metadata
license: apache-2.0
base_model: facebook/bart-base
tags:
- generated_from_trainer
datasets:
- xsum
model-index:
- name: finetuned-bart-xsum
results: []
finetuned-bart-xsum
This model is a fine-tuned version of facebook/bart-base on the xsum dataset. It achieves the following results on the evaluation set:
- Loss: 0.4374
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: 10
- eval_batch_size: 10
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 3
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.8225 | 0.56 | 1000 | 0.4618 |
| 0.4988 | 1.11 | 2000 | 0.4510 |
| 0.4358 | 1.67 | 3000 | 0.4439 |
| 0.4073 | 2.22 | 4000 | 0.4435 |
| 0.3748 | 2.78 | 5000 | 0.4392 |
Framework versions
- Transformers 4.38.2
- Pytorch 2.1.0+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2