bart-large-cnn-weaksup-1000-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: 1.9082
• Rouge1: 26.9663
• Rouge2: 11.3027
• Rougel: 20.7327
• Rougelsum: 23.5965
• Gen Len: 67.19

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
1.4775	1.0	1000	1.6796	27.208	12.01	20.8401	24.1333	66.06
0.6972	2.0	2000	1.9082	26.9663	11.3027	20.7327	23.5965	67.19

Framework versions

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