small-mlm-glue-stsb-target-glue-sst2
This model is a fine-tuned version of muhtasham/small-mlm-glue-stsb on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.3917
- Accuracy: 0.8876
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 |
|---|---|---|---|---|
| 0.3935 | 0.24 | 500 | 0.3629 | 0.8429 |
| 0.3047 | 0.48 | 1000 | 0.3266 | 0.8624 |
| 0.2593 | 0.71 | 1500 | 0.3357 | 0.8624 |
| 0.2399 | 0.95 | 2000 | 0.3098 | 0.8739 |
| 0.1845 | 1.19 | 2500 | 0.3578 | 0.8761 |
| 0.1687 | 1.43 | 3000 | 0.3724 | 0.8716 |
| 0.1668 | 1.66 | 3500 | 0.3827 | 0.8842 |
| 0.1672 | 1.9 | 4000 | 0.3403 | 0.8899 |
| 0.1361 | 2.14 | 4500 | 0.5571 | 0.8372 |
| 0.1256 | 2.38 | 5000 | 0.3917 | 0.8876 |
Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.0+cu116
- Datasets 2.8.1.dev0
- Tokenizers 0.13.2