60-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: 230.9055
Accuracy: 0.855
Brier Loss: 0.2199
Nll: 1.0336
F1 Micro: 0.855
F1 Macro: 0.8458
Ece: 0.1335
Aurc: 0.0336

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	238.9631	0.285	0.8828	6.5847	0.285	0.1874	0.2964	0.5019
No log	2.0	100	237.5668	0.385	0.7452	3.0464	0.3850	0.2403	0.2776	0.4072
No log	3.0	150	236.0767	0.615	0.5834	2.4900	0.615	0.4713	0.2629	0.2189
No log	4.0	200	235.7085	0.635	0.5104	2.9259	0.635	0.4933	0.2605	0.1700
No log	5.0	250	234.8680	0.66	0.4494	2.2358	0.66	0.4974	0.2021	0.1292
No log	6.0	300	235.0645	0.68	0.4615	2.3164	0.68	0.5568	0.2601	0.1346
No log	7.0	350	234.3820	0.735	0.3741	1.8624	0.735	0.6245	0.2265	0.0894
No log	8.0	400	233.9906	0.72	0.3954	1.7623	0.72	0.6479	0.1779	0.0980
No log	9.0	450	234.0194	0.755	0.3410	1.6963	0.755	0.6965	0.1730	0.0718
234.3621	10.0	500	234.2365	0.705	0.3859	2.4174	0.705	0.6957	0.1759	0.0944
234.3621	11.0	550	233.5580	0.76	0.3331	1.5730	0.76	0.7135	0.1817	0.0709
234.3621	12.0	600	233.3485	0.815	0.2725	1.6179	0.815	0.7769	0.1835	0.0497
234.3621	13.0	650	233.6395	0.805	0.3038	1.7563	0.805	0.7813	0.1976	0.0698
234.3621	14.0	700	233.1443	0.805	0.2928	1.4722	0.805	0.7779	0.1513	0.0546
234.3621	15.0	750	233.1237	0.83	0.2569	1.7889	0.83	0.8231	0.1361	0.0460
234.3621	16.0	800	232.9007	0.825	0.2675	1.7492	0.825	0.8125	0.1742	0.0558
234.3621	17.0	850	233.0284	0.79	0.2861	1.4243	0.79	0.7698	0.1418	0.0520
234.3621	18.0	900	232.9831	0.79	0.3009	1.4085	0.79	0.7440	0.1962	0.0643
234.3621	19.0	950	232.9837	0.825	0.2618	1.4839	0.825	0.7949	0.1444	0.0457
231.5126	20.0	1000	232.9143	0.825	0.2599	1.5299	0.825	0.8086	0.1570	0.0463
231.5126	21.0	1050	232.5251	0.835	0.2495	1.2311	0.835	0.8279	0.1511	0.0472
231.5126	22.0	1100	232.6748	0.855	0.2165	1.2547	0.855	0.8448	0.1450	0.0299
231.5126	23.0	1150	232.6610	0.83	0.2450	1.2944	0.83	0.8113	0.1545	0.0403
231.5126	24.0	1200	232.7660	0.83	0.2480	1.4783	0.83	0.8101	0.1645	0.0397
231.5126	25.0	1250	232.5843	0.855	0.2336	1.1227	0.855	0.8303	0.1672	0.0397
231.5126	26.0	1300	232.3482	0.84	0.2321	1.1720	0.8400	0.8346	0.1527	0.0380
231.5126	27.0	1350	232.3758	0.84	0.2353	1.1684	0.8400	0.8327	0.1520	0.0372
231.5126	28.0	1400	232.3022	0.82	0.2460	1.0888	0.82	0.7920	0.1746	0.0392
231.5126	29.0	1450	232.1077	0.845	0.2355	1.3720	0.845	0.8306	0.1314	0.0372
230.5624	30.0	1500	232.3631	0.825	0.2443	1.3202	0.825	0.7971	0.1404	0.0373
230.5624	31.0	1550	232.1099	0.84	0.2509	1.2935	0.8400	0.8199	0.1552	0.0453
230.5624	32.0	1600	232.1548	0.855	0.2275	1.0948	0.855	0.8360	0.1428	0.0347
230.5624	33.0	1650	232.0168	0.84	0.2284	1.1303	0.8400	0.8260	0.1460	0.0374
