pegasus-newsroom-headline_writer_57k

This model is a fine-tuned version of google/pegasus-newsroom on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.3599
• Rouge1: 42.2586
• Rouge2: 23.2731
• Rougel: 35.8685
• Rougelsum: 36.0581
• Gen Len: 34.3651

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: 3
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
1.5213	1.0	5670	1.4040	41.8648	22.8205	35.3983	35.535	34.8817
1.4171	2.0	11340	1.3672	42.26	23.2611	35.8016	35.9753	34.3492
1.3722	3.0	17010	1.3599	42.2586	23.2731	35.8685	36.0581	34.3651

Framework versions

• Transformers 4.23.1
• Pytorch 1.12.1+cu113
• Datasets 2.6.0
• Tokenizers 0.13.1