mobilebert_add_GLUE_Experiment_qqp_256
This model is a fine-tuned version of google/mobilebert-uncased on the GLUE QQP dataset. It achieves the following results on the evaluation set:
- Loss: 0.5069
- Accuracy: 0.7558
- F1: 0.6391
- Combined Score: 0.6975
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: 128
- eval_batch_size: 128
- 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
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Combined Score |
|---|---|---|---|---|---|---|
| 0.6505 | 1.0 | 2843 | 0.6497 | 0.6321 | 0.0012 | 0.3166 |
| 0.6473 | 2.0 | 5686 | 0.6479 | 0.6321 | 0.0012 | 0.3166 |
| 0.5376 | 3.0 | 8529 | 0.5167 | 0.7486 | 0.5879 | 0.6682 |
| 0.4943 | 4.0 | 11372 | 0.5069 | 0.7558 | 0.6391 | 0.6975 |
| 0.4816 | 5.0 | 14215 | 0.5072 | 0.7547 | 0.6574 | 0.7061 |
| 0.4738 | 6.0 | 17058 | nan | 0.7588 | 0.6526 | 0.7057 |
| 0.4646 | 7.0 | 19901 | nan | 0.6318 | 0.0 | 0.3159 |
| 0.0 | 8.0 | 22744 | nan | 0.6318 | 0.0 | 0.3159 |
| 0.0 | 9.0 | 25587 | nan | 0.6318 | 0.0 | 0.3159 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.14.0a0+410ce96
- Datasets 2.8.0
- Tokenizers 0.13.2