114-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: 4.1549
Accuracy: 0.83
Brier Loss: 0.2823
Nll: 1.5555
F1 Micro: 0.83
F1 Macro: 0.8165
Ece: 0.1345
Aurc: 0.0505

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	4.6466	0.205	0.8898	8.5236	0.205	0.1171	0.2921	0.7976
No log	2.0	26	4.1877	0.37	0.7948	4.6881	0.37	0.2765	0.3019	0.4618
No log	3.0	39	4.0470	0.495	0.7046	4.2682	0.495	0.3642	0.2930	0.3256
No log	4.0	52	3.9713	0.535	0.6402	3.2866	0.535	0.3996	0.2908	0.2421
No log	5.0	65	3.8631	0.645	0.5467	2.7568	0.645	0.5172	0.3135	0.1523
No log	6.0	78	3.9102	0.65	0.5148	2.6608	0.65	0.5059	0.2764	0.1533
No log	7.0	91	3.8517	0.67	0.4656	2.6499	0.67	0.5649	0.2541	0.1313
No log	8.0	104	4.1021	0.645	0.4936	2.7236	0.645	0.5425	0.2230	0.1622
No log	9.0	117	3.9369	0.66	0.4363	2.1539	0.66	0.5584	0.2057	0.1246
No log	10.0	130	4.0468	0.675	0.4564	2.9594	0.675	0.5976	0.2109	0.1319
No log	11.0	143	3.8168	0.735	0.3788	2.3105	0.735	0.6427	0.2121	0.0768
No log	12.0	156	3.9052	0.72	0.3890	2.7166	0.72	0.6500	0.1906	0.0978
No log	13.0	169	3.8460	0.735	0.3741	2.3764	0.735	0.6513	0.1941	0.0904
No log	14.0	182	3.7683	0.78	0.3418	2.0497	0.78	0.7024	0.1980	0.0656
No log	15.0	195	3.7220	0.78	0.3242	2.1470	0.78	0.7176	0.1780	0.0624
No log	16.0	208	3.7405	0.785	0.3170	1.6643	0.785	0.7151	0.1874	0.0642
No log	17.0	221	3.8003	0.775	0.3264	1.8059	0.775	0.7434	0.1699	0.0740
No log	18.0	234	3.7431	0.795	0.3134	1.6740	0.795	0.7780	0.1891	0.0625
No log	19.0	247	3.7844	0.8	0.3366	1.6058	0.8000	0.7896	0.2193	0.0707
No log	20.0	260	3.7887	0.81	0.3142	1.8916	0.81	0.7984	0.1869	0.0660
No log	21.0	273	3.7557	0.82	0.2997	1.6302	0.82	0.8003	0.1653	0.0548
No log	22.0	286	3.8324	0.815	0.3303	2.0226	0.815	0.8124	0.1903	0.0742
No log	23.0	299	3.7161	0.815	0.2916	1.9124	0.815	0.8080	0.1504	0.0554
No log	24.0	312	3.8438	0.785	0.3072	1.9052	0.785	0.7848	0.1823	0.0688
No log	25.0	325	3.7427	0.82	0.2859	1.9856	0.82	0.8014	0.1583	0.0546
No log	26.0	338	3.7653	0.81	0.2838	1.6727	0.81	0.8053	0.1527	0.0565
No log	27.0	351	3.7667	0.82	0.2927	1.8255	0.82	0.8134	0.1655	0.0601
No log	28.0	364	3.7104	0.815	0.2747	1.7848	0.815	0.8052	0.1732	0.0513
No log	29.0	377	3.7794	0.825	0.2884	1.7768	0.825	0.8194	0.1616	0.0581
No log	30.0	390	3.7582	0.81	0.2732	1.8177	0.81	0.7965	0.1358	0.0520
No log	31.0	403	3.7540	0.82	0.2790	1.7627	0.82	0.8027	0.1399	0.0520
No log	32.0	416	3.7146	0.82	0.2644	1.7950	0.82	0.8097	0.1532	0.0487
No log	33.0	429	3.8304	0.79	0.2983	2.3232	0.79	0.7817	0.1596	0.0599
No log	34.0	442	3.7604	0.82	0.2714	1.7869	0.82	0.7976	0.1429	0.0474
No log	35.0	455	3.8126	0.815	0.2768	1.8130	0.815	0.8075	0.1388	0.0510
No log	36.0	468	3.7828	0.825	0.2648	1.5233	0.825	0.8101	0.1603	0.0471
No log	37.0	481	3.8297	0.82	0.2781	1.6717	0.82	0.8114	0.1557	0.0491
No log	38.0	494	3.8217	0.82	0.2704	1.5368	0.82	0.8058	0.1461	0.0498
3.4744	39.0	507	3.8171	0.845	0.2639	1.7121	0.845	0.8325	0.1270	0.0468
3.4744	40.0	520	3.8336	0.83	0.2691	1.6086	0.83	0.8158	0.1289	0.0486
3.4744	41.0	533	3.8612	0.815	0.2699	1.6193	0.815	0.8052	0.1393	0.0516
3.4744	42.0	546	3.8801	0.825	0.2716	1.6084	0.825	0.8139	0.1309	0.0499
3.4744	43.0	559	3.8851	0.82	0.2744	1.6179	0.82	0.8124	0.1320	0.0504
3.4744	44.0	572	3.8818	0.825	0.2708	1.5941	0.825	0.8107	0.1295	0.0492
3.4744	45.0	585	3.8843	0.83	0.2656	1.5945	0.83	0.8207	0.1301	0.0485
3.4744	46.0	598	3.9274	0.82	0.2749	1.6790	0.82	0.8079	0.1391	0.0504
3.4744	47.0	611	3.9210	0.83	0.2736	1.6600	0.83	0.8137	0.1198	0.0486
3.4744	48.0	624	3.9345	0.82	0.2746	1.7165	0.82	0.8056	0.1393	0.0488
3.4744	49.0	637	3.9491	0.835	0.2706	1.5963	0.835	0.8240	0.1355	0.0495
3.4744	50.0	650	3.9607	0.815	0.2767	1.6837	0.815	0.8029	0.1407	0.0503
3.4744	51.0	663	3.9657	0.825	0.2727	1.6016	0.825	0.8125	0.1338	0.0500
3.4744	52.0	676	3.9763	0.825	0.2766	1.6100	0.825	0.8125	0.1419	0.0501
3.4744	53.0	689	3.9726	0.83	0.2746	1.5261	0.83	0.8189	0.1296	0.0495
3.4744	54.0	702	3.9863	0.825	0.2765	1.6034	0.825	0.8125	0.1320	0.0502
3.4744	55.0	715	3.9922	0.815	0.2767	1.7367	0.815	0.8029	0.1305	0.0494
3.4744	56.0	728	4.0013	0.825	0.2745	1.5431	0.825	0.8133	0.1239	0.0496
3.4744	57.0	741	4.0124	0.825	0.2773	1.6129	0.825	0.8118	0.1291	0.0504
3.4744	58.0	754	4.0031	0.83	0.2740	1.5309	0.83	0.8174	0.1353	0.0489
3.4744	59.0	767	4.0280	0.825	0.2781	1.6088	0.825	0.8107	0.1419	0.0505
3.4744	60.0	780	4.0266	0.83	0.2757	1.5379	0.83	0.8170	0.1375	0.0501
3.4744	61.0	793	4.0333	0.82	0.2769	1.5573	0.82	0.8060	0.1430	0.0504
3.4744	62.0	806	4.0370	0.835	0.2766	1.5287	0.835	0.8221	0.1284	0.0498
3.4744	63.0	819	4.0418	0.825	0.2763	1.5459	0.825	0.8107	0.1381	0.0500
3.4744	64.0	832	4.0393	0.83	0.2745	1.5433	0.83	0.8174	0.1371	0.0493
3.4744	65.0	845	4.0636	0.83	0.2779	1.5378	0.83	0.8174	0.1369	0.0505
3.4744	66.0	858	4.0627	0.825	0.2790	1.6074	0.825	0.8107	0.1378	0.0504
3.4744	67.0	871	4.0715	0.825	0.2797	1.5569	0.825	0.8107	0.1338	0.0507
3.4744	68.0	884	4.0698	0.835	0.2770	1.5355	0.835	0.8221	0.1362	0.0498
3.4744	69.0	897	4.0808	0.825	0.2798	1.5505	0.825	0.8107	0.1304	0.0506
3.4744	70.0	910	4.0837	0.825	0.2794	1.5387	0.825	0.8118	0.1432	0.0502
3.4744	71.0	923	4.0868	0.825	0.2793	1.6048	0.825	0.8107	0.1343	0.0507
3.4744	72.0	936	4.0912	0.825	0.2793	1.5487	0.825	0.8118	0.1482	0.0504
3.4744	73.0	949	4.0962	0.825	0.2795	1.5468	0.825	0.8118	0.1414	0.0502
3.4744	74.0	962	4.1011	0.83	0.2802	1.5485	0.83	0.8165	0.1528	0.0501
3.4744	75.0	975	4.1044	0.825	0.2801	1.6062	0.825	0.8107	0.1410	0.0504
3.4744	76.0	988	4.1094	0.825	0.2804	1.5688	0.825	0.8107	0.1435	0.0506
3.1992	77.0	1001	4.1127	0.825	0.2805	1.6066	0.825	0.8107	0.1496	0.0506
3.1992	78.0	1014	4.1164	0.83	0.2804	1.5433	0.83	0.8165	0.1384	0.0503
3.1992	79.0	1027	4.1198	0.83	0.2805	1.5485	0.83	0.8165	0.1530	0.0504
3.1992	80.0	1040	4.1228	0.83	0.2808	1.5610	0.83	0.8165	0.1385	0.0504
3.1992	81.0	1053	4.1257	0.83	0.2809	1.5505	0.83	0.8165	0.1475	0.0504
3.1992	82.0	1066	4.1301	0.83	0.2812	1.5504	0.83	0.8165	0.1474	0.0504
3.1992	83.0	1079	4.1310	0.83	0.2811	1.5544	0.83	0.8165	0.1390	0.0504
3.1992	84.0	1092	4.1345	0.83	0.2813	1.5513	0.83	0.8165	0.1476	0.0504
3.1992	85.0	1105	4.1373	0.83	0.2814	1.5530	0.83	0.8165	0.1340	0.0504
3.1992	86.0	1118	4.1401	0.83	0.2818	1.5528	0.83	0.8165	0.1427	0.0504
3.1992	87.0	1131	4.1416	0.83	0.2816	1.5521	0.83	0.8165	0.1341	0.0504
3.1992	88.0	1144	4.1439	0.83	0.2819	1.5527	0.83	0.8165	0.1341	0.0504
3.1992	89.0	1157	4.1453	0.83	0.2819	1.5536	0.83	0.8165	0.1343	0.0504
3.1992	90.0	1170	4.1475	0.83	0.2820	1.5534	0.83	0.8165	0.1343	0.0504
3.1992	91.0	1183	4.1488	0.83	0.2820	1.5514	0.83	0.8165	0.1344	0.0504
3.1992	92.0	1196	4.1495	0.83	0.2820	1.5561	0.83	0.8165	0.1392	0.0503
3.1992	93.0	1209	4.1510	0.83	0.2821	1.5510	0.83	0.8165	0.1432	0.0506
3.1992	94.0	1222	4.1518	0.83	0.2821	1.5548	0.83	0.8165	0.1344	0.0504
3.1992	95.0	1235	4.1529	0.83	0.2822	1.5544	0.83	0.8165	0.1345	0.0505
3.1992	96.0	1248	4.1536	0.83	0.2822	1.5564	0.83	0.8165	0.1344	0.0504
3.1992	97.0	1261	4.1542	0.83	0.2822	1.5542	0.83	0.8165	0.1345	0.0505
3.1992	98.0	1274	4.1545	0.83	0.2823	1.5574	0.83	0.8165	0.1345	0.0505
3.1992	99.0	1287	4.1548	0.83	0.2823	1.5553	0.83	0.8165	0.1345	0.0505
3.1992	100.0	1300	4.1549	0.83	0.2823	1.5555	0.83	0.8165	0.1345	0.0505

Framework versions

Transformers 4.26.1
Pytorch 1.13.1.post200
Datasets 2.9.0
Tokenizers 0.13.2

114-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