114-tiny_tobacco3482_kd

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: 0.1493
Accuracy: 0.71
Brier Loss: 0.5252
Nll: 1.5452
F1 Micro: 0.7100
F1 Macro: 0.6059
Ece: 0.3891
Aurc: 0.0888

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: 64
eval_batch_size: 64
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	13	0.9704	0.23	0.8929	7.3757	0.23	0.1649	0.2840	0.7790
No log	2.0	26	0.4936	0.34	0.8279	4.5780	0.34	0.2115	0.3215	0.5666
No log	3.0	39	0.3759	0.455	0.7568	3.7623	0.455	0.3073	0.3262	0.3769
No log	4.0	52	0.3165	0.545	0.7159	2.7078	0.545	0.4004	0.3944	0.2710
No log	5.0	65	0.2807	0.6	0.6537	2.7061	0.6	0.4506	0.3539	0.2091
No log	6.0	78	0.2714	0.575	0.6476	2.6202	0.575	0.4324	0.3576	0.2081
No log	7.0	91	0.2473	0.64	0.6163	2.4990	0.64	0.5080	0.3882	0.1616
No log	8.0	104	0.2752	0.625	0.5837	2.5795	0.625	0.5022	0.3291	0.1892
No log	9.0	117	0.3128	0.59	0.6027	2.7893	0.59	0.4859	0.2944	0.2296
No log	10.0	130	0.2292	0.66	0.5612	2.3152	0.66	0.5253	0.3822	0.1155
No log	11.0	143	0.2676	0.665	0.5632	2.6937	0.665	0.5479	0.3608	0.1422
No log	12.0	156	0.2512	0.65	0.5543	2.2519	0.65	0.5533	0.3324	0.1300
No log	13.0	169	0.2053	0.67	0.5555	1.9904	0.67	0.5739	0.3659	0.1162
No log	14.0	182	0.2281	0.68	0.5613	2.2343	0.68	0.5508	0.3683	0.1193
No log	15.0	195	0.2029	0.705	0.5511	1.6184	0.705	0.5984	0.4175	0.0937
No log	16.0	208	0.2090	0.71	0.5459	1.9750	0.7100	0.6052	0.3911	0.0983
No log	17.0	221	0.1828	0.705	0.5385	1.8272	0.705	0.5973	0.3700	0.0969
No log	18.0	234	0.1739	0.73	0.5358	1.6202	0.7300	0.6180	0.4115	0.0962
No log	19.0	247	0.1847	0.685	0.5300	2.1083	0.685	0.5717	0.3582	0.1047
No log	20.0	260	0.1839	0.69	0.5390	1.8560	0.69	0.5932	0.3708	0.1090
No log	21.0	273	0.1756	0.72	0.5417	1.7203	0.72	0.6132	0.4000	0.0855
No log	22.0	286	0.1727	0.69	0.5212	1.9503	0.69	0.5853	0.3574	0.1041
No log	23.0	299	0.1684	0.72	0.5333	1.5951	0.72	0.6229	0.3922	0.0943
No log	24.0	312	0.1652	0.735	0.5263	1.6768	0.735	0.6519	0.4001	0.0920
No log	25.0	325	0.1637	0.735	0.5363	1.6879	0.735	0.6514	0.4079	0.0819
No log	26.0	338	0.1609	0.675	0.5299	1.5660	0.675	0.5602	0.3593	0.0989
No log	27.0	351	0.1581	0.725	0.5210	1.5886	0.7250	0.6206	0.3739	0.0847
No log	28.0	364	0.1591	0.71	0.5286	1.7728	0.7100	0.6076	0.3868	0.0921
No log	29.0	377	0.1544	0.715	0.5251	1.6215	0.715	0.6201	0.3813	0.0948
No log	30.0	390	0.1618	0.705	0.5340	1.5824	0.705	0.6064	0.3853	0.1000
No log	31.0	403	0.1580	0.705	0.5202	1.7228	0.705	0.5949	0.3710	0.0963
No log	32.0	416	0.1531	0.72	0.5257	1.6330	0.72	0.6137	0.3857	0.0904
