bart-base-summarization-medical_on_cnn-47
This model is a fine-tuned version of facebook/bart-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 3.3826
- Rouge1: 0.2498
- Rouge2: 0.0937
- Rougel: 0.1992
- Rougelsum: 0.2218
- Gen Len: 18.217
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: 4
- eval_batch_size: 1
- seed: 47
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 6
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| 2.7195 | 1.0 | 1250 | 3.3721 | 0.2478 | 0.0885 | 0.1948 | 0.219 | 18.949 |
| 2.6054 | 2.0 | 2500 | 3.3830 | 0.251 | 0.0932 | 0.1977 | 0.2222 | 18.61 |
| 2.572 | 3.0 | 3750 | 3.3801 | 0.251 | 0.092 | 0.1978 | 0.2222 | 18.472 |
| 2.5529 | 4.0 | 5000 | 3.3811 | 0.2495 | 0.0927 | 0.1992 | 0.2215 | 18.222 |
| 2.5453 | 5.0 | 6250 | 3.3833 | 0.2496 | 0.0918 | 0.1983 | 0.2219 | 18.318 |
| 2.5151 | 6.0 | 7500 | 3.3826 | 0.2498 | 0.0937 | 0.1992 | 0.2218 | 18.217 |
Framework versions
- PEFT 0.12.0
- Transformers 4.42.4
- Pytorch 2.3.1+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1
- Downloads last month
- 1
Model tree for zbigi/bart-base-summarization-medical_on_cnn-47
Base model
facebook/bart-base