Edit model card

HealthScienceBARTMainSections

This model is a fine-tuned version of facebook/bart-large-cnn on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 3.8753
  • Rouge1: 57.7808
  • Rouge2: 23.8942
  • Rougel: 42.2385
  • Rougelsum: 54.2817
  • Bertscore Precision: 83.4992
  • Bertscore Recall: 85.0665
  • Bertscore F1: 84.2729
  • Bleu: 0.1888
  • Gen Len: 234.2871

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: 5e-05
  • train_batch_size: 1
  • eval_batch_size: 1
  • seed: 42
  • gradient_accumulation_steps: 16
  • total_train_batch_size: 16
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 500
  • num_epochs: 1

Training results

Training Loss Epoch Step Validation Loss Rouge1 Rouge2 Rougel Rougelsum Bertscore Precision Bertscore Recall Bertscore F1 Bleu Gen Len
5.7855 0.0835 100 5.6539 48.5675 17.0972 33.6039 45.6212 80.1089 82.0896 81.0838 0.1318 234.2871
5.2884 0.1671 200 5.1333 51.1112 18.5297 35.438 47.6657 80.5123 82.7897 81.6311 0.1454 234.2871
4.9931 0.2506 300 4.8074 52.1485 19.5231 36.7572 48.8489 81.0844 83.2739 82.1612 0.1534 234.2871
4.6538 0.3342 400 4.5657 52.6901 20.3073 37.8168 49.4253 81.3851 83.5259 82.438 0.1580 234.2871
4.435 0.4177 500 4.3963 54.6727 20.9559 39.1501 51.7226 82.5534 83.8521 83.1954 0.1618 234.2871
4.4327 0.5013 600 4.2367 55.3497 21.849 39.8267 51.9765 82.6498 84.2237 83.4265 0.1698 234.2871
4.2704 0.5848 700 4.1312 56.2031 22.5317 40.6962 52.8736 82.966 84.4887 83.7178 0.1762 234.2871
4.2211 0.6684 800 4.0373 56.1405 22.9558 41.2482 52.772 82.9397 84.6224 83.7695 0.1800 234.2871
4.0727 0.7519 900 3.9672 57.5881 23.6311 41.712 53.9676 83.2595 84.8583 84.0486 0.1850 234.2871
4.0741 0.8355 1000 3.9182 57.2156 23.6916 42.074 53.7327 83.3537 84.9605 84.1466 0.1868 234.2871
3.8563 0.9190 1100 3.8753 57.7808 23.8942 42.2385 54.2817 83.4992 85.0665 84.2729 0.1888 234.2871

Framework versions

  • Transformers 4.41.2
  • Pytorch 2.3.1+cu121
  • Datasets 2.20.0
  • Tokenizers 0.19.1
Downloads last month
2
Safetensors
Model size
406M params
Tensor type
F32
·
Inference API
This model can be loaded on Inference API (serverless).

Finetuned from