model update

README.md

@@ -0,0 +1,176 @@
+---
+datasets:
+- tner/tweetner7
+metrics:
+- f1
+- precision
+- recall
+model-index:
+- name: tner/twitter-roberta-base-2019-90m-tweetner7-2020
+  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: 0.6427956619039422
+    - name: Precision
+      type: precision
+      value: 0.63799977218362
+    - name: Recall
+      type: recall
+      value: 0.6476641998149861
+    - name: F1 (macro)
+      type: f1_macro
+      value: 0.5931418933396797
+    - name: Precision (macro)
+      type: precision_macro
+      value: 0.5885274267802955
+    - name: Recall (macro)
+      type: recall_macro
+      value: 0.6003736375632336
+    - name: F1 (entity span)
+      type: f1_entity_span
+      value: 0.778950992769425
+    - name: Precision (entity span)
+      type: precision_entity_span
+      value: 0.773094885522269
+    - name: Recall (entity span)
+      type: recall_entity_span
+      value: 0.7848964958945299
+  - 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: 0.6541700624830209
+    - name: Precision
+      type: precision
+      value: 0.6864310148232611
+    - name: Recall
+      type: recall
+      value: 0.6248053969901401
+    - name: F1 (macro)
+      type: f1_macro
+      value: 0.6111250287364248
+    - name: Precision (macro)
+      type: precision_macro
+      value: 0.6418894762960121
+    - name: Recall (macro)
+      type: recall_macro
+      value: 0.5881138886316531
+    - name: F1 (entity span)
+      type: f1_entity_span
+      value: 0.7655528389024722
+    - name: Precision (entity span)
+      type: precision_entity_span
+      value: 0.8033067274800456
+    - name: Recall (entity span)
+      type: recall_entity_span
+      value: 0.7311883757135443
+
+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-2019-90m-tweetner7-2020
+
+This model is a fine-tuned version of [cardiffnlp/twitter-roberta-base-2019-90m](https://huggingface.co/cardiffnlp/twitter-roberta-base-2019-90m) on the 
+[tner/tweetner7](https://huggingface.co/datasets/tner/tweetner7) dataset (`train_2020` 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): 0.6427956619039422
+- Precision (micro): 0.63799977218362
+- Recall (micro): 0.6476641998149861
+- F1 (macro): 0.5931418933396797
+- Precision (macro): 0.5885274267802955
+- Recall (macro): 0.6003736375632336
+
+
+
+The per-entity breakdown of the F1 score on the test set are below:
+- corporation: 0.48535564853556484
+- creative_work: 0.46893787575150303
+- event: 0.4369260512324794
+- group: 0.5908798972382787
+- location: 0.6701366297983083
+- person: 0.8399633363886344
+- product: 0.6597938144329897 
+
+For F1 scores, the confidence interval is obtained by bootstrap as below:
+- F1 (micro): 
+    - 90%: [0.6341597289758578, 0.6524372908527413]
+    - 95%: [0.6326184232151462, 0.6539625614316887] 
+- F1 (macro): 
+    - 90%: [0.6341597289758578, 0.6524372908527413]
+    - 95%: [0.6326184232151462, 0.6539625614316887] 
+
+Full evaluation can be found at [metric file of NER](https://huggingface.co/tner/twitter-roberta-base-2019-90m-tweetner7-2020/raw/main/eval/metric.json) 
+and [metric file of entity span](https://huggingface.co/tner/twitter-roberta-base-2019-90m-tweetner7-2020/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-2019-90m-tweetner7-2020")
+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_2020
+ - dataset_name: None
+ - local_dataset: None
+ - model: cardiffnlp/twitter-roberta-base-2019-90m
+ - 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/twitter-roberta-base-2019-90m-tweetner7-2020/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 @@
-{"2020.dev": {"micro/f1": 0.6341463414634146, "micro/f1_ci": {}, "micro/recall": 0.6044932079414838, "micro/precision": 0.6668587896253603, "macro/f1": 0.5725336564500102, "macro/f1_ci": {}, "macro/recall": 0.547299581336728, "macro/precision": 0.6045770051810889, "per_entity_metric": {"corporation": {"f1": 0.4741144414168937, "f1_ci": {}, "precision": 0.5304878048780488, "recall": 0.42857142857142855}, "creative_work": {"f1": 0.4736842105263158, "f1_ci": {}, "precision": 0.5232558139534884, "recall": 0.4326923076923077}, "event": {"f1": 0.3905579399141631, "f1_ci": {}, "precision": 0.43333333333333335, "recall": 0.35546875}, "group": {"f1": 0.5396825396825395, "f1_ci": {}, "precision": 0.5560747663551402, "recall": 0.5242290748898678}, "location": {"f1": 0.6077922077922078, "f1_ci": {}, "precision": 0.5735294117647058, "recall": 0.6464088397790055}, "person": {"f1": 0.8705281090289607, "f1_ci": {}, "precision": 0.8871527777777778, "recall": 0.8545150501672241}, "product": {"f1": 0.6513761467889908, "f1_ci": {}, "precision": 0.7282051282051282, "recall": 0.5892116182572614}}}, "2021.test": {"micro/f1": 0.6427956619039422, "micro/f1_ci": {"90": [0.6341597289758578, 0.6524372908527413], "95": [0.6326184232151462, 0.6539625614316887]}, "micro/recall": 0.6476641998149861, "micro/precision": 0.63799977218362, "macro/f1": 0.5931418933396797, "macro/f1_ci": {"90": [0.5829882769964541, 0.6029253957111215], "95": [0.581816657712503, 0.6041961940759853]}, "macro/recall": 0.6003736375632336, "macro/precision": 0.5885274267802955, "per_entity_metric": {"corporation": {"f1": 0.48535564853556484, "f1_ci": {"90": [0.46085441774110664, 0.5119725803305396], "95": [0.45520505216640084, 0.5171770970744771]}, "precision": 0.45849802371541504, "recall": 0.5155555555555555}, "creative_work": {"f1": 0.46893787575150303, "f1_ci": {"90": [0.4397129644242311, 0.49841951445170957], "95": [0.43447079440029895, 0.5046330558125194]}, "precision": 0.45822454308093996, "recall": 0.4801641586867305}, "event": {"f1": 0.4369260512324794, "f1_ci": {"90": [0.41453756735790565, 0.46084953195682354], "95": [0.4096712538810093, 0.46454062974413185]}, "precision": 0.465979381443299, "recall": 0.41128298453139217}, "group": {"f1": 0.5908798972382787, "f1_ci": {"90": [0.5695393272947947, 0.6123359251552747], "95": [0.5664627459133553, 0.6174932887262934]}, "precision": 0.5764411027568922, "recall": 0.6060606060606061}, "location": {"f1": 0.6701366297983083, "f1_ci": {"90": [0.6429253017545784, 0.6962477707064031], "95": [0.6379536594421498, 0.7007031745845312]}, "precision": 0.6272838002436053, "recall": 0.7192737430167597}, "person": {"f1": 0.8399633363886344, "f1_ci": {"90": [0.8290721665958414, 0.8508194267661195], "95": [0.8262450150763545, 0.8526792684560923]}, "precision": 0.8352169157856362, "recall": 0.8447640117994101}, "product": {"f1": 0.6597938144329897, "f1_ci": {"90": [0.6385839770940236, 0.6815940704640066], "95": [0.6339610512677912, 0.6852420327037805]}, "precision": 0.6980482204362801, "recall": 0.6255144032921811}}}, "2020.test": {"micro/f1": 0.6541700624830209, "micro/f1_ci": {"90": [0.635059380660795, 0.6729375721264049], "95": [0.6307634421882851, 0.6758834108900299]}, "micro/recall": 0.6248053969901401, "micro/precision": 0.6864310148232611, "macro/f1": 0.6111250287364248, "macro/f1_ci": {"90": [0.5905920126042988, 0.6311945655032225], "95": [0.5847070260170446, 0.6346146083812573]}, "macro/recall": 0.5881138886316531, "macro/precision": 0.6418894762960121, "per_entity_metric": {"corporation": {"f1": 0.5628140703517588, "f1_ci": {"90": [0.5040526283125426, 0.6157834502662088], "95": [0.49274048209056087, 0.6271466174858398]}, "precision": 0.5410628019323671, "recall": 0.5863874345549738}, "creative_work": {"f1": 0.5368731563421829, "f1_ci": {"90": [0.479085175656292, 0.5886075949367089], "95": [0.4668668548987869, 0.6]}, "precision": 0.56875, "recall": 0.5083798882681564}, "event": {"f1": 0.43388429752066116, "f1_ci": {"90": [0.383236701245796, 0.48343508343508346], "95": [0.37419282994467995, 0.4941783112677321]}, "precision": 0.4794520547945205, "recall": 0.39622641509433965}, "group": {"f1": 0.5622775800711743, "f1_ci": {"90": [0.5072004976892996, 0.6153846153846153], "95": [0.4948220528910802, 0.6231760502760004]}, "precision": 0.6294820717131474, "recall": 0.5080385852090032}, "location": {"f1": 0.64756446991404, "f1_ci": {"90": [0.5840220385674931, 0.7058823529411764], "95": [0.5744330305431932, 0.7183161845323438]}, "precision": 0.6141304347826086, "recall": 0.6848484848484848}, "person": {"f1": 0.8472821397756686, "f1_ci": {"90": [0.8225270104159926, 0.8691725550928349], "95": [0.818687006214862, 0.8734017382906808]}, "precision": 0.872113676731794, "recall": 0.8238255033557047}, "product": {"f1": 0.6871794871794871, "f1_ci": {"90": [0.6374695863746959, 0.7352148543737329], "95": [0.6271313488554868, 0.7454203021919558]}, "precision": 0.788235294117647, "recall": 0.6090909090909091}}}, "2021.test (span detection)": {"micro/f1": 0.778950992769425, "micro/f1_ci": {}, "micro/recall": 0.7848964958945299, "micro/precision": 0.773094885522269, "macro/f1": 0.778950992769425, "macro/f1_ci": {}, "macro/recall": 0.7848964958945299, "macro/precision": 0.773094885522269}, "2020.test (span detection)": {"micro/f1": 0.7655528389024722, "micro/f1_ci": {}, "micro/recall": 0.7311883757135443, "micro/precision": 0.8033067274800456, "macro/f1": 0.7655528389024722, "macro/f1_ci": {}, "macro/recall": 0.7311883757135443, "macro/precision": 0.8033067274800456}}

eval/metric.test_2020.json

@@ -0,0 +1 @@
+{"micro/f1": 0.6541700624830209, "micro/f1_ci": {"90": [0.635059380660795, 0.6729375721264049], "95": [0.6307634421882851, 0.6758834108900299]}, "micro/recall": 0.6248053969901401, "micro/precision": 0.6864310148232611, "macro/f1": 0.6111250287364248, "macro/f1_ci": {"90": [0.5905920126042988, 0.6311945655032225], "95": [0.5847070260170446, 0.6346146083812573]}, "macro/recall": 0.5881138886316531, "macro/precision": 0.6418894762960121, "per_entity_metric": {"corporation": {"f1": 0.5628140703517588, "f1_ci": {"90": [0.5040526283125426, 0.6157834502662088], "95": [0.49274048209056087, 0.6271466174858398]}, "precision": 0.5410628019323671, "recall": 0.5863874345549738}, "creative_work": {"f1": 0.5368731563421829, "f1_ci": {"90": [0.479085175656292, 0.5886075949367089], "95": [0.4668668548987869, 0.6]}, "precision": 0.56875, "recall": 0.5083798882681564}, "event": {"f1": 0.43388429752066116, "f1_ci": {"90": [0.383236701245796, 0.48343508343508346], "95": [0.37419282994467995, 0.4941783112677321]}, "precision": 0.4794520547945205, "recall": 0.39622641509433965}, "group": {"f1": 0.5622775800711743, "f1_ci": {"90": [0.5072004976892996, 0.6153846153846153], "95": [0.4948220528910802, 0.6231760502760004]}, "precision": 0.6294820717131474, "recall": 0.5080385852090032}, "location": {"f1": 0.64756446991404, "f1_ci": {"90": [0.5840220385674931, 0.7058823529411764], "95": [0.5744330305431932, 0.7183161845323438]}, "precision": 0.6141304347826086, "recall": 0.6848484848484848}, "person": {"f1": 0.8472821397756686, "f1_ci": {"90": [0.8225270104159926, 0.8691725550928349], "95": [0.818687006214862, 0.8734017382906808]}, "precision": 0.872113676731794, "recall": 0.8238255033557047}, "product": {"f1": 0.6871794871794871, "f1_ci": {"90": [0.6374695863746959, 0.7352148543737329], "95": [0.6271313488554868, 0.7454203021919558]}, "precision": 0.788235294117647, "recall": 0.6090909090909091}}}

