predict-perception-xlmr-blame-none

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: 0.8941
• Rmse: 1.1259
• Rmse Blame::a Nessuno: 1.1259
• Mae: 0.8559
• Mae Blame::a Nessuno: 0.8559
• R2: 0.2847
• R2 Blame::a Nessuno: 0.2847
• Cos: 0.3043
• Pair: 0.0
• Rank: 0.5
• Neighbors: 0.3537
• 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 Blame::a Nessuno	Mae	Mae Blame::a Nessuno	R2	R2 Blame::a Nessuno	Cos	Pair	Rank	Neighbors	Rsa
1.042	1.0	15	1.2746	1.3443	1.3443	1.1788	1.1788	-0.0197	-0.0197	0.0435	0.0	0.5	0.2970	nan
0.9994	2.0	30	1.3264	1.3714	1.3714	1.1967	1.1967	-0.0612	-0.0612	-0.0435	0.0	0.5	0.2961	nan
0.9123	3.0	45	1.2511	1.3319	1.3319	1.0932	1.0932	-0.0009	-0.0009	0.1304	0.0	0.5	0.2681	nan
0.741	4.0	60	1.0204	1.2028	1.2028	0.9818	0.9818	0.1836	0.1836	0.3043	0.0	0.5	0.3686	nan
0.6337	5.0	75	0.8607	1.1047	1.1047	0.8145	0.8145	0.3115	0.3115	0.3913	0.0	0.5	0.4044	nan
0.4974	6.0	90	0.8574	1.1026	1.1026	0.8095	0.8095	0.3140	0.3140	0.3913	0.0	0.5	0.4044	nan
0.4929	7.0	105	0.8548	1.1009	1.1009	0.8560	0.8560	0.3161	0.3161	0.3043	0.0	0.5	0.3686	nan
0.4378	8.0	120	0.6974	0.9944	0.9944	0.7503	0.7503	0.4421	0.4421	0.3043	0.0	0.5	0.3686	nan
0.3999	9.0	135	0.7955	1.0620	1.0620	0.7907	0.7907	0.3636	0.3636	0.3913	0.0	0.5	0.4044	nan
0.3715	10.0	150	0.8954	1.1267	1.1267	0.8036	0.8036	0.2837	0.2837	0.4783	0.0	0.5	0.4058	nan
0.3551	11.0	165	0.8449	1.0945	1.0945	0.8748	0.8748	0.3241	0.3241	0.3913	0.0	0.5	0.3931	nan
0.3428	12.0	180	0.7960	1.0624	1.0624	0.8000	0.8000	0.3632	0.3632	0.3913	0.0	0.5	0.4044	nan
0.2923	13.0	195	0.9027	1.1313	1.1313	0.8441	0.8441	0.2778	0.2778	0.3043	0.0	0.5	0.3537	nan
0.2236	14.0	210	0.8914	1.1242	1.1242	0.8998	0.8998	0.2869	0.2869	0.2174	0.0	0.5	0.3324	nan
0.2553	15.0	225	0.9184	1.1411	1.1411	0.8633	0.8633	0.2652	0.2652	0.3043	0.0	0.5	0.3537	nan
0.2064	16.0	240	0.9284	1.1473	1.1473	0.8919	0.8919	0.2573	0.2573	0.3043	0.0	0.5	0.3537	nan
0.1972	17.0	255	0.9495	1.1602	1.1602	0.8768	0.8768	0.2404	0.2404	0.3043	0.0	0.5	0.3537	nan
0.1622	18.0	270	0.9850	1.1818	1.1818	0.9303	0.9303	0.2120	0.2120	0.2174	0.0	0.5	0.3324	nan
0.1685	19.0	285	0.9603	1.1669	1.1669	0.8679	0.8679	0.2317	0.2317	0.3043	0.0	0.5	0.3537	nan
0.1773	20.0	300	0.9269	1.1464	1.1464	0.8391	0.8391	0.2585	0.2585	0.3043	0.0	0.5	0.3537	nan
0.1716	21.0	315	0.8936	1.1256	1.1256	0.8357	0.8357	0.2851	0.2851	0.3043	0.0	0.5	0.3537	nan
0.161	22.0	330	0.8894	1.1230	1.1230	0.8593	0.8593	0.2884	0.2884	0.3043	0.0	0.5	0.3537	nan
0.1297	23.0	345	0.8997	1.1294	1.1294	0.8568	0.8568	0.2802	0.2802	0.3043	0.0	0.5	0.3537	nan
0.15	24.0	360	0.8748	1.1137	1.1137	0.8541	0.8541	0.3002	0.3002	0.2174	0.0	0.5	0.3324	nan
0.1149	25.0	375	0.9264	1.1461	1.1461	0.8682	0.8682	0.2588	0.2588	0.3913	0.0	0.5	0.3901	nan
0.1354	26.0	390	0.8829	1.1188	1.1188	0.8608	0.8608	0.2937	0.2937	0.2174	0.0	0.5	0.3324	nan
0.1321	27.0	405	0.9137	1.1382	1.1382	0.8656	0.8656	0.2691	0.2691	0.3043	0.0	0.5	0.3537	nan
0.1154	28.0	420	0.8774	1.1154	1.1154	0.8488	0.8488	0.2980	0.2980	0.2174	0.0	0.5	0.3324	nan
0.1112	29.0	435	0.8985	1.1287	1.1287	0.8562	0.8562	0.2812	0.2812	0.3043	0.0	0.5	0.3537	nan
0.1525	30.0	450	0.8941	1.1259	1.1259	0.8559	0.8559	0.2847	0.2847	0.3043	0.0	0.5	0.3537	nan

Framework versions

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