bart-large-cnn-finetuned-roundup-2-8

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: 1.1855
• Rouge1: 53.552
• Rouge2: 34.9077
• Rougel: 38.0158
• Rougelsum: 50.7179
• Gen Len: 142.0

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: 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: 8
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	167	1.2085	50.8706	32.0069	32.9241	47.9805	142.0
No log	2.0	334	1.0897	53.2218	34.1317	34.4827	50.4795	139.4286
1.0256	3.0	501	1.0535	50.8882	30.2514	31.5051	47.9856	141.9048
1.0256	4.0	668	1.0515	54.9414	35.2309	36.006	52.0331	142.0
1.0256	5.0	835	1.0829	53.0709	33.4587	36.4223	50.1627	140.7619
0.4579	6.0	1002	1.1310	51.5274	30.7069	32.4146	48.8851	142.0
0.4579	7.0	1169	1.1670	52.1536	31.7158	35.7483	49.2678	142.0
0.4579	8.0	1336	1.1855	53.552	34.9077	38.0158	50.7179	142.0

Framework versions

• Transformers 4.18.0
• Pytorch 1.11.0+cu113
• Datasets 2.1.0
• Tokenizers 0.12.1