distilbert_add_GLUE_Experiment_sst2_96
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: 0.5220
- Accuracy: 0.7683
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 |
|---|---|---|---|---|
| 0.6871 | 1.0 | 264 | 0.6971 | 0.5092 |
| 0.6867 | 2.0 | 528 | 0.6968 | 0.5092 |
| 0.6865 | 3.0 | 792 | 0.6978 | 0.5092 |
| 0.6864 | 4.0 | 1056 | 0.6951 | 0.5092 |
| 0.514 | 5.0 | 1320 | 0.5220 | 0.7683 |
| 0.3536 | 6.0 | 1584 | 0.5425 | 0.7764 |
| 0.297 | 7.0 | 1848 | 0.5412 | 0.7901 |
| 0.2677 | 8.0 | 2112 | 0.5740 | 0.7729 |
| 0.245 | 9.0 | 2376 | 0.5970 | 0.7741 |
| 0.2313 | 10.0 | 2640 | 0.6024 | 0.7856 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.8.0
- Tokenizers 0.13.2