smids_5x_deit_tiny_adamax_00001_fold3

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

• Loss: 0.8422
• Accuracy: 0.8983

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: 1e-05
• train_batch_size: 32
• eval_batch_size: 32
• 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: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.2997	1.0	375	0.3235	0.88
0.2668	2.0	750	0.2793	0.8967
0.1125	3.0	1125	0.2572	0.9117
0.1056	4.0	1500	0.2703	0.9117
0.1075	5.0	1875	0.3070	0.8983
0.0954	6.0	2250	0.3649	0.8917
0.0479	7.0	2625	0.3675	0.91
0.0335	8.0	3000	0.4528	0.905
0.0032	9.0	3375	0.4970	0.9
0.0009	10.0	3750	0.5488	0.9167
0.0002	11.0	4125	0.5799	0.8983
0.0004	12.0	4500	0.6150	0.9067
0.0001	13.0	4875	0.6403	0.9083
0.0001	14.0	5250	0.6886	0.9017
0.0008	15.0	5625	0.6997	0.9083
0.0	16.0	6000	0.7289	0.9067
0.0	17.0	6375	0.7468	0.905
0.0	18.0	6750	0.7378	0.905
0.0	19.0	7125	0.7534	0.9033
0.0	20.0	7500	0.7571	0.9083
0.0	21.0	7875	0.7624	0.9033
0.0	22.0	8250	0.7704	0.9083
0.0049	23.0	8625	0.8162	0.9017
0.0	24.0	9000	0.7799	0.9033
0.0	25.0	9375	0.8193	0.9033
0.0	26.0	9750	0.7928	0.9033
0.0	27.0	10125	0.7850	0.9017
0.0	28.0	10500	0.8132	0.9
0.0	29.0	10875	0.8205	0.8983
0.0038	30.0	11250	0.8084	0.905
0.0	31.0	11625	0.8179	0.9017
0.0	32.0	12000	0.8194	0.8983
0.0	33.0	12375	0.8163	0.9017
0.0	34.0	12750	0.8152	0.9067
0.0	35.0	13125	0.8374	0.8983
0.0	36.0	13500	0.8315	0.8983
0.0	37.0	13875	0.8335	0.8967
0.0	38.0	14250	0.8285	0.8983
0.0	39.0	14625	0.8274	0.9033
0.0022	40.0	15000	0.8347	0.9017
0.0	41.0	15375	0.8356	0.9
0.0	42.0	15750	0.8391	0.9
0.0	43.0	16125	0.8395	0.8983
0.0	44.0	16500	0.8400	0.8983
0.0	45.0	16875	0.8413	0.8983
0.0	46.0	17250	0.8418	0.8983
0.0	47.0	17625	0.8416	0.8983
0.0	48.0	18000	0.8421	0.8983
0.0	49.0	18375	0.8423	0.8983
0.0	50.0	18750	0.8422	0.8983

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.1+cu121
• Datasets 2.12.0
• Tokenizers 0.13.2