300-tiny_tobacco3482_kd_CEKD_t2.5_a0.5

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: 0.4770
Accuracy: 0.82
Brier Loss: 0.2875
Nll: 1.3922
F1 Micro: 0.82
F1 Macro: 0.8020
Ece: 0.2219
Aurc: 0.0517

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	1.7938	0.235	0.8938	7.8599	0.235	0.1394	0.3127	0.7433
No log	2.0	26	1.2738	0.455	0.6913	3.6965	0.455	0.3679	0.2904	0.3352
No log	3.0	39	1.0682	0.555	0.5748	2.0566	0.555	0.4968	0.2475	0.2296
No log	4.0	52	0.8509	0.655	0.4621	1.7782	0.655	0.6085	0.2242	0.1405
No log	5.0	65	0.7670	0.71	0.4142	1.4993	0.7100	0.6560	0.2271	0.1082
No log	6.0	78	0.7285	0.735	0.3857	1.5730	0.735	0.6874	0.2098	0.0996
No log	7.0	91	0.7052	0.72	0.3804	1.4916	0.72	0.6974	0.2249	0.0959
No log	8.0	104	0.7590	0.71	0.3925	1.8047	0.7100	0.6641	0.1956	0.1008
No log	9.0	117	0.7657	0.71	0.4006	1.8296	0.7100	0.7169	0.2330	0.1025
No log	10.0	130	0.6512	0.755	0.3514	1.5899	0.755	0.7256	0.1863	0.0853
No log	11.0	143	0.6615	0.775	0.3638	1.8180	0.775	0.7564	0.2106	0.0911
No log	12.0	156	0.6195	0.785	0.3398	1.6998	0.785	0.7419	0.2337	0.0643
No log	13.0	169	0.6065	0.78	0.3471	1.5917	0.78	0.7550	0.2280	0.0793
No log	14.0	182	0.6314	0.75	0.3486	1.9235	0.75	0.7315	0.2105	0.0755
No log	15.0	195	0.6426	0.745	0.3686	1.8633	0.745	0.7100	0.2099	0.0891
No log	16.0	208	0.5849	0.765	0.3476	1.3466	0.765	0.7505	0.1978	0.0827
No log	17.0	221	0.5604	0.79	0.3311	1.3948	0.79	0.7581	0.2258	0.0705
No log	18.0	234	0.5504	0.78	0.3230	1.4757	0.78	0.7712	0.2104	0.0624
No log	19.0	247	0.5586	0.785	0.3247	1.5297	0.785	0.7642	0.2159	0.0655
No log	20.0	260	0.5879	0.78	0.3366	1.5348	0.78	0.7727	0.2162	0.0716
No log	21.0	273	0.5558	0.805	0.3113	1.5720	0.805	0.7945	0.2161	0.0652
No log	22.0	286	0.5439	0.795	0.3258	1.7373	0.795	0.7883	0.2307	0.0745
No log	23.0	299	0.5155	0.795	0.3094	1.4183	0.795	0.7725	0.2221	0.0625
No log	24.0	312	0.5039	0.81	0.2994	1.4458	0.81	0.7830	0.2114	0.0624
No log	25.0	325	0.5142	0.81	0.3101	1.2798	0.81	0.7928	0.2205	0.0624
No log	26.0	338	0.5007	0.8	0.3100	1.2390	0.8000	0.7730	0.2038	0.0645
No log	27.0	351	0.4779	0.815	0.2865	1.3312	0.815	0.7863	0.2061	0.0518
No log	28.0	364	0.4893	0.825	0.2927	1.3993	0.825	0.8009	0.2219	0.0555
No log	29.0	377	0.4938	0.82	0.2996	1.4038	0.82	0.7888	0.2138	0.0586
No log	30.0	390	0.4668	0.82	0.2795	1.3366	0.82	0.7944	0.2217	0.0495
No log	31.0	403	0.4662	0.8	0.2805	1.1721	0.8000	0.7761	0.2009	0.0494
No log	32.0	416	0.4787	0.82	0.2887	1.3872	0.82	0.8043	0.2161	0.0542
No log	33.0	429	0.4842	0.81	0.2909	1.4774	0.81	0.7854	0.2246	0.0562
No log	34.0	442	0.4899	0.81	0.2979	1.4419	0.81	0.7843	0.2155	0.0607
No log	35.0	455	0.4832	0.815	0.2920	1.3892	0.815	0.7907	0.2296	0.0552
No log	36.0	468	0.4739	0.815	0.2869	1.2603	0.815	0.7932	0.2385	0.0532
No log	37.0	481	0.4747	0.81	0.2877	1.4390	0.81	0.7848	0.2163	0.0526
No log	38.0	494	0.4710	0.815	0.2842	1.3024	0.815	0.7885	0.2153	0.0516
0.2992	39.0	507	0.4712	0.81	0.2839	1.3676	0.81	0.7860	0.2282	0.0518
0.2992	40.0	520	0.4772	0.815	0.2883	1.3845	0.815	0.7953	0.2216	0.0527
0.2992	41.0	533	0.4751	0.82	0.2877	1.3207	0.82	0.8018	0.2177	0.0521
0.2992	42.0	546	0.4724	0.82	0.2860	1.3075	0.82	0.8018	0.2183	0.0508
0.2992	43.0	559	0.4745	0.82	0.2869	1.3079	0.82	0.8020	0.2184	0.0522
0.2992	44.0	572	0.4779	0.815	0.2884	1.4039	0.815	0.7922	0.2142	0.0531
0.2992	45.0	585	0.4738	0.82	0.2859	1.3153	0.82	0.8018	0.2079	0.0516
0.2992	46.0	598	0.4755	0.815	0.2874	1.3273	0.815	0.7922	0.2279	0.0526
0.2992	47.0	611	0.4736	0.82	0.2858	1.3190	0.82	0.8018	0.2182	0.0515
0.2992	48.0	624	0.4753	0.82	0.2876	1.3170	0.82	0.8018	0.2274	0.0521
0.2992	49.0	637	0.4755	0.82	0.2866	1.4452	0.82	0.8018	0.2245	0.0516
0.2992	50.0	650	0.4754	0.815	0.2869	1.2915	0.815	0.7924	0.2336	0.0523
0.2992	51.0	663	0.4747	0.82	0.2861	1.3336	0.82	0.8020	0.2309	0.0517
0.2992	52.0	676	0.4765	0.815	0.2880	1.3456	0.815	0.7924	0.2137	0.0524
0.2992	53.0	689	0.4756	0.82	0.2866	1.3288	0.82	0.8020	0.2236	0.0518
0.2992	54.0	702	0.4757	0.82	0.2873	1.3860	0.82	0.8018	0.2085	0.0516
0.2992	55.0	715	0.4753	0.815	0.2866	1.3284	0.815	0.7922	0.2100	0.0515
0.2992	56.0	728	0.4759	0.82	0.2870	1.3199	0.82	0.8020	0.2240	0.0518
0.2992	57.0	741	0.4764	0.82	0.2874	1.3901	0.82	0.8020	0.2241	0.0517
0.2992	58.0	754	0.4754	0.815	0.2870	1.3246	0.815	0.7924	0.2260	0.0520
0.2992	59.0	767	0.4759	0.815	0.2870	1.3862	0.815	0.7924	0.2176	0.0520
0.2992	60.0	780	0.4765	0.815	0.2874	1.3873	0.815	0.7924	0.2266	0.0523
0.2992	61.0	793	0.4763	0.82	0.2873	1.3851	0.82	0.8020	0.2161	0.0517
0.2992	62.0	806	0.4768	0.815	0.2878	1.3903	0.815	0.7924	0.2128	0.0522
0.2992	63.0	819	0.4767	0.82	0.2876	1.3866	0.82	0.8020	0.2120	0.0521
0.2992	64.0	832	0.4762	0.82	0.2872	1.3910	0.82	0.8020	0.2157	0.0516
0.2992	65.0	845	0.4765	0.82	0.2874	1.3892	0.82	0.8020	0.2178	0.0519
0.2992	66.0	858	0.4767	0.82	0.2875	1.3462	0.82	0.8020	0.2180	0.0519
0.2992	67.0	871	0.4764	0.82	0.2872	1.3894	0.82	0.8020	0.2252	0.0518
0.2992	68.0	884	0.4767	0.82	0.2874	1.3860	0.82	0.8020	0.2118	0.0518
0.2992	69.0	897	0.4766	0.82	0.2874	1.3894	0.82	0.8020	0.2180	0.0519
0.2992	70.0	910	0.4765	0.82	0.2872	1.3882	0.82	0.8020	0.2280	0.0517
0.2992	71.0	923	0.4766	0.82	0.2874	1.3875	0.82	0.8020	0.2177	0.0519
0.2992	72.0	936	0.4765	0.82	0.2874	1.3880	0.82	0.8020	0.2148	0.0517
0.2992	73.0	949	0.4766	0.82	0.2873	1.3915	0.82	0.8020	0.2109	0.0516
0.2992	74.0	962	0.4765	0.82	0.2872	1.3900	0.82	0.8020	0.2110	0.0517
0.2992	75.0	975	0.4769	0.82	0.2875	1.3913	0.82	0.8020	0.2251	0.0520
0.2992	76.0	988	0.4770	0.82	0.2876	1.3909	0.82	0.8020	0.2196	0.0520
0.0695	77.0	1001	0.4768	0.82	0.2875	1.3890	0.82	0.8020	0.2212	0.0517
0.0695	78.0	1014	0.4767	0.82	0.2873	1.3935	0.82	0.8020	0.2281	0.0518
0.0695	79.0	1027	0.4767	0.82	0.2874	1.3897	0.82	0.8020	0.2282	0.0517
0.0695	80.0	1040	0.4770	0.82	0.2876	1.3889	0.82	0.8020	0.2174	0.0518
0.0695	81.0	1053	0.4770	0.82	0.2875	1.3935	0.82	0.8020	0.2221	0.0518
0.0695	82.0	1066	0.4766	0.82	0.2873	1.3901	0.82	0.8020	0.2283	0.0517
0.0695	83.0	1079	0.4768	0.82	0.2874	1.3902	0.82	0.8020	0.2283	0.0517
0.0695	84.0	1092	0.4770	0.82	0.2874	1.3917	0.82	0.8020	0.2217	0.0518
0.0695	85.0	1105	0.4769	0.82	0.2875	1.3913	0.82	0.8020	0.2283	0.0518
0.0695	86.0	1118	0.4769	0.82	0.2874	1.3916	0.82	0.8020	0.2282	0.0517
0.0695	87.0	1131	0.4769	0.82	0.2874	1.3912	0.82	0.8020	0.2218	0.0517
0.0695	88.0	1144	0.4770	0.82	0.2875	1.3923	0.82	0.8020	0.2218	0.0517
0.0695	89.0	1157	0.4768	0.82	0.2874	1.3905	0.82	0.8020	0.2283	0.0518
0.0695	90.0	1170	0.4769	0.82	0.2875	1.3924	0.82	0.8020	0.2219	0.0517
0.0695	91.0	1183	0.4769	0.82	0.2874	1.3923	0.82	0.8020	0.2219	0.0517
0.0695	92.0	1196	0.4768	0.82	0.2874	1.3908	0.82	0.8020	0.2219	0.0517
0.0695	93.0	1209	0.4770	0.82	0.2875	1.3909	0.82	0.8020	0.2219	0.0518
0.0695	94.0	1222	0.4768	0.82	0.2873	1.3918	0.82	0.8020	0.2219	0.0517
0.0695	95.0	1235	0.4769	0.82	0.2874	1.3914	0.82	0.8020	0.2219	0.0517
0.0695	96.0	1248	0.4770	0.82	0.2875	1.3917	0.82	0.8020	0.2219	0.0517
0.0695	97.0	1261	0.4769	0.82	0.2874	1.3918	0.82	0.8020	0.2219	0.0517
0.0695	98.0	1274	0.4770	0.82	0.2875	1.3920	0.82	0.8020	0.2219	0.0517
0.0695	99.0	1287	0.4770	0.82	0.2875	1.3922	0.82	0.8020	0.2219	0.0517
0.0695	100.0	1300	0.4770	0.82	0.2875	1.3922	0.82	0.8020	0.2219	0.0517

Framework versions

Transformers 4.26.1
Pytorch 1.13.1.post200
Datasets 2.9.0
Tokenizers 0.13.2

300-tiny_tobacco3482_kd_CEKD_t2.5_a0.5

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

NSDT 3DConvert

UnrealSynth

DreamTexture.js