225-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: 4.3111
Accuracy: 0.82
Brier Loss: 0.2977
Nll: 1.6959
F1 Micro: 0.82
F1 Macro: 0.8150
Ece: 0.1454
Aurc: 0.0488

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	5.2368	0.225	0.8876	8.2751	0.225	0.1306	0.3140	0.7919
No log	2.0	26	4.6617	0.385	0.7736	4.0165	0.3850	0.3071	0.3295	0.4033
No log	3.0	39	4.4343	0.525	0.6609	3.4855	0.525	0.4017	0.3068	0.2761
No log	4.0	52	4.2677	0.59	0.5775	2.7458	0.59	0.4879	0.3037	0.1850
No log	5.0	65	4.1495	0.67	0.5044	2.2848	0.67	0.6081	0.3100	0.1336
No log	6.0	78	4.1699	0.71	0.4412	2.9360	0.7100	0.6211	0.2407	0.1076
No log	7.0	91	4.0527	0.725	0.4198	2.1169	0.7250	0.6606	0.2359	0.0993
No log	8.0	104	4.0491	0.715	0.4001	2.1794	0.715	0.6343	0.1955	0.1013
No log	9.0	117	4.2070	0.715	0.4096	2.1137	0.715	0.6363	0.1968	0.1104
No log	10.0	130	4.2307	0.715	0.4030	2.4228	0.715	0.6467	0.1977	0.1054
No log	11.0	143	4.0841	0.73	0.3673	2.2764	0.7300	0.6697	0.1840	0.0781
No log	12.0	156	3.9980	0.74	0.3569	1.7264	0.74	0.6752	0.1822	0.0779
No log	13.0	169	4.0921	0.735	0.3704	1.8601	0.735	0.6818	0.1835	0.0888
No log	14.0	182	3.9026	0.755	0.3362	1.6596	0.755	0.7128	0.1684	0.0757
No log	15.0	195	4.0542	0.765	0.3472	2.0096	0.765	0.7051	0.1789	0.0783
No log	16.0	208	4.0180	0.75	0.3634	1.6543	0.75	0.7364	0.1958	0.0890
No log	17.0	221	3.9665	0.8	0.3330	1.4940	0.8000	0.7935	0.1919	0.0793
No log	18.0	234	3.9523	0.785	0.3225	1.6353	0.785	0.7825	0.1598	0.0719
No log	19.0	247	3.9298	0.79	0.3262	1.8606	0.79	0.7757	0.1785	0.0749
No log	20.0	260	3.9484	0.8	0.3106	1.6615	0.8000	0.8034	0.1692	0.0763
No log	21.0	273	3.9056	0.785	0.2930	1.6180	0.785	0.7499	0.1542	0.0609
No log	22.0	286	3.8094	0.82	0.2765	1.3116	0.82	0.8028	0.1784	0.0532
No log	23.0	299	3.8352	0.81	0.2939	1.5765	0.81	0.7971	0.1592	0.0559
No log	24.0	312	3.9996	0.79	0.3192	1.6863	0.79	0.7914	0.1678	0.0742
No log	25.0	325	3.8680	0.805	0.2932	1.4217	0.805	0.8052	0.1505	0.0578
No log	26.0	338	3.8913	0.8	0.3025	1.6254	0.8000	0.7971	0.1370	0.0607
No log	27.0	351	3.8603	0.815	0.2893	1.6578	0.815	0.8094	0.1659	0.0570
No log	28.0	364	3.9414	0.795	0.2990	1.9161	0.795	0.7900	0.1504	0.0593
No log	29.0	377	3.8802	0.815	0.2836	1.7091	0.815	0.7943	0.1395	0.0565
No log	30.0	390	3.9025	0.8	0.2957	1.7376	0.8000	0.7894	0.1373	0.0594
No log	31.0	403	3.8744	0.835	0.2785	1.5096	0.835	0.8185	0.1405	0.0550
No log	32.0	416	3.8670	0.8	0.2813	1.5817	0.8000	0.7825	0.1279	0.0500
No log	33.0	429	3.9197	0.8	0.2852	1.5082	0.8000	0.7802	0.1488	0.0540
No log	34.0	442	3.9589	0.795	0.3005	1.9897	0.795	0.7872	0.1487	0.0563
No log	35.0	455	3.9669	0.82	0.2863	1.7012	0.82	0.8161	0.1483	0.0551
No log	36.0	468	3.8924	0.81	0.2803	1.5552	0.81	0.7961	0.1322	0.0484
No log	37.0	481	3.9455	0.81	0.2838	1.6590	0.81	0.7989	0.1423	0.0531
No log	38.0	494	3.8957	0.82	0.2726	1.5431	0.82	0.8072	0.1409	0.0482
3.5636	39.0	507	3.9710	0.81	0.2979	1.7156	0.81	0.7989	0.1399	0.0524
3.5636	40.0	520	3.8789	0.83	0.2606	1.5452	0.83	0.8227	0.1323	0.0478
3.5636	41.0	533	3.9488	0.81	0.2839	1.6447	0.81	0.8016	0.1326	0.0509
3.5636	42.0	546	3.9774	0.815	0.2937	1.6907	0.815	0.8111	0.1291	0.0488
3.5636	43.0	559	3.9991	0.805	0.2877	1.7106	0.805	0.7979	0.1504	0.0518
3.5636	44.0	572	3.9634	0.815	0.2798	1.5063	0.815	0.8048	0.1272	0.0493
3.5636	45.0	585	4.0229	0.82	0.2904	1.6439	0.82	0.8156	0.1392	0.0511
3.5636	46.0	598	4.0206	0.82	0.2836	1.5407	0.82	0.8150	0.1233	0.0497
3.5636	47.0	611	4.0351	0.81	0.2835	1.7627	0.81	0.8003	0.1338	0.0486
3.5636	48.0	624	4.0646	0.82	0.2889	1.7694	0.82	0.8150	0.1341	0.0499
3.5636	49.0	637	4.0496	0.815	0.2828	1.7548	0.815	0.8071	0.1391	0.0477
3.5636	50.0	650	4.0914	0.815	0.2917	1.6381	0.815	0.8053	0.1310	0.0502
3.5636	51.0	663	4.0748	0.82	0.2866	1.5646	0.82	0.8148	0.1325	0.0483
3.5636	52.0	676	4.0921	0.82	0.2871	1.5732	0.82	0.8148	0.1381	0.0487
3.5636	53.0	689	4.1093	0.82	0.2886	1.6448	0.82	0.8147	0.1506	0.0481
3.5636	54.0	702	4.1200	0.82	0.2910	1.6446	0.82	0.8150	0.1335	0.0493
3.5636	55.0	715	4.1250	0.815	0.2901	1.5641	0.815	0.8098	0.1386	0.0491
3.5636	56.0	728	4.1340	0.82	0.2893	1.6575	0.82	0.8148	0.1298	0.0489
3.5636	57.0	741	4.1575	0.82	0.2935	1.6360	0.82	0.8150	0.1402	0.0499
3.5636	58.0	754	4.1495	0.82	0.2895	1.6349	0.82	0.8148	0.1398	0.0486
3.5636	59.0	767	4.1582	0.82	0.2909	1.6327	0.82	0.8150	0.1341	0.0487
3.5636	60.0	780	4.1720	0.82	0.2923	1.5746	0.82	0.8150	0.1386	0.0493
3.5636	61.0	793	4.1848	0.825	0.2940	1.6424	0.825	0.8181	0.1380	0.0494
3.5636	62.0	806	4.1880	0.82	0.2939	1.6323	0.82	0.8148	0.1389	0.0488
3.5636	63.0	819	4.1825	0.82	0.2916	1.6920	0.82	0.8150	0.1421	0.0483
3.5636	64.0	832	4.2037	0.82	0.2946	1.6365	0.82	0.8148	0.1393	0.0493
3.5636	65.0	845	4.2096	0.82	0.2948	1.5852	0.82	0.8150	0.1462	0.0493
3.5636	66.0	858	4.2191	0.82	0.2962	1.6349	0.82	0.8150	0.1491	0.0495
3.5636	67.0	871	4.2189	0.82	0.2948	1.6389	0.82	0.8150	0.1313	0.0489
3.5636	68.0	884	4.2243	0.82	0.2947	1.6322	0.82	0.8150	0.1398	0.0491
3.5636	69.0	897	4.2334	0.82	0.2957	1.6398	0.82	0.8150	0.1355	0.0491
3.5636	70.0	910	4.2312	0.82	0.2943	1.6395	0.82	0.8148	0.1419	0.0484
3.5636	71.0	923	4.2376	0.82	0.2956	1.6389	0.82	0.8150	0.1372	0.0490
3.5636	72.0	936	4.2420	0.82	0.2951	1.6368	0.82	0.8150	0.1427	0.0489
3.5636	73.0	949	4.2464	0.82	0.2946	1.6375	0.82	0.8150	0.1449	0.0488
3.5636	74.0	962	4.2540	0.82	0.2956	1.6364	0.82	0.8150	0.1476	0.0489
3.5636	75.0	975	4.2579	0.82	0.2955	1.6361	0.82	0.8150	0.1361	0.0491
3.5636	76.0	988	4.2638	0.82	0.2960	1.6368	0.82	0.8150	0.1483	0.0490
3.1969	77.0	1001	4.2653	0.82	0.2956	1.6950	0.82	0.8150	0.1509	0.0487
3.1969	78.0	1014	4.2708	0.82	0.2965	1.6365	0.82	0.8150	0.1398	0.0490
3.1969	79.0	1027	4.2761	0.82	0.2968	1.6400	0.82	0.8150	0.1399	0.0490
3.1969	80.0	1040	4.2792	0.82	0.2969	1.6381	0.82	0.8150	0.1425	0.0490
3.1969	81.0	1053	4.2801	0.82	0.2963	1.6949	0.82	0.8148	0.1477	0.0487
3.1969	82.0	1066	4.2841	0.82	0.2968	1.6459	0.82	0.8150	0.1425	0.0488
3.1969	83.0	1079	4.2864	0.82	0.2968	1.6378	0.82	0.8150	0.1421	0.0489
3.1969	84.0	1092	4.2918	0.82	0.2973	1.6398	0.82	0.8150	0.1373	0.0491
3.1969	85.0	1105	4.2930	0.82	0.2970	1.6408	0.82	0.8150	0.1486	0.0490
3.1969	86.0	1118	4.2956	0.82	0.2973	1.6420	0.82	0.8150	0.1427	0.0489
3.1969	87.0	1131	4.2988	0.82	0.2976	1.6390	0.82	0.8150	0.1374	0.0491
3.1969	88.0	1144	4.2995	0.82	0.2974	1.6509	0.82	0.8150	0.1427	0.0489
3.1969	89.0	1157	4.3026	0.82	0.2976	1.6418	0.82	0.8150	0.1375	0.0490
3.1969	90.0	1170	4.3028	0.82	0.2974	1.6445	0.82	0.8150	0.1453	0.0488
3.1969	91.0	1183	4.3054	0.82	0.2976	1.6443	0.82	0.8150	0.1402	0.0488
3.1969	92.0	1196	4.3060	0.82	0.2975	1.6530	0.82	0.8150	0.1454	0.0488
3.1969	93.0	1209	4.3074	0.82	0.2975	1.6961	0.82	0.8150	0.1453	0.0488
3.1969	94.0	1222	4.3078	0.82	0.2975	1.6638	0.82	0.8150	0.1454	0.0488
3.1969	95.0	1235	4.3092	0.82	0.2976	1.6959	0.82	0.8150	0.1454	0.0488
3.1969	96.0	1248	4.3094	0.82	0.2976	1.6957	0.82	0.8150	0.1454	0.0488
3.1969	97.0	1261	4.3106	0.82	0.2977	1.6961	0.82	0.8150	0.1455	0.0489
3.1969	98.0	1274	4.3110	0.82	0.2977	1.6960	0.82	0.8150	0.1455	0.0488
3.1969	99.0	1287	4.3111	0.82	0.2977	1.6959	0.82	0.8150	0.1454	0.0488
3.1969	100.0	1300	4.3111	0.82	0.2977	1.6959	0.82	0.8150	0.1454	0.0488

Framework versions

Transformers 4.26.1
Pytorch 1.13.1.post200
Datasets 2.9.0
Tokenizers 0.13.2

225-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