model update

README.md

@@ -0,0 +1,176 @@
+---
+datasets:
+- tner/tweetner7
+metrics:
+- f1
+- precision
+- recall
+model-index:
+- name: tner/bertweet-large-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: 0.6641431520991053
+    - name: Precision
+      type: precision
+      value: 0.6588529813381885
+    - name: Recall
+      type: recall
+      value: 0.6695189639222942
+    - name: F1 (macro)
+      type: f1_macro
+      value: 0.6165782134695219
+    - name: Precision (macro)
+      type: precision_macro
+      value: 0.6102975783874098
+    - name: Recall (macro)
+      type: recall_macro
+      value: 0.6256153624327598
+    - name: F1 (entity span)
+      type: f1_entity_span
+      value: 0.7896759392027531
+    - name: Precision (entity span)
+      type: precision_entity_span
+      value: 0.783340919435594
+    - name: Recall (entity span)
+      type: recall_entity_span
+      value: 0.7961142592806754
+  - 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: 0.6587912087912088
+    - name: Precision
+      type: precision
+      value: 0.6999416228838296
+    - name: Recall
+      type: recall
+      value: 0.6222106901920083
+    - name: F1 (macro)
+      type: f1_macro
+      value: 0.6182374585427982
+    - name: Precision (macro)
+      type: precision_macro
+      value: 0.6571485734047059
+    - name: Recall (macro)
+      type: recall_macro
+      value: 0.5865594344408018
+    - name: F1 (entity span)
+      type: f1_entity_span
+      value: 0.7641561297416162
+    - name: Precision (entity span)
+      type: precision_entity_span
+      value: 0.8123904149620105
+    - name: Recall (entity span)
+      type: recall_entity_span
+      value: 0.7213284898806435
+
+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/bertweet-large-tweetner7-2020-2021-continuous
+
+This model is a fine-tuned version of [tner/bertweet-large-tweetner-2020](https://huggingface.co/tner/bertweet-large-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): 0.6641431520991053
+- Precision (micro): 0.6588529813381885
+- Recall (micro): 0.6695189639222942
+- F1 (macro): 0.6165782134695219
+- Precision (macro): 0.6102975783874098
+- Recall (macro): 0.6256153624327598
+
+
+
+The per-entity breakdown of the F1 score on the test set are below:
+- corporation: 0.5507246376811594
+- creative_work: 0.4684914067472947
+- event: 0.4815724815724816
+- group: 0.6143572621035058
+- location: 0.6886731391585761
+- person: 0.8404178674351586
+- product: 0.6718106995884774 
+
+For F1 scores, the confidence interval is obtained by bootstrap as below:
+- F1 (micro): 
+    - 90%: [0.6551977421192867, 0.6726790034801573]
+    - 95%: [0.6537478870999098, 0.6745822333244045] 
+- F1 (macro): 
+    - 90%: [0.6551977421192867, 0.6726790034801573]
+    - 95%: [0.6537478870999098, 0.6745822333244045] 
+
+Full evaluation can be found at [metric file of NER](https://huggingface.co/tner/bertweet-large-tweetner7-2020-2021-continuous/raw/main/eval/metric.json) 
+and [metric file of entity span](https://huggingface.co/tner/bertweet-large-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/bertweet-large-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/bertweet-large-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: 0.15
+ - max_grad_norm: 1
+
+The full configuration can be found at [fine-tuning parameter file](https://huggingface.co/tner/bertweet-large-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": 0.6532663316582915, "micro/f1_ci": {}, "micro/recall": 0.65, "micro/precision": 0.6565656565656566, "macro/f1": 0.6130509850715328, "macro/f1_ci": {}, "macro/recall": 0.615356967051296, "macro/precision": 0.613081734826762, "per_entity_metric": {"corporation": {"f1": 0.5904761904761905, "f1_ci": {}, "precision": 0.5740740740740741, "recall": 0.6078431372549019}, "creative_work": {"f1": 0.49673202614379086, "f1_ci": {}, "precision": 0.4810126582278481, "recall": 0.5135135135135135}, "event": {"f1": 0.4170212765957447, "f1_ci": {}, "precision": 0.47115384615384615, "recall": 0.37404580152671757}, "group": {"f1": 0.618510158013544, "f1_ci": {}, "precision": 0.6342592592592593, "recall": 0.6035242290748899}, "location": {"f1": 0.6845637583892618, "f1_ci": {}, "precision": 0.6623376623376623, "recall": 0.7083333333333334}, "person": {"f1": 0.8324697754749568, "f1_ci": {}, "precision": 0.8141891891891891, "recall": 0.8515901060070671}, "product": {"f1": 0.6515837104072398, "f1_ci": {}, "precision": 