vit-tiny_tobacco3482_hint_

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: 56.8340
Accuracy: 0.85
Brier Loss: 0.2391
Nll: 1.2700
F1 Micro: 0.85
F1 Macro: 0.8414
Ece: 0.1228
Aurc: 0.0436

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	59.9929	0.25	0.8706	4.9000	0.25	0.1856	0.3000	0.7819
No log	2.0	50	59.2119	0.54	0.5951	2.8481	0.54	0.4540	0.2868	0.2585
No log	3.0	75	58.7608	0.675	0.4417	1.6161	0.675	0.6086	0.2356	0.1335
No log	4.0	100	58.5280	0.75	0.3897	1.7733	0.75	0.7378	0.2211	0.1173
No log	5.0	125	58.2236	0.81	0.3119	1.5196	0.81	0.7653	0.1573	0.0769
No log	6.0	150	58.0804	0.805	0.3048	1.6467	0.805	0.7766	0.1355	0.0599
No log	7.0	175	58.3912	0.645	0.5320	2.1424	0.645	0.6017	0.2645	0.1497
No log	8.0	200	57.7380	0.755	0.3634	1.4581	0.755	0.7173	0.1913	0.0920
No log	9.0	225	57.5060	0.805	0.2939	1.4346	0.805	0.7860	0.1603	0.0636
No log	10.0	250	57.5367	0.805	0.3345	1.7441	0.805	0.7872	0.1748	0.0727
No log	11.0	275	57.7193	0.725	0.4195	1.7223	0.7250	0.6706	0.2121	0.0984
No log	12.0	300	57.5447	0.785	0.3435	1.7574	0.785	0.7772	0.1667	0.0678
No log	13.0	325	57.4991	0.795	0.3335	1.5337	0.795	0.7724	0.1685	0.0895
No log	14.0	350	57.3536	0.835	0.2942	1.5660	0.835	0.8089	0.1451	0.0594
No log	15.0	375	57.5013	0.815	0.3096	1.7294	0.815	0.7978	0.1579	0.0696
No log	16.0	400	57.3610	0.835	0.2763	1.7078	0.835	0.8182	0.1354	0.0526
No log	17.0	425	57.3415	0.825	0.3002	1.3895	0.825	0.8249	0.1455	0.0651
No log	18.0	450	57.4573	0.79	0.3649	1.6263	0.79	0.7688	0.1936	0.0943
No log	19.0	475	57.5728	0.8	0.3261	1.7616	0.8000	0.7461	0.1669	0.0659
57.5035	20.0	500	57.3014	0.835	0.2982	1.3501	0.835	0.8175	0.1442	0.0549
57.5035	21.0	525	57.3792	0.82	0.3206	1.4779	0.82	0.7957	0.1698	0.0568
57.5035	22.0	550	57.6968	0.795	0.3627	1.8539	0.795	0.7468	0.1809	0.0698
57.5035	23.0	575	57.4539	0.8	0.3465	1.4210	0.8000	0.7787	0.1736	0.0776
57.5035	24.0	600	57.1479	0.815	0.2998	1.3434	0.815	0.7989	0.1513	0.0607
57.5035	25.0	625	57.3390	0.79	0.3482	1.2595	0.79	0.7803	0.1635	0.0652
57.5035	26.0	650	57.4046	0.82	0.3130	1.5825	0.82	0.8060	0.1495	0.0692
57.5035	27.0	675	57.1398	0.835	0.2643	1.5441	0.835	0.8185	0.1317	0.0485
57.5035	28.0	700	57.4322	0.82	0.3217	1.5260	0.82	0.7912	0.1576	0.0643
57.5035	29.0	725	57.1362	0.87	0.2291	1.3903	0.87	0.8621	0.1104	0.0539
57.5035	30.0	750	57.1131	0.855	0.2511	1.1423	0.855	0.8451	0.1292	0.0492
57.5035	31.0	775	56.8690	0.845	0.2454	1.2822	0.845	0.8265	0.1254	0.0473
57.5035	32.0	800	56.8384	0.855	0.2220	1.2243	0.855	0.8453	0.1119	0.0382
57.5035	33.0	825	56.9461	0.855	0.2450	1.2192	0.855	0.8537	0.1276	0.0395
57.5035	34.0	850	56.9061	0.85	0.2450	1.2097	0.85	0.8337	0.1216	0.0408
57.5035	35.0	875	56.9100	0.86	0.2413	1.2197	0.8600	0.8546	0.1189	0.0387
57.5035	36.0	900	56.9087	0.855	0.2444	1.2098	0.855	0.8460	0.1281	0.0379
57.5035	37.0	925	56.8923	0.86	0.2438	1.2156	0.8600	0.8509	0.1259	0.0378
57.5035	38.0	950	56.8908	0.85	0.2450	1.2187	0.85	0.8397	0.1236	0.0390
57.5035	39.0	975	56.8591	0.855	0.2404	1.2063	0.855	0.8411	0.1326	0.0383
56.2493	40.0	1000	56.8479	0.86	0.2352	1.2142	0.8600	0.8539	0.1257	0.0386
56.2493	41.0	1025	56.8762	0.86	0.2365	1.2166	0.8600	0.8492	0.1171	0.0383
56.2493	42.0	1050	56.8551	0.865	0.2321	1.2078	0.865	0.8547	0.1183	0.0383
56.2493	43.0	1075	56.8913	0.87	0.2360	1.2013	0.87	0.8617	0.1187	0.0410
56.2493	44.0	1100	56.8523	0.86	0.2359	1.2094	0.8600	0.8506	0.1214	0.0399
56.2493	45.0	1125	56.8546	0.87	0.2330	1.2028	0.87	0.8650	0.1233	0.0384
56.2493	46.0	1150	56.8536	0.865	0.2323	1.2099	0.865	0.8579	0.1230	0.0389
56.2493	47.0	1175	56.8490	0.865	0.2346	1.2095	0.865	0.8580	0.1139	0.0411
56.2493	48.0	1200	56.8569	0.88	0.2295	1.1277	0.88	0.8693	0.1238	0.0381
56.2493	49.0	1225	56.8528	0.875	0.2292	1.1966	0.875	0.8681	0.1264	0.0394
56.2493	50.0	1250	56.8462	0.875	0.2309	1.1215	0.875	0.8690	0.1178	0.0424
56.2493	51.0	1275	56.8438	0.87	0.2285	1.1259	0.87	0.8631	0.1304	0.0404
56.2493	52.0	1300	56.8660	0.865	0.2334	1.1231	0.865	0.8623	0.1343	0.0416
56.2493	53.0	1325	56.8802	0.885	0.2273	1.1220	0.885	0.8784	0.1202	0.0415
56.2493	54.0	1350	56.8581	0.885	0.2261	1.1938	0.885	0.8772	0.1286	0.0408
56.2493	55.0	1375	56.8348	0.875	0.2301	1.1953	0.875	0.8675	0.1267	0.0404
56.2493	56.0	1400	56.8489	0.87	0.2336	1.1974	0.87	0.8589	0.1079	0.0393
56.2493	57.0	1425	56.8508	0.87	0.2307	1.1928	0.87	0.8648	0.1143	0.0399
56.2493	58.0	1450	56.8172	0.875	0.2198	1.2030	0.875	0.8675	0.1062	0.0378
56.2493	59.0	1475	56.8662	0.87	0.2287	1.1354	0.87	0.8659	0.1198	0.0410
56.0552	60.0	1500	56.8437	0.88	0.2241	1.1968	0.88	0.8735	0.1127	0.0407
56.0552	61.0	1525	56.8468	0.87	0.2303	1.1998	0.87	0.8648	0.1170	0.0428
56.0552	62.0	1550	56.8575	0.865	0.2299	1.1940	0.865	0.8599	0.1171	0.0411
56.0552	63.0	1575	56.8451	0.875	0.2291	1.1981	0.875	0.8697	0.1225	0.0395
56.0552	64.0	1600	56.8355	0.875	0.2255	1.2009	0.875	0.8675	0.1137	0.0394
56.0552	65.0	1625	56.8490	0.875	0.2338	1.2653	0.875	0.8697	0.1259	0.0413
56.0552	66.0	1650	56.8206	0.87	0.2271	1.2860	0.87	0.8603	0.1197	0.0379
56.0552	67.0	1675	56.8592	0.87	0.2307	1.1992	0.87	0.8636	0.1210	0.0408
56.0552	68.0	1700	56.8339	0.865	0.2290	1.1968	0.865	0.8599	0.1101	0.0386
56.0552	69.0	1725	56.8533	0.87	0.2336	1.2625	0.87	0.8636	0.1154	0.0406
56.0552	70.0	1750	56.8281	0.87	0.2328	1.2012	0.87	0.8659	0.1229	0.0406
56.0552	71.0	1775	56.8431	0.875	0.2335	1.2557	0.875	0.8697	0.1318	0.0398
56.0552	72.0	1800	56.8396	0.865	0.2321	1.2659	0.865	0.8599	0.1165	0.0397
56.0552	73.0	1825	56.8229	0.86	0.2302	1.2615	0.8600	0.8524	0.1258	0.0402
56.0552	74.0	1850	56.8445	0.87	0.2344	1.2371	0.87	0.8659	0.1202	0.0393
56.0552	75.0	1875	56.8475	0.865	0.2341	1.2660	0.865	0.8599	0.1202	0.0423
56.0552	76.0	1900	56.8338	0.86	0.2320	1.2643	0.8600	0.8524	0.1296	0.0422
56.0552	77.0	1925	56.8481	0.87	0.2353	1.2665	0.87	0.8659	0.1266	0.0426
56.0552	78.0	1950	56.8328	0.865	0.2323	1.2584	0.865	0.8599	0.1128	0.0424
56.0552	79.0	1975	56.8382	0.86	0.2363	1.2658	0.8600	0.8553	0.1273	0.0425
55.9822	80.0	2000	56.8260	0.86	0.2354	1.2710	0.8600	0.8553	0.1129	0.0430
55.9822	81.0	2025	56.8474	0.86	0.2398	1.2679	0.8600	0.8553	0.1212	0.0433
55.9822	82.0	2050	56.8105	0.855	0.2320	1.2655	0.855	0.8478	0.1269	0.0423
55.9822	83.0	2075	56.8240	0.86	0.2347	1.2651	0.8600	0.8524	0.1312	0.0425
55.9822	84.0	2100	56.8350	0.86	0.2353	1.2690	0.8600	0.8553	0.1225	0.0435
55.9822	85.0	2125	56.8317	0.855	0.2371	1.2674	0.855	0.8478	0.1211	0.0433
55.9822	86.0	2150	56.8270	0.855	0.2364	1.2646	0.855	0.8478	0.1270	0.0433
55.9822	87.0	2175	56.8275	0.855	0.2359	1.2660	0.855	0.8478	0.1167	0.0423
55.9822	88.0	2200	56.8426	0.855	0.2385	1.2683	0.855	0.8478	0.1239	0.0428
55.9822	89.0	2225	56.8376	0.855	0.2368	1.2676	0.855	0.8478	0.1239	0.0426
55.9822	90.0	2250	56.8358	0.855	0.2382	1.2670	0.855	0.8451	0.1213	0.0435
55.9822	91.0	2275	56.8254	0.86	0.2374	1.2687	0.8600	0.8536	0.1308	0.0432
55.9822	92.0	2300	56.8269	0.855	0.2359	1.2684	0.855	0.8476	0.1223	0.0425
55.9822	93.0	2325	56.8324	0.85	0.2381	1.2708	0.85	0.8414	0.1224	0.0432
55.9822	94.0	2350	56.8344	0.85	0.2384	1.2682	0.85	0.8414	0.1222	0.0433
55.9822	95.0	2375	56.8344	0.85	0.2387	1.2708	0.85	0.8414	0.1228	0.0434
55.9822	96.0	2400	56.8342	0.85	0.2389	1.2687	0.85	0.8414	0.1222	0.0432
55.9822	97.0	2425	56.8339	0.85	0.2387	1.2713	0.85	0.8414	0.1222	0.0433
55.9822	98.0	2450	56.8342	0.85	0.2393	1.2708	0.85	0.8414	0.1228	0.0437
55.9822	99.0	2475	56.8325	0.85	0.2390	1.2703	0.85	0.8414	0.1228	0.0436
55.9517	100.0	2500	56.8340	0.85	0.2391	1.2700	0.85	0.8414	0.1228	0.0436

Framework versions

Transformers 4.26.1
Pytorch 1.13.1.post200
Datasets 2.9.0
Tokenizers 0.13.2

jordyvl/vit-tiny_tobacco3482_hint_

vit-tiny_tobacco3482_hint_

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions