dit-base_tobacco-small_tobacco3482_og_simkd

This model is a fine-tuned version of WinKawaks/vit-small-patch16-224 on the None dataset. It achieves the following results on the evaluation set:

Loss: 309.8690
Accuracy: 0.815
Brier Loss: 0.3313
Nll: 1.1190
F1 Micro: 0.815
F1 Macro: 0.7825
Ece: 0.2569
Aurc: 0.0659

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	328.2257	0.365	0.8441	5.5835	0.3650	0.2447	0.3371	0.4390
No log	2.0	100	325.5961	0.58	0.6442	1.8499	0.58	0.4727	0.3233	0.2414
No log	3.0	150	323.3813	0.66	0.4759	1.5815	0.66	0.5348	0.2704	0.1489
No log	4.0	200	322.5013	0.715	0.4234	1.6240	0.715	0.6009	0.2382	0.1142
No log	5.0	250	321.7315	0.755	0.3532	1.2868	0.755	0.6596	0.2141	0.0687
No log	6.0	300	320.5884	0.775	0.3668	1.4106	0.775	0.7233	0.2284	0.0922
No log	7.0	350	320.8456	0.775	0.3638	1.4833	0.775	0.7172	0.2487	0.0666
No log	8.0	400	319.6829	0.785	0.3308	1.3914	0.785	0.7203	0.1959	0.0674
No log	9.0	450	319.7741	0.815	0.3459	1.3920	0.815	0.7832	0.2541	0.0681
325.907	10.0	500	319.4605	0.775	0.3162	1.2997	0.775	0.6987	0.2140	0.0575
325.907	11.0	550	318.6996	0.81	0.3190	1.2271	0.81	0.7670	0.2110	0.0614
325.907	12.0	600	318.0233	0.81	0.3183	1.2432	0.81	0.7673	0.2134	0.0624
325.907	13.0	650	318.2606	0.79	0.3259	1.2187	0.79	0.7457	0.2299	0.0591
325.907	14.0	700	317.7428	0.83	0.3183	1.3279	0.83	0.8035	0.2449	0.0512
325.907	15.0	750	317.7053	0.81	0.3251	1.2097	0.81	0.7604	0.2193	0.0566
325.907	16.0	800	317.3470	0.84	0.3142	1.2606	0.8400	0.8132	0.2272	0.0484
325.907	17.0	850	316.8029	0.815	0.3202	1.1571	0.815	0.7748	0.2316	0.0563
325.907	18.0	900	316.9777	0.805	0.3442	1.1453	0.805	0.7645	0.2432	0.0625
325.907	19.0	950	316.2359	0.815	0.3219	1.1399	0.815	0.7717	0.2404	0.0603
320.1296	20.0	1000	316.1051	0.8	0.3220	1.1807	0.8000	0.7500	0.2412	0.0576
320.1296	21.0	1050	315.8117	0.845	0.3099	1.0976	0.845	0.8084	0.2530	0.0547
320.1296	22.0	1100	315.7457	0.82	0.3238	1.1904	0.82	0.7663	0.2507	0.0548
320.1296	23.0	1150	315.6591	0.82	0.3357	1.4044	0.82	0.7925	0.2639	0.0586
320.1296	24.0	1200	315.4048	0.82	0.3270	1.0817	0.82	0.7681	0.2575	0.0629
320.1296	25.0	1250	314.9790	0.81	0.3309	1.2002	0.81	0.7732	0.2279	0.0656
320.1296	26.0	1300	314.6778	0.79	0.3189	1.1219	0.79	0.7464	0.2014	0.0576
320.1296	27.0	1350	314.7844	0.8	0.3345	1.0655	0.8000	0.7555	0.2398	0.0661
320.1296	28.0	1400	314.4464	0.815	0.3175	1.1116	0.815	0.7636	0.2426	0.0532
320.1296	29.0	1450	314.3737	0.845	0.3271	1.1042	0.845	0.8072	0.2595	0.0531
317.1926	30.0	1500	313.9464	0.82	0.3225	1.1270	0.82	0.7841	0.2087	0.0609
317.1926	31.0	1550	314.0068	0.835	0.3187	1.1834	0.835	0.8070	0.2470	0.0522
317.1926	32.0	1600	313.8198	0.81	0.3271	1.0324	0.81	0.7642	0.2484	0.0609
