Edit model card

bart-base-summarization-medical-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: 2.1308
  • Rouge1: 0.4191
  • Rouge2: 0.2227
  • Rougel: 0.3552
  • Rougelsum: 0.3555
  • Gen Len: 18.216

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
2.7197 1.0 1250 2.2012 0.4139 0.2191 0.3517 0.3522 17.835
2.6119 2.0 2500 2.1664 0.416 0.2218 0.3517 0.3519 18.075
2.5724 3.0 3750 2.1502 0.4162 0.2197 0.3522 0.3526 18.298
2.548 4.0 5000 2.1390 0.4144 0.2197 0.3509 0.3511 18.138
2.5441 5.0 6250 2.1296 0.4199 0.224 0.3558 0.356 18.267
2.5115 6.0 7500 2.1308 0.4191 0.2227 0.3552 0.3555 18.216

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
6
Inference API
Unable to determine this model’s pipeline type. Check the docs .

Model tree for zbigi/bart-base-summarization-medical-47

Base model

facebook/bart-base
Adapter
(31)
this model