indosum-lora-0

This model is a fine-tuned version of LazarusNLP/IndoNanoT5-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.4876
• Rouge1: 73.5772
• Rouge2: 66.7059
• Rougel: 70.6615
• Rougelsum: 72.7397
• Gen Len: 102.9213

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: 0.001
• train_batch_size: 16
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 5.0

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
0.7749	1.0	892	0.5290	71.3679	64.0441	68.2584	70.462	103.084
0.5963	2.0	1784	0.5043	71.9204	64.8766	68.9265	71.053	105.9827
0.5524	3.0	2676	0.4928	72.0196	65.0022	69.1173	71.1808	103.2227
0.5236	4.0	3568	0.4948	72.454	65.5465	69.6508	71.6454	105.364
0.5017	5.0	4460	0.4876	73.194	66.308	70.335	72.3629	102.86

Framework versions

• Transformers 4.40.2
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1