distilbert_add_GLUE_Experiment_logit_kd_qqp
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: 0.6623
- Accuracy: 0.6425
- F1: 0.0601
- Combined Score: 0.3513
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 |
|---|---|---|---|---|---|---|
| 0.7968 | 1.0 | 1422 | 0.7159 | 0.6323 | 0.0030 | 0.3176 |
| 0.6542 | 2.0 | 2844 | 0.6925 | 0.6338 | 0.0115 | 0.3226 |
| 0.5893 | 3.0 | 4266 | 0.6695 | 0.6348 | 0.0172 | 0.3260 |
| 0.5538 | 4.0 | 5688 | 0.7068 | 0.6386 | 0.0393 | 0.3390 |
| 0.5323 | 5.0 | 7110 | 0.6670 | 0.6500 | 0.1014 | 0.3757 |
| 0.5181 | 6.0 | 8532 | 0.6738 | 0.6420 | 0.0573 | 0.3497 |
| 0.5082 | 7.0 | 9954 | 0.6623 | 0.6425 | 0.0601 | 0.3513 |
| 0.5012 | 8.0 | 11376 | 0.6995 | 0.6412 | 0.0536 | 0.3474 |
| 0.4957 | 9.0 | 12798 | 0.6836 | 0.6472 | 0.0858 | 0.3665 |
| 0.4911 | 10.0 | 14220 | 0.6778 | 0.6484 | 0.0922 | 0.3703 |
| 0.4874 | 11.0 | 15642 | 0.7183 | 0.6415 | 0.0550 | 0.3483 |
| 0.484 | 12.0 | 17064 | 0.6730 | 0.6451 | 0.0744 | 0.3598 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.9.0
- Tokenizers 0.13.2