317.1926	33.0	1650	313.7599	0.83	0.3193	1.0993	0.83	0.7910	0.2382	0.0536
317.1926	34.0	1700	313.4889	0.82	0.3224	1.1743	0.82	0.7823	0.2587	0.0546
317.1926	35.0	1750	313.2496	0.825	0.3324	1.1041	0.825	0.7988	0.2404	0.0652
317.1926	36.0	1800	313.1823	0.83	0.3207	1.0900	0.83	0.8007	0.2505	0.0581
317.1926	37.0	1850	313.1304	0.83	0.3367	1.2073	0.83	0.7973	0.2615	0.0571
317.1926	38.0	1900	313.2971	0.815	0.3398	1.1045	0.815	0.7709	0.2411	0.0641
317.1926	39.0	1950	313.0526	0.815	0.3352	1.1023	0.815	0.7744	0.2455	0.0616
315.1897	40.0	2000	312.7858	0.84	0.3231	1.0983	0.8400	0.8096	0.2619	0.0538
315.1897	41.0	2050	312.5119	0.815	0.3290	1.1174	0.815	0.7858	0.2540	0.0604
315.1897	42.0	2100	312.5961	0.82	0.3305	1.2144	0.82	0.7787	0.2480	0.0572
315.1897	43.0	2150	312.3510	0.825	0.3357	1.1367	0.825	0.7936	0.2398	0.0658
315.1897	44.0	2200	312.4015	0.81	0.3303	1.1015	0.81	0.7837	0.2488	0.0598
315.1897	45.0	2250	312.2003	0.825	0.3286	1.1810	0.825	0.7953	0.2480	0.0614
315.1897	46.0	2300	312.1683	0.825	0.3283	1.1112	0.825	0.7881	0.2414	0.0587
315.1897	47.0	2350	312.2554	0.815	0.3433	1.1313	0.815	0.7709	0.2579	0.0694
315.1897	48.0	2400	312.0919	0.825	0.3364	1.1074	0.825	0.7963	0.2471	0.0636
315.1897	49.0	2450	312.0760	0.82	0.3412	1.1076	0.82	0.7859	0.2554	0.0661
313.7276	50.0	2500	311.7450	0.83	0.3245	1.1723	0.83	0.7994	0.2512	0.0558
313.7276	51.0	2550	311.5801	0.835	0.3236	1.1056	0.835	0.7954	0.2581	0.0576
313.7276	52.0	2600	311.7016	0.83	0.3235	1.1182	0.83	0.7988	0.2462	0.0560
313.7276	53.0	2650	311.0808	0.81	0.3308	1.0526	0.81	0.7716	0.2401	0.0687
313.7276	54.0	2700	311.3835	0.81	0.3304	1.1210	0.81	0.7803	0.2442	0.0604
313.7276	55.0	2750	311.1007	0.825	0.3285	1.1265	0.825	0.7931	0.2569	0.0639
313.7276	56.0	2800	311.3446	0.81	0.3273	1.1896	0.81	0.7810	0.2342	0.0622
313.7276	57.0	2850	311.0753	0.825	0.3327	1.1225	0.825	0.7929	0.2659	0.0668
313.7276	58.0	2900	311.3600	0.825	0.3320	1.1142	0.825	0.8000	0.2524	0.0640
313.7276	59.0	2950	310.8636	0.83	0.3242	1.1157	0.83	0.8022	0.2416	0.0633
312.6368	60.0	3000	310.7809	0.815	0.3386	1.2166	0.815	0.7820	0.2571	0.0702
312.6368	61.0	3050	310.9625	0.825	0.3273	1.1168	0.825	0.7923	0.2362	0.0608
312.6368	62.0	3100	311.1122	0.81	0.3369	1.1021	0.81	0.7700	0.2433	0.0633
312.6368	63.0	3150	311.1530	0.82	0.3351	1.1108	0.82	0.7780	0.2584	0.0615
312.6368	64.0	3200	310.9366	0.8	0.3288	1.1112	0.8000	0.7616	0.2545	0.0609
312.6368	65.0	3250	310.7639	0.82	0.3379	1.0992	0.82	0.7898	0.2407	0.0710
312.6368	66.0	3300	310.5876	0.81	0.3287	1.1197	0.81	0.7763	0.2270	0.0654
312.6368	67.0	3350	310.7344	0.805	0.3387	1.1279	0.805	0.7646	0.2354	0.0679
312.6368	68.0	3400	310.2750	0.825	0.3323	1.1367	0.825	0.7971	0.2514	0.0673
312.6368	69.0	3450	310.5080	0.815	0.3298	1.1049	0.815	0.7845	0.2329	0.0664
311.7616	70.0	3500	310.6353	0.81	0.3305	1.1098	0.81	0.7745	0.2346	0.0633
311.7616	71.0	3550	310.3249	0.825	0.3286	1.1117	0.825	0.7951	0.2455	0.0641
311.7616	72.0	3600	310.5689	0.825	0.3248	1.1079	0.825	0.7911	0.2388	0.0586
311.7616	73.0	3650	310.4175	0.82	0.3298	1.1169	0.82	0.7859	0.2338	0.0630
311.7616	74.0	3700	310.1338	0.815	0.3313	1.1236	0.815	0.7902	0.2558	0.0677
311.7616	75.0	3750	310.4428	0.825	0.3310	1.1269	0.825	0.7972	0.2458	0.0606
311.7616	76.0	3800	310.3477	0.81	0.3317	1.1060	0.81	0.7775	0.2392	0.0654
311.7616	77.0	3850	310.2144	0.815	0.3294	1.1076	0.815	0.7857	0.2387	0.0627
311.7616	78.0	3900	310.1073	0.82	0.3296	1.1246	0.82	0.7993	0.2496	0.0634
311.7616	79.0	3950	310.1449	0.805	0.3246	1.1134	0.805	0.7734	0.2277	0.0627
311.1587	80.0	4000	310.1684	0.81	0.3327	1.1094	0.81	0.7781	0.2493	0.0660
311.1587	81.0	4050	310.1772	0.815	0.3311	1.1129	0.815	0.7876	0.2447	0.0668
311.1587	82.0	4100	309.9326	0.805	0.3295	1.1172	0.805	0.7716	0.2508	0.0666
311.1587	83.0	4150	310.1067	0.805	0.3330	1.1209	0.805	0.7756	0.2252	0.0653
311.1587	84.0	4200	309.9362	0.825	0.3288	1.1150	0.825	0.8024	0.2500	0.0637
311.1587	85.0	4250	309.6593	0.81	0.3292	1.1226	0.81	0.7723	0.2306	0.0680
311.1587	86.0	4300	309.9828	0.8	0.3310	1.1252	0.8000	0.7643	0.2474	0.0662
311.1587	87.0	4350	310.0325	0.825	0.3322	1.1136	0.825	0.7935	0.2633	0.0634
311.1587	88.0	4400	309.8688	0.815	0.3320	1.1145	0.815	0.7824	0.2478	0.0675
311.1587	89.0	4450	310.0577	0.81	0.3324	1.1160	0.81	0.7810	0.2475	0.0648
310.732	90.0	4500	309.8999	0.81	0.3273	1.1120	0.81	0.7720	0.2356	0.0624
310.732	91.0	4550	309.7399	0.815	0.3256	1.1164	0.815	0.7824	0.2502	0.0649
310.732	92.0	4600	309.9419	0.805	0.3287	1.1183	0.805	0.7751	0.2353	0.0640
310.732	93.0	4650	309.9055	0.81	0.3268	1.1194	0.81	0.7761	0.2429	0.0613
310.732	94.0	4700	309.7320	0.82	0.3275	1.1117	0.82	0.7914	0.2408	0.0654
310.732	95.0	4750	309.9635	0.81	0.3334	1.1067	0.81	0.7747	0.2317	0.0637
310.732	96.0	4800	309.9630	0.805	0.3304	1.1165	0.805	0.7712	0.2316	0.0631
310.732	97.0	4850	309.8564	0.815	0.3263	1.1130	0.815	0.7870	0.2355	0.0619
310.732	98.0	4900	309.7815	0.815	0.3298	1.1198	0.815	0.7857	0.2386	0.0634
310.732	99.0	4950	309.8337	0.81	0.3354	1.0806	0.81	0.7818	0.2480	0.0672
310.5225	100.0	5000	309.8690	0.815	0.3313	1.1190	0.815	0.7825	0.2569	0.0659

Framework versions

Transformers 4.36.0.dev0
Pytorch 2.2.0.dev20231112+cu118
Datasets 2.14.5
Tokenizers 0.14.1

jordyvl/dit-base_tobacco-small_tobacco3482_og_simkd

dit-base_tobacco-small_tobacco3482_og_simkd

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions