bart-base-summarization-medical-49
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: 2.1283
- Rouge1: 0.4194
- Rouge2: 0.2246
- Rougel: 0.3563
- Rougelsum: 0.356
- Gen Len: 18.24
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: 49
- 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 |
---|---|---|---|---|---|---|---|---|
2.7018 | 1.0 | 1250 | 2.1985 | 0.4123 | 0.2198 | 0.352 | 0.3522 | 17.961 |
2.6001 | 2.0 | 2500 | 2.1649 | 0.4125 | 0.2205 | 0.3526 | 0.3526 | 17.963 |
2.577 | 3.0 | 3750 | 2.1418 | 0.4189 | 0.222 | 0.3547 | 0.3548 | 18.185 |
2.5295 | 4.0 | 5000 | 2.1347 | 0.4213 | 0.2256 | 0.3564 | 0.3559 | 18.174 |
2.5513 | 5.0 | 6250 | 2.1299 | 0.4174 | 0.2224 | 0.3545 | 0.3542 | 18.118 |
2.5347 | 6.0 | 7500 | 2.1283 | 0.4194 | 0.2246 | 0.3563 | 0.356 | 18.24 |
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-49
Base model
facebook/bart-base
Adapter
this model