mt0

This model is a fine-tuned version of bigscience/mt0-base on the Multi-Document Vietnamese dataset. It achieves the following results on the evaluation set:

• Loss: 0.7265121936798096
• Rouge-1: 0.2592
• Rouge-2: 0.1412
• Rouge-4: 0.0788
• Rouge-l: 0.2212
• Rouge-w-1.2: 0.0901
• Rouge-s4: 0.1056
• Rouge-su4: 0.1323
• R: 0.1982
• P: 0.3745
• Bleu: 10.0657

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: 1
• eval_batch_size: 1
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge-1	Rouge-2	Rouge-4	Rouge-l	Rouge-w-1.2	Rouge-s4	Rouge-su4	R	P	Bleu
17.8707	1.0	628	15.0238	0.1453	0.078	0.0336	0.1293	0.0477	0.0524	0.0682	0.0838	0.5448	2.131
0.9125	2.0	1256	0.8100	0.2335	0.1331	0.0727	0.2029	0.0739	0.0978	0.1211	0.148	0.5531	6.1757
0.8117	3.0	1884	0.7965	0.2785	0.1583	0.0898	0.2429	0.0949	0.1195	0.1469	0.1994	0.4618	11.2862
0.881	4.0	2512	0.7945	0.2816	0.1597	0.0897	0.2457	0.0975	0.1195	0.1475	0.2099	0.4276	10.3668

Framework versions

• Transformers 4.26.1
• Pytorch 2.0.0
• Datasets 2.1.0
• Tokenizers 0.13.3