hBERTv1_new_pretrain_48_KD_w_init_wnli
This model is a fine-tuned version of gokuls/bert_12_layer_model_v1_complete_training_new_48_KD_wt_init on the GLUE WNLI dataset. It achieves the following results on the evaluation set:
- Loss: 0.6851
- Accuracy: 0.5634
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: 4e-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 |
|---|---|---|---|---|
| 0.9459 | 1.0 | 5 | 0.7251 | 0.4366 |
| 0.7037 | 2.0 | 10 | 0.6928 | 0.5211 |
| 0.6983 | 3.0 | 15 | 0.6927 | 0.5634 |
| 0.7148 | 4.0 | 20 | 0.6853 | 0.5634 |
| 0.6998 | 5.0 | 25 | 0.7046 | 0.4366 |
| 0.7209 | 6.0 | 30 | 0.6895 | 0.5634 |
| 0.7165 | 7.0 | 35 | 0.6867 | 0.5634 |
| 0.7012 | 8.0 | 40 | 0.6851 | 0.5634 |
| 0.6996 | 9.0 | 45 | 0.7126 | 0.4366 |
| 0.7131 | 10.0 | 50 | 0.6867 | 0.5634 |
| 0.7036 | 11.0 | 55 | 0.6879 | 0.5634 |
| 0.7026 | 12.0 | 60 | 0.6931 | 0.5634 |
| 0.7033 | 13.0 | 65 | 0.6938 | 0.4366 |
Framework versions
- Transformers 4.30.2
- Pytorch 1.14.0a0+410ce96
- Datasets 2.13.0
- Tokenizers 0.13.3