my_awesome_dailymail_long_model

This model is a fine-tuned version of google/long-t5-tglobal-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.6759
• Rouge1: 0.2345
• Rouge2: 0.0986
• Rougel: 0.1918
• Rougelsum: 0.1921
• Gen Len: 18.999

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: 2e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 4
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
2.5452	0.5	500	1.7971	0.227	0.0911	0.1815	0.1815	18.9695
2.2571	1.0	1000	1.7336	0.2315	0.0941	0.1865	0.1865	18.995
2.1671	1.5	1500	1.7149	0.2322	0.0958	0.1885	0.1887	18.995
2.1123	2.0	2000	1.6973	0.2343	0.0967	0.1903	0.1904	18.999
2.0778	2.5	2500	1.6900	0.2349	0.0974	0.1913	0.1914	18.999
2.0609	3.0	3000	1.6833	0.2342	0.098	0.1914	0.1916	18.999
2.0466	3.5	3500	1.6780	0.2344	0.0983	0.1918	0.192	18.999
2.0454	4.0	4000	1.6759	0.2345	0.0986	0.1918	0.1921	18.999

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.0
• Datasets 3.0.0
• Tokenizers 0.19.1