exp3_10partition_modeloorig

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: 1.1626
• Bleu Msl: 0.0
• Bleu 1 Msl: 0.52
• Bleu 2 Msl: 0.0132
• Bleu 3 Msl: 0.0041
• Bleu 4 Msl: 0.0021
• Ter Msl: {'score': 24.526515151515152, 'num_edits': 259, 'ref_length': 1056.0}
• 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	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	1.3893	0.0	0.4267	0.0119	0.0038	0.0020	{'score': 36.93181818181818, 'num_edits': 390, 'ref_length': 1056.0}	0	0	0	0	0	100
No log	2.0	150	1.0562	0.0	0.4533	0.0123	0.0039	0.0020	{'score': 68.84469696969697, 'num_edits': 727, 'ref_length': 1056.0}	0	0	0	0	0	100
No log	3.0	225	1.0520	0.0	0.5267	0.0133	0.0041	0.0021	{'score': 33.61742424242424, 'num_edits': 355, 'ref_length': 1056.0}	0	0	0	0	0	100
No log	4.0	300	1.0702	0.0	0.5167	0.0131	0.0041	0.0021	{'score': 27.84090909090909, 'num_edits': 294, 'ref_length': 1056.0}	0	0	0	0	0	100
No log	5.0	375	1.0654	0.0	0.5367	0.0134	0.0041	0.0021	{'score': 25.47348484848485, 'num_edits': 269, 'ref_length': 1056.0}	0	0	0	0	0	100
No log	6.0	450	1.1181	0.0	0.4667	0.0125	0.0040	0.0020	{'score': 28.219696969696972, 'num_edits': 298, 'ref_length': 1056.0}	0	0	0	0	0	100
0.5741	7.0	525	1.0898	0.0	0.4933	0.0128	0.0040	0.0021	{'score': 25.189393939393938, 'num_edits': 266, 'ref_length': 1056.0}	0	0	0	0	0	100
0.5741	8.0	600	1.0755	0.0	0.5	0.0129	0.0040	0.0021	{'score': 25.568181818181817, 'num_edits': 270, 'ref_length': 1056.0}	0	0	0	0	0	100
0.5741	9.0	675	1.1166	0.0	0.4933	0.0128	0.0040	0.0021	{'score': 25.28409090909091, 'num_edits': 267, 'ref_length': 1056.0}	0	0	0	0	0	100
0.5741	10.0	750	1.0323	0.0	0.51	0.0131	0.0041	0.0021	{'score': 24.33712121212121, 'num_edits': 257, 'ref_length': 1056.0}	0	0	0	0	0	100
0.5741	11.0	825	1.0523	0.0	0.4933	0.0128	0.0040	0.0021	{'score': 25.28409090909091, 'num_edits': 267, 'ref_length': 1056.0}	0	0	0	0	0	100
0.5741	12.0	900	1.0616	0.0	0.5133	0.0131	0.0041	0.0021	{'score': 24.242424242424242, 'num_edits': 256, 'ref_length': 1056.0}	0	0	0	0	0	100
0.5741	13.0	975	1.1157	0.0	0.4833	0.0127	0.0040	0.0021	{'score': 27.178030303030305, 'num_edits': 287, 'ref_length': 1056.0}	0	0	0	0	0	100
0.0462	14.0	1050	1.1162	0.0	0.5133	0.0131	0.0041	0.0021	{'score': 24.71590909090909, 'num_edits': 261, 'ref_length': 1056.0}	0	0	0	0	0	100
0.0462	15.0	1125	1.1162	0.0	0.5	0.0129	0.0040	0.0021	{'score': 25.66287878787879, 'num_edits': 271, 'ref_length': 1056.0}	0	0	0	0	0	100
0.0462	16.0	1200	1.1202	0.0	0.52	0.0132	0.0041	0.0021	{'score': 24.810606060606062, 'num_edits': 262, 'ref_length': 1056.0}	0	0	0	0	0	100
0.0462	17.0	1275	1.0917	0.0	0.51	0.0131	0.0041	0.0021	{'score': 23.863636363636363, 'num_edits': 252, 'ref_length': 1056.0}	0	0	0	0	0	100
0.0462	18.0	1350	1.1124	0.0	0.5167	0.0131	0.0041	0.0021	{'score': 24.147727272727273, 'num_edits': 255, 'ref_length': 1056.0}	0	0	0	0	0	100
0.0462	19.0	1425	1.1285	0.0	0.5033	0.0130	0.0041	0.0021	{'score': 25.189393939393938, 'num_edits': 266, 'ref_length': 1056.0}	0	0	0	0	0	100
0.0201	20.0	1500	1.1367	0.0	0.5167	0.0131	0.0041	0.0021	{'score': 24.242424242424242, 'num_edits': 256, 'ref_length': 1056.0}	0	0	0	0	0	100
0.0201	21.0	1575	1.1441	0.0	0.5167	0.0131	0.0041	0.0021	{'score': 24.147727272727273, 'num_edits': 255, 'ref_length': 1056.0}	0	0	0	0	0	100
0.0201	22.0	1650	1.1455	0.0	0.5133	0.0131	0.0041	0.0021	{'score': 23.768939393939394, 'num_edits': 251, 'ref_length': 1056.0}	0	0	0	0	0	100
0.0201	23.0	1725	1.1541	0.0	0.5167	0.0131	0.0041	0.0021	{'score': 23.768939393939394, 'num_edits': 251, 'ref_length': 1056.0}	0	0	0	0	0	100
0.0201	24.0	1800	1.1756	0.0	0.5333	0.0134	0.0041	0.0021	{'score': 23.863636363636363, 'num_edits': 252, 'ref_length': 1056.0}	0	0	0	0	0	100
0.0201	25.0	1875	1.1610	0.0	0.5133	0.0131	0.0041	0.0021	{'score': 24.526515151515152, 'num_edits': 259, 'ref_length': 1056.0}	0	0	0	0	0	100
0.0201	26.0	1950	1.1643	0.0	0.5267	0.0133	0.0041	0.0021	{'score': 23.768939393939394, 'num_edits': 251, 'ref_length': 1056.0}	0	0	0	0	0	100
0.0144	27.0	2025	1.1606	0.0	0.5233	0.0132	0.0041	0.0021	{'score': 24.053030303030305, 'num_edits': 254, 'ref_length': 1056.0}	0	0	0	0	0	100
0.0144	28.0	2100	1.1621	0.0	0.52	0.0132	0.0041	0.0021	{'score': 24.62121212121212, 'num_edits': 260, 'ref_length': 1056.0}	0	0	0	0	0	100
0.0144	29.0	2175	1.1637	0.0	0.52	0.0132	0.0041	0.0021	{'score': 24.62121212121212, 'num_edits': 260, 'ref_length': 1056.0}	0	0	0	0	0	100
0.0144	30.0	2250	1.1626	0.0	0.52	0.0132	0.0041	0.0021	{'score': 24.526515151515152, 'num_edits': 259, 'ref_length': 1056.0}	0	0	0	0	0	100

Framework versions

• Transformers 4.49.0
• Pytorch 2.5.1+cu124
• Datasets 3.3.2
• Tokenizers 0.21.0