raptorkwok's picture
update model card README.md
d72bdca
|
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: 0.2373
  • Bleu: 58.9213
  • Chrf: 57.6665
  • Gen Len: 12.8396

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
0.3565 0.48 1000 0.2624 58.3152 56.9278 12.8539
0.3077 0.96 2000 0.2403 58.4429 57.226 12.8036
0.2297 1.44 3000 0.2469 58.6654 57.3437 12.8374
0.2256 1.92 4000 0.2373 58.9213 57.6665 12.8396
0.1711 2.39 5000 0.2427 58.8291 57.4604 12.8506
0.1694 2.87 6000 0.2500 58.4154 57.0752 12.813
0.1336 3.35 7000 0.2575 58.4311 57.0237 12.8415

Framework versions

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