bricklerex/lane-detect-jds

segformer-b0-finetuned-segments-sidewalk-oct-22

This model is a fine-tuned version of nvidia/mit-b0 on the bricklerex/lane-detect dataset. It achieves the following results on the evaluation set:

• Loss: 0.0646
• Mean Iou: 0.4681
• Mean Accuracy: 0.9362
• Overall Accuracy: 0.9362
• Accuracy Unlabelled: nan
• Accuracy Lane: 0.9362
• Iou Unlabelled: 0.0
• Iou Lane: 0.9362

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: 6e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Unlabelled	Accuracy Lane	Iou Unlabelled	Iou Lane
0.0823	0.35	20	0.0933	0.4675	0.9349	0.9349	nan	0.9349	0.0	0.9349
0.0676	0.7	40	0.0737	0.4547	0.9093	0.9093	nan	0.9093	0.0	0.9093
0.0639	1.05	60	0.0659	0.4583	0.9166	0.9166	nan	0.9166	0.0	0.9166
0.0584	1.4	80	0.0831	0.4715	0.9429	0.9429	nan	0.9429	0.0	0.9429
0.0541	1.75	100	0.0646	0.4681	0.9362	0.9362	nan	0.9362	0.0	0.9362

Framework versions

• Transformers 4.38.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2