sumarize_model_pegasus_v1

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

• Loss: 1.3379
• Rouge1: 0.6034
• Rouge2: 0.4459
• Rougel: 0.5685
• Rougelsum: 0.5681
• Gen Len: 32.8647

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: 3.419313942464226e-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: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	239	1.4418	0.6747	0.5033	0.6338	0.6335	43.9549
No log	2.0	478	1.3434	0.6869	0.5148	0.646	0.6459	44.938
1.8531	3.0	717	1.2791	0.6843	0.5141	0.6451	0.645	44.7556
1.8531	4.0	956	1.2358	0.6868	0.5168	0.6473	0.647	44.4305
1.4419	5.0	1195	1.2654	0.6858	0.5172	0.6467	0.6464	43.7857
1.4419	6.0	1434	1.2838	0.6686	0.4999	0.6291	0.6288	39.9549
1.4368	7.0	1673	1.3379	0.6034	0.4459	0.5685	0.5681	32.8647

Framework versions

• Transformers 4.38.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2