small-mlm-glue-mnli-target-glue-qqp
This model is a fine-tuned version of muhtasham/small-mlm-glue-mnli on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.3263
- Accuracy: 0.8535
- F1: 0.8134
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 |
|---|---|---|---|---|---|
| 0.4778 | 0.04 | 500 | 0.4286 | 0.7863 | 0.7468 |
| 0.4182 | 0.09 | 1000 | 0.3862 | 0.8142 | 0.7696 |
| 0.4014 | 0.13 | 1500 | 0.3732 | 0.8225 | 0.7767 |
| 0.3851 | 0.18 | 2000 | 0.3686 | 0.8234 | 0.7887 |
| 0.3784 | 0.22 | 2500 | 0.3600 | 0.8338 | 0.7974 |
| 0.36 | 0.26 | 3000 | 0.3438 | 0.8406 | 0.7995 |
| 0.3583 | 0.31 | 3500 | 0.3361 | 0.8475 | 0.7970 |
| 0.3528 | 0.35 | 4000 | 0.3316 | 0.8472 | 0.8076 |
| 0.3567 | 0.4 | 4500 | 0.3307 | 0.8494 | 0.8089 |
| 0.3428 | 0.44 | 5000 | 0.3263 | 0.8535 | 0.8134 |
Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.0+cu116
- Datasets 2.8.1.dev0
- Tokenizers 0.13.2