vit-small_tobacco3482_hint_

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: 56.9670
Accuracy: 0.835
Brier Loss: 0.2969
Nll: 1.1900
F1 Micro: 0.835
F1 Macro: 0.8377
Ece: 0.1545
Aurc: 0.0499

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: 32
eval_batch_size: 32
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	25	59.9015	0.37	0.7765	4.3056	0.37	0.2185	0.2975	0.4588
No log	2.0	50	58.9173	0.66	0.4866	2.0758	0.66	0.5717	0.2732	0.1578
No log	3.0	75	58.3604	0.745	0.3466	1.5077	0.745	0.7135	0.1846	0.0854
No log	4.0	100	58.0585	0.75	0.3628	1.5044	0.75	0.7674	0.2058	0.1052
No log	5.0	125	57.8363	0.76	0.3782	1.7066	0.76	0.7657	0.2174	0.1039
No log	6.0	150	57.4894	0.75	0.3593	1.5137	0.75	0.7377	0.1724	0.0800
No log	7.0	175	57.5188	0.76	0.3631	1.9770	0.76	0.7514	0.1968	0.0874
No log	8.0	200	57.4349	0.74	0.3947	1.8766	0.74	0.7412	0.1753	0.0777
No log	9.0	225	57.1764	0.765	0.3481	1.1532	0.765	0.7411	0.1956	0.0829
No log	10.0	250	57.6192	0.765	0.3943	1.8998	0.765	0.7755	0.1850	0.0981
No log	11.0	275	57.2121	0.77	0.3531	1.2685	0.7700	0.7643	0.1739	0.0689
No log	12.0	300	57.2250	0.795	0.3279	1.7553	0.795	0.7816	0.1596	0.0660
No log	13.0	325	57.4911	0.785	0.3678	2.0499	0.785	0.7857	0.1788	0.0945
No log	14.0	350	57.1481	0.77	0.3542	1.4834	0.7700	0.7649	0.1892	0.0636
No log	15.0	375	57.1701	0.825	0.3041	1.6075	0.825	0.8223	0.1609	0.0621
No log	16.0	400	57.4059	0.805	0.3343	1.7348	0.805	0.8080	0.1654	0.0822
No log	17.0	425	57.9813	0.72	0.4616	2.6345	0.72	0.7252	0.2263	0.1101
No log	18.0	450	57.2677	0.825	0.2953	1.5836	0.825	0.8171	0.1590	0.0572
No log	19.0	475	57.6052	0.765	0.4023	1.7463	0.765	0.7333	0.2052	0.0822
57.2084	20.0	500	57.4249	0.79	0.3653	1.5564	0.79	0.7941	0.1818	0.0845
57.2084	21.0	525	57.2631	0.845	0.2704	1.7326	0.845	0.8312	0.1358	0.0628
57.2084	22.0	550	57.1520	0.845	0.2723	1.2743	0.845	0.8386	0.1402	0.0551
57.2084	23.0	575	57.2977	0.82	0.3137	1.3068	0.82	0.8029	0.1578	0.0621
57.2084	24.0	600	57.2030	0.81	0.3107	1.5814	0.81	0.7870	0.1594	0.0688
57.2084	25.0	625	57.1500	0.82	0.3027	1.4128	0.82	0.8229	0.1584	0.0436
57.2084	26.0	650	57.1619	0.855	0.2735	1.5164	0.855	0.8558	0.1404	0.0530
57.2084	27.0	675	57.1504	0.845	0.2832	1.5742	0.845	0.8507	0.1500	0.0516
57.2084	28.0	700	57.1829	0.835	0.2932	1.4010	0.835	0.8410	0.1489	0.0496
57.2084	29.0	725	57.1899	0.83	0.2953	1.4038	0.83	0.8338	0.1497	0.0511
57.2084	30.0	750	57.1644	0.835	0.2948	1.3923	0.835	0.8374	0.1509	0.0507
57.2084	31.0	775	57.1720	0.83	0.2958	1.4622	0.83	0.8296	0.1502	0.0509
57.2084	32.0	800	57.1365	0.835	0.3024	1.2976	0.835	0.8374	0.1575	0.0509
