<!-- 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. -->
robeta-large-batch128
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: 60.4167
- F1: 67.7400
- Loss: 1.8242
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: 16
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 15
- mixed_precision_training: Native AMP
- label_smoothing_factor: 0.1
Training results
| Training Loss | Epoch | Step | Exact Match | F1 | Validation Loss |
|---|---|---|---|---|---|
| 2.5177 | 2.08 | 100 | 62.9167 | 75.0527 | 1.2793 |
| 0.7171 | 4.17 | 200 | 60.8333 | 69.9682 | 1.2480 |
| 0.3504 | 6.25 | 300 | 66.6667 | 76.6871 | 1.2969 |
| 0.1727 | 8.33 | 400 | 64.1667 | 73.7049 | 1.2617 |
| 0.3793 | 10.42 | 500 | 58.3333 | 68.6892 | 1.5420 |
| 0.1726 | 12.5 | 600 | 66.25 | 74.7943 | 1.8848 |
| 0.0974 | 14.58 | 700 | 60.4167 | 67.7400 | 1.8242 |
Framework versions
- Transformers 4.25.1
- Pytorch 1.13.0+cu117
- Datasets 2.7.1
- Tokenizers 0.13.2
