<!-- 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. -->
distilbert_sa_GLUE_Experiment_logit_kd_qqp_384
This model is a fine-tuned version of distilbert-base-uncased on the GLUE QQP dataset. It achieves the following results on the evaluation set:
- Loss: 0.6771
- Accuracy: 0.6454
- F1: 0.0788
- Combined Score: 0.3621
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: 5e-05
- train_batch_size: 256
- eval_batch_size: 256
- seed: 10
- distributed_type: multi-GPU
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 50
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Combined Score |
|---|---|---|---|---|---|---|
| 0.7984 | 1.0 | 1422 | 0.7600 | 0.6318 | 0.0 | 0.3159 |
| 0.7388 | 2.0 | 2844 | 0.7348 | 0.6318 | 0.0 | 0.3159 |
| 0.7037 | 3.0 | 4266 | 0.7082 | 0.6329 | 0.0056 | 0.3192 |
| 0.6717 | 4.0 | 5688 | 0.7014 | 0.6474 | 0.0908 | 0.3691 |
| 0.6462 | 5.0 | 7110 | 0.6841 | 0.6377 | 0.0339 | 0.3358 |
| 0.6259 | 6.0 | 8532 | 0.6795 | 0.6382 | 0.0364 | 0.3373 |
| 0.6092 | 7.0 | 9954 | 0.6782 | 0.6408 | 0.0513 | 0.3461 |
| 0.5941 | 8.0 | 11376 | 0.6771 | 0.6454 | 0.0788 | 0.3621 |
| 0.5812 | 9.0 | 12798 | 0.6841 | 0.6492 | 0.0991 | 0.3741 |
| 0.5703 | 10.0 | 14220 | 0.6774 | 0.6452 | 0.0776 | 0.3614 |
| 0.5604 | 11.0 | 15642 | 0.6791 | 0.6464 | 0.0831 | 0.3647 |
| 0.5523 | 12.0 | 17064 | 0.6817 | 0.6520 | 0.1143 | 0.3831 |
| 0.5448 | 13.0 | 18486 | 0.6774 | 0.6477 | 0.0905 | 0.3691 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.9.0
- Tokenizers 0.13.2
