metadata
base_model: klue/roberta-base
tags:
- generated_from_trainer
datasets:
- klue
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: klue_ner_roberta_model
results:
- task:
name: Token Classification
type: token-classification
dataset:
name: klue
type: klue
config: ner
split: validation
args: ner
metrics:
- name: Precision
type: precision
value: 0.9545986426398315
- name: Recall
type: recall
value: 0.9557169634489222
- name: F1
type: f1
value: 0.955157475705421
- name: Accuracy
type: accuracy
value: 0.9883703228112445
klue_ner_roberta_model
This model is a fine-tuned version of klue/roberta-base on the klue dataset. It achieves the following results on the evaluation set:
- Loss: 0.0487
- Precision: 0.9546
- Recall: 0.9557
- F1: 0.9552
- Accuracy: 0.9884
Model description
Pretrained RoBERTa Model on Korean Language. See Github and Paper for more details.
Intended uses & limitations
How to use
NOTE: Use BertTokenizer instead of RobertaTokenizer. (AutoTokenizer will load BertTokenizer)
from transformers import AutoModel, AutoTokenizer
model = AutoModel.from_pretrained("klue/roberta-base")
tokenizer = AutoTokenizer.from_pretrained("klue/roberta-base")
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.0449 | 1.0 | 2626 | 0.0601 | 0.9361 | 0.9176 | 0.9267 | 0.9830 |
| 0.0262 | 2.0 | 5252 | 0.0469 | 0.9484 | 0.9510 | 0.9497 | 0.9874 |
| 0.0144 | 3.0 | 7878 | 0.0487 | 0.9546 | 0.9557 | 0.9552 | 0.9884 |
Framework versions
- Transformers 4.31.0
- Pytorch 2.0.1+cu118
- Datasets 2.13.1
- Tokenizers 0.13.3