377f91c43b0284c9eaf52daa10aaa9ac

This model is a fine-tuned version of facebook/mbart-large-50 on the Helsinki-NLP/opus_books [en-no] dataset. It achieves the following results on the evaluation set:

• Loss: 3.6648
• Data Size: 1.0
• Epoch Runtime: 24.8851
• Bleu: 7.9423

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	10.6337	0	2.3724	0.2620
No log	1	87	9.6939	0.0078	2.7900	0.5333
No log	2	174	8.7310	0.0156	4.3838	0.6226
No log	3	261	8.1748	0.0312	5.7734	1.1240
No log	4	348	7.4870	0.0625	7.1214	1.3134
0.3363	5	435	6.0535	0.125	9.3449	1.8334
1.4493	6	522	4.6613	0.25	12.0517	2.7687
1.586	7	609	3.7135	0.5	15.6271	4.3622
1.9341	8.0	696	3.0631	1.0	27.2023	9.1069
3.2925	9.0	783	2.9832	1.0	26.2388	10.5508
1.8198	10.0	870	2.9413	1.0	25.0626	11.1426
1.2999	11.0	957	3.1459	1.0	25.5639	8.9503
0.915	12.0	1044	3.3633	1.0	25.9204	12.6457
0.5695	13.0	1131	3.5643	1.0	24.9361	9.8916
0.4108	14.0	1218	3.6648	1.0	24.8851	7.9423

Framework versions

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