LazarusNLP/IndoNanoT5-base-Liputan6-Canonical

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

• Loss: 1.1194
• Rouge1: 0.3976
• Rouge2: 0.2229
• Rougel: 0.3346
• Rougelsum: 0.3345
• Gen Len: 43.3808

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: 1e-05
• train_batch_size: 8
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
0.9693	1.0	24236	1.3245	0.3082	0.1585	0.2687	0.2688	18.9956
0.9338	2.0	48472	1.2759	0.3105	0.159	0.2705	0.2706	18.9985
0.8632	3.0	72708	1.2698	0.3094	0.1586	0.2701	0.2702	18.9995
0.8257	4.0	96944	1.2631	0.312	0.1603	0.2716	0.2715	18.9993
0.7789	5.0	121180	1.2642	0.3149	0.1625	0.2748	0.2747	18.9998
0.7595	6.0	145416	1.2587	0.3202	0.1658	0.279	0.2791	18.9995
0.7343	7.0	169652	1.2644	0.3183	0.1647	0.2773	0.2773	18.9996
0.7165	8.0	193888	1.2635	0.3141	0.1605	0.2732	0.2732	18.9993
0.6697	9.0	218124	1.2856	0.316	0.162	0.275	0.275	18.9998
0.6729	10.0	242360	1.2809	0.3195	0.164	0.2775	0.2776	18.9992
0.6471	11.0	266596	1.2833	0.3185	0.1636	0.2769	0.277	18.9982

Framework versions

• Transformers 4.37.2
• Pytorch 2.2.0+cu118
• Datasets 2.16.1
• Tokenizers 0.15.1