.test": {"micro/f1": 0.6154274575327208, "micro/f1_ci": {"90": [0.5947241888019572, 0.6356762437073024], "95": [0.5903024143830322, 0.6396733847327379]}, "micro/recall": 0.5734302023871303, "micro/precision": 0.6640625, "macro/f1": 0.5709159092071027, "macro/f1_ci": {"90": [0.5485436010515674, 0.5911412312900881], "95": [0.54348814668388, 0.5952871224560826]}, "macro/recall": 0.5321784485961766, "macro/precision": 0.6168953196783556, "per_entity_metric": {"corporation": {"f1": 0.5633802816901409, "f1_ci": {"90": [0.5108901703453305, 0.6114260304694146], "95": [0.5013283378746595, 0.6201379487863938]}, "precision": 0.6097560975609756, "recall": 0.5235602094240838}, "creative_work": {"f1": 0.4281150159744409, "f1_ci": {"90": [0.3629106221339231, 0.4872107567229519], "95": [0.35123805851861367, 0.49881329442767974]}, "precision": 0.5, "recall": 0.3743016759776536}, "event": {"f1": 0.4268774703557312, "f1_ci": {"90": [0.3761945562305827, 0.47865295050923795], "95": [0.36749324466274025, 0.48812845288166423]}, "precision": 0.44813278008298757, "recall": 0.4075471698113208}, "group": {"f1": 0.5335689045936396, "f1_ci": {"90": [0.4815280555531712, 0.5886398331255172], "95": [0.47204810764639515, 0.5993434343434344]}, "precision": 0.592156862745098, "recall": 0.4855305466237942}, "location": {"f1": 0.6198083067092652, "f1_ci": {"90": [0.5512898975732075, 0.6781922675026124], "95": [0.534796626984127, 0.6899004797898276]}, "precision": 0.6554054054054054, "recall": 0.5878787878787879}, "person": {"f1": 0.810375670840787, "f1_ci": {"90": [0.7806688121728075, 0.8346625323604978], "95": [0.7761355449156406, 0.8400027100271004]}, "precision": 0.867816091954023, "recall": 0.7600671140939598}, "product": {"f1": 0.6142857142857142, "f1_ci": {"90": [0.5632209365970393, 0.6650651474676988], "95": [0.5506497316430371, 0.6756231123983224]}, "precision": 0.645, "recall": 0.5863636363636363}}}, "2021.test (span detection)": {"micro/f1": 0.7648665930360984, "micro/f1_ci": {}, "micro/recall": 0.760841910489187, "micro/precision": 0.7689340813464236, "macro/f1": 0.7648665930360984, "macro/f1_ci": {}, "macro/recall": 0.760841910489187, "macro/precision": 0.7689340813464236}, "2020.test (span detection)": {"micro/f1": 0.732943469785575, "micro/f1_ci": {}, "micro/recall": 0.6829268292682927, "micro/precision": 0.7908653846153846, "macro/f1": 0.732943469785575, "macro/f1_ci": {}, "macro/recall": 0.6829268292682927, "macro/precision": 0.7908653846153846}}
 
