beit-base-patch16-224-hasta-55-fold4

This model is a fine-tuned version of microsoft/beit-base-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 1.1593
• Accuracy: 0.7222

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: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• 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
No log	0.5714	1	1.2503	0.4444
No log	1.7143	3	1.1731	0.4167
No log	2.8571	5	1.0852	0.5278
No log	4.0	7	1.0787	0.5
No log	4.5714	8	1.1199	0.5278
1.0532	5.7143	10	1.0584	0.4722
1.0532	6.8571	12	1.0800	0.5278
1.0532	8.0	14	1.1635	0.4722
1.0532	8.5714	15	1.1171	0.4444
1.0532	9.7143	17	1.5254	0.3889
1.0532	10.8571	19	1.1236	0.4444
0.9087	12.0	21	1.0255	0.5556
0.9087	12.5714	22	1.1108	0.5278
0.9087	13.7143	24	1.0365	0.5278
0.9087	14.8571	26	1.0638	0.5
0.9087	16.0	28	1.1090	0.6111
0.9087	16.5714	29	1.1166	0.5556
0.7925	17.7143	31	1.0650	0.4722
0.7925	18.8571	33	1.3085	0.5556
0.7925	20.0	35	1.1624	0.5278
0.7925	20.5714	36	0.9994	0.5556
0.7925	21.7143	38	1.1054	0.4722
0.7472	22.8571	40	1.0926	0.5833
0.7472	24.0	42	1.1054	0.6111
0.7472	24.5714	43	1.0486	0.5556
0.7472	25.7143	45	1.0454	0.5556
0.7472	26.8571	47	1.0267	0.6389
0.7472	28.0	49	1.0684	0.6667
0.572	28.5714	50	1.0575	0.6111
0.572	29.7143	52	1.1591	0.5833
0.572	30.8571	54	1.1837	0.5833
0.572	32.0	56	1.0444	0.6667
0.572	32.5714	57	1.0450	0.6667
0.572	33.7143	59	1.0975	0.6667
0.471	34.8571	61	1.1131	0.6667
0.471	36.0	63	1.1204	0.5833
0.471	36.5714	64	1.0992	0.5833
0.471	37.7143	66	1.0879	0.6389
0.471	38.8571	68	1.0981	0.6111
0.3896	40.0	70	1.0576	0.6667
0.3896	40.5714	71	1.0612	0.6389
0.3896	41.7143	73	1.1195	0.6667
0.3896	42.8571	75	1.1974	0.6667
0.3896	44.0	77	1.1353	0.6667
0.3896	44.5714	78	1.1143	0.6667
0.3775	45.7143	80	1.1055	0.6667
0.3775	46.8571	82	1.1997	0.6667
0.3775	48.0	84	1.3267	0.6667
0.3775	48.5714	85	1.3027	0.6667
0.3775	49.7143	87	1.1593	0.7222
0.3775	50.8571	89	1.0970	0.6111
0.3623	52.0	91	1.0902	0.6111
0.3623	52.5714	92	1.0908	0.6111
0.3623	53.7143	94	1.1214	0.6389
0.3623	54.8571	96	1.1691	0.6944
0.3623	56.0	98	1.1914	0.6667
0.3623	56.5714	99	1.1949	0.6667
0.3455	57.1429	100	1.1951	0.6667

Framework versions

• Transformers 4.41.0
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1