e5540451fc864a7ff2c2e43201236549

This model is a fine-tuned version of facebook/mbart-large-50 on the Helsinki-NLP/opus_books [en-pl] dataset. It achieves the following results on the evaluation set:

• Loss: 2.7909
• Data Size: 1.0
• Epoch Runtime: 21.4718
• Bleu: 2.4250

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	8.8037	0	2.0774	0.5073
No log	1	70	7.4982	0.0078	2.6980	0.7815
No log	2	140	6.5341	0.0156	4.3990	0.7601
No log	3	210	6.2822	0.0312	5.4365	1.1194
No log	4	280	5.9489	0.0625	6.5229	1.4615
No log	5	350	5.5108	0.125	9.1531	1.6819
No log	6	420	4.4852	0.25	11.1909	1.9624
0.7002	7	490	2.7826	0.5	13.9265	2.2404
4.0454	8.0	560	17.5607	1.0	23.7272	0.0045
13.3049	9.0	630	6.4795	1.0	24.3997	0.0076
4.9579	10.0	700	3.0659	1.0	21.6058	2.1608
2.0529	11.0	770	2.7909	1.0	21.4718	2.4250

Framework versions

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