MarianMT-finetuned-ES-ZAP-v1

This model is a fine-tuned version of Helsinki-NLP/opus-mt-es-en on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3245
• Bleu: 5.6338
• Meteor: 0.2533
• Ter: 83.8078

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: 1e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Meteor	Ter
1.8463	1.0	102	0.6304	0.6015	0.1084	103.2028
0.573	2.0	204	0.5103	0.6829	0.1172	114.4128
0.4699	3.0	306	0.4442	0.7484	0.1519	120.4626
0.4027	4.0	408	0.4011	1.7813	0.1868	93.7722
0.3578	5.0	510	0.3685	1.8834	0.1952	89.5018
0.3243	6.0	612	0.3504	3.1038	0.2079	88.4342
0.2992	7.0	714	0.3377	4.8564	0.2314	86.1210
0.2852	8.0	816	0.3295	5.3120	0.2357	85.5872
0.2762	9.0	918	0.3259	5.6216	0.2532	83.9858
0.2716	10.0	1020	0.3245	5.6338	0.2533	83.8078

Framework versions

• Transformers 4.48.3
• Pytorch 2.6.0+cu124
• Datasets 3.4.1
• Tokenizers 0.21.1