metadata
license: apache-2.0
base_model: facebook/bart-base
tags:
- generated_from_trainer
model-index:
- name: pubmed-abs-sub-01
results: []
pubmed-abs-sub-01
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: 0.2157
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 |
|---|---|---|---|
| 0.6234 | 0.11 | 500 | 0.5441 |
| 0.4953 | 0.21 | 1000 | 0.4205 |
| 0.4357 | 0.32 | 1500 | 0.3782 |
| 0.4194 | 0.43 | 2000 | 0.3422 |
| 0.364 | 0.54 | 2500 | 0.3211 |
| 0.3064 | 0.64 | 3000 | 0.2959 |
| 0.3169 | 0.75 | 3500 | 0.2859 |
| 0.3169 | 0.86 | 4000 | 0.2703 |
| 0.3033 | 0.96 | 4500 | 0.2622 |
| 0.2401 | 1.07 | 5000 | 0.2635 |
| 0.2253 | 1.18 | 5500 | 0.2564 |
| 0.2214 | 1.28 | 6000 | 0.2571 |
| 0.2315 | 1.39 | 6500 | 0.2424 |
| 0.2524 | 1.5 | 7000 | 0.2386 |
| 0.2257 | 1.61 | 7500 | 0.2374 |
| 0.1949 | 1.71 | 8000 | 0.2344 |
| 0.2344 | 1.82 | 8500 | 0.2286 |
| 0.1896 | 1.93 | 9000 | 0.2265 |
| 0.1722 | 2.03 | 9500 | 0.2266 |
| 0.1697 | 2.14 | 10000 | 0.2293 |
| 0.1558 | 2.25 | 10500 | 0.2265 |
| 0.154 | 2.35 | 11000 | 0.2231 |
| 0.1476 | 2.46 | 11500 | 0.2205 |
| 0.1715 | 2.57 | 12000 | 0.2190 |
| 0.1735 | 2.68 | 12500 | 0.2175 |
| 0.1319 | 2.78 | 13000 | 0.2170 |
| 0.1556 | 2.89 | 13500 | 0.2163 |
| 0.1573 | 3.0 | 14000 | 0.2157 |
Framework versions
- Transformers 4.34.1
- Pytorch 2.1.0
- Datasets 2.14.6
- Tokenizers 0.14.1