metadata
license: apache-2.0
base_model: facebook/bart-base
tags:
- generated_from_trainer
model-index:
- name: pubmed-abs-sub-03
results: []
pubmed-abs-sub-03
This model is a fine-tuned version of facebook/bart-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.1905
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: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 10
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.53 | 0.11 | 500 | 0.4647 |
| 0.4464 | 0.21 | 1000 | 0.3745 |
| 0.4506 | 0.32 | 1500 | 0.3262 |
| 0.3944 | 0.43 | 2000 | 0.3019 |
| 0.3538 | 0.54 | 2500 | 0.2816 |
| 0.2626 | 0.64 | 3000 | 0.2692 |
| 0.2607 | 0.75 | 3500 | 0.2540 |
| 0.2967 | 0.86 | 4000 | 0.2357 |
| 0.2716 | 0.96 | 4500 | 0.2334 |
| 0.2065 | 1.07 | 5000 | 0.2286 |
| 0.19 | 1.18 | 5500 | 0.2271 |
| 0.1976 | 1.28 | 6000 | 0.2247 |
| 0.2223 | 1.39 | 6500 | 0.2164 |
| 0.2229 | 1.5 | 7000 | 0.2123 |
| 0.2018 | 1.61 | 7500 | 0.2106 |
| 0.1857 | 1.71 | 8000 | 0.2037 |
| 0.22 | 1.82 | 8500 | 0.2033 |
| 0.1793 | 1.93 | 9000 | 0.1993 |
| 0.1441 | 2.03 | 9500 | 0.2012 |
| 0.1515 | 2.14 | 10000 | 0.2011 |
| 0.1412 | 2.25 | 10500 | 0.2023 |
| 0.1505 | 2.35 | 11000 | 0.1978 |
| 0.1472 | 2.46 | 11500 | 0.1961 |
| 0.1526 | 2.57 | 12000 | 0.1916 |
| 0.1454 | 2.68 | 12500 | 0.1919 |
| 0.1011 | 2.78 | 13000 | 0.1920 |
| 0.1386 | 2.89 | 13500 | 0.1915 |
| 0.1368 | 3.0 | 14000 | 0.1905 |
Framework versions
- Transformers 4.34.1
- Pytorch 2.1.0
- Datasets 2.14.6
- Tokenizers 0.14.1