deit-base-distilled-patch16-224-hasta-75-fold3

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

Loss: 0.5699
Accuracy: 0.9167

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	1.0	1	1.0629	0.1667
No log	2.0	2	0.8596	0.5833
No log	3.0	3	0.5699	0.9167
No log	4.0	4	0.3576	0.9167
No log	5.0	5	0.2875	0.9167
No log	6.0	6	0.2993	0.9167
No log	7.0	7	0.3152	0.9167
No log	8.0	8	0.3399	0.9167
No log	9.0	9	0.4232	0.9167
0.3622	10.0	10	0.5129	0.9167
0.3622	11.0	11	0.4640	0.9167
0.3622	12.0	12	0.3709	0.9167
0.3622	13.0	13	0.3229	0.9167
0.3622	14.0	14	0.3183	0.9167
0.3622	15.0	15	0.3590	0.9167
0.3622	16.0	16	0.4468	0.9167
0.3622	17.0	17	0.5249	0.75
0.3622	18.0	18	0.4227	0.9167
0.3622	19.0	19	0.3300	0.9167
0.1749	20.0	20	0.3031	0.9167
0.1749	21.0	21	0.2836	0.9167
0.1749	22.0	22	0.2596	0.9167
0.1749	23.0	23	0.2215	0.9167
0.1749	24.0	24	0.1797	0.9167
0.1749	25.0	25	0.1658	0.9167
0.1749	26.0	26	0.1600	0.9167
0.1749	27.0	27	0.1567	0.9167
0.1749	28.0	28	0.1680	0.9167
0.1749	29.0	29	0.2079	0.9167
0.1002	30.0	30	0.2458	0.9167
0.1002	31.0	31	0.2637	0.9167
0.1002	32.0	32	0.2795	0.9167
0.1002	33.0	33	0.3174	0.9167
0.1002	34.0	34	0.3426	0.9167
0.1002	35.0	35	0.3858	0.9167
0.1002	36.0	36	0.4281	0.9167
0.1002	37.0	37	0.4447	0.9167
0.1002	38.0	38	0.4355	0.9167
0.1002	39.0	39	0.4306	0.9167
0.0662	40.0	40	0.4398	0.9167
0.0662	41.0	41	0.4657	0.9167
0.0662	42.0	42	0.4941	0.9167
0.0662	43.0	43	0.5061	0.9167
0.0662	44.0	44	0.5010	0.9167
0.0662	45.0	45	0.4816	0.9167
0.0662	46.0	46	0.4398	0.9167
0.0662	47.0	47	0.3948	0.9167
0.0662	48.0	48	0.3486	0.9167
0.0662	49.0	49	0.3180	0.9167
0.0543	50.0	50	0.3029	0.9167
0.0543	51.0	51	0.3053	0.9167
0.0543	52.0	52	0.3160	0.9167
0.0543	53.0	53	0.3449	0.9167
0.0543	54.0	54	0.3712	0.9167
0.0543	55.0	55	0.3785	0.9167
0.0543	56.0	56	0.4049	0.9167
0.0543	57.0	57	0.4094	0.9167
0.0543	58.0	58	0.4179	0.9167
0.0543	59.0	59	0.4083	0.9167
0.0473	60.0	60	0.3855	0.9167
0.0473	61.0	61	0.3758	0.9167
0.0473	62.0	62	0.3675	0.9167
0.0473	63.0	63	0.3660	0.9167
0.0473	64.0	64	0.3843	0.9167
0.0473	65.0	65	0.4092	0.9167
0.0473	66.0	66	0.4374	0.9167
0.0473	67.0	67	0.4666	0.9167
0.0473	68.0	68	0.4798	0.9167
0.0473	69.0	69	0.4869	0.9167
0.018	70.0	70	0.4853	0.9167
0.018	71.0	71	0.4783	0.9167
0.018	72.0	72	0.4649	0.9167
0.018	73.0	73	0.4525	0.9167
0.018	74.0	74	0.4411	0.9167
0.018	75.0	75	0.4355	0.9167
0.018	76.0	76	0.4310	0.9167
0.018	77.0	77	0.4330	0.9167
0.018	78.0	78	0.4296	0.9167
0.018	79.0	79	0.4264	0.9167
0.0195	80.0	80	0.4204	0.9167
0.0195	81.0	81	0.4182	0.9167
0.0195	82.0	82	0.4193	0.9167
0.0195	83.0	83	0.4225	0.9167
0.0195	84.0	84	0.4208	0.9167
0.0195	85.0	85	0.4193	0.9167
0.0195	86.0	86	0.4185	0.9167
0.0195	87.0	87	0.4186	0.9167
0.0195	88.0	88	0.4163	0.9167
0.0195	89.0	89	0.4076	0.9167
0.0203	90.0	90	0.4032	0.9167
0.0203	91.0	91	0.4000	0.9167
0.0203	92.0	92	0.4003	0.9167
0.0203	93.0	93	0.4033	0.9167
0.0203	94.0	94	0.4079	0.9167
0.0203	95.0	95	0.4112	0.9167
0.0203	96.0	96	0.4138	0.9167
0.0203	97.0	97	0.4155	0.9167
0.0203	98.0	98	0.4167	0.9167
0.0203	99.0	99	0.4169	0.9167
0.0252	100.0	100	0.4170	0.9167

Framework versions

Transformers 4.41.0
Pytorch 2.3.0+cu121
Datasets 2.19.1
Tokenizers 0.19.1

BilalMuftuoglu/deit-base-distilled-patch16-224-hasta-75-fold3

deit-base-distilled-patch16-224-hasta-75-fold3

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Seed this model