smids_5x_deit_tiny_rms_001_fold2

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

• Loss: 1.4964
• Accuracy: 0.8486

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: 0.001
• 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
1.0328	1.0	375	0.9569	0.4459
0.8929	2.0	750	0.8978	0.5374
0.8224	3.0	1125	0.7888	0.5574
0.8327	4.0	1500	0.8571	0.5641
0.7266	5.0	1875	1.1729	0.5025
0.6507	6.0	2250	0.7875	0.6456
0.6983	7.0	2625	0.6489	0.6972
0.6312	8.0	3000	0.7326	0.6789
0.641	9.0	3375	0.5505	0.7488
0.6354	10.0	3750	0.5766	0.7354
0.5813	11.0	4125	0.4910	0.7920
0.6084	12.0	4500	0.5458	0.7720
0.4944	13.0	4875	0.4657	0.8020
0.5555	14.0	5250	0.5401	0.7621
0.526	15.0	5625	0.4958	0.7837
0.3751	16.0	6000	0.4911	0.8037
0.4264	17.0	6375	0.5204	0.7837
0.4312	18.0	6750	0.5011	0.7953
0.3686	19.0	7125	0.4979	0.7970
0.3954	20.0	7500	0.4812	0.8120
0.3782	21.0	7875	0.4706	0.8120
0.3544	22.0	8250	0.4461	0.8353
0.3759	23.0	8625	0.4516	0.8319
0.3473	24.0	9000	0.4332	0.8270
0.2572	25.0	9375	0.5951	0.8203
0.3628	26.0	9750	0.5630	0.7887
0.2737	27.0	10125	0.5304	0.8336
0.2272	28.0	10500	0.5597	0.8319
0.2226	29.0	10875	0.5680	0.8419
0.1778	30.0	11250	0.6295	0.8170
0.2382	31.0	11625	0.6223	0.8270
0.1721	32.0	12000	0.6049	0.8469
0.219	33.0	12375	0.5556	0.8569
0.0972	34.0	12750	0.6389	0.8502
0.1781	35.0	13125	0.7873	0.8253
0.1052	36.0	13500	0.8815	0.8236
0.1087	37.0	13875	0.7444	0.8453
0.09	38.0	14250	0.9779	0.8253
0.0859	39.0	14625	0.8817	0.8386
0.0521	40.0	15000	0.9849	0.8453
0.081	41.0	15375	1.0555	0.8203
0.0225	42.0	15750	1.1081	0.8303
0.0521	43.0	16125	1.2294	0.8253
0.0259	44.0	16500	1.3035	0.8336
0.0403	45.0	16875	1.3613	0.8253
0.0225	46.0	17250	1.4500	0.8103
0.0235	47.0	17625	1.5096	0.8270
0.0002	48.0	18000	1.5022	0.8469
0.0101	49.0	18375	1.4968	0.8469
0.0029	50.0	18750	1.4964	0.8486

Framework versions

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