mt5-small-finetuned-icelandic-summary-finetuned-icelandic-summary

This model is a fine-tuned version of nozagleh/mt5-small-finetuned-icelandic-summary on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.0847
• Rouge1: 24.7758
• Rouge2: 13.6541
• Rougel: 22.0304
• Rougelsum: 22.8727

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: 5.6e-05
• train_batch_size: 6
• eval_batch_size: 6
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 12

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
2.6294	1.0	2552	2.1826	23.9594	13.1502	21.2044	22.1031
2.5328	2.0	5104	2.1888	24.0688	13.178	21.3606	22.1735
2.4571	3.0	7656	2.1371	24.1003	13.3883	21.4866	22.3277
2.4024	4.0	10208	2.1331	24.2949	13.2282	21.5826	22.4117
2.3513	5.0	12760	2.1198	24.1912	13.2633	21.5876	22.3797
2.3141	6.0	15312	2.1283	24.3672	13.2826	21.5934	22.472
2.2853	7.0	17864	2.0878	24.5056	13.3639	21.7807	22.6229
2.2567	8.0	20416	2.0952	24.4647	13.428	21.7303	22.6027
2.2373	9.0	22968	2.0908	24.5012	13.3905	21.7448	22.6278
2.2203	10.0	25520	2.0889	24.5345	13.4032	21.7559	22.6362
2.2033	11.0	28072	2.0857	24.7518	13.5923	21.9905	22.8425
2.199	12.0	30624	2.0847	24.7758	13.6541	22.0304	22.8727

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0