nvl-ca

This model is a fine-tuned version of t5-small on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.6425
• Rouge1: 36.2683
• Rouge2: 17.3571
• Rougel: 31.414
• Rougelsum: 33.3573
• Gen Len: 18.1

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 12

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
2.7351	1.0	50	2.0532	29.1549	10.8095	24.5213	27.1706	18.54
2.2954	2.0	100	1.8884	34.1103	15.1143	28.6964	30.6995	18.5
2.1461	3.0	150	1.7999	33.7268	15.3397	29.1248	30.7545	18.48
2.0402	4.0	200	1.7510	35.2811	16.3829	29.5922	31.3828	18.64
1.9727	5.0	250	1.7251	35.9939	17.0171	30.9116	32.514	18.3
1.9185	6.0	300	1.6982	36.1673	17.3892	31.4179	33.2171	18.06
1.8791	7.0	350	1.6809	36.0791	17.9475	31.6153	33.2867	18.2
1.8443	8.0	400	1.6631	36.3616	17.7432	31.9719	33.651	17.96
1.8322	9.0	450	1.6533	35.9061	16.9737	31.1291	33.1402	17.96
1.7978	10.0	500	1.6482	35.8366	17.0094	31.3893	33.3356	17.88
1.8037	11.0	550	1.6440	36.2683	17.3571	31.414	33.3573	18.1
1.7937	12.0	600	1.6425	36.2683	17.3571	31.414	33.3573	18.1

Framework versions

• Transformers 4.36.1
• Pytorch 2.1.0+cu121
• Datasets 2.15.0
• Tokenizers 0.15.0