exp1_10partition_modelo_asl6000

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

• Loss: 0.1417
• Model Preparation Time: 0.0033
• Bleu Msl: 0
• Bleu 1 Msl: 0
• Bleu 2 Msl: 0
• Bleu 3 Msl: 0
• Bleu 4 Msl: 0
• Ter Msl: 100
• Bleu Asl: 0
• Bleu 1 Asl: 0.9679
• Bleu 2 Asl: 0.9488
• Bleu 3 Asl: 0.9263
• Bleu 4 Asl: 0.8977
• Ter Asl: 3.8971

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: 0.0001
• train_batch_size: 32
• eval_batch_size: 64
• 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: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Model Preparation Time	Bleu Msl	Bleu 1 Msl	Bleu 2 Msl	Bleu 3 Msl	Bleu 4 Msl	Ter Msl	Bleu Asl	Bleu 1 Asl	Bleu 2 Asl	Bleu 3 Asl	Bleu 4 Asl	Ter Asl
No log	1.0	150	0.1841	0.0033	0	0	0	0	0	100	0	0.9454	0.9111	0.8754	0.8365	6.7797
No log	2.0	300	0.1321	0.0033	0	0	0	0	0	100	0	0.9637	0.9393	0.9134	0.8832	4.5078
No log	3.0	450	0.1157	0.0033	0	0	0	0	0	100	0	0.9653	0.9414	0.9167	0.8888	4.3996
0.2566	4.0	600	0.1099	0.0033	0	0	0	0	0	100	0	0.9657	0.9435	0.9195	0.8918	4.0750
0.2566	5.0	750	0.1093	0.0033	0	0	0	0	0	100	0	0.9694	0.9491	0.9275	0.9017	3.6783
0.2566	6.0	900	0.1180	0.0033	0	0	0	0	0	100	0	0.9627	0.9405	0.9170	0.8890	4.3996
0.0331	7.0	1050	0.1049	0.0033	0	0	0	0	0	100	0	0.9722	0.9520	0.9313	0.9071	3.5341
0.0331	8.0	1200	0.1063	0.0033	0	0	0	0	0	100	0	0.9668	0.9479	0.9273	0.9027	3.8586
0.0331	9.0	1350	0.1111	0.0033	0	0	0	0	0	100	0	0.9710	0.9512	0.9294	0.9042	3.6062
0.0165	10.0	1500	0.1093	0.0033	0	0	0	0	0	100	0	0.9713	0.9518	0.9306	0.9050	3.5701
0.0165	11.0	1650	0.1173	0.0033	0	0	0	0	0	100	0	0.9697	0.9490	0.9257	0.8989	3.7865
0.0165	12.0	1800	0.1140	0.0033	0	0	0	0	0	100	0	0.9720	0.9531	0.9320	0.9067	3.4259
0.0165	13.0	1950	0.1147	0.0033	0	0	0	0	0	100	0	0.9544	0.9327	0.9095	0.8812	5.4814
0.0094	14.0	2100	0.1118	0.0033	0	0	0	0	0	100	0	0.9700	0.9493	0.9273	0.9023	3.7144
0.0094	15.0	2250	0.1118	0.0033	0	0	0	0	0	100	0	0.9646	0.9440	0.9217	0.8953	4.2914
0.0094	16.0	2400	0.1099	0.0033	0	0	0	0	0	100	0	0.9701	0.9506	0.9293	0.9043	3.5341
0.0057	17.0	2550	0.1165	0.0033	0	0	0	0	0	100	0	0.9710	0.9520	0.9305	0.9051	3.6423
0.0057	18.0	2700	0.1135	0.0033	0	0	0	0	0	100	0	0.9682	0.9486	0.9271	0.9021	3.8226
0.0057	19.0	2850	0.1165	0.0033	0	0	0	0	0	100	0	0.9710	0.9510	0.9294	0.9046	3.4620
0.0039	20.0	3000	0.1169	0.0033	0	0	0	0	0	100	0	0.9659	0.9466	0.9255	0.9003	4.0389
0.0039	21.0	3150	0.1172	0.0033	0	0	0	0	0	100	0	0.9712	0.9512	0.9289	0.9031	3.6423
0.0039	22.0	3300	0.1194	0.0033	0	0	0	0	0	100	0	0.9685	0.9479	0.9250	0.8984	3.8586
0.0039	23.0	3450	0.1242	0.0033	0	0	0	0	0	100	0	0.9688	0.9477	0.9246	0.8982	3.8226
0.0027	24.0	3600	0.1178	0.0033	0	0	0	0	0	100	0	0.9695	0.9489	0.9265	0.9010	3.6062
0.0027	25.0	3750	0.1177	0.0033	0	0	0	0	0	100	0	0.9698	0.9495	0.9268	0.9006	3.6062
0.0027	26.0	3900	0.1154	0.0033	0	0	0	0	0	100	0	0.9683	0.9482	0.9260	0.9002	3.7144
0.0014	27.0	4050	0.1147	0.0033	0	0	0	0	0	100	0	0.9700	0.9493	0.9266	0.9006	3.6783
0.0014	28.0	4200	0.1123	0.0033	0	0	0	0	0	100	0	0.9710	0.9514	0.9300	0.9053	3.4980
0.0014	29.0	4350	0.1123	0.0033	0	0	0	0	0	100	0	0.9716	0.9525	0.9316	0.9072	3.3898
0.0016	30.0	4500	0.1124	0.0033	0	0	0	0	0	100	0	0.9713	0.9518	0.9304	0.9057	3.4620

Framework versions

• Transformers 4.50.2
• Pytorch 2.6.0+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1