vit-small_tobacco3482_kd_MSE

This model is a fine-tuned version of WinKawaks/vit-small-patch16-224 on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.2131
Accuracy: 0.84
Brier Loss: 0.2974
Nll: 0.8913
F1 Micro: 0.8400
F1 Macro: 0.8190
Ece: 0.2456
Aurc: 0.0512

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: 128
eval_batch_size: 128
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	7	1.4711	0.21	0.8898	6.2752	0.2100	0.1403	0.2702	0.7673
No log	2.0	14	1.0769	0.41	0.8120	5.2446	0.41	0.2713	0.3253	0.5170
No log	3.0	21	0.7901	0.51	0.7057	2.6186	0.51	0.4114	0.3359	0.3162
No log	4.0	28	0.6044	0.61	0.5736	1.8428	0.61	0.4989	0.3358	0.1889
No log	5.0	35	0.4605	0.7	0.5009	1.3395	0.7	0.6120	0.3587	0.1321
No log	6.0	42	0.4484	0.73	0.4373	1.4781	0.7300	0.6394	0.2751	0.1150
No log	7.0	49	0.4406	0.765	0.4180	1.1081	0.765	0.7193	0.3066	0.0981
No log	8.0	56	0.3421	0.82	0.3575	0.9309	0.82	0.7764	0.2867	0.0703
No log	9.0	63	0.4201	0.75	0.3973	1.5859	0.75	0.7562	0.2618	0.1051
No log	10.0	70	0.4086	0.795	0.3775	1.2870	0.795	0.7701	0.3104	0.0691
No log	11.0	77	0.2867	0.82	0.3251	1.2141	0.82	0.7996	0.2511	0.0683
No log	12.0	84	0.2964	0.825	0.3233	1.0042	0.825	0.8028	0.2801	0.0538
No log	13.0	91	0.3010	0.81	0.3351	1.0085	0.81	0.7735	0.2678	0.0584
No log	14.0	98	0.2741	0.835	0.3194	1.0574	0.835	0.8127	0.2982	0.0542
No log	15.0	105	0.2524	0.845	0.3228	1.1162	0.845	0.8225	0.2911	0.0568
No log	16.0	112	0.2652	0.83	0.3154	0.8145	0.83	0.8130	0.2786	0.0516
No log	17.0	119	0.2478	0.83	0.3241	1.1158	0.83	0.8034	0.2776	0.0683
No log	18.0	126	0.2526	0.85	0.3112	1.0132	0.85	0.8324	0.2757	0.0517
No log	19.0	133	0.2423	0.855	0.3023	1.0623	0.855	0.8382	0.2727	0.0561
No log	20.0	140	0.2294	0.83	0.3112	1.1134	0.83	0.8139	0.2697	0.0703
No log	21.0	147	0.2380	0.835	0.3080	0.9961	0.835	0.8190	0.2841	0.0489
No log	22.0	154	0.2362	0.84	0.3034	0.9586	0.8400	0.8145	0.2626	0.0520
No log	23.0	161	0.2252	0.86	0.2946	1.1006	0.8600	0.8471	0.2830	0.0495
No log	24.0	168	0.2325	0.85	0.2985	0.9069	0.85	0.8288	0.2681	0.0533
No log	25.0	175	0.2335	0.825	0.3005	0.8930	0.825	0.8000	0.2640	0.0496
No log	26.0	182	0.2309	0.845	0.2984	1.0007	0.845	0.8308	0.2573	0.0536
No log	27.0	189	0.2265	0.835	0.3051	1.0092	0.835	0.8158	0.2626	0.0603
No log	28.0	196	0.2192	0.83	0.2977	1.0186	0.83	0.8019	0.2516	0.0572
No log	29.0	203	0.2276	0.83	0.3017	0.9407	0.83	0.8179	0.2553	0.0480
No log	30.0	210	0.2131	0.84	0.2992	0.9232	0.8400	0.8195	0.2541	0.0546
No log	31.0	217	0.2197	0.845	0.2998	0.9012	0.845	0.8301	0.2537	0.0569
No log	32.0	224	0.2138	0.85	0.2972	0.9117	0.85	0.8349	0.2777	0.0551
No log	33.0	231	0.2167	0.85	0.2969	1.0176	0.85	0.8390	0.2676	0.0535
No log	34.0	238	0.2114	0.84	0.2959	0.8912	0.8400	0.8190	0.2512	0.0514
No log	35.0	245	0.2145	0.845	0.2952	0.8960	0.845	0.8216	0.2638	0.0492
No log	36.0	252	0.2146	0.845	0.2960	0.9093	0.845	0.8301	0.2841	0.0519
No log	37.0	259	0.2157	0.845	0.2973	0.9043	0.845	0.8216	0.2614	0.0520
No log	38.0	266	0.2116	0.84	0.2949	0.8871	0.8400	0.8190	0.2639	0.0512
No log	39.0	273	0.2138	0.845	0.2963	0.9002	0.845	0.8301	0.2497	0.0512
No log	40.0	280	0.2129	0.84	0.2960	0.9731	0.8400	0.8190	0.2500	0.0511
No log	41.0	287	0.2139	0.845	0.2966	1.0111	0.845	0.8301	0.2750	0.0523
No log	42.0	294	0.2134	0.84	0.2959	0.9515	0.8400	0.8190	0.2577	0.0506
No log	43.0	301	0.2134	0.84	0.2972	0.9022	0.8400	0.8190	0.2538	0.0517
No log	44.0	308	0.2131	0.84	0.2966	0.9569	0.8400	0.8190	0.2683	0.0519
No log	45.0	315	0.2131	0.84	0.2965	0.8931	0.8400	0.8190	0.2504	0.0513
No log	46.0	322	0.2119	0.84	0.2963	0.8998	0.8400	0.8190	0.2535	0.0513
No log	47.0	329	0.2129	0.84	0.2973	0.9017	0.8400	0.8190	0.2527	0.0514
No log	48.0	336	0.2130	0.84	0.2971	0.8947	0.8400	0.8190	0.2520	0.0510
No log	49.0	343	0.2123	0.84	0.2972	0.9482	0.8400	0.8190	0.2583	0.0515
No log	50.0	350	0.2124	0.84	0.2970	0.9083	0.8400	0.8190	0.2604	0.0513
No log	51.0	357	0.2130	0.84	0.2974	0.8978	0.8400	0.8190	0.2446	0.0513
No log	52.0	364	0.2127	0.84	0.2975	0.8932	0.8400	0.8190	0.2457	0.0513
No log	53.0	371	0.2125	0.84	0.2972	0.8935	0.8400	0.8190	0.2508	0.0512
No log	54.0	378	0.2130	0.84	0.2975	0.8989	0.8400	0.8190	0.2551	0.0513
No log	55.0	385	0.2128	0.84	0.2972	0.8941	0.8400	0.8190	0.2448	0.0511
No log	56.0	392	0.2128	0.84	0.2974	0.8944	0.8400	0.8190	0.2459	0.0515
No log	57.0	399	0.2128	0.84	0.2973	0.8934	0.8400	0.8190	0.2517	0.0512
No log	58.0	406	0.2130	0.84	0.2973	0.8936	0.8400	0.8190	0.2448	0.0513
No log	59.0	413	0.2129	0.84	0.2973	0.8951	0.8400	0.8190	0.2383	0.0513
No log	60.0	420	0.2128	0.84	0.2972	0.8921	0.8400	0.8190	0.2519	0.0512
No log	61.0	427	0.2125	0.84	0.2974	0.8959	0.8400	0.8190	0.2518	0.0515
No log	62.0	434	0.2128	0.84	0.2973	0.8937	0.8400	0.8190	0.2385	0.0513
No log	63.0	441	0.2131	0.84	0.2974	0.8933	0.8400	0.8190	0.2551	0.0512
No log	64.0	448	0.2129	0.84	0.2974	0.8930	0.8400	0.8190	0.2388	0.0512
No log	65.0	455	0.2129	0.84	0.2973	0.8927	0.8400	0.8190	0.2447	0.0513
No log	66.0	462	0.2129	0.84	0.2974	0.8930	0.8400	0.8190	0.2385	0.0513
No log	67.0	469	0.2129	0.84	0.2974	0.8929	0.8400	0.8190	0.2458	0.0512
