distilbert_sa_GLUE_Experiment_logit_kd_mnli_256
This model is a fine-tuned version of distilbert-base-uncased on the GLUE MNLI dataset. It achieves the following results on the evaluation set:
- Loss: 0.5305
- Accuracy: 0.5881
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 |
|---|---|---|---|---|
| 0.5834 | 1.0 | 1534 | 0.5611 | 0.5153 |
| 0.5545 | 2.0 | 3068 | 0.5469 | 0.5330 |
| 0.5418 | 3.0 | 4602 | 0.5420 | 0.5477 |
| 0.5323 | 4.0 | 6136 | 0.5382 | 0.5633 |
| 0.5235 | 5.0 | 7670 | 0.5333 | 0.5743 |
| 0.5153 | 6.0 | 9204 | 0.5315 | 0.5753 |
| 0.5078 | 7.0 | 10738 | 0.5295 | 0.5832 |
| 0.5005 | 8.0 | 12272 | 0.5313 | 0.5821 |
| 0.4939 | 9.0 | 13806 | 0.5281 | 0.5949 |
| 0.4873 | 10.0 | 15340 | 0.5298 | 0.5895 |
| 0.481 | 11.0 | 16874 | 0.5263 | 0.6010 |
| 0.4748 | 12.0 | 18408 | 0.5288 | 0.6041 |
| 0.469 | 13.0 | 19942 | 0.5288 | 0.6031 |
| 0.463 | 14.0 | 21476 | 0.5336 | 0.6056 |
| 0.4574 | 15.0 | 23010 | 0.5296 | 0.6072 |
| 0.4523 | 16.0 | 24544 | 0.5358 | 0.6096 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.9.0
- Tokenizers 0.13.2