bart-cnndm
This model is a fine-tuned version of facebook/bart-base on the cnn_dailymail dataset. It achieves the following results on the evaluation set:
- Loss: 1.6305
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: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 32
- 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 |
|---|---|---|---|
| 2.0521 | 0.06 | 500 | 1.8483 |
| 2.0187 | 0.11 | 1000 | 1.7939 |
| 1.9884 | 0.17 | 1500 | 1.7849 |
| 2.0118 | 0.22 | 2000 | 1.7372 |
| 1.9341 | 0.28 | 2500 | 1.7180 |
| 1.8866 | 0.33 | 3000 | 1.7186 |
| 1.9491 | 0.39 | 3500 | 1.6971 |
| 1.8668 | 0.45 | 4000 | 1.6930 |
| 1.9666 | 0.5 | 4500 | 1.6570 |
| 1.9386 | 0.56 | 5000 | 1.6703 |
| 1.9207 | 0.61 | 5500 | 1.6570 |
| 1.876 | 0.67 | 6000 | 1.6571 |
| 1.9118 | 0.72 | 6500 | 1.6541 |
| 1.8098 | 0.78 | 7000 | 1.6506 |
| 1.8564 | 0.84 | 7500 | 1.6391 |
| 1.8527 | 0.89 | 8000 | 1.6376 |
| 1.7987 | 0.95 | 8500 | 1.6324 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1+cu117
- Datasets 2.10.1
- Tokenizers 0.13.2