Mikask/indobart-indonlg-jv-id

indobart-combined-jv-id

This model is a fine-tuned version of indobenchmark/indobart-v2 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.6285
• Bleu: 16.9732
• Gen Len: 19.4429

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: 64
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
No log	1.0	94	0.8084	11.1865	19.5232
No log	2.0	188	0.7259	13.1223	19.478
No log	3.0	282	0.6905	14.6859	19.4366
No log	4.0	376	0.6682	15.2578	19.4216
No log	5.0	470	0.6536	15.9459	19.4366
0.8734	6.0	564	0.6439	16.3189	19.4115
0.8734	7.0	658	0.6367	16.7353	19.4228
0.8734	8.0	752	0.6327	16.9575	19.4529
0.8734	9.0	846	0.6300	17.093	19.4442
0.8734	10.0	940	0.6285	16.9732	19.4429

Framework versions

• Transformers 4.33.1
• Pytorch 2.0.1+cu118
• Datasets 2.14.5
• Tokenizers 0.13.3