bart-mawpnli-calcx

This model is a fine-tuned version of facebook/bart-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1721
• Rouge1: 95.182
• Rouge2: 88.0785
• Rougel: 95.0205
• Rougelsum: 95.0114
• Gen Len: 14.5892

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: 5e-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: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	426	0.1673	94.5945	86.1233	94.3034	94.3197	14.4437
0.2429	2.0	852	0.1457	94.7631	86.9333	94.5539	94.5633	14.4366
0.073	3.0	1278	0.1489	94.7349	87.5738	94.5739	94.5773	14.5833
0.0462	4.0	1704	0.1710	95.2312	88.2565	95.0116	94.9837	14.5646
0.0214	5.0	2130	0.1721	95.182	88.0785	95.0205	95.0114	14.5892

Framework versions

• Transformers 4.35.2
• Pytorch 1.12.1+cu113
• Datasets 2.15.0
• Tokenizers 0.15.0