metadata
base_model: facebook/bart-base
library_name: peft
license: apache-2.0
metrics:
- rouge
tags:
- generated_from_trainer
model-index:
- name: bart-base-summarization-medical_on_cnn-42
results: []
bart-base-summarization-medical_on_cnn-42
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: 3.3832
- Rouge1: 0.2502
- Rouge2: 0.0919
- Rougel: 0.1972
- Rougelsum: 0.2211
- Gen Len: 18.61
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: 42
- 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 |
|---|---|---|---|---|---|---|---|---|
| 2.7168 | 1.0 | 1250 | 3.3726 | 0.2491 | 0.0895 | 0.1955 | 0.221 | 18.952 |
| 2.6026 | 2.0 | 2500 | 3.3663 | 0.2511 | 0.092 | 0.1967 | 0.2219 | 18.8 |
| 2.5707 | 3.0 | 3750 | 3.3707 | 0.2505 | 0.0921 | 0.1967 | 0.2207 | 18.618 |
| 2.5606 | 4.0 | 5000 | 3.3795 | 0.2518 | 0.093 | 0.1981 | 0.222 | 18.687 |
| 2.5437 | 5.0 | 6250 | 3.3853 | 0.2517 | 0.0928 | 0.1975 | 0.2219 | 18.521 |
| 2.5435 | 6.0 | 7500 | 3.3832 | 0.2502 | 0.0919 | 0.1972 | 0.2211 | 18.61 |
Framework versions
- PEFT 0.12.0
- Transformers 4.42.4
- Pytorch 2.3.1+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1