small-mlm-glue-stsb-target-glue-qqp
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: 0.3294
- Accuracy: 0.8525
- F1: 0.8131
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 |
|---|---|---|---|---|---|
| 0.4739 | 0.04 | 500 | 0.4259 | 0.7919 | 0.7514 |
| 0.4186 | 0.09 | 1000 | 0.3841 | 0.8190 | 0.7709 |
| 0.3984 | 0.13 | 1500 | 0.3737 | 0.8228 | 0.7757 |
| 0.3853 | 0.18 | 2000 | 0.3725 | 0.8228 | 0.7878 |
| 0.3761 | 0.22 | 2500 | 0.3558 | 0.8362 | 0.7969 |
| 0.3616 | 0.26 | 3000 | 0.3434 | 0.8418 | 0.8010 |
| 0.3616 | 0.31 | 3500 | 0.3286 | 0.8504 | 0.8008 |
| 0.3528 | 0.35 | 4000 | 0.3293 | 0.8513 | 0.8110 |
| 0.358 | 0.4 | 4500 | 0.3213 | 0.8539 | 0.8104 |
| 0.3428 | 0.44 | 5000 | 0.3294 | 0.8525 | 0.8131 |
Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.0+cu116
- Datasets 2.8.1.dev0
- Tokenizers 0.13.2