Modelpublic

Svetlana0303/Regression_longformer

sourceHugging Faceupdated 4y agoView on Hugging Face

0likes17downloads

Model Card

Regression_longformer

This model is a fine-tuned version of allenai/longformer-base-4096 on an unknown dataset. It achieves the following results on the evaluation set:

—Loss: 0.0973
—Mse: 0.0973
—Mae: 0.2464
—R2: 0.9498
—Accuracy: 0.8571

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
—trainbatchsize: 8
—evalbatchsize: 8
—seed: 42
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Mse	Mae	R2	Accuracy
No log	1.0	4	3.1150	3.1150	1.4245	-1.9228	0.2857
No log	2.0	8	2.4934	2.4934	1.2570	-1.3396	0.4286
No log	3.0	12	1.5812	1.5812	1.1308	-0.4836	0.1429
No log	4.0	16	1.4004	1.4004	1.0230	-0.3140	0.4286
No log	5.0	20	2.4083	2.4083	1.2028	-1.2597	0.4286
No log	6.0	24	2.4795	2.4795	1.2186	-1.3265	0.2857
No log	7.0	28	1.9781	1.9781	1.0441	-0.8560	0.5714
No log	8.0	32	1.6604	1.6604	0.9238	-0.5579	0.5714
No log	9.0	36	1.5105	1.5105	0.9193	-0.4172	0.5714
No log	10.0	40	0.9750	0.9750	0.7766	0.0852	0.5714
No log	11.0	44	0.6147	0.6147	0.6262	0.4233	0.2857
No log	12.0	48	0.5429	0.5429	0.5941	0.4906	0.4286
No log	13.0	52	0.4183	0.4183	0.5999	0.6075	0.4286
No log	14.0	56	0.5570	0.5570	0.6746	0.4774	0.2857
No log	15.0	60	0.2450	0.2450	0.4159	0.7701	0.7143
No log	16.0	64	0.1650	0.1650	0.3241	0.8452	0.7143
No log	17.0	68	0.1367	0.1367	0.2701	0.8718	0.8571
No log	18.0	72	0.1225	0.1225	0.2905	0.8850	0.8571
No log	19.0	76	0.1169	0.1169	0.2825	0.8903	0.8571
No log	20.0	80	0.1309	0.1309	0.3046	0.8772	0.8571
No log	21.0	84	0.4585	0.4585	0.5426	0.5698	0.5714
No log	22.0	88	0.2451	0.2451	0.4216	0.7700	0.5714
No log	23.0	92	0.1123	0.1123	0.2678	0.8946	0.8571
No log	24.0	96	0.0942	0.0942	0.2816	0.9116	1.0
No log	25.0	100	0.1239	0.1239	0.2445	0.8838	0.8571
No log	26.0	104	0.2268	0.2268	0.3540	0.7872	0.8571
No log	27.0	108	0.3575	0.3575	0.4408	0.6645	0.5714
No log	28.0	112	0.3418	0.3418	0.4349	0.6793	0.5714
No log	29.0	116	0.2677	0.2677	0.3958	0.7488	0.5714
No log	30.0	120	0.2235	0.2235	0.3717	0.7903	0.5714

Framework versions

—Transformers 4.26.1
—Pytorch 1.13.1+cu116
—Datasets 2.9.0
—Tokenizers 0.13.2