6c68655df2a22d3b9c0dd9063ae71eb2

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

• Loss: 1.4914
• Data Size: 1.0
• Epoch Runtime: 376.7987
• Bleu: 6.4684

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	3.1643	0	28.3105	0.2899
No log	1	688	2.2539	0.0078	33.1415	2.6581
No log	2	1376	2.1601	0.0156	37.6501	3.5815
No log	3	2064	2.0682	0.0312	46.9903	3.8680
0.0944	4	2752	1.9749	0.0625	59.7987	3.8130
0.1857	5	3440	1.8820	0.125	79.6783	3.9899
2.128	6	4128	1.7873	0.25	122.4464	3.9539
1.9461	7	4816	1.6862	0.5	209.1218	4.8308
1.773	8.0	5504	1.5773	1.0	378.6916	5.2791
1.6221	9.0	6192	1.5133	1.0	377.2166	5.7557
1.5006	10.0	6880	1.4726	1.0	375.3109	5.7854
1.3912	11.0	7568	1.4527	1.0	376.6940	6.0125
1.2966	12.0	8256	1.4375	1.0	378.9779	6.2890
1.2177	13.0	8944	1.4357	1.0	376.3818	6.5334
1.1423	14.0	9632	1.4417	1.0	377.9785	6.3076
1.0593	15.0	10320	1.4536	1.0	378.0307	6.4652
0.9837	16.0	11008	1.4671	1.0	376.3978	6.4866
0.9372	17.0	11696	1.4914	1.0	376.7987	6.4684

Framework versions

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