hushem_40x_deit_small_sgd_001_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: 0.6398
• Accuracy: 0.7778

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
1.1952	1.0	215	1.4276	0.3778
1.0042	2.0	430	1.4162	0.2889
0.8488	3.0	645	1.3544	0.3778
0.7465	4.0	860	1.2674	0.4889
0.6036	5.0	1075	1.1735	0.5111
0.5355	6.0	1290	1.0934	0.5556
0.4712	7.0	1505	1.0208	0.6
0.3805	8.0	1720	0.9372	0.6222
0.3422	9.0	1935	0.8901	0.6444
0.2964	10.0	2150	0.8433	0.6667
0.2485	11.0	2365	0.7909	0.6889
0.2255	12.0	2580	0.7693	0.7111
0.1717	13.0	2795	0.7309	0.7556
0.1588	14.0	3010	0.7252	0.7556
0.1672	15.0	3225	0.6986	0.7333
0.1097	16.0	3440	0.6863	0.7333
0.1167	17.0	3655	0.6753	0.7556
0.0952	18.0	3870	0.6754	0.7556
0.0806	19.0	4085	0.6768	0.7556
0.0794	20.0	4300	0.6533	0.7556
0.0649	21.0	4515	0.6553	0.7556
0.0639	22.0	4730	0.6451	0.7556
0.0578	23.0	4945	0.6498	0.7556
0.0439	24.0	5160	0.6457	0.7556
0.0437	25.0	5375	0.6423	0.7556
0.038	26.0	5590	0.6342	0.7556
0.0346	27.0	5805	0.6184	0.7556
0.0278	28.0	6020	0.6299	0.7556
0.035	29.0	6235	0.6381	0.7556
0.0226	30.0	6450	0.6272	0.7556
0.0178	31.0	6665	0.6325	0.7556
0.019	32.0	6880	0.6409	0.7556
0.0184	33.0	7095	0.6323	0.7778
0.0238	34.0	7310	0.6091	0.7556
0.0126	35.0	7525	0.6363	0.7778
0.0156	36.0	7740	0.6253	0.7556
0.0165	37.0	7955	0.6280	0.7556
0.0106	38.0	8170	0.6294	0.7778
0.0189	39.0	8385	0.6262	0.7778
0.0098	40.0	8600	0.6454	0.7556
0.0098	41.0	8815	0.6342	0.7778
0.0112	42.0	9030	0.6356	0.7778
0.0128	43.0	9245	0.6416	0.7778
0.0115	44.0	9460	0.6374	0.7778
0.0087	45.0	9675	0.6423	0.7778
0.0077	46.0	9890	0.6446	0.7778
0.0087	47.0	10105	0.6388	0.7778
0.0071	48.0	10320	0.6394	0.7778
0.0087	49.0	10535	0.6404	0.7778
0.0108	50.0	10750	0.6398	0.7778

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.0+cu121
• Datasets 2.12.0
• Tokenizers 0.13.2