230.5624	34.0	1700	232.1122	0.875	0.2154	1.3044	0.875	0.8612	0.1410	0.0307
230.5624	35.0	1750	232.0452	0.84	0.2382	1.1732	0.8400	0.8199	0.1218	0.0376
230.5624	36.0	1800	231.8549	0.835	0.2390	1.1322	0.835	0.8227	0.1478	0.0375
230.5624	37.0	1850	231.8938	0.845	0.2558	1.0115	0.845	0.8261	0.1730	0.0447
230.5624	38.0	1900	231.8979	0.85	0.2292	1.1142	0.85	0.8361	0.1359	0.0348
230.5624	39.0	1950	231.7455	0.84	0.2217	1.0746	0.8400	0.8185	0.1294	0.0309
229.9089	40.0	2000	231.8694	0.85	0.2200	1.0529	0.85	0.8334	0.1435	0.0343
229.9089	41.0	2050	231.8044	0.84	0.2317	0.9332	0.8400	0.8204	0.1432	0.0337
229.9089	42.0	2100	231.6414	0.855	0.2165	1.1356	0.855	0.8393	0.1357	0.0321
229.9089	43.0	2150	231.5806	0.835	0.2378	1.0314	0.835	0.8124	0.1443	0.0382
229.9089	44.0	2200	231.6199	0.855	0.2287	1.0907	0.855	0.8463	0.1196	0.0362
229.9089	45.0	2250	231.5991	0.85	0.2208	1.0967	0.85	0.8350	0.1321	0.0339
229.9089	46.0	2300	231.5103	0.85	0.2249	1.0330	0.85	0.8270	0.1239	0.0332
229.9089	47.0	2350	231.4252	0.87	0.2126	1.1054	0.87	0.8618	0.1230	0.0312
229.9089	48.0	2400	231.4696	0.86	0.2136	1.0952	0.8600	0.8503	0.1304	0.0302
229.9089	49.0	2450	231.5416	0.84	0.2329	1.0155	0.8400	0.8256	0.1381	0.0356
229.4364	50.0	2500	231.4932	0.84	0.2215	1.1382	0.8400	0.8177	0.1557	0.0319
229.4364	51.0	2550	231.4270	0.84	0.2312	1.0191	0.8400	0.8253	0.1376	0.0371
229.4364	52.0	2600	231.3520	0.85	0.2233	1.2815	0.85	0.8289	0.1444	0.0334
229.4364	53.0	2650	231.3922	0.86	0.2223	1.0950	0.8600	0.8423	0.1309	0.0334
229.4364	54.0	2700	231.3504	0.855	0.2171	1.0165	0.855	0.8422	0.1304	0.0313
229.4364	55.0	2750	231.2676	0.87	0.2129	1.0704	0.87	0.8598	0.1334	0.0320
229.4364	56.0	2800	231.2823	0.84	0.2390	1.0982	0.8400	0.8226	0.1187	0.0372
229.4364	57.0	2850	231.2740	0.85	0.2251	0.9521	0.85	0.8271	0.1388	0.0320
229.4364	58.0	2900	231.2784	0.85	0.2284	1.0194	0.85	0.8320	0.1306	0.0360
229.4364	59.0	2950	231.2078	0.84	0.2265	1.0036	0.8400	0.8253	0.1359	0.0366
229.043	60.0	3000	231.2086	0.85	0.2380	1.0247	0.85	0.8391	0.1395	0.0374
229.043	61.0	3050	231.2673	0.845	0.2410	1.0205	0.845	0.8272	0.1466	0.0389
229.043	62.0	3100	231.1900	0.855	0.2219	1.0835	0.855	0.8449	0.1351	0.0346
229.043	63.0	3150	231.0561	0.845	0.2417	0.9740	0.845	0.8332	0.1503	0.0405
229.043	64.0	3200	231.1282	0.845	0.2387	1.1105	0.845	0.8270	0.1198	0.0379
229.043	65.0	3250	231.0782	0.85	0.2334	0.9838	0.85	0.8403	0.1282	0.0356
229.043	66.0	3300	231.0704	0.84	0.2442	1.0380	0.8400	0.8275	0.1543	0.0410
229.043	67.0	3350	231.0450	0.85	0.2246	1.0023	0.85	0.8394	0.1212	0.0353
229.043	68.0	3400	231.1017	0.85	0.2257	1.0437	0.85	0.8334	0.1232	0.0350
