114-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: 0.5018
Accuracy: 0.81
Brier Loss: 0.3510
Nll: 1.3312
F1 Micro: 0.81
F1 Macro: 0.7895
Ece: 0.2880
Aurc: 0.0537

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	1.4179	0.225	0.8892	8.1717	0.225	0.1452	0.3025	0.7575
No log	2.0	26	1.0222	0.44	0.7148	3.7941	0.44	0.3487	0.2946	0.3460
No log	3.0	39	0.8765	0.57	0.5993	2.4902	0.57	0.4787	0.2800	0.2411
No log	4.0	52	0.7663	0.645	0.5127	1.8672	0.645	0.5731	0.2922	0.1395
No log	5.0	65	0.7017	0.695	0.4672	1.3663	0.695	0.6567	0.3016	0.1153
No log	6.0	78	0.6744	0.725	0.4267	1.5639	0.7250	0.6508	0.2743	0.1066
No log	7.0	91	0.6457	0.735	0.4093	1.5745	0.735	0.7161	0.2659	0.0897
No log	8.0	104	0.6701	0.715	0.4180	1.7405	0.715	0.6768	0.2274	0.1123
No log	9.0	117	0.6394	0.735	0.4147	1.5335	0.735	0.6938	0.2380	0.1089
No log	10.0	130	0.6231	0.795	0.4000	1.4767	0.795	0.7795	0.3033	0.0702
No log	11.0	143	0.6335	0.74	0.3955	2.0438	0.74	0.7274	0.2632	0.0900
No log	12.0	156	0.5905	0.745	0.3898	1.6288	0.745	0.7078	0.2524	0.0808
No log	13.0	169	0.6359	0.72	0.4279	2.0889	0.72	0.7190	0.2676	0.1137
No log	14.0	182	0.5462	0.78	0.3627	1.5296	0.78	0.7550	0.2957	0.0678
No log	15.0	195	0.5639	0.76	0.3913	1.6057	0.76	0.7296	0.2769	0.0763
No log	16.0	208	0.5731	0.77	0.3957	1.7224	0.7700	0.7444	0.3190	0.0661
No log	17.0	221	0.5561	0.745	0.3842	1.3096	0.745	0.7130	0.2562	0.0784
No log	18.0	234	0.5559	0.755	0.3880	1.5937	0.755	0.7263	0.2810	0.0798
No log	19.0	247	0.5454	0.8	0.3824	1.4121	0.8000	0.7875	0.3284	0.0595
No log	20.0	260	0.5327	0.78	0.3638	1.4363	0.78	0.7462	0.2774	0.0633
No log	21.0	273	0.5291	0.775	0.3596	1.5236	0.775	0.7470	0.2902	0.0686
No log	22.0	286	0.5175	0.79	0.3597	1.2045	0.79	0.7583	0.3013	0.0547
No log	23.0	299	0.5160	0.83	0.3560	1.4152	0.83	0.8028	0.3339	0.0545
No log	24.0	312	0.5473	0.79	0.3736	1.5232	0.79	0.7719	0.2815	0.0842
No log	25.0	325	0.5270	0.805	0.3685	1.5013	0.805	0.7801	0.2881	0.0606
No log	26.0	338	0.5032	0.815	0.3444	1.2979	0.815	0.7913	0.2776	0.0544
No log	27.0	351	0.5245	0.79	0.3607	1.4567	0.79	0.7767	0.2932	0.0658
No log	28.0	364	0.5192	0.775	0.3636	1.3492	0.775	0.7445	0.2730	0.0627
No log	29.0	377	0.5219	0.79	0.3595	1.3888	0.79	0.7590	0.2956	0.0627
No log	30.0	390	0.5138	0.82	0.3548	1.4869	0.82	0.7911	0.3138	0.0579
No log	31.0	403	0.5110	0.8	0.3523	1.3410	0.8000	0.7800	0.3211	0.0535
No log	32.0	416	0.5059	0.825	0.3517	1.3191	0.825	0.8090	0.3256	0.0499
No log	33.0	429	0.5060	0.805	0.3487	1.4018	0.805	0.7850	0.2882	0.0568
No log	34.0	442	0.5063	0.79	0.3558	1.4104	0.79	0.7607	0.2863	0.0593
No log	35.0	455	0.4944	0.81	0.3426	1.3328	0.81	0.7918	0.2916	0.0465
No log	36.0	468	0.4999	0.815	0.3476	1.3893	0.815	0.7973	0.2954	0.0534
No log	37.0	481	0.5014	0.8	0.3487	1.3948	0.8000	0.7800	0.3012	0.0574
No log	38.0	494	0.5013	0.81	0.3500	1.3885	0.81	0.7896	0.2923	0.0537
0.3577	39.0	507	0.5016	0.81	0.3495	1.3905	0.81	0.7927	0.3037	0.0565
0.3577	40.0	520	0.4999	0.805	0.3491	1.3888	0.805	0.7850	0.2920	0.0547
0.3577	41.0	533	0.5014	0.805	0.3509	1.3877	0.805	0.7878	0.3036	0.0546
0.3577	42.0	546	0.4997	0.805	0.3487	1.3352	0.805	0.7850	0.2921	0.0537
0.3577	43.0	559	0.5012	0.81	0.3508	1.3924	0.81	0.7882	0.3171	0.0542
0.3577	44.0	572	0.5014	0.805	0.3504	1.3910	0.805	0.7878	0.2963	0.0545
0.3577	45.0	585	0.5008	0.805	0.3500	1.3281	0.805	0.7878	0.2871	0.0539
0.3577	46.0	598	0.5014	0.805	0.3512	1.3288	0.805	0.7850	0.3028	0.0542
0.3577	47.0	611	0.5005	0.805	0.3502	1.3282	0.805	0.7864	0.3115	0.0531
0.3577	48.0	624	0.5010	0.805	0.3510	1.3280	0.805	0.7878	0.3022	0.0542
0.3577	49.0	637	0.5014	0.805	0.3502	1.3334	0.805	0.7864	0.3023	0.0543
0.3577	50.0	650	0.5016	0.805	0.3509	1.3344	0.805	0.7878	0.3095	0.0547
0.3577	51.0	663	0.5015	0.805	0.3507	1.3325	0.805	0.7878	0.3089	0.0543
0.3577	52.0	676	0.5009	0.81	0.3505	1.3292	0.81	0.7908	0.3082	0.0543
0.3577	53.0	689	0.5012	0.81	0.3506	1.3313	0.81	0.7923	0.3008	0.0539
0.3577	54.0	702	0.5012	0.805	0.3506	1.3284	0.805	0.7892	0.2850	0.0539
0.3577	55.0	715	0.5012	0.81	0.3506	1.3288	0.81	0.7894	0.2940	0.0537
0.3577	56.0	728	0.5017	0.805	0.3508	1.3284	0.805	0.7892	0.2967	0.0544
0.3577	57.0	741	0.5017	0.81	0.3510	1.3316	0.81	0.7895	0.2838	0.0542
0.3577	58.0	754	0.5013	0.815	0.3508	1.3299	0.815	0.7940	0.2792	0.0539
0.3577	59.0	767	0.5015	0.81	0.3508	1.3326	0.81	0.7910	0.2894	0.0538
0.3577	60.0	780	0.5014	0.81	0.3505	1.3305	0.81	0.7910	0.2891	0.0540
0.3577	61.0	793	0.5016	0.81	0.3508	1.3305	0.81	0.7910	0.3138	0.0538
0.3577	62.0	806	0.5015	0.815	0.3507	1.3329	0.815	0.7952	0.2835	0.0538
0.3577	63.0	819	0.5017	0.81	0.3509	1.3288	0.81	0.7895	0.2841	0.0542
0.3577	64.0	832	0.5018	0.81	0.3511	1.3311	0.81	0.7895	0.2908	0.0540
0.3577	65.0	845	0.5015	0.815	0.3509	1.3310	0.815	0.7940	0.3124	0.0539
0.3577	66.0	858	0.5014	0.81	0.3507	1.3273	0.81	0.7923	0.3039	0.0536
0.3577	67.0	871	0.5014	0.81	0.3507	1.3287	0.81	0.7895	0.2907	0.0538
0.3577	68.0	884	0.5016	0.81	0.3509	1.3304	0.81	0.7910	0.3043	0.0540
0.3577	69.0	897	0.5017	0.81	0.3509	1.3316	0.81	0.7923	0.3041	0.0539
0.3577	70.0	910	0.5017	0.81	0.3510	1.3307	0.81	0.7882	0.2986	0.0538
0.3577	71.0	923	0.5019	0.81	0.3511	1.3315	0.81	0.7923	0.3047	0.0540
0.3577	72.0	936	0.5016	0.81	0.3510	1.3305	0.81	0.7910	0.3044	0.0539
0.3577	73.0	949	0.5017	0.81	0.3508	1.3308	0.81	0.7895	0.2951	0.0540
0.3577	74.0	962	0.5015	0.815	0.3509	1.3309	0.815	0.7952	0.3046	0.0537
0.3577	75.0	975	0.5018	0.81	0.3510	1.3308	0.81	0.7923	0.3099	0.0542
0.3577	76.0	988	0.5019	0.81	0.3511	1.3311	0.81	0.7923	0.3045	0.0543
0.197	77.0	1001	0.5017	0.81	0.3510	1.3308	0.81	0.7882	0.3045	0.0538
0.197	78.0	1014	0.5017	0.81	0.3509	1.3308	0.81	0.7923	0.2974	0.0538
0.197	79.0	1027	0.5018	0.81	0.3510	1.3313	0.81	0.7882	0.3028	0.0540
0.197	80.0	1040	0.5018	0.81	0.3511	1.3310	0.81	0.7910	0.3030	0.0539
0.197	81.0	1053	0.5019	0.815	0.3511	1.3307	0.815	0.7952	0.2952	0.0540
0.197	82.0	1066	0.5015	0.81	0.3507	1.3314	0.81	0.7882	0.2948	0.0535
0.197	83.0	1079	0.5018	0.81	0.3510	1.3305	0.81	0.7895	0.2802	0.0538
0.197	84.0	1092	0.5018	0.81	0.3510	1.3310	0.81	0.7882	0.2951	0.0539
0.197	85.0	1105	0.5018	0.81	0.3510	1.3314	0.81	0.7882	0.2879	0.0539
0.197	86.0	1118	0.5018	0.81	0.3510	1.3318	0.81	0.7910	0.3028	0.0539
0.197	87.0	1131	0.5018	0.81	0.3510	1.3311	0.81	0.7895	0.2802	0.0538
0.197	88.0	1144	0.5018	0.81	0.3511	1.3315	0.81	0.7895	0.2880	0.0539
0.197	89.0	1157	0.5018	0.81	0.3510	1.3310	0.81	0.7895	0.2880	0.0539
0.197	90.0	1170	0.5019	0.81	0.3511	1.3314	0.81	0.7923	0.2881	0.0538
0.197	91.0	1183	0.5018	0.81	0.3510	1.3316	0.81	0.7895	0.2880	0.0537
0.197	92.0	1196	0.5018	0.81	0.3510	1.3311	0.81	0.7882	0.2880	0.0538
0.197	93.0	1209	0.5019	0.81	0.3511	1.3315	0.81	0.7895	0.2935	0.0539
0.197	94.0	1222	0.5018	0.81	0.3510	1.3313	0.81	0.7895	0.2803	0.0537
0.197	95.0	1235	0.5017	0.81	0.3510	1.3312	0.81	0.7882	0.2880	0.0537
0.197	96.0	1248	0.5018	0.81	0.3510	1.3315	0.81	0.7895	0.2935	0.0538
0.197	97.0	1261	0.5018	0.81	0.3510	1.3313	0.81	0.7895	0.2880	0.0537
0.197	98.0	1274	0.5018	0.81	0.3510	1.3314	0.81	0.7895	0.2935	0.0538
0.197	99.0	1287	0.5018	0.81	0.3510	1.3313	0.81	0.7895	0.2880	0.0537
0.197	100.0	1300	0.5018	0.81	0.3510	1.3312	0.81	0.7895	0.2880	0.0537

Framework versions

Transformers 4.26.1
Pytorch 1.13.1.post200
Datasets 2.9.0
Tokenizers 0.13.2

114-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