README.md

metadata

license: apache-2.0
tags:
  - generated_from_keras_callback
model-index:
  - name: mariav/helsinki-opus-de-en-fine-tuned-wmt16
    results: []
datasets:
  - wmt16
language:
  - de
  - en
metrics:
  - bleu
pipeline_tag: translation

mariav/helsinki-opus-de-en-fine-tuned-wmt16

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

• Train Loss: 1.1917
• Validation Loss: 1.4099
• Epoch: 2

Model description

This model is a fine-tuned version of the Helsinki-NLP/opus-mt-de-en with the dataset wmt16 with the pair of languages de-en.

Intended uses & limitations

Limitations: schoolar purposes.

Training and evaluation data

The training was done with tranformers keras with just a few epochs due to lack of more GPU. With more epochs the performance could be better.

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• optimizer: {'name': 'AdamWeightDecay', 'learning_rate': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 5e-05, 'decay_steps': 747, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.01}
• training_precision: mixed_float16

Training results

Train Loss	Validation Loss	Epoch
1.5114	1.4037	0
1.2991	1.4064	1
1.1917	1.4099	2

Framework versions

• Transformers 4.27.4
• TensorFlow 2.11.0
• Datasets 2.11.0
• Tokenizers 0.13.2