229.043	69.0	3450	231.0068	0.85	0.2321	1.0075	0.85	0.8445	0.1271	0.0361
228.7748	70.0	3500	231.0666	0.85	0.2274	1.0133	0.85	0.8382	0.1334	0.0361
228.7748	71.0	3550	230.9450	0.85	0.2417	1.0738	0.85	0.8356	0.1294	0.0380
228.7748	72.0	3600	230.7952	0.85	0.2379	0.9779	0.85	0.8391	0.1309	0.0393
228.7748	73.0	3650	231.0920	0.86	0.2188	1.0154	0.8600	0.8538	0.1230	0.0335
228.7748	74.0	3700	230.9152	0.855	0.2408	1.1637	0.855	0.8486	0.1490	0.0400
228.7748	75.0	3750	230.9537	0.85	0.2195	1.0135	0.85	0.8301	0.1131	0.0321
228.7748	76.0	3800	230.9977	0.855	0.2208	1.0136	0.855	0.8484	0.1296	0.0334
228.7748	77.0	3850	230.9619	0.855	0.2348	1.0158	0.855	0.8526	0.1346	0.0371
228.7748	78.0	3900	230.9416	0.84	0.2315	1.0372	0.8400	0.8219	0.1290	0.0353
228.7748	79.0	3950	231.0093	0.85	0.2196	1.0981	0.85	0.8318	0.1380	0.0335
228.5779	80.0	4000	230.9455	0.845	0.2290	1.0193	0.845	0.8332	0.1459	0.0350
228.5779	81.0	4050	230.8672	0.845	0.2184	1.0164	0.845	0.8309	0.1560	0.0322
228.5779	82.0	4100	230.9410	0.855	0.2282	1.0116	0.855	0.8486	0.1309	0.0349
228.5779	83.0	4150	230.9393	0.855	0.2258	1.0168	0.855	0.8498	0.1227	0.0355
228.5779	84.0	4200	230.8770	0.845	0.2204	1.0105	0.845	0.8303	0.1259	0.0315
228.5779	85.0	4250	230.9236	0.845	0.2271	1.0079	0.845	0.8332	0.1274	0.0353
228.5779	86.0	4300	230.9332	0.845	0.2237	1.0410	0.845	0.8288	0.1480	0.0340
228.5779	87.0	4350	230.8607	0.845	0.2273	1.0343	0.845	0.8313	0.1070	0.0338
228.5779	88.0	4400	230.7832	0.85	0.2434	1.1399	0.85	0.8432	0.1483	0.0407
228.5779	89.0	4450	230.8136	0.85	0.2263	1.0002	0.85	0.8367	0.1283	0.0343
228.4285	90.0	4500	230.9154	0.86	0.2243	1.0189	0.8600	0.8552	0.1319	0.0332
228.4285	91.0	4550	230.8125	0.86	0.2320	0.9955	0.8600	0.8552	0.1437	0.0361
228.4285	92.0	4600	230.8634	0.855	0.2238	1.0341	0.855	0.8477	0.1252	0.0329
228.4285	93.0	4650	230.8527	0.84	0.2326	0.9819	0.8400	0.8242	0.1248	0.0362
228.4285	94.0	4700	230.8176	0.845	0.2347	1.0080	0.845	0.8332	0.1356	0.0369
228.4285	95.0	4750	230.7853	0.85	0.2286	1.0163	0.85	0.8445	0.1176	0.0356
228.4285	96.0	4800	230.9345	0.855	0.2218	1.0151	0.855	0.8458	0.1252	0.0328
228.4285	97.0	4850	230.8020	0.855	0.2300	1.0166	0.855	0.8462	0.1424	0.0349
228.4285	98.0	4900	230.8873	0.855	0.2240	1.0244	0.855	0.8477	0.1309	0.0343
228.4285	99.0	4950	230.8796	0.855	0.2276	1.0084	0.855	0.8498	0.1254	0.0344
228.3542	100.0	5000	230.9055	0.855	0.2199	1.0336	0.855	0.8458	0.1335	0.0336

Framework versions

Transformers 4.26.1
Pytorch 1.13.1.post200
Datasets 2.9.0
Tokenizers 0.13.2

jordyvl/60-tiny_tobacco3482_og_simkd

60-tiny_tobacco3482_og_simkd

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions