raptorkwok's picture
update model card README.md
c91ccb9
|
raw
history blame
1.95 kB
metadata
tags:
  - generated_from_trainer
metrics:
  - bleu
model-index:
  - name: cantonese-chinese-translation
    results: []

cantonese-chinese-translation

This model is a fine-tuned version of fnlp/bart-base-chinese on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.2349
  • Bleu: 60.8976
  • Chrf: 59.0374
  • Gen Len: 12.9296

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
0.3535 0.48 1000 0.2530 60.3526 58.5789 12.8843
0.3069 0.96 2000 0.2452 60.6498 58.8383 12.9006
0.2299 1.44 3000 0.2389 60.6464 58.9253 12.8851
0.2203 1.92 4000 0.2349 60.8976 59.0374 12.9296
0.1745 2.39 5000 0.2407 60.7308 58.8719 12.9298
0.1658 2.87 6000 0.2373 60.3476 58.6283 12.8618
0.1351 3.35 7000 0.2589 60.7838 58.8847 12.9313

Framework versions

  • Transformers 4.28.1
  • Pytorch 2.1.0+cu121
  • Datasets 2.14.6
  • Tokenizers 0.13.3