yesj1234/mbart-mmt_mid1_ko-ja

tst-translation-output

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: 0.8761
• Bleu: 11.9991
• Gen Len: 17.5742

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: 2
• eval_batch_size: 2
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 8
• total_train_batch_size: 64
• total_eval_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 35

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
1.014	0.91	1500	0.9561	5.9697	18.2307
0.7888	1.82	3000	0.8761	10.9388	17.4752
0.6166	2.73	4500	0.8778	11.0165	17.1855
0.474	3.64	6000	0.9008	11.8779	17.1838
0.3559	4.55	7500	0.9504	12.6024	17.3179
0.2588	5.47	9000	0.9978	13.4321	17.1999

Framework versions

• Transformers 4.34.0
• Pytorch 2.1.0+cu121
• Datasets 2.14.5
• Tokenizers 0.14.1