81-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: 0.1635
Accuracy: 0.815
Brier Loss: 0.3719
Nll: 0.7360
F1 Micro: 0.815
F1 Macro: 0.7973
Ece: 0.3345
Aurc: 0.0542

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	2.2942	0.16	1.0273	8.4554	0.16	0.0543	0.3538	0.8479
No log	2.0	14	1.1952	0.095	0.9063	7.1126	0.095	0.0819	0.2327	0.8615
No log	3.0	21	0.7862	0.305	0.8434	5.1628	0.305	0.1452	0.2965	0.6092
No log	4.0	28	0.6144	0.385	0.7594	3.3452	0.3850	0.2942	0.2978	0.4140
No log	5.0	35	0.5040	0.54	0.6858	2.6433	0.54	0.4419	0.3200	0.2853
No log	6.0	42	0.4291	0.625	0.6353	2.2564	0.625	0.5273	0.3772	0.1975
No log	7.0	49	0.3839	0.63	0.5815	2.0215	0.63	0.4970	0.3469	0.1795
No log	8.0	56	0.3390	0.705	0.5512	1.5239	0.705	0.5983	0.3704	0.1337
No log	9.0	63	0.3128	0.705	0.5015	1.7837	0.705	0.6172	0.3403	0.1138
No log	10.0	70	0.2868	0.705	0.4532	1.1243	0.705	0.6108	0.3221	0.1015
No log	11.0	77	0.2866	0.74	0.4532	1.0967	0.74	0.6909	0.3422	0.0932
No log	12.0	84	0.2625	0.765	0.4235	1.4646	0.765	0.6788	0.3240	0.0805
No log	13.0	91	0.2516	0.73	0.4141	1.4855	0.7300	0.6495	0.2770	0.0869
No log	14.0	98	0.2487	0.805	0.4440	1.2873	0.805	0.7319	0.3699	0.0670
No log	15.0	105	0.2103	0.77	0.4029	1.0834	0.7700	0.6891	0.3007	0.0677
No log	16.0	112	0.2216	0.8	0.4029	1.0621	0.8000	0.7247	0.3513	0.0611
No log	17.0	119	0.2230	0.745	0.3999	1.0367	0.745	0.6710	0.2992	0.0809
No log	18.0	126	0.2301	0.805	0.4044	0.9650	0.805	0.7473	0.3406	0.0538
No log	19.0	133	0.2024	0.795	0.3935	0.9429	0.795	0.7562	0.3327	0.0669
No log	20.0	140	0.1959	0.82	0.3935	0.8984	0.82	0.8000	0.3453	0.0571
No log	21.0	147	0.2020	0.815	0.3946	0.8507	0.815	0.7949	0.3669	0.0588
No log	22.0	154	0.2100	0.805	0.3758	0.8083	0.805	0.7743	0.3087	0.0625
No log	23.0	161	0.2002	0.81	0.3881	1.0046	0.81	0.7953	0.3276	0.0617
No log	24.0	168	0.2004	0.83	0.3987	0.8922	0.83	0.8034	0.3565	0.0531
No log	25.0	175	0.1914	0.785	0.3699	0.8571	0.785	0.7484	0.3117	0.0601
No log	26.0	182	0.1845	0.815	0.3764	0.8153	0.815	0.7970	0.3215	0.0539
No log	27.0	189	0.1821	0.835	0.3815	0.8488	0.835	0.8175	0.3441	0.0497
No log	28.0	196	0.1869	0.84	0.3808	0.8654	0.8400	0.8236	0.3639	0.0504
No log	29.0	203	0.1859	0.79	0.3752	0.7067	0.79	0.7661	0.3019	0.0618
No log	30.0	210	0.1842	0.795	0.3826	0.9031	0.795	0.7646	0.3170	0.0599
No log	31.0	217	0.1797	0.815	0.3714	0.8572	0.815	0.8038	0.3214	0.0588
No log	32.0	224	0.1754	0.805	0.3679	0.7412	0.805	0.7883	0.3063	0.0563
No log	33.0	231	0.1790	0.835	0.3736	0.8357	0.835	0.8128	0.3431	0.0497
No log	34.0	238	0.1761	0.81	0.3744	0.7360	0.81	0.7907	0.3292	0.0526
No log	35.0	245	0.1731	0.83	0.3744	0.7494	0.83	0.8202	0.3351	0.0540
No log	36.0	252	0.1758	0.79	0.3738	0.9441	0.79	0.7710	0.3099	0.0625
No log	37.0	259	0.1730	0.81	0.3785	0.9418	0.81	0.7939	0.3425	0.0577
No log	38.0	266	0.1727	0.815	0.3752	0.7539	0.815	0.8001	0.3290	0.0540
No log	39.0	273	0.1769	0.81	0.3754	0.9013	0.81	0.7888	0.3296	0.0563
No log	40.0	280	0.1770	0.805	0.3809	0.7637	0.805	0.7809	0.3291	0.0519
No log	41.0	287	0.1771	0.815	0.3814	0.7495	0.815	0.7993	0.3352	0.0565
No log	42.0	294	0.1745	0.8	0.3807	0.8386	0.8000	0.7865	0.3212	0.0601
No log	43.0	301	0.1700	0.825	0.3762	0.8216	0.825	0.8067	0.3420	0.0517
No log	44.0	308	0.1686	0.825	0.3706	0.7310	0.825	0.8024	0.3448	0.0509
No log	45.0	315	0.1692	0.81	0.3739	0.6805	0.81	0.7914	0.3397	0.0508
No log	46.0	322	0.1697	0.825	0.3709	0.7482	0.825	0.7995	0.3364	0.0531
No log	47.0	329	0.1681	0.805	0.3735	0.8810	0.805	0.7872	0.3388	0.0585
No log	48.0	336	0.1667	0.815	0.3731	0.6788	0.815	0.7968	0.3427	0.0567
No log	49.0	343	0.1690	0.82	0.3786	0.8693	0.82	0.7967	0.3507	0.0502
No log	50.0	350	0.1668	0.82	0.3761	0.7854	0.82	0.8032	0.3509	0.0526
No log	51.0	357	0.1659	0.82	0.3760	0.7437	0.82	0.8045	0.3374	0.0557
No log	52.0	364	0.1663	0.805	0.3682	0.7932	0.805	0.7863	0.3171	0.0560
No log	53.0	371	0.1649	0.805	0.3649	0.7249	0.805	0.7863	0.3139	0.0562
No log	54.0	378	0.1654	0.815	0.3725	0.7390	0.815	0.7935	0.3273	0.0560
No log	55.0	385	0.1667	0.84	0.3741	0.7324	0.8400	0.8279	0.3501	0.0513
No log	56.0	392	0.1659	0.825	0.3718	0.7278	0.825	0.8058	0.3288	0.0522
No log	57.0	399	0.1659	0.825	0.3731	0.7351	0.825	0.8084	0.3204	0.0537
No log	58.0	406	0.1645	0.825	0.3730	0.7356	0.825	0.8068	0.3468	0.0524
No log	59.0	413	0.1626	0.825	0.3715	0.7282	0.825	0.8040	0.3391	0.0527
No log	60.0	420	0.1628	0.825	0.3698	0.7297	0.825	0.8056	0.3212	0.0534
No log	61.0	427	0.1630	0.825	0.3719	0.7312	0.825	0.8065	0.3260	0.0533
No log	62.0	434	0.1642	0.82	0.3714	0.7335	0.82	0.8040	0.3388	0.0549
No log	63.0	441	0.1634	0.82	0.3733	0.7349	0.82	0.8011	0.3399	0.0553
No log	64.0	448	0.1628	0.815	0.3710	0.7301	0.815	0.7974	0.3249	0.0532
No log	65.0	455	0.1629	0.815	0.3704	0.7362	0.815	0.7981	0.3362	0.0541
No log	66.0	462	0.1630	0.82	0.3727	0.7354	0.82	0.8024	0.3559	0.0526
No log	67.0	469	0.1633	0.825	0.3733	0.7355	0.825	0.8038	0.3492	0.0526
No log	68.0	476	0.1638	0.82	0.3724	0.7369	0.82	0.8040	0.3468	0.0550
No log	69.0	483	0.1629	0.82	0.3717	0.7329	0.82	0.8040	0.3251	0.0536
No log	70.0	490	0.1629	0.815	0.3709	0.7312	0.815	0.7947	0.3221	0.0541
No log	71.0	497	0.1630	0.815	0.3718	0.7371	0.815	0.7973	0.3436	0.0545
0.1743	72.0	504	0.1631	0.815	0.3712	0.7350	0.815	0.7973	0.3264	0.0536
0.1743	73.0	511	0.1634	0.815	0.3721	0.7348	0.815	0.7981	0.3340	0.0543
0.1743	74.0	518	0.1631	0.815	0.3716	0.7332	0.815	0.7973	0.3279	0.0541
0.1743	75.0	525	0.1633	0.82	0.3720	0.7346	0.82	0.8008	0.3186	0.0542
0.1743	76.0	532	0.1631	0.815	0.3716	0.7342	0.815	0.7973	0.3189	0.0542
0.1743	77.0	539	0.1633	0.815	0.3718	0.7358	0.815	0.7973	0.3344	0.0542
0.1743	78.0	546	0.1633	0.815	0.3718	0.7363	0.815	0.7991	0.3342	0.0542
0.1743	79.0	553	0.1633	0.815	0.3718	0.7350	0.815	0.7973	0.3344	0.0543
0.1743	80.0	560	0.1634	0.815	0.3718	0.7354	0.815	0.7973	0.3136	0.0542
0.1743	81.0	567	0.1633	0.815	0.3721	0.7360	0.815	0.7973	0.3279	0.0543
0.1743	82.0	574	0.1634	0.815	0.3719	0.7358	0.815	0.7973	0.3276	0.0542
0.1743	83.0	581	0.1634	0.815	0.3720	0.7358	0.815	0.7973	0.3345	0.0542
0.1743	84.0	588	0.1633	0.815	0.3717	0.7352	0.815	0.7973	0.3275	0.0542
0.1743	85.0	595	0.1633	0.815	0.3720	0.7357	0.815	0.7973	0.3194	0.0541
0.1743	86.0	602	0.1634	0.815	0.3717	0.7361	0.815	0.7973	0.3276	0.0542
0.1743	87.0	609	0.1634	0.815	0.3719	0.7359	0.815	0.7973	0.3345	0.0542
0.1743	88.0	616	0.1634	0.815	0.3720	0.7359	0.815	0.7973	0.3278	0.0542
0.1743	89.0	623	0.1634	0.815	0.3718	0.7359	0.815	0.7973	0.3192	0.0543
0.1743	90.0	630	0.1635	0.815	0.3719	0.7359	0.815	0.7973	0.3278	0.0543
0.1743	91.0	637	0.1635	0.815	0.3719	0.7357	0.815	0.7973	0.3344	0.0541
0.1743	92.0	644	0.1635	0.815	0.3719	0.7361	0.815	0.7973	0.3278	0.0542
0.1743	93.0	651	0.1635	0.815	0.3719	0.7357	0.815	0.7973	0.3345	0.0542
0.1743	94.0	658	0.1635	0.815	0.3719	0.7356	0.815	0.7973	0.3261	0.0543
0.1743	95.0	665	0.1635	0.815	0.3719	0.7360	0.815	0.7973	0.3345	0.0542
0.1743	96.0	672	0.1635	0.815	0.3719	0.7357	0.815	0.7973	0.3278	0.0542
0.1743	97.0	679	0.1635	0.815	0.3719	0.7360	0.815	0.7973	0.3345	0.0542
0.1743	98.0	686	0.1635	0.815	0.3719	0.7360	0.815	0.7973	0.3278	0.0543
0.1743	99.0	693	0.1635	0.815	0.3719	0.7360	0.815	0.7973	0.3345	0.0542
0.1743	100.0	700	0.1635	0.815	0.3719	0.7360	0.815	0.7973	0.3345	0.0542

Framework versions

Transformers 4.26.1
Pytorch 1.13.1.post200
Datasets 2.9.0
Tokenizers 0.13.2

jordyvl/81-tiny_tobacco3482

81-tiny_tobacco3482

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions