94b30d9cc2315e0cd9ff484731b38f52

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

• Loss: 1.4633
• Data Size: 1.0
• Epoch Runtime: 50.3594
• Bleu: 12.4260

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	7.8812	0	3.1567	0.0112
No log	1	77	5.3106	0.0078	3.5268	0.0472
No log	2	154	3.6828	0.0156	11.8525	0.3424
No log	3	231	3.0095	0.0312	17.9511	0.7707
No log	4	308	2.7073	0.0625	26.0137	1.0505
No log	5	385	2.2687	0.125	25.5012	1.3031
0.336	6	462	1.7849	0.25	29.1275	13.0199
0.9198	7	539	1.5161	0.5	43.3328	9.2252
1.7349	8.0	616	1.3972	1.0	65.3745	11.4792
1.4925	9.0	693	1.3574	1.0	51.1325	11.9868
1.2101	10.0	770	1.3493	1.0	53.2914	12.1847
1.1172	11.0	847	1.3935	1.0	48.7534	12.2443
0.9201	12.0	924	1.3828	1.0	48.4278	12.5264
0.8138	13.0	1001	1.4273	1.0	51.6848	12.3103
0.7079	14.0	1078	1.4633	1.0	50.3594	12.4260

Framework versions

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