longt5_xl_gov_bp_15
This model is a fine-tuned version of /exports/eddie/scratch/s1970716/models/summarization/longt5_xl_gov_bp_10/checkpoint-680 on the learn3r/gov_report_bp dataset. It achieves the following results on the evaluation set:
- Loss: 1.7126
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: 0.001
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 64
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant
- num_epochs: 5.0
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.2051 | 1.0 | 136 | 1.7126 |
| 0.1732 | 1.99 | 272 | 1.8857 |
| 0.1777 | 3.0 | 409 | 1.9036 |
| 0.1122 | 4.0 | 545 | 1.9538 |
| 0.1098 | 4.99 | 680 | 2.1134 |
Framework versions
- Transformers 4.34.1
- Pytorch 2.1.0+cu121
- Datasets 2.14.5
- Tokenizers 0.14.1