smids_5x_deit_small_sgd_0001_fold2

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: 0.5006
• Accuracy: 0.8037

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: 0.0001
• 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
1.0629	1.0	375	1.0383	0.4592
1.0151	2.0	750	1.0009	0.4925
0.9588	3.0	1125	0.9619	0.5574
0.924	4.0	1500	0.9255	0.5890
0.8743	5.0	1875	0.8899	0.6290
0.8177	6.0	2250	0.8563	0.6522
0.7888	7.0	2625	0.8262	0.6755
0.7921	8.0	3000	0.7964	0.7005
0.7372	9.0	3375	0.7699	0.7138
0.7291	10.0	3750	0.7453	0.7221
0.7295	11.0	4125	0.7221	0.7255
0.6995	12.0	4500	0.7007	0.7288
0.621	13.0	4875	0.6811	0.7388
0.6398	14.0	5250	0.6638	0.7504
0.6383	15.0	5625	0.6483	0.7587
0.5747	16.0	6000	0.6341	0.7587
0.6097	17.0	6375	0.6214	0.7604
0.594	18.0	6750	0.6099	0.7604
0.5533	19.0	7125	0.5997	0.7654
0.5984	20.0	7500	0.5904	0.7687
0.5406	21.0	7875	0.5822	0.7720
0.525	22.0	8250	0.5743	0.7704
0.5434	23.0	8625	0.5673	0.7720
0.5253	24.0	9000	0.5609	0.7737
0.5143	25.0	9375	0.5549	0.7754
0.5351	26.0	9750	0.5494	0.7787
0.5716	27.0	10125	0.5444	0.7787
0.4849	28.0	10500	0.5399	0.7820
0.4878	29.0	10875	0.5357	0.7887
0.4887	30.0	11250	0.5319	0.7920
0.4866	31.0	11625	0.5283	0.7920
0.5025	32.0	12000	0.5250	0.7937
0.4672	33.0	12375	0.5219	0.7903
0.4395	34.0	12750	0.5192	0.7887
0.473	35.0	13125	0.5166	0.7920
0.4458	36.0	13500	0.5143	0.7920
0.4639	37.0	13875	0.5122	0.7937
0.4488	38.0	14250	0.5103	0.7953
0.4766	39.0	14625	0.5086	0.7970
0.4603	40.0	15000	0.5071	0.7987
0.4461	41.0	15375	0.5058	0.8003
0.4671	42.0	15750	0.5046	0.8003
0.4415	43.0	16125	0.5036	0.8020
0.4496	44.0	16500	0.5027	0.8020
0.4327	45.0	16875	0.5020	0.8020
0.5062	46.0	17250	0.5015	0.8020
0.4692	47.0	17625	0.5010	0.8037
0.426	48.0	18000	0.5008	0.8037
0.518	49.0	18375	0.5006	0.8037
0.4765	50.0	18750	0.5006	0.8037

Framework versions

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