distilbert_add_GLUE_Experiment_qnli_96
This model is a fine-tuned version of distilbert-base-uncased on the GLUE QNLI dataset. It achieves the following results on the evaluation set:
- Loss: 0.6575
- Accuracy: 0.6072
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 |
|---|---|---|---|---|
| 0.6932 | 1.0 | 410 | 0.6931 | 0.4946 |
| 0.6932 | 2.0 | 820 | 0.6932 | 0.4946 |
| 0.6932 | 3.0 | 1230 | 0.6931 | 0.5054 |
| 0.6826 | 4.0 | 1640 | 0.6659 | 0.5967 |
| 0.6539 | 5.0 | 2050 | 0.6575 | 0.6072 |
| 0.6403 | 6.0 | 2460 | 0.6608 | 0.6074 |
| 0.6288 | 7.0 | 2870 | 0.6702 | 0.6039 |
| 0.6186 | 8.0 | 3280 | 0.6730 | 0.6022 |
| 0.6094 | 9.0 | 3690 | 0.6740 | 0.6013 |
| 0.5995 | 10.0 | 4100 | 0.6906 | 0.5920 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.8.0
- Tokenizers 0.13.2