300-tiny_tobacco3482_kd_NKD_t1.0_g1.5

This model is a fine-tuned version of WinKawaks/vit-tiny-patch16-224 on the None dataset. It achieves the following results on the evaluation set:

Loss: 4.2183
Accuracy: 0.83
Brier Loss: 0.2913
Nll: 1.7770
F1 Micro: 0.83
F1 Macro: 0.8264
Ece: 0.1561
Aurc: 0.0404

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: 0.0001
train_batch_size: 64
eval_batch_size: 64
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Brier Loss	Nll	F1 Micro	F1 Macro	Ece	Aurc
No log	1.0	13	5.3145	0.225	0.8879	8.2660	0.225	0.1306	0.2937	0.7940
No log	2.0	26	4.7229	0.395	0.7730	4.0151	0.395	0.3145	0.3180	0.3989
No log	3.0	39	4.4853	0.52	0.6589	3.4597	0.52	0.3986	0.2918	0.2753
No log	4.0	52	4.3071	0.595	0.5739	2.6724	0.595	0.5030	0.3024	0.1820
No log	5.0	65	4.1849	0.67	0.5011	2.3415	0.67	0.6072	0.2760	0.1327
No log	6.0	78	4.2077	0.69	0.4390	2.6524	0.69	0.6081	0.2261	0.1082
No log	7.0	91	4.0758	0.705	0.4177	2.1183	0.705	0.6424	0.2219	0.1011
No log	8.0	104	4.1727	0.705	0.4036	2.0253	0.705	0.6324	0.2012	0.1048
No log	9.0	117	4.1369	0.72	0.4007	2.1063	0.72	0.6700	0.2155	0.1058
No log	10.0	130	4.2585	0.705	0.4100	2.2749	0.705	0.6445	0.2011	0.1081
No log	11.0	143	3.9977	0.765	0.3467	1.7895	0.765	0.6913	0.1865	0.0703
No log	12.0	156	4.1178	0.72	0.3606	1.5964	0.72	0.6535	0.1917	0.0879
No log	13.0	169	4.2015	0.69	0.4133	2.1808	0.69	0.6412	0.2060	0.1157
No log	14.0	182	3.9586	0.775	0.3258	1.6381	0.775	0.7529	0.1930	0.0637
No log	15.0	195	3.9068	0.775	0.3176	1.3371	0.775	0.7259	0.1640	0.0671
No log	16.0	208	3.9768	0.75	0.3493	1.6983	0.75	0.7460	0.1677	0.0827
No log	17.0	221	3.9648	0.77	0.3116	1.5883	0.7700	0.7352	0.1523	0.0632
No log	18.0	234	3.8706	0.8	0.2933	1.7099	0.8000	0.7800	0.1548	0.0568
No log	19.0	247	3.8302	0.82	0.2815	1.4136	0.82	0.8092	0.1606	0.0508
No log	20.0	260	3.9628	0.815	0.3061	1.4014	0.815	0.8032	0.1701	0.0645
No log	21.0	273	3.9288	0.81	0.3065	1.3082	0.81	0.8034	0.1744	0.0665
No log	22.0	286	3.8470	0.795	0.2934	1.5781	0.795	0.7626	0.1510	0.0557
No log	23.0	299	3.9932	0.785	0.3214	1.6262	0.785	0.7678	0.1626	0.0646
No log	24.0	312	3.9308	0.795	0.3165	1.7210	0.795	0.7907	0.1692	0.0661
No log	25.0	325	3.9077	0.82	0.2895	1.7608	0.82	0.8019	0.1556	0.0577
No log	26.0	338	3.8450	0.82	0.2816	1.6690	0.82	0.8035	0.1629	0.0557
No log	27.0	351	4.0009	0.815	0.3063	1.7843	0.815	0.8128	0.1705	0.0679
No log	28.0	364	3.8787	0.815	0.2839	1.6762	0.815	0.8071	0.1355	0.0535
No log	29.0	377	3.8246	0.825	0.2713	1.7012	0.825	0.8100	0.1428	0.0503
No log	30.0	390	3.8846	0.815	0.2758	1.7815	0.815	0.8025	0.1440	0.0490
