bart-dnc-booksum
This model is a fine-tuned version of facebook/bart-large on the None dataset. It achieves the following results on the evaluation set:
- Loss: 6.5809
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: 1
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 40
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.7554 | 0.16 | 150 | 4.3521 |
| 0.3115 | 0.32 | 300 | 4.9631 |
| 0.3536 | 0.48 | 450 | 4.8518 |
| 0.3419 | 0.64 | 600 | 4.8791 |
| 0.2082 | 0.8 | 750 | 5.8989 |
| 0.3123 | 0.97 | 900 | 4.8499 |
| 0.1517 | 1.13 | 1050 | 5.3516 |
| 0.2184 | 1.29 | 1200 | 5.5229 |
| 0.217 | 1.45 | 1350 | 5.7270 |
| 0.1375 | 1.61 | 1500 | 5.5408 |
| 0.1506 | 1.77 | 1650 | 5.3878 |
| 0.1904 | 1.93 | 1800 | 5.0405 |
| 0.0956 | 2.09 | 1950 | 6.0605 |
| 0.0856 | 2.25 | 2100 | 6.1505 |
| 0.1206 | 2.41 | 2250 | 5.4903 |
| 0.0872 | 2.58 | 2400 | 5.4324 |
| 0.1456 | 2.74 | 2550 | 6.1430 |
| 0.0848 | 2.9 | 2700 | 5.3999 |
| 0.079 | 3.06 | 2850 | 6.0992 |
| 0.0706 | 3.22 | 3000 | 6.2144 |
| 0.0768 | 3.38 | 3150 | 6.2338 |
| 0.0643 | 3.54 | 3300 | 6.0533 |
| 0.0576 | 3.7 | 3450 | 5.9912 |
| 0.0859 | 3.86 | 3600 | 6.3768 |
| 0.0615 | 4.02 | 3750 | 6.1106 |
| 0.072 | 4.19 | 3900 | 6.5398 |
| 0.0683 | 4.35 | 4050 | 6.5556 |
| 0.077 | 4.51 | 4200 | 6.6243 |
| 0.0811 | 4.67 | 4350 | 6.5221 |
| 0.0591 | 4.83 | 4500 | 6.5732 |
| 0.0699 | 4.99 | 4650 | 6.5809 |
Framework versions
- Transformers 4.28.0
- Pytorch 2.0.1+cu118
- Datasets 2.14.4
- Tokenizers 0.13.3
- Downloads last month
- 10
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.