hushem_1x_deit_tiny_rms_lr001_fold3

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: 1.7042
• Accuracy: 0.2791

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
No log	1.0	6	3.1061	0.2326
4.0184	2.0	12	1.7666	0.2558
4.0184	3.0	18	1.6279	0.2558
1.7385	4.0	24	1.9636	0.2558
1.583	5.0	30	1.6503	0.2558
1.583	6.0	36	1.4630	0.2326
1.4859	7.0	42	3.2936	0.2326
1.4859	8.0	48	2.0073	0.2558
2.0303	9.0	54	1.4859	0.2326
1.4062	10.0	60	1.6529	0.2326
1.4062	11.0	66	1.4259	0.2791
1.359	12.0	72	1.3892	0.2558
1.359	13.0	78	1.4650	0.3023
1.3464	14.0	84	1.4368	0.2558
1.262	15.0	90	1.4241	0.2558
1.262	16.0	96	1.6562	0.3023
1.2521	17.0	102	1.3729	0.3023
1.2521	18.0	108	1.5241	0.2093
1.2212	19.0	114	1.5032	0.3023
1.1882	20.0	120	1.4178	0.2558
1.1882	21.0	126	1.8156	0.3023
1.1382	22.0	132	1.5280	0.2558
1.1382	23.0	138	1.5037	0.2326
1.0802	24.0	144	1.5058	0.3488
1.1083	25.0	150	1.5421	0.2791
1.1083	26.0	156	1.5398	0.2558
1.0555	27.0	162	1.8560	0.2791
1.0555	28.0	168	1.9193	0.2558
1.0051	29.0	174	1.5934	0.3256
0.958	30.0	180	1.6481	0.2791
0.958	31.0	186	1.5950	0.2791
0.9855	32.0	192	1.5539	0.2558
0.9855	33.0	198	1.6644	0.2791
0.9482	34.0	204	1.6743	0.2326
0.9401	35.0	210	1.6352	0.3023
0.9401	36.0	216	1.6896	0.2791
0.9225	37.0	222	1.7369	0.2326
0.9225	38.0	228	1.6916	0.2558
0.8891	39.0	234	1.6919	0.2791
0.8732	40.0	240	1.7104	0.2791
0.8732	41.0	246	1.7028	0.2791
0.8715	42.0	252	1.7042	0.2791
0.8715	43.0	258	1.7042	0.2791
0.8826	44.0	264	1.7042	0.2791
0.8986	45.0	270	1.7042	0.2791
0.8986	46.0	276	1.7042	0.2791
0.8589	47.0	282	1.7042	0.2791
0.8589	48.0	288	1.7042	0.2791
0.9236	49.0	294	1.7042	0.2791
0.8539	50.0	300	1.7042	0.2791

Framework versions

• Transformers 4.35.0
• Pytorch 2.1.0+cu118
• Datasets 2.14.6
• Tokenizers 0.14.1