model update

README.md

@@ -0,0 +1,176 @@
+---
+datasets:
+- tner/tweetner7
+metrics:
+- f1
+- precision
+- recall
+model-index:
+- name: tner/twitter-roberta-base-dec2021-tweetner7-random
+  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: 0.6321284238886395
+    - name: Precision
+      type: precision
+      value: 0.6142015706806283
+    - name: Recall
+      type: recall
+      value: 0.6511332099907493
+    - name: F1 (macro)
+      type: f1_macro
+      value: 0.583682304736069
+    - name: Precision (macro)
+      type: precision_macro
+      value: 0.5654677691354458
+    - name: Recall (macro)
+      type: recall_macro
+      value: 0.6047150410746663
+    - name: F1 (entity span)
+      type: f1_entity_span
+      value: 0.7703620544484986
+    - name: Precision (entity span)
+      type: precision_entity_span
+      value: 0.7484729493891797
+    - name: Recall (entity span)
+      type: recall_entity_span
+      value: 0.7935700242858795
+  - 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: 0.6368775235531628
+    - name: Precision
+      type: precision
+      value: 0.6616331096196868
+    - name: Recall
+      type: recall
+      value: 0.6139076284379865
+    - name: F1 (macro)
+      type: f1_macro
+      value: 0.5976605759407211
+    - name: Precision (macro)
+      type: precision_macro
+      value: 0.6177069721428509
+    - name: Recall (macro)
+      type: recall_macro
+      value: 0.5812570646484104
+    - name: F1 (entity span)
+      type: f1_entity_span
+      value: 0.7542395693135936
+    - name: Precision (entity span)
+      type: precision_entity_span
+      value: 0.7835570469798657
+    - name: Recall (entity span)
+      type: recall_entity_span
+      value: 0.7270368448365335
+
+pipeline_tag: token-classification
+widget:
+- text: "Jacob Collier is a Grammy awarded artist from England."
+  example_title: "NER Example 1"
+---
+# tner/twitter-roberta-base-dec2021-tweetner7-random
+
+This model is a fine-tuned version of [cardiffnlp/twitter-roberta-base-dec2021](https://huggingface.co/cardiffnlp/twitter-roberta-base-dec2021) on the 
+[tner/tweetner7](https://huggingface.co/datasets/tner/tweetner7) dataset.
+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): 0.6321284238886395
+- Precision (micro): 0.6142015706806283
+- Recall (micro): 0.6511332099907493
+- F1 (macro): 0.583682304736069
+- Precision (macro): 0.5654677691354458
+- Recall (macro): 0.6047150410746663
+
+
+
+The per-entity breakdown of the F1 score on the test set are below:
+- corporation: 0.5019685039370079
+- creative_work: 0.41401273885350315
+- event: 0.4564727108705458
+- group: 0.5892444737710327
+- location: 0.6486486486486486
+- person: 0.8268075031870332
+- product: 0.6486215538847118 
+
+For F1 scores, the confidence interval is obtained by bootstrap as below:
+- F1 (micro): 
+    - 90%: [0.6245116881258609, 0.6411928894306437]
+    - 95%: [0.6221686986039963, 0.642603475030015] 
+- F1 (macro): 
+    - 90%: [0.6245116881258609, 0.6411928894306437]
+    - 95%: [0.6221686986039963, 0.642603475030015] 
+
+Full evaluation can be found at [metric file of NER](https://huggingface.co/tner/twitter-roberta-base-dec2021-tweetner7-random/raw/main/eval/metric.json) 
+and [metric file of entity span](https://huggingface.co/tner/twitter-roberta-base-dec2021-tweetner7-random/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-dec2021-tweetner7-random")
+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_random
+ - dataset_name: None
+ - local_dataset: None
+ - model: cardiffnlp/twitter-roberta-base-dec2021
+ - 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
+
+The full configuration can be found at [fine-tuning parameter file](https://huggingface.co/tner/twitter-roberta-base-dec2021-tweetner7-random/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 @@
-{"random.dev": {"micro/f1": 0.6390532544378698, "micro/f1_ci": {}, "micro/recall": 0.6342765616657768, "micro/precision": 0.6439024390243903, "macro/f1": 0.5948376345756554, "macro/f1_ci": {}, "macro/recall": 0.5917630600682918, "macro/precision": 0.5994147209089418, "per_entity_metric": {"corporation": {"f1": 0.5817307692307693, "f1_ci": {}, "precision": 0.5426008968609866, "recall": 0.6269430051813472}, "creative_work": {"f1": 0.4888888888888889, "f1_ci": {}, "precision": 0.5032679738562091, "recall": 0.47530864197530864}, "event": {"f1": 0.3907563025210084, "f1_ci": {}, "precision": 0.4025974025974026, "recall": 0.3795918367346939}, "group": {"f1": 0.6050670640834576, "f1_ci": {}, "precision": 0.6246153846153846, "recall": 0.5867052023121387}, "location": {"f1": 0.6191950464396285, "f1_ci": {}, "precision": 0.625, "recall": 0.6134969325153374}, "person": {"f1": 0.84593837535014, "f1_ci": {}, "precision": 0.8420074349442379, "recall": 0.849906191369606}, "product": {"f1": 0.632286995515695, "f1_ci": {}, "precision": 0.6558139534883721, "recall": 0.6103896103896104}}}, "2021.test": {"micro/f1": 0.6321284238886395, "micro/f1_ci": {"90": [0.6245116881258609, 0.6411928894306437], "95": [0.6221686986039963, 0.642603475030015]}, "micro/recall": 0.6511332099907493, "micro/precision": 0.6142015706806283, "macro/f1": 0.583682304736069, "macro/f1_ci": {"90": [0.5745215695064868, 0.5935538510061642], "95": [0.5721210274985685, 0.5949826139031175]}, "macro/recall": 0.6047150410746663, "macro/precision": 0.5654677691354458, "per_entity_metric": {"corporation": {"f1": 0.5019685039370079, "f1_ci": {"90": [0.4775046763935653, 0.5276263097468425], "95": [0.4708860124040466, 0.5322353978734399]}, "precision": 0.450530035335689, "recall": 0.5666666666666667}, "creative_work": {"f1": 0.41401273885350315, "f1_ci": {"90": [0.3838081951821538, 0.446030537527557], "95": [0.3785096669891205, 0.4516264083963575]}, "precision": 0.3873659117997616, "recall": 0.44459644322845415}, "event": {"f1": 0.4564727108705458, "f1_ci": {"90": [0.43317409571140686, 0.47845101312231], "95": [0.42696327577105797, 0.48275945643271434]}, "precision": 0.4525939177101968, "recall": 0.4604185623293904}, "group": {"f1": 0.5892444737710327, "f1_ci": {"90": [0.5689646859309015, 0.6108478406183305], "95": [0.564631734245271, 0.6146674362079695]}, "precision": 0.5902181097157965, "recall": 0.5882740447957839}, "location": {"f1": 0.6486486486486486, "f1_ci": {"90": [0.6209685619384042, 0.6762750470285591], "95": [0.6156787330316741, 0.6811199553135038]}, "precision": 0.6282722513089005, "recall": 0.6703910614525139}, "person": {"f1": 0.8268075031870332, "f1_ci": {"90": [0.8166678815298463, 0.8379175120851383], "95": [0.8152424612453091, 0.8395018891964325]}, "precision": 0.8168405901403383, "recall": 0.8370206489675516}, "product": {"f1": 0.6486215538847118, "f1_ci": {"90": [0.6270291403273055, 0.6710524317515882], "95": [0.6216769714725208, 0.674748088790642]}, "precision": 0.6324535679374389, "recall": 0.6656378600823045}}}, "2020.test": {"micro/f1": 0.6368775235531628, "micro/f1_ci": {"90": [0.6169064115810362, 0.6559384186705487], "95": [0.6141058003386577, 0.6594030324221372]}, "micro/recall": 0.6139076284379865, "micro/precision": 0.6616331096196868, "macro/f1": 0.5976605759407211, "macro/f1_ci": {"90": [0.5759632782880849, 0.6166251611368425], "95": [0.5736091476367275, 0.620115150872923]}, "macro/recall": 0.5812570646484104, "macro/precision": 0.6177069721428509, "per_entity_metric": {"corporation": {"f1": 0.56, "f1_ci": {"90": [0.5013616633738992, 0.6116487952836637], "95": [0.4910296600700221, 0.6227328614008941]}, "precision": 0.5358851674641149, "recall": 0.5863874345549738}, "creative_work": {"f1": 0.5166666666666666, "f1_ci": {"90": [0.46151629350476603, 0.5722086282622287], "95": [0.4514431946006749, 0.5835841303677111]}, "precision": 0.5138121546961326, "recall": 0.5195530726256983}, "event": {"f1": 0.4137931034482759, "f1_ci": {"90": [0.3618226809356626, 0.46421880458537457], "95": [0.35246839778317457, 0.47369129736369653]}, "precision": 0.4473684210526316, "recall": 0.3849056603773585}, "group": {"f1": 0.5693950177935941, "f1_ci": {"90": [0.5249356080118784, 0.6172953291232861], "95": [0.515244168484645, 0.6276404963608163]}, "precision": 0.6374501992031872, "recall": 0.5144694533762058}, "location": {"f1": 0.6747720364741642, "f1_ci": {"90": [0.6173401259613639, 0.7267984140530545], "95": [0.604807000719252, 0.7348546893498322]}, "precision": 0.676829268292683, "recall": 0.6727272727272727}, "person": {"f1": 0.8222029488291414, "f1_ci": {"90": [0.7938635459282011, 0.846929422548121], "95": [0.7895076867859065, 0.851171304347826]}, "precision": 0.8509874326750448, "recall": 0.7953020134228188}, "product": {"f1": 0.6267942583732058, "f1_ci": {"90": [0.5756898281700741, 0.6747046867175303], "95": [0.5657433380084151, 0.6810691357228119]}, "precision": 0.6616161616161617, "recall": 0.5954545454545455}}}, "2021.test (span detection)": {"micro/f1": 0.7703620544484986, "micro/f1_ci": {}, "micro/recall": 0.7935700242858795, "micro/precision": 0.7484729493891797, "macro/f1": 0.7703620544484986, "macro/f1_ci": {}, "macro/recall": 0.7935700242858795, "macro/precision": 0.7484729493891797}, "2020.test (span detection)": {"micro/f1": 0.7542395693135936, "micro/f1_ci": {}, "micro/recall": 0.7270368448365335, "micro/precision": 0.7835570469798657, "macro/f1": 0.7542395693135936, "macro/f1_ci": {}, "macro/recall": 0.7270368448365335, "macro/precision": 0.7835570469798657}}

eval/metric.test_2020.json

@@ -0,0 +1 @@
+{"micro/f1": 0.6368775235531628, "micro/f1_ci": {"90": [0.6169064115810362, 0.6559384186705487], "95": [0.6141058003386577, 0.6594030324221372]}, "micro/recall": 0.6139076284379865, "micro/precision": 0.6616331096196868, "macro/f1": 0.5976605759407211, "macro/f1_ci": {"90": [0.5759632782880849, 0.6166251611368425], "95": [0.5736091476367275, 0.620115150872923]}, "macro/recall": 0.5812570646484104, "macro/precision": 0.6177069721428509, "per_entity_metric": {"corporation": {"f1": 0.56, "f1_ci": {"90": [0.5013616633738992, 0.6116487952836637], "95": [0.4910296600700221, 0.6227328614008941]}, "precision": 0.5358851674641149, "recall": 0.5863874345549738}, "creative_work": {"f1": 0.5166666666666666, "f1_ci": {"90": [0.46151629350476603, 0.5722086282622287], "95": [0.4514431946006749, 0.5835841303677111]}, "precision": 0.5138121546961326, "recall": 0.5195530726256983}, "event": {"f1": 0.4137931034482759, "f1_ci": {"90": [0.3618226809356626, 0.46421880458537457], "95": [0.35246839778317457, 0.47369129736369653]}, "precision": 0.4473684210526316, "recall": 0.3849056603773585}, "group": {"f1": 0.5693950177935941, "f1_ci": {"90": [0.5249356080118784, 0.6172953291232861], "95": [0.515244168484645, 0.6276404963608163]}, "precision": 0.6374501992031872, "recall": 0.5144694533762058}, "location": {"f1": 0.6747720364741642, "f1_ci": {"90": [0.6173401259613639, 0.7267984140530545], "95": [0.604807000719252, 0.7348546893498322]}, "precision": 0.676829268292683, "recall": 0.6727272727272727}, "person": {"f1": 0.8222029488291414, "f1_ci": {"90": [0.7938635459282011, 0.846929422548121], "95": [0.7895076867859065, 0.851171304347826]}, "precision": 0.8509874326750448, "recall": 0.7953020134228188}, "product": {"f1": 0.6267942583732058, "f1_ci": {"90": [0.5756898281700741, 0.6747046867175303], "95": [0.5657433380084151, 0.6810691357228119]}, "precision": 0.6616161616161617, "recall": 0.5954545454545455}}}

eval/metric.test_2021.json

@@ -0,0 +1 @@
+{"micro/f1": 0.6321284238886395, "micro/f1_ci": {"90": [0.6245116881258609, 0.6411928894306437], "95": [0.6221686986039963, 0.642603475030015]}, "micro/recall": 0.6511332099907493, "micro/precision": 0.6142015706806283, "macro/f1": 0.583682304736069, "macro/f1_ci": {"90": [0.5745215695064868, 0.5935538510061642], "95": [0.5721210274985685, 0.5949826139031175]}, "macro/recall": 0.6047150410746663, "macro/precision": 0.5654677691354458, "per_entity_metric": {"corporation": {"f1": 0.5019685039370079, "f1_ci": {"90": [0.4775046763935653, 0.5276263097468425], "95": [0.4708860124040466, 0.5322353978734399]}, "precision": 0.450530035335689, "recall": 0.5666666666666667}, "creative_work": {"f1": 0.41401273885350315, "f1_ci": {"90": [0.3838081951821538, 0.446030537527557], "95": [0.3785096669891205, 0.4516264083963575]}, "precision": 0.3873659117997616, "recall": 0.44459644322845415}, "event": {"f1": 0.4564727108705458, "f1_ci": {"90": [0.43317409571140686, 0.47845101312231], "95": [0.42696327577105797, 0.48275945643271434]}, "precision": 0.4525939177101968, "recall": 0.4604185623293904}, "group": {"f1": 0.5892444737710327, "f1_ci": {"90": [0.5689646859309015, 0.6108478406183305], "95": [0.564631734245271, 0.6146674362079695]}, "precision": 0.5902181097157965, "recall": 0.5882740447957839}, "location": {"f1": 0.6486486486486486, "f1_ci": {"90": [0.6209685619384042, 0.6762750470285591], "95": [0.6156787330316741, 0.6811199553135038]}, "precision": 0.6282722513089005, "recall": 0.6703910614525139}, "person": {"f1": 0.8268075031870332, "f1_ci": {"90": [0.8166678815298463, 0.8379175120851383], "95": [0.8152424612453091, 0.8395018891964325]}, "precision": 0.8168405901403383, "recall": 0.8370206489675516}, "product": {"f1": 0.6486215538847118, "f1_ci": {"90": [0.6270291403273055, 0.6710524317515882], "95": [0.6216769714725208, 0.674748088790642]}, "precision": 0.6324535679374389, "recall": 0.6656378600823045}}}

eval/metric_span.test_2020.json

@@ -0,0 +1 @@
+{"micro/f1": 0.7542395693135936, "micro/f1_ci": {}, "micro/recall": 0.7270368448365335, "micro/precision": 0.7835570469798657, "macro/f1": 0.7542395693135936, "macro/f1_ci": {}, "macro/recall": 0.7270368448365335, "macro/precision": 0.7835570469798657}

eval/metric_span.test_2021.json

@@ -0,0 +1 @@
+{"micro/f1": 0.7703620544484986, "micro/f1_ci": {}, "micro/recall": 0.7935700242858795, "micro/precision": 0.7484729493891797, "macro/f1": 0.7703620544484986, "macro/f1_ci": {}, "macro/recall": 0.7935700242858795, "macro/precision": 0.7484729493891797}

trainer_config.json

@@ -1 +1 @@
-{"data_split": "random.train", "model": "cardiffnlp/twitter-roberta-base-dec2021", "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}
+{"dataset": ["tner/tweetner7"], "dataset_split": "train_random", "dataset_name": null, "local_dataset": null, "model": "cardiffnlp/twitter-roberta-base-dec2021", "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}