exp4_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: 0.7457
• Model Preparation Time: 0.0034
• Bleu Msl: 0
• Bleu 1 Msl: 0.7713
• Bleu 2 Msl: 0.6545
• Bleu 3 Msl: 0.4894
• Bleu 4 Msl: 0.3118
• Ter Msl: 30.6485
• Bleu Asl: 0
• Bleu 1 Asl: 0.9706
• Bleu 2 Asl: 0.9536
• Bleu 3 Asl: 0.9345
• Bleu 4 Asl: 0.9107
• Ter Asl: 3.5791

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	0.5878	0.0034	0	0.4067	0.3167	0.2269	0.1361	67.2431	0	0.9574	0.9276	0.8958	0.8591	5.1378
No log	2.0	450	0.5415	0.0034	0	0.6429	0.5363	0.4247	0.2908	38.4537	0	0.9619	0.9338	0.9025	0.8661	4.7655
0.4626	3.0	675	0.5018	0.0034	0	0.4290	0.3438	0.2557	0.1628	38.8606	0	0.9736	0.9532	0.9298	0.9016	3.1646
0.4626	4.0	900	0.5354	0.0034	0	0.6278	0.5219	0.4042	0.2706	36.6226	0	0.9700	0.9481	0.9241	0.8949	3.5741
0.0785	5.0	1125	0.5086	0.0034	0	0.7672	0.6649	0.5410	0.3939	33.3672	0	0.9739	0.9522	0.9274	0.8978	3.3507
0.0785	6.0	1350	0.5244	0.0034	0	0.7261	0.6332	0.5123	0.3576	34.3845	0	0.9759	0.9562	0.9341	0.9067	2.8295
0.0425	7.0	1575	0.5274	0.0034	0	0.7160	0.6355	0.5115	0.3657	32.4517	0	0.9765	0.9573	0.9353	0.9084	2.8667
0.0425	8.0	1800	0.5042	0.0034	0	0.7589	0.6720	0.5547	0.4135	29.6033	0	0.9766	0.9584	0.9380	0.9122	2.8295
0.0265	9.0	2025	0.5092	0.0034	0	0.7431	0.6477	0.5320	0.3886	31.8413	0	0.9772	0.9579	0.9362	0.9096	2.6806
0.0265	10.0	2250	0.4964	0.0034	0	0.7088	0.6158	0.4983	0.3573	34.4863	0	0.9759	0.9561	0.9343	0.9074	2.8295
0.0265	11.0	2475	0.5224	0.0034	0	0.7544	0.6589	0.5448	0.4046	33.2655	0	0.9769	0.9587	0.9378	0.9115	2.6806
0.0198	12.0	2700	0.5098	0.0034	0	0.7424	0.6521	0.5358	0.3927	32.9603	0	0.9769	0.9583	0.9370	0.9106	2.6433
0.0198	13.0	2925	0.5245	0.0034	0	0.7525	0.6619	0.5554	0.4124	33.4690	0	0.9756	0.9561	0.9337	0.9054	3.0529
0.0149	14.0	3150	0.5363	0.0034	0	0.7492	0.6579	0.5494	0.4025	32.7569	0	0.9720	0.9513	0.9273	0.8977	3.5741
0.0149	15.0	3375	0.5522	0.0034	0	0.7415	0.6470	0.5361	0.3884	35.7070	0	0.9762	0.9573	0.9359	0.9100	2.8295
0.0121	16.0	3600	0.5217	0.0034	0	0.7601	0.6709	0.5548	0.4012	32.5534	0	0.9749	0.9557	0.9339	0.9069	2.9412
0.0121	17.0	3825	0.5126	0.0034	0	0.7419	0.6459	0.5253	0.3763	34.5880	0	0.9769	0.9583	0.9366	0.9101	2.7178
0.0096	18.0	4050	0.5239	0.0034	0	0.7682	0.6787	0.5642	0.4118	32.9603	0	0.9778	0.9597	0.9385	0.9122	2.5689
0.0096	19.0	4275	0.5519	0.0034	0	0.7508	0.6708	0.5600	0.4058	31.4344	0	0.9759	0.9570	0.9352	0.9085	2.7923
0.0069	20.0	4500	0.5532	0.0034	0	0.7580	0.6743	0.5569	0.4011	31.2309	0	0.9765	0.9578	0.9360	0.9096	2.6806
0.0069	21.0	4725	0.5334	0.0034	0	0.7599	0.6797	0.5655	0.4159	31.4344	0	0.9762	0.9576	0.9370	0.9116	2.7550
0.0069	22.0	4950	0.5297	0.0034	0	0.7658	0.6861	0.5728	0.4200	30.6205	0	0.9772	0.9594	0.9393	0.9146	2.6806
0.0058	23.0	5175	0.5561	0.0034	0	0.7553	0.6710	0.5554	0.4022	32.1465	0	0.9769	0.9589	0.9387	0.9137	2.7178
0.0058	24.0	5400	0.5389	0.0034	0	0.7571	0.6740	0.5606	0.4079	31.4344	0	0.9762	0.9582	0.9375	0.9121	2.7550
0.0046	25.0	5625	0.5479	0.0034	0	0.7469	0.6613	0.5448	0.3959	32.2482	0	0.9769	0.9587	0.9379	0.9125	2.7550
0.0046	26.0	5850	0.5529	0.0034	0	0.7383	0.6483	0.5303	0.3805	33.0621	0	0.9756	0.9566	0.9356	0.9100	2.8667
0.0039	27.0	6075	0.5596	0.0034	0	0.7451	0.6579	0.5385	0.3832	32.3499	0	0.9778	0.9602	0.9402	0.9155	2.6061
0.0039	28.0	6300	0.5584	0.0034	0	0.7454	0.6628	0.5455	0.3947	31.7396	0	0.9772	0.9592	0.9388	0.9136	2.6433
0.0036	29.0	6525	0.5604	0.0034	0	0.7481	0.6622	0.5427	0.3904	32.3499	0	0.9772	0.9592	0.9388	0.9137	2.6806
0.0036	30.0	6750	0.5574	0.0034	0	0.7475	0.6640	0.5458	0.3931	31.6378	0	0.9769	0.9587	0.9381	0.9128	2.7178

Framework versions

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