exp2_10partition_modelo9000

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: 0.4128
• Model Preparation Time: 0.0038
• Bleu Msl: 0
• Bleu 1 Msl: 0.8509
• Bleu 2 Msl: 0.8059
• Bleu 3 Msl: 0.7419
• Bleu 4 Msl: 0.6107
• Ter Msl: 20.3548
• Bleu Asl: 0
• Bleu 1 Asl: 0.9663
• Bleu 2 Asl: 0.9440
• Bleu 3 Asl: 0.9193
• Bleu 4 Asl: 0.8908
• Ter Asl: 3.9450

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	225	0.3088	0.0038	0	0.2992	0.2390	0.1792	0.1099	88.3090	0	0.9517	0.9241	0.8923	0.8589	5.9826
No log	2.0	450	0.2691	0.0038	0	0.8584	0.7799	0.6792	0.4958	20.7724	0	0.9640	0.9423	0.9152	0.8855	4.6048
0.4907	3.0	675	0.2479	0.0038	0	0.8932	0.8218	0.7138	0.5437	17.9541	0	0.9705	0.9507	0.9270	0.9004	3.5533
0.4907	4.0	900	0.2521	0.0038	0	0.8189	0.7413	0.6413	0.4898	19.8330	0	0.9676	0.9460	0.9206	0.8919	4.0972
0.0885	5.0	1125	0.2573	0.0038	0	0.8680	0.8041	0.7113	0.5416	19.2067	0	0.9694	0.9507	0.9264	0.8986	3.6621
0.0885	6.0	1350	0.2423	0.0038	0	0.8640	0.7893	0.6870	0.5151	21.7119	0	0.9737	0.9557	0.9321	0.9050	3.1182
0.047	7.0	1575	0.2308	0.0038	0	0.8951	0.8277	0.7299	0.5528	19.9374	0	0.9768	0.9603	0.9389	0.9142	2.9369
0.047	8.0	1800	0.2880	0.0038	0	0.8596	0.7856	0.6788	0.4906	18.9979	0	0.9724	0.9534	0.9289	0.9022	3.3358
0.0296	9.0	2025	0.2530	0.0038	0	0.8804	0.8172	0.7277	0.5684	21.1900	0	0.9743	0.9561	0.9328	0.9067	3.2270
0.0296	10.0	2250	0.2573	0.0038	0	0.8697	0.7896	0.6886	0.5156	23.5908	0	0.9756	0.9579	0.9351	0.9096	2.9732
0.0296	11.0	2475	0.2681	0.0038	0	0.8682	0.7926	0.6893	0.5100	24.4259	0	0.9640	0.9453	0.9208	0.8930	4.3147
0.0208	12.0	2700	0.2789	0.0038	0	0.8891	0.8304	0.7371	0.5664	18.6848	0	0.9756	0.9587	0.9356	0.9096	2.9369
0.0208	13.0	2925	0.2463	0.0038	0	0.8941	0.8362	0.7425	0.5816	17.9541	0	0.9771	0.9590	0.9350	0.9081	2.8281
0.0153	14.0	3150	0.2701	0.0038	0	0.8518	0.7822	0.6871	0.5221	24.4259	0	0.9734	0.9559	0.9327	0.9067	3.1545
0.0153	15.0	3375	0.2805	0.0038	0	0.8769	0.8143	0.7193	0.5422	19.2067	0	0.9768	0.9597	0.9368	0.9113	2.8281
0.0128	16.0	3600	0.2906	0.0038	0	0.8824	0.8105	0.7106	0.5389	21.9207	0	0.9768	0.9595	0.9366	0.9116	2.9007
0.0128	17.0	3825	0.2870	0.0038	0	0.8693	0.7938	0.6917	0.5221	22.4426	0	0.9769	0.9603	0.9385	0.9140	2.7556
0.0101	18.0	4050	0.2758	0.0038	0	0.8821	0.8130	0.7135	0.5438	20.5637	0	0.9749	0.9563	0.9330	0.9069	3.1545
0.0101	19.0	4275	0.2702	0.0038	0	0.8837	0.8193	0.7213	0.5464	19.6242	0	0.9750	0.9583	0.9357	0.9094	2.9369
0.0074	20.0	4500	0.2731	0.0038	0	0.8653	0.7935	0.6921	0.5283	22.7557	0	0.9747	0.9577	0.9352	0.9094	3.0457
0.0074	21.0	4725	0.2785	0.0038	0	0.8894	0.8259	0.7280	0.5578	18.9979	0	0.9765	0.9601	0.9376	0.9127	2.8281
0.0074	22.0	4950	0.2906	0.0038	0	0.8912	0.8304	0.7331	0.5552	18.8935	0	0.9747	0.9583	0.9364	0.9115	2.9732
0.006	23.0	5175	0.2898	0.0038	0	0.8825	0.8233	0.7253	0.5507	18.3716	0	0.9759	0.9597	0.9378	0.9126	2.8644
0.006	24.0	5400	0.2851	0.0038	0	0.8795	0.8226	0.7266	0.5447	18.5804	0	0.9766	0.9610	0.9395	0.9146	2.7919
0.0054	25.0	5625	0.2709	0.0038	0	0.8864	0.8186	0.7177	0.5434	19.3111	0	0.9769	0.9615	0.9403	0.9159	2.7919
0.0054	26.0	5850	0.2767	0.0038	0	0.8884	0.8215	0.7243	0.5477	19.5198	0	0.9772	0.9616	0.9405	0.9161	2.7556
0.0044	27.0	6075	0.2802	0.0038	0	0.8760	0.8060	0.7042	0.5237	20.0418	0	0.9769	0.9616	0.9402	0.9153	2.6831
0.0044	28.0	6300	0.2765	0.0038	0	0.8789	0.8090	0.7084	0.5317	20.2505	0	0.9775	0.9622	0.9411	0.9168	2.6831
0.0036	29.0	6525	0.2785	0.0038	0	0.8808	0.8111	0.7094	0.5294	20.1461	0	0.9775	0.9622	0.9411	0.9166	2.6831
0.0036	30.0	6750	0.2801	0.0038	0	0.8789	0.8068	0.7054	0.5267	20.5637	0	0.9784	0.9635	0.9427	0.9188	2.5743

Framework versions

• Transformers 4.50.0
• Pytorch 2.6.0+cu124
• Datasets 3.4.1
• Tokenizers 0.21.1