ca69167df914e0f83ef6e29f311a765d

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

• Loss: 1.7214
• Data Size: 1.0
• Epoch Runtime: 35.9118
• Bleu: 9.0342

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	7.3636	0	1.7041	0.0590
No log	1	29	6.1402	0.0078	2.1097	0.0685
No log	2	58	5.0734	0.0156	9.4621	0.0979
No log	3	87	4.3698	0.0312	14.6535	0.1653
No log	4	116	3.7734	0.0625	19.8736	0.2722
No log	5	145	3.0995	0.125	25.1294	0.5249
0.557	6	174	2.5947	0.25	27.5075	0.9068
0.557	7	203	2.1346	0.5	25.4466	1.1695
0.557	8.0	232	1.7368	1.0	37.2337	7.0602
1.5595	9.0	261	1.6552	1.0	38.3491	7.6295
1.5595	10.0	290	1.6188	1.0	35.9982	8.1920
1.6322	11.0	319	1.6268	1.0	28.6706	8.8259
1.6322	12.0	348	1.6106	1.0	32.0709	9.1815
1.3657	13.0	377	1.6265	1.0	35.6865	9.0679
1.1379	14.0	406	1.6546	1.0	27.0253	9.3339
1.1379	15.0	435	1.6998	1.0	30.0941	9.1016
0.9634	16.0	464	1.7214	1.0	35.9118	9.0342

Framework versions

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