SocialScienceBARTPrincipal

This model is a fine-tuned version of facebook/bart-large-cnn on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 4.8587
• Rouge1: 48.4993
• Rouge2: 14.8435
• Rougel: 33.0264
• Rougelsum: 44.9256
• Bertscore Precision: 80.3517
• Bertscore Recall: 82.7128
• Bertscore F1: 81.5112
• Bleu: 0.1092
• Gen Len: 195.1640

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: 5e-05
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 16
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Bertscore Precision	Bertscore Recall	Bertscore F1	Bleu	Gen Len
6.5089	0.1314	100	6.2390	39.4898	11.0769	27.6002	36.497	75.7798	80.6901	78.1466	0.0800	195.1640
5.9338	0.2628	200	5.7540	41.6352	11.9524	29.0458	38.5778	77.0272	81.1993	79.0507	0.0882	195.1640
5.6077	0.3943	300	5.4443	41.5238	12.2762	29.4389	38.8683	77.5496	81.3713	79.4075	0.0894	195.1640
5.3997	0.5257	400	5.2541	44.1846	13.1247	30.5659	41.1211	78.8697	81.8978	80.3498	0.0962	195.1640
5.1614	0.6571	500	5.1269	44.5045	13.3887	31.1505	41.1205	78.727	82.0655	80.3557	0.0994	195.1640
5.0558	0.7885	600	4.9610	46.7823	14.4367	32.4159	43.2551	79.6807	82.5047	81.0632	0.1059	195.1640
4.9749	0.9199	700	4.8587	48.4993	14.8435	33.0264	44.9256	80.3517	82.7128	81.5112	0.1092	195.1640

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1