tiny-mlm-glue-cola-target-glue-mnli
This model is a fine-tuned version of muhtasham/tiny-mlm-glue-cola on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.8037
- Accuracy: 0.6427
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 | Accuracy |
|---|---|---|---|---|
| 1.0736 | 0.04 | 500 | 1.0266 | 0.4807 |
| 1.0005 | 0.08 | 1000 | 0.9516 | 0.5605 |
| 0.9517 | 0.12 | 1500 | 0.9140 | 0.5810 |
| 0.9271 | 0.16 | 2000 | 0.9009 | 0.5921 |
| 0.919 | 0.2 | 2500 | 0.8858 | 0.6014 |
| 0.9125 | 0.24 | 3000 | 0.8740 | 0.6069 |
| 0.8965 | 0.29 | 3500 | 0.8676 | 0.6134 |
| 0.89 | 0.33 | 4000 | 0.8547 | 0.6193 |
| 0.8754 | 0.37 | 4500 | 0.8516 | 0.6214 |
| 0.8779 | 0.41 | 5000 | 0.8448 | 0.6220 |
| 0.8698 | 0.45 | 5500 | 0.8396 | 0.6252 |
| 0.8653 | 0.49 | 6000 | 0.8371 | 0.6287 |
| 0.8692 | 0.53 | 6500 | 0.8304 | 0.6309 |
| 0.8579 | 0.57 | 7000 | 0.8307 | 0.6301 |
| 0.8528 | 0.61 | 7500 | 0.8151 | 0.6409 |
| 0.8538 | 0.65 | 8000 | 0.8153 | 0.6381 |
| 0.8451 | 0.69 | 8500 | 0.8264 | 0.6329 |
| 0.8497 | 0.73 | 9000 | 0.8002 | 0.6464 |
| 0.8401 | 0.77 | 9500 | 0.8125 | 0.6363 |
| 0.8299 | 0.81 | 10000 | 0.7968 | 0.6464 |
| 0.8343 | 0.86 | 10500 | 0.8037 | 0.6427 |
Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.0+cu116
- Datasets 2.8.1.dev0
- Tokenizers 0.13.2