small-mlm-glue-qnli-target-glue-qqp
This model is a fine-tuned version of muhtasham/small-mlm-glue-qnli on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.3296
- Accuracy: 0.8511
- F1: 0.8117
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 |
|---|---|---|---|---|---|
| 0.4762 | 0.04 | 500 | 0.4247 | 0.7897 | 0.7473 |
| 0.4188 | 0.09 | 1000 | 0.3880 | 0.8126 | 0.7702 |
| 0.4011 | 0.13 | 1500 | 0.3760 | 0.8194 | 0.7750 |
| 0.387 | 0.18 | 2000 | 0.3779 | 0.8189 | 0.7866 |
| 0.3802 | 0.22 | 2500 | 0.3642 | 0.8320 | 0.7958 |
| 0.3606 | 0.26 | 3000 | 0.3526 | 0.8358 | 0.7972 |
| 0.3604 | 0.31 | 3500 | 0.3337 | 0.8495 | 0.8010 |
| 0.3538 | 0.35 | 4000 | 0.3341 | 0.8483 | 0.8102 |
| 0.3582 | 0.4 | 4500 | 0.3293 | 0.8503 | 0.8106 |
| 0.345 | 0.44 | 5000 | 0.3296 | 0.8511 | 0.8117 |
Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.0+cu116
- Datasets 2.8.1.dev0
- Tokenizers 0.13.2