model update

README.md

@@ -0,0 +1,123 @@
+---
+datasets:
+- btc
+metrics:
+- f1
+- precision
+- recall
+model-index:
+- name: tner/roberta-large-btc
+  results:
+  - task:
+      name: Token Classification
+      type: token-classification
+    dataset:
+      name: btc
+      type: btc
+      args: btc
+    metrics:
+    - name: F1
+      type: f1
+      value: 0.8367557645979121
+    - name: Precision
+      type: precision
+      value: 0.8401290025339784
+    - name: Recall
+      type: recall
+      value: 0.8334095063985375
+    - name: F1 (macro)
+      type: f1_macro
+      value: 0.7830389304099722
+    - name: Precision (macro)
+      type: precision_macro
+      value: 0.7911560677795398
+    - name: Recall (macro)
+      type: recall_macro
+      value: 0.7756024849498971
+    - name: F1 (entity span)
+      type: f1_entity_span
+      value: 0.9113227027647126
+    - name: Precision (entity span)
+      type: precision_entity_span
+      value: 0.9149965445749827
+    - name: Recall (entity span)
+      type: recall_entity_span
+      value: 0.9076782449725777
+
+pipeline_tag: token-classification
+widget:
+- text: "Jacob Collier is a Grammy awarded artist from England."
+  example_title: "NER Example 1"
+---
+# tner/roberta-large-btc
+
+This model is a fine-tuned version of [roberta-large](https://huggingface.co/roberta-large) on the 
+[tner/btc](https://huggingface.co/datasets/tner/btc) 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): 0.8367557645979121
+- Precision (micro): 0.8401290025339784
+- Recall (micro): 0.8334095063985375
+- F1 (macro): 0.7830389304099722
+- Precision (macro): 0.7911560677795398
+- Recall (macro): 0.7756024849498971
+
+The per-entity breakdown of the F1 score on the test set are below:
+- location: 0.736756316218419
+- organization: 0.6927985414767548
+- person: 0.9195619335347431 
+
+For F1 scores, the confidence interval is obtained by bootstrap as below:
+- F1 (micro): 
+    - 90%: [0.8263755823738717, 0.8472678708881698]
+    - 95%: [0.8238362631404713, 0.8498613485265176] 
+- F1 (macro): 
+    - 90%: [0.8263755823738717, 0.8472678708881698]
+    - 95%: [0.8238362631404713, 0.8498613485265176] 
+
+Full evaluation can be found at [metric file of NER](https://huggingface.co/tner/roberta-large-btc/raw/main/eval/metric.json) 
+and [metric file of entity span](https://huggingface.co/tner/roberta-large-btc/raw/main/eval/metric_span.json).
+
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+ - dataset: ['tner/btc']
+ - dataset_split: train
+ - dataset_name: None
+ - local_dataset: None
+ - model: roberta-large
+ - crf: True
+ - max_length: 128
+ - epoch: 15
+ - batch_size: 64
+ - lr: 1e-05
+ - random_seed: 42
+ - gradient_accumulation_steps: 2
+ - 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/roberta-large-btc/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/btc_roberta_large/best_model",
+  "_name_or_path": "tner_ckpt/btc_roberta_large/model_cghqta/epoch_5",
   "architectures": [
     "RobertaForTokenClassification"
   ],

eval/metric.json

@@ -0,0 +1 @@
+{"micro/f1": 0.8367557645979121, "micro/f1_ci": {"90": [0.8263755823738717, 0.8472678708881698], "95": [0.8238362631404713, 0.8498613485265176]}, "micro/recall": 0.8334095063985375, "micro/precision": 0.8401290025339784, "macro/f1": 0.7830389304099722, "macro/f1_ci": {"90": [0.7696091150716922, 0.7967192841541696], "95": [0.7660108860238721, 0.7998833434490498]}, "macro/recall": 0.7756024849498971, "macro/precision": 0.7911560677795398, "per_entity_metric": {"location": {"f1": 0.736756316218419, "f1_ci": {"90": [0.7113860430705887, 0.7627252046055419], "95": [0.7051230501472048, 0.7673361281231106]}, "precision": 0.7648054145516074, "recall": 0.710691823899371}, "organization": {"f1": 0.6927985414767548, "f1_ci": {"90": [0.6678631913277484, 0.716819379942262], "95": [0.6645546888239872, 0.7213803488241352]}, "precision": 0.6884057971014492, "recall": 0.6972477064220184}, "person": {"f1": 0.9195619335347431, "f1_ci": {"90": [0.9104693517194988, 0.9284883241222021], "95": [0.9089856318984997, 0.9301216516668346]}, "precision": 0.9202569916855631, "recall": 0.9188679245283019}}}

eval/metric_span.json

@@ -0,0 +1 @@
+{"micro/f1": 0.9113227027647126, "micro/f1_ci": {"90": [0.9044179638988344, 0.9176975890486316], "95": [0.902687742764797, 0.9191195236935528]}, "micro/recall": 0.9076782449725777, "micro/precision": 0.9149965445749827, "macro/f1": 0.9113227027647126, "macro/f1_ci": {"90": [0.9044179638988344, 0.9176975890486316], "95": [0.902687742764797, 0.9191195236935528]}, "macro/recall": 0.9076782449725777, "macro/precision": 0.9149965445749827}

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7d09155c297e6c1193d96398366bb54cbc734f09ea44c80aad6e320840c37409
-size 1417400369
+oid sha256:065f2a6482fdb2615f5ca0f25c077ba12328882aba50137814b2aec25f5eab4c
+size 1417405809

tokenizer_config.json

@@ -6,7 +6,7 @@
   "errors": "replace",
   "mask_token": "<mask>",
   "model_max_length": 512,
-  "name_or_path": "tner_ckpt/btc_roberta_large/best_model",
+  "name_or_path": "tner_ckpt/btc_roberta_large/model_cghqta/epoch_5",
   "pad_token": "<pad>",
   "sep_token": "</s>",
   "special_tokens_map_file": "tner_ckpt/btc_roberta_large/model_cghqta/epoch_5/special_tokens_map.json",

trainer_config.json

@@ -0,0 +1 @@
+{"dataset": ["tner/btc"], "dataset_split": "train", "dataset_name": null, "local_dataset": null, "model": "roberta-large", "crf": true, "max_length": 128, "epoch": 15, "batch_size": 64, "lr": 1e-05, "random_seed": 42, "gradient_accumulation_steps": 2, "weight_decay": 1e-07, "lr_warmup_step_ratio": 0.1, "max_grad_norm": null}