6_e_200-tiny_tobacco3482_kd_CEKD_t5.0_a0.7

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: 0.6810
Accuracy: 0.795
Brier Loss: 0.3284
Nll: 1.4069
F1 Micro: 0.795
F1 Macro: 0.7636
Ece: 0.2215
Aurc: 0.0726

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: 1e-06
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	3.5123	0.19	1.2292	9.8836	0.19	0.0557	0.5319	0.8455
No log	2.0	50	3.3061	0.19	1.1746	9.5259	0.19	0.0557	0.4948	0.8502
No log	3.0	75	3.0163	0.18	1.1038	9.0069	0.18	0.0530	0.3959	0.8542
No log	4.0	100	2.7051	0.16	1.0425	8.3332	0.16	0.0685	0.3637	0.8603
No log	5.0	125	2.4101	0.145	0.9955	7.8875	0.145	0.0886	0.3276	0.8594
No log	6.0	150	2.1622	0.16	0.9581	7.4210	0.16	0.1094	0.3098	0.8418
No log	7.0	175	1.9665	0.17	0.9199	6.3502	0.17	0.1141	0.2939	0.8129
No log	8.0	200	1.8218	0.19	0.8852	4.6467	0.19	0.1580	0.2691	0.7908
No log	9.0	225	1.7041	0.205	0.8503	3.9915	0.205	0.1722	0.2746	0.7349
No log	10.0	250	1.6030	0.315	0.8144	3.8156	0.315	0.2716	0.2999	0.6443
No log	11.0	275	1.5078	0.39	0.7749	3.6634	0.39	0.3103	0.2992	0.5291
No log	12.0	300	1.4231	0.46	0.7351	3.5329	0.46	0.3550	0.3143	0.4339
No log	13.0	325	1.3487	0.465	0.6988	3.4534	0.465	0.3851	0.2929	0.3704
No log	14.0	350	1.2823	0.51	0.6657	3.2203	0.51	0.4164	0.2835	0.3215
No log	15.0	375	1.2282	0.54	0.6399	3.0431	0.54	0.4415	0.2943	0.2928
No log	16.0	400	1.1793	0.555	0.6170	2.8562	0.555	0.4502	0.2813	0.2676
No log	17.0	425	1.1466	0.565	0.6019	2.7630	0.565	0.4607	0.2450	0.2524
No log	18.0	450	1.1114	0.59	0.5832	2.7390	0.59	0.4721	0.2850	0.2289
No log	19.0	475	1.0835	0.6	0.5699	2.5928	0.6	0.5073	0.2760	0.2195
1.9045	20.0	500	1.0547	0.615	0.5539	2.6261	0.615	0.5273	0.2883	0.2044
1.9045	21.0	525	1.0294	0.625	0.5404	2.6118	0.625	0.5343	0.2703	0.1945
1.9045	22.0	550	1.0085	0.635	0.5300	2.4378	0.635	0.5381	0.2727	0.1842
1.9045	23.0	575	0.9900	0.64	0.5188	2.4290	0.64	0.5435	0.2781	0.1745
1.9045	24.0	600	0.9674	0.645	0.5071	2.3513	0.645	0.5527	0.2631	0.1647
1.9045	25.0	625	0.9522	0.64	0.4980	2.2815	0.64	0.5494	0.2687	0.1606
1.9045	26.0	650	0.9336	0.65	0.4883	2.2513	0.65	0.5603	0.2727	0.1515
1.9045	27.0	675	0.9175	0.665	0.4795	2.2466	0.665	0.5707	0.2848	0.1450
1.9045	28.0	700	0.9060	0.655	0.4731	2.2223	0.655	0.5655	0.2598	0.1426
1.9045	29.0	725	0.8924	0.67	0.4648	2.1571	0.67	0.5748	0.2504	0.1364
1.9045	30.0	750	0.8808	0.675	0.4580	2.1970	0.675	0.5804	0.2124	0.1302
1.9045	31.0	775	0.8698	0.675	0.4513	1.9818	0.675	0.5784	0.2413	0.1248
1.9045	32.0	800	0.8581	0.685	0.4451	2.0653	0.685	0.6062	0.2783	0.1221
1.9045	33.0	825	0.8493	0.68	0.4398	1.9229	0.68	0.5964	0.2430	0.1198
1.9045	34.0	850	0.8416	0.675	0.4351	1.9147	0.675	0.5901	0.2547	0.1181
1.9045	35.0	875	0.8329	0.69	0.4296	1.9727	0.69	0.6098	0.2498	0.1121
1.9045	36.0	900	0.8222	0.7	0.4234	1.8988	0.7	0.6185	0.2404	0.1084
1.9045	37.0	925	0.8178	0.69	0.4201	1.8900	0.69	0.6103	0.2338	0.1079
1.9045	38.0	950	0.8091	0.69	0.4153	1.9396	0.69	0.6100	0.2469	0.1058
1.9045	39.0	975	0.7992	0.705	0.4098	1.8177	0.705	0.6453	0.1971	0.1037
0.8325	40.0	1000	0.7962	0.715	0.4069	1.7962	0.715	0.6560	0.2474	0.1008
0.8325	41.0	1025	0.7890	0.715	0.4032	1.7233	0.715	0.6592	0.2417	0.1009
0.8325	42.0	1050	0.7842	0.715	0.3997	1.6669	0.715	0.6573	0.2441	0.1005
0.8325	43.0	1075	0.7788	0.71	0.3962	1.6468	0.7100	0.6515	0.2199	0.0998
0.8325	44.0	1100	0.7713	0.725	0.3918	1.6398	0.7250	0.6698	0.2363	0.0945
0.8325	45.0	1125	0.7687	0.725	0.3895	1.6397	0.7250	0.6677	0.2478	0.0939
0.8325	46.0	1150	0.7642	0.73	0.3867	1.6259	0.7300	0.6684	0.2538	0.0941
0.8325	47.0	1175	0.7571	0.74	0.3822	1.5033	0.74	0.6907	0.2269	0.0903
0.8325	48.0	1200	0.7547	0.74	0.3806	1.5595	0.74	0.6830	0.2448	0.0912
0.8325	49.0	1225	0.7516	0.75	0.3779	1.6264	0.75	0.6921	0.2489	0.0887
0.8325	50.0	1250	0.7477	0.75	0.3759	1.5568	0.75	0.7029	0.2388	0.0888
0.8325	51.0	1275	0.7431	0.755	0.3725	1.5037	0.755	0.6986	0.2254	0.0848
0.8325	52.0	1300	0.7418	0.755	0.3713	1.4951	0.755	0.7085	0.2261	0.0862
0.8325	53.0	1325	0.7360	0.77	0.3676	1.4881	0.7700	0.7241	0.2474	0.0825
0.8325	54.0	1350	0.7339	0.77	0.3665	1.5554	0.7700	0.7241	0.2646	0.0827
0.8325	55.0	1375	0.7294	0.775	0.3636	1.4885	0.775	0.7275	0.2283	0.0818
0.8325	56.0	1400	0.7265	0.78	0.3617	1.5387	0.78	0.7306	0.2416	0.0799
0.8325	57.0	1425	0.7247	0.77	0.3598	1.4382	0.7700	0.7241	0.2553	0.0806
0.8325	58.0	1450	0.7234	0.78	0.3589	1.4888	0.78	0.7306	0.2231	0.0796
0.8325	59.0	1475	0.7186	0.78	0.3565	1.5400	0.78	0.7306	0.2200	0.0790
0.618	60.0	1500	0.7174	0.78	0.3549	1.4823	0.78	0.7306	0.2340	0.0787
0.618	61.0	1525	0.7148	0.78	0.3534	1.4804	0.78	0.7306	0.2412	0.0785
0.618	62.0	1550	0.7122	0.785	0.3516	1.4334	0.785	0.7445	0.2353	0.0787
0.618	63.0	1575	0.7107	0.79	0.3501	1.4153	0.79	0.7516	0.2354	0.0777
0.618	64.0	1600	0.7091	0.78	0.3491	1.4698	0.78	0.7306	0.2324	0.0780
0.618	65.0	1625	0.7071	0.79	0.3481	1.4097	0.79	0.7516	0.2198	0.0785
0.618	66.0	1650	0.7047	0.785	0.3464	1.4088	0.785	0.7458	0.2325	0.0778
0.618	67.0	1675	0.7041	0.785	0.3457	1.4108	0.785	0.7458	0.2248	0.0781
0.618	68.0	1700	0.7025	0.79	0.3444	1.4145	0.79	0.7516	0.2195	0.0773
0.618	69.0	1725	0.7014	0.79	0.3436	1.4120	0.79	0.7516	0.2629	0.0771
0.618	70.0	1750	0.6992	0.785	0.3422	1.4046	0.785	0.7458	0.2294	0.0767
0.618	71.0	1775	0.6982	0.785	0.3412	1.4142	0.785	0.7458	0.2325	0.0761
0.618	72.0	1800	0.6954	0.79	0.3395	1.4009	0.79	0.7516	0.2253	0.0763
0.618	73.0	1825	0.6942	0.79	0.3389	1.3994	0.79	0.7559	0.2383	0.0763
0.618	74.0	1850	0.6937	0.785	0.3382	1.4061	0.785	0.7458	0.2213	0.0762
0.618	75.0	1875	0.6935	0.785	0.3378	1.4082	0.785	0.7458	0.2218	0.0762
0.618	76.0	1900	0.6910	0.795	0.3359	1.4098	0.795	0.7599	0.2689	0.0746
0.618	77.0	1925	0.6907	0.79	0.3356	1.4072	0.79	0.7541	0.2254	0.0741
0.618	78.0	1950	0.6896	0.795	0.3352	1.3996	0.795	0.7636	0.2226	0.0743
0.618	79.0	1975	0.6896	0.79	0.3349	1.4073	0.79	0.7541	0.2295	0.0742
0.516	80.0	2000	0.6874	0.79	0.3335	1.4089	0.79	0.7541	0.2287	0.0743
0.516	81.0	2025	0.6874	0.795	0.3333	1.3983	0.795	0.7636	0.2387	0.0742
0.516	82.0	2050	0.6867	0.795	0.3327	1.4098	0.795	0.7636	0.2162	0.0736
0.516	83.0	2075	0.6865	0.795	0.3323	1.4656	0.795	0.7636	0.2072	0.0738
0.516	84.0	2100	0.6857	0.795	0.3323	1.4107	0.795	0.7636	0.2138	0.0741
0.516	85.0	2125	0.6854	0.795	0.3316	1.4223	0.795	0.7636	0.2262	0.0732
0.516	86.0	2150	0.6846	0.795	0.3311	1.4138	0.795	0.7636	0.2224	0.0733
0.516	87.0	2175	0.6834	0.795	0.3302	1.4113	0.795	0.7636	0.2307	0.0731
0.516	88.0	2200	0.6831	0.795	0.3300	1.4088	0.795	0.7636	0.2256	0.0730
0.516	89.0	2225	0.6821	0.795	0.3295	1.4126	0.795	0.7636	0.2395	0.0728
0.516	90.0	2250	0.6821	0.795	0.3294	1.4123	0.795	0.7636	0.2237	0.0728
0.516	91.0	2275	0.6823	0.795	0.3294	1.4085	0.795	0.7636	0.2213	0.0728
0.516	92.0	2300	0.6819	0.795	0.3290	1.4105	0.795	0.7636	0.2332	0.0730
0.516	93.0	2325	0.6816	0.795	0.3289	1.4094	0.795	0.7636	0.2236	0.0729
0.516	94.0	2350	0.6812	0.795	0.3287	1.4092	0.795	0.7636	0.2235	0.0729
0.516	95.0	2375	0.6813	0.795	0.3286	1.4065	0.795	0.7636	0.2197	0.0727
0.516	96.0	2400	0.6811	0.795	0.3285	1.4079	0.795	0.7636	0.2247	0.0729
0.516	97.0	2425	0.6810	0.795	0.3284	1.4072	0.795	0.7636	0.2320	0.0729
0.516	98.0	2450	0.6810	0.795	0.3284	1.4062	0.795	0.7636	0.2148	0.0727
0.516	99.0	2475	0.6810	0.795	0.3284	1.4068	0.795	0.7636	0.2215	0.0726
0.4715	100.0	2500	0.6810	0.795	0.3284	1.4069	0.795	0.7636	0.2215	0.0726

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_t5.0_a0.7

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

NSDT 3DConvert

UnrealSynth

DreamTexture.js