23ea0df95097bbba58b6c7c938df866a

This model is a fine-tuned version of google/umt5-xl on the Helsinki-NLP/opus_books [es-it] dataset. It achieves the following results on the evaluation set:

• Loss: 2.1536
• Data Size: 1.0
• Epoch Runtime: 369.0926
• Bleu: 6.9930

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	5.5312	0	25.3861	1.7981
No log	1	721	4.1738	0.0078	27.9021	5.6825
No log	2	1442	3.3502	0.0156	35.3767	8.6623
0.0899	3	2163	2.9390	0.0312	45.5764	11.2225
0.2378	4	2884	2.5831	0.0625	60.6828	4.7548
2.9071	5	3605	2.3719	0.125	82.9186	5.0657
2.6677	6	4326	2.2692	0.25	121.6984	5.5592
2.5167	7	5047	2.1771	0.5	204.9287	5.9626
2.2863	8.0	5768	2.0880	1.0	376.8316	6.4973
2.0482	9.0	6489	2.0612	1.0	367.4358	6.7417
1.9162	10.0	7210	2.0510	1.0	371.9522	6.8470
1.7448	11.0	7931	2.0631	1.0	369.6082	6.9275
1.6574	12.0	8652	2.0783	1.0	369.6752	7.0343
1.5212	13.0	9373	2.1114	1.0	373.6983	7.0696
1.3695	14.0	10094	2.1536	1.0	369.0926	6.9930

Framework versions

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