ViT_LFW_Model4

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1287
• Accuracy: 0.9705
• Precision: 0.9054
• Recall: 0.9583
• F1: 0.8838

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: 0.0002
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
3.4756	0.41	100	2.8779	0.6015	0.8461	0.3406	0.2698
2.6524	0.83	200	1.8112	0.7749	0.8298	0.5915	0.5064
1.6994	1.24	300	1.1829	0.8450	0.8065	0.7112	0.6160
1.3097	1.66	400	0.6849	0.9225	0.8808	0.8486	0.7908
0.5976	2.07	500	0.4778	0.9336	0.9015	0.8803	0.8293
0.412	2.49	600	0.4110	0.9299	0.8555	0.8988	0.8000
0.3165	2.9	700	0.3295	0.9262	0.8108	0.8787	0.7350
0.1537	3.32	800	0.2427	0.9520	0.8792	0.9333	0.8405
0.087	3.73	900	0.2373	0.9520	0.8989	0.9308	0.8562
0.0728	4.15	1000	0.2068	0.9483	0.8815	0.9264	0.8297
0.0305	4.56	1100	0.1759	0.9557	0.8692	0.9391	0.8279
0.0277	4.98	1200	0.1879	0.9446	0.8328	0.9197	0.7856
0.0126	5.39	1300	0.1759	0.9594	0.87	0.9333	0.8193
0.0137	5.81	1400	0.1595	0.9631	0.8771	0.9440	0.8396
0.0083	6.22	1500	0.1287	0.9705	0.9054	0.9583	0.8838
0.0078	6.64	1600	0.1295	0.9668	0.8910	0.9511	0.8592
0.0064	7.05	1700	0.1322	0.9668	0.8910	0.9511	0.8592
0.0062	7.47	1800	0.1299	0.9668	0.8910	0.9511	0.8592
0.0053	7.88	1900	0.1307	0.9668	0.8910	0.9511	0.8592
0.0049	8.3	2000	0.1295	0.9668	0.8910	0.9511	0.8592
0.0041	8.71	2100	0.1302	0.9668	0.8910	0.9511	0.8592
0.0036	9.13	2200	0.1310	0.9668	0.8910	0.9511	0.8592
0.0037	9.54	2300	0.1311	0.9668	0.8910	0.9511	0.8592
0.0028	9.96	2400	0.1301	0.9668	0.8910	0.9511	0.8592
0.0031	10.37	2500	0.1308	0.9668	0.8910	0.9511	0.8592
0.0026	10.79	2600	0.1304	0.9668	0.8910	0.9511	0.8592
0.0023	11.2	2700	0.1299	0.9668	0.8910	0.9511	0.8592
0.0024	11.62	2800	0.1315	0.9668	0.8910	0.9511	0.8592
0.0022	12.03	2900	0.1321	0.9668	0.8910	0.9511	0.8592
0.002	12.45	3000	0.1321	0.9668	0.8910	0.9511	0.8592
0.002	12.86	3100	0.1332	0.9668	0.8910	0.9511	0.8592
0.0017	13.28	3200	0.1327	0.9668	0.8910	0.9511	0.8592
0.0016	13.69	3300	0.1328	0.9668	0.8910	0.9511	0.8592
0.0015	14.11	3400	0.1336	0.9668	0.8910	0.9511	0.8592
0.0015	14.52	3500	0.1343	0.9668	0.8910	0.9511	0.8592
0.0015	14.94	3600	0.1345	0.9668	0.8910	0.9511	0.8592
0.0014	15.35	3700	0.1344	0.9668	0.8910	0.9511	0.8592
0.0013	15.77	3800	0.1354	0.9668	0.8910	0.9511	0.8592
0.0013	16.18	3900	0.1357	0.9668	0.8910	0.9511	0.8592
0.0012	16.6	4000	0.1365	0.9668	0.8910	0.9511	0.8592
0.0011	17.01	4100	0.1357	0.9668	0.8910	0.9511	0.8592
0.001	17.43	4200	0.1361	0.9668	0.8910	0.9511	0.8592
0.001	17.84	4300	0.1364	0.9668	0.8910	0.9511	0.8592
0.001	18.26	4400	0.1379	0.9668	0.8910	0.9511	0.8592
0.001	18.67	4500	0.1375	0.9668	0.8910	0.9511	0.8592
0.0009	19.09	4600	0.1374	0.9668	0.8910	0.9511	0.8592
0.0009	19.5	4700	0.1374	0.9668	0.8910	0.9511	0.8592
0.0009	19.92	4800	0.1382	0.9668	0.8910	0.9511	0.8592
0.0008	20.33	4900	0.1385	0.9705	0.8963	0.9550	0.8666
0.0007	20.75	5000	0.1389	0.9705	0.8963	0.9550	0.8666
0.0007	21.16	5100	0.1391	0.9705	0.8963	0.9550	0.8666
0.0007	21.58	5200	0.1392	0.9705	0.8963	0.9550	0.8666
0.0007	21.99	5300	0.1397	0.9705	0.8963	0.9550	0.8666
0.0007	22.41	5400	0.1401	0.9668	0.8910	0.9511	0.8592
0.0007	22.82	5500	0.1404	0.9705	0.8963	0.9550	0.8666
0.0006	23.24	5600	0.1404	0.9705	0.8963	0.9550	0.8666
0.0006	23.65	5700	0.1402	0.9705	0.8963	0.9550	0.8666
0.0006	24.07	5800	0.1411	0.9705	0.8963	0.9550	0.8666
0.0006	24.48	5900	0.1411	0.9705	0.8963	0.9550	0.8666
0.0006	24.9	6000	0.1413	0.9705	0.8963	0.9550	0.8666
0.0005	25.31	6100	0.1418	0.9705	0.8963	0.9550	0.8666
0.0006	25.73	6200	0.1420	0.9705	0.8963	0.9550	0.8666
0.0005	26.14	6300	0.1421	0.9705	0.8963	0.9550	0.8666
0.0005	26.56	6400	0.1423	0.9705	0.8963	0.9550	0.8666
0.0005	26.97	6500	0.1424	0.9705	0.8963	0.9550	0.8666
0.0004	27.39	6600	0.1428	0.9705	0.8963	0.9550	0.8666
0.0005	27.8	6700	0.1429	0.9705	0.8963	0.9550	0.8666
0.0005	28.22	6800	0.1428	0.9705	0.8963	0.9550	0.8666
0.0005	28.63	6900	0.1430	0.9705	0.8963	0.9550	0.8666
0.0005	29.05	7000	0.1430	0.9705	0.8963	0.9550	0.8666
0.0005	29.46	7100	0.1430	0.9705	0.8963	0.9550	0.8666
0.0005	29.88	7200	0.1430	0.9705	0.8963	0.9550	0.8666

Framework versions

• Transformers 4.30.2
• Pytorch 2.0.1+cu118
• Datasets 2.13.0
• Tokenizers 0.13.3