model update

README.md

@@ -0,0 +1,137 @@
+---
+datasets:
+- bionlp2004
+metrics:
+- f1
+- precision
+- recall
+model-index:
+- name: tner/deberta-v3-large-bionlp2004
+  results:
+  - task:
+      name: Token Classification
+      type: token-classification
+    dataset:
+      name: bionlp2004
+      type: bionlp2004
+      args: bionlp2004
+    metrics:
+    - name: F1
+      type: f1
+      value: 0.758624442267929
+    - name: Precision
+      type: precision
+      value: 0.7174763277068753
+    - name: Recall
+      type: recall
+      value: 0.8047794966520434
+    - name: F1 (macro)
+      type: f1_macro
+      value: 0.7195387988303987
+    - name: Precision (macro)
+      type: precision_macro
+      value: 0.681309505763584
+    - name: Recall (macro)
+      type: recall_macro
+      value: 0.7691804743892025
+    - name: F1 (entity span)
+      type: f1_entity_span
+      value: 0.796539152201121
+    - name: Precision (entity span)
+      type: precision_entity_span
+      value: 0.7533710756562018
+    - name: Recall (entity span)
+      type: recall_entity_span
+      value: 0.8449549757561764
+
+pipeline_tag: token-classification
+widget:
+- text: "Jacob Collier is a Grammy awarded artist from England."
+  example_title: "NER Example 1"
+---
+# tner/deberta-v3-large-bionlp2004
+
+This model is a fine-tuned version of [microsoft/deberta-v3-large](https://huggingface.co/microsoft/deberta-v3-large) on the 
+[tner/bionlp2004](https://huggingface.co/datasets/tner/bionlp2004) 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): 0.758624442267929
+- Precision (micro): 0.7174763277068753
+- Recall (micro): 0.8047794966520434
+- F1 (macro): 0.7195387988303987
+- Precision (macro): 0.681309505763584
+- Recall (macro): 0.7691804743892025
+
+The per-entity breakdown of the F1 score on the test set are below:
+- cell_line: 0.6465517241379309
+- cell_type: 0.7562483203439935
+- dna: 0.7449506810709253
+- protein: 0.7757859652283577
+- rna: 0.6741573033707865 
+
+For F1 scores, the confidence interval is obtained by bootstrap as below:
+- F1 (micro): 
+    - 90%: [0.7508836679942893, 0.7667327003308145]
+    - 95%: [0.7498144548458301, 0.7680807868080707] 
+- F1 (macro): 
+    - 90%: [0.7508836679942893, 0.7667327003308145]
+    - 95%: [0.7498144548458301, 0.7680807868080707] 
+
+Full evaluation can be found at [metric file of NER](https://huggingface.co/tner/deberta-v3-large-bionlp2004/raw/main/eval/metric.json) 
+and [metric file of entity span](https://huggingface.co/tner/deberta-v3-large-bionlp2004/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-bionlp2004")
+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/bionlp2004']
+ - 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: 8
+ - weight_decay: 1e-07
+ - lr_warmup_step_ratio: 0.1
+ - max_grad_norm: None
+
+The full configuration can be found at [fine-tuning parameter file](https://huggingface.co/tner/deberta-v3-large-bionlp2004/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/bionlp2004_deberta_v3_large/best_model",
+  "_name_or_path": "tner_ckpt/bionlp2004_deberta_v3_large/model_cghqta/epoch_5",
   "architectures": [
     "DebertaV2ForTokenClassification"
   ],

eval/metric.json

@@ -0,0 +1 @@
+{"micro/f1": 0.758624442267929, "micro/f1_ci": {"90": [0.7508836679942893, 0.7667327003308145], "95": [0.7498144548458301, 0.7680807868080707]}, "micro/recall": 0.8047794966520434, "micro/precision": 0.7174763277068753, "macro/f1": 0.7195387988303987, "macro/f1_ci": {"90": [0.704160346303814, 0.734315970324062], "95": [0.7011867069041885, 0.7369915261120946]}, "macro/recall": 0.7691804743892025, "macro/precision": 0.681309505763584, "per_entity_metric": {"cell_line": {"f1": 0.6465517241379309, "f1_ci": {"90": [0.6089615709080467, 0.681785134696527], "95": [0.5990760017778193, 0.6864572985282207]}, "precision": 0.5681818181818182, "recall": 0.75}, "cell_type": {"f1": 0.7562483203439935, "f1_ci": {"90": [0.7410739312769105, 0.771535982706181], "95": [0.738658976728006, 0.7751383966368984]}, "precision": 0.7816666666666666, "recall": 0.7324310255075481}, "dna": {"f1": 0.7449506810709253, "f1_ci": {"90": [0.7242533665740931, 0.7645011134479772], "95": [0.7196239737811436, 0.766789450954474]}, "precision": 0.739049394221808, "recall": 0.7509469696969697}, "protein": {"f1": 0.7757859652283577, "f1_ci": {"90": [0.766777076592289, 0.7852282635428913], "95": [0.7643832732252224, 0.7869697999538529]}, "precision": 0.7136228041100431, "recall": 0.8498125123347148}, "rna": {"f1": 0.6741573033707865, "f1_ci": {"90": [0.6153561253561253, 0.7310969549109966], "95": [0.6008369098712445, 0.7422877290950747]}, "precision": 0.6040268456375839, "recall": 0.7627118644067796}}}

eval/metric_span.json

@@ -0,0 +1 @@
+{"micro/f1": 0.796539152201121, "micro/f1_ci": {"90": [0.7896444164716317, 0.8039723864322159], "95": [0.7886216488815425, 0.8053749479886999]}, "micro/recall": 0.8449549757561764, "micro/precision": 0.7533710756562018, "macro/f1": 0.796539152201121, "macro/f1_ci": {"90": [0.7896444164716317, 0.8039723864322159], "95": [0.7886216488815425, 0.8053749479886999]}, "macro/recall": 0.8449549757561764, "macro/precision": 0.7533710756562018}

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cf80a5659925b9c2ee6c1edf422f492c55032cde3599ec38f6e7d05e266bc68a
-size 1736225711
+oid sha256:c759ae61f9ee75129e6c3eb9cdba9e8fcb2840ce7bc4cf0357c64f317eb9695f
+size 1736231215

tokenizer_config.json

@@ -4,7 +4,7 @@
   "do_lower_case": false,
   "eos_token": "[SEP]",
   "mask_token": "[MASK]",
-  "name_or_path": "tner_ckpt/bionlp2004_deberta_v3_large/best_model",
+  "name_or_path": "tner_ckpt/bionlp2004_deberta_v3_large/model_cghqta/epoch_5",
   "pad_token": "[PAD]",
   "sep_token": "[SEP]",
   "sp_model_kwargs": {},

trainer_config.json

@@ -0,0 +1 @@
+{"dataset": ["tner/bionlp2004"], "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": 8, "weight_decay": 1e-07, "lr_warmup_step_ratio": 0.1, "max_grad_norm": null}