small-mlm-glue-mrpc
This model is a fine-tuned version of google/bert_uncased_L-4_H-512_A-8 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.6779
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: 3e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant
- num_epochs: 200
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 2.8599 | 1.09 | 500 | 2.5767 |
| 2.7017 | 2.18 | 1000 | 2.5086 |
| 2.535 | 3.27 | 1500 | 2.5969 |
| 2.4565 | 4.36 | 2000 | 2.6871 |
| 2.3714 | 5.45 | 2500 | 2.4743 |
| 2.3244 | 6.54 | 3000 | 2.5176 |
| 2.2136 | 7.63 | 3500 | 2.4796 |
| 2.1939 | 8.71 | 4000 | 2.6779 |
Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.0+cu116
- Datasets 2.8.1.dev0
- Tokenizers 0.13.2