<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
roberta-large-all-dataset
This model is a fine-tuned version of klue/roberta-large on the None dataset. It achieves the following results on the evaluation set:
- Exact Match: 77.1769
- F1: 85.6178
- Loss: 0.5347
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 32
- total_train_batch_size: 256
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 10
- mixed_precision_training: Native AMP
- label_smoothing_factor: 0.1
Training results
Training Loss | Epoch | Step | Exact Match | F1 | Validation Loss |
---|---|---|---|---|---|
3.7417 | 1.5 | 672 | 68.5244 | 78.5098 | 1.1172 |
0.889 | 3.0 | 1344 | 74.5613 | 83.7411 | 0.6406 |
0.6361 | 4.5 | 2016 | 76.3492 | 84.9826 | 0.5654 |
0.5696 | 6.0 | 2688 | 76.6472 | 85.2937 | 0.5493 |
0.5461 | 7.5 | 3360 | 76.7906 | 85.3519 | 0.5415 |
0.5349 | 9.0 | 4032 | 77.1769 | 85.6178 | 0.5347 |
Framework versions
- Transformers 4.25.1
- Pytorch 1.13.0+cu117
- Datasets 2.7.1
- Tokenizers 0.13.2