pubmed-mixed-noise-v5-0.1-large
This model is a fine-tuned version of facebook/bart-large on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.3064
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 |
|---|---|---|---|
| 0.4762 | 0.11 | 500 | 0.4936 |
| 0.4174 | 0.21 | 1000 | 0.4293 |
| 0.3835 | 0.32 | 1500 | 0.4280 |
| 0.3628 | 0.43 | 2000 | 0.4472 |
| 0.3927 | 0.54 | 2500 | 0.3898 |
| 0.3012 | 0.64 | 3000 | 0.3744 |
| 0.3189 | 0.75 | 3500 | 0.3784 |
| 0.2986 | 0.86 | 4000 | 0.3624 |
| 0.2493 | 0.96 | 4500 | 0.3588 |
| 0.2438 | 1.07 | 5000 | 0.3439 |
| 0.2465 | 1.18 | 5500 | 0.3448 |
| 0.268 | 1.28 | 6000 | 0.3476 |
| 0.2298 | 1.39 | 6500 | 0.3411 |
| 0.2587 | 1.5 | 7000 | 0.3322 |
| 0.2499 | 1.61 | 7500 | 0.3253 |
| 0.2296 | 1.71 | 8000 | 0.3177 |
| 0.2184 | 1.82 | 8500 | 0.3175 |
| 0.2245 | 1.93 | 9000 | 0.3573 |
| 0.164 | 2.03 | 9500 | 0.3292 |
| 0.1784 | 2.14 | 10000 | 0.3224 |
| 0.1487 | 2.25 | 10500 | 0.3209 |
| 0.1818 | 2.35 | 11000 | 0.3175 |
| 0.1521 | 2.46 | 11500 | 0.3190 |
| 0.1663 | 2.57 | 12000 | 0.3137 |
| 0.1604 | 2.68 | 12500 | 0.3113 |
| 0.1447 | 2.78 | 13000 | 0.3080 |
| 0.162 | 2.89 | 13500 | 0.3068 |
| 0.1414 | 3.0 | 14000 | 0.3064 |
Framework versions
- Transformers 4.37.2
- Pytorch 2.1.2+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1