171-tiny_tobacco3482_kd

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: 0.2409
Accuracy: 0.765
Brier Loss: 0.3934
Nll: 1.2941
F1 Micro: 0.765
F1 Macro: 0.6802
Ece: 0.2843
Aurc: 0.0716

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	1.3775	0.235	0.8906	7.4936	0.235	0.1630	0.2705	0.7664
No log	2.0	26	0.8552	0.38	0.8035	4.2966	0.38	0.2738	0.3086	0.4787
No log	3.0	39	0.6791	0.51	0.6994	3.1397	0.51	0.3754	0.3208	0.3181
No log	4.0	52	0.5747	0.59	0.6289	2.2673	0.59	0.4483	0.3491	0.2156
No log	5.0	65	0.4985	0.64	0.5620	2.3221	0.64	0.5323	0.3258	0.1637
No log	6.0	78	0.5387	0.665	0.5254	2.2659	0.665	0.5673	0.2991	0.1582
No log	7.0	91	0.5050	0.66	0.5100	2.4796	0.66	0.5360	0.2241	0.1675
No log	8.0	104	0.4247	0.685	0.4864	2.1512	0.685	0.6017	0.2848	0.1242
No log	9.0	117	0.5322	0.65	0.5101	1.9216	0.65	0.5658	0.2808	0.1775
No log	10.0	130	0.4128	0.705	0.4744	1.6466	0.705	0.6320	0.3126	0.1011
No log	11.0	143	0.4216	0.72	0.4621	2.3585	0.72	0.6365	0.2807	0.1181
No log	12.0	156	0.3925	0.72	0.4535	2.1047	0.72	0.6466	0.2976	0.0989
No log	13.0	169	0.3664	0.71	0.4406	1.6481	0.7100	0.6388	0.2789	0.1056
No log	14.0	182	0.4002	0.715	0.4396	1.8360	0.715	0.6437	0.2705	0.1096
No log	15.0	195	0.3324	0.72	0.4156	1.8256	0.72	0.6309	0.2857	0.0874
No log	16.0	208	0.3041	0.75	0.4277	1.5808	0.75	0.6746	0.3014	0.0843
No log	17.0	221	0.3178	0.78	0.4139	1.1784	0.78	0.7004	0.3016	0.0772
No log	18.0	234	0.2911	0.735	0.4210	1.4736	0.735	0.6754	0.2909	0.0844
No log	19.0	247	0.2988	0.76	0.4107	1.2700	0.76	0.6933	0.2904	0.0775
No log	20.0	260	0.2904	0.745	0.4215	1.4039	0.745	0.6686	0.2920	0.0852
No log	21.0	273	0.3022	0.77	0.4196	1.0212	0.7700	0.7040	0.3041	0.0714
No log	22.0	286	0.2748	0.73	0.4106	1.1826	0.7300	0.6715	0.2977	0.0854
No log	23.0	299	0.2835	0.745	0.4079	1.2464	0.745	0.6654	0.3083	0.0797
No log	24.0	312	0.2748	0.75	0.4089	1.1540	0.75	0.6797	0.2772	0.0802
No log	25.0	325	0.2818	0.735	0.4142	1.3465	0.735	0.6523	0.2693	0.0916
No log	26.0	338	0.2666	0.74	0.4076	1.3420	0.74	0.6560	0.2892	0.0831
No log	27.0	351	0.2693	0.745	0.4083	1.4070	0.745	0.6883	0.3074	0.0858
No log	28.0	364	0.2598	0.725	0.4007	1.3015	0.7250	0.6509	0.2874	0.0843
No log	29.0	377	0.2579	0.745	0.4023	1.2920	0.745	0.6770	0.2808	0.0788
No log	30.0	390	0.2606	0.745	0.4053	1.2203	0.745	0.6643	0.2838	0.0805
No log	31.0	403	0.2588	0.735	0.3982	1.3655	0.735	0.6743	0.2941	0.0852
No log	32.0	416	0.2524	0.74	0.3941	1.1515	0.74	0.6771	0.2735	0.0813
No log	33.0	429	0.2579	0.765	0.4002	1.3257	0.765	0.6993	0.2784	0.0753
No log	34.0	442	0.2448	0.775	0.3981	1.2289	0.775	0.7015	0.2923	0.0720
No log	35.0	455	0.2483	0.75	0.3987	1.2485	0.75	0.6645	0.2751	0.0751
No log	36.0	468	0.2417	0.765	0.3879	1.0562	0.765	0.6856	0.2827	0.0723
No log	37.0	481	0.2506	0.755	0.3944	1.2087	0.755	0.6855	0.3045	0.0744
No log	38.0	494	0.2427	0.765	0.3917	1.2356	0.765	0.6862	0.2822	0.0703
0.2351	39.0	507	0.2449	0.745	0.3958	1.2868	0.745	0.6750	0.2697	0.0762
0.2351	40.0	520	0.2413	0.755	0.3917	1.3279	0.755	0.6831	0.2720	0.0724
0.2351	41.0	533	0.2428	0.75	0.3924	1.2369	0.75	0.6781	0.2700	0.0766
0.2351	42.0	546	0.2412	0.76	0.3919	1.2235	0.76	0.6913	0.2945	0.0732
0.2351	43.0	559	0.2428	0.74	0.3968	1.3021	0.74	0.6648	0.2743	0.0773
0.2351	44.0	572	0.2400	0.75	0.3936	1.2410	0.75	0.6643	0.2789	0.0723
0.2351	45.0	585	0.2424	0.77	0.3949	1.2480	0.7700	0.7041	0.2813	0.0722
0.2351	46.0	598	0.2398	0.77	0.3931	1.2463	0.7700	0.7005	0.3050	0.0722
0.2351	47.0	611	0.2397	0.77	0.3919	1.2957	0.7700	0.6874	0.2961	0.0703
0.2351	48.0	624	0.2401	0.77	0.3926	1.2360	0.7700	0.7045	0.2945	0.0720
0.2351	49.0	637	0.2401	0.77	0.3927	1.2905	0.7700	0.6876	0.2825	0.0706
0.2351	50.0	650	0.2413	0.765	0.3936	1.2892	0.765	0.6978	0.3016	0.0743
0.2351	51.0	663	0.2410	0.77	0.3943	1.2913	0.7700	0.7005	0.2849	0.0728
0.2351	52.0	676	0.2391	0.765	0.3926	1.2846	0.765	0.6805	0.2777	0.0710
0.2351	53.0	689	0.2400	0.77	0.3929	1.2927	0.7700	0.6876	0.2698	0.0711
0.2351	54.0	702	0.2401	0.77	0.3929	1.2917	0.7700	0.6876	0.2775	0.0711
0.2351	55.0	715	0.2405	0.775	0.3934	1.2912	0.775	0.7074	0.2858	0.0709
0.2351	56.0	728	0.2403	0.775	0.3927	1.2912	0.775	0.7077	0.3059	0.0710
0.2351	57.0	741	0.2403	0.77	0.3932	1.2886	0.7700	0.6876	0.2914	0.0709
0.2351	58.0	754	0.2403	0.765	0.3932	1.2893	0.765	0.6805	0.2735	0.0717
0.2351	59.0	767	0.2405	0.77	0.3932	1.2923	0.7700	0.6876	0.2915	0.0714
0.2351	60.0	780	0.2403	0.77	0.3931	1.2893	0.7700	0.6874	0.2830	0.0708
0.2351	61.0	793	0.2405	0.77	0.3931	1.2910	0.7700	0.6876	0.2912	0.0711
0.2351	62.0	806	0.2405	0.77	0.3935	1.2905	0.7700	0.6874	0.2831	0.0711
0.2351	63.0	819	0.2405	0.765	0.3933	1.2918	0.765	0.6805	0.2776	0.0716
0.2351	64.0	832	0.2405	0.77	0.3932	1.2937	0.7700	0.6876	0.2747	0.0712
0.2351	65.0	845	0.2405	0.765	0.3934	1.2927	0.765	0.6802	0.2781	0.0717
0.2351	66.0	858	0.2406	0.77	0.3930	1.2926	0.7700	0.6876	0.2780	0.0708
0.2351	67.0	871	0.2407	0.77	0.3934	1.2933	0.7700	0.6876	0.2784	0.0712
0.2351	68.0	884	0.2407	0.77	0.3934	1.2942	0.7700	0.6876	0.2853	0.0714
