vit-transformer3

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

Loss: 1.8890
Accuracy: 0.6833
F1: 0.6840

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: 5e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 32
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	F1
1.8752	0.9552	16	0.9886	0.6667	0.6484
0.7728	1.9701	33	0.6862	0.5667	0.4099
0.7065	2.9851	50	0.6627	0.6333	0.6132
0.6845	4.0	67	0.7065	0.55	0.4922
0.6513	4.9552	83	0.7202	0.4667	0.3905
0.6567	5.9701	100	0.7677	0.5333	0.4667
0.6539	6.9851	117	0.6269	0.6167	0.6047
0.7025	8.0	134	0.6838	0.65	0.6107
0.6698	8.9552	150	0.6313	0.6667	0.6337
0.6986	9.9701	167	0.6200	0.6667	0.6484
0.6811	10.9851	184	0.5869	0.6833	0.6840
0.6132	12.0	201	0.5881	0.6833	0.6687
0.7235	12.9552	217	0.5732	0.65	0.6274
0.5768	13.9701	234	0.5802	0.6833	0.6825
0.5307	14.9851	251	0.6610	0.7	0.7010
0.552	16.0	268	0.6229	0.7333	0.7296
0.5548	16.9552	284	0.6186	0.7167	0.7036
0.4863	17.9701	301	0.8409	0.5667	0.5366
0.5048	18.9851	318	1.0019	0.4833	0.4015
0.4919	20.0	335	0.6475	0.7333	0.7333
0.4788	20.9552	351	0.6931	0.6333	0.6282
0.5076	21.9701	368	0.6798	0.7	0.6983
0.5047	22.9851	385	0.6784	0.7	0.7
0.3477	24.0	402	0.8261	0.7	0.6983
0.4508	24.9552	418	0.6846	0.6833	0.6825
0.4948	25.9701	435	0.7509	0.6833	0.6804
0.3661	26.9851	452	0.7321	0.6667	0.6678
0.3072	28.0	469	0.8338	0.6833	0.6839
0.3573	28.9552	485	0.9031	0.65	0.6434
0.3828	29.9701	502	0.8582	0.6667	0.6667
0.2931	30.9851	519	0.7648	0.65	0.6515
0.3193	32.0	536	0.9218	0.6333	0.6333
0.2783	32.9552	552	0.8452	0.7	0.7013
0.2816	33.9701	569	0.8310	0.6833	0.6735
0.3018	34.9851	586	0.8437	0.7	0.6960
0.2256	36.0	603	1.0684	0.65	0.6507
0.2609	36.9552	619	0.9117	0.65	0.6491
0.2198	37.9701	636	1.1688	0.5833	0.5652
0.306	38.9851	653	0.9001	0.6167	0.6130
0.2243	40.0	670	1.2253	0.6333	0.6313
0.3482	40.9552	686	1.0028	0.65	0.6491
0.196	41.9701	703	0.8747	0.6667	0.6682
0.2261	42.9851	720	1.3642	0.65	0.6468
0.2802	44.0	737	1.3271	0.5833	0.5704
0.1965	44.9552	753	1.3784	0.6	0.6018
0.2198	45.9701	770	1.3224	0.6667	0.6682
0.1852	46.9851	787	1.5364	0.6333	0.6243
0.197	48.0	804	1.5706	0.6167	0.6174
0.1932	48.9552	820	1.3610	0.6667	0.6648
0.1495	49.9701	837	1.4687	0.6167	0.6174
0.1404	50.9851	854	1.3438	0.7	0.6983
0.1275	52.0	871	1.4674	0.6	0.5978
0.1545	52.9552	887	1.3120	0.6167	0.6183
0.147	53.9701	904	1.5816	0.6167	0.6183
0.1541	54.9851	921	1.5117	0.6667	0.6678
0.1283	56.0	938	1.5965	0.6667	0.6678
0.1715	56.9552	954	1.6750	0.65	0.6491
0.1513	57.9701	971	1.9170	0.5333	0.5164
0.2349	58.9851	988	1.5358	0.6333	0.6346
0.1248	60.0	1005	1.6686	0.6833	0.6840
0.1076	60.9552	1021	1.7018	0.6333	0.6346
0.1431	61.9701	1038	1.9088	0.6333	0.6333
0.0838	62.9851	1055	1.8821	0.6333	0.6346
0.0989	64.0	1072	1.6053	0.65	0.6491
0.1323	64.9552	1088	1.7114	0.6333	0.6312
0.0908	65.9701	1105	1.7326	0.65	0.6491
0.2056	66.9851	1122	1.7166	0.6167	0.6130
0.0752	68.0	1139	1.8009	0.65	0.6467
0.1116	68.9552	1155	1.6964	0.6667	0.6678
0.0821	69.9701	1172	1.7557	0.6167	0.6094
0.1284	70.9851	1189	1.8039	0.65	0.6491
0.1905	72.0	1206	1.7951	0.6167	0.6094
0.1031	72.9552	1222	1.6888	0.6667	0.6648
0.0706	73.9701	1239	1.8992	0.65	0.6467
0.0944	74.9851	1256	1.6965	0.6833	0.6840
0.1042	76.0	1273	1.6756	0.6833	0.6825
0.1599	76.9552	1289	1.4360	0.7333	0.7342
0.0896	77.9701	1306	1.5759	0.65	0.6467
0.0674	78.9851	1323	1.7071	0.7	0.7010
0.1133	80.0	1340	1.6499	0.6833	0.6840
0.0506	80.9552	1356	1.6546	0.6833	0.6825
0.1015	81.9701	1373	1.6468	0.7	0.7013
0.0923	82.9851	1390	1.8567	0.6667	0.6622
0.0752	84.0	1407	1.8140	0.7	0.7010
0.0768	84.9552	1423	1.8225	0.6667	0.6678
0.0683	85.9701	1440	1.8094	0.6833	0.6840
0.0454	86.9851	1457	1.8892	0.65	0.6491
0.054	88.0	1474	1.8180	0.7	0.7010
0.0449	88.9552	1490	1.7891	0.7333	0.7345
0.0645	89.9701	1507	1.8262	0.7	0.7010
0.0632	90.9851	1524	1.8187	0.7167	0.7179
0.0795	92.0	1541	1.7941	0.7333	0.7345
0.0923	92.9552	1557	1.8340	0.6833	0.6840
0.0486	93.9701	1574	1.8843	0.6667	0.6667
0.0821	94.9851	1591	1.8907	0.6667	0.6667
0.0384	95.5224	1600	1.8890	0.6833	0.6840

Framework versions

Transformers 4.45.2
Pytorch 2.4.0+cu121
Datasets 3.1.0
Tokenizers 0.20.1

masafresh/vit-transformer3

vit-transformer3

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

为此模型做种