exp3_10partition_modelo_msl3000

This model is a fine-tuned version of vgaraujov/bart-base-spanish on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.5938
• Model Preparation Time: 0.0035
• Bleu Msl: 0
• Bleu 1 Msl: 0.5930
• Bleu 2 Msl: 0.5164
• Bleu 3 Msl: 0.4267
• Bleu 4 Msl: 0.3285
• Ter Msl: 46.9697
• Bleu Asl: 0
• Bleu 1 Asl: 0
• Bleu 2 Asl: 0
• Bleu 3 Asl: 0
• Bleu 4 Asl: 0
• Ter Asl: 100

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	75	3.7541	0.0035	0	0.3386	0.1627	0.0792	0.0357	93.4233	0	0	0	0	0	100
No log	2.0	150	3.5098	0.0035	0	0.4492	0.2826	0.1691	0.0710	72.6813	0	0	0	0	0	100
No log	3.0	225	3.1749	0.0035	0	0.4246	0.2689	0.1652	0.0976	69.6459	0	0	0	0	0	100
No log	4.0	300	3.0146	0.0035	0	0.4646	0.3286	0.2427	0.1793	72.2597	0	0	0	0	0	100
No log	5.0	375	2.9902	0.0035	0	0.4561	0.3293	0.2482	0.1764	64.6712	0	0	0	0	0	100
No log	6.0	450	3.0020	0.0035	0	0.4844	0.3456	0.2676	0.1995	64.9241	0	0	0	0	0	100
1.0832	7.0	525	2.9761	0.0035	0	0.4787	0.3231	0.2337	0.1726	76.5599	0	0	0	0	0	100
1.0832	8.0	600	3.0630	0.0035	0	0.4702	0.3253	0.2468	0.1830	65.0084	0	0	0	0	0	100
1.0832	9.0	675	2.9335	0.0035	0	0.4598	0.3093	0.2163	0.1436	64.8398	0	0	0	0	0	100
1.0832	10.0	750	2.7573	0.0035	0	0.5098	0.3634	0.2783	0.2075	63.8280	0	0	0	0	0	100
1.0832	11.0	825	2.9261	0.0035	0	0.4766	0.3426	0.2690	0.2069	66.4418	0	0	0	0	0	100
1.0832	12.0	900	2.9181	0.0035	0	0.4677	0.3359	0.2637	0.1976	68.5497	0	0	0	0	0	100
1.0832	13.0	975	2.9349	0.0035	0	0.4825	0.3465	0.2749	0.2106	64.5868	0	0	0	0	0	100
0.0742	14.0	1050	3.0945	0.0035	0	0.4727	0.3334	0.2604	0.1953	67.2850	0	0	0	0	0	100
0.0742	15.0	1125	2.8864	0.0035	0	0.5020	0.3536	0.2739	0.2062	68.5497	0	0	0	0	0	100
0.0742	16.0	1200	2.8654	0.0035	0	0.4962	0.3504	0.2735	0.2047	69.5616	0	0	0	0	0	100
0.0742	17.0	1275	3.0379	0.0035	0	0.4812	0.3469	0.2673	0.2027	67.0320	0	0	0	0	0	100
0.0742	18.0	1350	2.9964	0.0035	0	0.4894	0.3397	0.2635	0.1974	63.6594	0	0	0	0	0	100
0.0742	19.0	1425	2.9434	0.0035	0	0.4940	0.3501	0.2754	0.2091	66.5261	0	0	0	0	0	100
0.0294	20.0	1500	2.9587	0.0035	0	0.4924	0.3510	0.2785	0.2129	64.4182	0	0	0	0	0	100
0.0294	21.0	1575	3.0184	0.0035	0	0.5057	0.3476	0.2632	0.1921	68.8027	0	0	0	0	0	100
0.0294	22.0	1650	3.0982	0.0035	0	0.4923	0.3565	0.2836	0.2175	64.3339	0	0	0	0	0	100
0.0294	23.0	1725	2.9882	0.0035	0	0.4968	0.3474	0.2691	0.2001	68.4654	0	0	0	0	0	100
0.0294	24.0	1800	2.9938	0.0035	0	0.4944	0.3440	0.2655	0.1959	70.4890	0	0	0	0	0	100
0.0294	25.0	1875	3.0287	0.0035	0	0.5076	0.3625	0.2855	0.2171	65.9359	0	0	0	0	0	100
0.0294	26.0	1950	3.0334	0.0035	0	0.4995	0.3509	0.2737	0.2048	66.3575	0	0	0	0	0	100
0.0159	27.0	2025	3.0210	0.0035	0	0.5084	0.3615	0.2852	0.2166	65.5143	0	0	0	0	0	100
0.0159	28.0	2100	3.0197	0.0035	0	0.5128	0.3642	0.2852	0.2140	64.8398	0	0	0	0	0	100
0.0159	29.0	2175	3.0489	0.0035	0	0.5072	0.3610	0.2820	0.2132	65.1771	0	0	0	0	0	100
0.0159	30.0	2250	3.0483	0.0035	0	0.5117	0.3607	0.2810	0.2111	65.7673	0	0	0	0	0	100

Framework versions

• Transformers 4.50.3
• Pytorch 2.6.0+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1