MarianMT-finetuned-ES-ZAP-v3

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: 0.2356
• Bleu: 21.9740
• Meteor: 0.4674
• Ter: 62.6908
• Chrf: 45.9496

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_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Meteor	Ter	Chrf
1.3736	1.0	179	0.5978	0.3111	0.0882	216.7939	9.2974
0.4949	2.0	358	0.4382	3.6416	0.1889	95.1336	20.9865
0.3762	3.0	537	0.3719	6.2974	0.2275	88.7405	24.9969
0.3149	4.0	716	0.3382	8.2395	0.2822	84.8282	29.5532
0.2765	5.0	895	0.3156	8.4096	0.3200	79.4847	32.3501
0.2462	6.0	1074	0.2968	11.4876	0.3457	75.8588	35.1628
0.2217	7.0	1253	0.2841	13.5969	0.3650	77.0992	36.6277
0.2013	8.0	1432	0.2731	15.4573	0.3967	71.3740	38.5145
0.1842	9.0	1611	0.2638	16.6958	0.4135	70.1336	40.9340
0.1690	10.0	1790	0.2573	17.4951	0.4206	67.4618	41.4994
0.1560	11.0	1969	0.2531	18.0435	0.4313	67.6527	41.9272
0.1446	12.0	2148	0.2469	20.5832	0.4548	64.5992	43.7279
0.1331	13.0	2327	0.2449	18.6926	0.4423	66.6985	43.3962
0.1249	14.0	2506	0.2419	20.7880	0.4566	63.6450	44.3923
0.1166	15.0	2685	0.2401	21.5664	0.4608	61.7366	44.9293
0.1082	16.0	2864	0.2379	20.8974	0.4626	62.6908	45.3873
0.1019	17.0	3043	0.2356	21.9740	0.4674	62.6908	45.9496
0.0949	18.0	3222	0.2361	21.5536	0.4704	62.5	46.4209

Framework versions

• Transformers 5.0.0
• Pytorch 2.10.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.2