6_e_200-tiny_tobacco3482_kd_CEKD_t1.5_a0.9

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: 0.5536
Accuracy: 0.82
Brier Loss: 0.2571
Nll: 1.4560
F1 Micro: 0.82
F1 Macro: 0.7994
Ece: 0.1404
Aurc: 0.0578

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	2.0125	0.23	0.8650	4.4951	0.23	0.1799	0.2806	0.7660
No log	2.0	50	1.2756	0.555	0.5948	2.6781	0.555	0.4537	0.2800	0.2519
No log	3.0	75	0.9515	0.685	0.4392	1.9416	0.685	0.5937	0.2067	0.1288
No log	4.0	100	0.7861	0.72	0.3622	1.5125	0.72	0.6675	0.2050	0.0961
No log	5.0	125	0.7551	0.77	0.3362	1.5478	0.7700	0.7318	0.2043	0.0838
No log	6.0	150	0.8056	0.77	0.3525	1.4305	0.7700	0.7589	0.1943	0.0891
No log	7.0	175	0.7942	0.775	0.3310	1.8237	0.775	0.7454	0.1812	0.0924
No log	8.0	200	0.7735	0.77	0.3384	1.5161	0.7700	0.7530	0.1987	0.0931
No log	9.0	225	0.6992	0.79	0.3025	1.5664	0.79	0.7777	0.1631	0.0774
No log	10.0	250	0.6753	0.8	0.2955	1.5189	0.8000	0.7900	0.1654	0.0633
No log	11.0	275	0.7701	0.805	0.3018	1.4787	0.805	0.7932	0.1581	0.0881
No log	12.0	300	0.7164	0.79	0.3292	1.3527	0.79	0.7892	0.1946	0.0871
No log	13.0	325	0.6376	0.8	0.2901	1.4953	0.8000	0.7824	0.1770	0.0659
No log	14.0	350	0.7319	0.77	0.3247	1.6062	0.7700	0.7424	0.1803	0.0816
No log	15.0	375	0.5749	0.805	0.2738	0.8483	0.805	0.8010	0.1569	0.0647
No log	16.0	400	0.6879	0.775	0.3085	1.3379	0.775	0.7759	0.1909	0.0730
No log	17.0	425	0.5094	0.85	0.2241	1.4391	0.85	0.8360	0.1589	0.0441
No log	18.0	450	0.6826	0.8	0.3015	1.6933	0.8000	0.7969	0.1651	0.0792
No log	19.0	475	0.5677	0.825	0.2622	1.5426	0.825	0.8051	0.1600	0.0515
0.4493	20.0	500	0.5156	0.85	0.2312	1.5882	0.85	0.8471	0.1466	0.0427
0.4493	21.0	525	0.5743	0.83	0.2600	1.5702	0.83	0.8187	0.1604	0.0540
0.4493	22.0	550	0.5872	0.825	0.2712	1.6270	0.825	0.8056	0.1687	0.0572
0.4493	23.0	575	0.5770	0.81	0.2701	1.5089	0.81	0.7969	0.1559	0.0655
0.4493	24.0	600	0.5621	0.82	0.2590	1.3500	0.82	0.8052	0.1621	0.0587
0.4493	25.0	625	0.5480	0.805	0.2518	1.2519	0.805	0.7884	0.1483	0.0619
0.4493	26.0	650	0.5555	0.81	0.2575	1.3183	0.81	0.7926	0.1585	0.0598
0.4493	27.0	675	0.5449	0.82	0.2524	1.4400	0.82	0.8059	0.1713	0.0579
0.4493	28.0	700	0.5483	0.81	0.2545	1.4400	0.81	0.7894	0.1450	0.0580
0.4493	29.0	725	0.5448	0.81	0.2524	1.3070	0.81	0.7931	0.1447	0.0595
0.4493	30.0	750	0.5476	0.815	0.2538	1.3101	0.815	0.7982	0.1536	0.0582
0.4493	31.0	775	0.5433	0.82	0.2529	1.3812	0.82	0.8011	0.1637	0.0575
0.4493	32.0	800	0.5469	0.805	0.2528	1.2973	0.805	0.7905	0.1668	0.0600
0.4493	33.0	825	0.5443	0.815	0.2525	1.3020	0.815	0.7933	0.1768	0.0579
0.4493	34.0	850	0.5442	0.82	0.2521	1.3234	0.82	0.8011	0.1555	0.0580
0.4493	35.0	875	0.5434	0.82	0.2531	1.4362	0.82	0.8011	0.1430	0.0564
0.4493	36.0	900	0.5469	0.815	0.2534	1.3075	0.815	0.7933	0.1590	0.0578
0.4493	37.0	925	0.5468	0.815	0.2546	1.3204	0.815	0.7933	0.1623	0.0567
0.4493	38.0	950	0.5473	0.815	0.2540	1.3722	0.815	0.7933	0.1514	0.0582
0.4493	39.0	975	0.5453	0.82	0.2532	1.3874	0.82	0.8011	0.1751	0.0568
0.0581	40.0	1000	0.5475	0.815	0.2543	1.3116	0.815	0.7933	0.1654	0.0573
0.0581	41.0	1025	0.5452	0.815	0.2533	1.4421	0.815	0.7933	0.1459	0.0579
0.0581	42.0	1050	0.5467	0.815	0.2538	1.3730	0.815	0.7933	0.1642	0.0576
0.0581	43.0	1075	0.5478	0.815	0.2544	1.3086	0.815	0.7933	0.1657	0.0581
0.0581	44.0	1100	0.5482	0.815	0.2545	1.3744	0.815	0.7933	0.1629	0.0583
0.0581	45.0	1125	0.5493	0.815	0.2550	1.3676	0.815	0.7933	0.1638	0.0594
0.0581	46.0	1150	0.5478	0.82	0.2547	1.4645	0.82	0.8011	0.1631	0.0572
0.0581	47.0	1175	0.5487	0.815	0.2547	1.3795	0.815	0.7933	0.1634	0.0577
0.0581	48.0	1200	0.5471	0.825	0.2546	1.4421	0.825	0.8067	0.1436	0.0564
0.0581	49.0	1225	0.5489	0.815	0.2547	1.3676	0.815	0.7933	0.1663	0.0578
0.0581	50.0	1250	0.5482	0.82	0.2549	1.4346	0.82	0.7990	0.1481	0.0574
0.0581	51.0	1275	0.5472	0.82	0.2540	1.5012	0.82	0.8011	0.1565	0.0569
0.0581	52.0	1300	0.5489	0.825	0.2553	1.4351	0.825	0.8051	0.1608	0.0576
0.0581	53.0	1325	0.5486	0.815	0.2549	1.3799	0.815	0.7933	0.1483	0.0573
0.0581	54.0	1350	0.5498	0.815	0.2552	1.4434	0.815	0.7933	0.1542	0.0578
0.0581	55.0	1375	0.5508	0.82	0.2559	1.4394	0.82	0.7994	0.1562	0.0576
0.0581	56.0	1400	0.5492	0.825	0.2552	1.4368	0.825	0.8051	0.1483	0.0572
0.0581	57.0	1425	0.5501	0.815	0.2552	1.3874	0.815	0.7933	0.1390	0.0579
0.0581	58.0	1450	0.5497	0.82	0.2553	1.4365	0.82	0.7994	0.1437	0.0579
0.0581	59.0	1475	0.5507	0.82	0.2557	1.4343	0.82	0.7994	0.1389	0.0584
0.056	60.0	1500	0.5501	0.825	0.2555	1.4410	0.825	0.8051	0.1585	0.0583
0.056	61.0	1525	0.5510	0.82	0.2559	1.4380	0.82	0.7994	0.1395	0.0578
0.056	62.0	1550	0.5510	0.82	0.2558	1.4421	0.82	0.7994	0.1441	0.0573
0.056	63.0	1575	0.5508	0.82	0.2559	1.4369	0.82	0.7994	0.1395	0.0575
0.056	64.0	1600	0.5514	0.82	0.2560	1.4410	0.82	0.7994	0.1393	0.0579
0.056	65.0	1625	0.5519	0.825	0.2563	1.4544	0.825	0.8051	0.1427	0.0575
0.056	66.0	1650	0.5510	0.82	0.2560	1.4400	0.82	0.7994	0.1391	0.0576
0.056	67.0	1675	0.5520	0.825	0.2563	1.4396	0.825	0.8051	0.1422	0.0580
0.056	68.0	1700	0.5516	0.82	0.2561	1.4412	0.82	0.7994	0.1394	0.0580
0.056	69.0	1725	0.5512	0.82	0.2560	1.4433	0.82	0.7994	0.1393	0.0577
0.056	70.0	1750	0.5515	0.82	0.2561	1.4418	0.82	0.7994	0.1391	0.0576
0.056	71.0	1775	0.5517	0.82	0.2562	1.4448	0.82	0.7994	0.1449	0.0581
0.056	72.0	1800	0.5524	0.825	0.2566	1.4421	0.825	0.8051	0.1437	0.0579
0.056	73.0	1825	0.5518	0.82	0.2562	1.4403	0.82	0.7994	0.1469	0.0576
0.056	74.0	1850	0.5529	0.825	0.2568	1.4450	0.825	0.8051	0.1434	0.0580
0.056	75.0	1875	0.5528	0.82	0.2566	1.4475	0.82	0.7994	0.1447	0.0585
0.056	76.0	1900	0.5529	0.82	0.2568	1.4463	0.82	0.7994	0.1447	0.0578
0.056	77.0	1925	0.5528	0.82	0.2567	1.4469	0.82	0.7994	0.1401	0.0577
0.056	78.0	1950	0.5525	0.82	0.2565	1.4506	0.82	0.7994	0.1444	0.0576
0.056	79.0	1975	0.5527	0.825	0.2567	1.4479	0.825	0.8051	0.1423	0.0576
0.0559	80.0	2000	0.5530	0.825	0.2568	1.4429	0.825	0.8051	0.1423	0.0578
0.0559	81.0	2025	0.5529	0.825	0.2567	1.4489	0.825	0.8051	0.1422	0.0581
0.0559	82.0	2050	0.5529	0.82	0.2568	1.4550	0.82	0.7994	0.1401	0.0576
0.0559	83.0	2075	0.5534	0.82	0.2570	1.4458	0.82	0.7994	0.1399	0.0580
0.0559	84.0	2100	0.5530	0.82	0.2568	1.4497	0.82	0.7994	0.1399	0.0577
0.0559	85.0	2125	0.5533	0.82	0.2570	1.4507	0.82	0.7994	0.1401	0.0577
0.0559	86.0	2150	0.5531	0.825	0.2568	1.4515	0.825	0.8051	0.1428	0.0577
0.0559	87.0	2175	0.5534	0.82	0.2569	1.4503	0.82	0.7994	0.1404	0.0577
0.0559	88.0	2200	0.5534	0.82	0.2569	1.4532	0.82	0.7994	0.1399	0.0581
0.0559	89.0	2225	0.5533	0.825	0.2569	1.4499	0.825	0.8051	0.1423	0.0578
0.0559	90.0	2250	0.5534	0.82	0.2570	1.4517	0.82	0.7994	0.1404	0.0577
0.0559	91.0	2275	0.5533	0.82	0.2569	1.4526	0.82	0.7994	0.1405	0.0579
0.0559	92.0	2300	0.5534	0.825	0.2570	1.4533	0.825	0.8051	0.1424	0.0577
0.0559	93.0	2325	0.5535	0.82	0.2570	1.4527	0.82	0.7994	0.1399	0.0580
0.0559	94.0	2350	0.5536	0.82	0.2571	1.4533	0.82	0.7994	0.1404	0.0577
0.0559	95.0	2375	0.5536	0.82	0.2571	1.4547	0.82	0.7994	0.1400	0.0579
0.0559	96.0	2400	0.5535	0.82	0.2570	1.4567	0.82	0.7994	0.1400	0.0578
0.0559	97.0	2425	0.5536	0.82	0.2571	1.4523	0.82	0.7994	0.1404	0.0579
0.0559	98.0	2450	0.5536	0.82	0.2571	1.4570	0.82	0.7994	0.1404	0.0578
0.0559	99.0	2475	0.5536	0.82	0.2571	1.4570	0.82	0.7994	0.1404	0.0578
0.0559	100.0	2500	0.5536	0.82	0.2571	1.4560	0.82	0.7994	0.1404	0.0578

Framework versions

Transformers 4.30.2
Pytorch 1.13.1
Datasets 2.13.1
Tokenizers 0.13.3

6_e_200-tiny_tobacco3482_kd_CEKD_t1.5_a0.9

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

NSDT 3DConvert

UnrealSynth

DreamTexture.js