hushem_1x_deit_small_rms_0001_fold1

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: 2.9945
• Accuracy: 0.5111

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	1.6080	0.2444
2.1035	2.0	12	1.4698	0.2444
2.1035	3.0	18	1.4524	0.2444
1.4518	4.0	24	1.3705	0.4444
1.3954	5.0	30	1.3051	0.3333
1.3954	6.0	36	1.7248	0.2444
1.3016	7.0	42	1.2877	0.4444
1.3016	8.0	48	1.2531	0.3111
1.1687	9.0	54	2.0909	0.3333
1.0636	10.0	60	1.6618	0.4
1.0636	11.0	66	1.5040	0.3778
0.852	12.0	72	1.6151	0.3556
0.852	13.0	78	1.3188	0.4667
0.6455	14.0	84	1.7023	0.4
0.4371	15.0	90	1.6246	0.4889
0.4371	16.0	96	1.5721	0.5111
0.2911	17.0	102	1.7046	0.4889
0.2911	18.0	108	2.5082	0.4889
0.0854	19.0	114	2.6259	0.4444
0.0965	20.0	120	2.0620	0.6
0.0965	21.0	126	2.2433	0.4889
0.0234	22.0	132	1.7971	0.5333
0.0234	23.0	138	2.3712	0.4667
0.039	24.0	144	1.6644	0.6
0.016	25.0	150	2.8343	0.4222
0.016	26.0	156	2.1232	0.5111
0.0027	27.0	162	2.6607	0.4889
0.0027	28.0	168	2.7233	0.4889
0.0006	29.0	174	2.7692	0.4889
0.0004	30.0	180	2.8123	0.5111
0.0004	31.0	186	2.8470	0.5111
0.0004	32.0	192	2.8799	0.5111
0.0004	33.0	198	2.9035	0.5111
0.0003	34.0	204	2.9241	0.5111
0.0003	35.0	210	2.9413	0.5111
0.0003	36.0	216	2.9570	0.5111
0.0003	37.0	222	2.9705	0.5111
0.0003	38.0	228	2.9803	0.5111
0.0003	39.0	234	2.9868	0.5111
0.0003	40.0	240	2.9915	0.5111
0.0003	41.0	246	2.9938	0.5111
0.0003	42.0	252	2.9945	0.5111
0.0003	43.0	258	2.9945	0.5111
0.0002	44.0	264	2.9945	0.5111
0.0003	45.0	270	2.9945	0.5111
0.0003	46.0	276	2.9945	0.5111
0.0002	47.0	282	2.9945	0.5111
0.0002	48.0	288	2.9945	0.5111
0.0003	49.0	294	2.9945	0.5111
0.0002	50.0	300	2.9945	0.5111

Framework versions

• Transformers 4.35.1
• Pytorch 2.1.0+cu118
• Datasets 2.14.7
• Tokenizers 0.14.1