metadata
license: apache-2.0
base_model: cl-tohoku/bert-base-japanese-v3
tags:
- generated_from_trainer
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: bert-japanese-ner
results: []
bert-japanese-ner
This model is a fine-tuned version of cl-tohoku/bert-base-japanese-v3 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0372
- Precision: 0.9673
- Recall: 0.9682
- F1: 0.9678
- Accuracy: 0.9933
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.0553 | 1.0 | 848 | 0.0263 | 0.9683 | 0.9334 | 0.9505 | 0.9908 |
| 0.0133 | 2.0 | 1696 | 0.0241 | 0.9707 | 0.9560 | 0.9633 | 0.9928 |
| 0.0065 | 3.0 | 2544 | 0.0245 | 0.9631 | 0.9706 | 0.9668 | 0.9935 |
| 0.0027 | 4.0 | 3392 | 0.0321 | 0.9716 | 0.9659 | 0.9687 | 0.9936 |
| 0.0012 | 5.0 | 4240 | 0.0372 | 0.9673 | 0.9682 | 0.9678 | 0.9933 |
Framework versions
- Transformers 4.36.2
- Pytorch 2.0.1+cu118
- Datasets 2.16.1
- Tokenizers 0.15.0