metadata
license: mit
tags:
- generated_from_trainer
metrics:
- rouge
model-index:
- name: bart-large-cnn-10k-lit-evalMA-NOpad
results: []
bart-large-cnn-10k-lit-evalMA-NOpad
This model is a fine-tuned version of facebook/bart-large-cnn on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.9464
- Rouge1: 28.6721
- Rouge2: 13.8303
- Rougel: 22.458
- Rougelsum: 25.668
- Gen Len: 66.893
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: 1
- eval_batch_size: 1
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|---|---|---|---|---|---|---|---|---|
| 1.535 | 1.0 | 10000 | 1.7501 | 28.519 | 13.967 | 22.4854 | 25.4511 | 66.555 |
| 0.8754 | 2.0 | 20000 | 1.9464 | 28.6721 | 13.8303 | 22.458 | 25.668 | 66.893 |
Framework versions
- Transformers 4.16.2
- Pytorch 1.10.2
- Datasets 1.18.3
- Tokenizers 0.11.0