No log	33.0	429	0.1521	0.72	0.5248	1.6212	0.72	0.6349	0.3928	0.0898
No log	34.0	442	0.1526	0.71	0.5261	1.4652	0.7100	0.6141	0.3829	0.0905
No log	35.0	455	0.1529	0.7	0.5256	1.5784	0.7	0.5926	0.3885	0.0887
No log	36.0	468	0.1526	0.735	0.5268	1.5163	0.735	0.6514	0.3991	0.0878
No log	37.0	481	0.1497	0.695	0.5222	1.6068	0.695	0.5785	0.3918	0.0919
No log	38.0	494	0.1488	0.72	0.5248	1.5401	0.72	0.6115	0.3905	0.0891
0.1554	39.0	507	0.1504	0.715	0.5208	1.5917	0.715	0.6152	0.3730	0.0894
0.1554	40.0	520	0.1487	0.725	0.5260	1.5258	0.7250	0.6399	0.3998	0.0879
0.1554	41.0	533	0.1484	0.71	0.5250	1.6093	0.7100	0.6073	0.3908	0.0880
0.1554	42.0	546	0.1481	0.715	0.5245	1.5711	0.715	0.6096	0.3857	0.0860
0.1554	43.0	559	0.1493	0.705	0.5243	1.6261	0.705	0.6000	0.3727	0.0901
0.1554	44.0	572	0.1495	0.71	0.5242	1.5942	0.7100	0.6080	0.3808	0.0868
0.1554	45.0	585	0.1495	0.71	0.5242	1.5417	0.7100	0.6059	0.3813	0.0881
0.1554	46.0	598	0.1490	0.715	0.5239	1.5403	0.715	0.6134	0.3826	0.0893
0.1554	47.0	611	0.1486	0.715	0.5248	1.5387	0.715	0.6112	0.3754	0.0883
0.1554	48.0	624	0.1491	0.71	0.5252	1.5527	0.7100	0.6059	0.3761	0.0889
0.1554	49.0	637	0.1491	0.71	0.5249	1.5545	0.7100	0.6059	0.3880	0.0885
0.1554	50.0	650	0.1489	0.71	0.5247	1.5376	0.7100	0.6059	0.3900	0.0895
0.1554	51.0	663	0.1492	0.71	0.5257	1.5385	0.7100	0.6059	0.3857	0.0890
0.1554	52.0	676	0.1491	0.71	0.5251	1.5460	0.7100	0.6059	0.3816	0.0888
0.1554	53.0	689	0.1491	0.71	0.5248	1.5429	0.7100	0.6059	0.3806	0.0886
0.1554	54.0	702	0.1489	0.71	0.5247	1.5426	0.7100	0.6059	0.3949	0.0887
0.1554	55.0	715	0.1492	0.71	0.5258	1.5550	0.7100	0.6059	0.3921	0.0890
0.1554	56.0	728	0.1492	0.71	0.5248	1.5470	0.7100	0.6059	0.3859	0.0888
0.1554	57.0	741	0.1491	0.71	0.5251	1.5447	0.7100	0.6059	0.4035	0.0888
0.1554	58.0	754	0.1491	0.71	0.5248	1.5440	0.7100	0.6059	0.4033	0.0886
0.1554	59.0	767	0.1491	0.71	0.5246	1.5561	0.7100	0.6059	0.3920	0.0890
0.1554	60.0	780	0.1492	0.71	0.5251	1.5461	0.7100	0.6059	0.3847	0.0889
0.1554	61.0	793	0.1493	0.71	0.5251	1.5455	0.7100	0.6059	0.3931	0.0887
0.1554	62.0	806	0.1493	0.71	0.5252	1.5443	0.7100	0.6059	0.3912	0.0889
0.1554	63.0	819	0.1493	0.71	0.5253	1.5441	0.7100	0.6059	0.3944	0.0887
0.1554	64.0	832	0.1492	0.71	0.5249	1.5444	0.7100	0.6059	0.3891	0.0888
0.1554	65.0	845	0.1492	0.71	0.5255	1.5430	0.7100	0.6059	0.3995	0.0888
0.1554	66.0	858	0.1493	0.71	0.5250	1.5435	0.7100	0.6059	0.3991	0.0890
0.1554	67.0	871	0.1493	0.71	0.5252	1.5449	0.7100	0.6059	0.3991	0.0890
0.1554	68.0	884	0.1492	0.71	0.5251	1.5458	0.7100	0.6059	0.3968	0.0889
