distilbart-cnn-12-6-finetuned-30k-3epoch

This model is a fine-tuned version of sshleifer/distilbart-cnn-12-6 on the arxiv_summarization_dataset dataset. It achieves the following results on the evaluation set:

• Loss: 2.3411
• Rouge1: 43.696
• Rouge2: 15.6681
• Rougel: 25.6889
• Rougelsum: 38.574
• Gen Len: 121.98

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

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
2.7304	1.0	3750	2.4322	43.0913	15.1302	25.2555	38.0346	122.3755
2.3518	2.0	7500	2.3613	43.8799	15.6977	25.6984	38.7646	122.6945
2.2318	3.0	11250	2.3411	43.696	15.6681	25.6889	38.574	121.98

Framework versions

• Transformers 4.30.2
• Pytorch 2.0.0
• Datasets 2.1.0
• Tokenizers 0.13.3