mymodel_v2_4

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: 2.1383
• Rouge1: 0.5107
• Rouge2: 0.1818
• Rougel: 0.4557
• Rougelsum: 0.4753
• Gen Len: 19.4327

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: 0.0001
• train_batch_size: 10
• eval_batch_size: 10
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	111	1.6651	1.0836	0.9742	1.076	1.0681	19.4
No log	2.0	222	1.6632	0.5545	0.3924	0.5312	0.5302	19.5855
No log	3.0	333	1.7607	0.7463	0.5905	0.7663	0.7512	19.6982
No log	4.0	444	1.8583	0.8352	0.7153	0.8546	0.8534	19.7018
1.4574	5.0	555	1.9357	0.659	0.6196	0.6745	0.6962	19.3273
1.4574	6.0	666	2.0241	0.4785	0.4545	0.4878	0.4997	19.6036
1.4574	7.0	777	2.0663	0.2327	0.1818	0.2741	0.2741	19.2327
1.4574	8.0	888	2.0969	0.3755	0.2916	0.3915	0.3956	19.4545
1.4574	9.0	999	2.1291	0.7743	0.5592	0.7473	0.7881	19.3964
0.3529	10.0	1110	2.1383	0.5107	0.1818	0.4557	0.4753	19.4327

Framework versions

• Transformers 4.28.1
• Pytorch 2.0.0
• Datasets 2.1.0
• Tokenizers 0.13.3