nli_sts_roberta_large_lr1e_05_wd1e_03_ep5_lr1e-05_wd1e-03_ep5_ckpt
This model is a fine-tuned version of ys7yoo/sts_roberta_large_lr1e-05_wd1e-03_ep5 on the klue dataset. It achieves the following results on the evaluation set:
- Loss: 0.4971
- Accuracy: 0.8987
- F1: 0.8985
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 | Accuracy | F1 |
---|---|---|---|---|---|
0.5471 | 1.0 | 391 | 0.3522 | 0.876 | 0.8756 |
0.2379 | 2.0 | 782 | 0.3345 | 0.8983 | 0.8981 |
0.1215 | 3.0 | 1173 | 0.3708 | 0.8997 | 0.8995 |
0.0661 | 4.0 | 1564 | 0.4734 | 0.896 | 0.8958 |
0.0407 | 5.0 | 1955 | 0.4971 | 0.8987 | 0.8985 |
Framework versions
- Transformers 4.33.2
- Pytorch 2.0.1+cu117
- Datasets 2.13.0
- Tokenizers 0.13.3
- Downloads last month
- 5
Finetuned from
Dataset used to train ys7yoo/nli_sts_roberta_large_lr1e_05_wd1e_03_ep5_lr1e-05_wd1e-03_ep5_ckpt
Evaluation results
- Accuracy on kluevalidation set self-reported0.899
- F1 on kluevalidation set self-reported0.899