<!-- 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_data_aug_sst2_256
This model is a fine-tuned version of distilbert-base-uncased on the GLUE SST2 dataset. It achieves the following results on the evaluation set:
- Loss: 0.5583
- Accuracy: 0.8406
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 |
|---|---|---|---|---|
| 0.5009 | 1.0 | 4374 | 0.6370 | 0.8165 |
| 0.3329 | 2.0 | 8748 | 0.6643 | 0.8257 |
| 0.2804 | 3.0 | 13122 | 0.6192 | 0.8326 |
| 0.249 | 4.0 | 17496 | 0.6205 | 0.8372 |
| 0.2279 | 5.0 | 21870 | 0.6250 | 0.8349 |
| 0.2122 | 6.0 | 26244 | 0.6644 | 0.8280 |
| 0.2008 | 7.0 | 30618 | 0.5707 | 0.8440 |
| 0.1918 | 8.0 | 34992 | 0.5863 | 0.8360 |
| 0.1847 | 9.0 | 39366 | 0.5779 | 0.8394 |
| 0.1784 | 10.0 | 43740 | 0.5662 | 0.8349 |
| 0.1734 | 11.0 | 48114 | 0.5619 | 0.8394 |
| 0.169 | 12.0 | 52488 | 0.5583 | 0.8406 |
| 0.1653 | 13.0 | 56862 | 0.5830 | 0.8303 |
| 0.1619 | 14.0 | 61236 | 0.5773 | 0.8372 |
| 0.1591 | 15.0 | 65610 | 0.5728 | 0.8291 |
| 0.1564 | 16.0 | 69984 | 0.5631 | 0.8383 |
| 0.154 | 17.0 | 74358 | 0.5628 | 0.8452 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.9.0
- Tokenizers 0.13.2
