exp5_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: 0.7922
• Model Preparation Time: 0.0034
• Bleu Msl: 0
• Bleu 1 Msl: 0.6459
• Bleu 2 Msl: 0.5244
• Bleu 3 Msl: 0.4178
• Bleu 4 Msl: 0.3002
• Ter Msl: 43.3114
• Bleu Asl: 0
• Bleu 1 Asl: 0.9757
• Bleu 2 Asl: 0.9586
• Bleu 3 Asl: 0.9378
• Bleu 4 Asl: 0.9130
• Ter Asl: 2.7027

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	0.5492	0.0034	0	0.0151	0.0099	0.0068	0.0037	1375.0221	0	0.9518	0.9216	0.8897	0.8545	5.8389
No log	2.0	450	0.4142	0.0034	0	0.7246	0.6407	0.5573	0.4358	32.5066	0	0.9606	0.9351	0.9083	0.8774	5.0259
0.4717	3.0	675	0.4521	0.0034	0	0.6578	0.4969	0.3748	0.2422	45.5270	0	0.9677	0.9462	0.9225	0.8947	4.0281
0.4717	4.0	900	0.4779	0.0034	0	0.4113	0.3023	0.2256	0.1508	99.8229	0	0.9687	0.9488	0.9265	0.8995	3.7694
0.0864	5.0	1125	0.4282	0.0034	0	0.6941	0.5776	0.4779	0.3717	35.9610	0	0.9699	0.9498	0.9274	0.9007	3.8064
0.0864	6.0	1350	0.4464	0.0034	0	0.6896	0.5970	0.5255	0.4252	38.5297	0	0.9709	0.9522	0.9310	0.9061	3.6955
0.0427	7.0	1575	0.4092	0.0034	0	0.7219	0.6444	0.5706	0.4632	29.9380	0	0.9703	0.9503	0.9278	0.9014	3.6216
0.0427	8.0	1800	0.5207	0.0034	0	0.6261	0.4955	0.3885	0.2788	42.4269	0	0.9735	0.9548	0.9336	0.9088	3.3259
0.0293	9.0	2025	0.4797	0.0034	0	0.6827	0.5760	0.4826	0.3747	38.7068	0	0.9681	0.9475	0.9237	0.8958	3.9542
0.0293	10.0	2250	0.5031	0.0034	0	0.6852	0.5826	0.4838	0.3688	36.4925	0	0.9709	0.9516	0.9294	0.9035	3.6216
0.0293	11.0	2475	0.4935	0.0034	0	0.7109	0.6097	0.5191	0.4047	35.4296	0	0.9706	0.9515	0.9295	0.9036	3.6216
0.0222	12.0	2700	0.4795	0.0034	0	0.6672	0.5483	0.4408	0.3188	39.0611	0	0.9725	0.9540	0.9335	0.9092	3.5107
0.0222	13.0	2925	0.4913	0.0034	0	0.6836	0.5615	0.4618	0.3485	37.6439	0	0.9713	0.9526	0.9316	0.9065	3.5107
0.0172	14.0	3150	0.4531	0.0034	0	0.6956	0.5751	0.4666	0.3451	38.9725	0	0.9755	0.9586	0.9392	0.9160	3.0303
0.0172	15.0	3375	0.5068	0.0034	0	0.6971	0.5811	0.4846	0.3681	37.1125	0	0.9712	0.9528	0.9318	0.9065	3.6216
0.0099	16.0	3600	0.4841	0.0034	0	0.7237	0.6121	0.5218	0.4140	35.1639	0	0.9742	0.9564	0.9364	0.9123	3.3259
0.0099	17.0	3825	0.4761	0.0034	0	0.7201	0.6141	0.5301	0.4288	33.6581	0	0.9716	0.9529	0.9318	0.9064	3.4368
0.0091	18.0	4050	0.4839	0.0034	0	0.7	0.5790	0.4924	0.3924	36.7582	0	0.9719	0.9527	0.9309	0.9047	3.5107
0.0091	19.0	4275	0.4671	0.0034	0	0.6778	0.5702	0.4847	0.3867	36.3153	0	0.9739	0.9559	0.9353	0.9113	3.1781
0.0073	20.0	4500	0.4492	0.0034	0	0.7231	0.6240	0.5420	0.4406	31.7981	0	0.9726	0.9540	0.9332	0.9082	3.3629
0.0073	21.0	4725	0.4483	0.0034	0	0.7246	0.6142	0.5274	0.4187	34.2781	0	0.9751	0.9577	0.9374	0.9136	3.1042
0.0073	22.0	4950	0.4539	0.0034	0	0.7227	0.6176	0.5381	0.4357	32.9495	0	0.9758	0.9584	0.9387	0.9148	3.1042
0.0051	23.0	5175	0.4624	0.0034	0	0.6919	0.5739	0.4841	0.3769	37.2011	0	0.9745	0.9570	0.9369	0.9127	3.1412
0.0051	24.0	5400	0.4541	0.0034	0	0.7033	0.5894	0.5023	0.3949	34.7210	0	0.9758	0.9593	0.9401	0.9167	2.9194
0.0046	25.0	5625	0.4565	0.0034	0	0.7281	0.6193	0.5329	0.4322	32.5066	0	0.9764	0.9602	0.9413	0.9188	2.9194
0.0046	26.0	5850	0.4546	0.0034	0	0.7140	0.6116	0.5346	0.4398	32.4181	0	0.9742	0.9562	0.9355	0.9107	3.1412
0.004	27.0	6075	0.4779	0.0034	0	0.7184	0.6114	0.5351	0.4418	32.7724	0	0.9745	0.9570	0.9369	0.9129	3.1412
0.004	28.0	6300	0.4666	0.0034	0	0.7124	0.6024	0.5256	0.4307	33.6581	0	0.9729	0.9548	0.9340	0.9097	3.3259
0.003	29.0	6525	0.4661	0.0034	0	0.7105	0.6086	0.5339	0.4384	34.0124	0	0.9748	0.9571	0.9367	0.9127	3.1412
0.003	30.0	6750	0.4640	0.0034	0	0.7097	0.6071	0.5319	0.4360	34.1895	0	0.9748	0.9571	0.9367	0.9127	3.1412

Framework versions

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