predict-perception-xlmr-focus-victim

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: 0.2546
• Rmse: 0.6301
• Rmse Focus::a Sulla vittima: 0.6301
• Mae: 0.5441
• Mae Focus::a Sulla vittima: 0.5441
• R2: 0.7205
• R2 Focus::a Sulla vittima: 0.7205
• Cos: 0.8261
• Pair: 0.0
• Rank: 0.5
• Neighbors: 0.7802
• 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 Sulla vittima	Mae	Mae Focus::a Sulla vittima	R2	R2 Focus::a Sulla vittima	Cos	Pair	Rank	Neighbors	Rsa
1.0607	1.0	15	0.9261	1.2017	1.2017	0.9557	0.9557	-0.0166	-0.0166	0.4783	0.0	0.5	0.6332	nan
1.0107	2.0	30	0.9481	1.2159	1.2159	0.9861	0.9861	-0.0408	-0.0408	0.4783	0.0	0.5	0.6332	nan
0.9921	3.0	45	0.9068	1.1892	1.1892	0.9548	0.9548	0.0045	0.0045	0.4783	0.0	0.5	0.6332	nan
0.7769	4.0	60	0.5014	0.8842	0.8842	0.7121	0.7121	0.4496	0.4496	0.7391	0.0	0.5	0.6232	nan
0.5763	5.0	75	0.4019	0.7917	0.7917	0.6737	0.6737	0.5588	0.5588	0.8261	0.0	0.5	0.8155	nan
0.4378	6.0	90	0.3594	0.7486	0.7486	0.5957	0.5957	0.6055	0.6055	0.7391	0.0	0.5	0.4442	nan
0.3595	7.0	105	0.3452	0.7337	0.7337	0.6333	0.6333	0.6210	0.6210	0.5652	0.0	0.5	0.2649	nan
0.3192	8.0	120	0.3275	0.7147	0.7147	0.6205	0.6205	0.6405	0.6405	0.7391	0.0	0.5	0.6561	nan
0.2482	9.0	135	0.2978	0.6815	0.6815	0.5754	0.5754	0.6731	0.6731	0.7391	0.0	0.5	0.6715	nan
0.2416	10.0	150	0.3018	0.6860	0.6860	0.5954	0.5954	0.6687	0.6687	0.5652	0.0	0.5	0.2553	nan
0.2292	11.0	165	0.2764	0.6565	0.6565	0.5522	0.5522	0.6966	0.6966	0.9130	0.0	0.5	0.8408	nan
0.1752	12.0	180	0.3070	0.6920	0.6920	0.5680	0.5680	0.6629	0.6629	0.7391	0.0	0.5	0.6715	nan
0.1956	13.0	195	0.2923	0.6752	0.6752	0.5499	0.5499	0.6791	0.6791	0.8261	0.0	0.5	0.7843	nan
0.1424	14.0	210	0.3163	0.7023	0.7023	0.6060	0.6060	0.6528	0.6528	0.9130	0.0	0.5	0.8408	nan
0.152	15.0	225	0.2436	0.6164	0.6164	0.5127	0.5127	0.7326	0.7326	0.9130	0.0	0.5	0.8408	nan
0.1277	16.0	240	0.2471	0.6208	0.6208	0.5367	0.5367	0.7287	0.7287	0.8261	0.0	0.5	0.7802	nan
0.1269	17.0	255	0.2573	0.6334	0.6334	0.5329	0.5329	0.7175	0.7175	0.8261	0.0	0.5	0.7802	nan
0.1058	18.0	270	0.2538	0.6291	0.6291	0.5530	0.5530	0.7214	0.7214	0.7391	0.0	0.5	0.2347	nan
0.107	19.0	285	0.2568	0.6328	0.6328	0.5464	0.5464	0.7181	0.7181	0.8261	0.0	0.5	0.7802	nan
0.1185	20.0	300	0.2452	0.6183	0.6183	0.5317	0.5317	0.7309	0.7309	0.7391	0.0	0.5	0.2347	nan
0.1029	21.0	315	0.2419	0.6142	0.6142	0.5415	0.5415	0.7344	0.7344	0.7391	0.0	0.5	0.2347	nan
0.0908	22.0	330	0.2462	0.6196	0.6196	0.5261	0.5261	0.7297	0.7297	0.8261	0.0	0.5	0.7802	nan
0.0901	23.0	345	0.2528	0.6279	0.6279	0.5330	0.5330	0.7225	0.7225	0.8261	0.0	0.5	0.7802	nan
0.0979	24.0	360	0.2800	0.6607	0.6607	0.5682	0.5682	0.6927	0.6927	0.9130	0.0	0.5	0.8408	nan
0.0992	25.0	375	0.2502	0.6246	0.6246	0.5517	0.5517	0.7254	0.7254	0.6522	0.0	0.5	0.2372	nan
0.0846	26.0	390	0.2570	0.6331	0.6331	0.5524	0.5524	0.7178	0.7178	0.8261	0.0	0.5	0.7802	nan
0.0717	27.0	405	0.2562	0.6321	0.6321	0.5456	0.5456	0.7187	0.7187	0.8261	0.0	0.5	0.7802	nan
0.0739	28.0	420	0.2570	0.6330	0.6330	0.5471	0.5471	0.7179	0.7179	0.8261	0.0	0.5	0.7802	nan
0.0828	29.0	435	0.2553	0.6309	0.6309	0.5446	0.5446	0.7198	0.7198	0.8261	0.0	0.5	0.7802	nan
0.086	30.0	450	0.2546	0.6301	0.6301	0.5441	0.5441	0.7205	0.7205	0.8261	0.0	0.5	0.7802	nan

Framework versions

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