mbart-large-50-many-to-many-mmt-finetuned-test2

This model is a fine-tuned version of facebook/mbart-large-50-many-to-many-mmt on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.9283
• Rouge1: 28.015
• Rouge2: 11.5757
• Rougel: 23.4706
• Rougelsum: 27.303
• Gen Len: 48.8231

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: 2e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 4
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
2.1456	1.0	2130	2.0695	24.7024	9.6769	20.91	24.0793	45.739
1.8957	2.0	4261	1.9656	25.8455	10.4715	21.9971	25.1654	50.9408
1.7359	3.0	6391	1.9281	26.9356	11.0235	22.709	26.197	51.5054
1.6207	4.0	8520	1.9283	28.015	11.5757	23.4706	27.303	48.8231

Framework versions

• Transformers 4.27.2
• Pytorch 1.13.0+cu117
• Datasets 2.7.1
• Tokenizers 0.13.2