model update

README.md

@@ -0,0 +1,176 @@
+---
+datasets:
+- tner/tweetner7
+metrics:
+- f1
+- precision
+- recall
+model-index:
+- name: tner/twitter-roberta-base-dec2020-tweetner7-2020-2021-continuous
+  results:
+  - task:
+      name: Token Classification
+      type: token-classification
+    dataset:
+      name: tner/tweetner7/test_2021
+      type: tner/tweetner7/test_2021
+      args: tner/tweetner7/test_2021
+    metrics:
+    - name: F1
+      type: f1
+      value: 0.655146764318819
+    - name: Precision
+      type: precision
+      value: 0.6484313059236607
+    - name: Recall
+      type: recall
+      value: 0.6620027752081407
+    - name: F1 (macro)
+      type: f1_macro
+      value: 0.60565538970149
+    - name: Precision (macro)
+      type: precision_macro
+      value: 0.5978135601251405
+    - name: Recall (macro)
+      type: recall_macro
+      value: 0.6152969312272543
+    - name: F1 (entity span)
+      type: f1_entity_span
+      value: 0.7802700846875715
+    - name: Precision (entity span)
+      type: precision_entity_span
+      value: 0.7722278853777325
+    - name: Recall (entity span)
+      type: recall_entity_span
+      value: 0.7884815542962877
+  - task:
+      name: Token Classification
+      type: token-classification
+    dataset:
+      name: tner/tweetner7/test_2020
+      type: tner/tweetner7/test_2020
+      args: tner/tweetner7/test_2020
+    metrics:
+    - name: F1
+      type: f1
+      value: 0.6529060293318849
+    - name: Precision
+      type: precision
+      value: 0.6849002849002849
+    - name: Recall
+      type: recall
+      value: 0.6237675142708874
+    - name: F1 (macro)
+      type: f1_macro
+      value: 0.6127864056494463
+    - name: Precision (macro)
+      type: precision_macro
+      value: 0.6440791059118922
+    - name: Recall (macro)
+      type: recall_macro
+      value: 0.5885664058069695
+    - name: F1 (entity span)
+      type: f1_entity_span
+      value: 0.7588267246061923
+    - name: Precision (entity span)
+      type: precision_entity_span
+      value: 0.796011396011396
+    - name: Recall (entity span)
+      type: recall_entity_span
+      value: 0.724961079398028
+
+pipeline_tag: token-classification
+widget:
+- text: "Get the all-analog Classic Vinyl Edition of `Takin' Off` Album from {{@Herbie Hancock@}} via {{USERNAME}} link below: {{URL}}"
+  example_title: "NER Example 1"
+---
+# tner/twitter-roberta-base-dec2020-tweetner7-2020-2021-continuous
+
+This model is a fine-tuned version of [tner/twitter-roberta-base-dec2020-tweetner-2020](https://huggingface.co/tner/twitter-roberta-base-dec2020-tweetner-2020) on the 
+[tner/tweetner7](https://huggingface.co/datasets/tner/tweetner7) dataset (`train_2021` split). The model is first fine-tuned on `train_2020`, and then continuously fine-tuned on `train_2021`. 
+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): 0.655146764318819
+- Precision (micro): 0.6484313059236607
+- Recall (micro): 0.6620027752081407
+- F1 (macro): 0.60565538970149
+- Precision (macro): 0.5978135601251405
+- Recall (macro): 0.6152969312272543
+
+
+
+The per-entity breakdown of the F1 score on the test set are below:
+- corporation: 0.5356371490280778
+- creative_work: 0.4529526281635302
+- event: 0.4692272096251735
+- group: 0.610738255033557
+- location: 0.6627831715210356
+- person: 0.8433472499546196
+- product: 0.6649020645844361 
+
+For F1 scores, the confidence interval is obtained by bootstrap as below:
+- F1 (micro): 
+    - 90%: [0.646485917836786, 0.6644401423537809]
+    - 95%: [0.6449507873997479, 0.6659444015725502] 
+- F1 (macro): 
+    - 90%: [0.646485917836786, 0.6644401423537809]
+    - 95%: [0.6449507873997479, 0.6659444015725502] 
+
+Full evaluation can be found at [metric file of NER](https://huggingface.co/tner/twitter-roberta-base-dec2020-tweetner7-2020-2021-continuous/raw/main/eval/metric.json) 
+and [metric file of entity span](https://huggingface.co/tner/twitter-roberta-base-dec2020-tweetner7-2020-2021-continuous/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
+```
+and activate model as below.
+```python
+from tner import TransformersNER
+model = TransformersNER("tner/twitter-roberta-base-dec2020-tweetner7-2020-2021-continuous")
+model.predict(["Jacob Collier is a Grammy awarded English artist from London"])
+```
+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_2021
+ - dataset_name: None
+ - local_dataset: None
+ - model: tner/twitter-roberta-base-dec2020-tweetner-2020
+ - crf: True
+ - max_length: 128
+ - epoch: 30
+ - batch_size: 32
+ - lr: 1e-06
+ - random_seed: 0
+ - gradient_accumulation_steps: 1
+ - weight_decay: 1e-07
+ - lr_warmup_step_ratio: 0.3
+ - max_grad_norm: 1
+
+The full configuration can be found at [fine-tuning parameter file](https://huggingface.co/tner/twitter-roberta-base-dec2020-tweetner7-2020-2021-continuous/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.",
+}
+
+```

