metadata
license: apache-2.0
base_model: cl-tohoku/bert-large-japanese-v2
tags:
- generated_from_trainer
metrics:
- accuracy
- f1
model-index:
- name: bert-large-japanease-v2-gpt4-relevance-learned
results: []
bert-large-japanease-v2-gpt4-relevance-learned
This model is a fine-tuned version of cl-tohoku/bert-large-japanese-v2 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 3.2693
- Accuracy: 0.885
- F1: 0.8788
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: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|---|---|---|---|---|---|
| 3.3692 | 1.0 | 563 | 3.2122 | 0.872 | 0.8560 |
| 3.0963 | 2.0 | 1126 | 3.1045 | 0.866 | 0.8625 |
| 2.8698 | 3.0 | 1689 | 3.1410 | 0.882 | 0.8755 |
| 2.6212 | 4.0 | 2252 | 3.2119 | 0.876 | 0.8702 |
| 2.407 | 5.0 | 2815 | 3.2693 | 0.885 | 0.8788 |
Framework versions
- Transformers 4.33.2
- Pytorch 2.0.1+cu118
- Datasets 2.14.5
- Tokenizers 0.13.3