vit-base-patch16-224-in21k-FINALAsphaltLaneClassifier-detectorVIT30epochsTrainValAUGMENTED

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.0427
• Accuracy: {'accuracy': 0.9804996953077392}
• F1: {'f1': 0.9800980973913306}
• Precision: {'precision': 0.9820258378580791}
• Recall: {'recall': 0.9807653776798236}

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: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• 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	F1	Precision	Recall
0.2763	0.9994	1230	0.2269	{'accuracy': 0.9640463132236441}	{'f1': 0.9630811116832685}	{'precision': 0.9640657360845907}	{'recall': 0.964963663987968}
0.2271	1.9996	2461	0.0938	{'accuracy': 0.9764371318301849}	{'f1': 0.9761448410653392}	{'precision': 0.977818779128281}	{'recall': 0.9770355856317611}
0.1737	2.9998	3692	0.1073	{'accuracy': 0.9577493398334349}	{'f1': 0.9569000292933705}	{'precision': 0.959800239637741}	{'recall': 0.9552400643800896}
0.0361	4.0	4923	0.0532	{'accuracy': 0.9782652853950843}	{'f1': 0.9778377521587307}	{'precision': 0.9794461626122134}	{'recall': 0.9785602619414852}
0.1613	4.9994	6153	0.0574	{'accuracy': 0.9762340036563071}	{'f1': 0.9759632908989441}	{'precision': 0.9776775737495528}	{'recall': 0.9768318271751163}
0.098	5.9996	7384	0.0551	{'accuracy': 0.9774527726995734}	{'f1': 0.9771202466632837}	{'precision': 0.9788281272701719}	{'recall': 0.9779919889728328}
0.0867	6.9998	8615	0.0476	{'accuracy': 0.9794840544383506}	{'f1': 0.9791014077010869}	{'precision': 0.9810276957543254}	{'recall': 0.9798251744941444}
0.0855	8.0	9846	0.0657	{'accuracy': 0.9731870810481413}	{'f1': 0.9729473784067677}	{'precision': 0.9755572390111291}	{'recall': 0.973170601861673}
0.1041	8.9994	11076	0.0454	{'accuracy': 0.9802965671338615}	{'f1': 0.9799080834672663}	{'precision': 0.981846143338762}	{'recall': 0.9805652976478109}
0.0674	9.9996	12307	0.0471	{'accuracy': 0.9794840544383506}	{'f1': 0.9791012017629246}	{'precision': 0.9809386449900642}	{'recall': 0.9798657735056604}
0.1083	10.9998	13538	0.0437	{'accuracy': 0.9802965671338615}	{'f1': 0.9799080834672663}	{'precision': 0.981846143338762}	{'recall': 0.9805652976478109}
0.1197	12.0	14769	0.0560	{'accuracy': 0.9766402600040626}	{'f1': 0.9763088751876786}	{'precision': 0.9779331362017656}	{'recall': 0.9773126240278623}
0.0543	12.9994	15999	0.0455	{'accuracy': 0.9794840544383506}	{'f1': 0.9791012017629246}	{'precision': 0.9809386449900642}	{'recall': 0.9798657735056604}
0.0949	13.9996	17230	0.0440	{'accuracy': 0.9802965671338615}	{'f1': 0.9798985564872046}	{'precision': 0.981846143338762}	{'recall': 0.9805462716324967}
0.0764	14.9998	18461	0.0508	{'accuracy': 0.9782652853950843}	{'f1': 0.9779293131150244}	{'precision': 0.9797174734398995}	{'recall': 0.9787862484966643}
0.0482	16.0	19692	0.0625	{'accuracy': 0.9760308754824294}	{'f1': 0.9757209855091478}	{'precision': 0.9773628450645225}	{'recall': 0.9768071193135048}
0.0504	16.9994	20922	0.0428	{'accuracy': 0.9804996953077392}	{'f1': 0.9800980973913306}	{'precision': 0.9820258378580791}	{'recall': 0.9807653776798236}
0.1044	17.9996	22153	0.0428	{'accuracy': 0.9804996953077392}	{'f1': 0.9800980973913306}	{'precision': 0.9820258378580791}	{'recall': 0.9807653776798236}
0.07	18.9998	23384	0.0431	{'accuracy': 0.9804996953077392}	{'f1': 0.9800980973913306}	{'precision': 0.9820258378580791}	{'recall': 0.9807653776798236}
0.0804	20.0	24615	0.0439	{'accuracy': 0.9802965671338615}	{'f1': 0.9798986112130946}	{'precision': 0.9818070673483438}	{'recall': 0.980585456844991}
0.0334	20.9994	25845	0.0443	{'accuracy': 0.9800934389599838}	{'f1': 0.9796902317752838}	{'precision': 0.9816436079495018}	{'recall': 0.9803850961958226}
0.0839	21.9996	27076	0.0433	{'accuracy': 0.9802965671338615}	{'f1': 0.9798986112130946}	{'precision': 0.9818070673483438}	{'recall': 0.980585456844991}
0.0826	22.9998	28307	0.0443	{'accuracy': 0.9802965671338615}	{'f1': 0.9798918326004827}	{'precision': 0.98177108929345}	{'recall': 0.9805511990548504}
0.0897	24.0	29538	0.0434	{'accuracy': 0.9804996953077392}	{'f1': 0.9800980973913306}	{'precision': 0.9820258378580791}	{'recall': 0.9807653776798236}
0.0911	24.9994	30768	0.0437	{'accuracy': 0.9800934389599838}	{'f1': 0.9796991791604638}	{'precision': 0.9815889658615741}	{'recall': 0.9804055360101583}
0.0266	25.9996	31999	0.0426	{'accuracy': 0.9804996953077392}	{'f1': 0.9800980973913306}	{'precision': 0.9820258378580791}	{'recall': 0.9807653776798236}
0.0453	26.9998	33230	0.0453	{'accuracy': 0.9796871826122283}	{'f1': 0.9793004750340752}	{'precision': 0.9811547577187063}	{'recall': 0.980045694340493}
0.0509	28.0	34461	0.0427	{'accuracy': 0.9804996953077392}	{'f1': 0.9800980973913306}	{'precision': 0.9820258378580791}	{'recall': 0.9807653776798236}
0.0682	28.9994	35691	0.0429	{'accuracy': 0.9804996953077392}	{'f1': 0.9800980973913306}	{'precision': 0.9820258378580791}	{'recall': 0.9807653776798236}
0.0601	29.9817	36900	0.0427	{'accuracy': 0.9804996953077392}	{'f1': 0.9800980973913306}	{'precision': 0.9820258378580791}	{'recall': 0.9807653776798236}

Framework versions

• Transformers 4.43.3
• Pytorch 2.3.1
• Datasets 2.20.0
• Tokenizers 0.19.1