 
eval/metric.test_2020.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"micro/f1": 0.6154274575327208, "micro/f1_ci": {"90": [0.5947241888019572, 0.6356762437073024], "95": [0.5903024143830322, 0.6396733847327379]}, "micro/recall": 0.5734302023871303, "micro/precision": 0.6640625, "macro/f1": 0.5709159092071027, "macro/f1_ci": {"90": [0.5485436010515674, 0.5911412312900881], "95": [0.54348814668388, 0.5952871224560826]}, "macro/recall": 0.5321784485961766, "macro/precision": 0.6168953196783556, "per_entity_metric": {"corporation": {"f1": 0.5633802816901409, "f1_ci": {"90": [0.5108901703453305, 0.6114260304694146], "95": [0.5013283378746595, 0.6201379487863938]}, "precision": 0.6097560975609756, "recall": 0.5235602094240838}, "creative_work": {"f1": 0.4281150159744409, "f1_ci": {"90": [0.3629106221339231, 0.4872107567229519], "95": [0.35123805851861367, 0.49881329442767974]}, "precision": 0.5, "recall": 0.3743016759776536}, "event": {"f1": 0.4268774703557312, "f1_ci": {"90": [0.3761945562305827, 0.47865295050923795], "95": [0.36749324466274025, 0.48812845288166423]}, "precision": 0.44813278008298757, "recall": 0.4075471698113208}, "group": {"f1": 0.5335689045936396, "f1_ci": {"90": [0.4815280555531712, 0.5886398331255172], "95": [0.47204810764639515, 0.5993434343434344]}, "precision": 0.592156862745098, "recall": 0.4855305466237942}, "location": {"f1": 0.6198083067092652, "f1_ci": {"90": [0.5512898975732075, 0.6781922675026124], "95": [0.534796626984127, 0.6899004797898276]}, "precision": 0.6554054054054054, "recall": 0.5878787878787879}, "person": {"f1": 0.810375670840787, "f1_ci": {"90": [0.7806688121728075, 0.8346625323604978], "95": [0.7761355449156406, 0.8400027100271004]}, "precision": 0.867816091954023, "recall": 0.7600671140939598}, "product": {"f1": 0.6142857142857142, "f1_ci": {"90": [0.5632209365970393, 0.6650651474676988], "95": [0.5506497316430371, 0.6756231123983224]}, "precision": 0.645, "recall": 0.5863636363636363}}}
eval/metric.test_2021.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"micro/f1": 0.6238958623895862, "micro/f1_ci": {"90": [0.6141711573096726, 0.6329835528622229], "95": [0.6127880547187768, 0.6345997209553179]}, "micro/recall": 0.6207215541165587, "micro/precision": 0.6271028037383177, "macro/f1": 0.5754103658637805, "macro/f1_ci": {"90": [0.565187314674676, 0.5855711118073963], "95": [0.563829302210402, 0.5869723763958868]}, "macro/recall": 0.5751041088351385, "macro/precision": 0.5760445653768616, "per_entity_metric": {"corporation": {"f1": 0.49146005509641877, "f1_ci": {"90": [0.46653944279830356, 0.5156783797722426], "95": [0.4614505665499631, 0.5195247572390868]}, "precision": 0.48743169398907105, "recall": 0.4955555555555556}, "creative_work": {"f1": 0.3972413793103448, "f1_ci": {"90": [0.3643388832934659, 0.431475468975469], "95": [0.357765951797727, 0.4368833727344366]}, "precision": 0.40055632823365783, "recall": 0.39398084815321477}, "event": {"f1": 0.44788732394366193, "f1_ci": {"90": [0.4252921892767356, 0.4712090747178809], "95": [0.42052975286951233, 0.4741356529398811]}, "precision": 0.4626576139670223, "recall": 0.4340309372156506}, "group": {"f1": 0.5767073573078192, "f1_ci": {"90": [0.5553909855557407, 0.5986221015634983], "95": [0.5507048431429983, 0.6039963050868477]}, "precision": 0.5776602775941837, "recall": 0.5757575757575758}, "location": {"f1": 0.6721649484536083, "f1_ci": {"90": [0.6437046042939083, 0.6989743577005144], "95": [0.6382294156836619, 0.7034515355421547]}, "precision": 0.6617050067658998, "recall": 0.6829608938547486}, "person": {"f1": 0.8116810183451891, "f1_ci": {"90": [0.800425398613887, 0.8226651182198794], "95": [0.7974815277796216, 0.8242620017446516]}, "precision": 0.8243346007604563, "recall": 0.799410029498525}, "product": {"f1": 0.6307304785894207, "f1_ci": {"90": [0.6096523222472238, 0.6528996533628258], "95": [0.6046932305919911, 0.6557772793770326]}, "precision": 0.6179664363277394, "recall": 0.6440329218106996}}}
eval/metric_span.test_2020.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"micro/f1": 0.732943469785575, "micro/f1_ci": {}, "micro/recall": 0.6829268292682927, "micro/precision": 0.7908653846153846, "macro/f1": 0.732943469785575, "macro/f1_ci": {}, "macro/recall": 0.6829268292682927, "macro/precision": 0.7908653846153846}
eval/metric_span.test_2021.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"micro/f1": 0.7648665930360984, "micro/f1_ci": {}, "micro/recall": 0.760841910489187, "micro/precision": 0.7689340813464236, "macro/f1": 0.7648665930360984, "macro/f1_ci": {}, "macro/recall": 0.760841910489187, "macro/precision": 0.7689340813464236}
eval/prediction.2020.test.json DELETED
The diff for this file is too large to render. See raw diff
 
eval/prediction.2021.test.json DELETED
The diff for this file is too large to render. See raw diff
 
eval/prediction.random.dev.json DELETED
The diff for this file is too large to render. See raw diff
 
trainer_config.json CHANGED
@@ -1 +1 @@
1
- {"data_split": "random.train", "model": "bert-large-cased", "crf": true, "max_length": 128, "epoch": 30, "batch_size": 32, "lr": 0.0001, "random_seed": 0, "gradient_accumulation_steps": 1, "weight_decay": 1e-07, "lr_warmup_step_ratio": 0.15, "max_grad_norm": 1}
 
1
+ {"dataset": ["tner/tweetner7"], "dataset_split": "train_random", "dataset_name": null, "local_dataset": null, "model": "bert-large-cased", "crf": true, "max_length": 128, "epoch": 30, "batch_size": 32, "lr": 0.0001, "random_seed": 0, "gradient_accumulation_steps": 1, "weight_decay": 1e-07, "lr_warmup_step_ratio": 0.15, "max_grad_norm": 1}