<!-- 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. -->
cantonese-chinese-translation-bart-large
This model is a fine-tuned version of fnlp/bart-large-chinese on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.2866
- Bleu: 58.7602
- Chrf: 56.7526
- Gen Len: 12.8968
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: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 30
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Bleu | Chrf | Gen Len |
---|---|---|---|---|---|---|
0.3819 | 0.48 | 1000 | 0.3085 | 57.8706 | 55.6445 | 12.9583 |
0.3257 | 0.96 | 2000 | 0.2866 | 58.7602 | 56.7526 | 12.8968 |
0.2863 | 1.44 | 3000 | 0.4211 | 57.7522 | 55.039 | 13.5412 |
0.2996 | 1.92 | 4000 | 0.2845 | 58.585 | 56.3821 | 12.9442 |
0.1377 | 2.39 | 5000 | 0.3182 | 57.5651 | 55.3076 | 12.9231 |
Framework versions
- Transformers 4.28.1
- Pytorch 2.0.1+cu117
- Datasets 2.14.5
- Tokenizers 0.13.3