longt5_xl_gov_memsum_25

This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.3918

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: 0.001
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 32
• total_train_batch_size: 256
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant
• num_epochs: 15.0

Training results

Training Loss	Epoch	Step	Validation Loss
0.4372	1.0	68	1.6270
0.3678	1.99	136	1.8330
0.3026	2.99	204	1.8467
0.2785	3.99	272	1.9830
0.2489	5.0	341	2.1279
0.181	6.0	409	2.2981
0.1753	6.99	477	2.3683
0.1511	7.99	545	2.3130
0.1483	8.99	613	2.5342
0.2277	10.0	682	2.3054
0.1952	10.99	750	2.2331
0.1773	11.99	818	2.1944
0.1524	12.99	886	2.3607
0.1373	14.0	955	2.3946
0.1238	14.95	1020	2.3918

Framework versions

• Transformers 4.38.1
• Pytorch 2.2.1+cu121
• Datasets 2.17.1
• Tokenizers 0.15.2