eval/metric.json

@@ -1 +0,0 @@
-{"2021.dev": {"micro/f1": 0.6494949494949495, "micro/f1_ci": {}, "micro/recall": 0.643, "micro/precision": 0.6561224489795918, "macro/f1": 0.6068039027139369, "macro/f1_ci": {}, "macro/recall": 0.605166519640185, "macro/precision": 0.6118349239758869, "per_entity_metric": {"corporation": {"f1": 0.6176470588235294, "f1_ci": {}, "precision": 0.6176470588235294, "recall": 0.6176470588235294}, "creative_work": {"f1": 0.4713375796178344, "f1_ci": {}, "precision": 0.4457831325301205, "recall": 0.5}, "event": {"f1": 0.3951612903225807, "f1_ci": {}, "precision": 0.4188034188034188, "recall": 0.37404580152671757}, "group": {"f1": 0.6267281105990783, "f1_ci": {}, "precision": 0.6570048309178744, "recall": 0.5991189427312775}, "location": {"f1": 0.6233766233766234, "f1_ci": {}, "precision": 0.5853658536585366, "recall": 0.6666666666666666}, "person": {"f1": 0.8304498269896194, "f1_ci": {}, "precision": 0.8135593220338984, "recall": 0.8480565371024735}, "product": {"f1": 0.6829268292682927, "f1_ci": {}, "precision": 0.7446808510638298, "recall": 0.6306306306306306}}}, "2021.test": {"micro/f1": 0.655146764318819, "micro/f1_ci": {"90": [0.646485917836786, 0.6644401423537809], "95": [0.6449507873997479, 0.6659444015725502]}, "micro/recall": 0.6620027752081407, "micro/precision": 0.6484313059236607, "macro/f1": 0.60565538970149, "macro/f1_ci": {"90": [0.5958609226036335, 0.615029496768845], "95": [0.59347853460072, 0.6177128475619766]}, "macro/recall": 0.6152969312272543, "macro/precision": 0.5978135601251405, "per_entity_metric": {"corporation": {"f1": 0.5356371490280778, "f1_ci": {"90": [0.5114094871504585, 0.5608433653038019], "95": [0.5052604332153696, 0.5661445279866333]}, "precision": 0.5210084033613446, "recall": 0.5511111111111111}, "creative_work": {"f1": 0.4529526281635302, "f1_ci": {"90": [0.42236024844720493, 0.48234065608829313], "95": [0.4175525495344849, 0.48726058770701636]}, "precision": 0.4308641975308642, "recall": 0.4774281805745554}, "event": {"f1": 0.4692272096251735, "f1_ci": {"90": [0.44730383541051455, 0.4914571996286822], "95": [0.44300273768043796, 0.497628418609345]}, "precision": 0.4774011299435028, "recall": 0.46132848043676067}, "group": {"f1": 0.610738255033557, "f1_ci": {"90": [0.5911010424446823, 0.6317182638626475], "95": [0.587397323488694, 0.636247821945832]}, "precision": 0.6224350205198358, "recall": 0.5994729907773386}, "location": {"f1": 0.6627831715210356, "f1_ci": {"90": [0.6361094283387652, 0.6906877216895649], "95": [0.6284165623933567, 0.6944596639532455]}, "precision": 0.617611580217129, "recall": 0.7150837988826816}, "person": {"f1": 0.8433472499546196, "f1_ci": {"90": [0.8330629714737499, 0.8537574855967393], "95": [0.8310271369505687, 0.8553120874462692]}, "precision": 0.8305327136217375, "recall": 0.8565634218289085}, "product": {"f1": 0.6649020645844361, "f1_ci": {"90": [0.6436639052703277, 0.6861324260947763], "95": [0.639253505828489, 0.6900270969352551]}, "precision": 0.6848418756815703, "recall": 0.6460905349794238}}}, "2020.test": {"micro/f1": 0.6529060293318849, "micro/f1_ci": {"90": [0.6320560339514061, 0.6714840239575703], "95": [0.6277113779725418, 0.6759792334494775]}, "micro/recall": 0.6237675142708874, "micro/precision": 0.6849002849002849, "macro/f1": 0.6127864056494463, "macro/f1_ci": {"90": [0.5897169937547521, 0.6334131005387055], "95": [0.5857108200385485, 0.6376942274138916]}, "macro/recall": 0.5885664058069695, "macro/precision": 0.6440791059118922, "per_entity_metric": {"corporation": {"f1": 0.5925925925925924, "f1_ci": {"90": [0.533721414937729, 0.6480675262204496], "95": [0.5207716966045063, 0.6541185932818168]}, "precision": 0.5989304812834224, "recall": 0.5863874345549738}, "creative_work": {"f1": 0.5558739255014327, "f1_ci": {"90": [0.4984564943253468, 0.607918557688304], "95": [0.4885970084990079, 0.6177219060891107]}, "precision": 0.5705882352941176, "recall": 0.5418994413407822}, "event": {"f1": 0.4383561643835617, "f1_ci": {"90": [0.3847710023194851, 0.48764285714285716], "95": [0.37720025056234163, 0.49720332993010286]}, "precision": 0.45528455284552843, "recall": 0.4226415094339623}, "group": {"f1": 0.5437616387337058, "f1_ci": {"90": [0.48760765869239764, 0.5993162024857116], "95": [0.47615165631469974, 0.6082547602450746]}, "precision": 0.6460176991150443, "recall": 0.4694533762057878}, "location": {"f1": 0.641399416909621, "f1_ci": {"90": [0.5723443613683027, 0.7030686612965094], "95": [0.5566328196906268, 0.7147551120186363]}, "precision": 0.6179775280898876, "recall": 0.6666666666666666}, "person": {"f1": 0.8407534246575342, "f1_ci": {"90": [0.8133493861413243, 0.8663783854306567], "95": [0.8073137196568275, 0.8716676932880203]}, "precision": 0.8583916083916084, "recall": 0.8238255033557047}, "product": {"f1": 0.6767676767676768, "f1_ci": {"90": [0.6232405642394205, 0.7267706131078224], "95": [0.6133611279563371, 0.7349278036430734]}, "precision": 0.7613636363636364, "recall": 0.6090909090909091}}}, "2021.test (span detection)": {"micro/f1": 0.7802700846875715, "micro/f1_ci": {}, "micro/recall": 0.7884815542962877, "micro/precision": 0.7722278853777325, "macro/f1": 0.7802700846875715, "macro/f1_ci": {}, "macro/recall": 0.7884815542962877, "macro/precision": 0.7722278853777325}, "2020.test (span detection)": {"micro/f1": 0.7588267246061923, "micro/f1_ci": {}, "micro/recall": 0.724961079398028, "micro/precision": 0.796011396011396, "macro/f1": 0.7588267246061923, "macro/f1_ci": {}, "macro/recall": 0.724961079398028, "macro/precision": 0.796011396011396}}