0.6545454545454545, "recall": 0.6486486486486487}}}, "2021.test": {"micro/f1": 0.6641431520991053, "micro/f1_ci": {"90": [0.6551977421192867, 0.6726790034801573], "95": [0.6537478870999098, 0.6745822333244045]}, "micro/recall": 0.6695189639222942, "micro/precision": 0.6588529813381885, "macro/f1": 0.6165782134695219, "macro/f1_ci": {"90": [0.6067197256596341, 0.6256044510960733], "95": [0.6051019712561241, 0.6272725844616904]}, "macro/recall": 0.6256153624327598, "macro/precision": 0.6102975783874098, "per_entity_metric": {"corporation": {"f1": 0.5507246376811594, "f1_ci": {"90": [0.5254082699108996, 0.5773166181256275], "95": [0.5208478779438392, 0.5823103234936461]}, "precision": 0.5525727069351231, "recall": 0.5488888888888889}, "creative_work": {"f1": 0.4684914067472947, "f1_ci": {"90": [0.43769556974104906, 0.4987696892294594], "95": [0.4306276297645928, 0.503900885472692]}, "precision": 0.4380952380952381, "recall": 0.5034199726402189}, "event": {"f1": 0.4815724815724816, "f1_ci": {"90": [0.45748383888612637, 0.5051356400612352], "95": [0.453626352887228, 0.5078142035880663]}, "precision": 0.5235042735042735, "recall": 0.445859872611465}, "group": {"f1": 0.6143572621035058, "f1_ci": {"90": [0.5931312033684751, 0.6362758250713497], "95": [0.5898185538332443, 0.640990290362259]}, "precision": 0.6228842247799594, "recall": 0.6060606060606061}, "location": {"f1": 0.6886731391585761, "f1_ci": {"90": [0.6638351146228115, 0.7148122115942915], "95": [0.6584092995447562, 0.7185025217834956]}, "precision": 0.6417370325693607, "recall": 0.7430167597765364}, "person": {"f1": 0.8404178674351586, "f1_ci": {"90": [0.8298100715749012, 0.8504605055787255], "95": [0.828366934999076, 0.851877851877852]}, "precision": 0.8214788732394366, "recall": 0.8602507374631269}, "product": {"f1": 0.6718106995884774, "f1_ci": {"90": [0.65082156900923, 0.6922712449645116], "95": [0.6453161615534301, 0.6960348412597874]}, "precision": 0.6718106995884774, "recall": 0.6718106995884774}}}, "2020.test": {"micro/f1": 0.6587912087912088, "micro/f1_ci": {"90": [0.6388349564310538, 0.6775372411164946], "95": [0.6347700739178536, 0.6817724596099083]}, "micro/recall": 0.6222106901920083, "micro/precision": 0.6999416228838296, "macro/f1": 0.6182374585427982, "macro/f1_ci": {"90": [0.5956469999846385, 0.6396562173710457], "95": [0.5924717479435003, 0.643052479903031]}, "macro/recall": 0.5865594344408018, "macro/precision": 0.6571485734047059, "per_entity_metric": {"corporation": {"f1": 0.5837837837837839, "f1_ci": {"90": [0.5257238173647558, 0.6340269047020367], "95": [0.5147848132574822, 0.6415194434712977]}, "precision": 0.6033519553072626, "recall": 0.5654450261780105}, "creative_work": {"f1": 0.5464788732394367, "f1_ci": {"90": [0.4858406810117715, 0.6012249962680998], "95": [0.4752830687830688, 0.6118356854309848]}, "precision": 0.5511363636363636, "recall": 0.5418994413407822}, "event": {"f1": 0.4612244897959184, "f1_ci": {"90": [0.40925298655360004, 0.5121951219512195], "95": [0.4, 0.5224982057102179]}, "precision": 0.5022222222222222, "recall": 0.42641509433962266}, "group": {"f1": 0.5639097744360902, "f1_ci": {"90": [0.5085325152068667, 0.6129582440493426], "95": [0.4999765478424015, 0.6242067115955933]}, "precision": 0.6787330316742082, "recall": 0.48231511254019294}, "location": {"f1": 0.6666666666666666, "f1_ci": {"90": [0.5992845117845118, 0.725033185840708], "95": [0.5871509772636617, 0.7378266550522647]}, "precision": 0.6792452830188679, "recall": 0.6545454545454545}, "person": {"f1": 0.8389319552110249, "f1_ci": {"90": [0.8120086435076491, 0.8618636353151934], "95": [0.8049340069594505, 0.8669009775259731]}, "precision": 0.8619469026548673, "recall": 0.8171140939597316}, "product": {"f1": 0.6666666666666667, "f1_ci": {"90": [0.6168948944597402, 0.7127101210767604], "95": [0.604633351992265, 0.7208783153405305]}, "precision": 0.723404255319149, "recall": 0.6181818181818182}}}, "2021.test (span detection)": {"micro/f1": 0.7896759392027531, "micro/f1_ci": {}, "micro/recall": 0.7961142592806754, "micro/precision": 0.783340919435594, "macro/f1": 0.7896759392027531, "macro/f1_ci": {}, "macro/recall": 0.7961142592806754, "macro/precision": 0.783340919435594}, "2020.test (span detection)": {"micro/f1": 0.7641561297416162, "micro/f1_ci": {}, "micro/recall": 0.7213284898806435, "micro/precision": 0.8123904149620105, "macro/f1": 0.7641561297416162, "macro/f1_ci": {}, "macro/recall": 0.7213284898806435, "macro/precision": 0.8123904149620105}}

