led-base-16384-biolaysum-both-wiki_lucene
This model is a fine-tuned version of allenai/led-base-16384 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.1548
- Rouge1: 0.4543
- Rouge2: 0.1550
- Rougel: 0.2425
- Rougelsum: 0.2425
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: 4
- eval_batch_size: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|---|---|---|---|---|---|---|---|
| 2.2704 | 0.69 | 5000 | 2.2177 | 0.4481 | 0.1522 | 0.2414 | 0.2415 |
| 2.0364 | 1.37 | 10000 | 2.1548 | 0.4543 | 0.1550 | 0.2425 | 0.2425 |
| 1.9201 | 2.06 | 15000 | 2.1200 | 0.4551 | 0.1530 | 0.2410 | 0.2410 |
| 1.9007 | 2.75 | 20000 | 2.1060 | 0.4554 | 0.1540 | 0.2400 | 0.2400 |
| 1.7766 | 3.43 | 25000 | 2.0941 | 0.4536 | 0.1513 | 0.2397 | 0.2399 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.13.1
- Datasets 2.10.1
- Tokenizers 0.12.1