exp1_10partition_modelo12000

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: 0.5307
• Model Preparation Time: 0.0033
• Bleu Msl: 0
• Bleu 1 Msl: 0.7236
• Bleu 2 Msl: 0.5880
• Bleu 3 Msl: 0.4898
• Bleu 4 Msl: 0.3772
• Ter Msl: 41.4894
• Bleu Asl: 0
• Bleu 1 Asl: 0.9423
• Bleu 2 Asl: 0.9134
• Bleu 3 Asl: 0.8825
• Bleu 4 Asl: 0.8473
• Ter Asl: 7.6042

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	300	0.2824	0.0033	0	0.3970	0.3437	0.2930	0.2280	59.9496	0	0.9410	0.9115	0.8792	0.8422	7.2709
0.4472	2.0	600	0.2520	0.0033	0	0.6272	0.5542	0.4799	0.3809	36.6919	0	0.9524	0.9274	0.8990	0.8663	6.1255
0.4472	3.0	900	0.2422	0.0033	0	0.7249	0.6515	0.5749	0.4743	31.9060	0	0.9464	0.9220	0.8935	0.8595	6.6484
0.0958	4.0	1200	0.2221	0.0033	0	0.8251	0.7588	0.6904	0.5818	24.7691	0	0.9631	0.9420	0.9166	0.8865	4.9552
0.0543	5.0	1500	0.2228	0.0033	0	0.7720	0.7056	0.6371	0.5350	27.7918	0	0.9561	0.9341	0.9088	0.8782	5.9761
0.0543	6.0	1800	0.2289	0.0033	0	0.8046	0.7289	0.6549	0.5539	26.3644	0	0.9444	0.9218	0.8964	0.8665	7.6195
0.0339	7.0	2100	0.2397	0.0033	0	0.7718	0.6886	0.6044	0.4846	32.4097	0	0.9181	0.8922	0.8640	0.8301	12.3506
0.0339	8.0	2400	0.2501	0.0033	0	0.7656	0.6927	0.6164	0.5111	30.7305	0	0.9371	0.9147	0.8900	0.8595	8.5408
0.0279	9.0	2700	0.2256	0.0033	0	0.8464	0.7802	0.7114	0.6030	23.0898	0	0.9340	0.9123	0.8887	0.8589	8.4910
0.0197	10.0	3000	0.2364	0.0033	0	0.8317	0.7717	0.7034	0.5941	24.0134	0	0.8519	0.8202	0.7860	0.7427	19.8456
0.0197	11.0	3300	0.2382	0.0033	0	0.8315	0.7693	0.7025	0.5981	23.0059	0	0.9456	0.9243	0.9004	0.8705	7.3705
0.0181	12.0	3600	0.2367	0.0033	0	0.8180	0.7518	0.6761	0.5627	27.1201	0	0.9318	0.8973	0.8611	0.8211	8.9890
0.0181	13.0	3900	0.2518	0.0033	0	0.7694	0.7032	0.6303	0.5205	28.9673	0	0.9285	0.8971	0.8646	0.8290	9.5369
0.0142	14.0	4200	0.2287	0.0033	0	0.8210	0.7598	0.6943	0.5892	24.5172	0	0.9420	0.9153	0.8880	0.8565	7.9681
0.0126	15.0	4500	0.2474	0.0033	0	0.8212	0.7575	0.6856	0.5748	25.1050	0	0.9259	0.8985	0.8707	0.8386	9.8606
0.0126	16.0	4800	0.2625	0.0033	0	0.7986	0.7340	0.6661	0.5594	26.1125	0	0.9557	0.9332	0.9067	0.8762	6.0010
0.0086	17.0	5100	0.2426	0.0033	0	0.8044	0.7325	0.6572	0.5571	26.0285	0	0.9539	0.9296	0.9023	0.8703	6.0010
0.0086	18.0	5400	0.2505	0.0033	0	0.7994	0.7270	0.6517	0.5444	27.7078	0	0.9459	0.9204	0.8927	0.8604	7.6693
0.0084	19.0	5700	0.2514	0.0033	0	0.8252	0.7541	0.6807	0.5752	25.9446	0	0.9150	0.8761	0.8384	0.7990	11.4791
0.007	20.0	6000	0.2542	0.0033	0	0.7730	0.7018	0.6303	0.5289	30.3946	0	0.9068	0.8760	0.8430	0.8054	12.4253
0.007	21.0	6300	0.2555	0.0033	0	0.7984	0.7275	0.6544	0.5479	28.5474	0	0.9444	0.9152	0.8869	0.8541	7.8934
0.0054	22.0	6600	0.2483	0.0033	0	0.7941	0.7271	0.6557	0.5480	27.6238	0	0.9478	0.9223	0.8944	0.8618	7.2460
0.0054	23.0	6900	0.2480	0.0033	0	0.8055	0.7460	0.6826	0.5782	25.1889	0	0.9364	0.9063	0.8754	0.8396	9.0139
0.0056	24.0	7200	0.2541	0.0033	0	0.8139	0.7519	0.6847	0.5792	24.7691	0	0.9461	0.9212	0.8947	0.8628	7.2211
0.0037	25.0	7500	0.2527	0.0033	0	0.8172	0.7523	0.6841	0.5763	26.7842	0	0.9294	0.9079	0.8829	0.8519	9.3376
0.0037	26.0	7800	0.2594	0.0033	0	0.8075	0.7417	0.6699	0.5596	26.8682	0	0.9366	0.9163	0.8920	0.8619	8.4163
0.0037	27.0	8100	0.2485	0.0033	0	0.8050	0.7420	0.6714	0.5611	25.7767	0	0.9345	0.9071	0.8783	0.8439	8.9392
0.0037	28.0	8400	0.2485	0.0033	0	0.8028	0.7389	0.6691	0.5607	26.5323	0	0.9385	0.9125	0.8847	0.8510	8.5408
0.0032	29.0	8700	0.2499	0.0033	0	0.8034	0.7409	0.6721	0.5655	26.0285	0	0.9421	0.9171	0.8901	0.8570	8.0428
0.0027	30.0	9000	0.2506	0.0033	0	0.8018	0.7398	0.6719	0.5648	26.1965	0	0.9404	0.9147	0.8877	0.8547	8.2918

Framework versions

• Transformers 4.49.0
• Pytorch 2.6.0+cu124
• Datasets 3.4.1
• Tokenizers 0.21.1