hushem_1x_deit_base_rms_0001_fold1

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

• Loss: 3.1229
• Accuracy: 0.5778

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
No log	1.0	6	1.8872	0.2444
1.9577	2.0	12	1.4317	0.2444
1.9577	3.0	18	1.4241	0.2444
1.42	4.0	24	1.4307	0.2444
1.3849	5.0	30	1.3737	0.4222
1.3849	6.0	36	1.3932	0.3111
1.4386	7.0	42	1.7764	0.2444
1.4386	8.0	48	1.3602	0.2667
1.258	9.0	54	2.0427	0.3778
1.1111	10.0	60	1.1310	0.4889
1.1111	11.0	66	2.7936	0.3778
0.8089	12.0	72	1.1048	0.5333
0.8089	13.0	78	1.5229	0.4
0.6344	14.0	84	2.4918	0.4222
0.3674	15.0	90	1.2248	0.5333
0.3674	16.0	96	1.9743	0.5333
0.1742	17.0	102	2.3841	0.4444
0.1742	18.0	108	3.0361	0.4222
0.1094	19.0	114	2.4775	0.5778
0.0149	20.0	120	2.7557	0.6
0.0149	21.0	126	2.8306	0.6
0.0005	22.0	132	2.8795	0.6
0.0005	23.0	138	2.9199	0.6
0.0004	24.0	144	2.9515	0.6
0.0003	25.0	150	2.9748	0.6
0.0003	26.0	156	2.9936	0.6
0.0002	27.0	162	3.0112	0.6
0.0002	28.0	168	3.0280	0.6
0.0002	29.0	174	3.0414	0.6
0.0002	30.0	180	3.0544	0.6
0.0002	31.0	186	3.0664	0.6
0.0002	32.0	192	3.0752	0.6
0.0002	33.0	198	3.0851	0.5778
0.0002	34.0	204	3.0929	0.5778
0.0002	35.0	210	3.0988	0.5778
0.0002	36.0	216	3.1066	0.5778
0.0002	37.0	222	3.1117	0.5778
0.0002	38.0	228	3.1165	0.5778
0.0001	39.0	234	3.1194	0.5778
0.0001	40.0	240	3.1215	0.5778
0.0001	41.0	246	3.1226	0.5778
0.0001	42.0	252	3.1229	0.5778
0.0001	43.0	258	3.1229	0.5778
0.0001	44.0	264	3.1229	0.5778
0.0001	45.0	270	3.1229	0.5778
0.0001	46.0	276	3.1229	0.5778
0.0001	47.0	282	3.1229	0.5778
0.0001	48.0	288	3.1229	0.5778
0.0001	49.0	294	3.1229	0.5778
0.0001	50.0	300	3.1229	0.5778

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0