my_awesome_billsum_model_36

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: 0.4601
• Rouge1: 0.9721
• Rouge2: 0.8819
• Rougel: 0.9256
• Rougelsum: 0.9271
• Gen Len: 4.9167

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	1.9874	0.4145	0.2913	0.3883	0.3891	17.6042
No log	2.0	24	1.4300	0.4322	0.3091	0.4061	0.4068	17.0833
No log	3.0	36	0.9451	0.5076	0.3886	0.4814	0.48	14.75
No log	4.0	48	0.6345	0.8401	0.7297	0.7858	0.7884	7.625
No log	5.0	60	0.5226	0.9591	0.8586	0.8998	0.9042	5.125
No log	6.0	72	0.4907	0.9701	0.8736	0.9129	0.9167	4.8958
No log	7.0	84	0.4783	0.9701	0.8736	0.9129	0.9167	4.8958
No log	8.0	96	0.4697	0.9721	0.8819	0.9256	0.9271	4.9167
No log	9.0	108	0.4627	0.9721	0.8819	0.9256	0.9271	4.9167
No log	10.0	120	0.4601	0.9721	0.8819	0.9256	0.9271	4.9167

Framework versions

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