7b76ba995d5fc3a95be7325936017a22

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

• Loss: 2.2376
• Data Size: 1.0
• Epoch Runtime: 44.6259
• Bleu: 3.4514

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	7.5703	0	2.8772	0.0084
No log	1	70	6.4477	0.0078	3.3209	0.0138
No log	2	140	4.5614	0.0156	7.9915	0.0562
No log	3	210	4.0719	0.0312	12.8319	0.0990
No log	4	280	3.7587	0.0625	21.5003	0.1498
No log	5	350	3.2352	0.125	22.6846	0.2207
No log	6	420	2.5822	0.25	23.2475	1.2481
0.579	7	490	2.3332	0.5	30.5218	2.1480
2.5736	8.0	560	2.1905	1.0	51.6880	2.3556
2.3462	9.0	630	2.1524	1.0	41.5992	2.6776
2.0395	10.0	700	2.1307	1.0	45.9313	2.9177
1.8266	11.0	770	2.1372	1.0	42.2034	3.1032
1.685	12.0	840	2.1640	1.0	45.1173	3.5080
1.4818	13.0	910	2.2000	1.0	41.5260	3.5106
1.3597	14.0	980	2.2376	1.0	44.6259	3.4514

Framework versions

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