dit-base_tobacco-small_tobacco3482_hint

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: 1.9099
• Accuracy: 0.85
• Brier Loss: 0.2772
• Nll: 1.4757
• F1 Micro: 0.85
• F1 Macro: 0.8366
• Ece: 0.1392
• Aurc: 0.0460

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	3.2233	0.44	0.7001	2.8339	0.44	0.3067	0.2724	0.3661
No log	2.0	100	2.3954	0.705	0.4016	1.5814	0.705	0.6657	0.2046	0.1093
No log	3.0	150	2.1938	0.735	0.3560	1.5685	0.735	0.7026	0.1879	0.0858
No log	4.0	200	2.0989	0.74	0.3533	1.5416	0.74	0.7058	0.2015	0.0896
No log	5.0	250	2.0203	0.795	0.3169	1.5407	0.795	0.7861	0.1773	0.0919
No log	6.0	300	2.1849	0.675	0.4531	1.6333	0.675	0.6701	0.2207	0.1166
No log	7.0	350	2.2223	0.745	0.4113	1.4333	0.745	0.7293	0.2045	0.0980
No log	8.0	400	2.1696	0.715	0.4221	1.6537	0.715	0.6723	0.2069	0.1040
No log	9.0	450	2.4443	0.735	0.4291	1.5392	0.735	0.7458	0.2236	0.1323
1.8536	10.0	500	2.0474	0.775	0.3649	1.6156	0.775	0.7528	0.1915	0.0844
1.8536	11.0	550	2.0046	0.81	0.3170	1.6225	0.81	0.7920	0.1547	0.0639
1.8536	12.0	600	2.4864	0.725	0.4602	1.5678	0.7250	0.7308	0.2415	0.1218
1.8536	13.0	650	1.8413	0.83	0.2698	1.6361	0.83	0.8117	0.1349	0.0674
1.8536	14.0	700	2.1304	0.815	0.3281	1.5685	0.815	0.7936	0.1715	0.0703
1.8536	15.0	750	2.5075	0.71	0.4652	1.9297	0.7100	0.6877	0.2281	0.1099
1.8536	16.0	800	2.4854	0.73	0.4462	1.5241	0.7300	0.7176	0.2282	0.1097
1.8536	17.0	850	2.1252	0.805	0.3210	1.5685	0.805	0.7907	0.1650	0.0804
1.8536	18.0	900	1.9249	0.86	0.2473	1.7031	0.8600	0.8689	0.1244	0.0528
1.8536	19.0	950	2.0943	0.835	0.2840	1.4696	0.835	0.8267	0.1439	0.0652
1.0941	20.0	1000	1.8548	0.845	0.2566	1.3059	0.845	0.8403	0.1333	0.0558
1.0941	21.0	1050	2.1487	0.805	0.3362	1.4556	0.805	0.8051	0.1665	0.0764
1.0941	22.0	1100	2.2147	0.81	0.3149	1.4884	0.81	0.8081	0.1710	0.0984
1.0941	23.0	1150	2.1111	0.84	0.2898	1.5426	0.8400	0.8410	0.1489	0.0848
1.0941	24.0	1200	2.2432	0.85	0.2884	1.7273	0.85	0.8482	0.1532	0.0765
1.0941	25.0	1250	2.3105	0.75	0.4190	1.4648	0.75	0.7396	0.2177	0.1074
1.0941	26.0	1300	2.0587	0.795	0.3444	1.6181	0.795	0.7960	0.1641	0.0799
1.0941	27.0	1350	2.4465	0.8	0.3517	2.0076	0.8000	0.7770	0.1731	0.0849
1.0941	28.0	1400	2.1351	0.825	0.3132	1.5650	0.825	0.8315	0.1631	0.0553
1.0941	29.0	1450	1.9746	0.86	0.2451	1.5908	0.8600	0.8374	0.1267	0.0537
0.9575	30.0	1500	2.0257	0.855	0.2737	1.6541	0.855	0.8121	0.1352	0.0480
0.9575	31.0	1550	1.9631	0.84	0.3037	1.7341	0.8400	0.8201	0.1515	0.0423
0.9575	32.0	1600	2.4215	0.785	0.3909	1.4042	0.785	0.7740	0.2018	0.0708
0.9575	33.0	1650	2.2159	0.795	0.3492	1.7639	0.795	0.7716	0.1721	0.0537
0.9575	34.0	1700	2.3363	0.82	0.3132	1.9858	0.82	0.7993	0.1610	0.0845
0.9575	35.0	1750	2.2187	0.84	0.2884	1.5376	0.8400	0.8182	0.1523	0.0803
0.9575	36.0	1800	2.3407	0.825	0.3206	1.8292	0.825	0.8028	0.1588	0.0719
0.9575	37.0	1850	2.4302	0.815	0.3353	1.7611	0.815	0.8091	0.1654	0.0920
0.9575	38.0	1900	2.3307	0.815	0.3269	1.8263	0.815	0.8043	0.1675	0.0876
0.9575	39.0	1950	2.2905	0.825	0.3217	1.7612	0.825	0.8116	0.1639	0.0841
0.8923	40.0	2000	2.2699	0.83	0.3225	1.7537	0.83	0.8186	0.1655	0.0792
0.8923	41.0	2050	2.2327	0.83	0.3179	1.7534	0.83	0.8186	0.1559	0.0764
0.8923	42.0	2100	2.2852	0.825	0.3230	1.6737	0.825	0.8150	0.1611	0.0760
0.8923	43.0	2150	2.2597	0.825	0.3221	1.6727	0.825	0.8147	0.1610	0.0734
0.8923	44.0	2200	2.2492	0.83	0.3176	1.6692	0.83	0.8169	0.1619	0.0720
0.8923	45.0	2250	2.2208	0.825	0.3182	1.6737	0.825	0.8124	0.1627	0.0707
0.8923	46.0	2300	2.2192	0.825	0.3209	1.6771	0.825	0.8121	0.1650	0.0712
0.8923	47.0	2350	2.2127	0.825	0.3198	1.6187	0.825	0.8124	0.1636	0.0684
0.8923	48.0	2400	2.2079	0.825	0.3208	1.6760	0.825	0.8121	0.1632	0.0707
0.8923	49.0	2450	2.1995	0.825	0.3187	1.5377	0.825	0.8124	0.1656	0.0702
0.8511	50.0	2500	2.1877	0.825	0.3158	1.6098	0.825	0.8124	0.1600	0.0690
0.8511	51.0	2550	2.1698	0.825	0.3167	1.5353	0.825	0.8124	0.1607	0.0695
0.8511	52.0	2600	2.1667	0.825	0.3133	1.5303	0.825	0.8121	0.1596	0.0680
0.8511	53.0	2650	2.1791	0.83	0.3170	1.5332	0.83	0.8149	0.1608	0.0690
0.8511	54.0	2700	2.1621	0.83	0.3148	1.5274	0.83	0.8146	0.1551	0.0693
0.8511	55.0	2750	2.1572	0.83	0.3119	1.5318	0.83	0.8149	0.1532	0.0680
0.8511	56.0	2800	2.1587	0.83	0.3100	1.5232	0.83	0.8148	0.1524	0.0712
0.8511	57.0	2850	2.1596	0.83	0.3101	1.5234	0.83	0.8146	0.1560	0.0696
0.8511	58.0	2900	2.1048	0.835	0.3047	1.5231	0.835	0.8189	0.1442	0.0676
0.8511	59.0	2950	2.4279	0.76	0.4096	1.4535	0.76	0.7538	0.2078	0.0731
0.8335	60.0	3000	2.2098	0.775	0.4036	1.4180	0.775	0.7565	0.2010	0.0870
0.8335	61.0	3050	2.0122	0.85	0.2596	1.5903	0.85	0.8272	0.1349	0.0779
0.8335	62.0	3100	2.2465	0.815	0.3311	1.6852	0.815	0.7899	0.1672	0.0658
0.8335	63.0	3150	2.1239	0.84	0.2963	1.6390	0.8400	0.8305	0.1458	0.0878
0.8335	64.0	3200	2.1931	0.82	0.3181	1.7037	0.82	0.8199	0.1654	0.0719
0.8335	65.0	3250	1.8262	0.855	0.2493	1.4845	0.855	0.8335	0.1297	0.0456
0.8335	66.0	3300	1.9467	0.845	0.2657	1.4217	0.845	0.8326	0.1361	0.0498
0.8335	67.0	3350	1.9371	0.85	0.2680	1.4175	0.85	0.8405	0.1293	0.0506
0.8335	68.0	3400	1.9172	0.85	0.2656	1.4203	0.85	0.8405	0.1331	0.0503
0.8335	69.0	3450	1.8872	0.845	0.2664	1.4327	0.845	0.8324	0.1360	0.0493
0.8281	70.0	3500	1.9045	0.845	0.2715	1.4920	0.845	0.8324	0.1377	0.0496
0.8281	71.0	3550	1.8954	0.845	0.2684	1.4919	0.845	0.8338	0.1385	0.0499
0.8281	72.0	3600	1.9222	0.85	0.2698	1.4870	0.85	0.8375	0.1356	0.0499
0.8281	73.0	3650	1.9004	0.845	0.2691	1.4912	0.845	0.8335	0.1377	0.0484
0.8281	74.0	3700	1.9168	0.85	0.2693	1.4903	0.85	0.8375	0.1338	0.0495
0.8281	75.0	3750	1.8970	0.85	0.2700	1.4908	0.85	0.8366	0.1416	0.0477
0.8281	76.0	3800	1.9089	0.85	0.2705	1.4867	0.85	0.8366	0.1373	0.0480
0.8281	77.0	3850	1.8902	0.85	0.2697	1.4896	0.85	0.8366	0.1407	0.0464
0.8281	78.0	3900	1.8889	0.85	0.2710	1.4882	0.85	0.8366	0.1421	0.0472
0.8281	79.0	3950	1.9080	0.85	0.2712	1.4876	0.85	0.8366	0.1345	0.0476
0.8047	80.0	4000	1.9011	0.85	0.2703	1.4864	0.85	0.8366	0.1373	0.0472
0.8047	81.0	4050	1.9112	0.85	0.2735	1.4867	0.85	0.8366	0.1379	0.0465
0.8047	82.0	4100	1.8850	0.85	0.2728	1.4872	0.85	0.8366	0.1419	0.0462
0.8047	83.0	4150	1.9074	0.85	0.2740	1.4862	0.85	0.8366	0.1369	0.0463
0.8047	84.0	4200	1.8804	0.85	0.2714	1.4818	0.85	0.8366	0.1376	0.0461
0.8047	85.0	4250	1.9092	0.85	0.2757	1.4825	0.85	0.8366	0.1437	0.0463
0.8047	86.0	4300	1.8985	0.85	0.2745	1.4827	0.85	0.8366	0.1390	0.0460
0.8047	87.0	4350	1.9091	0.85	0.2731	1.4808	0.85	0.8366	0.1403	0.0466
0.8047	88.0	4400	1.9037	0.85	0.2754	1.4836	0.85	0.8366	0.1383	0.0459
0.8047	89.0	4450	1.8950	0.85	0.2750	1.4798	0.85	0.8366	0.1386	0.0452
0.7971	90.0	4500	1.9115	0.85	0.2755	1.4785	0.85	0.8366	0.1387	0.0461
0.7971	91.0	4550	1.9061	0.85	0.2757	1.4791	0.85	0.8366	0.1451	0.0460
0.7971	92.0	4600	1.9058	0.85	0.2757	1.4785	0.85	0.8366	0.1392	0.0464
0.7971	93.0	4650	1.9128	0.85	0.2724	1.4769	0.85	0.8366	0.1341	0.0468
0.7971	94.0	4700	1.9115	0.85	0.2770	1.4771	0.85	0.8366	0.1388	0.0463
0.7971	95.0	4750	1.9097	0.85	0.2761	1.4761	0.85	0.8366	0.1382	0.0462
0.7971	96.0	4800	1.9025	0.85	0.2761	1.4759	0.85	0.8366	0.1385	0.0460
0.7971	97.0	4850	1.9153	0.85	0.2775	1.4757	0.85	0.8366	0.1394	0.0463
0.7971	98.0	4900	1.9084	0.85	0.2765	1.4755	0.85	0.8366	0.1388	0.0460
0.7971	99.0	4950	1.9087	0.85	0.2772	1.4757	0.85	0.8366	0.1392	0.0460
0.7931	100.0	5000	1.9099	0.85	0.2772	1.4757	0.85	0.8366	0.1392	0.0460

Framework versions

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