marian-finetuned-kde4-en-to-zh

This model is a fine-tuned version of Helsinki-NLP/opus-mt-en-zh on the kde4 dataset.

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: 32
• eval_batch_size: 64
• 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
• num_epochs: 3

Training results

{'eval_loss': 2.487135171890259, 'eval_model_preparation_time': 0.0014, 'eval_bleu': 28.03812027146028, 'eval_runtime': 5952.7499, 'eval_samples_per_second': 2.346, 'eval_steps_per_second': 0.037, 'epoch': 0}

{'eval_loss': 0.9341261982917786, 'eval_model_preparation_time': 0.0014, 'eval_bleu': 41.49392304308398, 'eval_runtime': 4947.72, 'eval_samples_per_second': 2.823, 'eval_steps_per_second': 0.044, 'epoch': 3.0}

Framework versions

• Transformers 5.4.0
• Pytorch 2.11.0
• Datasets 2.21.0
• Tokenizers 0.22.2