FLAN-T5-Base-Finetune-Remarks

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

• Loss: 0.7651
• Rouge1: 40.7065
• Rouge2: 22.2783
• Rougel: 34.3087
• Rougelsum: 34.3031
• Gen Len: 19.0

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: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
1.1215	1.0	1212	0.8802	39.125	19.456	32.7394	32.7269	19.0
0.9791	2.0	2424	0.8018	40.5473	21.5742	34.193	34.1905	19.0
0.9085	3.0	3636	0.7728	40.3545	21.7619	33.8864	33.8738	19.0
0.8884	4.0	4848	0.7651	40.7065	22.2783	34.3087	34.3031	19.0

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.0
• Datasets 2.21.0
• Tokenizers 0.19.1