|
--- |
|
license: apache-2.0 |
|
base_model: facebook/bart-large |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- rouge |
|
- wer |
|
model-index: |
|
- name: bart_extractive_1024_750 |
|
results: [] |
|
--- |
|
|
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You |
|
should probably proofread and complete it, then remove this comment. --> |
|
|
|
# bart_extractive_1024_750 |
|
|
|
This model is a fine-tuned version of [facebook/bart-large](https://huggingface.co./facebook/bart-large) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.8901 |
|
- Rouge1: 0.7176 |
|
- Rouge2: 0.4726 |
|
- Rougel: 0.6632 |
|
- Rougelsum: 0.6633 |
|
- Wer: 0.4177 |
|
|
|
## 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: 6 |
|
- eval_batch_size: 6 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 2 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Wer | |
|
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|:---------:|:------:| |
|
| No log | 0.13 | 250 | 1.1639 | 0.6758 | 0.4064 | 0.6138 | 0.6136 | 0.4827 | |
|
| 2.044 | 0.27 | 500 | 1.0693 | 0.6853 | 0.4267 | 0.6258 | 0.6256 | 0.4594 | |
|
| 2.044 | 0.4 | 750 | 1.0210 | 0.6982 | 0.4409 | 0.6399 | 0.6399 | 0.452 | |
|
| 1.1195 | 0.53 | 1000 | 0.9865 | 0.6989 | 0.4442 | 0.64 | 0.64 | 0.4449 | |
|
| 1.1195 | 0.66 | 1250 | 0.9697 | 0.7007 | 0.4476 | 0.643 | 0.6429 | 0.4407 | |
|
| 1.0531 | 0.8 | 1500 | 0.9680 | 0.7009 | 0.4495 | 0.6451 | 0.645 | 0.4384 | |
|
| 1.0531 | 0.93 | 1750 | 0.9346 | 0.7099 | 0.4587 | 0.6538 | 0.6539 | 0.4323 | |
|
| 1.0109 | 1.06 | 2000 | 0.9249 | 0.7066 | 0.4589 | 0.6519 | 0.6518 | 0.4295 | |
|
| 1.0109 | 1.2 | 2250 | 0.9221 | 0.7092 | 0.4627 | 0.6541 | 0.654 | 0.427 | |
|
| 0.9199 | 1.33 | 2500 | 0.9117 | 0.7134 | 0.4668 | 0.6583 | 0.6582 | 0.424 | |
|
| 0.9199 | 1.46 | 2750 | 0.9064 | 0.7147 | 0.4676 | 0.6593 | 0.6592 | 0.4225 | |
|
| 0.9164 | 1.6 | 3000 | 0.8996 | 0.7164 | 0.4701 | 0.6612 | 0.6611 | 0.4212 | |
|
| 0.9164 | 1.73 | 3250 | 0.9006 | 0.714 | 0.4695 | 0.6602 | 0.6601 | 0.4201 | |
|
| 0.8861 | 1.86 | 3500 | 0.8893 | 0.7176 | 0.4735 | 0.6635 | 0.6635 | 0.4176 | |
|
| 0.8861 | 1.99 | 3750 | 0.8901 | 0.7176 | 0.4726 | 0.6632 | 0.6633 | 0.4177 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.38.2 |
|
- Pytorch 2.2.1+cu121 |
|
- Datasets 2.18.0 |
|
- Tokenizers 0.15.2 |
|
|