70435a114d9151ba6ee9c3a4921f4481

This model is a fine-tuned version of google/long-t5-tglobal-xl on the Helsinki-NLP/opus_books [es-pt] dataset. It achieves the following results on the evaluation set:

• Loss: 1.5273
• Data Size: 1.0
• Epoch Runtime: 39.2990
• Bleu: 5.2520

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 32
• total_eval_batch_size: 32
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Data Size	Epoch Runtime	Bleu
No log	0	0	3.0413	0	2.2961	0.5017
No log	1	33	2.6037	0.0078	3.0387	0.9330
No log	2	66	2.2827	0.0156	6.6236	2.0581
No log	3	99	2.1202	0.0312	13.8058	3.2753
0.1559	4	132	2.0323	0.0625	17.2287	2.7933
0.1559	5	165	1.9536	0.125	22.8074	2.5654
0.1559	6	198	1.8495	0.25	24.0448	2.7841
0.4483	7	231	1.7570	0.5	25.6411	3.0927
1.4034	8.0	264	1.6347	1.0	40.8148	3.8026
1.4034	9.0	297	1.5785	1.0	40.2604	4.1157
1.7971	10.0	330	1.5634	1.0	40.5761	4.3332
1.5849	11.0	363	1.5438	1.0	32.7149	4.5080
1.5849	12.0	396	1.5254	1.0	33.8319	4.5447
1.4254	13.0	429	1.5281	1.0	37.3783	4.8787
1.3017	14.0	462	1.5260	1.0	33.4821	5.0125
1.3017	15.0	495	1.5308	1.0	36.5327	5.0939
1.1771	16.0	528	1.5273	1.0	39.2990	5.2520

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.2.0
• Tokenizers 0.22.1