metadata
license: mit
tags:
- generated_from_trainer
metrics:
- rouge
model-index:
- name: bart-base-cnn-xsum-swe
results: []
bart-base-cnn-xsum-swe
This model is a fine-tuned version of Gabriel/bart-base-cnn-swe on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.1140
- Rouge1: 30.7101
- Rouge2: 11.9532
- Rougel: 25.1864
- Rougelsum: 25.2227
- Gen Len: 19.7448
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: 3.75e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- 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: 3
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| 2.3087 | 1.0 | 6375 | 2.1997 | 29.7666 | 11.0222 | 24.2659 | 24.2915 | 19.7172 |
| 2.0793 | 2.0 | 12750 | 2.1285 | 30.4447 | 11.7671 | 24.9238 | 24.9622 | 19.7051 |
| 1.9186 | 3.0 | 19125 | 2.1140 | 30.7101 | 11.9532 | 25.1864 | 25.2227 | 19.7448 |
Framework versions
- Transformers 4.22.1
- Pytorch 1.12.1+cu113
- Datasets 2.5.1
- Tokenizers 0.12.1