metadata
base_model: facebook/bart-large
library_name: peft
license: apache-2.0
metrics:
- rouge
tags:
- generated_from_trainer
model-index:
- name: bart-large-summarization-medical-46
results: []
bart-large-summarization-medical-46
This model is a fine-tuned version of facebook/bart-large on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.8378
- Rouge1: 0.4404
- Rouge2: 0.2412
- Rougel: 0.3768
- Rougelsum: 0.3769
- Gen Len: 18.977
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 |
|---|---|---|---|---|---|---|---|---|
| 2.2273 | 1.0 | 1250 | 1.9018 | 0.4342 | 0.2347 | 0.3676 | 0.3677 | 19.319 |
| 2.1445 | 2.0 | 2500 | 1.8668 | 0.4394 | 0.2388 | 0.3744 | 0.3743 | 18.977 |
| 2.0968 | 3.0 | 3750 | 1.8556 | 0.4406 | 0.2411 | 0.3767 | 0.3769 | 18.689 |
| 2.0883 | 4.0 | 5000 | 1.8502 | 0.4398 | 0.2391 | 0.3758 | 0.376 | 18.757 |
| 2.0638 | 5.0 | 6250 | 1.8393 | 0.4416 | 0.2406 | 0.3779 | 0.3777 | 18.88 |
| 2.0453 | 6.0 | 7500 | 1.8378 | 0.4404 | 0.2412 | 0.3768 | 0.3769 | 18.977 |
Framework versions
- PEFT 0.11.1
- Transformers 4.42.4
- Pytorch 2.3.1+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1