DanSumT5-largeV_38143

This model is a fine-tuned version of Danish-summarisation/DanSumT5-large on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.9493
• Rouge1: 36.0009
• Rouge2: 12.4957
• Rougel: 22.4757
• Rougelsum: 33.603
• Gen Len: 125.1519

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: 5e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 11

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	0.99	118	2.1442	34.5887	11.0101	20.7128	32.0179	126.1603
No log	2.0	237	2.0645	34.9243	11.2939	21.4174	32.5111	125.7342
No log	3.0	356	2.0220	35.3258	11.7644	21.6971	32.9952	125.3713
No log	4.0	475	1.9962	35.4098	11.883	21.6822	33.0259	124.7553
2.2257	4.99	593	1.9807	36.0239	12.4173	22.5759	33.6794	125.038
2.2257	6.0	712	1.9656	35.847	12.3976	22.3612	33.5817	125.1646
2.2257	7.0	831	1.9606	35.6499	12.1119	22.3208	33.3221	124.6751
2.2257	8.0	950	1.9533	35.6641	12.3948	22.4477	33.3256	124.6878
1.9052	8.99	1068	1.9559	35.7509	12.4988	22.6593	33.4865	124.4937
1.9052	10.0	1187	1.9508	35.9176	12.548	22.6092	33.6489	125.0295
1.9052	10.93	1298	1.9493	36.0009	12.4957	22.4757	33.603	125.1519

Framework versions

• Transformers 4.30.2
• Pytorch 1.12.1+git7548e2f
• Datasets 2.13.2
• Tokenizers 0.13.3