en2zh_opus_100 / README.md
HAJIWEE's picture
update model card README.md
f953c7e
metadata
license: apache-2.0
tags:
  - generated_from_trainer
datasets:
  - opus100
metrics:
  - bleu
model-index:
  - name: en2zh_opus_100
    results:
      - task:
          name: Sequence-to-sequence Language Modeling
          type: text2text-generation
        dataset:
          name: opus100
          type: opus100
          config: en-zh
          split: test
          args: en-zh
        metrics:
          - name: Bleu
            type: bleu
            value: 12.9203

en2zh_opus_100

This model is a fine-tuned version of HAJIWEE/en2zh_opus_100 on the opus100 dataset. It achieves the following results on the evaluation set:

  • Loss: 2.0799
  • Bleu: 12.9203
  • Gen Len: 15.2965

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: 6
  • eval_batch_size: 6
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 2

Training results

Training Loss Epoch Step Validation Loss Bleu Gen Len
2.6252 1.0 166667 2.1221 12.7341 15.3295
2.5579 2.0 333334 2.0799 12.9203 15.2965

Framework versions

  • Transformers 4.27.4
  • Pytorch 2.0.0+cu117
  • Datasets 2.11.0
  • Tokenizers 0.13.3