bart-large-cnn-samsum-icsi-ami-v2

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

• Loss: 5.8987
• Rouge1: 39.0928
• Rouge2: 10.8408
• Rougel: 21.9138
• Rougelsum: 35.5067
• Gen Len: 138.7941

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
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
No log	1.0	135	3.2308	38.4274	13.6461	22.366	35.2353	185.7941
No log	2.0	270	3.3026	40.1748	11.7944	23.2655	36.4718	146.8529
No log	3.0	405	3.5199	39.8209	12.1621	22.7772	36.4967	141.7647
2.2131	4.0	540	4.0508	40.4325	11.6547	22.9958	36.8782	131.4412
2.2131	5.0	675	4.6988	38.4097	9.8309	20.3894	34.1967	145.9706
2.2131	6.0	810	4.9590	38.5758	9.6335	20.865	35.0321	169.2353
2.2131	7.0	945	5.4264	38.2813	9.5764	21.1406	34.5989	148.0294
0.401	8.0	1080	5.4887	38.3014	9.6881	21.2398	34.1584	139.3529
0.401	9.0	1215	5.8044	39.9603	10.4329	22.6895	36.2406	145.2353
0.401	10.0	1350	5.8987	39.0928	10.8408	21.9138	35.5067	138.7941

Framework versions

• Transformers 4.26.1
• Pytorch 1.13.1+cu117
• Datasets 2.10.1
• Tokenizers 0.13.2