|
--- |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- rouge |
|
model-index: |
|
- name: bart-large-chinese-cnhdwriter |
|
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-large-chinese-cnhdwriter |
|
|
|
This model is a fine-tuned version of [fnlp/bart-large-chinese](https://huggingface.co/fnlp/bart-large-chinese) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 1.7252 |
|
- Rouge1: 15.5844 |
|
- Rouge2: 2.1522 |
|
- Rougel: 15.5443 |
|
- Rougelsum: 15.5603 |
|
- Gen Len: 19.3469 |
|
|
|
## 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 | |
|
|:-------------:|:-----:|:------:|:---------------:|:-------:|:------:|:-------:|:---------:|:-------:| |
|
| 1.2431 | 1.0 | 32000 | 1.1646 | 15.6512 | 2.0244 | 15.6212 | 15.6283 | 18.7906 | |
|
| 0.8739 | 2.0 | 64000 | 1.1694 | 15.5784 | 2.255 | 15.5413 | 15.5504 | 19.2405 | |
|
| 0.6178 | 3.0 | 96000 | 1.2520 | 15.5932 | 2.2286 | 15.5458 | 15.5546 | 19.3576 | |
|
| 0.4016 | 4.0 | 128000 | 1.5014 | 15.3701 | 2.0511 | 15.3595 | 15.3654 | 19.2426 | |
|
| 0.235 | 5.0 | 160000 | 1.7252 | 15.5844 | 2.1522 | 15.5443 | 15.5603 | 19.3469 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.15.0 |
|
- Pytorch 1.10.0+cu111 |
|
- Datasets 1.17.0 |
|
- Tokenizers 0.10.3 |
|
|