No log	31.0	403	3.9277	0.81	0.2935	1.5897	0.81	0.8011	0.1523	0.0586
No log	32.0	416	3.8308	0.805	0.2744	1.5821	0.805	0.7951	0.1290	0.0475
No log	33.0	429	3.8453	0.82	0.2668	1.6016	0.82	0.8070	0.1307	0.0476
No log	34.0	442	3.8837	0.81	0.2767	1.6053	0.81	0.8003	0.1368	0.0464
No log	35.0	455	3.8103	0.83	0.2561	1.6922	0.83	0.8328	0.1285	0.0434
No log	36.0	468	3.8702	0.81	0.2712	1.8361	0.81	0.8011	0.1387	0.0465
No log	37.0	481	3.8080	0.81	0.2635	1.7576	0.81	0.7993	0.1247	0.0398
No log	38.0	494	3.8234	0.81	0.2672	1.7059	0.81	0.8008	0.1349	0.0420
3.5835	39.0	507	3.8962	0.81	0.2789	1.7265	0.81	0.8106	0.1459	0.0448
3.5835	40.0	520	3.8600	0.805	0.2731	1.7552	0.805	0.7917	0.1328	0.0415
3.5835	41.0	533	3.9201	0.815	0.2799	1.8308	0.815	0.8072	0.1416	0.0428
3.5835	42.0	546	3.9188	0.815	0.2721	1.7661	0.815	0.8102	0.1153	0.0416
3.5835	43.0	559	3.8939	0.82	0.2660	1.6435	0.82	0.8171	0.1349	0.0414
3.5835	44.0	572	3.9485	0.8	0.2806	1.8239	0.8000	0.7896	0.1515	0.0432
3.5835	45.0	585	3.9293	0.805	0.2731	1.7495	0.805	0.7980	0.1468	0.0408
3.5835	46.0	598	3.9623	0.815	0.2744	1.8305	0.815	0.8102	0.1397	0.0423
3.5835	47.0	611	3.9737	0.815	0.2776	1.8283	0.815	0.8132	0.1394	0.0415
3.5835	48.0	624	4.0066	0.825	0.2839	1.8282	0.825	0.8232	0.1422	0.0420
3.5835	49.0	637	4.0039	0.82	0.2789	1.8268	0.82	0.8166	0.1371	0.0423
3.5835	50.0	650	4.0113	0.82	0.2812	1.7108	0.82	0.8166	0.1352	0.0411
3.5835	51.0	663	4.0170	0.82	0.2810	1.7028	0.82	0.8197	0.1308	0.0404
3.5835	52.0	676	4.0325	0.82	0.2806	1.7729	0.82	0.8197	0.1364	0.0413
3.5835	53.0	689	4.0359	0.815	0.2824	1.7656	0.815	0.8065	0.1372	0.0417
3.5835	54.0	702	4.0536	0.82	0.2833	1.7665	0.82	0.8197	0.1390	0.0414
3.5835	55.0	715	4.0646	0.825	0.2864	1.6994	0.825	0.8232	0.1505	0.0405
3.5835	56.0	728	4.0602	0.83	0.2830	1.7634	0.83	0.8264	0.1543	0.0408
3.5835	57.0	741	4.0725	0.825	0.2850	1.7749	0.825	0.8220	0.1521	0.0411
3.5835	58.0	754	4.0659	0.83	0.2819	1.7018	0.83	0.8252	0.1426	0.0400
3.5835	59.0	767	4.0838	0.83	0.2832	1.8336	0.83	0.8264	0.1441	0.0416
3.5835	60.0	780	4.0928	0.825	0.2850	1.7666	0.825	0.8232	0.1425	0.0411
3.5835	61.0	793	4.0945	0.83	0.2841	1.7751	0.83	0.8252	0.1442	0.0412
3.5835	62.0	806	4.1037	0.83	0.2863	1.7089	0.83	0.8252	0.1389	0.0404
3.5835	63.0	819	4.1134	0.825	0.2879	1.6997	0.825	0.8220	0.1299	0.0404
3.5835	64.0	832	4.1176	0.83	0.2869	1.7757	0.83	0.8264	0.1378	0.0411
3.5835	65.0	845	4.1232	0.83	0.2862	1.7746	0.83	0.8264	0.1424	0.0406
3.5835	66.0	858	4.1335	0.83	0.2866	1.8341	0.83	0.8252	0.1439	0.0414
3.5835	67.0	871	4.1352	0.825	0.2865	1.7806	0.825	0.8229	0.1495	0.0413
3.5835	68.0	884	4.1348	0.83	0.2864	1.7896	0.83	0.8252	0.1427	0.0405
3.5835	69.0	897	4.1445	0.83	0.2883	1.7743	0.83	0.8252	0.1404	0.0406
3.5835	70.0	910	4.1455	0.83	0.2868	1.8347	0.83	0.8252	0.1445	0.0409
3.5835	71.0	923	4.1512	0.83	0.2876	1.7757	0.83	0.8252	0.1450	0.0409
3.5835	72.0	936	4.1579	0.83	0.2884	1.7730	0.83	0.8252	0.1463	0.0407
3.5835	73.0	949	4.1602	0.83	0.2879	1.7744	0.83	0.8252	0.1400	0.0406
3.5835	74.0	962	4.1661	0.83	0.2886	1.7724	0.83	0.8252	0.1464	0.0407
3.5835	75.0	975	4.1724	0.83	0.2894	1.7735	0.83	0.8252	0.1465	0.0407
3.5835	76.0	988	4.1726	0.83	0.2888	1.7730	0.83	0.8252	0.1455	0.0406
3.1987	77.0	1001	4.1784	0.83	0.2894	1.7734	0.83	0.8252	0.1460	0.0404
3.1987	78.0	1014	4.1805	0.83	0.2891	1.7744	0.83	0.8264	0.1455	0.0406
3.1987	79.0	1027	4.1858	0.83	0.2898	1.7749	0.83	0.8252	0.1551	0.0406
3.1987	80.0	1040	4.1881	0.83	0.2898	1.7749	0.83	0.8264	0.1549	0.0406
3.1987	81.0	1053	4.1907	0.83	0.2899	1.7744	0.83	0.8252	0.1562	0.0405
3.1987	82.0	1066	4.1930	0.83	0.2900	1.7763	0.83	0.8252	0.1555	0.0406
3.1987	83.0	1079	4.1954	0.83	0.2900	1.7761	0.83	0.8264	0.1552	0.0406
3.1987	84.0	1092	4.2000	0.83	0.2905	1.7773	0.83	0.8264	0.1553	0.0408
3.1987	85.0	1105	4.2015	0.83	0.2905	1.7746	0.83	0.8252	0.1501	0.0404
3.1987	86.0	1118	4.2041	0.83	0.2909	1.7765	0.83	0.8252	0.1565	0.0405
3.1987	87.0	1131	4.2060	0.83	0.2908	1.7768	0.83	0.8264	0.1556	0.0405
3.1987	88.0	1144	4.2070	0.83	0.2908	1.7756	0.83	0.8252	0.1565	0.0404
3.1987	89.0	1157	4.2093	0.83	0.2908	1.7770	0.83	0.8264	0.1556	0.0407
3.1987	90.0	1170	4.2099	0.83	0.2909	1.7763	0.83	0.8252	0.1559	0.0404
3.1987	91.0	1183	4.2123	0.83	0.2910	1.7776	0.83	0.8264	0.1559	0.0407
3.1987	92.0	1196	4.2138	0.83	0.2912	1.7764	0.83	0.8252	0.1512	0.0404
3.1987	93.0	1209	4.2142	0.83	0.2911	1.7767	0.83	0.8252	0.1505	0.0403
3.1987	94.0	1222	4.2154	0.83	0.2912	1.7765	0.83	0.8252	0.1561	0.0404
3.1987	95.0	1235	4.2162	0.83	0.2912	1.7766	0.83	0.8264	0.1560	0.0404
3.1987	96.0	1248	4.2165	0.83	0.2912	1.7769	0.83	0.8264	0.1561	0.0404
3.1987	97.0	1261	4.2180	0.83	0.2914	1.7773	0.83	0.8264	0.1561	0.0404
3.1987	98.0	1274	4.2180	0.83	0.2913	1.7768	0.83	0.8252	0.1561	0.0404
3.1987	99.0	1287	4.2183	0.83	0.2914	1.7772	0.83	0.8264	0.1561	0.0404
3.1987	100.0	1300	4.2183	0.83	0.2913	1.7770	0.83	0.8264	0.1561	0.0404

Framework versions

Transformers 4.26.1
Pytorch 1.13.1.post200
Datasets 2.9.0
Tokenizers 0.13.2

300-tiny_tobacco3482_kd_NKD_t1.0_g1.5

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

NSDT 3DConvert

UnrealSynth

DreamTexture.js