amb-dataset-factor

This model is a fine-tuned version of facebook/bart-large-cnn on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.8578
• Rouge1: 0.6039
• Rouge2: 0.3487
• Rougel: 0.4805
• Rougelsum: 0.4805
• Gen Len: 101.0

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: 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: 8
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	1	1.1889	0.5926	0.2701	0.3828	0.3828	65.5
No log	2.0	2	1.0179	0.6489	0.3333	0.458	0.458	77.5
No log	3.0	3	0.9405	0.6084	0.2627	0.3783	0.3783	82.5
No log	4.0	4	0.8990	0.6241	0.3058	0.4054	0.4054	86.0
No log	5.0	5	0.8814	0.6746	0.3882	0.4842	0.4842	95.0
No log	6.0	6	0.8679	0.5554	0.3111	0.4127	0.4127	94.5
No log	7.0	7	0.8607	0.5799	0.3016	0.4153	0.4153	100.0
No log	8.0	8	0.8578	0.6039	0.3487	0.4805	0.4805	101.0

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.1
• Tokenizers 0.19.1