bart-abs-1509-0313-lr-0.0003-bs-2-maxep-10

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: 7.6999
• Rouge/rouge1: 0.2439
• Rouge/rouge2: 0.0504
• Rouge/rougel: 0.2065
• Rouge/rougelsum: 0.2067
• Bertscore/bertscore-precision: 0.8544
• Bertscore/bertscore-recall: 0.8581
• Bertscore/bertscore-f1: 0.8562
• Meteor: 0.229
• Gen Len: 44.0

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: 0.0003
• train_batch_size: 2
• eval_batch_size: 2
• 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	Rouge/rouge1	Rouge/rouge2	Rouge/rougel	Rouge/rougelsum	Bertscore/bertscore-precision	Bertscore/bertscore-recall	Bertscore/bertscore-f1	Meteor	Gen Len
2.1693	1.0	434	4.5206	0.2693	0.0703	0.2315	0.232	0.8864	0.8585	0.8722	0.2188	29.0
1.3403	2.0	868	5.0395	0.3061	0.0778	0.251	0.2513	0.8875	0.864	0.8755	0.239	32.0
1.1783	3.0	1302	5.1339	0.2426	0.0523	0.1835	0.1835	0.8501	0.8566	0.8533	0.248	52.0
0.8203	4.0	1736	5.6678	0.3347	0.0996	0.2675	0.2678	0.8793	0.8662	0.8727	0.2663	27.0
0.623	5.0	2170	6.1732	0.2961	0.0668	0.2313	0.2314	0.8628	0.8608	0.8617	0.2421	52.0
0.5051	6.0	2604	6.1011	0.2953	0.0542	0.2213	0.2211	0.8685	0.8588	0.8636	0.2403	34.0
0.4004	7.0	3038	6.8848	0.2613	0.0803	0.214	0.2142	0.8711	0.8469	0.8588	0.2102	26.0
0.3371	8.0	3472	7.2987	0.2132	0.0353	0.1717	0.1717	0.8605	0.8522	0.8563	0.1839	27.0
0.2954	9.0	3906	7.4692	0.244	0.063	0.1986	0.1986	0.8544	0.8608	0.8575	0.211	52.0
0.2576	10.0	4340	7.6999	0.2439	0.0504	0.2065	0.2067	0.8544	0.8581	0.8562	0.229	44.0

Framework versions

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