pubmed-mixed-noise-v3-0.2
This model is a fine-tuned version of facebook/bart-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.4140
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: 5e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 10
- num_epochs: 3
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.7328 | 0.11 | 500 | 0.6952 |
| 0.6994 | 0.21 | 1000 | 0.6005 |
| 0.6684 | 0.32 | 1500 | 0.5670 |
| 0.6305 | 0.43 | 2000 | 0.5402 |
| 0.6224 | 0.54 | 2500 | 0.5175 |
| 0.6009 | 0.64 | 3000 | 0.5001 |
| 0.5126 | 0.75 | 3500 | 0.4935 |
| 0.5698 | 0.86 | 4000 | 0.4793 |
| 0.497 | 0.96 | 4500 | 0.4715 |
| 0.3716 | 1.07 | 5000 | 0.4689 |
| 0.4522 | 1.18 | 5500 | 0.4551 |
| 0.3958 | 1.28 | 6000 | 0.4556 |
| 0.4398 | 1.39 | 6500 | 0.4502 |
| 0.4572 | 1.5 | 7000 | 0.4425 |
| 0.4339 | 1.61 | 7500 | 0.4424 |
| 0.4289 | 1.71 | 8000 | 0.4322 |
| 0.3771 | 1.82 | 8500 | 0.4337 |
| 0.3668 | 1.93 | 9000 | 0.4265 |
| 0.3342 | 2.03 | 9500 | 0.4316 |
| 0.3465 | 2.14 | 10000 | 0.4244 |
| 0.32 | 2.25 | 10500 | 0.4226 |
| 0.3493 | 2.35 | 11000 | 0.4244 |
| 0.3549 | 2.46 | 11500 | 0.4216 |
| 0.3281 | 2.57 | 12000 | 0.4192 |
| 0.3259 | 2.68 | 12500 | 0.4181 |
| 0.3444 | 2.78 | 13000 | 0.4156 |
| 0.3201 | 2.89 | 13500 | 0.4146 |
| 0.3315 | 3.0 | 14000 | 0.4140 |
Framework versions
- Transformers 4.36.1
- Pytorch 2.0.1
- Datasets 2.16.1
- Tokenizers 0.15.0