Large-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: 0.8394
• Rouge1: 0.6016
• Rouge2: 0.3238
• Rougel: 0.3867
• Rougelsum: 0.3867
• Gen Len: 142.0

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	1.2175	0.4598	0.2293	0.3085	0.3085	75.5
No log	2.0	2	1.0135	0.5862	0.3326	0.432	0.432	114.5
No log	3.0	3	0.9291	0.5584	0.2891	0.3831	0.3831	142.0
No log	4.0	4	0.8851	0.5572	0.2773	0.3739	0.3739	142.0
No log	5.0	5	0.8642	0.5822	0.3125	0.3886	0.3886	142.0
No log	6.0	6	0.8517	0.5725	0.2977	0.3692	0.3692	142.0
No log	7.0	7	0.8427	0.6016	0.3238	0.3867	0.3867	142.0
No log	8.0	8	0.8394	0.6016	0.3238	0.3867	0.3867	142.0

Framework versions

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