eval/metric.test_2020.json

@@ -0,0 +1 @@
+{"micro/f1": 0.6529060293318849, "micro/f1_ci": {"90": [0.6320560339514061, 0.6714840239575703], "95": [0.6277113779725418, 0.6759792334494775]}, "micro/recall": 0.6237675142708874, "micro/precision": 0.6849002849002849, "macro/f1": 0.6127864056494463, "macro/f1_ci": {"90": [0.5897169937547521, 0.6334131005387055], "95": [0.5857108200385485, 0.6376942274138916]}, "macro/recall": 0.5885664058069695, "macro/precision": 0.6440791059118922, "per_entity_metric": {"corporation": {"f1": 0.5925925925925924, "f1_ci": {"90": [0.533721414937729, 0.6480675262204496], "95": [0.5207716966045063, 0.6541185932818168]}, "precision": 0.5989304812834224, "recall": 0.5863874345549738}, "creative_work": {"f1": 0.5558739255014327, "f1_ci": {"90": [0.4984564943253468, 0.607918557688304], "95": [0.4885970084990079, 0.6177219060891107]}, "precision": 0.5705882352941176, "recall": 0.5418994413407822}, "event": {"f1": 0.4383561643835617, "f1_ci": {"90": [0.3847710023194851, 0.48764285714285716], "95": [0.37720025056234163, 0.49720332993010286]}, "precision": 0.45528455284552843, "recall": 0.4226415094339623}, "group": {"f1": 0.5437616387337058, "f1_ci": {"90": [0.48760765869239764, 0.5993162024857116], "95": [0.47615165631469974, 0.6082547602450746]}, "precision": 0.6460176991150443, "recall": 0.4694533762057878}, "location": {"f1": 0.641399416909621, "f1_ci": {"90": [0.5723443613683027, 0.7030686612965094], "95": [0.5566328196906268, 0.7147551120186363]}, "precision": 0.6179775280898876, "recall": 0.6666666666666666}, "person": {"f1": 0.8407534246575342, "f1_ci": {"90": [0.8133493861413243, 0.8663783854306567], "95": [0.8073137196568275, 0.8716676932880203]}, "precision": 0.8583916083916084, "recall": 0.8238255033557047}, "product": {"f1": 0.6767676767676768, "f1_ci": {"90": [0.6232405642394205, 0.7267706131078224], "95": [0.6133611279563371, 0.7349278036430734]}, "precision": 0.7613636363636364, "recall": 0.6090909090909091}}}

