metadata
base_model: klue/roberta-large
tags:
- generated_from_trainer
datasets:
- klue
model-index:
- name: sts_roberta_large_lr1e-05_wd1e-03_ep5_ckpt
results: []
sts_roberta_large_lr1e-05_wd1e-03_ep5_ckpt
This model is a fine-tuned version of klue/roberta-large on the klue dataset. It achieves the following results on the evaluation set:
- Loss: 0.3718
- Mse: 0.3718
- Mae: 0.4467
- R2: 0.8298
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: 1e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Mse | Mae | R2 |
|---|---|---|---|---|---|---|
| 1.4636 | 1.0 | 183 | 0.6353 | 0.6353 | 0.5960 | 0.7091 |
| 0.1439 | 2.0 | 366 | 0.4268 | 0.4268 | 0.4874 | 0.8046 |
| 0.1016 | 3.0 | 549 | 0.4241 | 0.4241 | 0.4769 | 0.8058 |
| 0.0796 | 4.0 | 732 | 0.4218 | 0.4218 | 0.4811 | 0.8069 |
| 0.0651 | 5.0 | 915 | 0.3718 | 0.3718 | 0.4467 | 0.8298 |
Framework versions
- Transformers 4.33.2
- Pytorch 2.0.1+cu117
- Datasets 2.13.0
- Tokenizers 0.13.3