bart-large-cnn-weaksup-10k-NOpad-early

This model is a fine-tuned version of facebook/bart-large-cnn on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.7883
• Rouge1: 26.9755
• Rouge2: 12.4975
• Rougel: 21.0743
• Rougelsum: 23.9303
• Gen Len: 69.549

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: 1
• eval_batch_size: 1
• 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
1.4657	1.0	10000	1.7295	27.973	13.2818	21.8493	25.0101	67.831
0.8522	2.0	20000	1.7883	26.9755	12.4975	21.0743	23.9303	69.549

Framework versions

• Transformers 4.16.2
• Pytorch 1.10.2
• Datasets 1.18.3
• Tokenizers 0.11.0