predict-perception-xlmr-cause-object

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: 0.3069
• Rmse: 0.8927
• Rmse Cause::a Causata da un oggetto (es. una pistola): 0.8927
• Mae: 0.5854
• Mae Cause::a Causata da un oggetto (es. una pistola): 0.5854
• R2: 0.5410
• R2 Cause::a Causata da un oggetto (es. una pistola): 0.5410
• Cos: 0.4783
• Pair: 0.0
• Rank: 0.5
• Neighbors: 0.6177
• 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 Cause::a Causata da un oggetto (es. una pistola)	Mae	Mae Cause::a Causata da un oggetto (es. una pistola)	R2	R2 Cause::a Causata da un oggetto (es. una pistola)	Cos	Pair	Rank	Neighbors	Rsa
1.0329	1.0	15	0.8168	1.4564	1.4564	1.2947	1.2947	-0.2216	-0.2216	-0.5652	0.0	0.5	0.5993	nan
1.0096	2.0	30	0.7432	1.3893	1.3893	1.1883	1.1883	-0.1116	-0.1116	-0.3913	0.0	0.5	0.6499	nan
0.9323	3.0	45	0.6879	1.3366	1.3366	1.1054	1.1054	-0.0289	-0.0289	-0.1304	0.0	0.5	0.5471	nan
0.8636	4.0	60	0.6378	1.2870	1.2870	1.0477	1.0477	0.0461	0.0461	0.2174	0.0	0.5	0.3007	nan
0.8041	5.0	75	0.5494	1.1945	1.1945	0.9499	0.9499	0.1783	0.1783	0.6522	0.0	0.5	0.6695	nan
0.7413	6.0	90	0.5526	1.1980	1.1980	0.9503	0.9503	0.1735	0.1735	0.5652	0.0	0.5	0.3898	nan
0.6397	7.0	105	0.4726	1.1078	1.1078	0.7826	0.7826	0.2932	0.2932	0.5652	0.0	0.5	0.3257	nan
0.5556	8.0	120	0.7728	1.4167	1.4167	1.1528	1.1528	-0.1558	-0.1558	0.1304	0.0	0.5	0.4027	nan
0.4972	9.0	135	0.4375	1.0659	1.0659	0.7577	0.7577	0.3457	0.3457	0.5652	0.0	0.5	0.5683	nan
0.3691	10.0	150	0.4990	1.1383	1.1383	0.8272	0.8272	0.2537	0.2537	0.4783	0.0	0.5	0.4781	nan
0.3381	11.0	165	0.4401	1.0690	1.0690	0.7319	0.7319	0.3418	0.3418	0.5652	0.0	0.5	0.5683	nan
0.2966	12.0	180	0.4794	1.1158	1.1158	0.7835	0.7835	0.2830	0.2830	0.5652	0.0	0.5	0.5683	nan
0.2324	13.0	195	0.4013	1.0208	1.0208	0.6873	0.6873	0.3998	0.3998	0.4783	0.0	0.5	0.5796	nan
0.1848	14.0	210	0.4305	1.0574	1.0574	0.7372	0.7372	0.3561	0.3561	0.4783	0.0	0.5	0.5796	nan
0.1621	15.0	225	0.3652	0.9738	0.9738	0.6164	0.6164	0.4538	0.4538	0.4783	0.0	0.5	0.6177	nan
0.1762	16.0	240	0.3335	0.9307	0.9307	0.6458	0.6458	0.5012	0.5012	0.4783	0.0	0.5	0.5796	nan
0.1404	17.0	255	0.3420	0.9424	0.9424	0.6599	0.6599	0.4886	0.4886	0.3913	0.0	0.5	0.5831	nan
0.1379	18.0	270	0.2853	0.8608	0.8608	0.6063	0.6063	0.5733	0.5733	0.3913	0.0	0.5	0.5831	nan
0.1322	19.0	285	0.3261	0.9203	0.9203	0.6548	0.6548	0.5123	0.5123	0.4783	0.0	0.5	0.5796	nan
0.1067	20.0	300	0.3328	0.9296	0.9296	0.5535	0.5535	0.5023	0.5023	0.6522	0.0	0.5	0.6695	nan
0.1038	21.0	315	0.3066	0.8924	0.8924	0.6266	0.6266	0.5414	0.5414	0.4783	0.0	0.5	0.5796	nan
0.094	22.0	330	0.2924	0.8714	0.8714	0.5792	0.5792	0.5626	0.5626	0.4783	0.0	0.5	0.6177	nan
0.1078	23.0	345	0.3161	0.9060	0.9060	0.6022	0.6022	0.5272	0.5272	0.3913	0.0	0.5	0.5831	nan
0.0976	24.0	360	0.3118	0.8998	0.8998	0.6011	0.6011	0.5337	0.5337	0.3913	0.0	0.5	0.5831	nan
0.0911	25.0	375	0.3123	0.9005	0.9005	0.5811	0.5811	0.5330	0.5330	0.4783	0.0	0.5	0.6177	nan
0.1039	26.0	390	0.3122	0.9005	0.9005	0.5956	0.5956	0.5330	0.5330	0.4783	0.0	0.5	0.6177	nan
0.0775	27.0	405	0.3191	0.9103	0.9103	0.6124	0.6124	0.5228	0.5228	0.3913	0.0	0.5	0.5831	nan
0.0789	28.0	420	0.3135	0.9023	0.9023	0.5825	0.5825	0.5311	0.5311	0.4783	0.0	0.5	0.6177	nan
0.0778	29.0	435	0.3075	0.8936	0.8936	0.5837	0.5837	0.5401	0.5401	0.4783	0.0	0.5	0.6177	nan
0.082	30.0	450	0.3069	0.8927	0.8927	0.5854	0.5854	0.5410	0.5410	0.4783	0.0	0.5	0.6177	nan

Framework versions

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