README.md

metadata

license: apache-2.0
tags:
  - generated_from_trainer
datasets:
  - wmt14
metrics:
  - bleu
model-index:
  - name: t5-small-finetuned-de-en-wd-01
    results:
      - task:
          name: Sequence-to-sequence Language Modeling
          type: text2text-generation
        dataset:
          name: wmt14
          type: wmt14
          args: de-en
        metrics:
          - name: Bleu
            type: bleu
            value: 9.6027

t5-small-finetuned-de-en-wd-01

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

• Loss: 2.0482
• Bleu: 9.6027
• Gen Len: 17.3776

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: 0.0002
• 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: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
No log	1.0	188	2.0502	9.3675	17.3983
No log	2.0	376	2.0590	9.4393	17.3869
1.6509	3.0	564	2.0639	9.3886	17.3806
1.6509	4.0	752	2.0498	9.5802	17.3846
1.6509	5.0	940	2.0482	9.6027	17.3776

Framework versions

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