bart-abs-1509-0313-lr-3e-05-bs-8-maxep-6

This model is a fine-tuned version of sshleifer/distilbart-xsum-12-6 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 3.3957
• Rouge/rouge1: 0.4646
• Rouge/rouge2: 0.2089
• Rouge/rougel: 0.3939
• Rouge/rougelsum: 0.3945
• Bertscore/bertscore-precision: 0.8956
• Bertscore/bertscore-recall: 0.8935
• Bertscore/bertscore-f1: 0.8944
• Meteor: 0.4132
• Gen Len: 37.5

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: 3e-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: 6
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge/rouge1	Rouge/rouge2	Rouge/rougel	Rouge/rougelsum	Bertscore/bertscore-precision	Bertscore/bertscore-recall	Bertscore/bertscore-f1	Meteor	Gen Len
0.6107	1.0	109	2.4784	0.449	0.1974	0.3774	0.3776	0.8943	0.8904	0.8922	0.3981	36.9182
0.3993	2.0	218	2.7984	0.4656	0.2145	0.3954	0.3965	0.8975	0.8914	0.8943	0.408	35.1364
0.2779	3.0	327	3.0563	0.4669	0.2112	0.3981	0.3995	0.8961	0.8905	0.8931	0.4088	36.0545
0.2038	4.0	436	3.2410	0.4639	0.2052	0.3895	0.3904	0.896	0.8949	0.8953	0.4109	37.9
0.1606	5.0	545	3.3263	0.4582	0.2063	0.391	0.392	0.8961	0.893	0.8944	0.4033	36.5545
0.1282	6.0	654	3.3957	0.4646	0.2089	0.3939	0.3945	0.8956	0.8935	0.8944	0.4132	37.5

Framework versions

• Transformers 4.44.0
• Pytorch 2.4.0
• Datasets 2.21.0
• Tokenizers 0.19.1