eval/metric.test_2021.json

@@ -0,0 +1 @@
+{"micro/f1": 0.6427956619039422, "micro/f1_ci": {"90": [0.6341597289758578, 0.6524372908527413], "95": [0.6326184232151462, 0.6539625614316887]}, "micro/recall": 0.6476641998149861, "micro/precision": 0.63799977218362, "macro/f1": 0.5931418933396797, "macro/f1_ci": {"90": [0.5829882769964541, 0.6029253957111215], "95": [0.581816657712503, 0.6041961940759853]}, "macro/recall": 0.6003736375632336, "macro/precision": 0.5885274267802955, "per_entity_metric": {"corporation": {"f1": 0.48535564853556484, "f1_ci": {"90": [0.46085441774110664, 0.5119725803305396], "95": [0.45520505216640084, 0.5171770970744771]}, "precision": 0.45849802371541504, "recall": 0.5155555555555555}, "creative_work": {"f1": 0.46893787575150303, "f1_ci": {"90": [0.4397129644242311, 0.49841951445170957], "95": [0.43447079440029895, 0.5046330558125194]}, "precision": 0.45822454308093996, "recall": 0.4801641586867305}, "event": {"f1": 0.4369260512324794, "f1_ci": {"90": [0.41453756735790565, 0.46084953195682354], "95": [0.4096712538810093, 0.46454062974413185]}, "precision": 0.465979381443299, "recall": 0.41128298453139217}, "group": {"f1": 0.5908798972382787, "f1_ci": {"90": [0.5695393272947947, 0.6123359251552747], "95": [0.5664627459133553, 0.6174932887262934]}, "precision": 0.5764411027568922, "recall": 0.6060606060606061}, "location": {"f1": 0.6701366297983083, "f1_ci": {"90": [0.6429253017545784, 0.6962477707064031], "95": [0.6379536594421498, 0.7007031745845312]}, "precision": 0.6272838002436053, "recall": 0.7192737430167597}, "person": {"f1": 0.8399633363886344, "f1_ci": {"90": [0.8290721665958414, 0.8508194267661195], "95": [0.8262450150763545, 0.8526792684560923]}, "precision": 0.8352169157856362, "recall": 0.8447640117994101}, "product": {"f1": 0.6597938144329897, "f1_ci": {"90": [0.6385839770940236, 0.6815940704640066], "95": [0.6339610512677912, 0.6852420327037805]}, "precision": 0.6980482204362801, "recall": 0.6255144032921811}}}

eval/metric_span.test_2020.json

@@ -0,0 +1 @@
+{"micro/f1": 0.7655528389024722, "micro/f1_ci": {}, "micro/recall": 0.7311883757135443, "micro/precision": 0.8033067274800456, "macro/f1": 0.7655528389024722, "macro/f1_ci": {}, "macro/recall": 0.7311883757135443, "macro/precision": 0.8033067274800456}

eval/metric_span.test_2021.json

@@ -0,0 +1 @@
+{"micro/f1": 0.778950992769425, "micro/f1_ci": {}, "micro/recall": 0.7848964958945299, "micro/precision": 0.773094885522269, "macro/f1": 0.778950992769425, "macro/f1_ci": {}, "macro/recall": 0.7848964958945299, "macro/precision": 0.773094885522269}

trainer_config.json

@@ -1 +1 @@
-{"data_split": "2020.train", "model": "cardiffnlp/twitter-roberta-base-2019-90m", "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}
+{"dataset": ["tner/tweetner7"], "dataset_split": "train_2020", "dataset_name": null, "local_dataset": null, "model": "cardiffnlp/twitter-roberta-base-2019-90m", "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}