mobilevit-x-small

This model is a fine-tuned version of apple/mobilevit-x-small on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.0196
• Accuracy: 0.9959

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: 5e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6911	1.0	34	0.6932	0.5083
0.6584	2.0	68	0.6287	0.7510
0.5388	3.0	102	0.4852	0.8734
0.3891	4.0	136	0.3065	0.9357
0.2915	5.0	170	0.2005	0.9647
0.2319	6.0	204	0.1498	0.9689
0.2038	7.0	238	0.1228	0.9710
0.1641	8.0	272	0.0892	0.9855
0.1525	9.0	306	0.0778	0.9834
0.1584	10.0	340	0.0565	0.9896
0.1194	11.0	374	0.0491	0.9917
0.1222	12.0	408	0.0436	0.9896
0.1229	13.0	442	0.0360	0.9979
0.1334	14.0	476	0.0326	0.9959
0.122	15.0	510	0.0425	0.9896
0.096	16.0	544	0.0315	0.9959
0.0989	17.0	578	0.0303	0.9938
0.1085	18.0	612	0.0262	0.9959
0.0957	19.0	646	0.0232	0.9959
0.1129	20.0	680	0.0266	0.9959
0.0843	21.0	714	0.0234	0.9959
0.0868	22.0	748	0.0217	0.9959
0.0867	23.0	782	0.0233	0.9959
0.0947	24.0	816	0.0204	0.9959
0.0786	25.0	850	0.0199	0.9959
0.1009	26.0	884	0.0212	0.9959
0.0785	27.0	918	0.0204	0.9959
0.0811	28.0	952	0.0180	0.9959
0.0883	29.0	986	0.0193	0.9959
0.0988	30.0	1020	0.0196	0.9959

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.2.0
• Tokenizers 0.19.1