led-arxiv-10240
This model is a fine-tuned version of allenai/led-base-16384 on the arxiv-summarization dataset. It achieves the following results on the evaluation set:
- Loss: 2.0623
- Rouge1: 0.4423
- Rouge2: 0.1739
- Rougel: 0.2521
- Rougelsum: 0.4004
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: 1e-05
- train_batch_size: 8
- eval_batch_size: 2
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 1
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
|---|---|---|---|---|---|---|---|
| 2.2112 | 0.39 | 10000 | 2.1196 | 0.4367 | 0.1702 | 0.2468 | 0.3963 |
| 2.1591 | 0.79 | 20000 | 2.0623 | 0.4423 | 0.1739 | 0.2521 | 0.4004 |
Framework versions
- Transformers 4.25.1
- Pytorch 1.13.1+cu116
- Datasets 2.8.0
- Tokenizers 0.13.2