|
--- |
|
license: mit |
|
base_model: facebook/bart-large-xsum |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- rouge |
|
model-index: |
|
- name: LLM_Teached_Bart |
|
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. --> |
|
|
|
# LLM_Teached_Bart |
|
|
|
This model is a fine-tuned version of [facebook/bart-large-xsum](https://huggingface.co./facebook/bart-large-xsum) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 1.7715 |
|
- Rouge1: 0.4781 |
|
- Rouge2: 0.2085 |
|
- Rougel: 0.3718 |
|
- Rougelsum: 0.372 |
|
- Gen Len: 41.3245 |
|
|
|
## 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: 16 |
|
- eval_batch_size: 8 |
|
- 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 | |
|
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|:---------:|:-------:| |
|
| 1.6623 | 1.0 | 1250 | 1.6705 | 0.4681 | 0.2057 | 0.3632 | 0.3631 | 43.4718 | |
|
| 1.2986 | 2.0 | 2500 | 1.6330 | 0.476 | 0.2105 | 0.3732 | 0.3737 | 39.9745 | |
|
| 1.0401 | 3.0 | 3750 | 1.7081 | 0.4792 | 0.2134 | 0.3762 | 0.3763 | 40.6155 | |
|
| 0.8853 | 4.0 | 5000 | 1.7715 | 0.4781 | 0.2085 | 0.3718 | 0.372 | 41.3245 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.36.0 |
|
- Pytorch 2.0.1+cu117 |
|
- Datasets 2.14.5 |
|
- Tokenizers 0.15.0 |
|
|
