exp4_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: 0.1530
• Model Preparation Time: 0.0034
• 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: 0.9722
• Bleu 2 Asl: 0.9564
• Bleu 3 Asl: 0.9375
• Bleu 4 Asl: 0.9145
• Ter Asl: 3.2928

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	0.1387	0.0034	0	0	0	0	0	100	0	0.9596	0.9308	0.8999	0.8642	5.1005
No log	2.0	300	0.1095	0.0034	0	0	0	0	0	100	0	0.9698	0.9459	0.9200	0.8888	3.6113
No log	3.0	450	0.1068	0.0034	0	0	0	0	0	100	0	0.9677	0.9438	0.9175	0.8867	3.8347
0.2642	4.0	600	0.1017	0.0034	0	0	0	0	0	100	0	0.9743	0.9538	0.9319	0.9050	2.9412
0.2642	5.0	750	0.1086	0.0034	0	0	0	0	0	100	0	0.9680	0.9459	0.9221	0.8928	2.9784
0.2642	6.0	900	0.1100	0.0034	0	0	0	0	0	100	0	0.9007	0.8736	0.8435	0.8049	3.0529
0.0312	7.0	1050	0.1057	0.0034	0	0	0	0	0	100	0	0.9720	0.9530	0.9303	0.9033	3.2390
0.0312	8.0	1200	0.1105	0.0034	0	0	0	0	0	100	0	0.9782	0.9604	0.9405	0.9150	2.6806
0.0312	9.0	1350	0.1035	0.0034	0	0	0	0	0	100	0	0.9782	0.9602	0.9400	0.9147	2.5689
0.0156	10.0	1500	0.1028	0.0034	0	0	0	0	0	100	0	0.9749	0.9549	0.9325	0.9055	3.0529
0.0156	11.0	1650	0.1015	0.0034	0	0	0	0	0	100	0	0.9769	0.9580	0.9368	0.9110	2.6806
0.0156	12.0	1800	0.1001	0.0034	0	0	0	0	0	100	0	0.9798	0.9621	0.9419	0.9166	2.4944
0.0156	13.0	1950	0.1126	0.0034	0	0	0	0	0	100	0	0.9769	0.9576	0.9361	0.9094	2.8295
0.01	14.0	2100	0.1059	0.0034	0	0	0	0	0	100	0	0.9725	0.9532	0.9318	0.9046	3.2390
0.01	15.0	2250	0.1028	0.0034	0	0	0	0	0	100	0	0.9786	0.9609	0.9406	0.9149	2.5316
0.01	16.0	2400	0.1026	0.0034	0	0	0	0	0	100	0	0.9799	0.9635	0.9448	0.9205	2.3827
0.0072	17.0	2550	0.1032	0.0034	0	0	0	0	0	100	0	0.9763	0.9586	0.9383	0.9119	2.7550
0.0072	18.0	2700	0.1053	0.0034	0	0	0	0	0	100	0	0.9799	0.9631	0.9439	0.9193	2.2338
0.0072	19.0	2850	0.1063	0.0034	0	0	0	0	0	100	0	0.9802	0.9637	0.9444	0.9199	2.3083
0.0031	20.0	3000	0.1084	0.0034	0	0	0	0	0	100	0	0.9789	0.9632	0.9445	0.9205	2.4200
0.0031	21.0	3150	0.1098	0.0034	0	0	0	0	0	100	0	0.9799	0.9631	0.9439	0.9196	2.3455
0.0031	22.0	3300	0.1061	0.0034	0	0	0	0	0	100	0	0.9805	0.9643	0.9455	0.9215	2.2338
0.0031	23.0	3450	0.1085	0.0034	0	0	0	0	0	100	0	0.9795	0.9626	0.9431	0.9189	2.3455
0.0031	24.0	3600	0.1086	0.0034	0	0	0	0	0	100	0	0.9792	0.9622	0.9426	0.9178	2.4572
0.0031	25.0	3750	0.1086	0.0034	0	0	0	0	0	100	0	0.9792	0.9629	0.9438	0.9193	2.4200
0.0031	26.0	3900	0.1069	0.0034	0	0	0	0	0	100	0	0.9773	0.9606	0.9411	0.9164	2.5689
0.0017	27.0	4050	0.1080	0.0034	0	0	0	0	0	100	0	0.9805	0.9645	0.9458	0.9218	2.1966
0.0017	28.0	4200	0.1077	0.0034	0	0	0	0	0	100	0	0.9802	0.9639	0.9449	0.9206	2.2338
0.0017	29.0	4350	0.1069	0.0034	0	0	0	0	0	100	0	0.9805	0.9645	0.9456	0.9215	2.1593
0.0013	30.0	4500	0.1069	0.0034	0	0	0	0	0	100	0	0.9805	0.9645	0.9458	0.9218	2.1593

Framework versions

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