eval/metric.test_2020.json

@@ -0,0 +1 @@
+{"micro/f1": 0.6587912087912088, "micro/f1_ci": {"90": [0.6388349564310538, 0.6775372411164946], "95": [0.6347700739178536, 0.6817724596099083]}, "micro/recall": 0.6222106901920083, "micro/precision": 0.6999416228838296, "macro/f1": 0.6182374585427982, "macro/f1_ci": {"90": [0.5956469999846385, 0.6396562173710457], "95": [0.5924717479435003, 0.643052479903031]}, "macro/recall": 0.5865594344408018, "macro/precision": 0.6571485734047059, "per_entity_metric": {"corporation": {"f1": 0.5837837837837839, "f1_ci": {"90": [0.5257238173647558, 0.6340269047020367], "95": [0.5147848132574822, 0.6415194434712977]}, "precision": 0.6033519553072626, "recall": 0.5654450261780105}, "creative_work": {"f1": 0.5464788732394367, "f1_ci": {"90": [0.4858406810117715, 0.6012249962680998], "95": [0.4752830687830688, 0.6118356854309848]}, "precision": 0.5511363636363636, "recall": 0.5418994413407822}, "event": {"f1": 0.4612244897959184, "f1_ci": {"90": [0.40925298655360004, 0.5121951219512195], "95": [0.4, 0.5224982057102179]}, "precision": 0.5022222222222222, "recall": 0.42641509433962266}, "group": {"f1": 0.5639097744360902, "f1_ci": {"90": [0.5085325152068667, 0.6129582440493426], "95": [0.4999765478424015, 0.6242067115955933]}, "precision": 0.6787330316742082, "recall": 0.48231511254019294}, "location": {"f1": 0.6666666666666666, "f1_ci": {"90": [0.5992845117845118, 0.725033185840708], "95": [0.5871509772636617, 0.7378266550522647]}, "precision": 0.6792452830188679, "recall": 0.6545454545454545}, "person": {"f1": 0.8389319552110249, "f1_ci": {"90": [0.8120086435076491, 0.8618636353151934], "95": [0.8049340069594505, 0.8669009775259731]}, "precision": 0.8619469026548673, "recall": 0.8171140939597316}, "product": {"f1": 0.6666666666666667, "f1_ci": {"90": [0.6168948944597402, 0.7127101210767604], "95": [0.604633351992265, 0.7208783153405305]}, "precision": 0.723404255319149, "recall": 0.6181818181818182}}}

