CeLLaTe3.0_Base_no_vague_pubmed
This model is a fine-tuned version of microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract-fulltext on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.1384
- Precision: 0.7700
- Recall: 0.8469
- F1: 0.8066
- Accuracy: 0.9659
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: 2e-05
- train_batch_size: 32
- eval_batch_size: 16
- seed: 3407
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 20
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 1.4387 | 1.0 | 96 | 0.5374 | 0.0 | 0.0 | 0.0 | 0.8636 |
| 0.3318 | 2.0 | 192 | 0.1596 | 0.6288 | 0.7567 | 0.6868 | 0.9582 |
| 0.1398 | 3.0 | 288 | 0.1472 | 0.6731 | 0.7479 | 0.7085 | 0.9601 |
| 0.0908 | 4.0 | 384 | 0.1394 | 0.7700 | 0.8469 | 0.8066 | 0.9659 |
| 0.0657 | 5.0 | 480 | 0.1599 | 0.7181 | 0.7852 | 0.7502 | 0.9625 |
| 0.0511 | 6.0 | 576 | 0.1721 | 0.6924 | 0.8347 | 0.7569 | 0.9584 |
| 0.0398 | 7.0 | 672 | 0.1592 | 0.7721 | 0.8440 | 0.8064 | 0.9665 |
Framework versions
- Transformers 4.48.2
- Pytorch 2.4.1+cu121
- Datasets 3.0.2
- Tokenizers 0.21.0
- Downloads last month
- 2