bart-base-summarization-medical_on_cnn-46
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: 3.3801
- Rouge1: 0.2487
- Rouge2: 0.0923
- Rougel: 0.1972
- Rougelsum: 0.2207
- Gen Len: 18.133
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: 46
- 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 |
|---|---|---|---|---|---|---|---|---|
| 2.7108 | 1.0 | 1250 | 3.3782 | 0.25 | 0.0912 | 0.1969 | 0.2194 | 18.87 |
| 2.6125 | 2.0 | 2500 | 3.3660 | 0.2524 | 0.092 | 0.1978 | 0.2216 | 18.677 |
| 2.57 | 3.0 | 3750 | 3.3810 | 0.2493 | 0.0925 | 0.1967 | 0.2211 | 18.378 |
| 2.5502 | 4.0 | 5000 | 3.3807 | 0.2517 | 0.0938 | 0.1984 | 0.2216 | 18.211 |
| 2.5356 | 5.0 | 6250 | 3.3808 | 0.249 | 0.0923 | 0.197 | 0.2206 | 18.051 |
| 2.5134 | 6.0 | 7500 | 3.3801 | 0.2487 | 0.0923 | 0.1972 | 0.2207 | 18.133 |
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
- 2
Model tree for zbigi/bart-base-summarization-medical_on_cnn-46
Base model
facebook/bart-base