exp2_10partition_modelo6000

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: 0.7177
• Model Preparation Time: 0.0055
• Bleu Msl: 0
• Bleu 1 Msl: 0.8613
• Bleu 2 Msl: 0.8246
• Bleu 3 Msl: 0.7705
• Bleu 4 Msl: 0.6471
• Ter Msl: 19.5145
• Bleu Asl: 0
• Bleu 1 Asl: 0.9566
• Bleu 2 Asl: 0.9339
• Bleu 3 Asl: 0.9089
• Bleu 4 Asl: 0.8790
• Ter Asl: 5.3897

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	0.5531	0.0055	0	0.8827	0.7822	0.6773	0.5163	22.9645	0	0.9342	0.8941	0.8566	0.8168	8.3744
No log	2.0	300	0.4809	0.0055	0	0.8034	0.7089	0.6134	0.4689	33.1942	0	0.9424	0.9112	0.8790	0.8425	6.8262
No log	3.0	450	0.3903	0.0055	0	0.8921	0.8197	0.7195	0.5580	19.1023	0	0.9501	0.9212	0.8910	0.8579	6.2632
0.5309	4.0	600	0.4627	0.0055	0	0.8615	0.7749	0.6636	0.4943	25.7829	0	0.9468	0.9181	0.8873	0.8520	6.5447
0.5309	5.0	750	0.4495	0.0055	0	0.8941	0.8267	0.7307	0.5639	18.6848	0	0.9544	0.9290	0.9017	0.8707	5.6298
0.5309	6.0	900	0.4334	0.0055	0	0.8580	0.7896	0.6922	0.5148	23.7996	0	0.9531	0.9273	0.8990	0.8675	5.6298
0.0697	7.0	1050	0.4502	0.0055	0	0.8456	0.7671	0.6754	0.5213	25.4697	0	0.9586	0.9348	0.9085	0.8782	5.2076
0.0697	8.0	1200	0.4398	0.0055	0	0.8704	0.7887	0.6880	0.5276	23.7996	0	0.9562	0.9309	0.9034	0.8711	5.4187
0.0697	9.0	1350	0.4880	0.0055	0	0.8646	0.7850	0.6854	0.5280	24.2171	0	0.9538	0.9287	0.9015	0.8695	5.4187
0.0347	10.0	1500	0.4460	0.0055	0	0.8800	0.7952	0.6902	0.5186	23.4864	0	0.9525	0.9289	0.9044	0.8751	5.4891
0.0347	11.0	1650	0.4399	0.0055	0	0.8599	0.7817	0.6885	0.5307	23.6952	0	0.9531	0.9287	0.9023	0.8710	5.7002
0.0347	12.0	1800	0.4512	0.0055	0	0.8430	0.7587	0.6638	0.5103	26.4092	0	0.9586	0.9388	0.9174	0.8906	4.7854
0.0347	13.0	1950	0.4328	0.0055	0	0.9009	0.8295	0.7326	0.5642	19.2067	0	0.9580	0.9368	0.9142	0.8864	4.9261
0.0192	14.0	2100	0.4584	0.0055	0	0.8582	0.7826	0.6892	0.5327	23.9040	0	0.9569	0.9338	0.9089	0.8790	5.0669
0.0192	15.0	2250	0.4435	0.0055	0	0.8741	0.7961	0.7016	0.5362	23.0689	0	0.9529	0.9302	0.9060	0.8767	5.4891
0.0192	16.0	2400	0.4444	0.0055	0	0.8720	0.7965	0.7007	0.5298	23.3820	0	0.9593	0.9376	0.9132	0.8841	4.8557
0.013	17.0	2550	0.4424	0.0055	0	0.8866	0.8089	0.7089	0.5366	22.2338	0	0.9617	0.9414	0.9192	0.8915	4.5742
0.013	18.0	2700	0.4504	0.0055	0	0.8710	0.7954	0.6969	0.5273	23.9040	0	0.9623	0.9412	0.9173	0.8885	4.6446
0.013	19.0	2850	0.4472	0.0055	0	0.8780	0.8056	0.7023	0.5262	21.5031	0	0.9599	0.9371	0.9115	0.8800	4.9261
0.0098	20.0	3000	0.4748	0.0055	0	0.8770	0.8067	0.7067	0.5298	21.7119	0	0.9612	0.9412	0.9185	0.8902	4.6446
0.0098	21.0	3150	0.4512	0.0055	0	0.8704	0.7949	0.6956	0.5270	23.4864	0	0.9612	0.9412	0.9189	0.8904	4.6446
0.0098	22.0	3300	0.4549	0.0055	0	0.8527	0.7754	0.6816	0.5131	25.4697	0	0.9606	0.9398	0.9167	0.8878	4.7854
0.0098	23.0	3450	0.4658	0.0055	0	0.8541	0.7802	0.6885	0.5277	24.7390	0	0.9606	0.9394	0.9157	0.8863	4.7854
0.0076	24.0	3600	0.4437	0.0055	0	0.8746	0.8003	0.7011	0.5268	23.0689	0	0.9624	0.9421	0.9196	0.8918	4.5039
0.0076	25.0	3750	0.4636	0.0055	0	0.8654	0.7940	0.7017	0.5403	23.3820	0	0.9624	0.9425	0.9202	0.8923	4.5039
0.0076	26.0	3900	0.4425	0.0055	0	0.8741	0.8019	0.7051	0.5345	22.9645	0	0.9612	0.9404	0.9170	0.8886	4.6446
0.0067	27.0	4050	0.4544	0.0055	0	0.8547	0.7792	0.6867	0.5270	24.9478	0	0.9618	0.9415	0.9183	0.8900	4.5742
0.0067	28.0	4200	0.4502	0.0055	0	0.8626	0.7914	0.6990	0.5378	23.9040	0	0.9618	0.9415	0.9183	0.8900	4.5742
0.0067	29.0	4350	0.4536	0.0055	0	0.8616	0.7880	0.6934	0.5309	24.2171	0	0.9618	0.9415	0.9183	0.8900	4.5742
0.005	30.0	4500	0.4535	0.0055	0	0.8616	0.7880	0.6934	0.5309	24.2171	0	0.9618	0.9415	0.9183	0.8900	4.5742

Framework versions

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