exp2_10partition_modelo3000

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: 1.0948
• Model Preparation Time: 0.0033
• Bleu Msl: 0
• Bleu 1 Msl: 0.8845
• Bleu 2 Msl: 0.8370
• Bleu 3 Msl: 0.7723
• Bleu 4 Msl: 0.6451
• Ter Msl: 17.9272
• Bleu Asl: 0
• Bleu 1 Asl: 0
• Bleu 2 Asl: 0
• Bleu 3 Asl: 0
• Bleu 4 Asl: 0
• Ter Asl: 100

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	75	0.9046	0.0033	0	0.3703	0.2970	0.2268	0.1442	78.0793	0	0	0	0	0	100
No log	2.0	150	0.6660	0.0033	0	0.8752	0.8014	0.7035	0.5440	20.5637	0	0	0	0	0	100
No log	3.0	225	0.7901	0.0033	0	0.8289	0.7487	0.6546	0.5102	27.5574	0	0	0	0	0	100
No log	4.0	300	0.6702	0.0033	0	0.8519	0.7724	0.6515	0.4803	20.9812	0	0	0	0	0	100
No log	5.0	375	0.6241	0.0033	0	0.8704	0.8036	0.7068	0.5441	20.5637	0	0	0	0	0	100
No log	6.0	450	0.6839	0.0033	0	0.8641	0.7903	0.6911	0.5385	24.8434	0	0	0	0	0	100
0.6205	7.0	525	0.6668	0.0033	0	0.8694	0.8052	0.7063	0.5480	20.2505	0	0	0	0	0	100
0.6205	8.0	600	0.6896	0.0033	0	0.8639	0.7863	0.6773	0.5052	22.5470	0	0	0	0	0	100
0.6205	9.0	675	0.7457	0.0033	0	0.8305	0.7641	0.6696	0.5162	25.6785	0	0	0	0	0	100
0.6205	10.0	750	0.7002	0.0033	0	0.8723	0.8149	0.7157	0.5472	17.9541	0	0	0	0	0	100
0.6205	11.0	825	0.6898	0.0033	0	0.8603	0.7944	0.6951	0.5361	22.1294	0	0	0	0	0	100
0.6205	12.0	900	0.6749	0.0033	0	0.8759	0.8201	0.7297	0.5663	18.1628	0	0	0	0	0	100
0.6205	13.0	975	0.6903	0.0033	0	0.8761	0.8184	0.7254	0.5674	17.9541	0	0	0	0	0	100
0.0488	14.0	1050	0.6715	0.0033	0	0.8676	0.8067	0.7131	0.5525	20.3549	0	0	0	0	0	100
0.0488	15.0	1125	0.6555	0.0033	0	0.8666	0.8139	0.7199	0.5503	17.2234	0	0	0	0	0	100
0.0488	16.0	1200	0.7132	0.0033	0	0.8789	0.8168	0.7237	0.5601	19.2067	0	0	0	0	0	100
0.0488	17.0	1275	0.6874	0.0033	0	0.8837	0.8264	0.7310	0.5636	18.2672	0	0	0	0	0	100
0.0488	18.0	1350	0.7008	0.0033	0	0.8875	0.8278	0.7335	0.5729	18.7891	0	0	0	0	0	100
0.0488	19.0	1425	0.6363	0.0033	0	0.8681	0.8200	0.7279	0.5545	16.1795	0	0	0	0	0	100
0.0215	20.0	1500	0.6844	0.0033	0	0.8789	0.8182	0.7231	0.5615	19.3111	0	0	0	0	0	100
0.0215	21.0	1575	0.6466	0.0033	0	0.8718	0.8139	0.7213	0.5575	18.2672	0	0	0	0	0	100
0.0215	22.0	1650	0.6860	0.0033	0	0.8742	0.8151	0.7227	0.5584	19.6242	0	0	0	0	0	100
0.0215	23.0	1725	0.6485	0.0033	0	0.8700	0.8126	0.7211	0.5592	18.3716	0	0	0	0	0	100
0.0215	24.0	1800	0.6594	0.0033	0	0.8788	0.8196	0.7280	0.5653	19.1023	0	0	0	0	0	100
0.0215	25.0	1875	0.6605	0.0033	0	0.8758	0.8193	0.7286	0.5654	18.0585	0	0	0	0	0	100
0.0215	26.0	1950	0.6482	0.0033	0	0.8699	0.8133	0.7198	0.5562	18.2672	0	0	0	0	0	100
0.014	27.0	2025	0.6410	0.0033	0	0.8689	0.8123	0.7185	0.5545	18.2672	0	0	0	0	0	100
0.014	28.0	2100	0.6564	0.0033	0	0.8729	0.8158	0.7238	0.5611	18.3716	0	0	0	0	0	100
0.014	29.0	2175	0.6571	0.0033	0	0.8690	0.8113	0.7175	0.5538	18.4760	0	0	0	0	0	100
0.014	30.0	2250	0.6563	0.0033	0	0.8690	0.8113	0.7175	0.5538	18.4760	0	0	0	0	0	100

Framework versions

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