jeongyoun/t5-FairytaleQA-QG

t5-base-finetuned-FairytaleQA-QuestionGeneration

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

• Loss: 0.1497
• Rouge1: 41.8335
• Rouge2: 22.9223
• Rougel: 39.7489
• Rougelsum: 39.7165
• F1: 1.0732
• Exact Match: 1.0732
• Gen Len: 15.721

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: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	F1	Exact Match	Gen Len
0.1477	1.0	2137	0.1519	41.844	22.9178	39.6972	39.6635	1.1707	1.1707	15.7502
0.156	2.0	4274	0.1499	41.7453	22.8275	39.6485	39.6119	1.0732	1.0732	15.76
0.1539	3.0	6411	0.1497	41.8402	22.9432	39.7493	39.7177	1.0732	1.0732	15.7463
0.1529	4.0	8548	0.1497	41.8242	22.9165	39.7328	39.7163	1.0732	1.0732	15.7249
0.1574	5.0	10685	0.1497	41.8335	22.9223	39.7489	39.7165	1.0732	1.0732	15.721

Framework versions

• Transformers 4.42.4
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1