No log	68.0	476	0.2130	0.84	0.2975	0.8930	0.8400	0.8190	0.2455	0.0512
No log	69.0	483	0.2130	0.84	0.2973	0.8917	0.8400	0.8190	0.2459	0.0513
No log	70.0	490	0.2129	0.84	0.2973	0.8913	0.8400	0.8190	0.2520	0.0513
No log	71.0	497	0.2131	0.84	0.2974	0.8919	0.8400	0.8190	0.2519	0.0513
0.1234	72.0	504	0.2130	0.84	0.2973	0.8917	0.8400	0.8190	0.2457	0.0511
0.1234	73.0	511	0.2129	0.84	0.2974	0.8917	0.8400	0.8190	0.2455	0.0512
0.1234	74.0	518	0.2129	0.84	0.2974	0.8913	0.8400	0.8190	0.2455	0.0512
0.1234	75.0	525	0.2130	0.84	0.2973	0.8917	0.8400	0.8190	0.2519	0.0513
0.1234	76.0	532	0.2129	0.84	0.2974	0.8921	0.8400	0.8190	0.2455	0.0512
0.1234	77.0	539	0.2130	0.84	0.2973	0.8919	0.8400	0.8190	0.2455	0.0511
0.1234	78.0	546	0.2130	0.84	0.2973	0.8924	0.8400	0.8190	0.2455	0.0511
0.1234	79.0	553	0.2130	0.84	0.2974	0.8919	0.8400	0.8190	0.2456	0.0512
0.1234	80.0	560	0.2130	0.84	0.2973	0.8915	0.8400	0.8190	0.2515	0.0512
0.1234	81.0	567	0.2130	0.84	0.2973	0.8915	0.8400	0.8190	0.2456	0.0511
0.1234	82.0	574	0.2130	0.84	0.2974	0.8915	0.8400	0.8190	0.2456	0.0512
0.1234	83.0	581	0.2130	0.84	0.2973	0.8916	0.8400	0.8190	0.2516	0.0512
0.1234	84.0	588	0.2130	0.84	0.2974	0.8920	0.8400	0.8190	0.2456	0.0512
0.1234	85.0	595	0.2130	0.84	0.2974	0.8915	0.8400	0.8190	0.2456	0.0512
0.1234	86.0	602	0.2130	0.84	0.2974	0.8917	0.8400	0.8190	0.2456	0.0512
0.1234	87.0	609	0.2130	0.84	0.2974	0.8913	0.8400	0.8190	0.2517	0.0512
0.1234	88.0	616	0.2130	0.84	0.2973	0.8916	0.8400	0.8190	0.2456	0.0512
0.1234	89.0	623	0.2130	0.84	0.2974	0.8912	0.8400	0.8190	0.2456	0.0512
0.1234	90.0	630	0.2130	0.84	0.2973	0.8914	0.8400	0.8190	0.2517	0.0512
0.1234	91.0	637	0.2131	0.84	0.2974	0.8915	0.8400	0.8190	0.2456	0.0512
0.1234	92.0	644	0.2130	0.84	0.2973	0.8912	0.8400	0.8190	0.2456	0.0512
0.1234	93.0	651	0.2130	0.84	0.2974	0.8915	0.8400	0.8190	0.2456	0.0512
0.1234	94.0	658	0.2130	0.84	0.2973	0.8913	0.8400	0.8190	0.2456	0.0512
0.1234	95.0	665	0.2130	0.84	0.2973	0.8913	0.8400	0.8190	0.2456	0.0512
0.1234	96.0	672	0.2131	0.84	0.2974	0.8915	0.8400	0.8190	0.2456	0.0512
0.1234	97.0	679	0.2131	0.84	0.2973	0.8914	0.8400	0.8190	0.2517	0.0512
0.1234	98.0	686	0.2130	0.84	0.2974	0.8912	0.8400	0.8190	0.2456	0.0512
0.1234	99.0	693	0.2131	0.84	0.2974	0.8913	0.8400	0.8190	0.2456	0.0512
0.1234	100.0	700	0.2131	0.84	0.2974	0.8913	0.8400	0.8190	0.2456	0.0512

Framework versions

Transformers 4.26.1
Pytorch 1.13.1.post200
Datasets 2.9.0
Tokenizers 0.13.2

jordyvl/vit-small_tobacco3482_kd_MSE

vit-small_tobacco3482_kd_MSE

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions