39-tiny_tobacco3482_og_simkd

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: 239.0946
Accuracy: 0.835
Brier Loss: 0.3084
Nll: 1.2679
F1 Micro: 0.835
F1 Macro: 0.8047
Ece: 0.2467
Aurc: 0.0448

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: 16
eval_batch_size: 16
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	50	246.9268	0.22	0.8877	7.2845	0.22	0.1442	0.2508	0.5734
No log	2.0	100	245.5493	0.355	0.7851	4.2469	0.3550	0.2408	0.2761	0.4302
No log	3.0	150	244.0258	0.58	0.6569	2.7197	0.58	0.4698	0.3335	0.2529
No log	4.0	200	243.8010	0.615	0.5800	3.1495	0.615	0.4701	0.2865	0.1927
No log	5.0	250	242.9670	0.64	0.5193	2.2510	0.64	0.4843	0.2774	0.1631
No log	6.0	300	243.1274	0.62	0.5595	2.8430	0.62	0.4881	0.2980	0.1558
No log	7.0	350	242.1450	0.675	0.4620	2.2479	0.675	0.5389	0.2322	0.1266
No log	8.0	400	242.5806	0.63	0.5222	2.6597	0.63	0.5539	0.2649	0.1584
No log	9.0	450	241.9880	0.755	0.4313	1.8660	0.755	0.6923	0.2780	0.1028
241.4671	10.0	500	241.9459	0.73	0.4274	1.9762	0.7300	0.6550	0.2539	0.0843
241.4671	11.0	550	241.4543	0.74	0.4410	1.7736	0.74	0.6867	0.2589	0.1197
241.4671	12.0	600	241.5311	0.79	0.3750	1.6579	0.79	0.7193	0.2754	0.0582
241.4671	13.0	650	241.6211	0.77	0.4231	1.4441	0.7700	0.7350	0.3134	0.0761
241.4671	14.0	700	241.2984	0.795	0.3898	1.5615	0.795	0.7650	0.2879	0.0721
241.4671	15.0	750	241.2752	0.81	0.3641	1.5167	0.81	0.7692	0.2956	0.0515
241.4671	16.0	800	240.9899	0.78	0.3821	1.7296	0.78	0.7434	0.2935	0.0688
241.4671	17.0	850	241.3208	0.81	0.3850	1.6487	0.81	0.7628	0.2946	0.0657
241.4671	18.0	900	241.0431	0.82	0.3678	1.5667	0.82	0.7701	0.3109	0.0504
241.4671	19.0	950	240.9436	0.815	0.3615	1.1826	0.815	0.7624	0.3019	0.0557
238.6853	20.0	1000	240.7993	0.83	0.3475	1.2078	0.83	0.7887	0.3062	0.0541
238.6853	21.0	1050	240.6124	0.825	0.3577	1.1927	0.825	0.7956	0.2883	0.0560
238.6853	22.0	1100	240.7105	0.815	0.3535	1.4564	0.815	0.7550	0.2814	0.0553
238.6853	23.0	1150	240.6886	0.815	0.3584	1.1556	0.815	0.7815	0.2993	0.0564
238.6853	24.0	1200	240.5978	0.84	0.3451	1.1351	0.8400	0.8002	0.2937	0.0485
238.6853	25.0	1250	240.3825	0.815	0.3453	1.2408	0.815	0.7918	0.2853	0.0525
238.6853	26.0	1300	240.2633	0.83	0.3431	1.3299	0.83	0.7883	0.2715	0.0562
238.6853	27.0	1350	240.5535	0.82	0.3475	1.4406	0.82	0.7874	0.2897	0.0536
238.6853	28.0	1400	240.3554	0.835	0.3447	1.2483	0.835	0.7947	0.3078	0.0464
238.6853	29.0	1450	240.2271	0.82	0.3274	1.5224	0.82	0.7862	0.2691	0.0467
237.7411	30.0	1500	240.2261	0.825	0.3388	1.2921	0.825	0.7939	0.2913	0.0485
237.7411	31.0	1550	240.4772	0.83	0.3474	1.3063	0.83	0.7823	0.2995	0.0516
237.7411	32.0	1600	240.2594	0.85	0.3509	1.3993	0.85	0.8184	0.3142	0.0496
237.7411	33.0	1650	239.9501	0.805	0.3413	1.2018	0.805	0.7630	0.2619	0.0592
237.7411	34.0	1700	240.2149	0.86	0.3409	1.2633	0.8600	0.8490	0.3229	0.0467
237.7411	35.0	1750	240.0215	0.82	0.3264	1.1435	0.82	0.7705	0.2538	0.0499
237.7411	36.0	1800	239.9490	0.85	0.3204	1.4364	0.85	0.8147	0.2752	0.0432
237.7411	37.0	1850	239.9360	0.825	0.3262	1.1487	0.825	0.7921	0.2706	0.0482
237.7411	38.0	1900	240.0647	0.855	0.3322	1.1655	0.855	0.8339	0.3053	0.0460
237.7411	39.0	1950	239.7761	0.825	0.3237	1.2252	0.825	0.7862	0.2732	0.0499
237.1023	40.0	2000	239.9426	0.83	0.3220	1.2096	0.83	0.7906	0.2748	0.0457
237.1023	41.0	2050	239.8765	0.835	0.3241	1.3547	0.835	0.7929	0.2661	0.0462
237.1023	42.0	2100	239.8423	0.835	0.3329	1.2423	0.835	0.8000	0.2955	0.0527
237.1023	43.0	2150	239.6674	0.83	0.3215	1.1217	0.83	0.8069	0.2597	0.0486
237.1023	44.0	2200	239.7237	0.81	0.3206	1.1322	0.81	0.7724	0.2500	0.0473
237.1023	45.0	2250	239.7981	0.82	0.3263	1.4167	0.82	0.7754	0.2786	0.0485
237.1023	46.0	2300	239.4825	0.83	0.3149	1.2610	0.83	0.7879	0.2436	0.0487
237.1023	47.0	2350	239.6016	0.83	0.3195	1.2602	0.83	0.7903	0.2631	0.0529
237.1023	48.0	2400	239.6693	0.84	0.3248	1.2935	0.8400	0.7992	0.2581	0.0505
237.1023	49.0	2450	239.6932	0.845	0.3235	1.1698	0.845	0.8024	0.2600	0.0459
236.6301	50.0	2500	239.5297	0.845	0.3151	1.4464	0.845	0.8057	0.2775	0.0467
236.6301	51.0	2550	239.6262	0.83	0.3236	1.2198	0.83	0.7879	0.2680	0.0466
236.6301	52.0	2600	239.4326	0.855	0.3111	1.2320	0.855	0.8338	0.2785	0.0452
236.6301	53.0	2650	239.5918	0.825	0.3171	1.2295	0.825	0.7966	0.2517	0.0440
236.6301	54.0	2700	239.5557	0.845	0.3215	1.1543	0.845	0.8027	0.2723	0.0453
236.6301	55.0	2750	239.4767	0.83	0.3113	1.0283	0.83	0.8107	0.2457	0.0446
236.6301	56.0	2800	239.3879	0.85	0.3112	1.2353	0.85	0.8146	0.2547	0.0449
236.6301	57.0	2850	239.3733	0.85	0.3104	1.3730	0.85	0.8236	0.2593	0.0451
236.6301	58.0	2900	239.4988	0.835	0.3174	1.3295	0.835	0.7984	0.2572	0.0468
236.6301	59.0	2950	239.3514	0.815	0.3103	1.2168	0.815	0.7746	0.2382	0.0453
236.2404	60.0	3000	239.3031	0.835	0.3083	1.2398	0.835	0.7944	0.2577	0.0479
236.2404	61.0	3050	239.3242	0.855	0.3095	1.2921	0.855	0.8307	0.2739	0.0438
236.2404	62.0	3100	239.3217	0.815	0.3135	1.2385	0.815	0.7799	0.2559	0.0481
236.2404	63.0	3150	239.2695	0.835	0.3102	1.0724	0.835	0.8068	0.2276	0.0471
236.2404	64.0	3200	239.3596	0.83	0.3124	1.1534	0.83	0.7911	0.2440	0.0442
236.2404	65.0	3250	239.2498	0.825	0.3097	1.2191	0.825	0.7804	0.2450	0.0459
236.2404	66.0	3300	239.1448	0.82	0.3020	1.0196	0.82	0.7825	0.2559	0.0452
236.2404	67.0	3350	239.1730	0.825	0.3017	1.1825	0.825	0.8065	0.2494	0.0444
236.2404	68.0	3400	239.3466	0.835	0.3159	1.2392	0.835	0.7973	0.2617	0.0448
236.2404	69.0	3450	239.1867	0.825	0.3058	1.1611	0.825	0.7940	0.2596	0.0450
235.9626	70.0	3500	239.2792	0.84	0.3106	1.1812	0.8400	0.8134	0.2590	0.0444
235.9626	71.0	3550	239.0603	0.82	0.3055	1.1710	0.82	0.7895	0.2251	0.0466
235.9626	72.0	3600	238.9947	0.815	0.3069	1.1766	0.815	0.7905	0.2315	0.0485
235.9626	73.0	3650	239.2322	0.83	0.3108	1.1724	0.83	0.8015	0.2595	0.0450
235.9626	74.0	3700	239.0134	0.825	0.3087	1.2661	0.825	0.7854	0.2352	0.0467
235.9626	75.0	3750	239.1055	0.825	0.3090	1.1748	0.825	0.7946	0.2483	0.0458
235.9626	76.0	3800	239.0925	0.825	0.3133	1.1843	0.825	0.7918	0.2466	0.0490
235.9626	77.0	3850	239.1586	0.835	0.3115	1.1877	0.835	0.8140	0.2498	0.0455
235.9626	78.0	3900	239.1394	0.83	0.3103	1.2698	0.83	0.7897	0.2424	0.0467
235.9626	79.0	3950	239.2314	0.83	0.3121	1.2519	0.83	0.7938	0.2378	0.0453
235.7667	80.0	4000	239.1433	0.83	0.3076	1.1725	0.83	0.7924	0.2412	0.0459
235.7667	81.0	4050	239.0533	0.83	0.3026	1.2551	0.83	0.7994	0.2661	0.0448
235.7667	82.0	4100	239.0847	0.825	0.3123	1.1798	0.825	0.7964	0.2611	0.0483
235.7667	83.0	4150	239.1199	0.835	0.3089	1.2642	0.835	0.8047	0.2551	0.0471
235.7667	84.0	4200	239.0148	0.835	0.3055	1.1264	0.835	0.7975	0.2341	0.0439
235.7667	85.0	4250	239.0657	0.825	0.3107	1.1645	0.825	0.7924	0.2363	0.0452
235.7667	86.0	4300	239.1023	0.83	0.3134	1.2785	0.83	0.7949	0.2353	0.0462
235.7667	87.0	4350	239.0670	0.83	0.3028	1.2849	0.83	0.7927	0.2295	0.0428
235.7667	88.0	4400	239.0111	0.82	0.3097	1.1759	0.82	0.7865	0.2447	0.0462
235.7667	89.0	4450	238.9677	0.845	0.2995	1.1189	0.845	0.8231	0.2402	0.0423
235.6149	90.0	4500	239.1255	0.83	0.3063	1.1282	0.83	0.8028	0.2695	0.0431
235.6149	91.0	4550	238.9880	0.815	0.3102	1.1766	0.815	0.7851	0.2388	0.0486
235.6149	92.0	4600	239.0212	0.84	0.3079	1.1941	0.8400	0.8121	0.2627	0.0442
235.6149	93.0	4650	238.9948	0.82	0.3077	1.1702	0.82	0.7874	0.2533	0.0468
235.6149	94.0	4700	239.0357	0.82	0.3103	1.1823	0.82	0.7878	0.2639	0.0471
235.6149	95.0	4750	238.9936	0.81	0.2995	1.1251	0.81	0.7738	0.2519	0.0447
235.6149	96.0	4800	239.1319	0.835	0.3138	1.1957	0.835	0.8045	0.2806	0.0455
235.6149	97.0	4850	239.0160	0.815	0.3061	1.1852	0.815	0.7756	0.2281	0.0453
235.6149	98.0	4900	239.1168	0.83	0.3152	1.2637	0.83	0.7997	0.2509	0.0474
235.6149	99.0	4950	239.0691	0.83	0.3085	1.1799	0.83	0.7998	0.2524	0.0439
235.5384	100.0	5000	239.0946	0.835	0.3084	1.2679	0.835	0.8047	0.2467	0.0448

Framework versions

Transformers 4.26.1
Pytorch 1.13.1.post200
Datasets 2.9.0
Tokenizers 0.13.2

jordyvl/39-tiny_tobacco3482_og_simkd

39-tiny_tobacco3482_og_simkd

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions