171-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: 0.4688
Accuracy: 0.815
Brier Loss: 0.3067
Nll: 1.4679
F1 Micro: 0.815
F1 Macro: 0.7970
Ece: 0.2440
Aurc: 0.0500

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	1.6008	0.23	0.8921	8.0367	0.23	0.1380	0.3153	0.7486
No log	2.0	26	1.1383	0.445	0.6997	3.6320	0.445	0.3583	0.2866	0.3390
No log	3.0	39	0.9781	0.555	0.5896	2.1989	0.555	0.4763	0.2856	0.2440
No log	4.0	52	0.7953	0.65	0.4796	1.7904	0.65	0.5880	0.2308	0.1417
No log	5.0	65	0.7282	0.705	0.4370	1.4923	0.705	0.6654	0.2538	0.1123
No log	6.0	78	0.6794	0.73	0.3987	1.5706	0.7300	0.6928	0.2386	0.1041
No log	7.0	91	0.6813	0.73	0.4024	1.6519	0.7300	0.6984	0.2553	0.1027
No log	8.0	104	0.6669	0.72	0.3910	1.6057	0.72	0.6811	0.2234	0.0990
No log	9.0	117	0.7152	0.72	0.4167	1.9716	0.72	0.7201	0.2259	0.1091
No log	10.0	130	0.6722	0.745	0.3751	1.9561	0.745	0.7290	0.2362	0.0849
No log	11.0	143	0.6263	0.75	0.3817	1.8594	0.75	0.7238	0.2511	0.0980
No log	12.0	156	0.6259	0.725	0.3946	1.8363	0.7250	0.6835	0.2186	0.0974
No log	13.0	169	0.5756	0.77	0.3487	1.3847	0.7700	0.7171	0.2271	0.0723
No log	14.0	182	0.5670	0.76	0.3492	1.7986	0.76	0.7323	0.2201	0.0713
No log	15.0	195	0.5538	0.785	0.3532	1.6319	0.785	0.7608	0.2479	0.0629
No log	16.0	208	0.5634	0.75	0.3582	1.5131	0.75	0.7397	0.2333	0.0747
No log	17.0	221	0.5348	0.77	0.3378	1.5843	0.7700	0.7421	0.2193	0.0646
No log	18.0	234	0.5306	0.78	0.3310	1.6298	0.78	0.7618	0.2290	0.0644
No log	19.0	247	0.5185	0.805	0.3400	1.4945	0.805	0.7755	0.2627	0.0622
No log	20.0	260	0.5335	0.76	0.3402	1.5758	0.76	0.7108	0.2372	0.0699
No log	21.0	273	0.5191	0.76	0.3389	1.3860	0.76	0.7413	0.2587	0.0661
No log	22.0	286	0.5198	0.785	0.3423	1.4790	0.785	0.7607	0.2513	0.0649
No log	23.0	299	0.5155	0.79	0.3344	1.5003	0.79	0.7648	0.2393	0.0671
No log	24.0	312	0.5156	0.775	0.3380	1.5898	0.775	0.7388	0.2295	0.0667
No log	25.0	325	0.4808	0.815	0.3033	1.4602	0.815	0.7837	0.2520	0.0520
No log	26.0	338	0.4975	0.785	0.3325	1.3864	0.785	0.7563	0.2298	0.0673
No log	27.0	351	0.4988	0.785	0.3257	1.5206	0.785	0.7717	0.2156	0.0638
No log	28.0	364	0.4928	0.795	0.3209	1.3717	0.795	0.7719	0.2303	0.0612
No log	29.0	377	0.4660	0.81	0.3022	1.2190	0.81	0.7864	0.2285	0.0485
No log	30.0	390	0.4777	0.815	0.3123	1.4266	0.815	0.7926	0.2535	0.0562
No log	31.0	403	0.4695	0.82	0.3067	1.3425	0.82	0.8000	0.2338	0.0528
No log	32.0	416	0.4701	0.815	0.3026	1.3247	0.815	0.7893	0.2259	0.0522
No log	33.0	429	0.4625	0.82	0.3023	1.2646	0.82	0.7915	0.2441	0.0486
No log	34.0	442	0.4684	0.81	0.3080	1.3468	0.81	0.7846	0.2373	0.0521
No log	35.0	455	0.4629	0.81	0.3000	1.3441	0.81	0.7869	0.2375	0.0492
No log	36.0	468	0.4680	0.81	0.3074	1.2158	0.81	0.7894	0.2417	0.0508
No log	37.0	481	0.4672	0.81	0.3053	1.3329	0.81	0.7866	0.2320	0.0508
No log	38.0	494	0.4716	0.805	0.3091	1.2975	0.805	0.7863	0.2361	0.0545
0.3111	39.0	507	0.4703	0.805	0.3081	1.2855	0.805	0.7863	0.2473	0.0534
0.3111	40.0	520	0.4692	0.81	0.3073	1.2833	0.81	0.7894	0.2361	0.0525
0.3111	41.0	533	0.4681	0.81	0.3068	1.2804	0.81	0.7890	0.2386	0.0517
0.3111	42.0	546	0.4672	0.81	0.3058	1.4597	0.81	0.7898	0.2276	0.0521
0.3111	43.0	559	0.4691	0.81	0.3080	1.4136	0.81	0.7894	0.2280	0.0520
0.3111	44.0	572	0.4664	0.815	0.3048	1.4593	0.815	0.7921	0.2459	0.0509
0.3111	45.0	585	0.4684	0.81	0.3069	1.4071	0.81	0.7894	0.2415	0.0514
0.3111	46.0	598	0.4688	0.81	0.3066	1.4084	0.81	0.7890	0.2174	0.0516
0.3111	47.0	611	0.4683	0.81	0.3061	1.4052	0.81	0.7890	0.2406	0.0515
0.3111	48.0	624	0.4677	0.81	0.3065	1.4045	0.81	0.7890	0.2346	0.0508
0.3111	49.0	637	0.4679	0.81	0.3058	1.4072	0.81	0.7890	0.2177	0.0507
0.3111	50.0	650	0.4679	0.81	0.3061	1.4681	0.81	0.7890	0.2619	0.0510
0.3111	51.0	663	0.4688	0.81	0.3068	1.4662	0.81	0.7890	0.2325	0.0513
0.3111	52.0	676	0.4679	0.81	0.3063	1.4062	0.81	0.7890	0.2257	0.0508
0.3111	53.0	689	0.4682	0.81	0.3064	1.4667	0.81	0.7890	0.2279	0.0512
0.3111	54.0	702	0.4674	0.81	0.3058	1.4075	0.81	0.7890	0.2269	0.0507
0.3111	55.0	715	0.4689	0.81	0.3069	1.4674	0.81	0.7890	0.2428	0.0511
0.3111	56.0	728	0.4678	0.81	0.3062	1.4081	0.81	0.7890	0.2402	0.0507
0.3111	57.0	741	0.4691	0.81	0.3069	1.4691	0.81	0.7890	0.2279	0.0511
0.3111	58.0	754	0.4686	0.81	0.3067	1.4114	0.81	0.7890	0.2647	0.0510
0.3111	59.0	767	0.4688	0.81	0.3069	1.4130	0.81	0.7890	0.2416	0.0510
0.3111	60.0	780	0.4685	0.81	0.3065	1.4206	0.81	0.7890	0.2278	0.0509
0.3111	61.0	793	0.4688	0.81	0.3069	1.4145	0.81	0.7890	0.2307	0.0513
0.3111	62.0	806	0.4690	0.81	0.3070	1.4681	0.81	0.7890	0.2437	0.0510
0.3111	63.0	819	0.4688	0.81	0.3068	1.4680	0.81	0.7890	0.2465	0.0510
0.3111	64.0	832	0.4681	0.81	0.3062	1.4670	0.81	0.7890	0.2565	0.0507
0.3111	65.0	845	0.4690	0.81	0.3069	1.4675	0.81	0.7890	0.2444	0.0510
0.3111	66.0	858	0.4688	0.81	0.3069	1.4673	0.81	0.7890	0.2433	0.0510
0.3111	67.0	871	0.4686	0.81	0.3066	1.4676	0.81	0.7890	0.2560	0.0507
0.3111	68.0	884	0.4684	0.81	0.3064	1.4667	0.81	0.7890	0.2496	0.0506
0.3111	69.0	897	0.4686	0.81	0.3066	1.4675	0.81	0.7890	0.2407	0.0507
0.3111	70.0	910	0.4689	0.81	0.3068	1.4679	0.81	0.7890	0.2502	0.0508
0.3111	71.0	923	0.4690	0.81	0.3071	1.4687	0.81	0.7890	0.2445	0.0507
0.3111	72.0	936	0.4688	0.81	0.3068	1.4678	0.81	0.7890	0.2500	0.0506
0.3111	73.0	949	0.4689	0.81	0.3068	1.4685	0.81	0.7890	0.2662	0.0510
0.3111	74.0	962	0.4687	0.81	0.3067	1.4679	0.81	0.7890	0.2496	0.0507
0.3111	75.0	975	0.4688	0.81	0.3067	1.4683	0.81	0.7890	0.2468	0.0508
0.3111	76.0	988	0.4688	0.81	0.3067	1.4676	0.81	0.7890	0.2511	0.0508
0.1126	77.0	1001	0.4689	0.81	0.3068	1.4672	0.81	0.7890	0.2365	0.0506
0.1126	78.0	1014	0.4688	0.81	0.3066	1.4681	0.81	0.7890	0.2507	0.0507
0.1126	79.0	1027	0.4688	0.81	0.3068	1.4680	0.81	0.7890	0.2498	0.0508
0.1126	80.0	1040	0.4689	0.81	0.3068	1.4676	0.81	0.7890	0.2497	0.0507
0.1126	81.0	1053	0.4690	0.81	0.3068	1.4682	0.81	0.7890	0.2338	0.0506
0.1126	82.0	1066	0.4686	0.81	0.3065	1.4682	0.81	0.7890	0.2541	0.0505
0.1126	83.0	1079	0.4689	0.815	0.3067	1.4675	0.815	0.7970	0.2503	0.0501
0.1126	84.0	1092	0.4687	0.815	0.3065	1.4676	0.815	0.7970	0.2567	0.0501
0.1126	85.0	1105	0.4689	0.81	0.3067	1.4680	0.81	0.7890	0.2678	0.0507
0.1126	86.0	1118	0.4689	0.815	0.3067	1.4684	0.815	0.7970	0.2566	0.0502
0.1126	87.0	1131	0.4687	0.815	0.3066	1.4672	0.815	0.7970	0.2529	0.0501
0.1126	88.0	1144	0.4689	0.815	0.3067	1.4680	0.815	0.7970	0.2569	0.0502
0.1126	89.0	1157	0.4688	0.815	0.3067	1.4678	0.815	0.7970	0.2527	0.0500
0.1126	90.0	1170	0.4689	0.815	0.3067	1.4681	0.815	0.7970	0.2527	0.0501
0.1126	91.0	1183	0.4688	0.815	0.3067	1.4683	0.815	0.7970	0.2527	0.0500
0.1126	92.0	1196	0.4688	0.815	0.3066	1.4675	0.815	0.7970	0.2528	0.0500
0.1126	93.0	1209	0.4689	0.815	0.3068	1.4680	0.815	0.7970	0.2527	0.0500
0.1126	94.0	1222	0.4688	0.815	0.3066	1.4678	0.815	0.7970	0.2440	0.0499
0.1126	95.0	1235	0.4688	0.815	0.3066	1.4677	0.815	0.7970	0.2440	0.0499
0.1126	96.0	1248	0.4688	0.815	0.3067	1.4681	0.815	0.7970	0.2528	0.0500
0.1126	97.0	1261	0.4688	0.815	0.3066	1.4679	0.815	0.7970	0.2440	0.0500
0.1126	98.0	1274	0.4689	0.815	0.3067	1.4680	0.815	0.7970	0.2440	0.0500
0.1126	99.0	1287	0.4689	0.815	0.3067	1.4679	0.815	0.7970	0.2440	0.0500
0.1126	100.0	1300	0.4688	0.815	0.3067	1.4679	0.815	0.7970	0.2440	0.0500

Framework versions

Transformers 4.26.1
Pytorch 1.13.1.post200
Datasets 2.9.0
Tokenizers 0.13.2

171-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