metadata
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- samsum
metrics:
- rouge
model-index:
- name: bart-samsum
results:
- task:
name: Sequence-to-sequence Language Modeling
type: text2text-generation
dataset:
name: samsum
type: samsum
config: samsum
split: train
args: samsum
metrics:
- name: Rouge1
type: rouge
value: 0.4835
bart-samsum
This model is a fine-tuned version of facebook/bart-base on the samsum dataset. It achieves the following results on the evaluation set:
- Loss: 1.5071
- Rouge1: 0.4835
- Rouge2: 0.2546
- Rougel: 0.4128
- Rougelsum: 0.4131
- Gen Len: 17.9817
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: 2e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| 1.8082 | 1.0 | 2947 | 1.5613 | 0.4763 | 0.2412 | 0.4043 | 0.4041 | 17.9332 |
| 1.5609 | 2.0 | 5894 | 1.5206 | 0.4827 | 0.2485 | 0.4082 | 0.4085 | 18.3169 |
| 1.4228 | 3.0 | 8841 | 1.5008 | 0.4851 | 0.2557 | 0.4138 | 0.4137 | 17.9851 |
| 1.3131 | 4.0 | 11788 | 1.5071 | 0.4835 | 0.2546 | 0.4128 | 0.4131 | 17.9817 |
Framework versions
- Transformers 4.28.1
- Pytorch 2.0.0+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3