metadata
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- wmt14
model-index:
- name: opus-mt-de-en-finetuned-de-to-en
results: []
opus-mt-de-en-finetuned-de-to-en
This model is a fine-tuned version of Helsinki-NLP/opus-mt-de-en on the wmt14 dataset. It achieves the following results on the evaluation set:
- eval_loss: 1.3411
- eval_bleu: 32.4395
- eval_gen_len: 29.6925
- eval_runtime: 2250.0489
- eval_samples_per_second: 19.998
- eval_steps_per_second: 0.625
- epoch: 3.0
- step: 4221
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-06
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 11
- mixed_precision_training: Native AMP
Framework versions
- Transformers 4.19.2
- Pytorch 1.7.1+cu110
- Datasets 2.2.2
- Tokenizers 0.12.1