6b7e8dbcfa325f2dff0be9a3c2a42d52

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

• Loss: 1.5780
• Data Size: 1.0
• Epoch Runtime: 792.0022
• Bleu: 18.1848

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	4.6643	0	64.8101	2.3672
No log	1	3177	3.0936	0.0078	70.9656	4.4480
0.0499	2	6354	2.6493	0.0156	77.8951	5.1797
2.4554	3	9531	2.3092	0.0312	89.5178	6.4319
2.1419	4	12708	2.0331	0.0625	113.0354	8.0996
1.8453	5	15885	1.7854	0.125	157.7180	9.8873
1.608	6	19062	1.5961	0.25	246.8548	12.1049
1.3993	7	22239	1.4368	0.5	428.2038	14.0453
1.2302	8.0	25416	1.3191	1.0	788.4445	21.4663
1.0124	9.0	28593	1.2916	1.0	787.8098	19.4540
0.8958	10.0	31770	1.3239	1.0	789.8504	20.1527
0.7418	11.0	34947	1.3893	1.0	789.6639	16.8191
0.5937	12.0	38124	1.4621	1.0	790.5627	14.4724
0.4807	13.0	41301	1.5780	1.0	792.0022	18.1848

Framework versions

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