distilbert_sa_GLUE_Experiment_stsb_192
This model is a fine-tuned version of distilbert-base-uncased on the GLUE STSB dataset. It achieves the following results on the evaluation set:
- Loss: 2.2586
- Pearson: -0.0814
- Spearmanr: -0.0816
- Combined Score: -0.0815
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 | Pearson | Spearmanr | Combined Score |
|---|---|---|---|---|---|---|
| 6.966 | 1.0 | 23 | 4.0539 | -0.0244 | -0.0244 | -0.0244 |
| 4.4237 | 2.0 | 46 | 3.1176 | -0.0508 | -0.0503 | -0.0505 |
| 3.3768 | 3.0 | 69 | 2.5232 | -0.1303 | -0.1323 | -0.1313 |
| 2.6486 | 4.0 | 92 | 2.2586 | -0.0814 | -0.0816 | -0.0815 |
| 2.2539 | 5.0 | 115 | 2.3547 | 0.0512 | 0.0505 | 0.0508 |
| 2.1692 | 6.0 | 138 | 2.3367 | 0.0642 | 0.0568 | 0.0605 |
| 2.1268 | 7.0 | 161 | 2.4285 | 0.0444 | 0.0649 | 0.0546 |
| 1.9924 | 8.0 | 184 | 2.6031 | 0.0781 | 0.0846 | 0.0814 |
| 1.8254 | 9.0 | 207 | 2.6306 | 0.1155 | 0.1187 | 0.1171 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.8.0
- Tokenizers 0.13.2