my_awesome_billsum_model_32

This model is a fine-tuned version of google-t5/t5-small on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.6953
• Rouge1: 0.9328
• Rouge2: 0.7819
• Rougel: 0.8546
• Rougelsum: 0.857
• Gen Len: 5.2083

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: 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: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	12	2.1371	0.3875	0.2627	0.3679	0.3672	17.5625
No log	2.0	24	1.6293	0.4137	0.286	0.3928	0.3924	16.5625
No log	3.0	36	1.1556	0.478	0.347	0.4538	0.4505	14.6458
No log	4.0	48	0.8702	0.7792	0.6558	0.7344	0.7311	8.2708
No log	5.0	60	0.7539	0.9328	0.7819	0.8546	0.857	5.2083
No log	6.0	72	0.7230	0.9328	0.7819	0.8546	0.857	5.2083
No log	7.0	84	0.7094	0.9328	0.7819	0.8546	0.857	5.2083
No log	8.0	96	0.7026	0.9328	0.7819	0.8546	0.857	5.2083
No log	9.0	108	0.6978	0.9328	0.7819	0.8546	0.857	5.2083
No log	10.0	120	0.6953	0.9328	0.7819	0.8546	0.857	5.2083

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1