small-mlm-glue-rte-target-glue-qqp
This model is a fine-tuned version of muhtasham/small-mlm-glue-rte on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.3294
- Accuracy: 0.8496
- F1: 0.8112
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
- training_steps: 5000
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|---|---|---|---|---|---|
| 0.4764 | 0.04 | 500 | 0.4288 | 0.7863 | 0.7498 |
| 0.4172 | 0.09 | 1000 | 0.3936 | 0.8089 | 0.7701 |
| 0.4017 | 0.13 | 1500 | 0.3707 | 0.8236 | 0.7785 |
| 0.3865 | 0.18 | 2000 | 0.3751 | 0.8197 | 0.7857 |
| 0.3788 | 0.22 | 2500 | 0.3682 | 0.8292 | 0.7938 |
| 0.364 | 0.26 | 3000 | 0.3517 | 0.8351 | 0.7969 |
| 0.3616 | 0.31 | 3500 | 0.3324 | 0.8496 | 0.8043 |
| 0.3533 | 0.35 | 4000 | 0.3348 | 0.8457 | 0.8071 |
| 0.3599 | 0.4 | 4500 | 0.3362 | 0.8451 | 0.8094 |
| 0.3465 | 0.44 | 5000 | 0.3294 | 0.8496 | 0.8112 |
Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.0+cu116
- Datasets 2.8.1.dev0
- Tokenizers 0.13.2