t5-small-finetuned-de-en-256-epochs2

This model is a fine-tuned version of t5-small on the wmt14 dataset. It achieves the following results on the evaluation set:

• Loss: 2.1073
• Bleu: 7.8579
• Gen Len: 17.3896

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: 2
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
No log	1.0	188	2.1179	7.8498	17.382
No log	2.0	376	2.1073	7.8579	17.3896

Framework versions

• Transformers 4.12.5
• Pytorch 1.10.0+cu111
• Datasets 1.16.1
• Tokenizers 0.10.3