wikisum

This model is a fine-tuned version of t5-small on an wikisum dataset. It achieves the following results on the evaluation set:

• Loss: 2.2922
• Rouge1: 0.1811
• Rouge2: 0.0673
• Rougel: 0.147
• Rougelsum: 0.147
• Gen Len: 19.0

Model description

t5-small model fine-tuned on wikisum dataset.

Intended uses & limitations

Intended use: sumamrization of informatic articles. Limitations : may generate misleading information.

Training and evaluation data

check out wikisum dataset

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 1
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
2.5807	0.2236	500	2.3647	0.1813	0.0635	0.1452	0.1453	19.0
2.5059	0.4472	1000	2.3190	0.1823	0.0663	0.1473	0.1473	19.0
2.4945	0.6708	1500	2.3003	0.1808	0.0666	0.1468	0.1467	19.0
2.4963	0.8945	2000	2.2922	0.1811	0.0673	0.147	0.147	19.0

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.2
• Tokenizers 0.19.1