distilbert_sa_GLUE_Experiment_mrpc_256
This model is a fine-tuned version of distilbert-base-uncased on the GLUE MRPC dataset. It achieves the following results on the evaluation set:
- Loss: 0.5996
- Accuracy: 0.6814
- F1: 0.8105
- Combined Score: 0.7459
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 | F1 | Combined Score |
|---|---|---|---|---|---|---|
| 0.6343 | 1.0 | 15 | 0.6246 | 0.6838 | 0.8122 | 0.7480 |
| 0.6276 | 2.0 | 30 | 0.6234 | 0.6838 | 0.8122 | 0.7480 |
| 0.6306 | 3.0 | 45 | 0.6243 | 0.6838 | 0.8122 | 0.7480 |
| 0.6279 | 4.0 | 60 | 0.6205 | 0.6838 | 0.8122 | 0.7480 |
| 0.6168 | 5.0 | 75 | 0.5996 | 0.6814 | 0.8105 | 0.7459 |
| 0.5632 | 6.0 | 90 | 0.6020 | 0.6936 | 0.7954 | 0.7445 |
| 0.5021 | 7.0 | 105 | 0.6094 | 0.6936 | 0.7841 | 0.7389 |
| 0.4263 | 8.0 | 120 | 0.6844 | 0.6299 | 0.7113 | 0.6706 |
| 0.3476 | 9.0 | 135 | 0.7218 | 0.6373 | 0.7098 | 0.6735 |
| 0.2966 | 10.0 | 150 | 0.7759 | 0.7010 | 0.7953 | 0.7481 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.8.0
- Tokenizers 0.13.2