eval/metric.test_2021.json

@@ -0,0 +1 @@
+{"micro/f1": 0.655146764318819, "micro/f1_ci": {"90": [0.646485917836786, 0.6644401423537809], "95": [0.6449507873997479, 0.6659444015725502]}, "micro/recall": 0.6620027752081407, "micro/precision": 0.6484313059236607, "macro/f1": 0.60565538970149, "macro/f1_ci": {"90": [0.5958609226036335, 0.615029496768845], "95": [0.59347853460072, 0.6177128475619766]}, "macro/recall": 0.6152969312272543, "macro/precision": 0.5978135601251405, "per_entity_metric": {"corporation": {"f1": 0.5356371490280778, "f1_ci": {"90": [0.5114094871504585, 0.5608433653038019], "95": [0.5052604332153696, 0.5661445279866333]}, "precision": 0.5210084033613446, "recall": 0.5511111111111111}, "creative_work": {"f1": 0.4529526281635302, "f1_ci": {"90": [0.42236024844720493, 0.48234065608829313], "95": [0.4175525495344849, 0.48726058770701636]}, "precision": 0.4308641975308642, "recall": 0.4774281805745554}, "event": {"f1": 0.4692272096251735, "f1_ci": {"90": [0.44730383541051455, 0.4914571996286822], "95": [0.44300273768043796, 0.497628418609345]}, "precision": 0.4774011299435028, "recall": 0.46132848043676067}, "group": {"f1": 0.610738255033557, "f1_ci": {"90": [0.5911010424446823, 0.6317182638626475], "95": [0.587397323488694, 0.636247821945832]}, "precision": 0.6224350205198358, "recall": 0.5994729907773386}, "location": {"f1": 0.6627831715210356, "f1_ci": {"90": [0.6361094283387652, 0.6906877216895649], "95": [0.6284165623933567, 0.6944596639532455]}, "precision": 0.617611580217129, "recall": 0.7150837988826816}, "person": {"f1": 0.8433472499546196, "f1_ci": {"90": [0.8330629714737499, 0.8537574855967393], "95": [0.8310271369505687, 0.8553120874462692]}, "precision": 0.8305327136217375, "recall": 0.8565634218289085}, "product": {"f1": 0.6649020645844361, "f1_ci": {"90": [0.6436639052703277, 0.6861324260947763], "95": [0.639253505828489, 0.6900270969352551]}, "precision": 0.6848418756815703, "recall": 0.6460905349794238}}}

eval/metric_span.test_2020.json

@@ -0,0 +1 @@
+{"micro/f1": 0.7588267246061923, "micro/f1_ci": {}, "micro/recall": 0.724961079398028, "micro/precision": 0.796011396011396, "macro/f1": 0.7588267246061923, "macro/f1_ci": {}, "macro/recall": 0.724961079398028, "macro/precision": 0.796011396011396}

eval/metric_span.test_2021.json

@@ -0,0 +1 @@
+{"micro/f1": 0.7802700846875715, "micro/f1_ci": {}, "micro/recall": 0.7884815542962877, "micro/precision": 0.7722278853777325, "macro/f1": 0.7802700846875715, "macro/f1_ci": {}, "macro/recall": 0.7884815542962877, "macro/precision": 0.7722278853777325}

trainer_config.json

@@ -1 +1 @@
-{"data_split": "2021.train", "model": "tner/twitter-roberta-base-dec2020-tweetner-2020", "crf": true, "max_length": 128, "epoch": 30, "batch_size": 32, "lr": 1e-06, "random_seed": 0, "gradient_accumulation_steps": 1, "weight_decay": 1e-07, "lr_warmup_step_ratio": 0.3, "max_grad_norm": 1}
+{"dataset": ["tner/tweetner7"], "dataset_split": "train_2021", "dataset_name": null, "local_dataset": null, "model": "tner/twitter-roberta-base-dec2020-tweetner-2020", "crf": true, "max_length": 128, "epoch": 30, "batch_size": 32, "lr": 1e-06, "random_seed": 0, "gradient_accumulation_steps": 1, "weight_decay": 1e-07, "lr_warmup_step_ratio": 0.3, "max_grad_norm": 1}