0.2351	69.0	897	0.2406	0.77	0.3932	1.2920	0.7700	0.6876	0.2780	0.0709
0.2351	70.0	910	0.2403	0.77	0.3931	1.2937	0.7700	0.6874	0.2748	0.0709
0.2351	71.0	923	0.2407	0.77	0.3934	1.2929	0.7700	0.6874	0.2855	0.0710
0.2351	72.0	936	0.2405	0.77	0.3930	1.2944	0.7700	0.6874	0.2779	0.0708
0.2351	73.0	949	0.2407	0.765	0.3931	1.2919	0.765	0.6802	0.2799	0.0715
0.2351	74.0	962	0.2408	0.765	0.3933	1.2937	0.765	0.6802	0.2647	0.0716
0.2351	75.0	975	0.2407	0.765	0.3932	1.2935	0.765	0.6802	0.2728	0.0716
0.2351	76.0	988	0.2407	0.77	0.3933	1.2933	0.7700	0.6874	0.2773	0.0709
0.0004	77.0	1001	0.2407	0.77	0.3932	1.2941	0.7700	0.6874	0.2892	0.0709
0.0004	78.0	1014	0.2408	0.77	0.3933	1.2936	0.7700	0.6874	0.2820	0.0709
0.0004	79.0	1027	0.2410	0.77	0.3934	1.2931	0.7700	0.6874	0.2892	0.0710
0.0004	80.0	1040	0.2409	0.77	0.3933	1.2929	0.7700	0.6874	0.2855	0.0711
0.0004	81.0	1053	0.2409	0.77	0.3933	1.2937	0.7700	0.6874	0.2820	0.0709
0.0004	82.0	1066	0.2409	0.77	0.3934	1.2947	0.7700	0.6874	0.2819	0.0709
0.0004	83.0	1079	0.2408	0.77	0.3934	1.2933	0.7700	0.6874	0.2893	0.0709
0.0004	84.0	1092	0.2409	0.765	0.3934	1.2934	0.765	0.6802	0.2843	0.0716
0.0004	85.0	1105	0.2408	0.77	0.3933	1.2933	0.7700	0.6874	0.2893	0.0710
0.0004	86.0	1118	0.2409	0.77	0.3933	1.2940	0.7700	0.6874	0.2819	0.0710
0.0004	87.0	1131	0.2409	0.77	0.3934	1.2944	0.7700	0.6874	0.2820	0.0709
0.0004	88.0	1144	0.2409	0.77	0.3934	1.2936	0.7700	0.6874	0.2893	0.0709
0.0004	89.0	1157	0.2409	0.765	0.3933	1.2936	0.765	0.6802	0.2844	0.0717
0.0004	90.0	1170	0.2409	0.77	0.3934	1.2939	0.7700	0.6874	0.2893	0.0709
0.0004	91.0	1183	0.2409	0.765	0.3934	1.2943	0.765	0.6802	0.2843	0.0716
0.0004	92.0	1196	0.2409	0.77	0.3934	1.2942	0.7700	0.6874	0.2893	0.0709
0.0004	93.0	1209	0.2410	0.765	0.3934	1.2939	0.765	0.6802	0.2843	0.0716
0.0004	94.0	1222	0.2409	0.765	0.3934	1.2937	0.765	0.6802	0.2843	0.0716
0.0004	95.0	1235	0.2409	0.765	0.3934	1.2938	0.765	0.6802	0.2844	0.0716
0.0004	96.0	1248	0.2409	0.765	0.3934	1.2939	0.765	0.6802	0.2770	0.0716
0.0004	97.0	1261	0.2409	0.765	0.3934	1.2941	0.765	0.6802	0.2843	0.0715
0.0004	98.0	1274	0.2409	0.765	0.3934	1.2940	0.765	0.6802	0.2844	0.0715
0.0004	99.0	1287	0.2409	0.765	0.3934	1.2941	0.765	0.6802	0.2843	0.0716
0.0004	100.0	1300	0.2409	0.765	0.3934	1.2941	0.765	0.6802	0.2843	0.0716

Framework versions

Transformers 4.26.1
Pytorch 1.13.1.post200
Datasets 2.9.0
Tokenizers 0.13.2

jordyvl/171-tiny_tobacco3482_kd

171-tiny_tobacco3482_kd

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions