1678df87f10419907d2e000544ffa40a

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

• Loss: 2.9845
• Data Size: 1.0
• Epoch Runtime: 3.1285
• Bleu: 3.5119

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	7.2877	0	0.8417	0.2936
No log	1	29	6.7048	0.0078	1.1268	0.2841
No log	2	58	6.4060	0.0156	0.9913	0.2813
No log	3	87	6.2081	0.0312	1.0765	0.2774
No log	4	116	5.9462	0.0625	1.9175	0.3064
No log	5	145	5.5637	0.125	1.5853	0.5214
0.5562	6	174	5.0549	0.25	1.8226	0.7445
0.5562	7	203	4.4647	0.5	2.2366	1.2109
0.5562	8.0	232	3.9015	1.0	3.3107	1.8861
2.7734	9.0	261	3.6161	1.0	3.0992	2.1291
2.7734	10.0	290	3.4240	1.0	2.8096	2.0725
3.3225	11.0	319	3.2781	1.0	3.2782	2.6376
3.3225	12.0	348	3.1764	1.0	3.1320	2.8693
2.9416	13.0	377	3.1274	1.0	3.1755	2.9883
2.5954	14.0	406	3.0734	1.0	3.1548	3.0889
2.5954	15.0	435	3.0385	1.0	3.7395	3.1426
2.3357	16.0	464	3.0012	1.0	3.2131	3.1732
2.3357	17.0	493	2.9924	1.0	3.3260	3.2905
2.1228	18.0	522	2.9740	1.0	2.8296	3.2611
1.9145	19.0	551	2.9821	1.0	3.0264	3.2914
1.9145	20.0	580	2.9791	1.0	2.8122	3.3617
1.708	21.0	609	2.9789	1.0	2.9019	3.5669
1.708	22.0	638	2.9845	1.0	3.1285	3.5119

Framework versions

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