README.md

metadata

license: apache-2.0
tags:
  - translation
  - generated_from_trainer
datasets:
  - news_commentary
metrics:
  - bleu
model-index:
  - name: pt-opus-news
    results:
      - task:
          name: Sequence-to-sequence Language Modeling
          type: text2text-generation
        dataset:
          name: news_commentary
          type: news_commentary
          config: en-pt
          split: train
          args: en-pt
        metrics:
          - name: Bleu
            type: bleu
            value: 37.5501808262607

pt-opus-news

This model is a fine-tuned version of Helsinki-NLP/opus-mt-en-mul on the news_commentary dataset. It achieves the following results on the evaluation set:

• Loss: 1.0975
• Bleu: 37.5502

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: 32
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3
• mixed_precision_training: Native AMP

Training results

Framework versions

• Transformers 4.22.0
• Pytorch 1.12.1+cu113
• Datasets 2.4.0
• Tokenizers 0.12.1