mobilebert_sa_GLUE_Experiment_logit_kd_data_aug_mrpc

This model is a fine-tuned version of google/mobilebert-uncased on the GLUE MRPC dataset. It achieves the following results on the evaluation set:

Loss: 0.1256
Accuracy: 0.9877
F1: 0.9911
Combined Score: 0.9894

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: 128
eval_batch_size: 128
seed: 10
distributed_type: multi-GPU
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Combined Score
0.2964	1.0	1959	0.2026	0.9608	0.9718	0.9663
0.2307	2.0	3918	0.1943	0.9706	0.9789	0.9748
0.2221	3.0	5877	0.1874	0.9804	0.9858	0.9831
0.2163	4.0	7836	0.1703	0.9853	0.9894	0.9873
0.2115	5.0	9795	0.1805	0.9853	0.9894	0.9873
0.2071	6.0	11754	0.1682	0.9804	0.9859	0.9831
0.2036	7.0	13713	0.1583	0.9877	0.9911	0.9894
0.2007	8.0	15672	0.1628	0.9926	0.9947	0.9936
0.1985	9.0	17631	0.1548	0.9853	0.9894	0.9873
0.1965	10.0	19590	0.1583	0.9877	0.9911	0.9894
0.195	11.0	21549	0.1527	0.9902	0.9928	0.9915
0.1938	12.0	23508	0.1512	0.9902	0.9929	0.9915
0.1926	13.0	25467	0.1426	0.9951	0.9964	0.9958
0.1917	14.0	27426	0.1436	0.9951	0.9964	0.9958
0.191	15.0	29385	0.1503	0.9926	0.9946	0.9936
0.1901	16.0	31344	0.1461	0.9951	0.9964	0.9958
0.1894	17.0	33303	0.1498	0.9975	0.9982	0.9979
0.1888	18.0	35262	0.1402	0.9902	0.9929	0.9915
0.1882	19.0	37221	0.1420	0.9926	0.9946	0.9936
0.1876	20.0	39180	0.1346	0.9902	0.9929	0.9915
0.1871	21.0	41139	0.1396	0.9951	0.9964	0.9958
0.1867	22.0	43098	0.1443	0.9951	0.9964	0.9958
0.1862	23.0	45057	0.1346	0.9926	0.9947	0.9936
0.1857	24.0	47016	0.1361	0.9951	0.9964	0.9958
0.1854	25.0	48975	0.1318	0.9926	0.9947	0.9936
0.185	26.0	50934	0.1310	0.9902	0.9929	0.9915
0.1846	27.0	52893	0.1302	0.9926	0.9947	0.9936
0.1842	28.0	54852	0.1329	0.9951	0.9964	0.9958
0.1839	29.0	56811	0.1300	0.9902	0.9929	0.9915
0.1836	30.0	58770	0.1328	0.9902	0.9929	0.9915
0.1832	31.0	60729	0.1327	0.9902	0.9929	0.9915
0.1829	32.0	62688	0.1308	0.9902	0.9929	0.9915
0.1826	33.0	64647	0.1287	0.9902	0.9929	0.9915
0.1824	34.0	66606	0.1309	0.9926	0.9947	0.9936
0.1821	35.0	68565	0.1309	0.9926	0.9947	0.9936
0.1818	36.0	70524	0.1271	0.9902	0.9929	0.9915
0.1816	37.0	72483	0.1278	0.9877	0.9911	0.9894
0.1813	38.0	74442	0.1280	0.9902	0.9929	0.9915
0.1811	39.0	76401	0.1289	0.9902	0.9929	0.9915
0.1809	40.0	78360	0.1290	0.9877	0.9911	0.9894
0.1807	41.0	80319	0.1256	0.9877	0.9911	0.9894
0.1805	42.0	82278	0.1268	0.9926	0.9947	0.9936
0.1803	43.0	84237	0.1274	0.9926	0.9947	0.9936
0.1801	44.0	86196	0.1277	0.9926	0.9947	0.9936
0.1799	45.0	88155	0.1264	0.9926	0.9947	0.9936
0.1797	46.0	90114	0.1274	0.9902	0.9929	0.9915

Framework versions

Transformers 4.26.0
Pytorch 1.14.0a0+410ce96
Datasets 2.9.0
Tokenizers 0.13.2

gokuls/mobilebert_sa_GLUE_Experiment_logit_kd_data_aug_mrpc

mobilebert_sa_GLUE_Experiment_logit_kd_data_aug_mrpc

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions