gqa-opus-mt-de-en

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

IMNPORTANT: This is an experimental model that uses a custom modified architecture for the MarianMT models. (It has also seen insufficient data to learn properly so will perform badly)

gqa-opus-mt-de-en

This model is a fine-tuned version of sbartlett97/gqa-opus-mt-de-en on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.6134
• Bleu: 0.2252
• Gen Len: 13.0925

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: 5e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 25
• mixed_precision_training: Native AMP

Training results

Framework versions

• Transformers 5.0.0.dev0
• Pytorch 2.9.1+cu128
• Datasets 4.4.1
• Tokenizers 0.22.1