model update

README.md

@@ -0,0 +1,135 @@
+---
+datasets:
+- tner/ttc
+metrics:
+- f1
+- precision
+- recall
+model-index:
+- name: tner/deberta-v3-large-ttc
+  results:
+  - task:
+      name: Token Classification
+      type: token-classification
+    dataset:
+      name: tner/ttc
+      type: tner/ttc
+      args: tner/ttc
+    metrics:
+    - name: F1
+      type: f1
+      value: 0.8266925817946227
+    - name: Precision
+      type: precision
+      value: 0.8264248704663213
+    - name: Recall
+      type: recall
+      value: 0.8269604666234608
+    - name: F1 (macro)
+      type: f1_macro
+      value: 0.8267742072572187
+    - name: Precision (macro)
+      type: precision_macro
+      value: 0.8278533291801137
+    - name: Recall (macro)
+      type: recall_macro
+      value: 0.8257668793195109
+    - name: F1 (entity span)
+      type: f1_entity_span
+      value: 0.8713961775186264
+    - name: Precision (entity span)
+      type: precision_entity_span
+      value: 0.8711139896373057
+    - name: Recall (entity span)
+      type: recall_entity_span
+      value: 0.8716785482825664
+
+pipeline_tag: token-classification
+widget:
+- text: "Jacob Collier is a Grammy awarded artist from England."
+  example_title: "NER Example 1"
+---
+# tner/deberta-v3-large-ttc
+
+This model is a fine-tuned version of [microsoft/deberta-v3-large](https://huggingface.co/microsoft/deberta-v3-large) on the 
+[tner/ttc](https://huggingface.co/datasets/tner/ttc) 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:
+- F1 (micro): 0.8266925817946227
+- Precision (micro): 0.8264248704663213
+- Recall (micro): 0.8269604666234608
+- F1 (macro): 0.8267742072572187
+- Precision (macro): 0.8278533291801137
+- Recall (macro): 0.8257668793195109
+
+The per-entity breakdown of the F1 score on the test set are below:
+- location: 0.7862266857962696
+- organization: 0.7770320656226697
+- person: 0.9170638703527169 
+
+For F1 scores, the confidence interval is obtained by bootstrap as below:
+- F1 (micro): 
+    - 90%: [0.8124223893760291, 0.8416139230675236]
+    - 95%: [0.8098712905029445, 0.8440240645643514] 
+- F1 (macro): 
+    - 90%: [0.8124223893760291, 0.8416139230675236]
+    - 95%: [0.8098712905029445, 0.8440240645643514] 
+
+Full evaluation can be found at [metric file of NER](https://huggingface.co/tner/deberta-v3-large-ttc/raw/main/eval/metric.json) 
+and [metric file of entity span](https://huggingface.co/tner/deberta-v3-large-ttc/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/deberta-v3-large-ttc")
+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/ttc']
+ - dataset_split: train
+ - dataset_name: None
+ - local_dataset: None
+ - model: microsoft/deberta-v3-large
+ - crf: True
+ - max_length: 128
+ - epoch: 15
+ - batch_size: 16
+ - lr: 1e-05
+ - random_seed: 42
+ - gradient_accumulation_steps: 4
+ - weight_decay: 1e-07
+ - lr_warmup_step_ratio: 0.1
+ - max_grad_norm: 10.0
+
+The full configuration can be found at [fine-tuning parameter file](https://huggingface.co/tner/deberta-v3-large-ttc/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.",
+}
+
+```

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "tner_ckpt/ttc_deberta_v3_large/best_model",
+  "_name_or_path": "microsoft/deberta-v3-large",
   "architectures": [
     "DebertaV2ForTokenClassification"
   ],

eval/metric.json

@@ -0,0 +1 @@
+{"micro/f1": 0.8266925817946227, "micro/f1_ci": {"90": [0.8124223893760291, 0.8416139230675236], "95": [0.8098712905029445, 0.8440240645643514]}, "micro/recall": 0.8269604666234608, "micro/precision": 0.8264248704663213, "macro/f1": 0.8267742072572187, "macro/f1_ci": {"90": [0.8124381534173242, 0.8415633373861329], "95": [0.8091911900905925, 0.845009292494298]}, "macro/recall": 0.8257668793195109, "macro/precision": 0.8278533291801137, "per_entity_metric": {"location": {"f1": 0.7862266857962696, "f1_ci": {"90": [0.7537985559174355, 0.8180537895612209], "95": [0.747742347424999, 0.8240248027848741]}, "precision": 0.7896253602305475, "recall": 0.7828571428571428}, "organization": {"f1": 0.7770320656226697, "f1_ci": {"90": [0.754573031135531, 0.7996870719423494], "95": [0.7482457440846576, 0.8046029849290348]}, "precision": 0.7707100591715976, "recall": 0.7834586466165413}, "person": {"f1": 0.9170638703527169, "f1_ci": {"90": [0.8987457734702097, 0.9367336323051129], "95": [0.8954628546358585, 0.940273320608835]}, "precision": 0.9232245681381958, "recall": 0.9109848484848485}}}

eval/metric_span.json

@@ -0,0 +1 @@
+{"micro/f1": 0.8713961775186264, "micro/f1_ci": {"90": [0.8592197178260848, 0.8844864406930985], "95": [0.8564147820469139, 0.8869424677617551]}, "micro/recall": 0.8716785482825664, "micro/precision": 0.8711139896373057, "macro/f1": 0.8713961775186264, "macro/f1_ci": {"90": [0.8592197178260848, 0.8844864406930985], "95": [0.8564147820469139, 0.8869424677617551]}, "macro/recall": 0.8716785482825664, "macro/precision": 0.8711139896373057}

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:90c2023335908809242c4b9207c65063d8f9cceca272df44d6876de33fb697d4
-size 1736209327
+oid sha256:a34eb4892bb85383e6dc32136557291dacfd2866545b271e324547a950b08a21
+size 1736214831

tokenizer_config.json

@@ -4,7 +4,7 @@
   "do_lower_case": false,
   "eos_token": "[SEP]",
   "mask_token": "[MASK]",
-  "name_or_path": "tner_ckpt/ttc_deberta_v3_large/best_model",
+  "name_or_path": "microsoft/deberta-v3-large",
   "pad_token": "[PAD]",
   "sep_token": "[SEP]",
   "sp_model_kwargs": {},

trainer_config.json

@@ -0,0 +1 @@
+{"dataset": ["tner/ttc"], "dataset_split": "train", "dataset_name": null, "local_dataset": null, "model": "microsoft/deberta-v3-large", "crf": true, "max_length": 128, "epoch": 15, "batch_size": 16, "lr": 1e-05, "random_seed": 42, "gradient_accumulation_steps": 4, "weight_decay": 1e-07, "lr_warmup_step_ratio": 0.1, "max_grad_norm": 10.0}