eval/metric.test_2021.json

@@ -0,0 +1 @@
+{"micro/f1": 0.6641431520991053, "micro/f1_ci": {"90": [0.6551977421192867, 0.6726790034801573], "95": [0.6537478870999098, 0.6745822333244045]}, "micro/recall": 0.6695189639222942, "micro/precision": 0.6588529813381885, "macro/f1": 0.6165782134695219, "macro/f1_ci": {"90": [0.6067197256596341, 0.6256044510960733], "95": [0.6051019712561241, 0.6272725844616904]}, "macro/recall": 0.6256153624327598, "macro/precision": 0.6102975783874098, "per_entity_metric": {"corporation": {"f1": 0.5507246376811594, "f1_ci": {"90": [0.5254082699108996, 0.5773166181256275], "95": [0.5208478779438392, 0.5823103234936461]}, "precision": 0.5525727069351231, "recall": 0.5488888888888889}, "creative_work": {"f1": 0.4684914067472947, "f1_ci": {"90": [0.43769556974104906, 0.4987696892294594], "95": [0.4306276297645928, 0.503900885472692]}, "precision": 0.4380952380952381, "recall": 0.5034199726402189}, "event": {"f1": 0.4815724815724816, "f1_ci": {"90": [0.45748383888612637, 0.5051356400612352], "95": [0.453626352887228, 0.5078142035880663]}, "precision": 0.5235042735042735, "recall": 0.445859872611465}, "group": {"f1": 0.6143572621035058, "f1_ci": {"90": [0.5931312033684751, 0.6362758250713497], "95": [0.5898185538332443, 0.640990290362259]}, "precision": 0.6228842247799594, "recall": 0.6060606060606061}, "location": {"f1": 0.6886731391585761, "f1_ci": {"90": [0.6638351146228115, 0.7148122115942915], "95": [0.6584092995447562, 0.7185025217834956]}, "precision": 0.6417370325693607, "recall": 0.7430167597765364}, "person": {"f1": 0.8404178674351586, "f1_ci": {"90": [0.8298100715749012, 0.8504605055787255], "95": [0.828366934999076, 0.851877851877852]}, "precision": 0.8214788732394366, "recall": 0.8602507374631269}, "product": {"f1": 0.6718106995884774, "f1_ci": {"90": [0.65082156900923, 0.6922712449645116], "95": [0.6453161615534301, 0.6960348412597874]}, "precision": 0.6718106995884774, "recall": 0.6718106995884774}}}

eval/metric_span.test_2020.json

@@ -0,0 +1 @@
+{"micro/f1": 0.7641561297416162, "micro/f1_ci": {}, "micro/recall": 0.7213284898806435, "micro/precision": 0.8123904149620105, "macro/f1": 0.7641561297416162, "macro/f1_ci": {}, "macro/recall": 0.7213284898806435, "macro/precision": 0.8123904149620105}

eval/metric_span.test_2021.json

@@ -0,0 +1 @@
+{"micro/f1": 0.7896759392027531, "micro/f1_ci": {}, "micro/recall": 0.7961142592806754, "micro/precision": 0.783340919435594, "macro/f1": 0.7896759392027531, "macro/f1_ci": {}, "macro/recall": 0.7961142592806754, "macro/precision": 0.783340919435594}

trainer_config.json

@@ -1 +1 @@
-{"data_split": "2021.train", "model": "tner/bertweet-large-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": 0.15, "max_grad_norm": 1}
+{"dataset": ["tner/tweetner7"], "dataset_split": "train_2021", "dataset_name": null, "local_dataset": null, "model": "tner/bertweet-large-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": 0.15, "max_grad_norm": 1}