60-tiny_tobacco3482_kd_NKD_t1.0_g1.5

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: 4.0655
Accuracy: 0.82
Brier Loss: 0.2655
Nll: 1.0777
F1 Micro: 0.82
F1 Macro: 0.7953
Ece: 0.1364
Aurc: 0.0539

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	5.7153	0.09	1.0182	8.6654	0.09	0.0745	0.3226	0.8984
No log	2.0	14	4.7081	0.13	0.9028	8.0605	0.13	0.0949	0.2402	0.8297
No log	3.0	21	4.3197	0.215	0.8419	6.1917	0.2150	0.1338	0.2566	0.5994
No log	4.0	28	4.1502	0.405	0.7493	4.1855	0.405	0.3261	0.3035	0.4065
No log	5.0	35	4.0310	0.545	0.6643	2.9984	0.545	0.4451	0.3080	0.2731
No log	6.0	42	3.9472	0.64	0.6089	2.4473	0.64	0.5047	0.3290	0.1849
No log	7.0	49	3.8616	0.655	0.5435	2.3233	0.655	0.5434	0.3086	0.1536
No log	8.0	56	3.8289	0.72	0.5177	2.1266	0.72	0.6162	0.3496	0.1202
No log	9.0	63	3.7812	0.715	0.4461	1.7718	0.715	0.6276	0.2464	0.1099
No log	10.0	70	3.7599	0.715	0.3960	1.7795	0.715	0.6158	0.2262	0.0879
No log	11.0	77	3.7506	0.74	0.3758	1.9447	0.74	0.6284	0.2212	0.0847
No log	12.0	84	3.7261	0.79	0.3515	1.7839	0.79	0.7107	0.2419	0.0698
No log	13.0	91	3.8045	0.72	0.3765	1.8827	0.72	0.6249	0.1943	0.0941
No log	14.0	98	3.6964	0.79	0.3309	1.8042	0.79	0.6873	0.2162	0.0629
No log	15.0	105	3.7058	0.78	0.3285	1.8229	0.78	0.6827	0.1999	0.0672
No log	16.0	112	3.8395	0.74	0.3502	1.9683	0.74	0.6554	0.1777	0.0880
No log	17.0	119	3.7568	0.795	0.3120	1.3060	0.795	0.7121	0.1656	0.0633
No log	18.0	126	3.7399	0.795	0.3099	1.6225	0.795	0.7133	0.1721	0.0615
No log	19.0	133	3.7813	0.82	0.2884	1.3165	0.82	0.7475	0.1698	0.0565
No log	20.0	140	3.7907	0.8	0.3074	1.2638	0.8000	0.7701	0.1862	0.0654
No log	21.0	147	3.7687	0.83	0.2860	1.2646	0.83	0.8077	0.1730	0.0569
No log	22.0	154	3.7182	0.82	0.2813	1.3271	0.82	0.7848	0.1762	0.0551
No log	23.0	161	3.7801	0.83	0.2679	1.1397	0.83	0.8015	0.1574	0.0568
No log	24.0	168	3.7659	0.82	0.2752	1.1301	0.82	0.7893	0.1581	0.0582
No log	25.0	175	3.7960	0.825	0.2755	1.2565	0.825	0.8017	0.1663	0.0540
No log	26.0	182	3.7630	0.82	0.2715	1.0265	0.82	0.7971	0.1439	0.0536
No log	27.0	189	3.8411	0.82	0.2822	1.2532	0.82	0.7929	0.1621	0.0623
No log	28.0	196	3.7869	0.83	0.2608	1.1781	0.83	0.8013	0.1461	0.0531
No log	29.0	203	3.7798	0.81	0.2724	1.0950	0.81	0.7785	0.1436	0.0562
No log	30.0	210	3.8412	0.82	0.2724	1.3168	0.82	0.7892	0.1473	0.0596
No log	31.0	217	3.7937	0.83	0.2683	1.1818	0.83	0.8137	0.1523	0.0515
No log	32.0	224	3.7964	0.825	0.2655	1.1605	0.825	0.8028	0.1580	0.0551
No log	33.0	231	3.7950	0.81	0.2685	1.2514	0.81	0.7854	0.1275	0.0534
No log	34.0	238	3.8086	0.83	0.2599	0.9670	0.83	0.8022	0.1309	0.0529
No log	35.0	245	3.8393	0.83	0.2649	0.9615	0.83	0.8050	0.1445	0.0543
No log	36.0	252	3.8140	0.84	0.2528	1.0685	0.8400	0.8194	0.1371	0.0528
No log	37.0	259	3.8164	0.825	0.2665	1.0001	0.825	0.8030	0.1501	0.0548
No log	38.0	266	3.8298	0.82	0.2669	1.0166	0.82	0.8007	0.1413	0.0549
No log	39.0	273	3.8346	0.815	0.2646	1.1609	0.815	0.7876	0.1453	0.0551
No log	40.0	280	3.8409	0.825	0.2610	1.1811	0.825	0.8054	0.1432	0.0551
No log	41.0	287	3.8106	0.835	0.2642	1.0262	0.835	0.8078	0.1471	0.0547
No log	42.0	294	3.8402	0.81	0.2591	0.9683	0.81	0.7863	0.1638	0.0546
No log	43.0	301	3.8480	0.82	0.2604	1.1408	0.82	0.7933	0.1448	0.0538
No log	44.0	308	3.8627	0.835	0.2717	1.0142	0.835	0.8113	0.1472	0.0550
No log	45.0	315	3.8274	0.83	0.2589	1.0148	0.83	0.8138	0.1333	0.0518
No log	46.0	322	3.8349	0.83	0.2563	0.9773	0.83	0.8086	0.1237	0.0537
No log	47.0	329	3.8656	0.82	0.2668	0.9728	0.82	0.7947	0.1330	0.0546
No log	48.0	336	3.8455	0.84	0.2514	0.9784	0.8400	0.8175	0.1519	0.0509
No log	49.0	343	3.8827	0.81	0.2740	0.9515	0.81	0.7852	0.1325	0.0565
No log	50.0	350	3.8617	0.815	0.2659	1.2129	0.815	0.7885	0.1430	0.0565
No log	51.0	357	3.8555	0.825	0.2601	0.9544	0.825	0.8023	0.1500	0.0535
No log	52.0	364	3.8794	0.825	0.2611	1.1499	0.825	0.7998	0.1520	0.0550
No log	53.0	371	3.8563	0.815	0.2592	1.0856	0.815	0.7896	0.1198	0.0525
No log	54.0	378	3.8851	0.825	0.2579	0.8909	0.825	0.8042	0.1320	0.0543
No log	55.0	385	3.8682	0.815	0.2589	1.1381	0.815	0.7906	0.1376	0.0553
No log	56.0	392	3.8646	0.83	0.2582	1.0602	0.83	0.8144	0.1330	0.0532
No log	57.0	399	3.9205	0.82	0.2678	1.0739	0.82	0.7991	0.1403	0.0569
No log	58.0	406	3.8878	0.835	0.2613	0.9320	0.835	0.8112	0.1575	0.0527
No log	59.0	413	3.8944	0.82	0.2662	1.0116	0.82	0.7966	0.1304	0.0544
No log	60.0	420	3.8819	0.825	0.2597	1.1330	0.825	0.8019	0.1471	0.0529
No log	61.0	427	3.9219	0.82	0.2719	1.0206	0.82	0.7957	0.1465	0.0563
No log	62.0	434	3.8903	0.815	0.2583	0.9983	0.815	0.7945	0.1428	0.0532
No log	63.0	441	3.9258	0.82	0.2674	1.0722	0.82	0.7941	0.1312	0.0552
No log	64.0	448	3.9086	0.83	0.2546	1.1474	0.83	0.8049	0.1375	0.0522
No log	65.0	455	3.9260	0.83	0.2655	1.0795	0.83	0.8074	0.1300	0.0537
No log	66.0	462	3.9424	0.82	0.2634	1.0220	0.82	0.7978	0.1298	0.0542
No log	67.0	469	3.9315	0.815	0.2605	1.0206	0.815	0.7882	0.1404	0.0534
No log	68.0	476	3.9583	0.82	0.2634	1.0018	0.82	0.7991	0.1321	0.0532
No log	69.0	483	3.9447	0.825	0.2606	1.0840	0.825	0.8003	0.1326	0.0533
No log	70.0	490	3.9698	0.815	0.2646	0.9721	0.815	0.7940	0.1359	0.0546
No log	71.0	497	3.9621	0.82	0.2615	1.0207	0.82	0.7953	0.1312	0.0540
3.4582	72.0	504	3.9687	0.81	0.2628	1.0078	0.81	0.7861	0.1379	0.0549
3.4582	73.0	511	3.9699	0.825	0.2575	1.0739	0.825	0.8024	0.1257	0.0534
3.4582	74.0	518	3.9791	0.82	0.2638	1.0290	0.82	0.7978	0.1337	0.0543
3.4582	75.0	525	3.9893	0.82	0.2664	1.0711	0.82	0.7978	0.1323	0.0543
3.4582	76.0	532	3.9843	0.825	0.2620	0.9559	0.825	0.7999	0.1307	0.0535
3.4582	77.0	539	3.9983	0.82	0.2626	1.0239	0.82	0.7953	0.1333	0.0543
3.4582	78.0	546	4.0068	0.82	0.2663	1.0205	0.82	0.7953	0.1421	0.0542
3.4582	79.0	553	4.0096	0.82	0.2636	1.0204	0.82	0.7953	0.1296	0.0541
3.4582	80.0	560	4.0211	0.82	0.2655	1.0147	0.82	0.7953	0.1365	0.0541
3.4582	81.0	567	4.0206	0.815	0.2631	1.0781	0.815	0.7902	0.1378	0.0543
3.4582	82.0	574	4.0270	0.82	0.2652	1.0752	0.82	0.7953	0.1407	0.0539
3.4582	83.0	581	4.0301	0.82	0.2642	1.0763	0.82	0.7953	0.1444	0.0537
3.4582	84.0	588	4.0369	0.82	0.2652	1.0758	0.82	0.7953	0.1355	0.0539
3.4582	85.0	595	4.0395	0.82	0.2648	1.0244	0.82	0.7953	0.1300	0.0539
3.4582	86.0	602	4.0428	0.82	0.2651	1.0768	0.82	0.7953	0.1209	0.0538
3.4582	87.0	609	4.0461	0.82	0.2653	1.0780	0.82	0.7953	0.1295	0.0536
3.4582	88.0	616	4.0485	0.82	0.2650	1.0780	0.82	0.7953	0.1375	0.0537
3.4582	89.0	623	4.0521	0.82	0.2652	1.0259	0.82	0.7953	0.1296	0.0537
3.4582	90.0	630	4.0540	0.82	0.2652	1.0466	0.82	0.7953	0.1282	0.0538
3.4582	91.0	637	4.0562	0.82	0.2653	1.0778	0.82	0.7953	0.1304	0.0538
3.4582	92.0	644	4.0587	0.82	0.2655	1.0774	0.82	0.7953	0.1241	0.0537
3.4582	93.0	651	4.0607	0.82	0.2657	1.0771	0.82	0.7953	0.1357	0.0538
3.4582	94.0	658	4.0615	0.82	0.2655	1.0775	0.82	0.7953	0.1268	0.0538
3.4582	95.0	665	4.0625	0.82	0.2654	1.0300	0.82	0.7953	0.1211	0.0538
3.4582	96.0	672	4.0641	0.82	0.2657	1.0771	0.82	0.7953	0.1365	0.0538
3.4582	97.0	679	4.0641	0.82	0.2654	1.0312	0.82	0.7953	0.1210	0.0538
3.4582	98.0	686	4.0652	0.82	0.2656	1.0776	0.82	0.7953	0.1269	0.0538
3.4582	99.0	693	4.0654	0.82	0.2656	1.0775	0.82	0.7953	0.1364	0.0539
3.4582	100.0	700	4.0655	0.82	0.2655	1.0777	0.82	0.7953	0.1364	0.0539

Framework versions

Transformers 4.26.1
Pytorch 1.13.1.post200
Datasets 2.9.0
Tokenizers 0.13.2

60-tiny_tobacco3482_kd_NKD_t1.0_g1.5

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

NSDT 3DConvert

UnrealSynth

DreamTexture.js