vit-tiny_tobacco3482

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: 0.2012
Accuracy: 0.825
Brier Loss: 0.3279
Nll: 1.1568
F1 Micro: 0.825
F1 Macro: 0.7904
Ece: 0.2679
Aurc: 0.0635

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	1.7837	0.1	1.0203	8.4259	0.1000	0.0850	0.3286	0.9000
No log	2.0	14	1.0637	0.25	0.8576	5.3606	0.25	0.1988	0.2806	0.7500
No log	3.0	21	0.7848	0.41	0.7314	3.9906	0.41	0.3109	0.2967	0.3924
No log	4.0	28	0.5442	0.565	0.5856	2.7941	0.565	0.4762	0.2765	0.2125
No log	5.0	35	0.4420	0.68	0.4947	1.7142	0.68	0.6058	0.2692	0.1565
No log	6.0	42	0.3671	0.7	0.4278	1.6759	0.7	0.6223	0.2685	0.1186
No log	7.0	49	0.3163	0.765	0.3959	1.6600	0.765	0.6862	0.2639	0.1117
No log	8.0	56	0.3540	0.77	0.3688	1.4505	0.7700	0.7148	0.2634	0.0998
No log	9.0	63	0.3875	0.725	0.3934	1.4205	0.7250	0.6918	0.2622	0.1161
No log	10.0	70	0.3546	0.775	0.3523	1.3197	0.775	0.7619	0.2558	0.0835
No log	11.0	77	0.3046	0.79	0.3515	1.2730	0.79	0.7594	0.2526	0.0978
No log	12.0	84	0.3059	0.8	0.3313	1.2700	0.8000	0.7796	0.2520	0.0783
No log	13.0	91	0.3075	0.775	0.3397	1.2980	0.775	0.7501	0.2632	0.0702
No log	14.0	98	0.2917	0.79	0.3618	1.3528	0.79	0.7715	0.2744	0.0843
No log	15.0	105	0.2393	0.825	0.3268	1.1212	0.825	0.8022	0.2498	0.0656
No log	16.0	112	0.2862	0.8	0.3444	1.2675	0.8000	0.7622	0.2611	0.0689
No log	17.0	119	0.2539	0.79	0.3390	1.2317	0.79	0.7632	0.2492	0.0739
No log	18.0	126	0.2338	0.82	0.3359	1.2455	0.82	0.7838	0.2865	0.0914
No log	19.0	133	0.2357	0.825	0.3197	1.0057	0.825	0.7935	0.2656	0.0727
No log	20.0	140	0.2711	0.81	0.3525	1.1392	0.81	0.7852	0.2961	0.0670
No log	21.0	147	0.2341	0.795	0.3534	1.3905	0.795	0.7645	0.2634	0.0999
No log	22.0	154	0.2635	0.795	0.3382	1.2001	0.795	0.7860	0.2625	0.0635
No log	23.0	161	0.2176	0.82	0.3271	0.9072	0.82	0.7972	0.2703	0.0680
No log	24.0	168	0.2512	0.835	0.3329	1.2192	0.835	0.8160	0.2980	0.0626
No log	25.0	175	0.2169	0.805	0.3414	1.1117	0.805	0.7912	0.2798	0.0776
No log	26.0	182	0.2227	0.84	0.3264	1.0267	0.8400	0.8267	0.2985	0.0669
No log	27.0	189	0.2302	0.79	0.3342	1.1603	0.79	0.7708	0.2589	0.0680
No log	28.0	196	0.2215	0.805	0.3324	1.1168	0.805	0.7786	0.2826	0.0655
No log	29.0	203	0.2022	0.82	0.3217	0.9587	0.82	0.7874	0.2865	0.0646
No log	30.0	210	0.2142	0.805	0.3287	1.1199	0.805	0.7855	0.2526	0.0599
No log	31.0	217	0.2035	0.795	0.3272	1.0385	0.795	0.7717	0.2599	0.0777
No log	32.0	224	0.2079	0.835	0.3246	0.9399	0.835	0.8045	0.2974	0.0586
No log	33.0	231	0.2071	0.81	0.3173	1.2784	0.81	0.7848	0.2520	0.0652
No log	34.0	238	0.2070	0.815	0.3217	1.1020	0.815	0.7855	0.2633	0.0634
No log	35.0	245	0.2128	0.82	0.3235	1.2763	0.82	0.7800	0.2771	0.0593
No log	36.0	252	0.2093	0.825	0.3221	1.1203	0.825	0.8030	0.2666	0.0580
No log	37.0	259	0.1995	0.815	0.3240	1.0387	0.815	0.7831	0.2712	0.0659
No log	38.0	266	0.1977	0.82	0.3207	1.0955	0.82	0.7846	0.2589	0.0629
No log	39.0	273	0.2062	0.82	0.3235	1.0691	0.82	0.7911	0.2666	0.0616
No log	40.0	280	0.1993	0.825	0.3266	1.0812	0.825	0.7973	0.2755	0.0671
No log	41.0	287	0.1976	0.82	0.3288	1.1043	0.82	0.7948	0.2646	0.0688
No log	42.0	294	0.2040	0.825	0.3308	1.2371	0.825	0.7964	0.2782	0.0629
No log	43.0	301	0.2000	0.835	0.3224	1.0857	0.835	0.8041	0.2882	0.0584
No log	44.0	308	0.1987	0.83	0.3222	1.0746	0.83	0.7959	0.2837	0.0631
No log	45.0	315	0.2026	0.82	0.3248	1.1471	0.82	0.7887	0.2843	0.0633
No log	46.0	322	0.2014	0.825	0.3258	1.1310	0.825	0.7916	0.2915	0.0627
No log	47.0	329	0.1988	0.83	0.3237	1.0291	0.83	0.7959	0.2811	0.0633
No log	48.0	336	0.1989	0.82	0.3273	1.1741	0.82	0.7871	0.2699	0.0640
No log	49.0	343	0.1995	0.82	0.3251	1.1518	0.82	0.7869	0.2742	0.0631
No log	50.0	350	0.1996	0.825	0.3241	1.0873	0.825	0.7900	0.2659	0.0616
No log	51.0	357	0.1995	0.83	0.3248	1.1532	0.83	0.7933	0.2651	0.0618
No log	52.0	364	0.1983	0.825	0.3251	1.2117	0.825	0.7904	0.2967	0.0634
No log	53.0	371	0.1984	0.825	0.3254	1.1566	0.825	0.7904	0.2695	0.0635
No log	54.0	378	0.1996	0.825	0.3249	1.1259	0.825	0.7904	0.2841	0.0625
No log	55.0	385	0.1996	0.825	0.3252	1.1424	0.825	0.7904	0.2790	0.0616
No log	56.0	392	0.2004	0.825	0.3243	1.1391	0.825	0.7904	0.2857	0.0623
No log	57.0	399	0.2004	0.825	0.3259	1.2109	0.825	0.7904	0.2788	0.0617
No log	58.0	406	0.2007	0.825	0.3262	1.1473	0.825	0.7900	0.2842	0.0630
No log	59.0	413	0.2000	0.82	0.3264	1.2066	0.82	0.7871	0.2698	0.0638
No log	60.0	420	0.1994	0.82	0.3263	1.1542	0.82	0.7871	0.2822	0.0640
No log	61.0	427	0.1994	0.825	0.3261	1.1506	0.825	0.7904	0.2683	0.0635
No log	62.0	434	0.2005	0.82	0.3276	1.1754	0.82	0.7871	0.2767	0.0639
No log	63.0	441	0.2006	0.82	0.3269	1.2127	0.82	0.7871	0.2811	0.0635
No log	64.0	448	0.2003	0.825	0.3265	1.1547	0.825	0.7904	0.2814	0.0630
No log	65.0	455	0.2005	0.825	0.3268	1.1078	0.825	0.7904	0.3069	0.0629
No log	66.0	462	0.2006	0.825	0.3268	1.0998	0.825	0.7904	0.3012	0.0627
No log	67.0	469	0.2009	0.825	0.3265	1.1526	0.825	0.7904	0.2946	0.0623
No log	68.0	476	0.2006	0.825	0.3269	1.1500	0.825	0.7904	0.2983	0.0631
No log	69.0	483	0.2007	0.825	0.3272	1.1005	0.825	0.7904	0.2756	0.0635
No log	70.0	490	0.2003	0.825	0.3271	1.0947	0.825	0.7904	0.2697	0.0635
No log	71.0	497	0.2006	0.825	0.3272	1.1587	0.825	0.7904	0.2770	0.0636
0.1375	72.0	504	0.2008	0.825	0.3272	1.1549	0.825	0.7904	0.2749	0.0632
0.1375	73.0	511	0.2009	0.825	0.3272	1.1528	0.825	0.7904	0.2748	0.0624
0.1375	74.0	518	0.2013	0.825	0.3275	1.1528	0.825	0.7904	0.2757	0.0626
0.1375	75.0	525	0.2009	0.825	0.3277	1.1529	0.825	0.7904	0.2763	0.0632
0.1375	76.0	532	0.2009	0.825	0.3271	1.1526	0.825	0.7904	0.2754	0.0633
0.1375	77.0	539	0.2006	0.825	0.3274	1.1559	0.825	0.7904	0.2699	0.0636
0.1375	78.0	546	0.2005	0.825	0.3272	1.1499	0.825	0.7904	0.2755	0.0634
0.1375	79.0	553	0.2009	0.825	0.3277	1.1539	0.825	0.7904	0.2833	0.0634
0.1375	80.0	560	0.2009	0.825	0.3275	1.1551	0.825	0.7904	0.2751	0.0632
0.1375	81.0	567	0.2013	0.825	0.3276	1.1563	0.825	0.7904	0.2809	0.0634
0.1375	82.0	574	0.2010	0.825	0.3277	1.1545	0.825	0.7904	0.2752	0.0633
0.1375	83.0	581	0.2009	0.825	0.3275	1.1565	0.825	0.7904	0.2753	0.0634
0.1375	84.0	588	0.2009	0.825	0.3277	1.1564	0.825	0.7904	0.2817	0.0636
0.1375	85.0	595	0.2010	0.825	0.3277	1.1560	0.825	0.7904	0.2686	0.0633
0.1375	86.0	602	0.2010	0.825	0.3278	1.1560	0.825	0.7904	0.2755	0.0633
0.1375	87.0	609	0.2010	0.825	0.3277	1.1544	0.825	0.7904	0.2661	0.0634
0.1375	88.0	616	0.2012	0.825	0.3278	1.1571	0.825	0.7904	0.2612	0.0633
0.1375	89.0	623	0.2010	0.825	0.3278	1.1558	0.825	0.7904	0.2747	0.0635
0.1375	90.0	630	0.2010	0.825	0.3278	1.1564	0.825	0.7904	0.2687	0.0634
0.1375	91.0	637	0.2011	0.825	0.3278	1.1551	0.825	0.7904	0.2678	0.0633
0.1375	92.0	644	0.2011	0.825	0.3278	1.1559	0.825	0.7904	0.2759	0.0635
0.1375	93.0	651	0.2011	0.825	0.3279	1.1565	0.825	0.7904	0.2750	0.0633
0.1375	94.0	658	0.2011	0.825	0.3278	1.1567	0.825	0.7904	0.2760	0.0635
0.1375	95.0	665	0.2011	0.825	0.3279	1.1572	0.825	0.7904	0.2679	0.0633
0.1375	96.0	672	0.2012	0.825	0.3279	1.1574	0.825	0.7904	0.2679	0.0634
0.1375	97.0	679	0.2012	0.825	0.3279	1.1563	0.825	0.7904	0.2678	0.0634
0.1375	98.0	686	0.2012	0.825	0.3279	1.1567	0.825	0.7904	0.2678	0.0634
0.1375	99.0	693	0.2012	0.825	0.3279	1.1569	0.825	0.7904	0.2679	0.0635
0.1375	100.0	700	0.2012	0.825	0.3279	1.1568	0.825	0.7904	0.2679	0.0635

Framework versions

Transformers 4.26.1
Pytorch 1.13.1.post200
Datasets 2.9.0
Tokenizers 0.13.2

jordyvl/vit-tiny_tobacco3482

vit-tiny_tobacco3482

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions