test_sum_abs_bart-base_wasa_coref_stops

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: 0.2847
• Rouge1: 0.3924
• Rouge2: 0.2979
• Rougel: 0.3606
• Rougelsum: 0.3603
• Gen Len: 20.0

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
0.3752	1.0	1632	0.3126	0.3887	0.2965	0.3586	0.3582	19.9997
0.3192	2.0	3264	0.2995	0.3852	0.2901	0.3536	0.3532	20.0
0.2879	3.0	4896	0.2863	0.3933	0.2989	0.3621	0.362	19.9997
0.2625	4.0	6528	0.2847	0.3924	0.2979	0.3606	0.3603	20.0

Framework versions

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