57.2084	33.0	825	57.1499	0.835	0.2995	1.3654	0.835	0.8308	0.1574	0.0523
57.2084	34.0	850	57.1064	0.83	0.3022	1.3606	0.83	0.8251	0.1578	0.0526
57.2084	35.0	875	57.0901	0.835	0.3003	1.2803	0.835	0.8336	0.1554	0.0516
57.2084	36.0	900	57.0922	0.835	0.3047	1.2749	0.835	0.8336	0.1571	0.0517
57.2084	37.0	925	57.0673	0.83	0.3034	1.2533	0.83	0.8344	0.1559	0.0509
57.2084	38.0	950	57.0810	0.83	0.3024	1.2718	0.83	0.8344	0.1620	0.0526
57.2084	39.0	975	57.1040	0.835	0.3041	1.2522	0.835	0.8392	0.1571	0.0506
56.1387	40.0	1000	57.0542	0.835	0.3024	1.3210	0.835	0.8392	0.1525	0.0501
56.1387	41.0	1025	57.0554	0.83	0.3037	1.3231	0.83	0.8344	0.1534	0.0508
56.1387	42.0	1050	57.0724	0.83	0.2989	1.2517	0.83	0.8344	0.1485	0.0495
56.1387	43.0	1075	57.0429	0.835	0.3010	1.3082	0.835	0.8401	0.1557	0.0506
56.1387	44.0	1100	57.0208	0.835	0.3001	1.2428	0.835	0.8392	0.1583	0.0496
56.1387	45.0	1125	57.0700	0.835	0.2996	1.3149	0.835	0.8454	0.1601	0.0509
56.1387	46.0	1150	57.0054	0.835	0.2950	1.3019	0.835	0.8407	0.1476	0.0492
56.1387	47.0	1175	57.0516	0.825	0.3000	1.2344	0.825	0.8317	0.1485	0.0511
56.1387	48.0	1200	57.0373	0.835	0.3008	1.3016	0.835	0.8434	0.1611	0.0498
56.1387	49.0	1225	57.0154	0.83	0.2982	1.2376	0.83	0.8329	0.1515	0.0501
56.1387	50.0	1250	57.0000	0.835	0.2982	1.2196	0.835	0.8434	0.1535	0.0493
56.1387	51.0	1275	57.0054	0.825	0.2987	1.2217	0.825	0.8352	0.1517	0.0505
56.1387	52.0	1300	57.0347	0.835	0.2996	1.2239	0.835	0.8407	0.1643	0.0486
56.1387	53.0	1325	57.0183	0.835	0.2989	1.2208	0.835	0.8411	0.1604	0.0495
56.1387	54.0	1350	57.0094	0.845	0.2925	1.1545	0.845	0.8494	0.1515	0.0486
56.1387	55.0	1375	57.0027	0.83	0.2974	1.2161	0.83	0.8380	0.1538	0.0491
56.1387	56.0	1400	57.0060	0.835	0.2975	1.2215	0.835	0.8407	0.1546	0.0505
56.1387	57.0	1425	56.9898	0.835	0.2959	1.1432	0.835	0.8411	0.1483	0.0501
56.1387	58.0	1450	56.9907	0.835	0.2963	1.1437	0.835	0.8406	0.1527	0.0485
56.1387	59.0	1475	56.9578	0.84	0.2935	1.1583	0.8400	0.8439	0.1513	0.0488
55.9877	60.0	1500	57.0032	0.84	0.2957	1.2160	0.8400	0.8439	0.1460	0.0502
55.9877	61.0	1525	56.9880	0.835	0.2990	1.2836	0.835	0.8406	0.1475	0.0489
55.9877	62.0	1550	56.9920	0.83	0.2973	1.2071	0.83	0.8349	0.1519	0.0494
55.9877	63.0	1575	56.9681	0.835	0.2978	1.2076	0.835	0.8406	0.1465	0.0483
55.9877	64.0	1600	56.9772	0.835	0.3003	1.1997	0.835	0.8406	0.1567	0.0489
55.9877	65.0	1625	56.9705	0.835	0.2973	1.2038	0.835	0.8406	0.1520	0.0495
55.9877	66.0	1650	56.9682	0.835	0.2977	1.2005	0.835	0.8406	0.1576	0.0488
55.9877	67.0	1675	56.9775	0.835	0.2981	1.2093	0.835	0.8406	0.1497	0.0501
55.9877	68.0	1700	56.9762	0.835	0.2989	1.2061	0.835	0.8406	0.1626	0.0491
55.9877	69.0	1725	56.9807	0.84	0.2978	1.2023	0.8400	0.8434	0.1503	0.0481
55.9877	70.0	1750	56.9705	0.835	0.2988	1.1987	0.835	0.8406	0.1564	0.0487
55.9877	71.0	1775	56.9752	0.83	0.2987	1.2027	0.83	0.8349	0.1593	0.0497
55.9877	72.0	1800	56.9957	0.83	0.2996	1.2060	0.83	0.8349	0.1607	0.0496
55.9877	73.0	1825	56.9697	0.84	0.2966	1.1977	0.8400	0.8434	0.1510	0.0487
55.9877	74.0	1850	56.9644	0.83	0.2997	1.2055	0.83	0.8349	0.1528	0.0506
55.9877	75.0	1875	56.9677	0.84	0.2968	1.1969	0.8400	0.8434	0.1536	0.0495
55.9877	76.0	1900	56.9609	0.84	0.2958	1.1921	0.8400	0.8434	0.1531	0.0495
55.9877	77.0	1925	56.9663	0.835	0.2965	1.1950	0.835	0.8406	0.1576	0.0494
55.9877	78.0	1950	56.9796	0.83	0.2968	1.2049	0.83	0.8349	0.1525	0.0496
55.9877	79.0	1975	56.9648	0.835	0.2966	1.1944	0.835	0.8406	0.1545	0.0494
55.9237	80.0	2000	56.9596	0.845	0.2944	1.1912	0.845	0.8480	0.1543	0.0492
55.9237	81.0	2025	56.9596	0.84	0.2951	1.1878	0.8400	0.8434	0.1546	0.0492
55.9237	82.0	2050	56.9737	0.84	0.2958	1.1954	0.8400	0.8434	0.1521	0.0498
55.9237	83.0	2075	56.9725	0.835	0.2974	1.1963	0.835	0.8377	0.1512	0.0500
55.9237	84.0	2100	56.9743	0.835	0.2978	1.1928	0.835	0.8406	0.1554	0.0500
55.9237	85.0	2125	56.9788	0.835	0.2971	1.1952	0.835	0.8377	0.1493	0.0500
55.9237	86.0	2150	56.9705	0.84	0.2968	1.1933	0.8400	0.8434	0.1541	0.0499
55.9237	87.0	2175	56.9684	0.835	0.2966	1.1926	0.835	0.8377	0.1517	0.0497
55.9237	88.0	2200	56.9725	0.835	0.2979	1.1934	0.835	0.8377	0.1548	0.0497
55.9237	89.0	2225	56.9704	0.84	0.2959	1.1934	0.8400	0.8434	0.1527	0.0495
55.9237	90.0	2250	56.9681	0.84	0.2950	1.1907	0.8400	0.8434	0.1503	0.0498
55.9237	91.0	2275	56.9763	0.835	0.2979	1.1934	0.835	0.8377	0.1516	0.0501
55.9237	92.0	2300	56.9649	0.835	0.2959	1.1889	0.835	0.8377	0.1501	0.0495
55.9237	93.0	2325	56.9687	0.835	0.2959	1.1871	0.835	0.8377	0.1519	0.0501
55.9237	94.0	2350	56.9663	0.835	0.2963	1.1901	0.835	0.8377	0.1533	0.0496
55.9237	95.0	2375	56.9674	0.84	0.2955	1.1895	0.8400	0.8434	0.1534	0.0498
55.9237	96.0	2400	56.9661	0.835	0.2966	1.1907	0.835	0.8377	0.1520	0.0496
55.9237	97.0	2425	56.9623	0.84	0.2958	1.1871	0.8400	0.8434	0.1532	0.0499
55.9237	98.0	2450	56.9694	0.835	0.2969	1.1897	0.835	0.8377	0.1543	0.0499
55.9237	99.0	2475	56.9698	0.835	0.2967	1.1906	0.835	0.8377	0.1543	0.0499
55.8955	100.0	2500	56.9670	0.835	0.2969	1.1900	0.835	0.8377	0.1545	0.0499

Framework versions

Transformers 4.26.1
Pytorch 1.13.1.post200
Datasets 2.9.0
Tokenizers 0.13.2

jordyvl/vit-small_tobacco3482_hint_

vit-small_tobacco3482_hint_

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions