model update

README.md

@@ -0,0 +1,176 @@
+---
+datasets:
+- tner/tweetner7
+metrics:
+- f1
+- precision
+- recall
+model-index:
+- name: tner/roberta-base-tweetner7-2021
+  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: 0.6175553918513061
+    - name: Precision
+      type: precision
+      value: 0.6055117235248361
+    - name: Recall
+      type: recall
+      value: 0.6300878815911193
+    - name: F1 (macro)
+      type: f1_macro
+      value: 0.5699978267978356
+    - name: Precision (macro)
+      type: precision_macro
+      value: 0.5590687883112516
+    - name: Recall (macro)
+      type: recall_macro
+      value: 0.5841701622550579
+    - name: F1 (entity span)
+      type: f1_entity_span
+      value: 0.7692483045534848
+    - name: Precision (entity span)
+      type: precision_entity_span
+      value: 0.758314606741573
+    - name: Recall (entity span)
+      type: recall_entity_span
+      value: 0.7805019081762461
+  - 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: 0.604995882514411
+    - name: Precision
+      type: precision
+      value: 0.6421911421911422
+    - name: Recall
+      type: recall
+      value: 0.5718733783082511
+    - name: F1 (macro)
+      type: f1_macro
+      value: 0.5612313427645093
+    - name: Precision (macro)
+      type: precision_macro
+      value: 0.5945742705318462
+    - name: Recall (macro)
+      type: recall_macro
+      value: 0.5335324219465825
+    - name: F1 (entity span)
+      type: f1_entity_span
+      value: 0.7375103505382279
+    - name: Precision (entity span)
+      type: precision_entity_span
+      value: 0.7877358490566038
+    - name: Recall (entity span)
+      type: recall_entity_span
+      value: 0.6933056564608199
+
+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/roberta-base-tweetner7-2021
+
+This model is a fine-tuned version of [roberta-base](https://huggingface.co/roberta-base) on the 
+[tner/tweetner7](https://huggingface.co/datasets/tner/tweetner7) dataset (`train_2021` 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): 0.6175553918513061
+- Precision (micro): 0.6055117235248361
+- Recall (micro): 0.6300878815911193
+- F1 (macro): 0.5699978267978356
+- Precision (macro): 0.5590687883112516
+- Recall (macro): 0.5841701622550579
+
+
+
+The per-entity breakdown of the F1 score on the test set are below:
+- corporation: 0.4889844169801182
+- creative_work: 0.3800116211504939
+- event: 0.4550669216061186
+- group: 0.5702400000000001
+- location: 0.6506189821182944
+- person: 0.8133776792313377
+- product: 0.6316851664984864 
+
+For F1 scores, the confidence interval is obtained by bootstrap as below:
+- F1 (micro): 
+    - 90%: [0.6085886303107021, 0.6270245504530414]
+    - 95%: [0.6071586866067074, 0.6287436113239784] 
+- F1 (macro): 
+    - 90%: [0.6085886303107021, 0.6270245504530414]
+    - 95%: [0.6071586866067074, 0.6287436113239784] 
+
+Full evaluation can be found at [metric file of NER](https://huggingface.co/tner/roberta-base-tweetner7-2021/raw/main/eval/metric.json) 
+and [metric file of entity span](https://huggingface.co/tner/roberta-base-tweetner7-2021/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/roberta-base-tweetner7-2021")
+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: roberta-base
+ - crf: False
+ - 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
+
+The full configuration can be found at [fine-tuning parameter file](https://huggingface.co/tner/roberta-base-tweetner7-2021/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.test_2020.json

@@ -0,0 +1 @@
+{"micro/f1": 0.604995882514411, "micro/f1_ci": {"90": [0.5840748362685845, 0.6255164199352362], "95": [0.5803974312612263, 0.629783727608571]}, "micro/recall": 0.5718733783082511, "micro/precision": 0.6421911421911422, "macro/f1": 0.5612313427645093, "macro/f1_ci": {"90": [0.5392933879031333, 0.5843882413380433], "95": [0.5339723783579367, 0.5886659409585682]}, "macro/recall": 0.5335324219465825, "macro/precision": 0.5945742705318462, "per_entity_metric": {"corporation": {"f1": 0.498567335243553, "f1_ci": {"90": [0.4432367697038356, 0.5527248563561412], "95": [0.430937450036418, 0.5627151060479765]}, "precision": 0.5506329113924051, "recall": 0.45549738219895286}, "creative_work": {"f1": 0.45333333333333337, "f1_ci": {"90": [0.39745898298944954, 0.5064813035214726], "95": [0.3861220200232313, 0.5190568654646326]}, "precision": 0.4336734693877551, "recall": 0.4748603351955307}, "event": {"f1": 0.39828693790149894, "f1_ci": {"90": [0.3415730337078652, 0.456629460510739], "95": [0.3301304866742681, 0.4646474034020503]}, "precision": 0.4603960396039604, "recall": 0.35094339622641507}, "group": {"f1": 0.5281090289608178, "f1_ci": {"90": [0.48241536800880896, 0.577548236739413], "95": [0.4707699919291534, 0.5884499238915168]}, "precision": 0.5615942028985508, "recall": 0.4983922829581994}, "location": {"f1": 0.6095238095238096, "f1_ci": {"90": [0.5387194963046357, 0.6712802768166091], "95": [0.5269345238095238, 0.6860918021485609]}, "precision": 0.64, "recall": 0.5818181818181818}, "person": {"f1": 0.7992895204262878, "f1_ci": {"90": [0.7700797991571775, 0.8255599326386835], "95": [0.7649718398835049, 0.8306304806824568]}, "precision": 0.8490566037735849, "recall": 0.7550335570469798}, "product": {"f1": 0.6415094339622642, "f1_ci": {"90": [0.5941294695876195, 0.6870398266786277], "95": [0.5813202020202021, 0.6948054459658081]}, "precision": 0.6666666666666666, "recall": 0.6181818181818182}}}

eval/{metric.json → metric.test_2021.json}

@@ -1 +1 @@
-{"2021.dev": {"micro/f1": 0.6177062374245473, "micro/f1_ci": {}, "micro/recall": 0.614, "micro/precision": 0.6214574898785425, "macro/f1": 0.5772674113968216, "macro/f1_ci": {}, "macro/recall": 0.5791036438020786, "macro/precision": 0.5795388967201307, "per_entity_metric": {"corporation": {"f1": 0.5765765765765766, "f1_ci": {}, "precision": 0.5333333333333333, "recall": 0.6274509803921569}, "creative_work": {"f1": 0.4457831325301205, "f1_ci": {}, "precision": 0.40217391304347827, "recall": 0.5}, "event": {"f1": 0.3868312757201646, "f1_ci": {}, "precision": 0.41964285714285715, "recall": 0.35877862595419846}, "group": {"f1": 0.5887640449438203, "f1_ci": {}, "precision": 0.6009174311926605, "recall": 0.5770925110132159}, "location": {"f1": 0.6056338028169014, "f1_ci": {}, "precision": 0.6142857142857143, "recall": 0.5972222222222222}, "person": {"f1": 0.8057040998217468, "f1_ci": {}, "precision": 0.8129496402877698, "recall": 0.7985865724381626}, "product": {"f1": 0.6315789473684211, "f1_ci": {}, "precision": 0.673469387755102, "recall": 0.5945945945945946}}}, "2021.test": {"micro/f1": 0.6175553918513061, "micro/f1_ci": {"90": [0.6085886303107021, 0.6270245504530414], "95": [0.6071586866067074, 0.6287436113239784]}, "micro/recall": 0.6300878815911193, "micro/precision": 0.6055117235248361, "macro/f1": 0.5699978267978356, "macro/f1_ci": {"90": [0.5605132986273212, 0.5797067599839523], "95": [0.5588124392557708, 0.5811940101640996]}, "macro/recall": 0.5841701622550579, "macro/precision": 0.5590687883112516, "per_entity_metric": {"corporation": {"f1": 0.4889844169801182, "f1_ci": {"90": [0.4640154790729851, 0.515102305102305], "95": [0.458960569692277, 0.5184799847126035]}, "precision": 0.47346514047866806, "recall": 0.5055555555555555}, "creative_work": {"f1": 0.3800116211504939, "f1_ci": {"90": [0.35191852653953354, 0.40933711169196746], "95": [0.34682573875113965, 0.41311093888923844]}, "precision": 0.3303030303030303, "recall": 0.4473324213406293}, "event": {"f1": 0.4550669216061186, "f1_ci": {"90": [0.4322288263190161, 0.4776987543942918], "95": [0.4279424228309153, 0.4810260222978927]}, "precision": 0.47935548841893255, "recall": 0.43312101910828027}, "group": {"f1": 0.5702400000000001, "f1_ci": {"90": [0.5503755696583996, 0.591894081025951], "95": [0.5461277611277611, 0.5958909853906346]}, "precision": 0.5544492843808339, "recall": 0.5869565217391305}, "location": {"f1": 0.6506189821182944, "f1_ci": {"90": [0.6223673974540311, 0.678031547405726], "95": [0.6154868663693391, 0.6833961095707151]}, "precision": 0.6409214092140921, "recall": 0.6606145251396648}, "person": {"f1": 0.8133776792313377, "f1_ci": {"90": [0.802455084121242, 0.8241577387325482], "95": [0.8002960866175137, 0.8264256405395821]}, "precision": 0.8151851851851852, "recall": 0.8115781710914455}, "product": {"f1": 0.6316851664984864, "f1_ci": {"90": [0.6088324911643448, 0.6536741705155935], "95": [0.6046596973132213, 0.6580478190391984]}, "precision": 0.6198019801980198, "recall": 0.6440329218106996}}}, "2020.test": {"micro/f1": 0.604995882514411, "micro/f1_ci": {"90": [0.5840748362685845, 0.6255164199352362], "95": [0.5803974312612263, 0.629783727608571]}, "micro/recall": 0.5718733783082511, "micro/precision": 0.6421911421911422, "macro/f1": 0.5612313427645093, "macro/f1_ci": {"90": [0.5392933879031333, 0.5843882413380433], "95": [0.5339723783579367, 0.5886659409585682]}, "macro/recall": 0.5335324219465825, "macro/precision": 0.5945742705318462, "per_entity_metric": {"corporation": {"f1": 0.498567335243553, "f1_ci": {"90": [0.4432367697038356, 0.5527248563561412], "95": [0.430937450036418, 0.5627151060479765]}, "precision": 0.5506329113924051, "recall": 0.45549738219895286}, "creative_work": {"f1": 0.45333333333333337, "f1_ci": {"90": [0.39745898298944954, 0.5064813035214726], "95": [0.3861220200232313, 0.5190568654646326]}, "precision": 0.4336734693877551, "recall": 0.4748603351955307}, "event": {"f1": 0.39828693790149894, "f1_ci": {"90": [0.3415730337078652, 0.456629460510739], "95": [0.3301304866742681, 0.4646474034020503]}, "precision": 0.4603960396039604, "recall": 0.35094339622641507}, "group": {"f1": 0.5281090289608178, "f1_ci": {"90": [0.48241536800880896, 0.577548236739413], "95": [0.4707699919291534, 0.5884499238915168]}, "precision": 0.5615942028985508, "recall": 0.4983922829581994}, "location": {"f1": 0.6095238095238096, "f1_ci": {"90": [0.5387194963046357, 0.6712802768166091], "95": [0.5269345238095238, 0.6860918021485609]}, "precision": 0.64, "recall": 0.5818181818181818}, "person": {"f1": 0.7992895204262878, "f1_ci": {"90": [0.7700797991571775, 0.8255599326386835], "95": [0.7649718398835049, 0.8306304806824568]}, "precision": 0.8490566037735849, "recall": 0.7550335570469798}, "product": {"f1": 0.6415094339622642, "f1_ci": {"90": [0.5941294695876195, 0.6870398266786277], "95": [0.5813202020202021, 0.6948054459658081]}, "precision": 0.6666666666666666, "recall": 0.6181818181818182}}}, "2021.test (span detection)": {"micro/f1": 0.7692483045534848, "micro/f1_ci": {}, "micro/recall": 0.7805019081762461, "micro/precision": 0.758314606741573, "macro/f1": 0.7692483045534848, "macro/f1_ci": {}, "macro/recall": 0.7805019081762461, "macro/precision": 0.758314606741573}, "2020.test (span detection)": {"micro/f1": 0.7375103505382279, "micro/f1_ci": {}, "micro/recall": 0.6933056564608199, "micro/precision": 0.7877358490566038, "macro/f1": 0.7375103505382279, "macro/f1_ci": {}, "macro/recall": 0.6933056564608199, "macro/precision": 0.7877358490566038}}
+{"micro/f1": 0.6175553918513061, "micro/f1_ci": {"90": [0.6085886303107021, 0.6270245504530414], "95": [0.6071586866067074, 0.6287436113239784]}, "micro/recall": 0.6300878815911193, "micro/precision": 0.6055117235248361, "macro/f1": 0.5699978267978356, "macro/f1_ci": {"90": [0.5605132986273212, 0.5797067599839523], "95": [0.5588124392557708, 0.5811940101640996]}, "macro/recall": 0.5841701622550579, "macro/precision": 0.5590687883112516, "per_entity_metric": {"corporation": {"f1": 0.4889844169801182, "f1_ci": {"90": [0.4640154790729851, 0.515102305102305], "95": [0.458960569692277, 0.5184799847126035]}, "precision": 0.47346514047866806, "recall": 0.5055555555555555}, "creative_work": {"f1": 0.3800116211504939, "f1_ci": {"90": [0.35191852653953354, 0.40933711169196746], "95": [0.34682573875113965, 0.41311093888923844]}, "precision": 0.3303030303030303, "recall": 0.4473324213406293}, "event": {"f1": 0.4550669216061186, "f1_ci": {"90": [0.4322288263190161, 0.4776987543942918], "95": [0.4279424228309153, 0.4810260222978927]}, "precision": 0.47935548841893255, "recall": 0.43312101910828027}, "group": {"f1": 0.5702400000000001, "f1_ci": {"90": [0.5503755696583996, 0.591894081025951], "95": [0.5461277611277611, 0.5958909853906346]}, "precision": 0.5544492843808339, "recall": 0.5869565217391305}, "location": {"f1": 0.6506189821182944, "f1_ci": {"90": [0.6223673974540311, 0.678031547405726], "95": [0.6154868663693391, 0.6833961095707151]}, "precision": 0.6409214092140921, "recall": 0.6606145251396648}, "person": {"f1": 0.8133776792313377, "f1_ci": {"90": [0.802455084121242, 0.8241577387325482], "95": [0.8002960866175137, 0.8264256405395821]}, "precision": 0.8151851851851852, "recall": 0.8115781710914455}, "product": {"f1": 0.6316851664984864, "f1_ci": {"90": [0.6088324911643448, 0.6536741705155935], "95": [0.6046596973132213, 0.6580478190391984]}, "precision": 0.6198019801980198, "recall": 0.6440329218106996}}}

eval/metric_span.test_2020.json

@@ -0,0 +1 @@
+{"micro/f1": 0.7375103505382279, "micro/f1_ci": {}, "micro/recall": 0.6933056564608199, "micro/precision": 0.7877358490566038, "macro/f1": 0.7375103505382279, "macro/f1_ci": {}, "macro/recall": 0.6933056564608199, "macro/precision": 0.7877358490566038}

eval/metric_span.test_2021.json

@@ -0,0 +1 @@
+{"micro/f1": 0.7692483045534848, "micro/f1_ci": {}, "micro/recall": 0.7805019081762461, "micro/precision": 0.758314606741573, "macro/f1": 0.7692483045534848, "macro/f1_ci": {}, "macro/recall": 0.7805019081762461, "macro/precision": 0.758314606741573}

trainer_config.json

@@ -1 +1 @@
-{"data_split": "2021.train", "model": "roberta-base", "crf": false, "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}
+{"dataset": ["tner/tweetner7"], "dataset_split": "train_2021", "dataset_name": null, "local_dataset": null, "model": "roberta-base", "crf": false, "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}