4638a57cf9c284cf2d9e584796df0e34

This model is a fine-tuned version of google/mt5-xl on the Helsinki-NLP/opus_books [es-it] dataset. It achieves the following results on the evaluation set:

• Loss: 1.8483
• Data Size: 1.0
• Epoch Runtime: 340.3263
• Bleu: 6.0787

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	8.0956	0	22.6938	0.0155
No log	1	721	3.8000	0.0078	25.2927	0.2047
No log	2	1442	2.8711	0.0156	35.7791	0.4507
0.0886	3	2163	2.3693	0.0312	41.9169	0.7841
0.1979	4	2884	2.1639	0.0625	55.1904	3.6362
2.5077	5	3605	2.0440	0.125	79.4744	4.0698
2.322	6	4326	1.9617	0.25	115.1786	4.5888
2.1802	7	5047	1.8773	0.5	190.4928	5.4892
1.9764	8.0	5768	1.8104	1.0	349.7535	5.8973
1.7811	9.0	6489	1.7846	1.0	341.7203	5.9074
1.6643	10.0	7210	1.7706	1.0	344.9861	6.1212
1.5112	11.0	7931	1.7809	1.0	339.9801	6.1245
1.4353	12.0	8652	1.8014	1.0	339.7979	6.1819
1.3143	13.0	9373	1.8295	1.0	343.6585	6.1379
1.2033	14.0	10094	1.8483	1.0	340.3263	6.0787

Framework versions

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