0.1554	69.0	897	0.1493	0.71	0.5250	1.5468	0.7100	0.6059	0.4036	0.0888
0.1554	70.0	910	0.1494	0.71	0.5253	1.5464	0.7100	0.6059	0.3889	0.0887
0.1554	71.0	923	0.1493	0.71	0.5251	1.5452	0.7100	0.6059	0.3888	0.0887
0.1554	72.0	936	0.1493	0.71	0.5250	1.5457	0.7100	0.6059	0.3928	0.0888
0.1554	73.0	949	0.1494	0.71	0.5253	1.5455	0.7100	0.6059	0.3946	0.0889
0.1554	74.0	962	0.1493	0.71	0.5251	1.5441	0.7100	0.6059	0.3928	0.0888
0.1554	75.0	975	0.1493	0.71	0.5252	1.5455	0.7100	0.6059	0.3929	0.0891
0.1554	76.0	988	0.1493	0.71	0.5252	1.5449	0.7100	0.6059	0.3940	0.0886
0.0002	77.0	1001	0.1493	0.71	0.5253	1.5455	0.7100	0.6059	0.3891	0.0887
0.0002	78.0	1014	0.1493	0.71	0.5251	1.5468	0.7100	0.6059	0.3889	0.0887
0.0002	79.0	1027	0.1494	0.71	0.5252	1.5462	0.7100	0.6059	0.3891	0.0888
0.0002	80.0	1040	0.1493	0.71	0.5252	1.5443	0.7100	0.6059	0.3994	0.0886
0.0002	81.0	1053	0.1493	0.71	0.5251	1.5451	0.7100	0.6059	0.3890	0.0887
0.0002	82.0	1066	0.1493	0.71	0.5251	1.5448	0.7100	0.6059	0.3938	0.0886
0.0002	83.0	1079	0.1493	0.71	0.5251	1.5453	0.7100	0.6059	0.3890	0.0886
0.0002	84.0	1092	0.1493	0.71	0.5252	1.5462	0.7100	0.6059	0.3890	0.0888
0.0002	85.0	1105	0.1493	0.71	0.5252	1.5454	0.7100	0.6059	0.3891	0.0887
0.0002	86.0	1118	0.1493	0.71	0.5252	1.5452	0.7100	0.6059	0.3890	0.0887
0.0002	87.0	1131	0.1494	0.71	0.5252	1.5452	0.7100	0.6059	0.3891	0.0888
0.0002	88.0	1144	0.1494	0.71	0.5252	1.5453	0.7100	0.6059	0.3891	0.0886
0.0002	89.0	1157	0.1493	0.71	0.5252	1.5451	0.7100	0.6059	0.3891	0.0887
0.0002	90.0	1170	0.1493	0.71	0.5252	1.5449	0.7100	0.6059	0.3891	0.0887
0.0002	91.0	1183	0.1493	0.71	0.5252	1.5454	0.7100	0.6059	0.3890	0.0887
0.0002	92.0	1196	0.1493	0.71	0.5252	1.5450	0.7100	0.6059	0.3891	0.0887
0.0002	93.0	1209	0.1493	0.71	0.5252	1.5454	0.7100	0.6059	0.3891	0.0888
0.0002	94.0	1222	0.1493	0.71	0.5252	1.5453	0.7100	0.6059	0.3890	0.0886
0.0002	95.0	1235	0.1493	0.71	0.5252	1.5454	0.7100	0.6059	0.3891	0.0887
0.0002	96.0	1248	0.1493	0.71	0.5252	1.5450	0.7100	0.6059	0.3890	0.0887
0.0002	97.0	1261	0.1493	0.71	0.5252	1.5455	0.7100	0.6059	0.3891	0.0888
0.0002	98.0	1274	0.1494	0.71	0.5252	1.5453	0.7100	0.6059	0.3891	0.0886
0.0002	99.0	1287	0.1493	0.71	0.5252	1.5452	0.7100	0.6059	0.3891	0.0887
0.0002	100.0	1300	0.1493	0.71	0.5252	1.5452	0.7100	0.6059	0.3891	0.0888

Framework versions

Transformers 4.26.1
Pytorch 1.13.1.post200
Datasets 2.9.0
Tokenizers 0.13.2

jordyvl/114-tiny_tobacco3482_kd

114-tiny_tobacco3482_kd

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions