flan-t5-base-flant5-apple-support

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

• Loss: 3.0475
• Rouge1: 12.4139
• Rouge2: 2.0562
• Rougel: 9.4938
• Rougelsum: 11.0524
• Gen Len: 18.9589

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: 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: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	232	3.0886	12.844	2.1734	9.8971	11.3641	18.8876
No log	2.0	464	3.0639	12.2909	2.1209	9.4999	10.9458	18.9416
3.3185	3.0	696	3.0538	12.4154	2.0984	9.4989	11.0684	18.9492
3.3185	4.0	928	3.0489	12.7043	2.1969	9.7356	11.3629	18.9481
3.187	5.0	1160	3.0475	12.4139	2.0562	9.4938	11.0524	18.9589

Framework versions

• Transformers 4.25.1
• Pytorch 1.13.1+cu117
• Datasets 2.8.0
• Tokenizers 0.13.2