test_sum_bart_base_model

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: 0.7789
• Rouge1: 0.4137
• Rouge2: 0.3037
• Rougel: 0.3749
• Rougelsum: 0.3747
• Gen Len: 19.9959

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: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 4
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
0.9855	1.0	1764	0.8474	0.4122	0.303	0.3726	0.3726	19.9908
0.8409	2.0	3528	0.7938	0.4138	0.3044	0.3752	0.3751	19.9946
0.7872	3.0	5292	0.7776	0.4174	0.308	0.3783	0.3782	19.9928
0.7485	4.0	7056	0.7789	0.4137	0.3037	0.3749	0.3747	19.9959

Framework versions

• Transformers 4.39.3
• Pytorch 2.1.2
• Datasets 2.18.0
• Tokenizers 0.15.2