distilbert_add_GLUE_Experiment_mnli_384
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: 0.9341
- Accuracy: 0.5653
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 |
|---|---|---|---|---|
| 1.074 | 1.0 | 1534 | 1.0374 | 0.4592 |
| 1.0289 | 2.0 | 3068 | 1.0214 | 0.4736 |
| 1.0052 | 3.0 | 4602 | 0.9957 | 0.4889 |
| 0.9869 | 4.0 | 6136 | 0.9924 | 0.5111 |
| 0.9709 | 5.0 | 7670 | 0.9683 | 0.5246 |
| 0.9562 | 6.0 | 9204 | 0.9628 | 0.5295 |
| 0.9403 | 7.0 | 10738 | 0.9524 | 0.5379 |
| 0.9244 | 8.0 | 12272 | 0.9547 | 0.5309 |
| 0.9071 | 9.0 | 13806 | 0.9358 | 0.5563 |
| 0.8914 | 10.0 | 15340 | 0.9286 | 0.5586 |
| 0.8771 | 11.0 | 16874 | 0.9384 | 0.5651 |
| 0.863 | 12.0 | 18408 | 0.9361 | 0.5616 |
| 0.849 | 13.0 | 19942 | 0.9401 | 0.5606 |
| 0.8346 | 14.0 | 21476 | 0.9441 | 0.5646 |
| 0.8199 | 15.0 | 23010 | 0.9276 | 0.5677 |
| 0.8048 | 16.0 | 24544 | 0.9444 | 0.5656 |
| 0.7883 | 17.0 | 26078 | 0.9477 | 0.5653 |
| 0.7707 | 18.0 | 27612 | 0.9491 | 0.5745 |
| 0.7543 | 19.0 | 29146 | 0.9613 | 0.5825 |
| 0.7371 | 20.0 | 30680 | 0.9645 | 0.5728 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.8.0
- Tokenizers 0.13.2