predict-perception-xlmr-focus-concept

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

• Loss: 0.8296
• Rmse: 1.0302
• Rmse Focus::a Su un concetto astratto o un'emozione: 1.0302
• Mae: 0.7515
• Mae Focus::a Su un concetto astratto o un'emozione: 0.7515
• R2: 0.1804
• R2 Focus::a Su un concetto astratto o un'emozione: 0.1804
• Cos: 0.4783
• Pair: 0.0
• Rank: 0.5
• Neighbors: 0.3415
• Rsa: nan

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: 1e-05
• train_batch_size: 20
• eval_batch_size: 8
• seed: 1996
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Rmse	Rmse Focus::a Su un concetto astratto o un'emozione	Mae	Mae Focus::a Su un concetto astratto o un'emozione	R2	R2 Focus::a Su un concetto astratto o un'emozione	Cos	Pair	Rank	Neighbors	Rsa
1.0355	1.0	15	0.9822	1.1209	1.1209	0.9649	0.9649	0.0296	0.0296	0.2174	0.0	0.5	0.3706	nan
1.0083	2.0	30	1.1378	1.2065	1.2065	0.9954	0.9954	-0.1241	-0.1241	0.2174	0.0	0.5	0.3309	nan
0.9823	3.0	45	0.9669	1.1121	1.1121	0.9315	0.9315	0.0448	0.0448	0.3043	0.0	0.5	0.3810	nan
0.9468	4.0	60	0.8856	1.0644	1.0644	0.8584	0.8584	0.1251	0.1251	0.3913	0.0	0.5	0.3803	nan
0.9294	5.0	75	0.8136	1.0202	1.0202	0.8396	0.8396	0.1963	0.1963	0.6522	0.0	0.5	0.4727	nan
0.881	6.0	90	0.7634	0.9882	0.9882	0.8192	0.8192	0.2458	0.2458	0.6522	0.0	0.5	0.4727	nan
0.7589	7.0	105	0.8139	1.0204	1.0204	0.8136	0.8136	0.1960	0.1960	0.5652	0.0	0.5	0.4120	nan
0.7217	8.0	120	0.9105	1.0792	1.0792	0.9394	0.9394	0.1005	0.1005	0.3913	0.0	0.5	0.4108	nan
0.8059	9.0	135	1.0322	1.1491	1.1491	0.9115	0.9115	-0.0197	-0.0197	0.5652	0.0	0.5	0.3738	nan
0.6483	10.0	150	0.7989	1.0109	1.0109	0.7899	0.7899	0.2108	0.2108	0.6522	0.0	0.5	0.4727	nan
0.5725	11.0	165	0.7175	0.9581	0.9581	0.7011	0.7011	0.2912	0.2912	0.5652	0.0	0.5	0.3738	nan
0.5091	12.0	180	0.8818	1.0621	1.0621	0.8775	0.8775	0.1289	0.1289	0.5652	0.0	0.5	0.4063	nan
0.4526	13.0	195	0.8451	1.0398	1.0398	0.7990	0.7990	0.1651	0.1651	0.5652	0.0	0.5	0.4063	nan
0.361	14.0	210	0.8632	1.0508	1.0508	0.8124	0.8124	0.1472	0.1472	0.4783	0.0	0.5	0.3699	nan
0.3582	15.0	225	0.8461	1.0404	1.0404	0.7923	0.7923	0.1641	0.1641	0.3913	0.0	0.5	0.3672	nan
0.2945	16.0	240	0.9142	1.0814	1.0814	0.8125	0.8125	0.0968	0.0968	0.3913	0.0	0.5	0.3672	nan
0.2891	17.0	255	0.8377	1.0352	1.0352	0.7718	0.7718	0.1724	0.1724	0.4783	0.0	0.5	0.3415	nan
0.2569	18.0	270	0.8106	1.0183	1.0183	0.7481	0.7481	0.1992	0.1992	0.4783	0.0	0.5	0.3415	nan
0.2583	19.0	285	0.8239	1.0266	1.0266	0.7597	0.7597	0.1861	0.1861	0.4783	0.0	0.5	0.3415	nan
0.2217	20.0	300	0.8485	1.0419	1.0419	0.7663	0.7663	0.1617	0.1617	0.4783	0.0	0.5	0.3415	nan
0.1927	21.0	315	0.8304	1.0307	1.0307	0.7536	0.7536	0.1797	0.1797	0.4783	0.0	0.5	0.3415	nan
0.176	22.0	330	0.8321	1.0317	1.0317	0.7539	0.7539	0.1780	0.1780	0.4783	0.0	0.5	0.3415	nan
0.1639	23.0	345	0.7914	1.0062	1.0062	0.7460	0.7460	0.2182	0.2182	0.4783	0.0	0.5	0.3415	nan
0.177	24.0	360	0.8619	1.0500	1.0500	0.7725	0.7725	0.1486	0.1486	0.4783	0.0	0.5	0.3415	nan
0.1473	25.0	375	0.8101	1.0180	1.0180	0.7587	0.7587	0.1997	0.1997	0.4783	0.0	0.5	0.3415	nan
0.181	26.0	390	0.8038	1.0141	1.0141	0.7433	0.7433	0.2059	0.2059	0.4783	0.0	0.5	0.3415	nan
0.1679	27.0	405	0.7982	1.0105	1.0105	0.7248	0.7248	0.2115	0.2115	0.4783	0.0	0.5	0.3415	nan
0.1529	28.0	420	0.8282	1.0293	1.0293	0.7454	0.7454	0.1818	0.1818	0.4783	0.0	0.5	0.3415	nan
0.1822	29.0	435	0.8310	1.0311	1.0311	0.7512	0.7512	0.1790	0.1790	0.4783	0.0	0.5	0.3415	nan
0.1442	30.0	450	0.8296	1.0302	1.0302	0.7515	0.7515	0.1804	0.1804	0.4783	0.0	0.5	0.3415	nan

Framework versions

• Transformers 4.16.2
• Pytorch 1.10.2+cu113
• Datasets 1.18.3
• Tokenizers 0.11.0