mBart50-tl-en2vi-lora

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

• Loss: 6.6172
• Bleu: 34.4063
• Gen Len: 32.7321

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: 0.0005
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
6.9999	0.21	2000	6.6880	32.0922	32.1749
6.9674	0.42	4000	6.6544	32.7344	32.7289
6.9357	0.63	6000	6.6310	34.0802	32.4878
6.9162	0.84	8000	6.6172	34.4063	32.7321

Framework versions

• PEFT 0.9.1.dev0
• Transformers 4.39.0.dev0
• Pytorch 2.1.2
• Datasets 2.1.0
• Tokenizers 0.15.2