exp3_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: 0.4406
• Model Preparation Time: 0.0034
• Bleu Msl: 0
• Bleu 1 Msl: 0.8075
• Bleu 2 Msl: 0.7256
• Bleu 3 Msl: 0.6059
• Bleu 4 Msl: 0.4829
• Ter Msl: 27.0833
• Bleu Asl: 0
• Bleu 1 Asl: 0.9752
• Bleu 2 Asl: 0.9560
• Bleu 3 Asl: 0.9364
• Bleu 4 Asl: 0.9137
• Ter Asl: 2.9835

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	1.0887	0.0034	0	0.4826	0.2722	0.1227	0.0381	77.7403	0	0.9557	0.9266	0.8963	0.8624	5.6523
No log	2.0	450	0.9520	0.0034	0	0.2741	0.1653	0.0996	0.0588	100.3373	0	0.9598	0.9337	0.9067	0.8761	5.0480
0.4636	3.0	675	0.8482	0.0034	0	0.0798	0.0517	0.0356	0.0223	319.4772	0	0.9711	0.9506	0.9278	0.9006	3.4838
0.4636	4.0	900	0.8447	0.0034	0	0.5112	0.3597	0.2465	0.1454	70.4047	0	0.9718	0.9515	0.9284	0.9011	3.4838
0.0795	5.0	1125	0.9296	0.0034	0	0.6046	0.4295	0.3023	0.1991	61.9730	0	0.9746	0.9569	0.9361	0.9110	3.1283
0.0795	6.0	1350	0.9205	0.0034	0	0.2045	0.1377	0.0924	0.0525	129.8482	0	0.9730	0.9533	0.9299	0.9022	3.4127
0.0444	7.0	1575	0.8835	0.0034	0	0.6634	0.4998	0.3640	0.2243	46.7960	0	0.9690	0.9487	0.9260	0.8997	3.6260
0.0444	8.0	1800	0.9059	0.0034	0	0.3985	0.2859	0.2046	0.1390	82.6307	0	0.9688	0.9503	0.9291	0.9036	3.8038
0.0295	9.0	2025	0.9123	0.0034	0	0.6423	0.4784	0.3442	0.2388	56.0708	0	0.9721	0.9523	0.9295	0.9023	3.3772
0.0295	10.0	2250	0.9518	0.0034	0	0.6213	0.4665	0.3392	0.2180	59.7808	0	0.9734	0.9549	0.9333	0.9069	3.1283
0.0295	11.0	2475	0.8786	0.0034	0	0.6458	0.4985	0.3877	0.2960	51.1804	0	0.9761	0.9584	0.9374	0.9122	2.8439
0.0191	12.0	2700	0.9260	0.0034	0	0.6155	0.4650	0.3479	0.2512	49.1568	0	0.9758	0.9586	0.9378	0.9128	2.9150
0.0191	13.0	2925	0.9082	0.0034	0	0.6272	0.4727	0.3485	0.2391	53.1197	0	0.9762	0.9588	0.9375	0.9123	2.8439
0.0156	14.0	3150	0.9289	0.0034	0	0.6416	0.4961	0.3760	0.2734	49.5784	0	0.9739	0.9550	0.9334	0.9076	3.1994
0.0156	15.0	3375	0.8962	0.0034	0	0.6708	0.5202	0.4034	0.3018	52.0236	0	0.9740	0.9548	0.9322	0.9052	3.1283
0.0135	16.0	3600	0.9321	0.0034	0	0.6439	0.5017	0.3966	0.3077	47.3862	0	0.9740	0.9548	0.9324	0.9059	3.0217
0.0135	17.0	3825	0.9052	0.0034	0	0.6490	0.4953	0.3741	0.2675	54.5531	0	0.9733	0.9545	0.9327	0.9068	3.2350
0.0104	18.0	4050	0.8770	0.0034	0	0.6561	0.5057	0.3883	0.2808	51.2648	0	0.9734	0.9545	0.9333	0.9079	3.1283
0.0104	19.0	4275	0.9140	0.0034	0	0.6677	0.5107	0.3817	0.2643	50.0	0	0.9706	0.9515	0.9295	0.9027	3.4483
0.0078	20.0	4500	0.9122	0.0034	0	0.6495	0.4980	0.3803	0.2745	47.4705	0	0.9755	0.9577	0.9373	0.9129	2.8795
0.0078	21.0	4725	0.8904	0.0034	0	0.6584	0.5046	0.3862	0.2857	50.3373	0	0.9768	0.9603	0.9413	0.9185	2.7728
0.0078	22.0	4950	0.8965	0.0034	0	0.6587	0.5085	0.3895	0.2876	50.7589	0	0.9771	0.9602	0.9404	0.9169	2.7728
0.0064	23.0	5175	0.9242	0.0034	0	0.6522	0.4916	0.3607	0.2477	53.1197	0	0.9771	0.9598	0.9395	0.9150	2.7728
0.0064	24.0	5400	0.8987	0.0034	0	0.6615	0.5090	0.3902	0.2835	48.5666	0	0.9765	0.9598	0.9397	0.9154	2.7728
0.0046	25.0	5625	0.9078	0.0034	0	0.6654	0.5086	0.3844	0.2725	48.6509	0	0.9768	0.9603	0.9407	0.9169	2.7373
0.0046	26.0	5850	0.8952	0.0034	0	0.6747	0.5153	0.3926	0.2776	53.4570	0	0.9762	0.9594	0.9394	0.9152	2.8084
0.0044	27.0	6075	0.8998	0.0034	0	0.6814	0.5294	0.4121	0.3009	49.4941	0	0.9765	0.9595	0.9390	0.9142	2.7728
0.0044	28.0	6300	0.9095	0.0034	0	0.6751	0.5218	0.4000	0.2880	51.0118	0	0.9768	0.9599	0.9395	0.9149	2.7728
0.0039	29.0	6525	0.9113	0.0034	0	0.6728	0.5201	0.4009	0.2895	50.2530	0	0.9768	0.9599	0.9399	0.9158	2.7728
0.0039	30.0	6750	0.9123	0.0034	0	0.6714	0.5183	0.3989	0.2878	50.9275	0	0.9768	0.9599	0.9397	0.9154	2.7728

Framework versions

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