Modelpublic

antonkurylo/t5-base-news_headlines_8

sourceHugging Faceapache-2.0updated 3y agoView on Hugging Face

0likes14downloads

Model Card

t5-base-news_headlines

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

—Loss: 0.9299
—Rouge1: 55.5197
—Rouge2: 37.289
—Rougel: 52.5423
—Rougelsum: 52.5447

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: 5.6e-05
—trainbatchsize: 8
—evalbatchsize: 8
—seed: 42
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—num_epochs: 8

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
1.9936	1.0	1531	1.5027	44.1546	22.1175	40.3432	40.2843
1.6014	2.0	3062	1.2953	46.5025	25.0728	43.0053	43.0156
1.4043	3.0	4593	1.2426	48.2614	27.7433	45.0913	45.0575
1.2605	4.0	6124	1.1145	50.9833	31.0849	47.7984	47.8105
1.1521	5.0	7655	1.0423	53.1185	33.8153	49.9944	50.0043
1.0663	6.0	9186	0.9820	54.3416	35.6212	51.3341	51.3527
1.0089	7.0	10717	0.9346	55.4265	37.1236	52.5535	52.5711
0.9703	8.0	12248	0.9299	55.5197	37.289	52.5423	52.5447

Framework versions

—Transformers 4.28.0
—Pytorch 2.0.1+cu118
—Datasets 2.12.0
—Tokenizers 0.13.3